ACNP 63rd Annual Meeting: Poster Abstracts P305-P608 (2025)

Lauren Bystrom, Nicole Kujas, Jose Guizan-Gonzalez, Adrian Parra, Florence Bourgain-Guglielmetti, Luis Tuesta*

Center for Therapeutic Innovation, University of Miami, Miami, Florida, United States

Background: Opioid use disorder (OUD) is a chronic, relapsing disease that induces persistent changes in neuronal signaling and gene expression. OUD is driven by the reinforcing properties of the drug that promote seeking and is perpetuated by the need to avoid the negative physical and emotional states associated with withdrawal and craving. While available OUD pharmacotherapies can be effective, there is an urgent need to identify novel molecular targets that may add to our current therapeutic repertoire. One such potential target is calcitonin gene related peptide (CGRP). CGRP is a neuropeptide synthesized centrally in neurons of the parabrachial nucleus (PBN), a brain region enriched with µ-opioid receptors (µOR). CGRP^PBN neurons express µOR, release glutamate, and project to the nucleus of the central amygdala (CeA), a brain region whose activity is known to influence opioid reward and withdrawal. Considering that CGRP can amplify glutamatergic currents in the CeA, we therefore hypothesized that CGRP^PBN neurons can shape the behavioral responses associated with opioid reinforcement and withdrawal.

Methods: To characterize the behavioral role of CGRP^PBN neurons in opioid reinforcement, male and female CalcaCre^+/- mice received bilateral intra-PBN injections of an AAV containing either an inhibitory hM4Di Cre-inducible DREADD or Cre-inducible mCherry (DIO-mCherry) control. Two weeks later, jugular IV catheters were implanted and mice entered a morphine intravenous self-administration (IVSA) protocol until stable morphine responding was achieved (FR3TO20). Then, to inhibit CGRPPBN neurons, mice received clozapine-N-oxide (CNO) (7.5 mg/kg, IP) or vehicle (saline, IP) using a cross-over design, 30 minutes prior to the IVSA session and number of infusions earned were recorded. To study the role of CGRP^PBN neurons in opioid withdrawal a similar chemogenetic approach was employed. CalcaCre^+/- mice received viral injections into the PBN containing an inhibitory DREADD (hm4Di) and, two weeks later, were implanted with 2x25 mg morphine pellets. After five days, morphine withdrawal was induced by removing the pellets. Mice were then administered either CNO (7.5 mg/kg, IP) or vehicle (saline, IP) 5.5 hours post pellet removal, and behavior was recorded at 6 and 24 hours post pellet removal. Two experimenters scored videos for signs of somatic withdrawal. Following this experiment, C57BL/6J mice were implanted with two 25 mg morphine or placebo pellets and removed after five days to induce spontaneous withdrawal, as above. Zavegepant (7.5 mg/kg, SC), an FDA-approved CGRP receptor inhibitor, or vehicle (saline, SC) was administered 5 hours after pellet removal. Behavior was again recorded at 6 and 24 hours, and videos were scored as described above. These experiments were approved by the Institutional Animal Care and Use Committee and were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals

Results: Inhibition of CGRP^PBN neurons by CNO significantly reduced morphine intake in hm4Di-injected animals (N = 12)(p < 0.001), while control mCherry mice (N = 7) failed to show a change in morphine intake. Additionally, both chemogenetic (wet dog shakes 6 hours: p < 0.005) and pharmacological (Zavegepant) (jumps 24 hours: p < 0.005, freezing 24 hours: p < 0.05, global withdrawal score total and z-score: p < 0.001 and p < 0.0005) inhibition of CGRP receptors significantly reduced signs of somatic withdrawal. In the chemogenetic experiment (CNO-injected N = 12, vehicle-injected N = 12), withdrawal behavior decreased at 6 hours, while in the pharmacological experiment (Zavegepant-injected N = 8, vehicle-injected N = 6), differences were instead seen at 24 hours after pellet removal.

Conclusions: These results suggest that CGRP^PBN neurons play a role in driving opioid reinforcement and withdrawal. Since M4 DREADDs inhibit CGPR and glutamate co-release, follow up genetic studies are underway to dissociate the behavioral effects of CGRP from those of glutamate at the level of the CeA. Importantly, given the popularity of small-molecule CGRP receptor inhibitors (gepants) as first-line treatments for migraine, these new results suggest that CGRP may be a novel, druggable target for the treatment of OUD.

Keywords: opioid use disorder, CGRP, morphine, intravenous drug self-administration, opioid withdrawal

Disclosure: Nothing to disclose.

Aaron Wozniak, Woochan Kim, Nathaniel Burkard, Abolghasem Bakhoda, Kelly O’Conor, Seth Eisenberg, Wenjing Zhao, Sung Won Kim*, Nora Volkow

National Institute on Alcohol Abuse and Alcoholism, National Institutes of Health, Bethesda, Maryland, United States

Background: Fentanyl is a synthetic opioid widely used in clinical settings to enhance the effects of anesthetics and to manage severe surgical pain or breakthrough pain. Widespread illicit manufacture of fentanyl however, has greatly increased in illegal drug markets and was responsible in 2021 for more than 70,000 overdose deaths. When compared to overdoses involving pure heroin, contamination of illicit drugs with fentanyl creates substantial difficulty when reversing opioid overdoses using naloxone due to inducing more rapid and longer lasting respiratory depression. In some instances, repeated naloxone doses are required to avoid sequential overdoses. Pharmacokinetics (PK) of fentanyl is essential to understand its prolonged duration of respiratory depression and its potential accumulation in adipose tissue or in other organs. PK of fentanyl has been characterized in plasma and urine, but there is currently no data on the PK in brain or other organs. Here we report the radiosynthesis of [11C] fentanyl to quantify its biodistribution in the brain and peripheral organs using dynamic positron emission tomography (PET) in rodent studies.

Methods: Radiosynthesis of [11C]fentanyl was performed in a commercially available two-pot system (Synthra MeIplus). Radiolabeling was achieved via the amidation of N-[1-(2-phenylehtyl)-4-piperidinyl]aniline (4-ANPP) using [11C]propionyl chloride that was produced by [11C]carbon dioxide. Cryogenic trapping of [11C]carbon dioxide was applied to optimize molar activity, followed by a subsequent in-loop [11C]carboxylation of ethylmagnesium bromide. [11C]fentanyl was intravenously administered in male Wistar rats (n = 3) to obtain baseline biodistribution data using PET (LFER150 PET/CT, Mediso). To observe the influence of brain efflux pumps, rats (n = 3) were pretreated with elacridar (3 mg/Kg, IV). Knock-out (KO) mice (n = 3) for p-glycoprotein (p-gp) and breast cancer resistance protein (BCRP) were also used to compare brain uptake with controls (n = 4). For image analysis, regions of interest (ROIs) were identified with each individual CT scan and each time-activity curve was extracted from dynamic PET data.

Results: Based on [11C]propionyl chloride collected the second reaction vessel, radiochemical yield was 60.5 ± 29% (n = 10) and averaged molar activity was 34.1 ± 17 Ci/µmol at end of bombardment with high radiochemical purity (> 99%). PET scans in rodents showed a rapid, high brain uptake of [11C]fentanyl (n = 3) (SUVmax, 2.713; Tmax, 1.7 min) accompanied by rapid washout (T1/2, 4.7 min). Elacridar pretreatment (n = 3) significantly increased [11C]fentanyl uptake (SUVmax, 3.59); more than 30% of peak brain uptake was consistently observed in p-gp and BCRP KO mice compared with controls. In peripheral organs, radioactivity accumulated in the liver, kindney, brown adipose tissue, and lung.

Conclusions: [11C]Fentanyl synthesis was fully automated for reliable production with high molar activity and radiochemical yield. Despite serving as a substrate for brain efflux pumps, preliminary PK PET studies demonstrated rapid and high brain penetrability, reflecting the fast-acting nature of fentanyl. This report also documents high deposition of [11C]fentanyl in adipose tissue, which may be related to the prolonged respiratory depression observed after fentanyl use in chronic opioid users.

Keywords: [11C]fentanyl, Positron emission tomography, Pharmacokinetics, Whole-Brain Rodent Imaging

Disclosure: Nothing to disclose.

Patricia Conrod*, Sherry Stewart, Samantha Lynch

Université de Montréal, Montreal, Canada

Background: Substance use disorders (SUDs) remain significantly above national targets for health promotion and disease prevention in Canada and the United States. This study investigates the five-year SUD outcomes following a selective drug and alcohol prevention program targeting personality risk factors for adolescent substance misuse. This study also investigate the intermediate processes that explain how early intervention translates to long-term prevention of substance use disorder in the context of a randomized trial.

Methods: The CoVenture Trial (NCT01655615) is a cluster randomized trial involving 31 secondary high schools from the greater Montreal area agreeing to conduct annual health behaviour surveys for five years on the entire 7th Grade cohort of assenting students enrolled at the school in 2012 or 2013. Half of all schools were randomly assigned to be trained and assisted in the delivery of the personality-targeted PreVenture® program to all eligible 7th grade participants.

The intervention consisted of a brief (2-session) group cognitive behavioural intervention that is delivered in a personality-matched fashion to youth who report elevated scores on one of four personality traits linked to early onset substance misuse: impulsivity, sensation seeking, anxiety sensitivity, or hopelessness.

Primary outcomes were substance use disorder, and intermediate mediators were onset and frequency of substance use (alcohol, cannabis and tobacco use), substance use social norms, mental health symptoms, and cognitive measures of executive functions (response inhibition and working memory), perceptual reasoning, and episodic memory.

Results: Mixed effects multi-level Bayesian models were used to estimate the effect of the intervention on the year-by-year change in probability of SUD. When controlling for baseline differences, a time by intervention interaction revealed positive growth in SUD for the control group (b = 1.380, S.E. 0.143, odds ratio (OR) = 3.97) and reduced growth for the intervention group (b = -0.423, S.E. = 0.173, CI = -0.771, -0.084, OR = 0.655), indicating a 35% reduction in the annual increase in SUD in the intervention condition relative to the control condition. Intermediate effects were revealed for cannabis use and response inhibition.

Conclusions: This study showed for the first time that personality-targeted interventions might protect against longer-term development of SUD and that such preventive effects might be explained by the delay of onset of cannabis use and improving response inhibition.

Keywords: Alcohol and substance use disorders, prevention, response inhibition, executive function, personality

Disclosure: Nothing to disclose.

Eamonn Duffy*, John Sterrett, Luanne Hale, Jonathan Ward, Kyle Brown, Andrew Kwilasz, Erika Mehrhoff, Daniel Frank, Laura Saba, Marissa Ehringer, Ryan Bachtell

The University of Colorado Boulder, Boulder, Colorado, United States

Background: Opioid use disorder (OUD) is an ongoing worldwide public health concern. Genetic factors contribute to multiple OUD-related phenotypes, including oxycodone-induced analgesia and self-administration. The gut microbiome also plays a major role in mediating behavioral responses to opioids. However, little is known about how interactions between host genetics and opioid use influence the gut microbiome. Here, we examine the effects of genetic background and oxycodone self-administration on the composition and structure of the fecal and cecal microbiomes in two genetically divergent inbred rat strains.

Methods: Rats from the ACI/EurMcwi (n = 25F/20M) and M520/N (n = 17F/14M) strains underwent an intravenous self-administration paradigm to measure the acquisition (ten 2-hr sessions) and escalation (ten 12-hr sessions) of oxycodone use. Saline-administering counterparts were used as controls. To assess the effects of oxycodone on the gut microbiome, fecal samples were collected one day following the last 12-hr self-administration session, and cecal contents were collected two days after the last 12-hr self-administration session. 16S rRNA amplicon sequencing was performed, and reads were processed using QIIME2. All behavioral data and processed sequencing data were analyzed using R. Three-way ANOVAs were run to test the effects of strain, sex, and treatment on total oxycodone intake and alpha diversity measures (Chao1, Simpson, and Shannon indices). Community structure (beta-diversity) was assessed using PERMANOVA tests using adonis2 from the R package vegan. ANCOM-BC was used for differential abundance analysis.

Results: Both strains acquired and escalated oxycodone consumption, although the M520/N strain consumed more oxycodone than the ACI/EurMcwi strain (F1,39 = 36.3 p < 0.001). Oxycodone altered the diversity and community structure of both the fecal and cecal microbiomes. Richness of the fecal microbiome, measured with the Chao1 index, was shaped by a strain-sex-treatment interaction (F1,57 = 4.8, p = 0.03), however no differences were observed in the Shannon nor Simpson indices. In the cecal microbiome, alpha diversity was shaped by a sex-treatment interaction (Chao1: F1,51 = 5.8, p = 0.02; Shannon: F1,51 = 7.9, p = 0.007; Simpson: F1,51 = 6.6, p = 0.01), with oxycodone-administering males showing decreased diversity compared to saline-administering males and oxycodone-administering females. ACI/EurMcwi rats also displayed greater alpha diversity than M520/N rats (Shannon: F1,51 = 7.8, p = 0.007; Simpson: F1,51 = 7.5, p = 0.008). Beta diversity analysis demonstrated that community structure of the fecal microbiome is shaped by a sex-treatment interaction (F1,57 = 4.0, p = 0.004) while the cecal microbiome community structure is influenced by oxycodone treatment (F1,51 = 2.0, p = 0.04). Differential abundance analysis was performed to identify specific bacterial taxa that respond to oxycodone, as compared to saline-administering controls. In the fecal microbiome, 8 taxa increased in prevalence and two decreased in response to oxycodone. In the cecal microbiome, 7 taxa increased in prevalence and 2 taxa decreased. Interestingly, Prevotalla increased in abundance in both microbiomes in response to oxycodone self-administration.

Conclusions: We demonstrated that genetic background influences oxycodone self-administration in two genetically diverse, inbred rat strains. Both host genetics and oxycodone self-administration altered the diversity and composition of the gut microbiome. ACI/EurMcwi rats displayed higher cecal, but not fecal, alpha diversity compared to M520/N rats. Although both sexes administered similar amounts of oxycodone, oxycodone-administering males displayed decreased cecal alpha diversity. Finally, oxycodone altered the composition of the gut microbiome, as indicated by beta diversity and differential abundance analyses. Together, these results demonstrate that interactions between genetic background, sex, and oxycodone exposure are important determinants of the fecal and cecal microbiomes.

Keywords: Gut Microbiome, Genetic variation, opioid use disorder

Disclosure: Nothing to disclose.

Mathew Schafer, Jasper Van Oort, Philip Kamilar-Britt, Vyoma Sahani, Yasmin L. Hurd, Daniela Schiller, Keren Bachi*

Icahn School of Medicine at Mount Sinai, New York, New York, United States

Background: Social behavior can influence the risk, trajectory and severity of substance use disorders, yet there remain gaps in knowledge of the related neurobehavioral underpinnings. Successfully navigating social interactions may depend on social mapping—the compact representations of how people relate to each other along abstract dimensions. It has been shown that the hippocampus constructs social maps; the prefrontal cortex may use these maps in decision-making. Cocaine use disorder (CUD) and childhood trauma, commonly co-occurring, feature concurrent social dysfunction and neural abnormalities in these regions, with prefrontal cortex alterations common in CUD and hippocampal alterations common in childhood trauma. We use a naturalistic social interaction task to test whether social relationships are tracked differently in the prefrontal cortex of individuals with CUD relative to healthy controls (HC), and differently in the hippocampus as a function of childhood trauma.

Methods: Participants (CUD n = 32 and HC n = 38) completed the social navigation task during functional magnetic resonance imaging (fMRI). In this task, participants interact with fictional characters to accomplish social goals (e.g., find a job). Participants’ interaction decisions with each character were modeled as a series of locations through a two-dimensional abstract space of affiliation by power to represent the “social navigation” of relationships. To probe the neural correlates of this navigation-like behavior, we used representational similarity analysis. For each participant, we estimated the voxel-wise correlations between character-level social trajectories and patterns of brain activity, controlling for the temporal distances between characters. This analysis captured the extent to which similar social trajectories across characters were represented by similar neural patterns, irrespective of where the characters appeared in the narrative. We used regions-of-interest analyses to test two predictions: 1) the social location effect for CUD participants would be reduced in prefrontal cortex, specifically in right inferior frontal gyrus, a region involved in processing navigational demands 2) individuals with higher childhood trauma would have a reduced social location representation in left hippocampus, given the region’s sensitivity to stress and its established role in this task. To test these predictions, we modeled the effect of diagnostic group (CUD vs. HC) and childhood trauma (Childhood Trauma Questionnaire, CTQ) on the trajectory analysis estimates.

Results: The behavioral estimates of social location tracked post-task subjective placements of the characters, suggesting the behavioral locations tracked perceptions of social locations (permutation test: p < 0.001); there was no effect of diagnosis or CTQ score (both ps > 0.5). As predicted, in the fMRI analysis there was a smaller pattern similarity effect in CUD than HC in the right inferior frontal gyrus (t = -2.16, right-tailed p < 0.05), suggesting altered engagement of this region during social navigation in CUD. Also as predicted, childhood trauma negatively correlated with the pattern similarity effect in the left hippocampus (t = -2.88, right-tailed p < 0.01), a region previously shown to track abstract social locations. This effect was independent of CUD status.

Conclusions: Our findings suggest that both CUD-related prefrontal dysfunction and childhood trauma-related hippocampal dysfunction extend to the neural tracking of social relationships. Childhood trauma had similar reductions in left hippocampal effects in both CUD and HC participants, suggesting that the effect of trauma on representing social relationships is separable from the effects of cocaine addiction. CUD and childhood trauma may thus impact the representation of social navigation in different but potentially complementary ways. More work is being done to fully characterize these effects – such as the interaction between these regions in both CUD and childhood trauma, as well as implications for real-world social behavior and clinical outcomes. Advancing knowledge of social neurobehavioral phenotypes may inform the development of novel social-target interventions in substance use disorders.

Keywords: social cognition, cocaine use disorder, Functional MRI (fMRI), Childhood trauma, Cognitive Neuroscience

Disclosure: Nothing to disclose.

Eric Woodcock*, Mark Greenwald, Leslie Lundahl, Paul Burghardt, Michael Petriello, Huailei Jiang, Otto Muzik

Wayne State University School of Medicine, Detroit, Michigan, United States

Background: Chronic substance use is known to modulate immune function, especially cigarette smoking. Recently, molecular imaging studies have investigated neuroimmune dysregulation across substance use disorders (SUDs) with one notable exception, opioid use disorder (OUD): no studies in OUD have been published to date. Preclinical opioid challenge studies, and postmortem analyses of deceased opioid users, show elevated levels of glial markers in the brain, suggestive of neuroinflammation. By contrast, peripheral immune markers are generally lower among living OUD patients relative to comparator subjects, though plasma markers may not reflect neuroimmune state. Additionally, case-control studies often do not match groups for cigarette smoking which is common among OUD patients and known to modulate immune function, thereby complicating interpretation. To address this research gap, we applied positron emission tomography (PET) imaging with α-[11C]methyl-L-tryptophan ([11C]AMT) to quantify brain kynurenine metabolic rate (K-complex), among OUD patients and well-matched comparators. In the human brain, ~95% of tryptophan is converted to kynurenine by the rate-limiting enzyme, indoleamine 2, 3-dioxygenase 1 (IDO1), which is highly sensitive to inflammatory stimuli. Pro-inflammatory stimuli, including endotoxin, interferon-γ (IFNγ), and interleukin (IL)-6, upregulate IDO1 expression thereby accelerating the metabolism of tryptophan to kynurenine, whereas, anti-inflammatory stimuli, e.g., IL-4 and IL-10, exert the opposite effect. Thus, quantification of regional brain kynurenine metabolic rate (K-complex) using PET [11C]AMT imaging yields an in vivo marker sensitive to neuroimmune state. Here, we hypothesized that OUD patients would exhibit higher PET [11C]AMT K-complex values across brain regions of interest (ROIs) relative to well-matched comparators, indicative of a neuroinflammatory state.

Methods: In a single day, OUD patients in outpatient methadone maintenance therapy (MMT; n = 20; DSM-5 criteria for OUD) and non-opioid-using comparator subjects (n = 23; matched for age, sex, body mass index [BMI], and cigarette smoking) underwent a T¬1-weighted magnetic resonance imaging (MRI) scan at 10:30am (for anatomical coregistration to PET data) and a 60min PET [11C]AMT scan from 13:30-14:30 (0.2 mCi/kg; max 20mCi) while fasted (6hr + ; to stabilize plasma tryptophan levels). Across seven a priori ROIs, unidirectional uptake rate constant, K-complex, was quantified via Patlak plot using a left ventricular and venous input function. Group differences in regional [11C]AMT K-complex values were evaluated using a repeated measures analysis of variance (rmANOVA) with Diagnostic Group as the fixed factor (OUD vs. comparators) and regional [11C]AMT K-complex values as the within-subject repeated factor. This study was powered to detect moderate effects (Cohen’s f = 0.34) with β > 0.80 and α ≤ 0.05.

Results: Overall, subjects were 42.9 ± 12.6yrs old, 67.4% male, and 45.2% African-American. Diagnostic groups were well-matched for age, sex, BMI, past-month cigarette smoking frequency, Fagerstrom Test for Nicotine Dependence (FTND) score, and expired breath carbon monoxide on scan day (ps > 0.20). Mean methadone dose among OUD patients was 96.1 ± 51.4mg/day PO. PET [11C]AMT findings indicate that OUD patients exhibited significantly higher regional [11C]AMT K-complex values than non-opioid comparators by 18% (11-21%; p = 0.029; partial η2 = 0.13; ‘large’ effect) with the largest effects in the nucleus accumbens and thalamus. Among OUD patients, [11C]AMT K-complex values were positively correlated with daily methadone dose (r = 0.64, p = 0.006). OUD patients exhibited higher levels of plasma kynurenine, kynurenine/tryptophan ratio, IFNγ, IL-6, and C-reactive protein (CRP) than comparators on scan day (all ps <0.05). Downstream of kynurenine, plasma kynurenic acid levels did not differ between groups (p = 0.22) whereas quinolinic acid results are pending.

Conclusions: Our findings show that OUD patients enrolled in outpatient MMT exhibit higher regional kynurenine metabolic rate (K-complex) than matched non-opioid-using comparators, with reward- and pain-related regions being particularly affected (nucleus accumbens and thalamus, respectively), thereby providing the first in vivo evidence of a neuroinflammatory phenotype in methadone-maintained OUD patients. Consistent with our PET data, OUD patients exhibited higher plasma levels of kynurenine, kynurenine/tryptophan ratio, IFNγ, IL-6, and CRP than non-opioid comparators suggesting OUD patients exhibit elevated IDO1 expression in the brain and body likely upregulated by pro-inflammatory signals. Plasma kynurenic acid levels did not differ between groups suggesting that the neuroprotective branch of the kynurenine pathway is not dysregulated in OUD whereas quinolinic acid results are pending (i.e., the neurotoxic branch). Importantly, neuroinflammation is a ‘druggable’ target for medications development in OUD and prior clinical research has shown that ibudilast, an anti-inflammatory glial modulator, reduced subjective opioid craving, some opioid withdrawal symptoms, and opioid-seeking behavior in non-treatment-seeking individuals with OUD. Our findings contribute to a growing literature showing neuroimmune dysregulation in SUDs and highlight the importance of studying kynurenine pathway dysfunction in OUD.

Keywords: opioid use disorder, neuroinflammation, Kynurenine pathway, Methadone maintenance treatment, Positron Emission Tomography (PET) Imaging

Disclosure: Nothing to disclose.

Juan Gomez*, Emilya Ventriglia, Zachary Frangos, Reece Budinich, Agnieszka Sulima, Michael Robertson, Michael Sacco, Ilinca Giosan, Anna Tischer, Grant Glatfelter, Jennifer Bossert, Hank Jedema, Charles Bradberry, Michael Baumann, Eugene Kiyatkin, Yavin Shaham, Venetia Zachariou, Georgios Skiniotis, Kenner Rice, Mike Michaelides

National Institute on Drug Abuse, Baltimore, Maryland, United States

Background: Mu opioid receptor (MOR) agonist medications are the most effective analgesics available. However, such medications have known adverse effects including respiratory depression and abuse liability. As such, there is a concerted effort to develop efficacious MOR agonists with lower adverse effects than available opioid medications. The nitazene class of MOR agonists, distinct from poppy-derived and other synthetic opioids, shows highly selective MOR binding. These compounds possess superagonist characteristics and to date have proven too potent to be utilized safely in clinical settings. Recently, nitazenes have been found in toxicology reports from overdose fatalities and it has been suggested that they contribute to the deleterious effects found in the poly-drug mixtures of street drugs. Here we describe a niazene analog and its metabolite, Fluornitrazene (FNZ) and N-desethyl-Fluornitrazene (DFNZ), respectively. Like etonitazene, both are superagonists and have superior in vivo efficacy. Furthermore, the data suggest that DFNZ may be peripherally restricted which may confer a favorable characteristic towards potential adverse effects.

Methods: To assess selectivity and propensity for brain entry, both FNZ and DFNZ were screened against a panel of > 100 receptors/enzymes and its drug transporter inhibition activity at 100 nM and 10 µM. Both were studied via competitive binding assays using rat brain membranes incubated in buffer solutions of 10 nM [3H]DAMGO (46 Ci/mmol) with increasing concentrations of unlabeled DAMGO, FNZ, and DFNZ. One-site competition curves were fitted, and Ki values were calculated using the Cheng-Prusoff equation. Both were also evaluated in their propensity to stimulate MOR Gi and β arrestin-signaling using HEK293 cells transfected with hMOR cDNA. To determine mu receptor occupancy, ex vivo experiments were designed to inject doses of Saline, FNZ, DFNZ, and Naloxone and tissue was collected 15-min post-injection for autoradiography visualization with [3H]DAMGO. The behavioral and physiological effects of both compounds were assessed using intravenous self-administration (IVSA), antinociception, mechanical sensitivity, catalepsy, body temperature, locomotion, and respiratory depression assays.

Results: The screening data found that at 100 nM, FNZ inhibited binding of only MOR. In the efflux transporter panel, at 100 nM, FNZ inhibited the activity of only OCT2 and MATE2-K, which are mainly peripheral efflux transporters. Competitive binding assays against [3H]DAMGO in whole rat brain tissue membranes showed that FNZ had a Ki=0.4 nM and DFNZ had a Ki =0.04 nM. Functional assays for MOR Gi found an FNZ EC50 = 0.8nM and DFNZ EC50 = 4.9nM. For β arrestin, FNZ EC50 = 0.6nM and DFNZ EC50 = 8.9nM. Occupancy data showed that doses of FNZ (10 and 100 µg/kg) significantly block [3H]DAMGO binding while DFNZ at low dosing (100 µg/kg) has no occupancy and requires supranormal dosing (1 mg/kg) to block [3H]DAMGO binding. The same effect for FNZ and DFNZ were found in all behavioral tests conducted. The main difference was that DFNZ required higher doses to elicit a similar response. Both compounds showed acquisition of self-administration but the dose response curve was right-shifted for DFNZ. Most importantly, during extinction, FNZ did not escalate their active lever pressing and rats did not extinguish responding, while DFNZ extinguished immediately on the first extinction session. Both compounds had analgesic effects on the hotplate, CFA model of pain, and Von Frey, but higher doses of DFNZ were required. Compared to fentanyl, FNZ resulted in a mild-transient depression in central oxygen levels that returned to baseline within minutes. However, DFNZ showed no significant change in central oxygen levels, again suggesting that the effects of DFNZ are peripherally mediated.

Conclusions: We show how a novel drug can be characterized by using several in vivo/ in vitro/ ex vivo radiometric and behavioral assays to gain an understanding of its mechanism as a potential analgesic. The etonitazene analog FNZ and its metabolite, DFNZ, had favorable qualities, such as, MOR selectivity, high therapeutic indices, and low potential for tolerance. To date we have not found evidence that the effects of DFNZ are centrally mediated, and we predict that the metabolite is restricted to the periphery. The importance of this finding is that we have discovered an efficacious opioid that is peripherally restricted and may provide long-lasting analgesia with low abuse liability and respiratory depression.

Keywords: Mu opioid receptor, Nitazene, PET Imaging, Fluornitrazene

Disclosure: Nothing to disclose.

Annaka Westphal*

Medical University of South Carolina, Charleston, South Carolina, United States

Background: Emerging data demonstrates that distinct neuronal ensembles within the dorsomedial prefrontal cortex (dmPFC) form during heroin self-administration (SA) and this coordinated dmPFC activity is required for heroin seeking following extinction training (Ext). Still, the precise neurobiological mechanisms underlying the formation and evolution of the dmPFC neuronal ensembles remains unknown. Astrocytes are understood to be key modulators of neuronal activity through mobilization of intracellular Ca2+ stores, and therefore represent a prime candidate for modulation of dmPFC neuronal ensemble dynamics. However, how Ca2+ activity within individual dmPFC astrocytes evolves across heroin use, and whether dmPFC astrocytic Ca2+ is required for heroin seeking has never been studied. To answer these questions, we first use 2-photon Ca2+ imaging to measure and track dmPFC astrocytic Ca2+ dynamics across heroin SA, extinction, and reinstatement tests. We then applied novel genetic tools to functionally manipulate dmPFC astrocytic Ca2+ throughout the same paradigm.

Methods: To measure astrocytic Ca2+ activity, we used a head-fixed heroin SA protocol with simultaneous 2-photon imaging of dmPFC astrocytic GCaMP6f fluorescence (AAV5-GfaABC1D-Lck-GCaMP6f). Astocytic Ca2+ dynamics were recorded across 7 mice (4 male and 3 female) during early (day 1), mid (day7) and late (day14) heroin SA, early (day 1) and late (day 10) Ext, and during cue-, drug-, and stress-induced reinstatement (Rein) of heroin seeking. Data were collected using a resonant scanning 2-photon microscope and processed with a custom Python pipeline. Changes in fluorescence of individual astrocytes relative to baseline were measured for lever press (active, inactive), cue presentation (or omission), and heroin infusion (or omission). To inhibit dmPFC astrocytic Ca2+ we delivered a virus encoding for either the plasma membrane Ca2+ ATP-ase (PMCA; AAV2/5 GfaABC1D-mCherry-hPMCA2w/b; 11 mice; 5 male and 6 female) or a control virus (AAV5 GfaABC1D-mCherry; 7 mice; 5 male and 2 female) to the dmPFC. Using the same head-fixed heroin SA paradigm we then tested whether PMCA-mediated inhibition of astrocytic Ca2+ activity impairs heroin seeking during cue-, drug-, or stress-induced reinstatement tests.

Results: Using an auROC-based discrimination score, astrocytic Ca2+ signals exhibited both sex- and trial type-dependent adaptations in their ability to discriminate baseline vs. the epoch containing active lever press (ALP) cue exposure and infusion. A two-way mixed ANOVA reveals effects of sex F(7, 2120) = 106.4, p < 0.0001) and SA phase F(7, 2120) = 175.2, p < 0.0001), with a significant interaction (F(7, 2120) = 35.52, p < 0.0001). Sidak post-hocs reveal higher astrocytic discrimination for this epoch in females relative to males at Early SA, Mid SA, Early Ext, Drug Rein, and TMT Rein (p < 0.0001), with higher discrimination in males at Late Ext (p < 0001) and Cue Rein (p < 0.001). Mixed sex binary decoding analyses demonstrate SA phase-dependent adaptations in the ability of astrocyte Ca2+ activity to predict ALP. One-way mixed ANOVA, main effect of trial type F(5,1638) = 47.8, p < 0.0001, with astrocyte Ca2+ activity best predicted ALP during CueRein relative to Mid SA, Late SA, Early Ext and Drug- and TMT-induced seeking (p < 0.0001). Following reduction of dimensionality, astrocytes were sorted into ensembles via spectral clustering. Astrocyte ensembles had the following properties: Ensemble 1 - Peak During Cue, Ensemble 2 - Peak at ALP, Ensemble 3 - Sustained weak response, and Ensemble 4 - Peak at infusion. During CueRein we did not observe weak response cells and did detect significant differences in the ability of astrocyte ensembles to predict ALP. A one-way mixed ANOVA showed a main effect of ensemble type F(2, 201) = 8.4, p < 0.0001. Tukey post-hocs reveal that Ensemble 1, which peaks during heroin conditioned cues, exhibits significantly better ability to predict ALP than Ensemble 2 (p < 0.01) and Ensemble 4 (p < 0.001). Additionally, while we found that PMCA-mediated inhibition of astrocytic Ca2+ activity did not affect lever pressing during heroin SA or Ext (mixed-model ANOVA), we found that PMCA did suppress active lever pressing in both cue- and drug-induced Rein relative to mCherry controls (unpaired t-tests, p < 0.05).

Conclusions: Our findings indicate that dmPFC astrocytes exhibit time locked Ca2+ responses to the epoch containing ALP, cues and heroin delivery during drug taking and seeking. We observe sex- and SA phase-dependent adaptations in astrocyte Ca2+ responses this epoch and find that astrocyte Ca2+ signals best predict ALP during cued heroin seeking. Additionally, inhibition of dmPFC astrocytic Ca2+ activity suppresses during cue- and drug-induced heroin seeking, an effect that has not previously been reported. Finally, akin to neurons, dmPFC astrocytes form discrete ensembles with distinct response profiles and varied abilities to decode or predict conditioned heroin seeking. Our data suggest that astrocytic Ca2+ activity is important in dmPFC ensemble formation underlying heroin seeking. Future directions are to functionally manipulate astrocytes while recording from dmPFC neurons to observe how astrocytes shape neuronal ensemble dynamics.

Keywords: astrocytes, heroin self-administration, heroin seeking

Disclosure: Nothing to disclose.

Sarah Lichenstein*, Brian Kiluk, Marc Potenza, Hugh Garavan, IMAGEN Consortium, Godfrey Pearlson, Sarah Yip

Yale University School of Medicine, New Haven, Connecticut, United States

Background: Cannabis is the most commonly used illicit drug, and substantial evidence indicates that its use is associated with clinically-significant harms for a subset of users. In the context of recent decriminalization/legalization of cannabis—in tandem with significant increases in potency—further work is urgently needed to (i) identify who is most vulnerable to cannabis-related harms and (ii) elucidate neurobiological mechanisms of risk. Existing studies examining neural correlates of cannabis use in adolescence yield inconsistent results, and provide insufficient information to distinguish neural risk factors of problem-level vs. recreational use. Our prior work applied connectome-based predictive modeling (CPM), a data-driven, whole-brain, machine learning approach, to identify a neuromarker of problem-level cannabis use in a non-clinical sample of adolescents/emerging adults and validated the clinical relevance of the identified network in an independent clinical sample of individuals entering treatment for cannabis use. The aims of the current study were to examine 1) whether the network identified would generalize to predict cannabis use in a third independent and demographically-distinct sample of European adolescents/emerging adults, and 2) whether the network is specific for predicting cannabis versus alcohol use outcomes across datasets.

Methods: The problem cannabis risk network was previously identified using reward task (Monetary Incentive Delay Task; MID) fMRI data from N = 191 participants from the NIAAA-funded Brain and Alcohol Research in College Students (BARCS) Study, a 2-year longitudinal study of alcohol and substance use behavior. Clinical relevance of the identified network was demonstrated in an independent clinical sample of N = 33 individuals entering cannabis use treatment who were enrolled in randomized clinical trials of cognitive behavioral treatments for substance use. The current analyses also include participants drawn from the IMAGEN study, a large-scale study of adolescent neurodevelopment including ~2000 youth from four European countries, for whom MID fMRI and cannabis use data at age 19 were available (N = 1320, 53% female). Generalizability of the problem cannabis risk network was assessed by examining the association between network strength and cannabis use frequency, quantified using the European School Survey Project on Alcohol and Drugs (ESPAD), among IMAGEN participants. Specificity to cannabis versus alcohol was examined by assessing associations between network strength and alcohol use outcomes (alcohol use frequency in BARCS and IMAGEN; alcohol problem severity on the Addiction Severity Index in the clinical replication sample) across all three datasets.

Results: Although rates of cannabis use were low in the IMAGEN sample relative to the BARCS sample, problem cannabis risk network strength was nonetheless significantly associated lifetime cannabis use (rho = 0.08, p = 0.003). Similarly, problem cannabis risk network strength was significantly higher among the subset of participants reporting ≥40 lifetime cannabis use occasions (i.e., the highest response option on the ESPAD; n = 186) relative to those with no lifetime cannabis use history (n = 652; t = 2.802, p = 0.005). In the BARCS study, there was no association between problem cannabis risk network strength and past-month alcohol use among a sub-sample of participants with no past-month cannabis use at baseline (n = 138, rho = -0.047, p = 0.582). Similarly, among IMAGEN participants, problem cannabis risk network strength was not significantly associated with lifetime alcohol use (n = 1321, rho = 0.05, p = 0.068), nor was there a significant difference between subgroups reporting no lifetime use versus those with 40 or more lifetime uses of alcohol (t = -0.249, p = 0.803). Finally, within the clinical replication sample, network strength was not associated with alcohol problems at treatment entry (n = 33, rho = 0.141, p = 0.434). Given the observed specificity of the problem cannabis risk network for predicting cannabis versus alcohol use outcomes across datasets, a comparison of network anatomy with an alcohol use risk network that was recently identified in the IMAGEN dataset and found to generalize to the BARCS sample is also presented.

Conclusions: These data demonstrate a second external replication of the problem cannabis risk network, as well as illustrating specificity to cannabis versus alcohol use outcomes across three independent samples. These findings provide insight into neural mechanisms of risk for problem-level cannabis use among adolescents/emerging adults and pave the way for future efforts to develop more targeted and effective prevention and treatment approaches for problem cannabis use.

Keywords: machine learning, Replication, Neural mechanism

Disclosure: Nothing to disclose.

Ryan Bachtell*, Eamonn Duffy, Chris Andrews, Mercedes Lopez, Luanne Hale, Jonathan Ward, Kyle Brown, Laura Saba, Marissa Ehringer

The University of Colorado Boulder, Boulder, Colorado, United States

Background: Prescription opioids are considered the gold standard in analgesic pain medication, yet consumption of opioid drugs, such as oxycodone, can lead to problematic drug use and the development of opioid use disorder (OUD). Relatively little work has systematically assessed the factors that contribute to the development of OUD. Twin and family studies indicate that up to 50% of OUD vulnerability is due to genetic factors. We recently employed a rodent genetic model to more precisely measure specific phenotypes associated with the initiation of oxycodone intake and the development of compulsive-like escalation of oxycodone use. Here, we present several novel oxycodone intake phenotypes from a self-administration (SA) model in Hybrid Rat Diversity Panel (HRDP) inbred strains that display high and low oxycodone intake.

Methods: Adult male (n = 107) and female (n = 104) rats from 15 inbred rat strains (n = 10-30/strain) from the HRDP were assessed for differences in oxycodone SA. Animals were surgically implanted with indwelling jugular catheters before beginning SA. To assess the acquisition of oxycodone SA, rats were initially trained to lever press for oxycodone on a fixed-ratio 1 (FR1) reinforcement schedule in 10 daily 2-hour (short-access; ShA) sessions. Rats then progressed to the escalation phase with 10 daily 12-hour (long-access; LgA) sessions under the same FR1:TO20 schedule of reinforcement. Based on initial profiling of oxycodone intake parameters across all 15 inbred strains, 3 strains were selected as “low-taking”, and 3 strains were selected as “high-taking” for additional phenotypic assessment. These six strains were evaluated for several intake parameters associated with compulsive and dysregulated oxycodone intake including episodes of burst infusions (3+ infusion/90 sec), burst duration (min), infusions per burst, and inter-infusion intervals (min).

Results: We find that genetic background and sex contributed to total oxycodone intake during the acquisition and escalation phases (Acquisition: F14,236 = 3.24, p < 0.001, Escalation: F13,176 = 4.27, p < 0.001). Based on total oxycodone intake, the BXH6/Cub, HXB23/lpcv, and HXB31/lpcvMcwi strains were deemed “low-taking” strains, and the HXB1/lpcv, HXB2/lpcvMcwi, and M520/N as “high-taking” strains. Self-administration intake parameters during the first and last LgA sessions were analyzed for patterns of dysregulated intake. Our results suggest that the low- and high-taking strains are significantly different in the number of burst episodes (F5,89 = 16.93, p < 0.001), burst duration (F5,88 = 8.99, p < 0.001), infusions per burst (F5,88 = 7.49, p < 0.001), and inter-infusion intervals (F5,88 = 12.19, p < 0.001). Among the high-taking strains, the M520/N displayed a significant increase in burst episodes from the first to the last LgA session (t89 = 6.68, p < 0.001), and had significantly more episodes of burst infusions compared with all other strains. Interestingly, the burst episodes for the high-taking HXB2/lpcvMcwi strain displayed the highest number of infusions per burst and longer burst durations compared with all other strains. The low-taking HXB23/lpcv strain, however, maintained controlled oxycodone intake across the escalation phase with no escalation in oxycodone intake, few burst episodes, and long inter-infusion intervals.

Conclusions: These findings suggest that genetic factors contribute to the acquisition and escalation of oxycodone intake in rats. Moreover, the manner in which rats volitionally consume oxycodone varies depending on genetic background. We find that even rat strains that are prone to higher levels of oxycodone intake display varying patterns of intake with some displaying more controlled intake while others engage in uncontrolled, dysregulated intake patterns. These findings suggest that the underlying genetic contributions to drug use may uniquely determine multiple features of opioid use disorders.

Keywords: behavioral genetics, Opioid abuse, Addiction phenotypes

Disclosure: Nothing to disclose.

Xuan Li*, Chloe Matheson, Xiang Luo, Adedayo Olaniran, Hongyu Lin

University of Maryland - College Park, Columbia, Maryland, United States

Background: Relapse is a major challenge in treating opioid addiction, including oxycodone, a commonly abused prescription opioid. In rats, cue-induced oxycodone seeking progressively increases during abstinence. Our previous work demonstrated that orbitofrontal cortex (OFC) plays a critical role in this incubation of oxycodone craving. However, the molecular mechanisms in OFC that contribute to this incubation are unknown. Here, we focus on mitochondrial dynamics in OFC and characterize the mitochondrial morphology in OFC neurons during incubation of oxycodone craving.

Methods: We used a dual-virus approach to sparsely label mitochondria in OFC neurons by injecting the adeno-associated virus (AAV)-hSyn-GFP together with AAV-CMV-mitoDsRed bilaterally into OFC. Next, we trained male rats to either self-administer saline (as the control group) or oxycodone (0.1 mg/kg/infusion) for 6 h/day over 10 days. On abstinence day 1 or 15, we perfused all rats and processed the brain for confocal microscopy.

Results: Our image analysis showed that in the somas of OFC neurons, there was a significant increase in the size-frequency of the smallest mitochondria, accompanied by overall increased mitochondria density, in oxycodone rats compared with saline rats on abstinence day 15, but not on abstinence day 1. In contrast, we found no differences in mitochondrial size-frequency but decreased mitochondrial densities in proximal dendrites of OFC neurons between the two groups on abstinence day 15.

Conclusions: Together, these findings demonstrated that incubation of oxycodone craving was associated with a time-dependent increase of mitochondrial fission in OFC soma in male rats. In addition, mitochondria were differentially distributed across soma and proximal dendrites of OFC neurons between saline and oxycodone rats on abstinence day 15. Studies are underway to examine the mitochondrial morphology in female rats during incubation of oxycodone craving and the causal role of mitochondrial dynamics in oxycodone relapse.

Keywords: incubation of oxycodone craving, orbitofrontal cortex (OFC), Mitochondrial dynamics

Disclosure: Nothing to disclose.

Shikun Hou, Elizabeth Glover*

University of Illinois at Chicago, Chicago, Illinois, United States

Background: Reward prediction error (RPE) is a computational neural process that compares outcomes with expectations to guide decision making and adaptive responding. Disrupted RPE signaling underlies impaired decision making, a consequence consistently observed in alcohol use disorder (AUD). The medial prefrontal cortex (mPFC) plays a critical role in decision making and emerging data suggests that distinct mPFC cell populations encode RPE. However, the effect of chronic ethanol exposure on RPE processing in the mPFC is unknown.

Methods: Adult male and female Long-Evans rats were classically conditioned to associate an audiovisual cue with delivery of either a sucrose pellet or foot shock. In vivo fiber photometry was used to measure intracellular calcium activity, a proxy for neural activity, in the prelimbic (PL) subregion of the mPFC during learning and in response to expected and unexpected outcomes. Rats were subsequently rendered dependent using a standard 14-d chronic intermittent ethanol (CIE) vapor exposure paradigm. Controls were exposed to room air (AIR). One week after their last vapor exposure, rats resumed behavioral testing to determine the effect of CIE exposure on behavioral responding and PL calcium signal.

Results: Before CIE, PL mPFC calcium signal increased significantly from baseline in response to both appetitive and aversive cues (*p < 0.0001) and outcomes (*p < 0.0001). Calcium signal did not reflect trial-by-trial changes in expected outcome for appetitive or aversive stimuli across learning or RPE probe sessions. However, it was significantly negatively modulated by reward seeking in the appetitive task (*p < 0.0001) and passive threat responding in the aversive task (*p < 0.0001). After vapor exposure, in AIR controls the increase in calcium signal in response to appetitive cues was attenuated whereas the response to aversive cues was unchanged. In contrast, calcium signal was maintained at pre-vapor levels in CIE-exposed rats in response to appetitive cues but attenuated in response to aversive cues (*p < 0.001).

Conclusions: These results indicate that bulk PL mPFC calcium activity does not encode RPE, but instead reflects a combined signal integrating the salience of environmental stimuli and behavioral responding to those stimuli. Our data further suggest that impaired mPFC engagement during salience processing occurs in a valence-specific manner following chronic ethanol exposure with heightened responding to appetitive stimuli and attenuated responding to aversive stimuli observed after CIE exposure.

Keywords: reward and aversion, Pavlovian conditioning, reward prediction error, Medial Prefrontal Cortex (mPFC), Salience

Disclosure: Nothing to disclose.

Nina Beltran, Vanessa Minervini, Katherine Serafine*

The University of Texas at El Paso, El Paso, Texas, United States

Background: Eating a high fat diet leads to negative health consequences such as obesity and type 2 diabetes. Recent evidence also suggests that eating a high fat diet can impact drug sensitivity. For example, eating a high fat/ high carbohydrate laboratory chow enhances the sensitivity of rats to the behavioral effects of cocaine and methamphetamine. While high fat/high carbohydrate diets are associated with weight gain, ketogenic diets that are high in fat, but very low in carbohydrates have been investigated for weight loss. We hypothesized that eating a high fat/high carbohydrate diet would increase sensitivity of rats to some of the effects of morphine (e.g., tolerance and constipation).

Methods: Female and male Sprague Dawley rats were given access (n = 8/dietary group) to either a high fat/high carbohydrate chow (60% kcal from fat), a ketogenic chow (90.5% kcal from fat) or standard chow (17% kcal from fat) for several weeks prior to once weekly testing with cumulative doses of morphine using an antinociception assay (experiment 1; n = 24/sex) or weekly testing with single doses of morphine in an assessment of constipation (experiment 2; n = 24/sex). In experiment 1, morphine-induced antinociception was assessed using the warm water tail withdrawal procedure (0.32 – 17.8 mg/kg IP). The procedure began with a saline injection, followed by 4 cumulative injections of morphine (0.32-17.8 mg/kg IP) administered every 15 minutes and 13 minutes after each injection, the latency for rats to remove their tail from hot water baths (40, 50, or 55°C) was recorded. Additionally, to study the effects of chronic morphine exposure, morphine was administered twice daily for 19 days (3.2-56 mg/kg IP) increasing in ¼ log doses every 3 days, followed by an additional warm water tail withdrawal assessment with morphine (3.2-56 mg/kg IP). In experiment 2, morphine-induced constipation was assessed by counting and weighing fecal output from female and male rats hourly, following an injection of saline or morphine (1, 3.2, or 10 mg/kg IP) after 2-hour access to wet homogenized chows. Warm water tail withdrawal latencies were converted to a % of the maximum possible effect and were averaged by each dietary group. Cumulative fecal count and weight were averaged by each dietary group. Data were analyzed using mixed model ANOVAs.

Results: In experiment 1, morphine-induced antinociception was comparable among rats eating different diets when tested under acute conditions and all rats developed tolerance to morphine-induced antinociception. In general, male rats were more sensitive to morphine-induced tolerance than females, regardless of diet. Further, while there were no group differences in morphine tolerance among female rats eating different diets, male rats eating high fat/high carbohydrate chow developed greater tolerance than rats eating ketogenic chow. In experiment 2, morphine induced constipation in all rats regardless of diet, and was comparable between sexes. However, for both sexes, rats eating either high fat/high carbohydrate or ketogenic chow experienced more severe constipation as compared to rats eating standard chow.

Conclusions: These results suggest that dietary intake might enhance sensitivity of individuals to some of the adverse effects of morphine, including tolerance and constipation. These results have implications for individuals who take opioids chronically for pain management, as well as for individuals who use opioids recreationally.

Keywords: Ketogenic diet, high fat diet, morphine, rats, opioid tolerance

Disclosure: Nothing to disclose.

Mackenzie Gamble*, Benjamin Williams, Samara Vilca, Ryan Logan

UMass Chan Medical School / Boston University School of Medicine, Worcester, Massachusetts, United States

Background: The opioid epidemic remains one of the greatest public health challenges in North America. In the past decade, fentanyl, a highly potent synthetic opioid, has risen as the greatest driver of overdose deaths, and chronic opioid use. Recently, the a2-adrenergic agonist, xylazine, has been deemed an emerging threat as it has become increasingly prevalent as a contaminant in the illicit fentanyl supply. Clinical reports indicate that co-administration of fentanyl and xylazine produces parasomnias, a group of sleep disorders associated with abnormal behaviors, suggesting unique sleep disruption effects from fentanyl-xylazine co-use. Indeed, most individuals that use chronic opioids experience persistent sleep disruption which fails to abate on maintenance therapy and can last months following opioid cessation. Furthermore, xylazine contamination has been shown to cause severe wounds in the periphery (distinct from the site of drug injection) and potentially exacerbate respiratory depression. Despite its alarming effects, to date, few studies have examined the effects of fentanyl-xylazine co-exposure. Utilizing a mouse model of fentanyl-xylazine addiction, we investigated the changes associated with fentanyl-xylazine co-administration and the molecular mechanisms that contribute to its effects on sleep-wake behavior, breathing, and skin.

Methods: Male and female C57BL/6J mice were administered saline, xylazine (5mg/kg), fentanyl (320ug/kg), or fentanyl-xylazine (6 mice per sex/group, n = 48 mice). Another cohort of fentanyl-xylazine mice was also administered saline (control) or yohimbine (a2-adrenergic antagonist) (4 mice per sex/group, n = 16 mice). Sleep-wake behavior was recorded simultaneously using non-invasive piezo sensors from intoxication through protracted withdrawal. Changes in percent as well as bout duration of wake, non-rapid eye movement sleep (NREMS), and rapid-eye movement sleep (REMS) were quantified. The following behaviors were also evaluated: open field, nesting, grooming, and spontaneous withdrawal. Brain and skin were extracted, flash frozen, and assessed with RNAscope and qPCR to investigate molecular changes associated with fentanyl-xylazine exposure. Given the known differences in opioid-use and sleep among males and females, all measures were evaluated for sex-differences. A two-way (repeated measures) ANOVA was used to assess significance (Treatment X Sex or Treatment X Zeitgeber or Time). All animal studies were approved by the Institutional Animal Care and Use Committee at University of Massachusetts Chan Medical School and were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Results: We show for the first time that fentanyl-xylazine exposure decreases NREMS during the dark period (p < 0.01), blunting the usual fentanyl sleep rebound during drug administration. Likewise, both wake (p < 0.001) and NREMS (p < 0.05) bouts were significantly decreased in the fentanyl-xylazine group compared to fentanyl alone. The average breathing rate per second was significantly decreased at the time of injection (p < 0.05) but not an hour after (p > 0.05) in fentanyl-xylazine mice compared to fentanyl only mice. Yohimbine was able to reverse the effects of xylazine in both cases. No differences in sleep-wake behavior (p > 0.05) were found during withdrawal between fentanyl and fentanyl-xylazine exposure. However, acute withdrawal showed that fentanyl-xylazine significantly decreased rearing (p < 0.05) with the largest reduction compared to all other groups. No differences were seen in measures of anhedonia in the nestlet test and sucrose splash testing among all groups. Furthermore, time spent in the center of open field was increased during acute withdrawal in fentanyl-xylazine mice compared to fentanyl only mice. RNAscope of the locus coeruleus showed differential pattern of expression of a2-adrenergic and mu-opioid transcripts. Skin showed no differences in markers of TNF-alpha and fibronectin across groups.

Conclusions: Our findings show fentanyl-xylazine exposure uniquely worsens sleep-wake behavior and spontaneous withdrawal behavior. Importantly, this work suggests that low-dose xylazine accelerates, but does not enhance opioid-induced respiratory depression at low-doses and that these effects are largely reversible via a2-adrenegic antagonism. These findings may help provide new or repurposed treatments for polysubstance use, especially opioid-xylazine co-use.

Keywords: Fentanyl, Xylazine, sleep architecture, circadian

Disclosure: Nothing to disclose.

Kelvin Lim*, Erich Kummerfeld, Jazmin Camchong

University of Minnesota, Minneapolis, Minnesota, United States

Background: Treatment for substance use disorders has been hampered by a one-size-fits-all approach, as it has been difficult to quantify the known heterogeneity found in the population that can be utilized for personalized treatment selection. Advancements in technology such as smartphones have made the collection of personal, intensive longitudinal data (ILD) a reality. Realizing benefits from the data, however, requires the ability to use it to develop useful models. Among the types of useful models, causal models are able to explicitly predict the effects of manipulations, making them key to developing and applying effective interventions. Most methods for learning causal effects require either experimentation or knowing the structure of the model to be known a priori. Experimentation is difficult and expensive in this population, however, and it is not only plausible but likely that differing structures is part of the heterogeneity of the substance use disorder population. This gap can be solved instead by using causal discovery analysis methods, which enable the identification of the best fitting causal model structure from observational data, including ILD. We applied causal discovery analysis to ILD collected from persons with substance use disorder, to quantify the individual heterogeneity in causal factors for the ability to resist using.

Methods: Data was collected on 12 persons with substance use disorder in treatment. Data collection was conducted over 4 weeks, 3-4 times/day and included ecological momentary assessment (EMA) surveys on craving, affect, drug use and agency. Following handling of missing values, transformations, and standardization, the Greedy Fast Causal Inference (GFCI) algorithm was applied to the processed data to search for the best-fitting causal model for each subject’s dataset. Structural equation modeling, using the causal model, is used to determine the effect sizes of the causal relationships among all of the included variables.

Results: From the individual causal models, feeling in control of life and positive affect were found to be positively causal for the ability to resist using in 7/12 and 3/12 subjects respectively. Craving and negative affect were found to be negatively causal for the ability to resist using in 5/12 and 5/12 subjects respectively.

Conclusions: In this pilot work, individual causal models from ILD demonstrated heterogeneity in causal factors for ability to resist using. These models may be useful in developing personalized interventions.

Keywords: Precision Medicne neuropsychiatric diseases, Alcohol and substance use disorders, EMA, causal discovery analyses

Disclosure: Nothing to disclose.

Miguel Angel Lujan Perez*, Reana Young-Morrison, Joseph Cheer

University of Maryland, School of Medicine, Baltimore, Maryland, United States

Background: The increasing prevalence of cannabis use during pregnancy has raised significant medical concerns, primarily related to the presence of Δ9-tetrahydrocannabinol (THC), which readily crosses the placenta and impacts fetal brain development. Previous research has identified midbrain dopaminergic neuronal alterations related to maternal THC consumption. However, the enduring consequences that prenatal cannabis exposure (PCE) has on striatum-based processing during voluntary reward pursuit have not been specifically determined.

Methods: To address this, we use in vivo multiple single-unit electrophysiological recordings in control and PCE rats, to monitor the activity of NAc neurons during effortful food pellet seeking. We next exploited different computational approaches to uncover maladaptations in stratum-based processing of rewards and its preceding cues.

Results: Using Bayesian logistic regression models, we determine that in utero cannabis caused a disproportional overrepresentation of reward-paired cues by NAc neuronal firing rates. We next asked whether differential encoding dynamics between treatment groups dynamically diverged as a function of reward cost. To address this question, we employed a novel unsupervised algorithm (Tensor component analysis) to uncover demixed, low-dimensional neuronal dynamics across multiple timescales. Our results link the higher propensity to work harder for food rewards with an overreactive representation of cues as effort increases. These alterations encompassed an increased willingness to work for natural and drug rewards, upregulated dopamine release responses to outcome-predictive cues and elevated impulsivity.

Conclusions: In conclusion, we provide evidence that prenatal exposure to THC leads to modifications in striatum-based processing consistent with an increased risk for excessive reinforcement. Collectively, the consequences of PCE here unveiled correspond to well-established neurodevelopmental risk factors for drug use disorders in the adult offspring.

Keywords: Prenatal Cannabis Exposure, Drug use disorders, Computational Neuroscience

Disclosure: Nothing to disclose.

Esther Visser*, Li Xu, Michele Petrella, Sanne Toivainen Eloff, Markus Heilig

Linköping University, Linköping, Sweden

Background: Alcohol use disorder (AUD), a major cause of morbidity and mortality worldwide 1,2, develops in a significant minority of alcohol users. It is characterised i.a. by choice of alcohol over healthy rewards, and continued alcohol use despite negative consequences (“compulsivity”) 2,3. In contrast, neurobiological research on AUD mechanisms typically relies on animal models where alcohol self-administration (SA) is studied in absence of alternative rewards or negative consequences4. This may explain why decades of research into the neurocircuitry of alcohol addiction have failed to produce effective treatments.

Previous research from our lab has shown that a minority of rats chooses alcohol over valuable alternatives and shows compulsive alcohol SA 5,6. Compulsivity was defined as continuation to lever press for alcohol rewards despite negative consequences, a mild electric foot-shock. This negative consequence suppresses responding for alcohol in most rats, while a substantial minority persists despite it. Neurobiological studies found that activity of central amygdala (CeA) neurons is necessary for compulsive alcohol SA 6–8.

Compulsive alcohol SA preferentially activated CeA neurons expressing protein kinase-C delta (PKC-δ). Both chemogenetic suppression of compulsivity-activated CeA neurons and knockdown of PKC-δ in the CeA reduced compulsive alcohol SA6. To gain access to this neuronal population in rats and to further delineate their role in compulsive alcohol SA, we recently developed a Prkcd::cre ratline using Crispr-Cas9 technology 9. We confirmed known projection targets of CeA PKC-δ neurons and replicated functional consequences of their activation previously found in mice. Here, we expressed an activating Designer Receptor Exclusively Activated by Designer Drugs (DREADD; hM3Dq) to investigate whether activation of the CeA PKC-δ population is sufficient to drive compulsive alcohol seeking.

Methods: The studies were conducted in accordance with the EU Directive 010/63/EU, as implemented in Swedish law, and approved by the National Committee for Animal Research in Sweden and the Local Ethics Committee for Animal Care and Use at Linköping University.

Adult male Prkcd::cre rats were group-housed and had ad libitum access to food and water in a temperature- and humidity-controlled facility. Animals were kept on a 12h light-dark cycle, and all experiments were performed in the dark phase of the cycle.

Animals were trained to lever press for an alcohol reward (100 μl of 20% alcohol) in standard operant chambers in 30 min sessions, starting on fixed-ratio (FR) 1 and progressing to FR2. Reward delivery started a 5 sec time-out in which lever presses were not rewarded, and this was signalled by a yellow cue-light above the alcohol-paired lever. Lever presses on a second lever had no programmed consequences (“inactive lever”). Sessions were conducted 5 days a week, until SA rates stabilised. Compulsive alcohol SA was assessed as responding for alcohol reward despite the contingent delivery of foot shock punishment (0.25mA). A resistance score was calculated as: punished alcohol deliveries / (punished alcohol deliveries + mean alcohol deliveries last 3 non-punished sessions).

After compulsive alcohol SA was established and stabilised, animals underwent stereotactic surgery. They were anesthetised by isoflurane and received bilateral microinjections targeted at the CeA (AP:-2.65, ML + -/4.5, DV -8.4mm from Bregma) of either 0,5 μl AAV5-hSyn::DIO-mCherry (control; n = 11) or AAV5-hSyn::DIO-hM3Dq-mCitrine (n = 10). Following two weeks of recovery from the surgery, animals went back to punished alcohol SA sessions. After twelve sessions, all animals received systemic injection of DREADD-agonist deschloroclozapine (DCZ; 0.1mg/kg, i.p.) 15 min before a punished alcohol SA session.

Finally, we assessed whether activation of PKC-δ neurons in the CeA modulated shock sensitivity, by injecting animals with DCZ before exposing them to a series of foot shocks with increasing intensity (every 30s + 0.05mA, starting at 0.15mA).

Results: Chemogenetic activation of PKC-δ neurons in the CeA significantly increased compulsive alcohol SA, while DCZ injection did not affect responding in the control group (RS: Group x Session F1,19 = 4.842, p = 0.040; Šídák’s multiple comparisons test S12 vs Test mCherry p = 0.910, hM3Dq p = 0.0058). Furthermore, the increase in responding was reversible; in the next session, under drug-free conditions, compulsive alcohol SA returned to baseline levels. Importantly, shock sensitivity was not affected. Of note, whether this effect also applies to female rats is the subject of ongoing studies.

Conclusions: This preliminary study shows that activation of PKC-δ neurons in the CeA using the newly developed Prkcd::cre rat line is sufficient to enhance compulsive alcohol SA. Thereby, we further emphasise the pivotal role of these neurons in alcohol use despite negative consequences, and this may therefore be an interesting target for the development of novel therapeutic strategies in AUD.

References:

1 World Health Organization (2019)

2 Carvalho AF et al. The Lancet 394, 781–792 (2019)

3 Black DW et al. (American Psychiatric Pub, 2014)

4 Venniro M et al. Nat Rev Neurosci 21, 625–643 (2020)

5 Augier E et al. Science 360, 1321–1326 (2018).

6 Domi E, Xu L et al. Sci. Adv. (2021).

7 Gilpin NW et al. Biol. Psychiatry 77, 859–869 (2015)

8 Lesscher HMB et al. Rev. Neurosci 23, 731–745 (2012)

9 Toivainen Eloff S et al. J. Neurosci. e0528242024 (2024)

Keywords: Alcohol, PKCdelta, Chemogenetics, Compulsivitiy

Disclosure: Nothing to disclose.

Samantha Ayoub*, Neal Jha, Sunitha Vemuri, Arpi Minassian, Adam Halberstadt, Jared Young

University of California San Diego, San Diego, California, United States

Background: Synthetic opioids, primarily fentanyl, are the leading cause of lethal drug overdoses in the US, which have dramatically increased 7.5 fold over the last decade. Acute relapse following opioid use detoxification drives a large proportion of overdose deaths, presumably through reduced tolerance. The drivers of relapse in abstinent individuals are not completely understood, though somatic (i.e. physiological) and affective (i.e. mood-altering, cognitive) withdrawal symptoms are thought to be leading contributors. Unlike somatic symptoms, which dissipate within weeks of abstinence, affective symptoms can occur for years after drug cessation and are therefore more likely to drive relapse during protracted withdrawal. Studying the causal impact of drug exposure on cognition during abstinence in the clinic remains difficult, however, since cognitive deficits may precede drug use. To help guide answers to such questions, we hypothesized that somatic withdrawal behaviors and increased risk-taking (measured by the cross-species Iowa Gambling Task; IGT) would occur during abstinence after fentanyl exposure in rats.

Methods: Twenty Fischer 344 rats (50% female) were trained to respond in operant chambers for milkshake reward. Animals were then exposed to intermittent fentanyl (n = 12) using a protocol previously shown to produce robust somatic withdrawal 16-hours post exposure: 9 days of exposure over 11 days, 0.1 mg/kg intraperitoneal injections twice-daily. Controls (n = 8) received saline injections. Animals were tested in the IGT to measure risk-based decision-making during drug exposure and at multiple time points during abstinence (17h post injection, and every 5 days thereafter). Difference scores (number of safe choices - risky choices) were the main outcome variable, with riskier behavior resulting in lower difference scores, as in the clinical IGT. Somatic withdrawal signs were also measured. Sex did not produce main or interactive effects (Fs < 2, ps > 0.1) in initial analysis, therefore data were collapsed by sex. One-tailed independent samples t-tests were utilized based on a priori hypotheses. One animal was removed from the IGT analyses due to insufficient trials on multiple testing days.

Results: IGT performance was comparable between saline- and fentanyl-treated animals 4h following drug administration on the 5th exposure day [t(17) = 0.308, ns]. At 16h following the last drug exposure, animals treated with fentanyl, but not saline, displayed robust somatic withdrawal [t(18) = 5.535, p < 0.001]. Importantly, at 17h post injection, animals in the fentanyl group exhibited lower difference scores (i.e. riskier behavior) in the IGT [t(17) = 1.97, p < 0.05]. After five days of abstinence, fentanyl-exposed animals no longer had observable somatic withdrawal signs [t(18) = 1.07, ns], and no longer differed in IGT performance [t(17) = 0.386, ns]. Interestingly, cognitive deficits in fentanyl-exposed animals tended to reappear in the IGT at 10 days of abstinence [t(17) = 1.53, p = 0.073], but were no longer observed at 15 days of abstinence [t(17) = 0.362, ns].

Conclusions: Fentanyl-exposed rats did not exhibit cognitive deficits in the IGT when drug was on board, but performed worse than saline treated rats at multiple time points during drug abstinence. Hence, the deficits in IGT performance in fentanyl-exposed rats were specific to drug withdrawal. Further, the IGT impairments observed during abstinence in fentanyl-treated animals were observed for longer than somatic withdrawal signs, consistent with the human literature indicating that former chronic opioid users experience protracted affective withdrawal. The cognitive deficits in the IGT were not always consistent across each timepoint, suggesting potential transient impairments. These inconsistencies could have also occurred from insufficient sample sizes, and follow-up experiments are necessary to validate the effects observed in this study. Future experiments will aim to identify neural markers of impaired IGT performance in animals, including contributors to their possibly transient nature. The results indicate that the fentanyl exposure protocol utilized was sufficient to induce cognitive deficits in rats performing a cross-species task during periods of acute and protracted opioid abstinence. Hence, this model can be used to explore the impact of opioid withdrawal on other translatable cognitive tasks and leveraged to explore neural mechanisms contributing to such behavior.

Keywords: Fentanyl withdrawal, cognition, cross-species translation, Risk-based decision making

Disclosure: Nothing to disclose.

Armando Salinas*, Charles Levy, Jacob Mitcham, Logan Slade

Louisiana State University Health Sciences Center Shreveport, Shreveport, Louisiana, United States

Background: Alcohol use disorder (AUD) affects over 29.5 million people in the United States and has adverse health and socioeconomic impacts totaling over $240 billion annually. FDA approved medications for AUD are limited and, thus, a further understanding of neurobiological mechanisms underlying AUD is required for the development of new treatments. Acetylcholine plays a pivotal role in normal decision making processes. Given the dysfunction in decision making processes in patients with acutely intoxicating blood ethanol levels as well as in patients with an AUD, acetylcholine signaling becomes an attractive pharmacotherapeutic target. Therefore, the experiments in this study aimed to determine how acute and chronic alcohol affect acetylcholine signaling in the dorsal striatum (DS).

Methods: To examine the acute effects of alcohol on acetylcholine, we performed cell-attached electrophysiological recordings of dorsal striatal cholinergic interneurons (CINs) from ChAT-Cre-TdTomato mice. We also assessed acetylcholine release following acute ethanol treatment using the intensity-based acetylcholine sensing fluorescent reporter, iAChSnFR in both in vitro and in vivo experiments using brain slice photometry and in vivo fiber photometry, respectively. To model AUD, we used the chronic intermittent ethanol vapor exposure model (CIE), which is considered a model of ethanol dependence. We again used the iAChSnFR biosensor and exposed mice to CIE treatment to induce dependence and then performed brain slice photometry three to five days following the final ethanol exposure period (to ensure that there was no residual ethanol on board at the time of these experiments). To assess any potential chronic ethanol-induced loss of CINs, we treated mice to CIE and perfused the brains three days following the final ethanol exposure. We then performed immunofluorescent staining for choline acetyltransferase (ChAT) and counted all ChAT-ir positive cells.

Results: We found that acute ethanol (40mM) treatment depressed CIN firing rate ~25% (n = 5 cells). Accordingly, we also found a decrease in evoked acetylcholine release following acute ethanol treatment in dorsal striatal brain slices (n = 6 slices) and in the dorsal striatum of awake behaving mice (2g/kg ethanol; n = 4 mice). Following CIE treatment, we found that evoked acetylcholine release was depressed in the medial, but not lateral, subdivision of the striatum (n = 6-10 slices). To assess the cause of this regional acetylcholine release deficit, we performed a stereological cell count and found that ChAT-ir positive neurons were decreased in the dorsomedial, but not dorsolateral, striatum of CIE treated mice relative to controls (n = 6-10 mice). Given the (minor) contribution of the pedunculopontine cholinergic nucleus to striatal acetylcholine release, we also counted ChAT-ir positive neurons in that region but found no significant difference between control and ethanol treated mice.

Conclusions: Altogether, our data show that both acute and chronic ethanol depress dorsal striatal acetylcholine release in subregion dependent manner.

Keywords: Alcohol use disorder and drug addiction, acetylcholine, Dorsal striatum

Disclosure: Nothing to disclose.

Kaitlin McManus*, Erica Grodin, Naomi Eisenberger, Lara Ray

University of California, Los Angeles, Los Angeles, California, United States

Background: Alcohol modulates the innate immune system. Chronic alcohol consumption is associated with increases in proinflammatory cytokines (interleukin (IL)-6, IL-8, IL-10, tumor necrosis factor alpha (TNF-α)). These increases in peripheral inflammatory markers may be associated with behavioral increases in alcohol consumption, craving for alcohol, and negative mood. To study the role of inflammation in alcohol use, experimental methods including an endotoxin administration can be utilized. Endotoxin can be safely administered to humans at low doses to produce a safe, dose-respondent, inflammatory response as seen by increases in cytokines and physiological symptoms. However, endotoxin has yet to be successfully administered in human subjects who drink alcohol heavily. Therefore, this randomized, placebo-controlled, clinical trial conducted an endotoxin challenge in individuals who drink alcohol heavily and leveraged data from matched controls to investigate the inflammatory, physiological, and behavioral responses to endotoxin.

Methods: A total of 40 participants completed the protocol. Participants included 20 individuals who drink alcohol heavily (n = 8 female) and 20 age and sex-matched control individuals (n = 8 female). Participants were randomly assigned to receive a bolus intravenous infusion of either low-dose endotoxin or placebo. Blood samples were collected at baseline, one hour, two hours, and four hours post-infusion to assess plasma levels of peripheral cytokines (IL-6, IL-8, IL-10, TNF-alpha, interferon gamma (IFN-gamma). Measures of sickness symptoms, physiological response (temperature, heart rate, blood pressure), and mood (self-reported negative and energetic mood) were collected for all participants at hourly intervals. Heavy drinking participants additionally completed measures of cue-induced craving at baseline and two hours post-infusion. Repeated measures of mixed-model analysis of variance were performed using SAS version 9.4. Group (heavy drinking vs. control) and treatment (endotoxin vs. placebo) were between-subjects factors and time (T0, T1, T2, T4) was a repeated measures factor. Additional treatment and time models were performed with individuals who drink alcohol heavily to assess cue-induced craving for alcohol.

Results: Cytokine values did not differ between group or treatment at baseline (ps > 0.05). Analyses revealed three-way interactive effects between group, treatment, and time on pro-inflammatory cytokine response for TNF-alpha, IFN-gamma, IL-6, and IL-8 (F3,107 = 2.91, ps < 0.039; F3,107 = 8.54, p < 0.001; F3,107 = 3.82, p = 0.012; F3,107 = 6.48, p < 0.001). Specifically, in participants who received endotoxin, TNF-alpha levels were significantly lower for individuals who drink alcohol heavily compared to matched controls at T2 (p = 0.005), while IFN-gamma, IL-6, and IL-8 levels were significantly lower for individuals who drink alcohol heavily compared to matched controls at T4 (ps < 0.014). Sickness symptoms were significantly increased at T2 in individuals who received endotoxin regardless of group (F1,107 = 15.96, p < 0.001). A significant main effect of treatment on systolic blood pressure emerged regardless of group or timepoint (F1,107 = 6.82, p = 0.010) such that systolic blood pressure was increased in individuals who received endotoxin. A significant treatment by time interaction emerged for diastolic blood pressure, temperature, and heart rate (F3,108 = 3.57, p = 0.017; F3,108 = 3.42, p = 0.020; F3,107 = 12.19, p < 0.001) such that diastolic blood pressure and heart rate were significantly increased by endotoxin at T2 regardless of group (ps < 0.010) and temperature and heart rate were significantly increased by endotoxin at T4 regardless of group (ps < 0.001). Moreover, a significant group by time interaction emerged for heart rate (F3,108 = 3.18, p = 0.027), such that individuals who drink alcohol heavily had marginally significantly lower heart rate at T1 compared to matched controls (p = 0.069). There were no significant group, treatment, or time effects on negative mood (ps > 0.05), though there was a significant time effect on energetic mood (F3,107 = 7.20, p < 0.001) such that energy decreased over time regardless of treatment or group. Lastly, within the individuals who drink alcohol heavily, a significant treatment by time interaction emerged (F1,52 = 5.02, p = 0.029) such that individuals who received endotoxin had significantly less cue-induced craving for alcohol at T2 compared to T0 (p < 0.001).

Conclusions: The acute endotoxin challenge successfully induced a safe inflammatory response in individuals who drink alcohol heavily that resulted in similar experiences of sickness symptoms, blood pressure, and temperature compared with age and sex-matched controls. In comparison to matched controls, individuals who drink alcohol heavily had a blunted proinflammatory response and faster recovery from the inflammatory challenge. Moreover, within heavy drinking individuals who received endotoxin, cue-induced craving for alcohol was decreased at the time of peak inflammatory response but persisted despite the experience of sickness symptoms. Individuals who drink alcohol heavily may experience consistent ramping up of their immune system from chronic drinking, leading to the appearance of a habituated inflammatory response following the endotoxin challenge compared to matched controls.

Keywords: Proinflammatory cytokines, Alcohol use, cue-induced craving

Disclosure: Nothing to disclose.

Sierra Stringfield*, Mary Torregrossa

University of Pittsburgh, Pittsburgh, Pennsylvania, United States

Background: Cannabis is the most widely used illicit substance among adolescents. Adolescent cannabis use is linked to an increased likelihood of developing multiple neuropsychiatric disorders, including schizophrenia, mood and anxiety disorders, and substance use disorders. Adolescent exposure to delta-9-tetrahydrocannabinol (THC), the primary psychoactive component in cannabis, has been shown to produce lasting alterations in key brain regions that contribute to these neuropsychiatric illnesses, which are also undergoing critical development during adolescence. Specifically, THC exposure may interfere with the activity of specific cell types within the prefrontal cortex, a brain region required for executive function and cognitively demanding tasks, such as working memory. This interference may disrupt the balance between excitatory and inhibitory neurotransmission, altering the ability of intracortical networks to support complex behaviors.

Methods: To investigate the long-term impacts of adolescent THC exposure on prefrontal cortex neuron activity, we conducted a series of experiments in adolescent male and female rats. Rats self-administered THC intravenously from postnatal day 32-51, while control rats self-administered a vehicle solution. One cohort of animals expressed the genetically encoded calcium indicator CamKII.GCaMP6f in the prelimbic prefrontal cortex and underwent miniscope recordings to measure the activity of individual principal neurons. A second cohort expressed DLX.GCaMP8f and underwent fiber photometry recordings to measure population-level activity of inhibitory neurons in the prelimbic prefrontal cortex. After recovery from surgery, all rats were trained to perform a delayed match-to-sample working memory task. Rats were required to maintain a representation of a specific response during a variable delay period before executing the correct response to receive a reward. Miniscope and fiber photometry recordings were conducted as rats performed the task at increasingly difficult delay lengths. All data were analyzed by t-test, one- or two-way ANOVA (α = 0.05 for all analyses), or chi-square test when appropriate. All experiments were approved by the Institutional Animal Care and Use Committee at the University of Pittsburgh.

Results: In the miniscope recordings, a total of 372 cells were recorded from n = 5 rats in the THC group, and 223 cells from n = 6 rats were recorded from control rats. We found that excitatory principal neurons from THC-exposed rats exhibited significantly larger increases in activity during the delay phase prior to a correct response on difficult trials, compared to control rats. Additionally, there was an increase in the proportion of cells that positively modulated their activity prior to an incorrect response in THC-exposed rats as the task became more difficult (p < 0.05 for all analyses). In the fiber photometry recordings, we found that inhibitory interneurons in THC-exposed rats showed blunted activity at reward delivery and were less responsive at the initiation of trials that would lead to an incorrect response, regardless of sex (p < 0.05).

Conclusions: These results demonstrate that adolescent THC exposure produces long-lasting effects on the activity patterns of both excitatory and inhibitory prefrontal cortical neurons during the performance of cognitively demanding tasks. Disruptions in the balance and coordination of these neuronal cell types may underlie the increased risk for neuropsychiatric disorders associated with adolescent cannabis use. The observed changes in neuronal activity, such as heightened excitatory responses during working memory demands and blunted inhibitory responses, could contribute to the cognitive deficits and psychiatric symptoms reported following adolescent THC exposure. Further research is needed to fully understand the mechanisms by which adolescent cannabis use alters the development and function of the prefrontal cortex, with the goal of informing prevention and intervention strategies.

Keywords: Adolescence, cannabinoid, working memory

Disclosure: Nothing to disclose.

Matthew Pomrenze*, Jason Tucciarone, Gavin Touponse, Nicholas Denomme, Robyn St. Laurent, Jinhee Baek, Allen Chen, BaDoi Phan, Julia Soares, Daniel Cardozo Pinto, Michaela Guo, Amei Shank, Andreas Pfenning, Zachary Freyberg, Neir Eshel, Robert Malenka

Stanford University, Stanford, California, United States

Background: The opioid crisis has surged over the past decade, leading to a critical need for effective treatments for opioid use disorder. The aversive nature of opioid withdrawal and its associated environmental cues act as potent drivers of relapse and increase the risk of future overdose. Changes in dopamine (DA) release in the nucleus accumbens (NAc) are known to be involved in opioid reward and withdrawal, with increased release during intoxication and decreased release during withdrawal. Despite these established dynamics, the underlying neural mechanisms remain controversial.

Methods: We used a combination of strategies in female and male wild-type and transgenic recombinase driver mice. Morphine was administered systemically to produce opioid dependence and withdrawal (n = 10-20 mice). Cell activity was monitored in vivo with genetically-encoded dopamine and calcium sensors (n = 6-8 mice), while activity was manipulated with optogenetic and chemogenetic tools (n = 8-14 mice). Genetic deletion of opioid receptor genes was achieved with site-specific delivery of Cre recombinase in specific driver lines (n = 10-16 mice). Single cell RNA sequencing data from NAc tissue collected from naïve mice and striatal tissue from post-mortem humans was analyzed for unique cell-types. Electrophysiology was performed in acute slices harvested from midbrain and NAc (n = 10-25 cells/4-8 mice). Opioid withdrawal aversion was tested in conditioned place aversion procedures with naloxone (n = 8-16 mice). All studies were analyzed with unpaired, two tailed-tests, two-way ANOVAs, repeated measures when appropriate, and mixed effect linear models when appropriate.

Results: Acute withdrawal from morphine resulted in strong avoidance of withdrawal-paired contexts and significant reductions in DA release in the NAc. Recordings of DA axonal activity in the NAc also showed strong inhibition during withdrawal. These effects were not mediated by GABA neurons of the midbrain. Analyses of single-cell RNA seq data from mouse NAc identified a unique cell-type enriched with the mu opioid receptor (Oprm1) and dopamine receptor 1 (Drd1), but mostly lacking dynorphin (Pdyn). This eccentric cell-type instead expressed high levels of the transcription factor Tshz1. This cell-type was also found in human striatum. Tshz1 neurons were inhibited by morphine but exhibited strong rebound activity during withdrawal. Chemogenetic inhibition of Tshz1 neurons or genetic deletion of Oprm1 specifically in these cells blocked behavioral aversion and reductions in DA release in the NAc during withdrawal. Stimulation of Tshz1 neurons in naive mice triggered strong inhibition of DA release in the NAc, mediated by local GABA release directly onto DA axons.

Conclusions: We have identified an eccentric, evolutionarily conserved cell-type in the NAc, enriched with Oprm1 and Tshz1, which exhibits robust control over local DA release. Through this mechanism, Tshz1 neurons mediate the aversive nature of opioid withdrawal.

Keywords: Opioid withdrawal, Nucelus accumbens, Dopamine, Mu opioid receptors

Disclosure: Nothing to disclose.

Tara Delorme*, Jennifer Ziegenfuss, Esha Sircar, Marianne Seney, Consuelo Walss-Bass, Michael Lodato, Ryan Logan

University of Massachusetts Medical School, Worcester, Massachusetts, United States

Background: The genome is constantly subjected to environmental and internal mutagens, leading to diverse forms of DNA damage in human tissues, including the brain. Uncorrected DNA damage can result in gene mutations or genomic variations, which threaten normal cellular function. Accumulated DNA damage in neurons has traditionally been associated with aging and neurodegenerative diseases. However, initial findings from our laboratory indicate a potential connection between opioid use disorder (OUD) and higher mutation rates in neurons. In our recent publication using single-cell RNA sequencing, we found neuronal-specific differences in pathways related to neurodegeneration, interferon response, and DNA damage were enriched in the dorsal striatum of postmortem brains of subjects with OUD, and in rhesus macaque monkeys chronically treated with either morphine or saline. The striatum is essential for controlling goal-directed, motivated, and reward behaviors, which are critically involved in opioid dependence, craving, and relapse. Here, we used an independent sample of postmortem human brains containing the striatum to validate the heightened DNA damage markers observed in OUD using classic markers of DNA damage and repair: 8-oxoguanine (8-oxoG) and 8-oxoGDNA glycosylase (OGG1). 8-oxoG is the oxidized version of guanine, is highly mutagenic and its effect on DNA is implicated in aging, and many neurodegenerative diseases. Thus, 8-oxoG can be used as a key biomarker for DNA damage due to oxidative stress. OGG1 is vital for removing 8-oxoG and preventing DNA mutations. Given that up to 60-70% of patients meet the criteria for sleep and circadian disorders, and altered sleep schedules affect oxidative DNA damage, we also examined 8oxoG and OGG1 immunofluorescence across time of day. Understanding DNA damage within the striatum of postmortem human brains from individuals with OUD can provide new insights into the molecular mechanisms driving OUD and potential therapeutic targets.

Methods: Postmortem dorsal striatum tissue punches (n = 6 subjects per sex per unaffected and OUD) were processed for immunofluorescence microscopy using antibodies targeting 8-oxoG and OGG1. Mean 8-oxoG and OGG1 immunofluorescence signal was quantified for 100 cells per case using Fiji (ImageJ) software. A line was drawn around the perimeter of the neuron cell body (visualized by pan-neuronal marker NeuN) and the mean grey value (absorbance units) was measured within the area in the 8-oxoG or OGG1 channels. This value, considered the ‘intracellular signal’, was normalized against the background immunofluorescence signal. A cosine regression analysis was conducted to access rhythmicity of the expression levels of DNA damage markers across the day, plotted using time of death. All experimental procedures were approved by the University of Massachusetts Chan Medical School’s Institutional User Committee.

Results: Our data indicates increased average nuclear 8oxoG and OGG1 intensity in OUD neurons compared to neurons from unaffected matched controls (Mann Whitney, 8oxoG p < 0.001 and OGG1 p < 0.001). Nuclear 8oxoG intensity did not differ by neuron area for OUD, but higher 8oxoG intensity was associated with smaller neurons in controls (linear regression slope, p = 0.025). Higher OGG1 intensity was associated with larger neurons in both OUD (p < 0.001) and controls (p < 0.001). Preliminary data revealed no significant rhythm in OUD or controls (p > 0.05). When comparing day versus night expression in mean nuclear 8oxoG and OGG1, no significant 2-way interactions (time of death x treatment) or main effects were observed (p > 0.05).

Conclusions: By describing the cellular levels of DNA damage in OUD striatal neurons, and how they may vary in the day versus night, we can leverage these data to gain insights into molecular alterations in human striatum associated with opioid addiction. The societal impact of OUD is profound, given its increasing prevalence across the United States. Persistent challenges in treatment efficacy and relapse prevention are evident, with over 90% of treated patients eventually experiencing relapse. This underscores the urgent need for deeper insights into the neurobiological underpinnings of OUD. Exploring DNA damage proteins can provide valuable insights into the molecular mechanisms and pathophysiology of OUD. This approach may offer new avenues for developing targeted therapies and biomarkers to improve treatment outcomes in individuals affected by OUD.

Keywords: Opioid addiction, DNA damage response, DNA breaks, Postmortem Human Brain Tissue, immunohistochemistry

Disclosure: Nothing to disclose.

Eun-Kyung Hwang, Amanda Wunsch, Marina Wolf*

Oregon Health and Science University, Portland, Oregon, United States

Background: Incubation of cocaine craving, a translationally relevant model for the persistence of drug craving during abstinence, ultimately depends on strengthening of nucleus accumbens core (NAcc) synapses through synaptic insertion of homomeric GluA1 Ca2 + -permeable AMPA receptors (CP-AMPARs). Here we tested the hypothesis that CP-AMPAR upregulation results from a form of homeostatic plasticity, previously characterized in vitro and in other brain regions, that depends on retinoic acid (RA) signaling in dendrites. Under normal conditions, ongoing synaptic transmission maintains intracellular Ca2+ at levels sufficient to suppress RA synthesis. Prolonged blockade of neuronal activity results in disinhibition of RA synthesis, leading to increased GluA1 translation and synaptic insertion of homomeric GluA1 CP-AMPARs.

Methods: Wildtype or transgenic rats (D1-Cre or A2a-Cre, crossed with a TdTomato reporter line) were used (approximately equal numbers of male and female rats in each group). Some rats self-administered saline or cocaine (6 h/d x 10 d; each infusion paired with a light cue) and underwent 40-60 d of home-cage abstinence before being used for whole-cell patch-clamp recording or tests for cue-induced cocaine seeking. This regimen elicits robust incubation of craving. Recordings were performed in NAcc medium spiny neurons (MSN) to assess the contribution of CP-AMPARs to synaptic transmission based on the rectification index (RI) or sensitivity to the selective antagonist Naspm. For assessment of cue-induced cocaine seeking, rats were returned to the operant box for a 1 h test during which responding in the previously active nose-poke hole delivered the cue but no cocaine infusion.

Results: Applying RA or a selective RARα agonist (1-2 h, in slice) to NAcc MSN from drug-naïve rats rapidly upregulated CP-AMPARs (elevated RI or Naspm sensitivity; p < 0.0001 or 0.01 RA or agonist vs vehicle; 13-18 cells/5-10 rats per group). This pathway was operative only in D1 receptor-expressing MSN (p = 0.02 D1+ vs D1- MSN; 11-15 cells/5-10 rats per group). In MSN recorded from rats that had undergone incubation of craving, this effect of RA was occluded; instead, interruption of RA signaling in the slice with the RA synthesis inhibitor DEAB normalized the incubation-associated elevation of synaptic CP-AMPARs measured 1-2 h after DEAB application (p < 0.0001 DEAB vs Veh or RA; 11-18 cells/5-6 rats per group). Related experiments showed that the incubation-associated elevation of synaptic CP-AMPARs and its reversal by DEAB were selective to D1 MSN (p < 0.0001, 7-16 cells/6-7 rats per group). Paralleling electrophysiological results, interruption of RA signaling in the NAcc of ‘incubated rats’ via intra-NAcc DEAB infusion 1 h prior to a seeking test normalized the incubation-associated elevation of cue-induced cocaine seeking (p = 0.01, DEAB vs veh; 7-9 rats/group). Although our studies were not adequately powered to detect sex differences, none were apparent.

Conclusions: These results implicate RA-mediated homeostatic plasticity in the incubation of cocaine craving. Specifically, they suggest that RA signaling becomes tonically active in the NAcc during cocaine withdrawal and, by maintaining elevated CP-AMPAR levels, contributes to the expression of the incubation of cocaine craving. They also show that CP-AMPAR elevation after incubation occurs exclusively in D1 receptor-expressing MSN.

Keywords: homeostatic plasticity, incubation of cocaine craving, nucleus accumbens core, calcium-permeable AMPA receptors, retinoic acid

Disclosure: Nothing to disclose.

Samara Vilca*, Tara Delorme, Ryan Logan

University of Massachusetts Chan Medical School, Worcester, Massachusetts, United States

Background: The rates of opioid use disorder (OUD) continue to increase. Indeed, many people with OUD will relapse withing the first few weeks and months. Despite significant research, the underlying mechanisms of opioid use and relapse are not fully understood. A better understanding of the molecular mechanisms contributing to OUD will aid in the development of new and improved treatments. Fentanyl, a synthetic opioid with far higher potency than other opioids, has led to an exponential increase in the number of deaths due to overdose. To study fentanyl use disorder (FUD), rodent models with face validity are necessary. Intravenous self-administration (IVSA), often termed the gold-standard of animal models for studying substance use disorders, has previously been implemented in rodent studies and have successfully recapitulated human patterns of addiction. However, many of these have used rats. Given the greater availability of genetic tools in mice, additional studies employing IVSA are key to interrogate the underlying cellular and molecular mechanisms governing FUD. Here, we established a model of FUD using mouse fentanyl IVSA. Additionally, we analyzed fentanyl-related behaviors to assess individual risk of FUD.

Methods: Adult male (n = 24) and female (n = 24) mice were implanted with an indwelling jugular catheter. Following recovery, mice were trained to self-administer fentanyl (or 0.9% saline) for 14 days in daily 2-hr sessions (5 d/week) under a fixed ratio 1 (FR1) schedule. Successful completion of criteria resulted in delivery of fentanyl (3 μg/kg/inf) or saline and activation of a cue-light, as well as a 20-sec timeout period in which active nose pokes were recorded but carried no scheduled consequences. Once responses for fentanyl stabilized, motivation was assessed using a progressive ratio schedule of reinforcement. Next, mice underwent 21 days of forced home-cage abstinence. On Day 1 and Day 21, mice underwent relapse testing, wherein active nose pokes yielded cue-light presentation, but no infusion was given. Twenty-four hours after the last IVSA session (n = 6 per Sex x Treatment) and following 21 days of abstinence (n = 6 per Sex x Treatment), brains were extracted and flash frozen. Then, the nucleus accumbens was micropunched, as this brain region is key in reward processing and addiction-like behaviors. Using RNA-seq, we determined differentially expressed genes between fentanyl and saline-administering mice, as well as abstinence. All experimental procedures were approved by the University of Massachusetts Chan Medical School’s IACUC and were conducted in accordance with the NIH’s Health Guide for the Care and Use of Laboratory Animals.

Results: Here we show both male and female mice will self-administer fentanyl, escalating their intake (F (13, 55) = 2.59, P = 0.01) and nose poking (F (1, 8) = 9.98, P = 0.01) over 14 sessions compared to saline. Mixed-effects models revealed that mice exhibit robust discrimination between the fentanyl active and inactive nose poke ports (F (2.67, 20.99) = 3.46, P = 0.04). Interestingly, mice did not display incubation of craving following 21 days of abstinence, which agrees with previous rodent and human studies. When components of fentanyl administration were operationalized, many were normally distributed. Shapiro-Wilk tests revealed normal distribution for fentanyl intake (p  =  0.87), fentanyl seeking (active nose pokes; p  =  0.44), escalation of intake (slope of intake change; p  =  0.96), motivation (progressive ratio breakpoint, p  =  0.83) and relapse potential (active nose pokes at 21-day relapse test; p  =  0.13), but not for strength of cue association (log of active nose poke latency; p  =  0.03). Additionally, Spearman’s rank-order correlation revealed that intake was strongly correlated with most other fentanyl-related behaviors, and motivation and escalation were also highly correlated. Z-scoring the behaviors for individual mice and calculating the sum of those Z-scores produced an FUD “Severity” score, which was also normally distributed (p  =  0.38. Furthermore, principal component analysis (PCA) of all fentanyl-related behaviors revealed “high-severity” mice exhibited different behavioral trajectories, with several mice displaying high intake, seeking, and cue-association, while others were highly motivated and displayed high risk for escalation and relapse. From RNA-seq, we found numerous differentially expressed genes related to pro-inflammatory signaling, neurotransmitter signaling, and extracellular matrix organization enriched in fentanyl-administering mice that persisted through prolonged abstinence. Furthermore, cell-type deconvolution indicated that microglia may play a significant role in molecular adaptations to fentanyl. Finally, we did not find any significant sex differences, neither behaviorally nor transcriptionally.

Conclusions: Our findings indicate that fentanyl IVSA in mice is a valid model for examining individual drug-taking and seeking-habits. Additionally, this model can be used to investigate and correlate other fentanyl-related maladaptive behaviors such as sleep disruption. Moreover, fentanyl-induced gene expression changes are robust and long-lasting, which has been shown in human postmortem studies. Finally, these studies further our understanding of the underlying molecular adaptations to fentanyl use in the hopes of developing novel therapies for this debilitating disease.

Keywords: opioid use disorder, intravenous drug self-administration, gene expression

Disclosure: Nothing to disclose.

Xinyi Li*, Peter Manza, Gene-Jack Wang, Natasha Giddens, Nora Volkow, Zhenhao Shi, Corinde Wiers

University of Pennsylvania, Philadelphia, Pennsylvania, United States

Background: Previous studies have suggested a link between opioid use and altered metabolic profiles that may increase the risk for metabolic syndrome, cardiovascular diseases, and type 2 diabetes. However, the study findings were widely inconsistent, possibly due to a lack of differentiation between synthetic opioids (i.e., heroin, fentanyl) vs. naturally derived opioids (i.e., crude opium and opium sap), and were limited to identifying potential underlying mechanisms. Here, we examined metabolic profiles in patients with opioid use disorder (OUD) and matched controls, and examined the degree to which an elevated inflammatory state statistically mediated the relationship between OUD and metabolic profiles.

Methods: We leveraged an active neuroimaging project at the NIH NIAAA intramural program studying the effects of medication for opioid use disorders in individuals with opioid use disorder (OUD) to compare metabolic parameters with non-OUD controls matched on demographics (age, sex, race, ethnicity, socioeconomic status, and education) using propensity score matching. Metabolic parameters included BMI, lipid profiles, blood glucose levels, and hemoglobin A1c, and markers of inflammation [i.e., C-reactive protein (CRP) and white blood cell count (WBC)]. We subsequently conducted mediation analyses to examine the role of CRP and WBC levels in mediating the relationship between OUD and altered metabolic parameters. Because the prevalence of other substance use disorders and smoking was higher in the OUD than in the control group, we examined the effects of OUD on metabolic profiles (1) unadjusted and (2) adjusted for other substance use disorders and smoking status.

Results: In the unadjusted model, the OUD (n = 110, mean age = 45.1 ± 12.2SD, 35 female) and control (n = 101, mean age = 44.8 ± 12.9SD, 27 female) groups did not differ in BMI, blood glucose levels, or hemoglobin A1c. There were significant group differences in lipid profiles, such that individuals with OUD demonstrated lower total cholesterol levels (t(209) = 3.0, p = 0.003) and high-density lipoprotein (HDL) cholesterol levels (t(209) = 2.5, p = 0.013), and higher triglyceride levels (t(187) = -2.1, p = 0.038) compared to controls. There were no group differences in low-density lipoprotein (LDL) cholesterol levels. Inflammatory markers, CRP (t(122) = -2.7, p = 0.009) and WBC (t(209) = -2.6, p = 0.01), were significantly higher in individuals with OUD than in controls and covaried negatively with HDL-cholesterol levels and positively with triglyceride levels. CRP and WBC significantly mediated the group differences in HDL-cholesterol and triglyceride levels but did not show significant associations with total or LDL cholesterol levels. In subsequent analyses that adjusted for all substance use disorders and smoking status, group differences in total- (F1,200 = 6.4, p = 0.012) and HDL-cholesterol (F1,200 = 6.7, p = 0.010) and triglycerides (F1,200 = 6.2, p = 0.014) remained significant. Higher CRP (F1,200 = 5.1, p = 0.026), but no differences in WBC, was observed in individuals with OUD than controls when adjusting for other SUDs and remained a significant mediator of the relationship between OUD and HDL-cholesterol.

Conclusions: Our results are consistent with prior findings of dyslipidemia in individuals with OUD, and expand to identify inflammation as a contributor. The findings may have implications in identifying targets for treatment to enhance the long-term health of individuals with OUD.

Keywords: opioid use disorder, Metabolic Biomarker, inflammation

Disclosure: Nothing to disclose.

William B. Lynch, Sophia A. Miracle*, Stanley I. Goldstein, Kelly K. Wingfield, Rhea Bhandari, Ethan T. Gerhardt, Ava Farnan, Binh-Minh Nguyen, Ida Kazerani, Gabriel A. Saavedra, Olga Averin, Britahny M. Baskin, Jacob A. Beierle, Martin T. Ferris, Christopher A. Reilly, Andrew Emili, Camron D. Bryant

Northeastern University, Boston, Massachusetts, United States

Background: Opioid use disorder (OUD) is a public health crisis. Prescription opioids have contributed to opioid deaths, especially oxycodone (OXY). OUD therapeutics comprise opioid substitution therapy and relapse rates remain high. Thus, there is a need for understanding OUD risk factors and improving therapeutics. OUD has a genetic component that remains undescribed; thus, gene discovery of opioid molecular and behavioral traits relevant to OUD will improve our understanding of drug action and prevention/treatment strategies. Using two nearly identical mouse substrains, BALB/cJ (J) and BALB/cByJ (By), we observed increased brain concentration of oxymorphone [OMOR] (OXY metabolite) in J vs. By females accompanied by increased state-dependent learning of OXY locomotion and OXY-CPP in J vs. By females. Importantly, OMOR is a much more potent and efficacious mu opioid receptor agonist and could mediate enhanced state-dependent OXY behaviors. OXY is metabolized to OMOR via cytochrome P450 2D (CYP2D) enzymes. Using a J x ByJ-F2 cross, we mapped a genome-wide significant quantitative trait locus (QTL) for brain [OMOR] accompanying state-dependent OXY-CPP peaking near Zhx2 (zinc fingers and homeoboxes 2). Co-mapping of a Zhx2 cis-eQTL from striatum and hippocampus further supported Zhx2’s candidacy. Zhx2 codes for a transcriptional repressor that decreases liver CYP expression in a sex-specific manner. The Zhx2 eQTL is mediated by a 6kb mouse endogenous retroviral insertion (MERV) in the J allele, causing vastly reduced Zhx2 expression. Here, we tested for validation of Zhx2 as a causal gene underlying brain [OMOR] associated with state-dependent OXY behaviors via constitutive Zhx2 knockout (KO) on the By background to model the Zhx2-MERV loss-of-function and via AAV-Zhx2 overexpression (OE) on the mutant J background. We predicted that Zhx2 loss-of-function would increase CYP2D expression, brain OMOR levels, and OXY state-dependent behaviors. Conversely, increasing Zhx2 OE would decrease CYP2D expression, brain OMOR levels, and OXY state-dependent behaviors.

Methods: For Zhx2 overexpression (OE), J mice with Zhx2-MERV loss-of-function received retro-orbital AAV (AAV8-TBG-mZhx2-P2A-eGFP) and intracerebroventricular (ICV) (AAV/F-CMV-mZhx2-P2A-eGFP.miR122) injections for liver vs. brain OE and behavioral testing commenced 3-5 weeks later. Zhx2 validation via constitutive gene targeting was accomplished via CRISPR-mediated Zhx2 Exon 3 knockout (KO) in By mice with the native wild-type Zhx2 genotype. All mice were assessed for OXY locomotion, state-free OXY-CPP, and state-dependent OXY-CPP. Following initial preference on Day (D) 1, mice received alternating OXY injections (1.25 mg/kg, IP) and saline (SAL; IP) on D2-D5. Mice were assessed for drug-free and state-dependent OXY-CPP on D8 (SAL, IP) and D9 (1.25 mg/kg OXY, IP). Brain hemispheres were extracted on D9 at 30 min post-OXY for molecular studies. Half-brains and liver lobes were used for immunoblotting and qPCR to validate Zhx2 and differential expression of Cyp2d’s, with some hemispheres used for Zhx2 immunohistochemical staining to validate brain OE. Half-brains from OXY-treated mice were used to quantify brain OXY and metabolite levels via mass-spectrometry, while half-brains from OXY-naive mice were used for transcriptomics via RNA-seq or proteomics via tandem mass-spectrometry. Statistics were run with t-tests or ANOVAs and post-hoc comparisons were run with Holms-Sidak corrections for significant effects or interactions. Benjamini and Hochberg’s false discovery rate was used to correct for multiple testing (-omics).

Results: Zhx2 liver OE resulted in > 100-fold increased expression (mRNA; p < 0.0001); however, the effects on brain [OMOR] and OXY behaviors were largely negative. Brain Zhx2 OE induced localized staining in lateral septum and hippocampus. Brain OE reduced state-dependent D9 OXY locomotion in females (AAV x Day: F2,60 = 9.33, p < 0.01, post-hoc: p < 0.0001), and males (AAV x Day: F2,58 = 3.94, p < 0.05, post-hoc: p < 0.05). Zhx2 KO in By females increased brain [OMOR] (t24 = 5.58, p = 0.0001), and a marked increase in OXY locomotion on D4 (p < 0.05), with no changes in KO males. Interestingly, KO females showed reduced D9 time on the OXY side at 10-15 min (Genotype x Treatment x Time Bin: F5,265 = 3.98, p < 0.01; post-hoc: p < 0.05). Conversely, KO males showed an increase in D9 time on the OXY side (Genotype x Treatment x Day: F2,62 = 5.29, p < 0.01; post-hoc: p < 0.001). Multiple pathways, e.g., small molecule and lipid metabolism, extracellular matrix process, inflammation, and cell activity were upregulated in KOs, plus astrocyte activity, extracellular matrix properties, and endothelial cell dysfunction.

Conclusions: Our findings support for Zhx2 as a quantitative trait gene underlying sex-specific brain [OMOR] associated with OXY state-dependent learning. However, the interplay between Zhx2 expression, background strain, brain region, and sex-specific behaviors are more complex than predicted and warrants further investigation of epistatic interactions between Zhx2 and background BALB/c genometype and potential social genetic effects. Pathways associated with astrocyte function, inflammation, and extracellular matrix structure suggest blood-brain-barrier disruption could contribute to the PK and behavioral profile of OXY and other drugs.

Keywords: Oxycodone, Proteomics, Immunohistochemistry

Disclosure: Nothing to disclose.

Utsav Gyawali*, Charlie Olson, Simar Monga, Shayna O’Connor, Michelle Bilotti, David De Sa Nogueira, Morgan James

Rutgers University, Piscataway, New Jersey, United States

Background: Sleep disturbances are common in individuals with cocaine use disorder (CUD) and worsen during withdrawal. Poor sleep is among the most commonly cited reasons for return to use (relapse), indicating a potential biological link between these phenomena. In preclinical models of CUD, the transition to uncontrolled drug use is marked by increased number and activity of orexin (hypocretin) neurons. Given the dual role for orexin neurons in driving reward and wakefulness, this system represents a potential therapeutic target for interventions designed to improve sleep and reduce relapse risk. Here, we tested the hypothesis that normalizing orexin signaling during cocaine abstinence in rats would improve both sleep and drug seeking outcomes. We also explored the ventral tegmental area (VTA) as a potential common site through which excessive orexin signaling mediates reward and arousal.

Methods: We first trained male (N = 8) and female (N = 9) rats to self-administer cocaine (0.055mg/kg/inf), after which lever pressing was extinguished (7d). In another group of rats, we trained rats do develop a conditioned place preference (CPP) for cocaine injections (10mg/kg), which was then extinguished (5d; male N = 10, female N = 11). Both groups were treated with the dual orexin receptor antagonist suvorexant (0 vs. 30mg/kg; p.o) 1h prior to the onset of the inactive period during extinction training. In a subset (male N = 3, female N = 5) of CPP rats, overnight sleep was monitored via EEG/EMG recordings throughout the experiment. In a separate group of cocaine CPP rats, we harvested samples from several reward and arousal brain regions including VTA and lateral hypothalamus (LH) following 48h cocaine abstinence (Cocaine N = 5M + 5F; Saline controls N = 5M + 5F;) and quantified changes in orexin and orexin receptor mRNA expression using qRT-PCR. In a final group of rats, we recorded the orexin fluorescent indicator (OxLight1) in VTA during waking. Cocaine extinction data were analyzed using day (days 1-5 or 1-7) x 2 dose (0, 30mg/kg) RM ANOVA. Similarly, time spent (in minutes) in Wake, NREM, and REM in 12h recording across extinction days was compared between suvorexant dose (0 vs 30mg/kg) using RM ANOVA. In qPCR experiments, orexin, orexin 1 (Ox1R) and 2 (Ox2R) receptor gene expression was compared between saline and cocaine CPP rats using Students t test. Peak OxLight fluorescence during baseline and wake was analyzed using the pMAT software. Significance for all tests was set at p < 0.05.

Results: Cocaine abstinence was associated with increased time in Wake and reduced time in NREM sleep during the first 3d of extinction; this was reversed by suvorexant treatment: D1 (W, p = 0.02, NREM, p = 0.043), D2 (W, p = 0.008, NREM, p = 0.009), D3 (W, p = 0.004, NREM, p = 0.007). Further, suvorexant administration during the inactive period facilitated next-day extinction of cocaine seeking in both self-administration (p = 0.034) and CPP (p = 0.040) rats. Abstinence in cocaine CPP rats was associated with increased expression of orexin mRNA in LH (p = 0.026), as well as increased Ox1R in VTA (p = 0.002). There was no change in orexin receptor levels in other reward and arousal regions. Preliminary fiber photometry data indicate that arousal events (waking) is associated with increased orexin binding compared to baseline in VTA (p = 0.002).

Conclusions: Blocking orexin signaling (via suvorexant administration) during cocaine abstinence normalizes sleep and reduces drug seeking. Changes in orexin receptor levels and peptide binding in VTA point to this region as a potential target of orexin signaling during abstinence. Ongoing studies are testing the causal role for orexin signaling in VTA in mediating sleep and drug-seeking outcomes.

Keywords: Dual Orexin Receptor Antagonists, Sleep disturbances, cocaine use disorder

Disclosure: Nothing to disclose.

Aida Mohammadkhani, Ijeoma Ifionu, Hadi Semizeh, Fatima Hamood, Min Qiao, Stephanie Borgland*

Hotchkiss Brain Institute, University of Calgary, Calgary, Canada

Background: The misuse of opioids and other prescription pain relievers has risen rapidly, becoming a major global health issue. Addiction is linked to neural circuit dysfunction, characterized by changes in synaptic transmission in the ventral tegmental area (VTA), a region involved in the incentive value of drug-related stimuli. Orexins (ox) and dynorphin (dyn) are co-expressed lateral hypothalamic (LH) neuropeptides that project to the VTA, having opposing effects on the firing activity of VTA dopamine (DA) neurons via orexin 1 (ox1) or kappa opioid (KOR) receptors, respectively. This study aimed to determine how chronic morphine alters the contributions of LHox/dyn to the firing of VTA DA neurons.

Methods: We expressed channelrhodopsin2 selectively in LHox/dyn neurons of orexin-cre mice. Mice were given 5 twice daily injections of escalating morphine or learned to self-administer oral fentanyl solution. Brain slices were extracted and we photostimulated terminals in the VTA while recording DA neuronal firing using patch clamp electrophysiology in neurons with identified projection targets.

Results: We previously showed that optical stimulation of LHox/dyn inputs in the VTA inhibited the firing of the majority of BLA-projecting VTADA neurons. This was replicated in saline-treated mice. In mice chronically exposed to morphine, regulation of LH neurons on DA-BLA neurons switches from inhibitory (71%) to primarily excitatory (63%). Similarly, in mice orally self-administering fentanyl, the effect of LHox/dyn release in BLA-projecting dopamine neurons becomes primarily excitatory (60%) compared to that in sucrose self-administering mice (30%). We found that this is associated with a reduced efficacy of dynorphin signaling in the VTA. Ongoing experiments will also address if there are changes in orexin signaling in this input as well.

Conclusions: Repeated opioid treatment shifts the LHox/dyn regulation of VTA dopamine neurons that project to the BLA to primarily excitatory. This may influence the responsiveness of the mouse to opioid-related cues.

Keywords: opioids, Ventral Tegmental Area (VTA), Dopamine, orexin, Dynorphin

Disclosure: Nothing to disclose.

Eden Evins*, Corinne Cather, Gladys Pachas, Kevin Potter, Harrison Reeder, Bryn Evohr, Meghan Costello, Jason Dufour, Kelly Casottana, Vanessa Iroegbulem, Jodi Gilman, Randi Schuster

Massachusetts General Hospital, Boston, Massachusetts, United States

Background: Electronic cigarette use among youth is common and associated with tobacco use disorder, initiation of combusted tobacco use, and negative health effects of vapor exposure. Few treatments have been tested in this population.

Methods: To evaluate the efficacy and acceptability of varenicline for nicotine vaping cessation in adolescents, we conducted a 3-group, randomized clinical trial (VIVA) of 12 weeks of intervention with follow up at 24 weeks in adolescents aged 16-25 with tobacco use disorder who vaped nicotine daily and reported no prior regular tobacco smoking. From June 2022 to May 2024, participants were randomized 1:1:1 to receive varenicline (n = 88) or identical placebo (n = 87) 0.5 mg per day for 3 days, 0.5 mg twice per day for 4 days, then 1 mg twice per day for 11 weeks, 12 weekly remote behavioral support sessions with lay counselors, and referral to TIQ text messaging support for vaping cessation or to TIQ texting support referral alone (n = 86). Outcomes were cotinine-verified continuous vaping abstinence in varenicline vs placebo groups for weeks 9 to 12 (primary) and weeks 9 to 24 (secondary) and continuous vaping abstinence from weeks 9-12 and 9-24 in varenicline vs TIQ texting support only and in placebo vs TIQ texting support only groups (secondary).

Results: Of 300 participants enrolled, 261 met inclusion criteria, were randomized and included in the analyses (mean age, 21.4 years, 53.3% female, 16.5% Hispanic), 256 completed the intervention, and 251 completed follow up. Four-week continuous abstinence rates at week 12 were 38.6% (34/88) in the varenicline group, 10.3% (9/87) in the placebo group, OR = 5.39, 95% CI: 2.35, 12.38, p < 0.001, and 5.8% (5/86) in the texting only group, varenicline vs. texting support only, OR = 10.18, 95% CI: 3.73, 27.83, p < 0.001; placebo vs texting support only, OR = 0.52, 95% CI: 0.17, 1.64, p = 0.264. Continuous abstinence rates from weeks 9 to 24 were 25% (22/88) in the varenicline group, 8.3% (7/87) in the placebo group, OR = 3.59, 95% CI: 1.43, 9.04, p = 0.007, and 2.3% (2/86) in the texting support only group, varenicline vs. TIQ texting support only, OR = 14.01, 95% CI 3.15, 62.29, p = 0.001, and placebo vs. texting support only, OR = 0.26, 95% CI 0.05, 1.29, p = 0.098.

Conclusions: Varenicline, with remote lay counselor behavioral support and TIQ texting support referral, demonstrated efficacy for vaping cessation in adolescents with tobacco use disorder.

Keywords: nicotine/substance use disorder, vaping, Adolescent, Varenicline

Disclosure: Nothing to disclose.

Mikala Zelows, Brooke Tyree, Anel Jaramillo*

University of Kentucky, Lexington, Kentucky, United States

Background: Forced abstinence in a mouse preclinical model of alcohol-use increases anxiety-like behavior and produces neuronal adaptations in the bed nucleus of the stria terminalis (BNST), a region critical for affective behavior. Activating BNST neurons innervated by the parabrachial nucleus (PBN), a sensory alarm, increases anxiety-like behavior in females relative to males. The PBN sends projections, containing the sexually-dimorphic neuropeptide pituitary adenylate cyclase activating peptide (PACAP), to the BNST that remain relatively unexplored in abstinence. We hypothesize an anxiogenic role for PBN(PACAP) neurons in alcohol abstinence. PACAP has a stimulatory role on the stress axis. Thus we hypothesize the anxiogenic role of PBN(PACAP) neurons in abstinence will be exacerbated with stress.

Methods: First to develop a model of stress that induces anxiety-like behavior, C57 female mice (n = 5/sex/stress group) received 30 min exposure to a daily restraint stressor. Following 4 restraint sessions, anxiety was measured with the novelty suppressed feeding task (NSFT). To investigate if inactivating PACAP neurons in the PBN decreased stress-induced anxiety in abstinence, PACAP-CRE female mice (n = 4-9/sex/genotype/stress group) received bilateral injections of CRE-dependent hM4D(Gi) DREADDs in the PBN. Mice performed home-cage voluntary drinking (10% w/v) or water 24/7 for 6 weeks using the two-bottle choice paradigm (2BC). Anxiety was measured in prolonged-abstinence (2 wks) from 2BC with NSFT. Next at 4wk abstinence mice were exposed to repeated restraint stress paired with CNO (3 mg/kg, IP). 3 days post restraint exposure anxiety was measured with NSFT.

Results: Repeated restraint stress increased latency to feed as measured with NSFT in alcohol-naïve males and females (t-test, p = 0.023). Total immobility during restraint varied across days (1-way RM ANOVA; day p = 0.043) but not sex. In 2BC prolonged withdrawal PBN(PACAP) inhibition induced a trend for decreased latency to feed in NSFT with no changes on RRS behavior. A history of RRS paired with PBN(PACAP) inhibition blunted RSS-induced increase in NSFT latency (2-way ANOVA; stress p = 0.004, genotype p = 0.0481).

Conclusions: These data demonstrate PBN(PACAP) neurons modulate anxiety-like behavior in alcohol-abstinence. Given the prominent role of the BNST in abstinence-induced anxiety current studies investigate changes in PBN(PACAP)♢BNST circuit activity. Given pharmaceutical treatments for migraines targeting PACAP inhibition are bioavailable, these studies inform a potential role for treatments in alcohol-abstinence.

Keywords: Anxiety and stress, Alcohol Abstinence, BNST, parabrachial nucleus, PACAP

Disclosure: Nothing to disclose.

Gorana Puzovic, Matthew Wolan, Gregory Collins*

University of Texas Health Science Center at San Antonio, San Antonio, Texas, United States

Background: Now in its “fourth wave”, the overdose epidemic is currently defined by rapid increases in deaths from multi-drug overdoses, including most recently fentanyl in combination with xylazine (a veterinary anesthetic). Though little is known about interactions between the toxic effects of fentanyl and xylazine, available evidence suggests that overdoses involving fentanyl+xylazine require larger and/or more frequent doses of naloxone. The goals of these studies are to define the nature of the drug-drug interactions between the cardiorespiratory effects of fentanyl and xylazine, and compare the effectiveness of naloxone to reverse the cardiorespiratory effects of fentanyl administered alone and in combination with xylazine.

Methods: Male and female Sprague-Dawley rats (n = 5m/5f per condition) were surgically prepared with an indwelling venous catheter and habituated to a collar-based pulse oximetry system capable of recording heart rate (HR) and blood oxygenation (SpO2). Dose-response curves (DRCs) for fentanyl (0.0056-0.56 mg/kg) and xylazine (0.1-10 mg/kg) were generated prior to evaluating the effectiveness of 5-min post-treatments with naloxone (1 mg/kg) and atipamezole (0.1-1 mg/kg) to reverse the effects of 0.56 mg/kg fentanyl or 3.2 mg/kg xylazine. To characterize interactions between fentanyl and xylazine, DRCs were generated for mixtures of fentanyl+xylazine, administered at 3:1, 1:1, and 1:3 ratios relative to their ED85% for HR. Dose-addition analyses were used to define the nature of the interactions (e.g., additivity, supra-additivity) for both HR and SpO2. Finally, naloxone (1 mg/kg), atipamezole (0.32 mg/kg), and a mixture of naloxone and atipamezole were evaluated for their effectiveness to reverse the cardiorespiratory effects of the largest dose-pair of the mixtures (3:1, 1:1, and 1:3) of fentanyl+xylazine.

Results: Fentanyl produced dose-related decreases in HR and SpO2 that were reversed by naloxone, but not atipamezole. Xylazine produced dose-related decreases in HR that were reversed by atipamezole, but not naloxone. Xylazine had minimal effects on SpO2. When evaluated as mixtures, fentanyl+xylazine produced dose-related decrease in both HR and SpO2, with dose-addition analyses indicating the interactions for HR and SpO2 were additive to sub-additive. When administered as post-treatments, naloxone and atipamezole each partially reversed the effects of the mixtures, however, the combination of naloxone+atipamezole resulted in a more complete reversal of the cardiorespiratory effects of fentanyl+xylazine. There were no differences between male and female rats.

Conclusions: These studies provide clear evidence that 1) the cardiorespiratory effects of fentanyl and xylazine are mediated by mu-opioid and alpha-2 adrenergic receptors, respectively; 2) mixtures of fentanyl+xylazine exhibit additive to sub-additive interactions with regard to their cardiorespiratory effects; and 3) naloxone is only able to partially reverse the cardiorespiratory effects of mixtures of fentanyl+xylazine. Together, these studies provide important new information about interactions between the overdose-related effects of fentanyl and xylazine and highlight the need to develop more effective treatments to rescue individuals who have overdosed on fentanyl+xylazine.

Keywords: Fentanyl, Xylazine, opioid overdose

Disclosure: Nothing to disclose.

Michael Johson, Thomas Scanlan, Andrey Ryabinin, Deena Walker*

Oregon Health and Science University, Portland, Oregon, United States

Background: In addition to their role in metabolism, thyroid hormones are necessary for neuronal development and proper monoaminergic system function. Dysregulation of thyroid hormones has been associated reward circuitry dysfunction and reward-related mood disorders including depression. With regard to substance use disorder, clinical data suggests a bidirectional relationship with thyroid hormone, in which drugs dysregulate thyroid hormones and dysregulation of thyroid hormones influence the rewarding properties of addictive drugs. This relationship is particularly true for alcohol. Dysregulation of thyroid hormone is associated with withdrawal from alcohol, craving, and severity of alcohol use disorder.

Methods: In order to determine how thyroid hormone might influence alcohol drinking, we measured mRNA expression of thyroid hormone receptors (Thra and Thrb) and two neuropeptides known to regulate alcohol drinking, oxytocin (Oxt) and vasopressin (Avp), in the paraventricular nucleus (PVN) of the hypothalamus and the medial amygdala (MeA) after 3 weeks of ethanol intake in C57BL/6J mice via a 2-bottle choice paradigm (EtOH = 12, 6M/6F; H2O = 12, 6M/6F). Next, to determine if thyroid hormone receptor beta can regulate alcohol drinking, we treated alcohol drinking male and female mice with the thyroid hormone receptor beta agonist, Sobetirome (Sob-AM2). This CNS-specific agonist is only converted into its active form when transported into the brain and is currently in clinical trials for the treatment of depression. We predicted that that its administration may decrease alcohol drinking. C57BL6J male and female mice that had established drinking in a 2-bottle choice ethanol drinking paradigm (EtOH = 24, 12M/12F; H2O = 24, 12M/12F) were injected with increasing doses of Sob-AM2 (6 subcutaneous injections with 0.3mg/kg followed by 4 injections of 1mg/kg, or Vehicle; 10% DMSO in Saline, once per day). Ethanol and water intake was monitored daily.

Results: In the RT-qPCR experiment, we found that while alcohol drinking had no effects on vasopressin expression in the PVN, it decreased oxytocin (F = 6.449; p < 0.05) and thyroid hormone receptor beta (Thrb) mRNA (F = 6.497; p < 0.05) expression in the PVN of males when compared to water drinking controls. Oxt mRNA was negatively correlated with alcohol intake (r = -0.83; p < 0.05) and positively correlated with Thrb mRNA (r = 0.85; p < 0.05). On the other hand, Oxt was increased in the MeA of females (F = 2.157; p < 0.01) but not males and was associated with alcohol intake (r = 0.91; p < 0.01). Importantly, alcohol drinking was not associated with changes in thyroid hormone receptors in the MeA of males or females suggesting a sex- and region-specific role for thyroid hormone in mediating alcohol drinking, possibly through regulation of oxytocin expression in males. In the Sob-AM2 experiment, we found that repeated treatment with Sob-AM2 significantly reduced alcohol intake and alcohol preference in male mice, particularly at the higher dose (1 mg/kg, p < 0.0001). In females, treatment with Sob-AM2 had more subtle effects on drinking but we did observe a decrease for the preference for alcohol in females with the higher dose (p < 0.05).

Conclusions: Together, these data suggest that alcohol-induced reductions in thyroid hormone receptor beta may be critical to influencing alcohol intake. Because oxytocin has been shown to have some translational value as a pharmacotherapeutic for alcohol use disorder, future work is focused on better understanding the sex- and region-specific effects of thyroid hormone receptor beta agonism on oxytocin expression in the PVN and MeA.

Keywords: Alcohol Use Disorder - Treatment, oxytocin and addiction, thyroid hormone

Disclosure: Nothing to disclose.

Ravinder Naik Dharavath*, Srija Das, Jonathan Hua, Cassandra Marceau-Linhares, Michael Marcotte, Paul Fletcher, Anh Le, Etienne Sibille, Thomas Prevot

CAMH Toronto, Toronto, Canada

Background: Impulsivity is a hallmark of substance use disorders (SUDs), indicating reduced control over addictive behaviors. Developing and testing new pharmacotherapies to reduce impulsivity requires the use of animal models. The 5-choice serial reaction time task (5-CSRTT) is widely used to measure impulsivity in rodents. Lengthening the inter-trial interval (ITI) in the 5-CSRTT induces reversible impulsivity without affecting task accuracy, making it a robust method for screening drugs for anti-impulsivity effects. In SUDs, multiple systems such as the adrenergic, opioid, and inflammatory systems are affected, making them potential targets for pharmacological interventions.

Methods: All animals were trained to respond to a light stimulus in one of five apertures with a short (5s) ITI and to perform the task with > 80% accuracy and < 20% omissions. Aim 1: Twenty-four male Wistar rats were administered Atomoxetine (anti-impulsive) and nicotine (pro-impulsive) to confirm the 5-choice task’s sensitivity. Aim 2: Thirty-six male Long Evans rats were used to screen three pharmacological agents with various modes of action (MCC950, Naltrexone, Guanfacine) for anti-impulsivity efficacy in a lengthened ITI (10s – LF-ITI) task.

Results: Nicotine increased premature responses at baseline compared to vehicle (p < 0.05), while atomoxetine reduced premature responses in LF-ITI (p < 0.05), confirming that the 5-CSRTT is sensitive to capturing increases and decreases in impulsivity. MCC950 and Naltrexone had no significant effects on impulsivity. Guanfacine significantly reduced LF-ITI-induced impulsivity at 0.2 and 0.6 mg/kg (p < 0.05).

Conclusions: The 5-choice assay is a sensitive tool for studying the pharmacological impact on impulsivity, with promising results for adrenergic drugs, to be confirmed in models of SUDs.

Keywords: Addictive behavior, Disinhibition, Impulsivity, Premature response

Disclosure: Nothing to disclose.

Malia Belnap*, Dylan Kirsch, Erica Grodin, Lara Ray

University of California, Los Angeles, Los Angeles, California, United States

Background: A critical aspect of alcohol use disorder (AUD) research involves identifying and understanding the risk factors that contribute to its development. Factors implicated in the etiology of AUD are subjective responses (SR) to alcohol, which encompass the subjective feelings and experiences elicited by alcohol consumption. Notably, SR to alcohol can vary widely between individuals, and individual differences in acute SR to alcohol impact the likelihood of experiencing subsequent alcohol-related problems, including the development of AUD. Understanding how these individual differences in SR are associated with drinking-related outcomes can provide valuable insights into the mechanisms underlying heightened risk for AUD. The current study examines the relationship between SR measures and the motivation to self-administer alcohol across multiple timepoints during a progressive ratio intravenous alcohol self-administration task. Specifically, we assessed how alcohol-induced stimulation, sedation, wanting, liking, positive mood, and negative mood influenced motivation to self-administer alcohol throughout the duration of the paradigm. Exploratory analyses tested the moderating effects of sex and family history of alcohol-related problems on self-administration behaviors.

Methods: Participants (n = 67; 36M/31F) who drank heavily (14+ drinks per week for males, 7+ for females) completed a progressive ratio alcohol intravenous self-administration paradigm that lasted 120 minutes. Every 15 minutes, participants were breathalyzed and completed self-report questionnaires to measure stimulation, sedation, alcohol wanting, alcohol liking, negative mood, and positive mood. To determine whether alcohol had been administered, we evaluated whether breath alcohol content (BrAC) increased from one timepoint to the subsequent timepoint. Self-administration outcomes were then categorized as either a value of ‘1’, indicating self-administration, or a value of ‘0’ indicating no self-administration had occurred. Time-lagged subjective response outcomes were examined as predictors of subsequent motivation to self-administer alcohol. A multilevel model was conducted using the ‘GLIMMIX’ procedure in SAS 9.4 to analyze the data, with the binary dependent variable of self-administration. The fixed effects included SR measures and time, as well as their interactions, to examine their influence on self-administration behavior. A random intercept was included for subject to account for repeated measures within the same subject. Covariates tested in each model included age, sex, and AUD severity. Sex and family history of alcohol-related problems were further investigated as potential moderators of the impact of SR measures on self-administration.

Results: Higher levels of stimulation, liking, and wanting were significantly associated with an increase in the likelihood of self-administration (p’s > 0.03). There were no significant main effects of sedation, positive mood, and negative mood on self-administration (p’s > 0.12). An increase in time was significantly associated with a decrease in the likelihood of self-administration in all models (p’s < 0.0001). There was a significant interaction between liking and time such that as task burden increased, the impact of liking on self-administration decreased (p = 0.04). Sex did not moderate the impact of SR measures on self-administration in any model (p’s > 0.17). There was a significant interaction between family history of alcohol-related problems and wanting such that higher wanting scores were associated with an increase in likelihood of self-administration in individuals without a family history (p = 0.03), but not in those with a family history. Additionally, there was a three-way interaction between negative mood, time, and family history of alcohol-related problems such that for individuals with no family history of alcohol-related problems, an increase in negative mood at earlier timepoints, but not later timepoints, was associated with an increased likelihood of self-administration compared to individuals with a family history (p = 0.04).

Conclusions: The human-laboratory alcohol intravenous progressive ratio self-administration paradigm provides a valuable framework for translating insights from preclinical operant-response studies to understand the motivational mechanisms driving alcohol consumption. By incorporating multiple observations per subject throughout course of the task, our analyses examined the interaction between SR and increasing task burden. Our findings indicate that stimulation, liking, and wanting are significantly associated with increased moment-to-moment self-administration under a progressive ratio schedule, and thus may be mechanisms underlying heightened risk for AUD. Our findings underscore the importance of family history of alcohol-related problems as a critical factor when investigating the relationships between SR and motivation to consume alcohol.

Keywords: alcohol self-administration, Subjective Response, Alcohol-seeking behavior, motivation

Disclosure: Nothing to disclose.

Lauren Beugelsdyk, Ellie Decker Ramirez, Jesse Schank*

University of Georgia, Athens, Georgia, United States

Background: Aversion-resistant alcohol consumption is often assessed using rodent models in which alcohol is paired with a negative stimulus such as adulteration with a bitter tastant or contingent footshock. These stimuli typically reduce alcohol consumption, and continuation of alcohol intake under these conditions is thought to model consumption of alcohol despite negative consequences in individuals with alcohol use disorder (AUD). However, people who suffer from AUD often experience negative consequences that are separated in time from their active alcohol drinking, and often impact their social life and interpersonal relationships.

Methods: In our model, we trained male and female mice to consume alcohol, and then exposed them to repeated social isolations following alcohol drinking sessions. We used the CD1 strain, as they have been found to be more responsive to social reward than C57BL6/J mice. However, CD1 mice do not drink as much alcohol as C57s, so we added 0.2% (w/v) saccharin to alcohol solution and provided this in a two bottle choice procedure with 0.2% saccharin without alcohol in the other bottle, to stimulate higher levels of alcohol consumption in the CD1 strain.

Results: We found that social isolation after drinking lead to decreased consumption relative to mice that were not isolated. This effect was observed in both males (p = 0.03) and females (p = 0.04). There was considerable individual variability in this response with some mice suppressing more strongly than others.

Conclusions: Current studies are examining if this variable suppression of consumption correlates with Fos activation in critical brain regions. We will next determine if interventions that induce aversion-resistance in other models, such as exposure to intermittent alcohol access or chronic stress, reduce sensitivity to social isolation-induced suppression of alcohol intake.

Keywords: Alcohol, aversion, social isolation

Disclosure: Nothing to disclose.

Catherine Moore*, Bryan Jenkins, Christopher Cunningham, Elise Weerts

Johns Hopkins University School of Medicine, Baltimore, Maryland, United States

Background: Abrupt discontinuation of chronic opioid use results in a severe, multidimensional withdrawal syndrome that includes symptoms such as aches/pain, tremors, gastrointestinal distress, restlessness, increased pain sensitivity (hyperalgesia), and negative emotional symptoms such as anxiety and irritability. Identifying non-opioid treatment strategies for opioid withdrawal management is of high public health significance. There is compelling preclinical evidence that medications targeting the endocannabinoid system can reduce the severity and duration of opioid withdrawal symptoms. In particular, indirect agonism of the cannabinoid receptor-1 (CB1R) via increasing circulating endocannabinoid (eCB) levels may have therapeutic benefit while minimizing cannabimimetic side effects of direct agonists, such as delta-9-THC. Recently, we determined sterol carrier protein-2 (SCP-2) acts as a binding and transport protein for eCBs and we have synthesized the first small molecule inhibitor of SCP-2 (SCPI-1) to block eCB transport and thereby increase eCB tone and CB1R activity. This project evaluated the effects of SCPI-1 using a multidimensional model of opioid withdrawal symptoms in rats.

Methods: Adult male and female Sprague-Dawley rats were administered ascending doses of morphine sulfate (10-50 mg/kg, subcutaneously) or saline twice daily for 10 days. Following abrupt discontinuation of morphine or saline, we assessed opioid withdrawal symptoms that included: 1) somatic signs indicative of withdrawal (e.g., ptosis, wet-dog shakes, grooms, hypophagia and weight loss); 2) negative emotional symptoms of anxiety- and irritability- like behavior in the elevated plus maze and bottle brush test, respectively; and 3) hyperalgesia using the von Frey test of mechanical pain sensitivity. Rats were administered SCPI-1 (10, 30 mg/kg, intraperitoneal injection) or vehicle 38-hrs post morphine discontinuation (peak withdrawal) under blinded conditions and outcomes were assessed.

Results: Following morphine discontinuation, somatic signs of withdrawal (e.g., diarrhea, vocalizing, wet-dog shakes, etc.) were evident 14h following morphine discontinuation, peaked at 38h, and lasted up to 5 days (p’s < 0.05 vs. nondependent controls). Anxiety-like behavior was observed for up to 3 days after morphine discontinuation (p’s < 0.05 vs. nondependent controls). Withdrawal-induced hyperalgesia peaked days 3-5 after morphine discontinuation (p’s < 0.05 vs. baseline). When administered at the time of peak somatic withdrawal, treatment with tSCPI-1 (10 mg/kg) alleviated somatic signs of withdrawal compared with vehicle treatment (p < 0.05). No effects of SCPI-1 were observed on anxiety- or irritability-like behavior nor on hyperalgesia.

Conclusions: Twice daily injections of morphine produced a robust spontaneous withdrawal syndrome in male and female rats, with certain symptoms persisting up to 8 days following morphine discontinuation. Treatment with the novel endocannabinoid transport inhibitor SCPI-1 alleviated somatic symptoms of withdrawal when administered in peak withdrawal. Research is ongoing to determine the underlying mechanisms of SCPI-1, including assessments of effects on brain eCB levels. Overall, these data support the endocannabinoid system as a promising therapeutic target for opioid withdrawal.

Keywords: opioid withdrawal, endocannabinoid, pharmacotherapy

Disclosure: Nothing to disclose.

Carolina Haass-Koffler*, Joshua Brown, Robert Swift

Brown University, Providence, Rhode Island, United States

Background: Intranasal (IN) administration of the neuropeptide oxytocin has been explored as a potential therapeutic agent for opioid use disorder (OUD). Oxytocin was effective in reducing stress and withdrawal symptoms in individuals with OUD.

Methods: This phase 1 cross-over, randomized, double-blind, placebo-controlled trial tested the safety, tolerability, and efficacy of IN oxytocin (80 IU) twice a day for 7 days in participants (n = 20) with OUD who were taking an opioid agonist therapy (OAT). Participants underwent an opioid cue-reactivity laboratory procedure, paired with noradrenergic activation produced by administration of yohimbine or placebo (also crossover, randomized, double-blind). Psychometric assessments included: 1) subjective response: craving, withdrawal, anxiety, and stress; 2) biomedical markers: hypothalamic-pituitary-adrenal (HPA) axis response (cortisol) and noradrenergic activation (α-amylase); and 3) safety measures: hemodynamics and adverse event evaluation. Generalized Linear Model with model-based estimator in the covariance matrix was used, with medication (oxytocin/placebo) and noradrenergic activation (yohimbine/placebo) as within-subject factors.

Results: Oxytocin significantly reduced opioid-like withdrawal, anxiety symptoms, and cortisol levels elicited by cue reactivity under noradrenergic activation produced by yohimbine. This effect was specific as oxytocin did not reduce craving, hemodynamics, or α-amylase levels increased by yohimbine administration. Overall, a single oral dose of yohimbine elicited the noradrenergic stimulation and 7-day oxytocin administration was safe and well-tolerated among individuals diagnosed with OUD and taking OAT.

Conclusions: The present study contributes to the existing knowledge regarding the efficacy of oxytocin in reducing opioid withdrawal and anxiety among individuals with OUD who are undergoing OAT. The findings suggest that oxytocin alleviates opioid-like withdrawal symptoms, including noradrenergic activation and anxiety by modulating the HPA axis.

Keywords: Opioid peptides, Noradrenergic System, oxytocin

Disclosure: Nothing to disclose.

Shuyi Qi, Jobe Ritchie, David Soto, Avery Pruitt, Dylan Reeves, Abigail Greenway, Rita Fuchs*

Washington State University, Pullman, Washington, United States

Background: Intractable craving upon exposure to drug-predictive environmental stimuli is a major impediment in the treatment of substance use disorders. Understanding the neural mechanisms that regulate the strength of the underlying contextual drug memories has the potential to facilitate the development of anti-relapse treatments. We have previously shown that the cornu ammonis (CA) 3 region of the dorsal hippocampus critically regulates the maintenance of contextual cocaine memories and subsequent context-induced relapse to cocaine-seeking behavior. The dorsal CA3 sends dense monosynaptic glutamatergic projections to the dorsal CA1 region. Therefore, we investigated the contribution of this CA3 to CA1 neural circuit to the regulation of cocaine memory strength using an optogenetic approach. We hypothesized that inhibition of this circuit at a time when cocaine memories are retrieved and become transiently susceptible to manipulation would diminish subsequent context-induced relapse of drug-seeking behavior.

Methods: Sprague-Dawley males and females (n = 5-7/sex/group) received jugular catheter implants and bilateral intra-CA3 microinfusions of an adeno-associated virus that either expressed the inhibitory opsin eNpHR3.0 and a fluorophore (AAV5-hSyn-eNpHR3.0-eYFP-WPRE-PA) or the fluorophore alone (AAV5-hSyn-eYFP). Optic fibers were aimed at CA3 terminals within the CA1. After surgery, rats were trained to self-administer intravenous cocaine infusions in a distinct context that consisted of salient auditory, visual, olfactory, and tactile stimuli over at least 10 days to establish context-response-cocaine associative memories. Lever presses resulted in cocaine reinforcement under a fixed ratio schedule. Next, the rats received extinction training over seven days in a different context to establish control memories. On the following day, the rats were re-exposed to the cocaine-predictive context for 15 minutes in order to elicit memory retrieval and destabilization. Immediately after or six hours after the memory retrieval session, all rats were transported to a third, familiar context where their optic fibers were connected to a solid-state laser. Rats received laser light treatment (10 mW, 5 s on, 5 s off) intermittently for one hour or remained in the context without laser treatment. Treatment group assignment was balanced based on behavioral history. Starting on the next day, the rats received at least two additional daily extinction sessions followed by a test session in the cocaine-predictive context. Lever presses were not reinforced during these sessions. Lever responding upon the first re-exposure to the control and cocaine-predictive contexts after treatment were used to index control and cocaine memory strength, respectively. Rats were perfused after the test session, and virus expression and optic fiber placement were evaluated. Data from rats with insufficient virus expression or misplaced optic fibers were excluded from data analysis. All behavioral data were analyzed using analyses of variance with sex, treatment, context, and time as factors, when appropriate. Significant interaction effects were further investigated using Bonferroni’s post-hoc tests. Alpha was set at 0.05 for all comparisons.

Results: Optogenetic inhibition of CA3 terminals in the CA1 immediately after the 15-minute memory retrieval session, when memories were expected to be labile, attenuated subsequent lever responding at test in the cocaine-predictive context, but not the control context, relative to no inhibition (ANOVA context x treatment interaction, p < 0.05). In the control experiments, exposure to the cocaine-predictive context resulted in more lever responding than exposure to the control context (all ANOVA context main effects, p < 0.05). However, optogenetic circuit inhibition six hours after the memory retrieval session, when memories were expected to be stable in long-term memory, did not alter subsequent lever responding in either context, compared to no inhibition (ANOVA treatment main and interaction effects, n.s.). Similarly, laser treatment did not alter lever responding in the absence of opsin expression, compared to no laser treatment (ANOVA treatment main and interaction effects, n.s.). All observed effects were independent of sex (all ANOVA sex main and interaction effects, n.s.).

Conclusions: Optogenetic inhibition of the dorsal CA3 to CA1 hippocampal circuit reduced the motivational effects of the cocaine-predictive context without altering responding in the control context. These effects were dependent on memory retrieval and destabilization, opsin expression, and laser activation, suggesting a weakening of cocaine memory strength and/or interference with memory reconsolidation. Together, these findings demonstrate that signaling in the CA3 to CA1 neural circuit is required for the maintenance of contextual cocaine memories that promote relapse to drug-seeking behavior. These findings support the idea that approaches selectively targeting maladaptive memories may be effective for the treatment of substance use disorder.

Keywords: cocaine, memory, optogenetics, hippocampus

Disclosure: Nothing to disclose.

Tanya Saraiya*, Sonali Singal, Priya Johal, Tamina Daruvala, Denise Hien, Sudie Back

Medical University of South Carolina, Charleston, South Carolina, United States

Background: Opioid use disorder (OUD) and posttraumatic stress disorder (PTSD) are chronic and debilitating psychiatric conditions that frequently co-occur. Among individuals seeking treatment for OUD, 30%-54% suffer from co-occurring PTSD. In comparison to individuals with OUD only, those with OUD and PTSD present with a more severe clinical profile, including more mental and physical health problems (e.g., depression, chronic pain), worse treatment outcomes, and higher rates of opioid-related overdose. Untreated PTSD is a risk factor for opioid craving, use, and overdose, as well as the development of other psychiatric problems (e.g., suicidal ideation). Further, compared to other substance use disorders and PTSD, individuals with OUD + PTSD show greater trauma exposure, higher PTSD severity, and a higher risk of opioid-related overdose. Effective treatments for co-occurring OUD + PTSD are lacking, and there is little empirical evidence available to guide the provision of clinical care. Concurrent Treatment of PTSD and Substance Use Disorders Using Prolonged Exposure (COPE) is one of the most efficacious behavioral treatments for substance use disorders and PTSD. However, it has not been tested on individuals with solely OUD + PTSD nor been augmented with medications for OUD.

Methods: We adapted COPE to Helping Opioid Use with Exposure (HOPE) following mixed methods feedback from formative stakeholders. In this open-label Stage Ia/Ib pilot trial, HOPE was piloted among N = 6 participants with OUD + PTSD. Eligible participants had a PTSD diagnosis from the Clinician Administered PTSD Scale (CAPS-5), had a diagnosis of OUD within the past five years on the Structured Clinical Interview for DSM-5, were stabilized on any medication for OUD in the last month, and has not received any prior evidence-based trauma therapy. Participants were excluded if they met bipolar II criteria, showed severe psychosis, or reported another non-opioid substance as their primary substance of concern. Eligible participants completed a baseline appointment, 10-12 60-minute weekly therapy sessions with a licensed clinical psychologist, and a one-month follow up. A total of n = 5 participants were treatment completers, defined as completing the intervention from baseline to end of treatment. At every appointment, participants completed a battery of self-report surveys including PTSD symptoms (PTSD Checklist for DSM-5; PCL-5) and a 12-panel urine drug screen. Data collection is ongoing and anticipated to be completed by September 2024.

Results: Repeated measures general linear models tested within group differences in self-reported PTSD symptoms and percent positive urine drug screens for all illicit substances from baseline to end of treatment. Greenhouse Geisser effects for non-sphericity showed significant within group reductions in PTSD scores from baseline to end of treatment, F(2.21, 8.86) = 7.63, p = 0.011, ηp2 = 0.66, 1-β = 0.86, with a moderate effect size. There was no change in the percent of positive urine drug screens from baseline to end of treatment, F(8,24) = 1.00, p = 0.46, ηp2 = 0.25, 1-β = 0.36. However, all participants were abstinent from opioids at baseline and at end of treatment. In addition, all participants stayed on medications for OUD throughout treatment.

Conclusions: HOPE is the first integrated behavioral treatment designed for OUD + PTSD. Preliminary findings suggest HOPE decreases PTSD symptoms and does not interfere with maintenance on medications for OUD. However, additional modifications are needed for HOPE to enhance reductions in polysubstance use. The HOPE intervention will be further tested in a larger randomized clinical trial to be launched in January 2025.

Keywords: Opioid addiction, PTSD, Psychotherapy

Disclosure: Nothing to disclose.

Heather Webber*, Joy Schmitz, Michael Weaver, Scott Lane

The University of Texas Health Science Center At Houston, Houston, Texas, United States

Background: Methamphetamine (MA) use disorder is an increasing public health concern and has no FDA-approved medications for treatment. Preclinical evidence suggests dual orexin receptor antagonists such as suvorexant may target key mechanisms in the recovery from MA use disorder by regulating sleep architecture, buffering hyper-arousal and stress responses, and modulating reward and drug-related behaviors. The current study aimed to assess 1) the feasibility of suvorexant administration and 2) preliminary effects on sleep, stress, and craving in a sample of individuals with MA use disorder.

Methods: We employed a within-subjects, case-series design. Non-treatment-seeking participants with MA use disorder completed weekly self-reported measures of sleep (insomnia severity index; ISI), stress (10-point visual analogue scale; VAS), craving (brief substance craving scale; BSCS), and side effects after 7 nights of receiving suvorexant, washout, and placebo. Two cases had suvorexant first and the other case had placebo first. Participants also wore a Fitbit Charge 5 to objectively measure sleep each night. The data were inspected visually to assess change in outcomes by condition, consistent with case-series designs.

Results: Three participants (mean age = 43; n = 2 females; n = 1 Hispanic, n = 3 White) attended all three in-person clinic visits. Cases A and C took all 7 doses of suvorexant; Case B took 4 doses. Case C reported severe drowsiness, otherwise, no severe side effects were found to be associated with the medication. Preliminary results point to more minutes slept on average per sleep session during the suvorexant week (Case A = 456; Case B = 258, Case C = 264) compared to the washout (Case A = 432; Case B = 228, Case C = 132), and placebo week (Case A = 377; Case B = 328, Case C = 204) as measured via the Fitbit. Results for subjective sleep quality on the ISI were more mixed, with Case A reporting worse subjective sleep quality, Case B reporting improved sleep quality; Case C reported similar sleep quality during the suvorexant week. In terms of stress, Case A reported no stress all weeks, but Case B and C reported lower stress during the suvorexant week compared to the placebo week (VAS 1 vs. 10; 6 vs. 7). Craving was slightly reduced in Case C for the suvorexant week, but was similar across weeks for Cases A and B on the BSCS.

Conclusions: Suvorexant was found to be safe and tolerable in individuals with MA use disorder. No unexpected side effects or adverse events occurred. Preliminary results regarding effects on objective measures of sleep were promising, with less consistent results for subjective sleep. There was also initial evidence of a small effect on stress and craving. Future research would benefit from assessing the effects of suvorexant on MA-related relapse factors in a randomized controlled trial in a diverse sample of treatment-seeking individuals.

Keywords: Methamphetamine use disorder, Dual Orexin Receptor Antagonists, Sleep disturbances, craving, stress

Disclosure: Nothing to disclose.

Patrick Mulholland*, Rachel Penrod, Kathy Lindquist

Medical University of South Carolina, Charleston, South Carolina, United States

Background: Alcohol use disorder (AUD) is a costly and devastating brain disease that affects regions that control reward seeking and decision-making. Individuals with AUD often display cognitive decline, poor decision-making, and risky behaviors. Excessive alcohol drinking alters multiple forms of neural plasticity, including adaptations in the plasticity of intrinsic excitability, which are thought to contribute to cognitive and behavioral impairments and persistent alcohol-seeking behaviors. Although there have been increased efforts to determine how alcohol alters neural membrane excitability, there is still a substantial gap in our understanding of the mechanisms that are responsible for maladaptive plasticity induced by chronic alcohol. Despite a prominent role in controlling membrane excitability, nothing is known about the contribution of Kv11.1 (Kcnh2) channels to alcohol-associated plasticity and alcohol-related behaviors.

Methods: To explore Kchn2-alcohol interactions and Kcnh2 expression differences, we performed bioinformatics analyses using publicly available genetic databases: GeneWeaver, GeneNetwork, and Mouse Genome Database. To determine if alcohol dependence alters Kcnh2 expression in the nucleus accumbens (NAc), samples were extracted from mice exposed to chronic intermittent ethanol (CIE), and we performed RT-PCR to measure Kcnh2a and Kcnh2b transcript expression. We then used electrophysiological and pharmacological approaches to manipulate Kv11.1 channels to determine their functional role in controlling NAc medium spiny neuron (MSN) cellular excitability and alcohol drinking behavior.

Results: Our search revealed that Kcnh2 was included in gene sets for alcohol-induced differential expression (6 gene sets) and QTLs for alcohol preference (4 gene sets) and tolerance (1 gene set) in mice and rats and AUD susceptibility in humans. In BXD recombinant inbred strains of mice, we found 42 data sets with significant correlations between NAc Kcnh2 expression and voluntary alcohol drinking/preference, all which were negative. After correcting for multiple comparisons using a 1% FDR, we observed significantly lower expression of Kcnh2 in the NAc core and shell in the C57BL/6J strain compared with the DBA/2J strain (P < 0.05), but not other regions implicated in alcohol-related behaviors. A search identified a compelling candidate SNP in Kcnh2 (rs29526580, missense variant) that differed between C57BL/6J and DBA/2J strains, as well as other strains with opposing drinking phenotypes. Because the SNP could explain drinking that involves deviation from the C57BL/6J strain, we used a reverse genetics approach to evaluate drinking phenotypes directly linked to the Kcnh2 interval by comparing alcohol intake in BXD RI strains that inherited the B or D allele. Male and female BXD RI strains that inherited the B allele at Kcnh2 consumed significantly more alcohol than strains that inherited the D allele (2-way ANOVA; main effect of genotype: F(1,42) = 16.50, p < 0.01; n = 23 strains). In alcohol-naïve mice, Kcnh2a and Kcnh2b expression was significantly higher in females compared with males (2-way ANOVA, sex: F(1,14) = 5.936, p < 0.05; n = 4-5 mice/sex). Consistent with our hypothesis, preliminary analysis revealed that average expression of Kcnh2a and Kcnh2b in the NAc of CIE exposed mice was lower than air controls. Pharmacological manipulation of Kv11 channels controlled the membrane excitability and action potential firing of NAc MSNs and reduced alcohol drinking in male and female C57BL/6J mice (2-way RM ANOVA; main effect of dose: F(2.249,40.49) = 3.21, p < 0.05; n - 10 mice/sex) without affecting locomotor activity or sucrose intake (p > 0.05).

Conclusions: In summary, our preliminary data demonstrate a compelling link between NAc Kcnh2 expression and alcohol-related behaviors, including alcohol drinking. To complement the genetics analyses, we showed that a pharmacological activator of KV11.1 channels reduced alcohol drinking in male and female C57BL/6J mice and that blockade of KV11.1 channels regulated membrane excitability of NAcc MSNs. Together, our results suggest that Kcnh2 could be a candidate gene that explains risk factors and individual vulnerability to alcohol misuse.

Keywords: Alcohol consumption, Nucelus accumbens, Kcnh2, synaptic plasticity, epigenetic, gene expression, action potentials

Disclosure: Nothing to disclose.

Laili Boozary*, Bidhan Lamichhane, Ashton R. Baltazar, Taylor A. Love, Sarah S. Tonkin, Adam M. Leventhal, Kyle Simmons, F. Joseph McClernon, Jason Oliver

University of Oklahoma Health Sciences Center, Oklahoma City, Oklahoma, United States

Background: Traditional perspectives on cigarette smoking emphasize the importance of smoking-related rewards (i.e., withdrawal relief, hedonic effects of nicotine) in driving use and addiction. However, recent theoretical models highlight the potential roles of non-smoking rewards (i.e. sources of pleasure or joy unrelated to smoking) in addictive behavior. For instance, smokers with fewer sources of reward outside of smoking appear more likely to relapse following a quit attempt. Further complicating these matters are the known interactions between nicotine and reward systems, including evidence that smoking abstinence may suppress neural responses to non-smoking rewards. Unfortunately, relatively few studies have examined smoking and non-smoking rewards within the same paradigm. To our knowledge, no study has examined personalized rewards, which may have distinct effects from the standard rewards common to most laboratory paradigms. The aim of this project was to develop and test a novel fMRI paradigm for examining neural responses to personal smoking and non-smoking rewards as well as examine the preliminary effects of smoking abstinence on responses to each reward type.

Methods: Daily cigarette smokers (≥ 5 cigarettes/day) not actively engaged in a cessation attempt were recruited. During a screening visit, participants completed a variety of self-report measures, including assessments of sociodemographic characteristics, smoking history, and nicotine dependence. A detailed semi-structured interview was used to gather personal narratives of each individual’s common smoking rewards, non-smoking rewards, and neutral scenarios. These were then rewritten into brief 15-second scripts and recorded for presentation during the fMRI task. Participants underwent two MRI scans, with one following 24-hours of biochemically-verified smoking abstinence and the other ad libitum smoking (session order randomly assigned/counterbalanced). Trials were presented in two pseudo-random orders (task version). During scans, participants were instructed to imagine each scenario as it was read to them. Imaging data analysis was performed using fMRIprep (version 23.0.1) and Analysis of Functional NeuroImages (AFNI, version 23.3.05). Primary analyses targeted 13 regions of interest (ROIs) based on their relation with addiction and reward in prior studies. These included: (1) Two regions in the dorsal anterior cingulate cortex (dACC) and (2) one region in the rostral ACC (rACC); (3) left and right ventromedial prefrontal cortex (vmPFC); (4) left and right ventral striatum (vSTR); (5) left and right dorsal striatum (dSTR); (6) left and right dorsal anterior insula (dAI); and (7) left and right ventral anterior insula (vAI). Linear mixed level models were employed to analyze ROI activation as a function of script condition (smoking reward, non-smoking reward, or neutral) and session type (smoking ad libitum vs. 24-hour abstinent), for task version and session order, and adjusting for age, nicotine dependence, sex, race, and annual income. A second model step included a script condition-by-session type interaction term to examine the effects of abstinence on the activation of each script condition.

Results: Participants (N = 35) were primarily white (74.3%) and female (65.7%) with an average age of 37.1 years. Outcomes revealed a significant increase in activation in response to non-smoking reward (vs. neutral) scripts in the dACC (p’s < 0.01), rACC (p < 0.01), vmPFC (p’s < 0.01), vSTR (p’s < 0.05), left dSTR (p = 0.03), and the left vAI (p < 0.01). There was also significant increase in activation for smoking reward (vs. neutral) scripts in one of the dACC regions (p = 0.01), as well as the rACC (p < 0.01), vmPFC (p’s < 0.05), and the left vSTR (p < 0.01). However, there was a significant decrease in activation for smoking reward (vs. non-smoking reward) scripts in one of the dACC regions (p = 0.01) and the left vmPFC (p = 0.01). Additionally, there was a significant decrease in activation in response to smoking abstinence (vs. smoking ad libitum) in the right vmPFC (p = 0.01), dSTR (p’s < 0.01), and the left and right dAI and left vAI (p’s ≤ 0.01). There were no significant interactions of script condition-by-session type.

Conclusions: Outcomes reveal consistent activation across several ROIs for non-smoking reward cues (compared to neutral) with somewhat less consistent activation for smoking reward cues (compared to neutral). Notably, only two ROIs had different activation for smoking compared to non-smoking reward. Accordingly, this study provides preliminary validation of a paradigm for simultaneously assessing neurobiological responses to personalized non-smoking and smoking rewards. Yet counter to hypotheses, the current paradigm did not find any evidence that abstinence impacted response to non-smoking or smoking reward scripts. Previous literature on this topic utilizing traditional lab-based task paradigms indicate that abstinence induces changes to reward-related activation. Thus, it may be that reward response remains intact when salient, personalized rewards are used. Additionally, outcomes found a consistent decrease in insular activation during 24-hour abstinence. Thus, future research should further explore the impacts of abstinence on reward-related activation in the insula.

Keywords: Neural Reward Circuitry, Functional MRI (fMRI), Addiction, Tobacco

Disclosure: Nothing to disclose.

George Uhl*, Balaji Kannan, Ed Levin, Yun Guan

VA, U Maryland, Baltimore, Maryland, United States

Background: Pentilludin is a novel small molecule candidate to aid treatment of stimulant use disorders, the allodynia of chronic neuropathic pain and (at higher doses) obesity. It is a potent irreversible inhibitor of the actions of the receptor type protein tyrosine phosphatase D (PTPRD). Pentilludin has moved forward through the drug development pipeline based on results of studies of carcinogenicity (Ames, micronucleus), hERG activities, PK/PD, bioavailability po and toxicities n mice, rats and dogs and a favorable preIND meeting with FDA. It reduces self administration of amphetamine and the allodynia of chronic constriction injury (CCI) models of chronic neuropathic pain at 20 mg/kg doses. At higher doses, it reduces mouse weight gain when fed normal or high fat diets. Pentilludin is a promising novel therapeutic for development to aid attempts to quit use of amphetamine and other stimulants. It may also have benefits for the allodynia of chronic neuropathic pain and as an adjuct to aid treatments for obesity.

Methods: In vitro toxicity: Ames testing. Rat micronucleus testing. hERG testing binding assays and voltage clamp electrophysiology.(Panlabs, Labcorp).

In vivo toxicity testing: Hematology, chemistry, histopathology and behavioral observations (Uhl lab, Bioduro)

PK/PD/bioavailability: LC/MS-MS (Bioduro)

Self Administration: Pentilludin (20 mg/kg) administration prior to alternating M-W-F sessions in rats trained for 6 amphetamine sessions prior to pentilludin administration.

Chronic constriction injury: Mice subjected to CCI and tested over time for withdrawal responses to von Frey hair and thermal stimuli with treatment with 20 mg/kg pentilludin.

Obesity: Mice treated with 100 mg/kg pentilludin during access to normal chow or high fat diet, weight monitored

Results: Toxicity: No positive result from studies of carcinogenicity, hERG activity, toxicity to doses up to 100 mg/kg in mice. At doses at or exceeding 150 mg/kg/d, mice lost weight and several died starting on day 4 of two week treatments, providing dose-limiting toxicity.

PK/PD: 2h plasma half life after po administration with modest ca 15% bioavailability in rats.

Self administration: Significant reduction in amphetamine self administration with biweekly dosing.

CCI: Significant reduction in mechanical or thermal allodynia following CCI in mice’

Obesity: Reduction in weight gain with access to either high far or regular chow @ 100 mg/kg/qod. Reduction in weight gained after exposure to high fat diet.

Conclusions: Pentilludin remains a strong candidate for development for reducing reward from psychomotor stimulants and thus aiding abstinence from stimulants. It may have indications for the allodynia that develops with chronic neuropathic pain and, at higher doses, may aid obesity treatments.

Keywords: Psychostimulant addiction, chronic pain treatment, High fat diet induced obesity, new drug development, Cell adhesion molecule

Disclosure: VA: Patent (Self).

Meghan Flanigan*, Thomas Kash

Medical University of South Carolina, Charleston, South Carolina, United States

Background: Alcohol Use Disorder (AUD) is associated with social cognition deficits that promote social withdrawal, decrease one’s capacity to seek out and receive social support, and worsen treatment outcomes. The lateral habenula (LHb) has been highly implicated in AUD and social behaviors, but to date no studies have investigated how it may contribute to social dysfunction induced by alcohol. Here, we investigated the role of LHb serotonin 5HT2c receptor signaling in alcohol-induced physiological adaptations that are associated with social recognition deficits in female mice.

Methods: Adult female Ai9x5HT2c-cre mice were exposed to the Drinking in the Dark (DiD) paradigm for three weeks, remained abstinent for one week, and then were sacrificed for LHb electrophysiology. Control mice were exposed to only water during this time. Whole-cell patch clamp electrophysiology was performed in tdTomato+ 5HT2c LHb neurons of Water and DiD mice (n = 7-10 mice/15-22 cells/group), where we measured both resting and evoked activity, serotonergic modulation of this activity, and the contributions of 5HT2c to these effects using the 5HT2c antagonist SB242084. Student’s unpaired t-tests or 2-way ANOVAs statistical tests were used to compare effects between groups.

Results: A history of binge alcohol consumption in the DiD model increased both the resting membrane potentials (RMPs) (p < 0.05) of LHb-5HT2c neurons and their tonic firing rates (p < 0.05) without affecting evoked activity (p > 0.05). In addition, evoked burst firing in these neurons was reduced following DiD (p < 0.05), suggesting that DiD constrains the dynamic range of LHb-5HT2c cell to trap them in a tonically-active but low bursting state. While 5HT strongly reduced burst firing of LHb-5HT2c in Water mice (p < 0.001), this effect was lost in DiD mice (p > 0.05). However, 5HT effectively reduced tonic activity of LHb-5HT2c in both DiD and Water mice (p < 0.001). Bath application of SB242084 normalized RMPs and tonic firing rates in DiD mice even in the absence of 5HT, suggesting that constitutive activation of 5HT2c may drive up LHb RMPs to enhance tonic firing after DiD. Furthermore, SB242084 blocked the effects of 5HT on burst firing in water mice, but enhanced bursting independently of serotonin in DiD mice. Critically, chemogenetic inhibition (Gi) of LHb-5HT2c neurons normalized social recognition deficits induced by DiD (p < 0.05).

Conclusions: These results support a model whereby binge alcohol consumption alters the physiology of LHb neurons via enhanced function of the 5HT2c receptor. This increase in LHb 5HT2c activation effectively depolarizes LHb neurons, enhances their tonic firing, and blunts their burst firing, which leads to a disruption in their capacity to recognize novel and familiar conspecifics during abstinence from alcohol.

Keywords: Serotonin 5-HT2C Receptor, Alcohol and substance use disorders, Lateral Habenula

Disclosure: Nothing to disclose.

Priscila Goncalves*, James Woodruff, Silvia Martins, Matthew Albaugh, Ardesheer Talati

Columbia University New York, New York, United States

Background: Cannabis use among women and during pregnancy has been increasing in the US. In 2022, 13.0% of women reported using cannabis in the past 30 days (compared to 5.6% in 2013). Recent studies have reported 50-62% increases in self-reported past-month cannabis use among pregnant people (e.g., 2.4% to 3.9% from 2002-2014 and 3.4% to 7.0% from 2002-2016). In tandem with the increased use is the growing perception that prenatal cannabis exposure (PCE) is not harmful despite limited awareness of its potential adverse effects. PCE may have short- and long-term consequences, including alteration in brain development. Current research suggests alterations in frontal, parietal, and temporal lobes associated with PCE. However, most of the research has been done cross-sectionally rather than looking at developmental changes over time. In the last decade, extensive studies such as the Adolescent Brain Cognitive Development Study have become available, enabling neuroimaging investigations across different age groups. Here, we aim to investigate the associations between PCE and changes in cortical structure across preadolescence (ages 9 to 13 years), a crucial period of development for the prefrontal circuitry governing risk- and reward-related behaviors.

Methods: We used data from 9,710 participants in the Adolescent Brain Cognitive Development Study(ABCD)(Release 4.0). Prenatal cannabis exposure (PCE) was defined as cannabis use at any time during pregnancy (before and/or after knowledge), by caregiver report (largely, mothers) at the ABCD study baseline. Outcomes were frontal, parietal, and temporal cortical structure (indexed by average surface area, thickness, and volume) obtained at both baseline (ages: 9–11) and Year 2 follow-up (ages: 11–13) based on the Desikan-Killiany cortical parcellation atlas. We used linear mixed-effects models to examine the extent to which PCE was related to cortical development over the age range studied, adjusting for baseline sociodemographic characteristics (e.g., sex assigned at birth, race/ ethnicity, parental marital status, household income), family history of alcohol and/or substance use problems (including for cannabis), prenatal exposure to other substances (e.g., tobacco, alcohol, and other drugs), and time-varying total intracranial volume. Family relatedness, MRI scanner, and participant ID were included as random intercepts. Analyses were conducted using the ‘lme4’ and ‘emmeans’ packages in R version 4.1.3.

Results: Among the 9,710 participants, 5.67% (N = 551) met the criteria for PCE. We ran models to test whether the effects of age on frontal, parietal, and temporal average surface area, cortical thickness, or volume varied as a function of PCE. We found significant interactions of PCE with age for average temporal cortical thickness (p = 0.02) but not for surface area (p = 0.48) or volume (p = 0.13). For example, at age 9, the overall mean of temporal cortical thickness (standardized mean) in the PCE group was 0.0135, and non-PCE was 0.0248, and at age 13, the PCE group had a std mean of -0.2501 while the non-PCE group’s mean was -0.3175, indicating a relatively slower thinning of overall temporal thickness among those with PCE. Finally, no differences were observed in age trajectories of the development of the frontal cortex (age and brain interactions: surface area, p = 0.93, thickness, p = 0.06; volume, p = 0.43) and parietal cortex (surface area: p = 0.83, thickness, p = 0.08; volume, p = 0.39).

Conclusions: PCE was associated with altered temporal cortical trajectories among preadolescents. Interestingly, portions of the temporal lobes are known to have particularly high CB1 receptor densities. Additionally, the structural development of temporal regions has been previously linked to aspects of learning and memory—two aspects of cognition consistently found to be impaired after chronic cannabis use. Whether these different trajectories lead to increased clinical risk will need to be tested in future studies as the children age. There are also limitations: 1) PCE was self-reported, which could include desirability and recall bias (as caregivers were asked to recall cannabis use from about ten years ago), 2) limited data on parental psychopathology (here, we included only history of alcohol/substance use), and 3) limited on detail on the type of cannabis use, including when in pregnancy, type of product, potency, frequency, duration of the exposure as well as motives for prenatal cannabis use (e.g., medical use for nausea symptoms or recreational).

These findings are in the opposite direction of what our research group had observed with the frontal cortical trajectory in individuals with a family history of alcohol/substance use problems (more rapid thinning among those who had a positive family history), which could suggest that different types of exposure (e.g., genetic and environmental versus in utero) have distinct biological mechanisms. Future studies could shed light on investigating individuals with multiple exposures (e.g., family history and in-utero exposure), time of cannabis exposure, and brain development. Finally, further research will have the opportunity to better understand PCE with available datasets, such as the Healthy Brain Development Study, focused on early life development, addressing some of the noted limitations in the present work.

Keywords: Prenatal Cannabis Exposure, Brain development, early adolescence, Adolescent Brain and Cognitive Development Study

Disclosure: Nothing to disclose.

Madison Marcus*, Tiffany Pignatello, Paul Howell, Shanshan Chen, F. Gerard Moeller, Gretchen Neigh, Caitlin Martin

Virginia Commonwealth University, School of Medicine, Williamsburg, Virginia, United States

Background: Individuals with opioid use disorder (OUD) have disproportionately high rates of early life stress (ELS) compared to the general population. ELS exposure is associated with an increase in both addictive drug-taking behaviors and the expression of inflammatory cytokines. Maintenance on medication for OUD, such as buprenorphine, can affect these behavioral and biological endpoints in an opposing manner. The aim of this study was to determine whether ELS exposure is associated with a distinct neurofunctional and/or inflammatory phenotype within a population of individuals maintained on buprenorphine for OUD.

Methods: The current secondary analysis (parent study: NCT03143855) consisted of 21 adults (16M/5F) maintained on buprenorphine who completed an abbreviated version of NIDA’s Phenotyping Assessment Battery, which assessed six neurofunctional domains: reward, cognition, negative emotionality, interoception, metacognition, and sleep. Participants also provided blood samples at three different timepoints relative to buprenorphine intake: 1 h before (7 AM), 2 h after (10 AM), and 8 h after (4 PM). Samples were used to assess participant levels (pg/mL) of 10 cytokines over time, and CRP levels were assessed at a single timepoint (7 AM). High and low ELS groups were determined by number of items endorsed on the Trauma History Questionnaire that occurred before 18 years old (above vs. below sample median). Scores on neurofunctional assessments were compared between ELS groups using two-sample t-tests. Cytokine levels were analyzed using a multivariate mixed-effects model adjusting for age, sex, BMI, and ELS group. Cytokine differences by ELS groups and time effects of cytokine changes were compared post-hoc with Tukey’s adjustment.

Results: There were no neurofunctional differences between individuals in the high ELS group (n = 10; 7M/3F; median traumatic events: 8.5 [range 4-13]) compared to the low ELS group (n = 11; 9M/2F; events: 3 [0-3]) across any of the six domains assessed. Proinflammatory cytokines IFNγ (p < 0.001) and IL-6 (p < 0.0001) were elevated in the high ELS group across timepoints compared to the low ELS group, whereas anti-inflammatory IL-13 was lower in the high ELS group (p = 0.012). Notably, ELS groups did not differ with respect to CRP (or other tested cytokines’) levels, age, sex, buprenorphine dose, BMI, smoking status, psychiatric or substance use disorder comorbidity. Participant levels of IL-6 were elevated at the end of the day (4 PM) compared to the beginning of the day (7 AM; p < 0.001) across ELS groups.

Conclusions: We identified a positive association between select proinflammatory markers and ELS in a buprenorphine-stabilized OUD treatment population. The results of this study extend previous findings of an ELS-associated pro-inflammatory state within the general population to our clinically relevant OUD treatment population. Future work with larger samples balanced by sex may help inform precision medicine strategies to tailor buprenorphine-based OUD treatment regimens to individuals’ biopsychosocial needs.

Keywords: opioid use disorder, Buprenorphine, inflammation, early life stress (ELS), neurofunction

Disclosure: Nothing to disclose.

Justin Matheson*, James MacKillop, Matthew Sloan, Marcos Sanches, Harseerat Saini, Rebecca Haines-Saah, Adam Zaweel, Ahmed Hassan, Leslie Buckley, Amy Porath, Christian Hendershot, Stefan Kloiber, Bernard Le Foll

Centre for Addiction and Mental Health, Toronto, Canada

Background: Cannabis use has emerged as a potential harm reduction approach for individuals with substance use disorders (SUDs), while cannabinoids may have promise for pharmacological treatment of SUDs. However, to date, there are very few studies examining cannabis use in relation to clinical mental health characteristics in patient samples. In order to better understand correlates of cannabis use among the SUD patient population, we aimed to characterize self-reported trauma history and mental health in relation to cannabis use status among a sample of patients seeking treatment for SUDs.

Methods: We conducted an online anonymous survey using REDCap (Research Electronic Data Capture) that recruited patients aged 18+ years accessing addiction treatment services in Ontario, Canada (inclusive of any sex at birth or gender identity). N = 544 participants were grouped based on their cannabis use status: any past-year use (Current Use [CU]; n = 363), lifetime use but no past-year use (Past Use [PU]; n = 109), and no lifetime use (Never Use [NU]; n = 72). We assessed trauma history with a researcher-derived measure (trauma before age 18 and trauma as an adult), anxiety with the Generalized Anxiety Disorder scale (GAD-7), and depression with the Patient Health Questionnaire (PHQ-9). Chi-squared tests examined associations between categorical variables (with Cramer’s V as an effect size estimate), while ANOVA evaluated cannabis group differences in GAD-7 and PHQ-9 scores (with partial-eta squared [η2] as an effect size). ANCOVA also examined if associations between cannabis use and mental health remain significant when controlling for trauma history.

Results: There were some significant demographic differences between groups (e.g., the CU and PU groups were slightly older than the NU group; the CU group had a higher proportion of participants identifying as sexual minorities and as white, and a lower proportion identifying as Asian, Indigenous, or Middle Eastern). We found a significant association between cannabis use group and both trauma before age 18 (V = 0.30, p < 0.001) and trauma as an adult (V = 0.34, p < 0.001). For both variables, all three groups significantly differed from one another: the CU group had the highest proportion of self-reported trauma (trauma before age 18: 50% and trauma as an adult: 57%), followed by the PU group (37% and 41%), then the NU group had the lowest (7% and 7%). All six types of trauma we assessed (physical aggression or assault, verbal aggression or abuse or threat, severe accident or injury, sexual trauma or assault or abuse, neglect, and other type) were significantly different between groups and always highest in the CU group, for both pre-age-18 and adult trauma. There were significant group differences in both GAD-7 scores (η2 = 0.084, p < 0.001) and PHQ-9 scores (η2 = 0.095, p < 0.001). Once again, for both outcomes, all three groups significantly differed from one another: the PU group had the highest scores (mean [SE] GAD-7 = 12.0 [0.69]; PHQ-9 = 14.8 [0.83]), followed by the CU group (GAD-7 = 9.0 [0.38]; PHQ-9 = 11.1 [0.46]), then the NU group (GAD-7 = 4.3 [0.88]; PHQ-9 = 5.0 [1.0]). Controlling for trauma history did not impact the results for GAD-7 or PHQ-9.

Conclusions: Any lifetime cannabis use was associated with increased likelihood of reporting trauma and with greater anxiety and depression in a sample of patients seeking treatment for SUDs. However, the relationships differed depending on cannabis use status: current use was associated with more self-report of trauma both before age 18 and as an adult, while past use was associated with higher GAD-7 (anxiety) and PHQ-9 (depression) scores. Importantly, the relationship between cannabis use and anxiety/depression scores remained significant when controlling for trauma history, suggesting that the group differences are not due simply to differences in self-report of trauma. As these data are cross-sectional, we are unable to determine the direction of this association (i.e., if patients with past trauma and more severe mental health symptoms seek out cannabis for symptomatic relief or whether cannabis use causes/exacerbates mental health symptoms). Future longitudinal and experimental research is needed to determine whether and under which conditions cannabis ameliorates or worsens mental health symptoms in patients with SUDs. Nevertheless, our results suggest that lifetime cannabis use is associated with a more complex clinical presentation in patients seeking SUD treatment.

Keywords: cannabis, substance use disorder, Trauma exposure

Disclosure: Nothing to disclose.

Luciano Voutour, Leo Pereira Sanabria, Victoria Braman, Christian Valade, Amy Arguello*

Michigan State University, East Lansing, Michigan, United States

Background: Drug use during adolescence has the potential to permanently alter brain processes and increase the life-long risk to develop substance use disorders (SUDs). Adolescents are particularly sensitive to peer-peer interactions- they are more likely to consume drugs with peers who engage in drug use. Preclinical studies have also shown that adult and young rats self-administer more cocaine in the presence of another cocaine-administering peer. However, less is known about the neural mechanisms by which adolescent cocaine exposure impacts social behavior. We find dense projections in adolescent rats from the perirhinal cortex to the dorsal hippocampus (PRh-- > DH). The perirhinal cortex plays a key role in object recognition and stimulant exposure impairs PRh-dependent object recognition, but the role of PRh-- > DH circuit in social interaction following cocaine exposure is unknown.

Methods: To determine the role of the PRh -- > DH circuit in social interaction following cocaine exposure, we passively administered cocaine (3.0 mg/kg) or saline to adolescent male rats, followed immediately by social interaction. Social interaction occurred for 15min/day across 5 days between the following pairs: cocaine-cocaine, saline-saline or saline-cocaine. Rats were perfused 90 minutes after the last interaction session to assess changes in neuronal activation (Cfos protein) within the PRh (a subset of rats also received retrograde tracer into the DH to examine PRh-- > DH specific activation). Social interaction was video recorded throughout the experiment.

Results: We find that cocaine-cocaine interacting pairs have reduced numbers of Cfos-positive cells in the PRh compared to saline-saline interacting pairs. Future work aims to determine whether PRh -- > DH circuit-specific activation patterns correlate with unique behavioral motifs in cocaine-cocaine, saline-saline, and saline-cocaine interacting pairs using the supervised machine learning programs DeepLabCut and Simple Behavioral Analysis.

Conclusions: Future work aims to determine whether PRh -- > DH circuit-specific activation patterns correlate with unique behavioral motifs in cocaine-cocaine, saline-saline, and saline-cocaine interacting pairs using the supervised machine learning programs DeepLabCut and Simple Behavioral Analysis.

Keywords: adolescent, cocaine, relapse, social interaction, machine learning

Disclosure: Nothing to disclose.

Laura Murray*, Maria Scavnicky, Blaise Frederick, Amy Janes

National Institute on Drug Abuse, Intramural Research Program, Baltimore, Maryland, United States

Background: Cue reactivity paradigms are common in addiction research. In nicotine dependence, cue-reactivity tasks using images of combustible cigarettes have been shown to robustly increase subjective ratings of craving, and activate regions of the default mode network (DMN) and anterior insula (AI), regions that play a role in internally-oriented attention and salience processing, respectively (Janes et al., 2020; Wang et al., 2020). However, as the prevalence of nicotine vaping continues to rise, it is essential to validate vaping-specific cue-reactivity tasks and test whether vaping and smoking groups have comparable neurobiological and behavioral responses to use-specific cues. If vaping and smoking groups do not differ on brain reactivity and behavioral responses to cue-reactivity paradigms, existing literature on smoking cue-reactivity could be used to inform research on electronic cigarette use, and combining cue-reactivity data across smoking and vaping groups could be supported. The aim of the current study was to test whether individuals who use combustible cigarettes and individuals who use electronic cigarettes differ in their: 1) brain response to use-specific smoking or vaping cues versus neutral stimuli, 2) subjective ratings of task stimuli, and 3) cue-induced subjective report of craving. We hypothesized that the brain and behavioral pattern of responses to nicotine cues versus neutral stimuli would be similar for both groups, with increased DMN and AI activity to cues compared to neutral stimuli and increased subjective craving after the cue-reactivity task.

Methods: 86 individuals who used nicotine daily completed a cue-reactivity task during fMRI. Analyses compared brain reactivity and behavior in 66 daily combustible cigarette users (smoking group: 60% male 40% female; mean age=29.18) and 19 daily electronic cigarette users (vaping group: 68% male, 32% female, mean age= 24.79). Individuals in the smoking group completed five, six-minute runs of a task with smoking cues vs. neutral images, and those in the vaping group completed an analogous 5-run task with electronic cigarette cues vs. neutral images. Following MRI, participants rated stimuli on craving and arousal. Nicotine craving was assessed pre and post MRI scan (Tiffany and Drobes, 1991). fMRI data were analyzed in FSL, and beta weights for the DMN and AI regions of interest (ROI) were extracted for cue and neutral stimuli conditions. Smoking and vaping groups were compared on brain response (a priori ROIs and exploratory whole brain), subjective ratings of stimuli, and cue-induced craving.

Results: Participants rated cues as significantly more craving-inducing (F = 75.94, p < 0.001) and arousing (F = 5.01, p = 0.03) compared to neutral images, and there was no effect of group (smoking versus vaping) on these ratings (F = 0.43, p = 0.52; F = 288.0, p = 0.09, respectively). There was a significant increase in craving following the cue-reactivity task (post-pre; F = 9.99, p = 0.002), and the groups did not differ on cue-induced craving (F = 0.02, p = 0.90). ROI analyses indicated a significantly greater DMN activation to cues versus neutral (F = 46.35, p < 0.001) and no differences between the two groups (F = 2.74, p = 0.10). Similarly, the AI had significantly higher activation to cues versus neutral (F = 13.51, p < 0.001), but did not differ between groups (F = 0.93, p = 0.34). To explore potential brain habituation to the task stimuli, we found that across the five runs, DMN and insula activity remained significantly higher for cues versus neutral (DMN: F = 36.97, p < 0.001; insula: F = 11.49, p = 0.001) and did not differ between vaping and smoking groups (DMN: F = 0.17, p = 0.96; insula: F = 1.44, p = 0.22). In exploratory whole-brain analyses, the smoking and vaping groups both showed robust activation in core nodes of the DMN to cues versus neutral and did not statistically differ from each other.

Conclusions: As hypothesized, both the smoking and vaping-specific versions of a cue-reactivity task robustly activated regions of the DMN and AI to nicotine (smoking or vaping, depending on cohort) cues relative to neutral images. There were no statistical differences between the smoking and vaping groups in the DMN and AI ROIs and across the exploratory whole-brain analysis. Across both tasks, nicotine cues were rated as more craving-inducing and arousing than neutral stimuli, and both groups showed significant increases in subjective nicotine craving following the cue-reactivity task. Finally, the relative activation to cues versus neutral remained robust over time, indicating there was no habituation to nicotine cue-reactivity in either group. Together, these findings indicate that brain and behavioral reactivity to smoking and vaping cues are comparable across individuals who use combustible cigarettes and those who use electronic cigarettes. As nicotine vaping becomes increasingly prevalent, our work suggests that samples of individuals who smoke and those who vape nicotine can be combined to characterize brain and behavioral responses to nicotine cues regardless of route of administration.

Keywords: nicotine dependence, fMRI, cue-reactivity, vaping, Smoking

Disclosure: Nothing to disclose.

Zhuzhu Zhang*

University of Chicago, Chicago, Illinois, United States

Background: The reward circuitry comprises a vast diversity of cell types, each defined by specific molecular and anatomical properties, including their locations and neuronal projections. These cell types perform distinct functions and respond differently to exposure to substances of abuse in the context of addiction. Thus, dissecting these cellular molecular responses, particularly gene expression alterations mediated by epigenomic remodeling, at cell-type resolution and within the context of connections between brain regions and cell types in the reward circuitry, is crucial for understanding the molecular underpinnings of addiction.

Methods: My lab adopts an interdisciplinary approach that combines the mouse model of intravenous self-administration (IVSA) of cocaine with innovative single-cell epigenomics assays. Specifically, I have developed Epi-Retro-Seq, a cutting-edge single-cell epigenomics method that merges retrograde tracing with single-cell DNA methylome and RNA sequencing. This approach allows for the simultaneous investigation of epigenetic and gene regulation in individual cells and their connectivity within the reward circuitry. We have proven the effectiveness of this strategy in exploring the principles that link neuronal projections to transcriptomics and epigenomics in the adult mouse brain. Moreover, we have integrated Epi-Retro-Seq with spatial transcriptomics, including multiplexed error-robust fluorescence in situ hybridization (MERFISH), to further refine the precise spatial locations of the cell types of interest.

Results: We are applying these methodologies in concert to the cocaine IVSA mouse model promises to reveal cell type-specific epigenetic mechanisms in addiction and to provide essential insights for subsequent targeted functional studies of key epigenomic regulators. In the pilot study, we analyzed the dynamic landscapes of single-cell epigenomes and gene regulation in the reward circuitry after cocaine IVSA and abstinence. We observed genome-wide epigenomic changes, the majority of which were cell-type specific. In addition, many of these epigenomic changes were enriched at cell-type specific distal DNA regulatory sequences such as putative enhancers, highlighting the complex gene regulatory machinery in response to cocaine IVSA.

Conclusions: Our ongoing studies combine single-cell multi-omics, spatial transcriptomic, mouse IVSA models and integrative computational analysis. These studies will illustrate cell-type and projection-specific epigenomic remodeling in drug self-administration and abstinence, and provide the crucial knowledge for follow-up targeted functional studies of the role of epigenomic regulations in addiction.

Keywords: Epigenetics, Genomics, brain epigenomics, gene regulation, addiction circuitry

Disclosure: Nothing to disclose.

Michelle Doyle, Maria Balaguer, Paola Campo, Thiago Sanches, Ran Qiao, Courtney Wood, Marsida Kallupi, Abraham Palmer, Giordano de Guglielmo*

University of California - San Diego, La Jolla, California, United States

Background: Alcohol use is prevalent among over 175 million Americans each year, with approximately 16% developing Alcohol Use Disorder (AUD). This indicates the presence of significant individual differences in the development of problematic drinking, potentially attributable to genetic variations.

Methods: In this extensive study, over 500 male and female HS rats were characterized for addiction-like behaviors. Rats self-administered oral ethanol (10% v/v) on a fixed ratio 1 schedule of reinforcement until a stable baseline of intake was measured. Subsequently, multiple addiction-related behaviors, such as preference for ethanol over water, progressive ratio responding, and level of quinine-adulterated ethanol intake, were assessed. Dependence was induced using chronic intermittent ethanol vapor exposure (14 hours/day x 4 weeks, achieving blood ethanol levels of 150-250 mg%) and behavioral experiments were conducted during acute withdrawal (6-8 hours after vapor) from ethanol. Assessments included escalation of ethanol intake, motivation to obtain ethanol, compulsivity, sensitivity to alcohol, somatic withdrawal signs, and withdrawal-induced hyperalgesia. An Addiction Index was computed from the mean Z scores of the five measurements.

Results: The robust individual differences in the Addiction Index suggest that HS rats present diverse AUD-related phenotypes, likely associated with genetic factors influencing AUD development. To identify genetic variants associated with compulsive alcohol use, we performed a genome-wide association study (GWAS) on these behaviorally characterized rats. Preliminary results highlighted the calcineurin gene as a potential hit, suggesting its involvement in AUD-related behaviors. To validate this finding, we conducted a pharmacological study using tacrolimus, a calcineurin modulator. Administration of tacrolimus resulted in a dose-dependent reduction in alcohol drinking in approximately half of the rat population, with particularly pronounced effects in heavy drinking subjects.

Conclusions: These findings, derived from characterizing addiction-like behaviors in over 500 rats and subsequent genetic and pharmacological analyses, emphasize the importance of pharmacogenetic studies and suggest the potential application of precision medicine in AUD treatment. The identification of calcineurin as a potential target and the differential response to tacrolimus highlight the complex interplay between genetic factors and treatment efficacy in AUD, paving the way for more personalized therapeutic approaches.

Keywords: Alcohol Use Disorder - Treatment, GWAS, Dependence

Disclosure: Nothing to disclose.

Cali Calarco*, Brian Herb, Isaiah Williamson, Symphanie Key, Christina Barrett, Gautam Kumar, Seth Ament, Mary Kay Lobo

University of Maryland School of Medicine, Baltimore, Maryland, United States

Background: Exposure to illicit drugs and subsequent chronic use profoundly impacts behavior, neuronal structure and firing, and gene expression in reward-related brain regions. Some of these changes are mediated by altered cellular energy homeostasis and mitochondrial function. Cocaine self-administration significantly reduces mitochondrial size in nucleus accumbens (NAc) neurons, and disruption of this process is sufficient to blunt cocaine seeking. Cocaine also has widespread impacts on transcription across the brain, including influencing mitochondrial-related genes, which may influence downstream changes in mitochondrial morphology and functioning. Understanding the cocaine-induced transcriptional changes across reward-related brain regions, and more specifically, understanding cocaine-induced transcriptional changes in a circuit-specific manner will improve out understanding of drug-induced plasticity. In this study we examined both bulk and circuit-specific mRNA sequencing data allowing us to examine both unique circuit markers as well as cocaine-induced transcriptional profiles.

Methods: Taking a brain-wide perspective, we examined bulk and circuit-specific transcriptional changes relating to cellular metabolism and mitochondria, concentrating on reward-related brain regions and inputs to NAc after IV cocaine self-administration C57Bl/6 mice. Using differentially expressed genes (DEG) identified in previously published bulk RNA sequencing data sets from mice that had undergone either cocaine self-administration or received acute cocaine (5-8 mice per group) we performed subsequent gene ontology analysis, with a focus on mitochondrial-related ontology terms. The sequencing data included tissue from the prefrontal cortex, NAc, dorsal striatum, ventral pallidum, amygdala, hippocampus, and ventral tegmental area. In a further circuit-specific analysis, we conducted ribotag-based labeling, isolation, and sequencing of mRNA from neurons in the prefrontal cortex, ventral hippocampus, and the ventral tegmental area that project into NAc as well as the input (bulk tissue) fraction of these samples after cocaine self-administration in male and female mice (n = 4 per group/sex). Similar DEG and gene ontology analyses we performed in addition to circuit-selective gene enrichment analysis.

Results: In bulk tissue samples we found significant representation of DEGs (corrected p < 0.05) in metabolism and mitochondrial-related ontology terms, with regional and cocaine exposure-related variability. Further, predictive analysis of transcription factors regulating mitochondrial-related genes identified multiple transcription factors that may control cocaine-related changes in metabolic function. Ongoing analysis is examining expression of metabolism-related genes (DEG FDR < 0.05), predicted transcription factors, and projection-specific characterization in the circuit-specific transcripts.

Conclusions: Together these data define the landscape of metabolism-related transcriptomic changes across reward regions in response to cocaine and provide the first circuit-selective transcriptomic characterization of NAc inputs after cocaine self-administration. Understanding circuit-specific transcriptional changes will inform how cellular metabolism supports responses to cocaine throughout reward circuits.

Keywords: substance use disorders, cocaine self-administration, Transcriptome, mesolimbic reward circuitry, Cellular metabolism

Disclosure: Nothing to disclose.

Alejandra Pulido Saavedra*, Mika Naganawa, Marcella Mignosa, Brian Pittman, Patrick Worhunsky, Nabulsi Nabeel, Yiyun Huang, Richard Carson, Marc Potenza, Gustavo Angarita

Yale University, New Haven, Connecticut, United States

Background: Cocaine use disorder (CUD) is recalcitrant to treatment. Understanding synaptic density in people with CUD may promote CUD treatment development. Preclinical data suggest increased dendritic branching and numbers of dendritic spines following cocaine exposure. Our previous Positron Emission Tomography (PET) study using the SV2A ligand [11C]UCB-J demonstrated a lower volume of distribution/plasma free fraction (VT/fP) in the anterior cingulate cortex (ACC), ventromedial prefrontal cortex (vmPFC), and medial orbitofrontal cortex (mOFC) in people with CUD compared to those without. In addition, we observed positive correlations between the frequency of cocaine use and VT/fP measures and a negative association between days of hospital-based abstinence and VT/fP measures. These findings suggest premorbid (i.e., before cocaine use) deficits in synaptic density in CUD. This aim of this study was to examine longitudinal changes in [11C]UCB-J SV2A binding with cocaine abstinence in people with CUD.

Methods: Seventeen participants with CUD were admitted as inpatients to complete two PET scans. The baseline [11C]UCB-J PET scan was 3.9 ± 1.7 days after the last cocaine use, and the second scan was performed after three-week (23.9 days ± 2.7) hospital-based abstinence from cocaine. A Siemens HRRT scanner was used for 60 minutes of dynamic PET data acquisition. Our primary outcome was VT/fP measures. Secondary outcome was the volume of distribution (VT), both derived using the 1-tissue compartment model in the following a priori regions of interest (ROIs): ACC, vmPFC, dorsomedial prefrontal cortex (dmPFC), lateral orbitofrontal cortex (IOFC), ventral striatum (VS), and mOFC. The centrum semiovale (CS), which serves as a reference region to measure non-displaceable uptake, was also evaluated. These ROIs were selected using an automated anatomical labeling template, and the VS was drawn by hand. A 30-minute magnetic resonance (MR) imaging was performed within the first week after admission using a 3 T Trio Scanner (Siemens Medical Systems, Erlangen, Germany) and rapid acquisition gradient echo (MPRAGE) sequence. MR images were transformed into Montreal Neurological Institute space. Pre-post changes in each outcome were analyzed using linear mixed models with time (baseline, 3 weeks abstinence) and region as within-subjects factors and random subject effects.

Results: Participants were 47 ± 12.5 [SD] years of age, primarily male (14 male/3 female), and diverse (10 African American, 6 Caucasian, 1 with more than one race). The average lifetime use of cocaine was 19 ± 11.8 years, and participants spent, on average, $93 ± 60 per day on cocaine. A significant effect of time was observed for both VT/fP (F(1, 176) = 4.3, p = 0.040) and VT (F(1, 176) = 3.94, p = 0.049). Post-hoc analysis showed significant [11C]UCB-J VT/fP reductions following abstinence in the vmPFC (reduction of 7%, p = 0.041), dmPFC (reduction of 7%, p = 0.011), IOFC (reduction of 6%, p = 0.021), and mOFC (reduction of 7%, p = 0.028). Similarly, [11C]UCB-J VT reductions were observed in the vmPFC (reduction of 6%, p = 0.047), dmPFC (reduction of 7%, p = 0.035), IOFC (reduction of 6%, p = 0.035), and mOFC (reduction of 6%, p = 0.011). For both outcomes, no significant effects of time were observed in the ACC, VS, and the reference CS.

Conclusions: These findings offer preliminary evidence suggesting synaptic density declines during the first three weeks of cocaine abstinence in CUD. Future analyses will compare changes in SV2A binding between people with and without CUD and explore potential associations between PET outcomes and cognitive and clinical variables. Results could motivate future studies with synaptogenic agents with limited misuse potential.

Keywords: PET Imaging, SV2A protein, cocaine use disorder

Disclosure: Nothing to disclose.

Marissa Borrego*, Nailyam Nasirova, Timothy O’Neal, Zackari Murphy, Susan Ferguson

University of Washington School of Medicine, Dept. of Psychiatry and Behv. Sci., Seattle, Washington, United States

Background: Men and women with opioid use disorder (OUD) show differences in their reasons for initiation, rates of use, and progression through the addiction cycle. While some differences can be attributed to gender-based social influence, preclinical data from animal models also suggest sex-based differences in reward circuitry and drug-related behaviors. As the importance of including sex as a biological variable becomes more apparent, it is sensible to consider which facets of OUD are particularly dependent on sex.

Methods: We directly compared adult male and female rats across an array of heroin-induced behaviors in order to provide insight into the nuances of sex differences in OUD. We first used a 6-hour intermittent access heroin self administration paradigm to quantify six distinct drug-taking and -seeking behaviors (N = 64). Based on the sum of the z-scores for each behavior, we classified rats as having high- or low-severity phenotypes. In a separate group of rats (N = 34), we adapted this classification system to a 1-hour continuous access self administration paradigm to better represent the timeframe of use common in people. Finally, we examined locomotor sensitization following daily heroin injections in two groups of rats. The first were given 2 mg/kg/day i.p. heroin for 9 days (N = 31) and the second were given 0.55 mg/kg/day i.v. heroin for 20 days (N = 11). Changes from baseline were analyzed using repeated measure 2-way ANOVAs (sex x time).

Results: In the 6-hour intermittent access, both sexes showed variability across individuals, but a greater proportion of females were classified as having a high-severity phenotype compared to males (OR = 2.14). This difference in severity distributions was also found in the 1-hour continuous access experiment (OR = 5.50). Consistent with the literature, in our sensitization experiments we found that males had a lower baseline level of locomotion compared to females (p < 0.001). Across sex and route of administration, rats treated with heroin initially decreased locomotion, but returned to baseline over the course of treatment. Females given i.v. infusions showed a rapid escalation (p < 0.01) of locomotion past baseline that was not seen in males or following i.p. injections.

Conclusions: Taken together, our results indicate a consistent pattern of females having a greater behavioral response to heroin compared to males. These findings support the need to consider sex as a biological variable in our development of treatments for OUD and other substance use disorders. Our ongoing research is focused on more fully fleshing out sex differences in locomotor sensitization that are specific to i.v. heroin. Additionally, we are measuring serum levels of gonadal hormones across this treatment period to analyze how chronic heroin may affect steroid hormone signaling.

Keywords: sex differences, Heroin, sensitization, locomotion, Self-Administration

Disclosure: Nothing to disclose.

Justin Strickland*, Cecilia Bergeria, Andrew Huhn, Kelly Dunn

Johns Hopkins University School of Medicine, Baltimore, Maryland, United States

Background: Epidemiological data document increased adulteration of the opioid drug supply by the veterinary anesthetic xylazine. Preclinical data suggest that xylazine exposure may alter the expression of opioid withdrawal including changes in the time course, severity, and number of specific withdrawal symptoms experienced. While these data have helped raised concerns about developing trends in xylazine exposure, the consequent impacts for treatment trajectories and courses of care for people with opioid use disorder remain unclear. This analysis leverages data from ongoing human laboratory studies to evaluate the association between xylazine exposure and opioid withdrawal in people with opioid use disorder within a controlled clinical environment.

Methods: Participants (n = 21; 13 male) were enrolled in a study evaluating opioid withdrawal phenotypes in a residential research setting. Participants completed a residential stay in a controlled setting in which monitored abstinence from illicit opioids (e.g., illicitly manufactured fentanyl) was conducted. Urine samples were collected at study admission and were tested for xylazine and norfentanyl concentrations using mass spectrometry. Opioid withdrawal was evaluated using the Clinical Opioid Withdrawal Scale (COWS). Physiological outcomes included systolic and blood pressure. Withdrawal and physiological measures were collected four times per day during induction on oral opioid agonist maintenance. Admission xylazine and norfentanyl concentrations were evaluated for their association with clinical outcome measures using general linear mixed effect models.

Results: Urinary xylazine concentrations above the limit of detection (5 ng/mL) were detected in 76% of samples with concentrations ranging from 5.7 to 30,900 ng/mL. Urinary norfentanyl concentrations were detected in all samples ranging from 63.1 to 28,531 ng/mL. Xylazine and norfentanyl concentrations were not significantly associated with each other (r = 0.11; p = 0.62). Higher xylazine concentrations were significantly associated with elevated withdrawal (p = 0.002) and diastolic blood pressure (p = 0.03) with a trend for associations with elevated systolic blood pressure (p = 0.07). These findings were not attributable to overall opioid exposure as associations remained similar when controlling for fentanyl concentration. Effect sizes for average withdrawal and blood pressure on the final day of testing were of a moderate to large effect size (r = 0.42 to 0.46). No differences in clinical associations or urinary concentrations were observed by biological sex or by gender.

Conclusions: These data indicate exposure-dependent alterations in opioid withdrawal severity and blood pressure related to xylazine. Such changes may lead to increasingly complex opioid withdrawal and alter the utility of existing pharmacological approaches for withdrawal management in clinical settings.

Keywords: opioid withdrawal, Xylazine, Opioid treatment

Disclosure: Nothing to disclose.

Kevin Braunscheidel*, Rohan Ghoshal, Mohammad Ishmam, Masago Ishikawa, Paul Kenny

Icahn School of Medicine At Mount Sinai, New York, New York, United States

Background: Nicotine, the major rewarding component of tobacco smoke, also has aversive properties that control the amounts of tobacco consumed by smokers. Sensitivity to nicotine aversion influences vulnerability to tobacco use disorder. Our laboratory has identified a network of hindbrain regions that regulate nicotine aversion including the nucleus of the solitary tract. This region receives dense afferent sensory input from peripheral organs via the nodose ganglia (NG) of the vagus nerve to help regulate feeding and other appetitive behaviors. This is important because nicotine is known to act at peripheral sites in human smokers (e.g., lung, heart, gut) to elicit sensory responses that regulate patterns and amounts of tobacco use. In previous studies, I have found that nicotine drastically alters the transcriptome of the NG and elevates postprandial cholecystokinin (CCK) levels in mice, similar to smokers. Here, we test the hypothesis that CCK released by enteroendocrine cells in the gut activates CCKR + NG to potentiate aversion signal transmittance centrally in order to regulate nicotine intake in mice.

Methods: Conditioned place aversion to an aversive dose of nicotine and a periphery-restricted nicotine receptor agonist, methylnicotinium was performed. Next, PHP.S-DIO-HM4Di was injected into TRAP2 animals that express Cre in an activity-dependent, temporally-restricted manner yielding inhibitory chemogenetic access to peripheral nervous tissue prior to nicotine place conditioning. We then recorded GCaMP activity ex vivo in CCKR + NG in response to nicotine in CCKR-ai96 mice. Next, we measured nicotine intravenous self-administration (IVSA) in mice following stimulation of peripheral CCKRs with CCK-8 or inhibition with dexamethoniun. In a separate cohort, CCKR + NG were ablated via intra-NG injection of CCK-saporin prior to nicotine IVSA. Finally, CCK-DTR mice were generated that express the diphtheria toxin receptor in CCK+ cells. These mice were injected with a custom ‘PEGylated’ diphtheria toxin, which does not cross the blood-brain barrier, to ablate only peripheral CCK-producing cells before nicotine IVSA. For IVSA studies, the number of nicotine infusions earned was compared using 2-way ANOVAs (treatment x nicotine dose) and post-hoc comparisons with corrections for multiple comparisons, where appropriate. All experimental protocols in animal studies were approved by the Institutional Animal Care and Use Committee and were conducted per the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Results: Peripheral nicotinic receptor activation induced conditioned place aversion in excess of an equimolar, aversive dose of nicotine. Nicotine directly stimulates CCKR + NG and contributes to the release of CCK from enteroendocrine cells. Increased nicotine intake resulted from systemic CCKR blockade (devazepide vs vehicle, 0.10 mg*kg-1 nicotine per infusion, t(36) = 4.007, p = 0.0003), CCKR + NG ablation (CCK-SAP vs Blank-SAP main effect, F(1,24) = 10.12, p = 0.0040), and CCK-producing enteroendocrine cell ablation (PEGyDT-treated vs vehicle, 0.10 mg*kg-1 nicotine per infusion, t(63) = 2.716, p = 0.0253). Conversely, stimulating peripheral CCKR signaling with CCK-8 decreased nicotine intake (CCK-8 vs vehicle main effect, F(1,56) = 37.17, p < 0.0001).

Conclusions: Peripheral actions of nicotine regulate nicotine intake due in part to their influence on CCK-releasing enteroendocrine cells and CCKR-expressing vagal sensory afferents. The existence of this novel “bottom-up” regulation of nicotine intake may prove useful for the development of novel addiction therapies with distinct therapeutic potential given that their activation does not require drug permeation of the blood-brain barrier.

Keywords: vagus, enteroendocrine cell, cholecystokinin, nicotine, peripheral nervous system

Disclosure: Nothing to disclose.

Gustavo Angarita*, Razi Kitaneh, Anna Borelli, Marcella Mignosa, Emalee Dauginikas, Celine Li, Nia Fogelman, Elcin Sakmar, Zachary Harvanek, Jane Taylor, Ralph DiLeone, Rajita Sinha

Yale University School of Medicine, New Haven, Connecticut, United States

Background: The rampant misuse of opioids and the high number of overdoses make medication development efforts for opioid use disorder (OUD) a public health priority. Albeit availability of FDA approved medications for OUD, significant attrition and relapse risk remains. Thus, development and testing of novel pharmacological targets remains a high priority. One of these targets is the glutamatergic system. The glutamatergic compound NYX-783 (Tenacia Biotechnologies), an NMDA positive allosteric modulator, has been shown to affect synaptic plasticity and learning/memory processes, and both are implicated in OUD and in the vulnerability to relapse. Previous preclinical research showed that NYX-783 is safe and has no significant effect on physiological signs of opioid withdrawal nor opioid self-administration, and NYX-783 safety of doses up to 600 mg was also examined and shown in healthy volunteers. However, no research has examined it’s safety and tolerability nor opioid drug interaction effects for OUD.

Thus, we undertook a phase 1 safety, tolerability and drug-drug interaction (DDI) study to assess two single doses of NYX-783 in combination with single 15 mg and 30 mg doses of Oxycodone in a randomized, placebo controlled, double blind study with recreational users of opioids.

Methods: Nine non-treatment seeking men with an average age of 33.4 ( ± 5.32) who use opioids recreationally were admitted to an inpatient research unit over the course of 3 weeks for 7 experimental sessions, after full medical and psychiatric screening.

After an initial safety session (Day 1) assessing Oxycodone drug response to 15 mg and 30 mg each, subjects started the double-blind treatment over 6 separate experimental sessions (days 2-7) utilizing a cross-over, counterbalanced experimental design. Subjects were randomized to receive Placebo or NYX-783 (NYX50 or NYX150) at 9:00 AM and then either 15 mg or 30 mg of Oxycodone at 11:00 am, for each session. Sessions were separated from each other by 48 hours.

Subjects were continuously monitored throughout each session by ACLS certified nurse and sessions were supervised by board certified study physicians.

A linear mixed-methods factorial design was performed to analyze physiological and subjective differences between Placebo, NYX-783 (NYX50; NYX150), and Oxycodone doses at a significance threshold of p < 0.05.

Results: No severe nor serious adverse events (AEs) were reported and 85 % of the AEs were mild.

In the safety session expected Oxycodone dose effects were observed with significant reductions in heart rate (HR), temperature and respiration rate (RR) for the 30 mg vs the 15 mg dose (p’s < 0.05) and significant increases in opioid drug effect of feeling high over time (p < 0.00001). For the drug-drug interaction sessions (Days 2-7), NYX-783 doses increased HR over placebo (p < 0.05) and there was an Oxycodone main effect (p < 0.05) for systolic blood pressure (SBP: lower after 30 mg vs. 15 mg; p < 0.01). Notably, NYX-783 doses X Oxycodone interactions were found for diastolic blood pressure (DBP: p < 0.0003) and with body temperature (p < 0.00001) where NYX150 normalized the lowering of DBP that occurred with NYX50 vs placebo (p < 0.02) and the increases in temperature that occurred with NYX50 relative to placebo (p < 0.003). There was also a DDI effect for subjective drug sedating/depressive effect in which both NYX doses reduced the sedative and depressive effect of 30 mg Oxycodone (p’s < 0.01). None of the participants experienced emergence of opioid withdrawal as a result of co-administration of Oxycodone with NYX-783.

Conclusions: Preliminary results support good safety and tolerability of NYX-783 in single administration of both 50 mg and 150 mg doses in combination with 15 mg and 30 mg of Oxycodone among recreational users of opioids. Pertaining to the DDI effects seen for DBP, body temperature and for sedating/depressive subjective effects, findings suggest a potential protective role of NYX-783 against Oxycodone effects, especially at the NYX 150 mg dose. Ongoing analyses will test pharmacokinetic effects of NYX-783 on oxycodone. Current results support additional future development of NYX-783 as a potential new treatment avenue for opioid relapse prevention in OUD.

Keywords: phase 1 study, oxycodone, NYX-783, opioid use disorder, drug-drug interaction

Disclosure: Nothing to disclose.

Gabriella Malamud, Lyne Baaj, Sophia Hanna, Mathilde Argote, Rachel Rabin*

McGill University, Montreal, Canada

Background: Despite increasing rates of cannabis use disorder (CUD), treatments remain largely ineffective. Targeting cannabis withdrawal may offer a promising therapeutic approach, as it strongly predicts relapse. Symptoms begin 24 hours after cessation, peak within one week, and subside following 28 days of abstinence. Some studies suggests that females experience more severe cannabis withdrawal compared to males. However, these data are derived from cross-sectional studies or studies that surveyed short abstinence periods (i.e., < 28 days). Since cannabis withdrawal can last for several weeks with fluctuations in severity, we investigated sex differences in cannabis withdrawal severity during 28 days of cannabis abstinence.

Methods: Males (n = 12) and females (n = 10) between the ages 18-55 diagnosed with DSM-5 CUD were recruited. Eligible participants had a positive cannabis urine toxicology at screen and had no current comorbid DSM-5 Axis 1 diagnoses (except for CUD). Participants underwent 28 days of cannabis abstinence supported by contingency management and weekly behavioral support sessions. Cannabis withdrawal symptoms were assessed at baseline and then weekly for four weeks using the Marijuana Withdrawal Checklist (MWC). Cannabis abstinence was verified using the Timeline Follow-Back, a self-report interview.

Results: Ten males and 10 females successfully abstained from cannabis for the 28-day study period; 2 males relapsed and were excluded from further analyses. Males and females did not significantly differ in age, race, or patterns of cannabis use (frequency, lifetime years of cannabis use, past 30-day cannabis use, CUD severity). Alcohol and nicotine/tobacco use were also similar between groups. A repeated measures ANOVA revealed a significant main effect of time on MWC [F (2.2, 38.7 = 4.49, p < 0.02]. Sex [F (1,18 = 1.33), p = 0.26]. and the time x sex interaction were not significant [F (2.2, 38.7) = 0.71, p = 0.51]. Cannabis withdrawal severity for both males and females followed the same trajectory: symptoms increased from baseline to day 7, peaked at day 7, and then dissipated to baseline levels or below baseline levels by day 28.

Conclusions: Our preliminary findings indicate that cannabis withdrawal severity follows the expected trajectory in both males and females, with no significant differences between the sexes. The lack of sex differences may be attributed to similar cannabis use patterns and comparable levels of CUD severity reported by both males and females in our sample. Our future studies will biochemically confirm self-reported cannabis abstinence and include larger samples. Nevertheless, cannabis withdrawal remains a viable and feasible treatment target for both sexes, particularly within the first seven days of cannabis cessation.

Keywords: Cannabis use disorder, Cannabis Withdrawal, Sex differences

Disclosure: Nothing to disclose.

Steven Nieto*, Suzanna Donato, Han Du, Wave-Ananda Baskerville, Kaitlin McManus, Lindsay Meredith, Molly Magill, Marcelo Lopez, Howard Becker, Lara Ray

University of California - Los Angeles, Los Angeles, California, United States

Background: Animal models have been used for preliminary testing of novel compounds for alcohol use disorder (AUD). However, the degree to which early efficacy in preclinical models reliably predicts clinical efficacy in human trials remains unclear.

Methods: To address this gap in the literature we employed a novel meta-analytic approach. We searched the literature for medications tested for AUD using both preclinical models (i.e., two-bottle choice and operant reinstatement) and randomized clinical trials (RCTs). For preclinical models, we computed medication effects on alcohol preference and consumption (k  =  77 studies, 14 medications) as well as operant reinstatement (k  = 18 studies, 8 medications). For RCTs, we computed medication effects on RCT endpoints including percent days abstinent, percent days heavy drinking, return to any drinking, and return to heavy drinking (k  = 139 studies, 19 medications). We applied random-effects meta-analysis to calculate a single effect size per medication within each paradigm. We used medication as the unit of analysis and applied the Williamson-York bivariate weighted least squares estimation to preserve the errors in both the independent and dependent variables.

Results: Medication effects on two-bottle choice alcohol preference (β  =  0.04, p  = 0.03) and reinstatement (β  =  0.20, p  = 0.05) were positively associated with medication effects on return to any drinking in human clinical trials but no associations were found on other clinical outcomes tested. Thus, medications that reduced alcohol preference and operant reinstatement in animals were associated with better clinical outcomes on the return to any drinking RCT endpoint.

Conclusions: This study demonstrates that preclinical endpoints of alcohol preference in two-bottle choice studies and operant reinstatement track medication effects from animal paradigms on select clinical trial outcomes, specifically return to any drinking. These results provide quantitative evidence regarding behavioral pharmacology methods that can be prioritized to screen novel compounds for AUD at the preclinical level of analysis. These preliminary results can help streamline the process of screening novel compounds for AUD.

Keywords: Alcohol Use Disorder - Treatment, Medication Development, Preclinical and clinical studies, meta-analysis, randomized clinical trial

Disclosure: Nothing to disclose.

Karl Bosque-Cordero*, Elizabeth Glover

Center for Alcohol Research in Epigenetics, University of Illinois at Chicago, Chicago, Illinois, United States

Background: Alcohol use disorder (AUD) is characterized by persistent excessive drinking despite adverse effects. Around 50% of those with AUD suffer from withdrawal symptoms, which frequently trigger relapse. The rostromedial tegmental nucleus (RMTg) signals aversion, and its function is implicated in withdrawal symptoms. The lateral habenula (LHb) provides dense glutamatergic input to the RMTg and parallels its function, including involvement in withdrawal symptoms. However, the exact mechanisms by which the LHb and RMTg regulate these symptoms remain unclear. Emerging data suggests that chronic ethanol exposure significantly alters astrocyte morphology in several brain regions. Given that astrocytes modulate neuronal activity at tripartite synapses it is likely that structural changes in these cells lead to alterations in synaptic transmission. The current experiment is designed to examine the impact of chronic ethanol exposure on astrocytes in the LHb and RMTg and their relationship to changes in neuronal physiology during acute withdrawal.

Methods: To assess the effects of withdrawal from chronic ethanol on RMTg and LHb astrocytes, Long-Evans rats underwent chronic intermittent ethanol (CIE) vapor exposure for 14 hours/day for 14 days. Controls were exposed to room air (AIR). Twelve hours after their final vapor session, a subset of rats were perfused with 4% paraformaldehyde, and their brains were extracted. Slices containing the RMTg were double labeled for GFAP (an astrocyte marker) and FoxP1 (a transcription factor enriched in the RMTg) using immunofluorescence. Slices containing the LHb were labeled for GFAP. Labeling was analyzed across both brain regions using Image J. Acute slices containing the LHb (220 µm) were made from a separate set of rats 24 hours after their final vapor exposure session in preparation for ex vivo whole-cell patch-clamp recordings. Current clamp recordings were made across increasing current steps (20 pA increments, 1 sec ea) from -80 pA to 200 pA, to analyze firing characteristics and intrinsic properties.

Results: Acute withdrawal from CIE exposure was associated with a significant increase in the expression of GFAP in the RMTg compared to AIR controls (*p < 0.05; nested t-test). This effect was primarily driven by greater GFAP expression at rostral levels of the RMTg in CIE- relative to AIR-exposed rats (*p < 0.05). GFAP expression was also greater in the LHb of CIE-exposed rats compared to AIR controls although this effect did not reach statistical significance. CIE exposure produced a significant reduction in the intrinsic excitability of LHb neurons (*p < 0.05) in the absence of significant changes in other intrinsic properties.

Conclusions: These data suggest that acute withdrawal increases astrocytic complexity in the RMTg, which may enhance astrocyte-neuron interactions and promote heightened neuronal activity. The unexpected decrease in LHb excitability during acute withdrawal may confer reduced LHb function (i.e., aversive signaling) following chronic ethanol exposure. Alternatively, it may represent a compensatory response to increased glutamatergic input to the LHb during acute withdrawal. Ongoing experiments are exploring how CIE-induced changes in astrocytic clearance of glutamate from the synapse may contribute to this effect. Altogether, these data shed light on the effects of chronic ethanol exposure on LHb and RMTg physiology and astrocyte morphology providing potential insight into the ways in which these brain regions regulate withdrawal symptoms. Supported by NIH grants R01 AA031003, P50 AA022538, and T32 AA02657.

Keywords: Alcohol, Lateral Habenula, astrocytes, RMTg, slice electrophysiology

Disclosure: Nothing to disclose.

Daniel Wood*, Evgeny Tsvetkov, Susana Comte-Walters, Colin Welsh, Rose Marie Akiki, Ethan Anderson, Rachel Penrod, Lalima Mada, Lauren Ball, Makoto Taniguchi, Christopher Cowan

Medical University of South Carolina, Charleston, South Carolina, United States

Background: Substance Use Disorder (SUD) is a relapsing disorder characterized by repeated seeking and use of psychoactive substances despite negative consequences to the individual. Epigenetic changes within the medium spiny neurons (MSNs) of the nucleus accumbens (NAc) are thought to underly the lasting propensity to relapse in SUD. Our previous studies have shown that the epigenetic enzyme histone deacetylase 5 (HDAC5) in NAc MSNs limits the formation of drug-cue associations in rodent models of drug seeking, leading to decreased relapse-like behavior. However, whether HDAC5 reduces drug seeking behavior through its own deacetylase activity or through the recruitment of co-repressors is unknown. Furthermore, these HDAC5-mediated behavioral changes are associated with a decrease in MSN intrinsic excitability and the downregulation of numerous genes associated with transmembrane cation transport. One of these genes, Scn4b, encodes a sodium channel auxiliary subunit that functions to increase action potential frequency in MSNs. We hypothesized that HDAC5 limits relapse-like behavior through reducing MSN intrinsic excitability via the deacetylase-dependent downregulation of Scn4b.

Methods: Long-Evans or Sprague-Dawley young adult male and female rats were used in all in vivo studies. Site-directed mutagenesis to create deacetylase-dead HDAC5 mutants and co-immunoprecipitation was used to assess HDAC5 binding partners. Targeted tandem mass spectrometry was used to examine HDAC5 post-translational modifications. Quantitative PCR was used to verify HDAC5’s regulation of Scn4b, and a virally expressed short-hairpin RNA targeting Scn4b (AAV2-shScn4b) was developed to directly reduce Scn4b transcript levels in NAc. AAV2-shScn4b was stereotactically infused into the NAc of male and female rats and allowed 3 weeks to expressed before ex vivo current-clamp recordings were taken from NAc MSNs. Using the same virally mediated Scn4b knockdown approach, male and female rats (n =) underwent two-weeks of intravenous cocaine self-administration followed by one week of forced abstinence, extinction training, and cue- and cocaine-primed reinstatement of drug-seeking.

Results: We found that NAc HDAC5 decreases MSN excitability and relapse-like behavior in an intrinsic deacetylase-dependent manner (n = 7-14 rats, One-Way ANOVA, F (3, 37) = 2.913, p < 0.05). Furthermore, this intrinsic deacetylase activity was required to reduce Scn4b expression (n = 5 rats, One-Way ANOVA, F (2, 12) = 8.367, p < 0.01). We also discovered that viral-mediated reduction of NAc Scn4b expression significantly decreased MSN intrinsic excitability (n = 34-37 cells/10 rats, Two-Way RM ANOVA, F (24, 1656) = 3.798, p < 0.0001), similar to the effects of HDAC5 overexpression. Further analysis of spike trains showed that underlying this reduction in intrinsic excitability was a robust increase in amplitude accommodation, reduced time to last spike, and reduced duration of the spike train, without altering rheobase or first spike amplitude. Lastly, viral-mediated Scn4b knockdown in NAc core had no effects on acquisition of cocaine self-administration, but it significantly decreased both context-associated cocaine seeking during early extinction training (n = 16-17 rats, Two-Way RM ANOVA, F (5, 155) = 4.173, p < 0.01) and cue-reinstated cocaine seeking n = 16-17 rats, unpaired t-test, p < 0.05). Although these studies were not powered to detect sex differences, we did not observe any clear sex differences in any of the experiments.

Conclusions: These results show that NAc SCN4B, an understudied voltage-gated sodium channel auxiliary subunit, is required to support drug-cue associations linked to relapse-like cocaine seeking, and that NAc HDAC5’s ability to suppress NAc MSN intrinsic excitability and limit drug-seeking behaviors is due, at least in part, to its deacetylase-dependent repression of NAc Scn4b. Together, these findings suggest that SCN4B, and MSN intrinsic excitability more broadly, may be important targets for the development of novel therapeutic strategies for the mitigation of relapse risk in SUD.

Keywords: cocaine self-administration and reinstatement, Epigenetics, Nucleus Accumbens, intrinsic excitability, substance use disorder

Disclosure: Nothing to disclose.

Che Liu, Francesca Filbey*

The University of Texas at Dallas, Dallas, Texas, United States

Background: There is high comorbidity/co-occurrence between cannabis use and depression. The relationship between cannabis use and depression appears to be complex and bidirectional. Heavy and/or chronic cannabis use may increase the risk of developing depression symptoms or exacerbate existing symptoms. On the other hand, some individuals with depression symptoms initiate and maintain cannabis use as a form of self-medication. Neuronal circuits can provide essential insights into the complex interplay between cannabis use and depression by understanding separate/individual and interactive influences of cannabis use and depression on brain networks. Previous studies have suggested that altered resting state functional connectivity (rsFC) within and between brain networks, such as default mode network (DMN), salience network (SN), frontal parietal network (FPN) and subcortical network (SCN), is respectively associated with cannabis use and depression, with mixed findings regarding the direction of these associations. However, studies focusing on interactive effects of cannabis and depression on network-level rsFC are sparse.

Methods: Resting state fMRI data from 395 adult participants (age range 18 to 55 years, 174 females) were analyzed to test the interaction between cannabis use and depression on rsFC. Cannabis use patterns were determined from self-reported history over the past 31 days or more. Beck’s Depression Inventory-II (BDI-II) was administered to the participants to score depressive severity. Resting-state fMRI data were preprocessed in CONN toolbox (v.22a) and analyzed in SPSS (version 28.0.1.1 (15)). Cortical regions of the DMN, SN, and FPN were defined based on default CONN network parcellation, while subcortical regions of the SCN were mapped using the Harvard-Oxford atlas. Within-network rsFC was estimated as the average of rsFC values between regions within the same network, while between-network rsFC was estimated as the average of rsFC values between regions in two different networks. General linear models were used to assess the main effects of times of cannabis use per week and BDI and their interaction on within- and between-network rsFC, controlling for age, sex, scanner (N = 4), years of education, age of cannabis use onset, number of drinks per week and number of cigarettes per week. Multiple comparison corrections were performed for 4 within-network rsFC and 6 between network rsFC respectively, with the false discovery rate (FDR) procedure. An FDR level of 0.05 was used to denote statistical significance.

Results: There was a significant association between within-SCN rsFC and BDI (B = 0.0017, F12, 382 = 8.507, p = 0.004, p-FDR = 0.015), indicating that higher within-SCN rsFC was linked to more depressive symptoms. The main effect of times of cannabis use per week on within-SCN rsFC was not significant (B = 0.0002, F12, 382 = 2.591, p = 0.108, p-FDR = 0.217). However, the interaction between BDI and times of cannabis use per week on the within-SCN rsFC showed a trend towards significance (B = -3.603E-05, F12, 382 = 5.600, p = 0.018, p-FDR = 0.074), suggesting that the association between within-SCN rsFC and BDI was more negative in individuals with more frequent cannabis use. The rsFC between SN and SCN was significantly associated with times of cannabis use per week across all participants (B = 0.0005, F12, 382 = 8.531, p = 0.004, p-FDR = 0.022), indicating that rsFC between these two networks was related to higher cannabis use frequency. The main effect of BDI on this rsFC was not significant (B = 0.0008, F12, 382 = 1.511, p = 0.220, p-FDR = 0.439). Nevertheless, the interaction between BDI and times of cannabis use per week was on the rsFC between SN and SCN (B = -3.325E-05, F12, 382 = 4.020, p = 0.046, p-FDR = 0.274). This trend in the interaction effect suggests that the association between rsFC between SN and SCN and times of cannabis use per week appears to be more negative in individuals with more depressive symptoms.

Conclusions: These findings indicate that within-SCN rsFC is associated with depression, while rsFC between SN and SCN is associated with cannabis use. More importantly, there is a trend suggesting that cannabis use and depression moderate each other’s relationship with network-level rsFC in the brain.

Keywords: Resting State Functional Connectivity, cannabis, Depression

Disclosure: Nothing to disclose.

Michael Nader*, Mia Allen

Wake Forest University School of Medicine, Winston-Salem, North Carolina, United States

Background: The goal of this research is to further enhance our understanding of the behavioral and neuropharmacological effects of cocaine in nonhuman primate models of cocaine use disorders (CUD). Estimates indicate that ~80% of people who use cocaine co-use nicotine and nearly 90% of people with CUD co-use alcohol. Despite this, very few preclinical studies incorporate either drug in animal models of CUD. When alcohol and cocaine are co-used, there is an active metabolite formed in the liver, cocaethylene, that has equal potency to cocaine and when studied under a concurrent schedule with food as the alternative reinforcer, appeared to have similar reinforcing strength to cocaine (Allen et al., 2023).

Methods: One method to better assess reinforcing strength is to manipulate the conditions and determine how resistant the subject is to changing their preference for drug over a non-drug alternative. For these studies, that environmental manipulation was delaying delivering of the drug reinforcer, whether that be cocaine alone, cocaine + nicotine, cocaethylene or cocaethylene + nicotine. The primary dependent variable for this study was an indifference point (IP value), or the delay needed to result in 50% choice between food and the delayed drug reinforcer. We hypothesized that under these conditions, cocaine alone would have the lowest reinforcing strength (i.e., the lowest IP value). In all experiments, monkeys responded on two manipulanda under a fixed-ratio (FR) 30 schedule of reinforcement; choice was between drug and 1.0 g banana-flavored food pellets.

Results: Cocaine and cocaethylene were equipotent, showing dose-dependent increases in drug choice. Adding nicotine shifted the cocaine and cocaethylene dose-response curves to the left, increasing reinforcing potency, although there was greater between subject variability with cocaethylene + nicotine choice. Delays on the drug reinforcer were studied at the lowest dose or dose combination in which drug was chosen > 80% of the trials. Although preliminary, as hypothesized, the lowest IP values were for cocaine alone, followed by cocaethylene. Cocaine + nicotine and cocaethylene + nicotine had equal reinforcing strength, and larger IP values, than cocaine and cocaethylene alone.

Conclusions: These findings suggest that preclinical studies designed to understand the neuropharmacology and, ultimately, development of novel treatments for CUD should include the co-use of alcohol and nicotine.

Keywords: cocaine self-administration, polysubstance use, Nonhuman Primate Models, Drug choice paradigm

Disclosure: Nothing to disclose.

E. Sherwood Brown*, Jayme Palka, Traci Holmes, Dante Durand, Ihsan Salloum

University of Texas Southwestern Medical Center, Dallas, Texas, United States

Background: Alcohol use disorder (AUD) is extremely common in people with bipolar disorder (BD). Over half of people with BD have lifetime AUD. When present, AUD is associated with increased rates of hospitalization, poor outcomes, violence towards self and others, and non-adherence to BD treatment. However, relatively few studies have examined the treatment of BD and co-occurring AUD. This two-site study examined the non-prescription supplements citicoline and pregnenolone in this patient population. Citicoline is a cholinergic compound that is involved in phospholipid metabolism. Prior research suggests it may decrease cocaine use in people with BD. Pregnenolone is a neurosteroid that is a precursor to other neurosteroids including allopregnanolone. Research suggests that pregnenolone may decrease alcohol self-administration in animal models and decrease depressive symptom severity in people with bipolar depression.

Methods: A 12-week, three-arm, randomized, double-blind, placebo-controlled, parallel-group, adjunctive trial of pregnenolone (500 mg/day) or citicoline (1000 mg/day) was conducted in 96 adults with a bipolar spectrum disorder (bipolar I, bipolar II, schizoaffective disorder, bipolar type) and AUD (≥ moderate severity, ≥ 28 drinks/week men and ≥ 21 drinks/week women). Enrollment was conducted at two sites (UT Southwestern in Dallas and the University of Miami). The primary outcome measure was drinks per drinking day based on the Timeline Followback (TLFB) method; secondary outcomes included drinks per day (over 7-day period), fraction drinking days, and fraction heavy drinking days. Although participants with a variety of mood states at baseline were enrolled, Young Mania Rating Scale (YMRS), Hamilton Rating Scale for Depression (HRSD) and Inventory of Depressive Symptomatology Self-Report (IDS-SR) scores were assessed in an exploratory fashion. Outcomes were evaluated using linear mixed-effects models.

Results: Participants were 44.8% women, 41.7% Black/African American, 13.5% Hispanic, with a mean age of 48.2 ± 9.7 years. At baseline, participants averaged 7.65 ± 4.99 drinks/day and 9.41 ± 5.27 drinks/drinking day. Of the 96 participants, 37 received pregnenolone, 30 citicoline, and 29 placebo. Treatment group main effects were significant for drinks/day (p = 0.009) and drinks/drinking day (p = 0.005), with both pregnenolone and citicoline groups having fewer drinks/day (mean difference = -0.93 [p = 0.012] and -0.83 [p = 0.047], respectively) and drinks/drinking day (mean difference = -1.29 [p = 0.011] and -1.32 [p = 0.015], respectively), relative to placebo. A significant treatment group effect also emerged for IDS-SR (p < 0.001), with pregnenolone demonstrating lower depressive symptoms in comparison to placebo (mean difference = -3.99 [p = 0.007]). For fraction drinking days, the treatment group × time interaction effect was significant (p < 0.001); relative to placebo, citicoline demonstrated lower fraction drinking days at weeks 6 (B = -0.211, p = 0.023) and 10 (B = -0.113, p = 0.045), while pregnenolone demonstrated lower fraction drinking days at weeks 8 (-0.164, p = 0.023), 10 (B = -0.082, p = 0.037), and 11 (B = -0.082, p = 0.037). Treatment group effects were non-significant for fraction heavy drinking days, HRSD, and YMRS. Both active treatments were well tolerated with 18.9% receiving pregnenolone, 30.0% receiving citicoline, compared to 37.9% receiving placebo, reporting an adverse event during the study.

Conclusions: Both pregnenolone and citicoline were associated with fewer drinks/day and drinks/drinking day than placebo. Relative to placebo, pregnenolone was also associated with lower self-reported, but not observer-rated, depressive symptom severity. These findings suggest that both supplements tested may be associated with a decrease in the quantity of alcohol consumed in people with BD and AUD. Both supplements were well tolerated. Larger trials are needed to confirm these findings.

Keywords: Neurosteroid, Bipolar Disorder, Alcohol Use Disorder - Treatment, Pregnenolone, Citicoline

Disclosure: Sage/Biogen: Advisory Board (Self)

Markku Lähteenvuo, Jari Tiihonen*, Anssi Solismaa, Antti Tanskanen, Ellenor Mittendorfer-Rutz, Heidi Taipale

Karolinska Institutet, Stockholm, Sweden

Background: Preliminary studies suggest that GLP-1-agonists, used to treat type-2 diabetes and obesity, may decrease alcohol consumption.

Methods: We tested whether the risk of hospitalization due to alcohol use disorder (AUD) is decreased during the use of semaglutide compared to non-use periods of GLP-1-agonists within the same individual, by using Swedish nationwide observational cohort data from January 2006 to December 2023. Our population-based cohort was identified from registers of inpatient care, specialized outpatient care, sickness absence, and disability pension. The cohort included all residents aged 16-64 years with a diagnosis of AUD. Primary exposure was use of individual GLP-1-agonists, and secondary exposure medications with indication for AUD (disulfiram, acamprosate, naltrexone, nalmefene). The primary outcome was adjusted Hazard Ratio (aHR) for AUD hospitalization analyzed in Cox regression within-individual model. Secondary outcomes were any substance use related hospitalization, somatic hospitalization and suicide attempt. Hazard Ratios were adjusted for time-varying use of psychotropic medications (antipsychotics N05A, antidepressants N06A, mood stabilizers including carbamazepine N03AF01, valproic acid N03AG01, lamotrigine N03AX09 and lithium N05AN01), benzodiazepines and related drugs (N05BA, N05CD, N05CF), ADHD medications (N06BA) and use of other antidiabetic drugs than GLP-1 agonists (A10 excluding A10BJ), temporal order of GLP-1 medication and time since cohort entry).

Results: The cohort included 227,866 individuals with AUD, of whom N = 144714 (63.5%) were male, with a mean age of 40.0 years (SD 15.7). Median (IQR) follow-up time was 8.8 (4.0–13.3) years. A total of 58.5% of the cohort experienced AUD hospitalization. Semaglutide (4321 users) was associated with the lowest risk of AUD hospitalization (aHR 0.64, 95% CI 0.50–0.83, p = 0.0007 compared with non-use of GLP-1 agonists). Use of any AUD medication was associated with a modestly decreased risk (aHR 0.98, 95% CI 0.96–1.00) compared with non-use of AUD medications. Concerning specific AUD medications, naltrexone use was associated with the lowest risk of AUD hospitalization (aHR 0.86, 95% CI 0.83–0.89). Semaglutide use was also associated with decreased risk of any substance use related hospitalizations (aHR 0.68, 95% CI 0.54–0.85) and somatic hospitalizations (aHR 0.78, 95% CI 0.68–0.90), and not associated with suicide attempts (aHR 0.55, 95% CI 0.23–1.30).

Conclusions: Among patients with AUD and comorbid obesity/type 2 diabetes, the use of semaglutide is associated with a substantially decreased risk of hospitalization due to AUD. This risk was lower than that of officially approved AUD medications. Semaglutide may be effective in treatment of AUD, and clinical trials are urgently needed to confirm these findings.

Keywords: Semaglutide, GLP-1 receptor agonist, Alcohol and substance use disorders

Disclosure: Nothing to disclose.

Fernando Cano, Kierra Smith, Sonya Adler, Asha Jotwani, Harita Udayashankar, Kiana Mohammadi, Vivian Dang, Christopher Denning, Alex Reed, Laura Huerta Sanchez, Diego Gonzalez, Karen Szumlinski*

University of California, Santa Barbara, Santa Barbara, California, United States

Background: The similar temporal profile of incubated food- vs. drug-seeking posits that these behavioral phenomena may involve common biochemical adaptations within neural circuits governing motivated behavior. Incubated cocaine-craving is associated with elevated mTOR activity within the prelimbic (PL) subregion of the prefrontal cortex and pretreatment with Everolimus, an FDA-approved mTOR inhibitor, dose-dependently blocks incubated cocaine-seeking and reverses incubation-related changes in mTOR activation within the PL. Herein, we conducted a series of studies to examine for the generalization of Everolimus’ effects on the incubation of sucrose- and methamphetamine (MA)-craving.

Methods: Male and female Sprague Dawley rats were trained to lever-press for either banana-flavored sucrose pellets or intravenous MA for 6 h/ day for 10 days. At 1 or 30 days withdrawal (WD), immunoblotting was conducted on the PL and infralimbic (IL) subregions to examine for indices of mTOR activity. In a follow-up study, male and female rats were gavage infused with either vehicle (VEH) or the maximally effective Everolimus dose from our published cocaine study (1.0 mg/kg), 30 min prior to a 2-h test for cue-reinforced responding.

Results: In all experiments, our self-administration procedures elicited an incubation of both sucrose- and MA-craving (Withdrawal effects: p’s < 0.01), the magnitude of which did not differ between male and female rats (Sex effects: p’s > 0.75). Incubated sucrose-craving was associated with elevated mTOR activity within the IL subregion (Group effect: p < 0.025), while no incubation-related change in mTOR activity was detected in either the PL or the IL in MA-experienced animals (Group effect: p > 0.50). In neither case did gavage infusion of 1.0 mg/kg Everolimus alter the magnitude of incubated craving (Pretreatment effects and interactions: p’s > 0.09).

Conclusions: The present results indicate that incubated cocaine-craving can be distinguished from both incubated sucrose- and MA-craving on both a biomolecular level (mTOR activity within prefrontal cortex subregions) and on a behavioral pharmacological level by differences in the efficacy of Everolimus to reduce craving. While our results for incubated sucrose-craving argue no overt off-target effects on conditioned motivation for a non-drug reinforcer, they do not support a generalization of “anti-craving” effects across psychomotor stimulant drugs of relevance to their potential repurposing for substance use disorder treatment.

Keywords: mTOR, cue-induced craving, incubation of methamphetamine craving, Medial Prefrontal Cortex (mPFC), incubation of sucrose craving

Disclosure: Nothing to disclose.

Damla Aksen*, Marc Potenza, Shashwath Meda, Sarah Lichenstein, Godfrey Pearlson

Yale School of Medicine, New Haven, Connecticut, United States

Background: Impulsivity is a complex and multifaceted construct associated with addictions, but its assessment and conceptualization often lack consistency, and sex differences further complicate effective targeting in addiction treatment. This investigation aims to identify the neural substrates of distinct domains of impulsivity and explore their relationships with alcohol use and stress experiences in both women and men.

Methods: We utilized a whole-brain machine learning strategy, connectome-based predictive modelling (CPM), to investigate brain networks linked to four composite impulsivity-related domains previously identified by principal components analysis in the NIAAA-funded Brain and Alcohol Research in College Students (BARCS) study data: (1) impulsive action; (2) approach/appetitive motivation; (3) impulsivity/compulsivity; and (4) behavioral inhibition/punishment sensitivity (BIPS). CPM (5-fold cross-validation, 100 repeats, and permutation testing) was applied to monetary incentive delay task (MIDT) fMRI data from 287 undergraduate students and neural features identified in successful models were examined in relation to alcohol and stress experiences across women and men.

Results: The CPM model predicting BIPS was significant (r = 0.24, p = 0.001). Results indicated that higher BIPS was associated with stronger connectivity between default mode, motor/sensory, and cerebellar networks, and reduced connectivity between the medial frontal, frontoparietal, default mode, and motor/sensory networks. BIPS network strength differed significantly between men and women (t(285) = 8.26, p < 0.001) and was negatively associated with alcohol use (p < 0.05) in men only. Conversely, BIPS network strength was positively associated with stressful life events (p < 0.05) in women only.

Conclusions: The identification of a neural network predicting BIPS in young adults has potential implications for how impulsive behaviors may be targeted. Future work should consider sex differences and whether targeted interventions (e.g., neuromodulation) may help individuals seeking treatment for addictive behaviors motivated by behavioral disinhibition and punishment sensitivity.

Keywords: impulsivity, Binge alcohol use, Stress, Machine learning, Sex differences

Disclosure: Nothing to disclose.

Carlos Blanco*, Leah Durbak, Shari Ling, Wilson Compton, Jan Losby, Christopher Jones

National Institute on Drug Abuse, Washington, District of Columbia, United States

Background: Changes in insurance payment coverage for medications to treat opioid use disorder (MOUD) may be an essential tool in reducing the number of overdose deaths. We sought to examine changes in retention in MOUD and receipt of behavioral health services after implementation of a new bundled payment Medicare benefit for OUD treatment provided by opioid treatment programs (OTP)s.

Methods: We used longitudinal cohort study from January 2020-December 2020 using CMS Medicare data to examine Medicare Fee-For-Service Beneficiaries aged ≥18 years with who received treatment at a Medicare-enrolled OTP based on at least one service claim containing an OTP-specific Healthcare Common Procedure Coding System (HCPCS) code at any time during 2020. The main outcome measures were receipt of MOUD and behavioral health telehealth services before and after the implementation of the new bundled payment benefit.

Results: The cohort consisted of 20,160 Medicare beneficiaries who received treatment at an OTP in 2020. The majority were male (57.1%), aged 45-64 years (47.7%), white (70.1%), living in metropolitan areas (52.2%), had both Medicare and Medicaid (86.8%), and were Medicare-eligible due to disability (71.6%). Most beneficiaries (96.4%) received methadone from OTPs. Mean time in treatment increased from 35.5 days in the 6 months preceding the first OTP service benefit to 155.5 in the following 6 months (p < 0.001). < 1 in 10 individuals (7.4%) received behavioral health telehealth services in the six months preceding the index visit compared to 18.2% in the six months after the index visit (p < 0.001)

Conclusions: Implementation of the new bundled payment Medicare policy was associated with significant increases in retention in OTP treatment. Policy changes, including changes in payment policies, may be an important tool to reduce the number of overdose deaths.

Keywords: methadone, opioid use disorder, Medicare

Disclosure: Nothing to disclose.

David Lovinger*, Yolanda Mateo, Andrew Kesner, Ilse Alonso-Vazquez

National Institute on Alcohol Abuse and Alcoholism, Rockville, Maryland, United States

Background: Disrupted sleep is an important symptom of Alcohol Use Disorder (AUD). Alcohol is sometimes used as a sleep aid, but ultimately alters the balance of sleep states. Tolerance to alcohol effects can also lead to escalation of ingestion as a sleep aid. Acute ethanol (EtOH) exposure enhances non-rapid-eye-movement (NREM) sleep in mice, but less is known about the neurobiological underpinnings of changes in sleep following chronic EtOH exposure. The neuromodulator dopamine is implicated in control of arousal and changes in sleep induced by misused substances, including alchool.

Methods: To elucidate these mechanisms, we examined changes in sleep state structure following 4 weeks of EtOH exposure and subsequent withdrawal using two well-established models in C57Bl6J mice, chronic intermittent EtOH vapor exposure (CIE) and drinking in the dark (DID). Polysomnography was used to measure sleep architecture 6 days following withdrawal after the two exposure paradigms. We also examined dopamine release in brain slices from dorsomedial striatum (DMS) following the same DID and withdrawal paradigm using fast-scan cyclic voltammetry.

Results: Surprisingly, disrupted sleep, especially decreased NREM sleep, was stronger in the DID model. Disruption was greatest in the light phase when mice normally sleep the most. The greater effect in the DID model may result from the circadian pattern of alcohol intake, with binge-like drinking in the dark phase. We did not observe changes in the amplitude or decay kinetics of electrical stimulation-induced dopamine release in DMS. However, acute EtOH inhibition of dopamine release was stronger following DID + withdrawal. We are currently exploring potential mechanisms underlying this increased inhibition as well as changes in dopamine in vivo during DID using the dLight genetically encoded fluorescent sensor.

Conclusions: These findings indicate that DID produces sleep disruption with altered NREM sleep, a pattern also observed following chronic, heavy alcohol use in humans. The alterations in acute ethanol inhibition of dopamine release may reflect a mechanism that compensates for increases in dopamine observed in response to initial EtOH exposure. Ultimately, we hope to uncover mechanisms involved in sleep disruption following chronic EtOH drinking/withdrawal that may lead to better therapies for AUD.

Keywords: addiction, alcohol withdrawal, Basal Ganglia, Dopamine

Disclosure: Nothing to disclose.

Gabriella Silva, Amber Kaplan, Celeste Park, Joseph Picone, Richard Kim, Natalie Truby, Rachael Neve, Xiaohong Cui, Peter Hamilton*

Virginia Commonwealth University, Medical College of Virginia, Richmond, Virginia, United States

Background: We have previously discovered that Zfp189, which encodes a Krüppel-associated box zinc finger protein (KZFP) transcription factor (TF), differentially accumulates in nucleus accumbens (NAc) Drd1+ and Drd2+ medium spiny neurons (MSNs) over the course of cocaine exposure and is causal in producing MSN functional and behavioral changes to cocaine. Here, we aimed to illuminate the brain cell-type specific molecular mechanisms through which this KZFP TF produces cocaine-related brain changes, with emphasis on investigating transposable elements (TEs), since KZFPs like ZFP189 are known regulators of TEs. To investigate this, we annotated TEs in existing single nuclei RNA-sequencing (snRNAseq) datasets of rodents that were exposed to either acute or repeated intraperitoneal (IP) cocaine. Additionally, we synthesized novel synthetic ZFP189 TFs, each capable of exerting distinct forms of transcriptional control at in vivo ZFP189 target genes, including TEs. These ZFP189 TFs were conditionally virally delivered to the NAc of Drd1- and Drd2-Cre+ mice and the subsequent effect on MSN dendritic spine morphology, cocaine-induced locomotion behaviors, and cell-type specific transcriptional response were characterized.

Methods: Three ZFP189 variants were synthesized: ZFP189WT which is identical to endogenous ZFP189 and contains a repressive KRAB domain; ZFP189VPR in which the KRAB domain is replaced with the transcriptional activator VP64-p65-Rta (VPR); and ZFP189NFD in which any transcriptional regulatory domain is removed. Each variant was sub-cloned into Cre-dependent (lox-STOP-lox (LSL) p1005) expression vectors and gene-targeted regulation was characterized in combination with a ZFP189 response element (RE) luciferase assay in Neuro2a cells (n = 3 in technical triplicate). Expression vectors were then packaged in herpes simplex virus (HSV) to enable brain-region targeted expression. Both male and female Drd1- and Drd2-Cre+ mice, aged 8-12 weeks, were utilized. HSV-LSL-ZFP189WT, -ZFP189VPR, or -ZFP189NFD were delivered by stereotaxic surgery bi-laterally to NAc (10°; +1.62 AP; ±1.5 ML; -4.4 DV, 1uL). After recovery, MSN dendritic spine morphology, cocaine-induced locomotion, and snRNAseq was performed. For locomotion, 10 mg/kg IP cocaine was delivered for seven sequential days and distance traveled was quantified in a 30-minute session. For snRNAseq, we extracted nuclei from virally infected NAc tissues and performed snRNAseq according to the 10x Genomics Chromium 3’ gene expression pipeline. We sequenced an average of 19,000 nuclei per treatment at a depth of > 50K reads/nuclei. Publicly available snRNAseq datasets were downloaded from NCBI GEO (GSE137763, GSE222418). TEs were identified using SoloTE. Differentially expressed genes (DEGs) by cell-cluster were generated against the saline or ZFP189NFD control conditions. DEGs with an FDR adjusted P-value < 0.05 and an absolute Log2(FoldChange) > 0.25 were considered significant. For all experiments, sample sizes are 10-15 mice per group and statistical analyses were performed with one- or two-way ANOVAs, including a post-test comparing the test groups (HSV-ZFP189WT, -ZFP189VPR) versus the control group (HSV-ZFP189NFD).

Results: By co-transfecting the ZFP189 RE luciferase plasmid and iterations of our ZFP189 variants, we observe, relative to GFP control: ZFP189VPR induces robust targeted gene activation (P < 0.0001), ZFP189WT induces gene-targeted repression (P < 0.05), and ZFP189NFD exerts no regulatory control (P > 0.5). By conditionally delivering our ZFP189 TFs to NAc MSNs, we observe that when ZFP189VPR was delivered to the Drd1+ MSNs of the NAc, animals demonstrated an immediate and persistent increase in cocaine-induced locomotion and an increase in mature mushroom spine morphology on Drd1+ MSNs relative to all other viral treatment conditions (P < 0.0001). Conversely, when ZFP189WT was delivered to the Drd2+ MSNs, animals demonstrated an increase in cocaine-induced locomotion and mature mushroom spine morphology on Drd2+ MSNs relative to other viral conditions (P < 0.0001). In the public snRNAseq data, TEs were released in NAc cell types over the course of cocaine exposure, most prominently in Drd1+ and Drd2+ MSNs. In our snRNAseq data, ZFP189WT permitted, and ZFP189VPR limited, gene expression in Drd1+ and Drd2+ MSNs, with particular impact on gene products involved in immune processes.

Conclusions: We discover that cocaine exposure releases TEs within NAc MSNs – a molecular phenomenon that rapidly and transiently occurs within Drd1+ MSNs yet accumulates within Drd2+ MSNs over the course of cocaine use. By creating synthetic ZFP189 TFs of opposing transcriptional control, we are able to intentionally either release or repress these TEs within the rodent brain. By conditionally delivering these ZFP189 TFs to the Drd1+ or Drd2+ MSNs of mouse NAc, we discover that opposite forms of KZFP-mediated transcriptional control in these opposing cell-types converge on producing behavioral and cell morphological adaptations to cocaine. We discover that normal KZFP function is critical to produce cocaine-induced transcription within NAc MSNs, and by inverting the normal function of a KZFP with ZFP189VPR, we are able to impede cocaine-induced gene expression in MSNs. This research demonstrates that KZFP-mediated transcriptional repression of TEs within NAc MSNs is a causal molecular step in enabling gene expression and subsequent cellular and behavioral responses to cocaine use.

Keywords: Cocaine, Transposable element, Cell-type specific transcription, Nucleus Accumbens

Disclosure: Nothing to disclose.

Lori Knackstedt*, Barry Setlow, Linda Cottler

University of Florida, Gainesville, Florida, United States

Background: Rodent models of cocaine use disorder have identified dozens of compounds that reduce relapse to cocaine-seeking after periods of drug abstinence, including compounds targeting glutamate homeostasis like ceftriaxone. Such compounds generally fail to translate into the clinic, however, possibly due to inadequate modeling of real-world patterns of cocaine use, including polysubstance use, in the lab. Our prior work has found that approximately 75% of people who use cocaine also use alcohol, both simultaneously and sequentially, with cocaine use reported on average 8 days per month. Here we developed a rat model in which intermittent (non-daily) intravenous self-administration (IVSA) of cocaine was employed either alone, or together with alcohol, in order to uncover the effects of these patterns on incubation of cocaine seeking (increased cued seeking on Day 21 compared to Day 1 of abstinence) and the ability of ceftriaxone to reduce such seeking.

Methods: Male and female rats underwent IVSA of cocaine alone (n = 38) or a cocaine+alcohol solution (n = 17) for 2 hr/day for 5 consecutive days, then 6 hr/day for 10 sessions. These 10 sessions occurred either daily (n = 10), or intermittently separated by 2-3 days (n = 35). Self-administration was followed by 1-hr tests for cued cocaine-seeking (relapse) at 1 and 21 days of abstinence. A subset of rats in the intermittent cocaine-only and cocaine+alcohol conditions (n = 18) received ceftriaxone (200 mg/kg IP) for 6 days prior to the Day 21 relapse test while the remainder received vehicle (saline IP). For all IVSA conditions, each infusion delivered 0.5 mg/kg cocaine; in the cocaine+alcohol condition, each infusion contained 0.5 mg/kg cocaine + 25 mg/kg alcohol.

Results: Following intermittent cocaine and cocaine+alcohol IVSA, incubation of cocaine seeking was not observed, due to robust seeking on Day 1 of abstinence. In addition, following intermittent IVSA, ceftriaxone attenuated cocaine-seeking on Day 21 of abstinence in the cocaine-only condition but not in the cocaine+alcohol condition.

Conclusions: Simultaneous cocaine+alcohol self-administration prevents the ability of ceftriaxone to attenuate cocaine seeking, in agreement with our prior results in a rat model of sequential daily cocaine+alcohol self-administration. While the intermittent self-administration of cocaine does not result in incubation of cocaine-seeking, it does not alter the ability of ceftriaxone to attenuate cued relapse. These findings highlight the importance of evaluating real-world patterns of drug use for screening potential interventions to reduce cocaine seeking.

Keywords: incubation of cocaine craving, polysubstance, glutamate

Disclosure: Nothing to disclose.

Quilla Flanagan-Burt, Byung Kwon Moon, Kerry Ressler, Junghyup Suh*

Harvard Medical School McLean Hospital, Belmont, Massachusetts, United States

Background: Alcohol seeking during abstinence is mediated in part by strong associations between the pharmacological effects of alcohol and the environment within which alcohol is administered. The amygdala, particularly the basolateral amygdala (BLA), is a key neural substrate of environmental context-reward associations since it is involved in associative learning and memory recall. However, we still lack a clear understanding of how the activity of molecularly distinct BLA neurons is affected by alcohol and encodes information that drives context-dependent, alcohol-related behaviors. Recent studies in mice have indicated that distinct subpopulations of neurons within the amygdala are differentially responsible for the activation and inhibition of fear memory. In addition, divergent ensemble activity from these subpopulations seems to mediate positive or negative valence coding. Specifically, we have found that a specific excitatory neuronal population in the BLA, marked by Thy1 expression (Thy1 + ), serve as ‘Fear-Off’ neurons. They also heavily project to the nucleus accumbens (NAcc), a core structure for reward-based learning and substance addiction, instead of the central amygdala, suggesting that these neurons directly inhibit fear and may support appetitive behavior. Understanding the roles of this molecularly identified population in alcohol-related behaviors is central to our study.

Methods: First, to determine whether Thy1+ neurons are involved in associative Pavlovian conditioning with alcohol, we employed a conditioned place preference (CPP) paradigm with systemic injection of alcohol (EtOH, 2g/kg BW) in male mice. Then, we have mapped EtOH-CPP-induced neuronal activity changes by quantifying the number of c-Fos protein-expressing (c-Fos + ) neurons in the BLA of Thy1-eYFP mice. Next, to determine the spatial activity of BLA Thy1+ neurons, we expressed genetically encoded calcium indicator, GCaMP6 in BLA Thy1+ neurons and subsequently performed fiber photometry while subjecting mice to CPP. Second, to investigate the roles of BLA Thy1+ neurons and their projections during conditioning or recall phase of CPP, we performed inhibitory optogenetic manipulation in Thy1+ neurons using Thy1-Cre driver mice and Cre-dependent halorhodopsin (eNpHR) or eYFP. AAV encoding eNpHR was injected to the BLA, but optic fibers were bilaterally implanted above the BLA, NAcc or prefrontal cortex (PFC). Third, to determine how the activity of BLA Thy1+ neurons contributes to pose dynamics and behavioral structures, we subjected Thy1-Cre mice injected with Cre-dependent inhibitory chemogenetic (DREADD) virus to instrumental conditioning paradigm and analyzed changes in mouse behaviors at the sub-second time scale using Keypoint-MoSeq, a machine learning algorithm-implemented platform.

Results: First, there was no difference in the number of total c-Fos+ neurons and c-Fos+ neurons in Thy1+ neurons in non-conditioned group and EtOH-conditioned group on the first day of conditioning. However, total c-Fos+ neurons and c-Fos+ neurons in Thy1+ neurons were significantly greater in EtOH-conditioned group than non-conditioned group on the last day of conditioning. In addition, In vivo calcium imaging indicated that the temporal activity profile of these neurons was also correlated with alcohol seeking behavior in response to environmental cues. Second, optogenetic inhibition of BLA Thy1+ neuronal activity disrupted both the formation and recall of alcohol conditioned place preference. Furthermore, selective axonal inhibition of BLA-Thy1+ neurons revealed that the activity of their NAcc and PFC projections were differentially necessary for alcohol cue association vs. recall, respectively. Third, the chemogenetic inhibition of the BLA Thy1+ neurons resulted in decreased conditioned responding for alcohol consumption and did not alter general locomotor activity or anxiety level. However, the Keypoint-MoSeq analysis revealed that behavioral syllables in approach behaviors to the reward port were altered by the chemogenetic inhibition.

Conclusions: Given the amygdala’s roles in fear, anxiety and substance abuse, the findings suggest that alcohol exposure alters the firing activity of subpopulations of amygdala neurons. Interactions between the BLA and other brain areas via its projections are differentially involved in the formation of memory associating alcohol experience and environmental cues and recall of the memory. In addition, the BLA activity contributes to reward-related approach behaviors.

Keywords: basolateral amygdala, Alcohol-seeking behavior, calcium imaging, circuit optogenetics, machine learning classification

Disclosure: Nothing to disclose.

Rajitha Narreddy, Sehwa Jung, Hannah Nonoguchi, Angelica Vandekerkhoff, Michael Jin, Chitra Mandyam*

University of California, San Diego, San Diego, California, United States

Background: The prefrontal cortex (PFC) is important for the development of alcohol addiction. Neuronal activity in this region is regulated by immune responses, and inflammatory responses are assisted by disruption of the blood-brain barrier (BBB). The detrimental effects of chronic alcohol consumption, abstinence and relapse to alcohol drinking on inflammatory markers and BBB integrity proteins in the PFC have been minimally explored. Moreover, how and if these effects are different or divergent in mild to moderate alcohol consumption or nondependent conditions is not known. Here we seek to answer these questions.

Methods: Adult female and male rats were made ethanol dependent by chronic intermittent ethanol vapor (CIE) and ethanol drinking (ED) procedure or only experienced ED as nondependent drinking. CIE rats were euthanized during acute withdrawal (6-8h after CIE), protracted abstinence (2 weeks) or after relapse session and nondependent rats were euthanized at matched time points. Plasma isolated from trunk blood were analyzed for estrogen, progesterone, cytokines and chemokines using a 9-plex panel from Meso Scale Discovery and brain tissue homogenate of the PFC were analyzed for BBB disruption via Western blotting and microglial activity via histology. This allowed us to associate the levels of gonadal hormones, individual cytokines to neuroimmune responses in the PFC.

Results: CIE increased ED, with rats having higher ethanol consumption during CIE and relapse to ethanol drinking sessions compared with pre-CIE sessions. ED was not altered in nondependent rats. Under both conditions, females consumed more ethanol than males. Plasma cytokine analysis revealed increases in anti-inflammatory cytokines in nondependent rats whereas increases in pro-inflammatory cytokines and estrogen in CIE rats, with significant changes observed during acute withdrawal and abstinence. Western blotting of PFC tissue indicated enhanced expression of adherens junction protein, Cdh5 in CIE rats, an effect that was significant during acute withdrawal and abstinence. 3D structural analysis showed an activated state of Iba-1 labeled microglial cells in the PFC in CIE rats during abstinence and this was not observed in nondependent rats.

Conclusions: These results reveal significant effects of ethanol dependence in peripheral estrogen and pro-inflammatory cytokine levels, which were associated with BBB disruption and microglial activation in the PFC during abstinence, and emphasize their possible role in relapse to ethanol seeking in ethanol dependent rats.

Keywords: Alcohol Use Disorder - Treatment, activated microglia, Blood-Brain-Barrier, Pro-inflammatory cytokines

Disclosure: Nothing to disclose.

Eleanor Towers*, Kyle Hsu, Shaniece Fraser, Wendy Lynch

University of Virginia, Charlottesville, Virginia, United States

Background: Drug-craving and relapse are major challenges in treating opioid use disorder (OUD). A better understanding of how these risk factors change over withdrawal, particularly with fentanyl, the major culprit in the opioid epidemic, could help advance treatment strategies and decrease fentanyl-related overdose deaths. Additional work is also needed to address sex differences considering that males have been the primary focus in studies on craving/relapse.

Methods: Here, we used a rat model of cue-induced relapse with fentanyl to evaluate sex differences in craving during early (day 0 and 1) versus protracted withdrawal (day 14), when levels of craving for other opioids are known to be low versus high, respectively. We also determined the efficacy of R-ketamine, a novel potential treatment for OUD, at reducing cue-induced relapse. Both time-course and R-ketamine effects were examined following extended, intermittent-access fentanyl self-administration (24-hr/day, 2, 5-min trials/hr, 0.25 ug/kg/infusion).

Results: Cue-induced craving was similarly high in males and females across all withdrawal timepoints; fentanyl-craving was not further enhanced by additional cycles of fentanyl self-administration and withdrawal and not diminished by repeated relapse testing. Despite this robust relapse phenotype, acute R-ketamine treatment (0 or 20 mg/kg) on withdrawal day 0 provided long-lasting protection against cue-induced fentanyl-seeking as assessed 14 days after treatment; however, this effect was only apparent in males.

Conclusions: Thus, fentanyl-craving did not incubate over withdrawal but was similarly high during both early and protracted withdrawal in males and females, and R-ketamine persistently protected against cue-induced relapse in males, but not females.

Keywords: sex differences, fentanyl, opioid use disorder, R-ketamine

Disclosure: Nothing to disclose.

Nicholas Russell, Daniela Carrizales, Marco Bortolato, Regina Mangieri*

The University of Texas At Austin, Austin, Texas, United States

Background: Alcohol and substance use disorders are often associated with pathological aggression, a recurrent pattern of angry, disruptive, and violent behavior. These conditions share multiple vulnerability factors, such as early-life trauma and alterations in monoamine transmitter systems, but the neurobiological foundations of comorbid aggression and alcohol misuse are not well understood. We previously developed a mouse model of pathological aggression in which mice with low-activity of the enzyme monoamine oxidase A (“MAOA-low”), when subjected to an early-life stress paradigm, exhibit a hyper-aggressive phenotype in adulthood. The goal of the present work was to examine alcohol consumption in this model to validate whether it may be used to study comorbid excessive aggression and alcohol use.

Methods: Genetically engineered, MAOA-low and wild-type MAOA mice underwent early-life stress (ELS) exposure (daily maternal separation of 1-3 hours and a flank needle poke, to mimic neglect and abuse) on postnatal days 1-7. In early adulthood, ELS-treated and control non-stress (NS) mice were tested for alcohol consumption in the “drinking-in-the-dark” model of binge drinking, in which access to drinking solutions was given for 4 hours/day during the dark (active) phase of the light cycle. Testing consisted of 4 sessions/week for 3 weeks, with 20% alcohol during the first week, and 9 mM saccharin and water during weeks 2 and 3; intake was measured at 2- and 4-hour timepoints of each session. Tail blood was collected at the end of the final alcohol session to measure blood alcohol concentration (BAC).

Results: For males (n = 10/group), there were no Genotype or Stress main or interaction effects on any measure of alcohol, saccharin, or water intake. However, despite consuming similar doses of alcohol during the final session, the mean BAC of ELS-MAOA-low was significantly > NS-MAOA-low males (p = 0.04). The relationship between intake and BAC could not be modeled using a single curve to fit all groups, but groups with the same stress condition did share best-fit curves. Thus, there may be a subtle effect of ELS overall to influence either microstructure of drinking or alcohol pharmacokinetics in males.

In females (n = 9-13/group), both Genotype and Stress were significant factors impacting drinking behaviors. There was a significant main effect of Stress (p = 0.004) on average alcohol intake during the first 2 hours of the session, and relative intake during this period (2hr/total intake, “front-loading”) was greater in ELS- than NS-MAOA-low females (p = 0.01). Total alcohol intake per session, averaged across all 4 days, tended to be greater for MAOA-low (p = 0.05, main effect of Genotype) and ELS-treated groups (p = 0.09, main effect of Stress), and there were clear main effects during the final day of testing (p = 0.002, Genotype, 0.03, Stress), with ELS-MAOA-low mice drinking more than every other group. The increased alcohol consumption by ELS-MAOA-low was not reflected in a significantly higher BAC at the end of the session, which may be a consequence of this group exhibiting more front-loading. Average saccharin and water intake were lower in ELS-treated groups (main effect of Stress, p = 0.03 for both fluids). Although there was a main effect of MAOA-low genotype overall to increase water intake (p = 0.009), this appeared to be driven by the NS-MAOA-low group.

Conclusions: Our findings suggest that male ELS-MAOA-low mice may self-administer to achieve higher BACs than controls, whereas female ELS-MAOA-low mice exhibited front-loading behavior and also self-administered a greater total dose of alcohol than all other groups by the final day of testing. In consideration with previous work, our findings indicate that ELS-treatment of MAOA-low mice increases propensity for both aggression and maladaptive alcohol use in both sexes. Future studies will use this model to probe the neurobiological links between aggression and alcohol use.

Keywords: aggression, Alcohol intake, Early life stress, Animal Models

Disclosure: Nothing to disclose.

Ginny Park, Alexandra Malena, Nicholas Glodosky, Savannah Lightfoot, Samantha Baglot, Zachary Fisher, Carrie Cuttler, Matthew Hill, Ryan McLaughlin*

Washington State University, Pullman, Washington, United States

Background: Approximately 9% of first-time cannabis users will become dependent on cannabis, yet there are no FDA-approved pharmacotherapies for managing cannabis use disorder (CUD). This is in part due to flawed diagnostic nosology resulting from a lack of understanding of the mechanisms that give rise to CUD, as well as a conspicuous lack of translationally relevant animal models of cannabis use. To address these gaps, we have developed a model of cannabis self-administration that delivers vaporized cannabis extract in a response-contingent manner via the pulmonary route of administration that is most common in humans. We used this model to identify behavioral and biological predictors of motivation to self-administer vaporized cannabis in rats.

Methods: We conducted an extensive battery of behavioral assays in female and male Long Evans rats (N = 48) prior to initiation of cannabis self-administration training and characterized endophenotypes using endpoints that correspond to the behavioral dimensions of the NIMH Research Domain Criteria (RDoC). We then used a series of linear regression analyses to determine whether behavioral and physiological parameters in the five RDoC dimensions (positive and negative valence systems, cognition, social processes, and arousal/regulatory systems) significantly predicted the number of cannabis vapor deliveries earned during a 3-hr progressive ratio test after four weeks of cannabis self-administration.

Results: The Arousal/Regulatory Systems model was significant, accounting for 59.3% of the variance in cannabis self-administration. Specifically, higher concentrations of basal corticosterone (CORT) predicted higher rates of cannabis vapor self-administration (p = 0.005), while lower concentrations of plasma anandamide (AEA) content predicted higher rates of self-administration (p = 0.008). The Cognition model was significant, accounting for 17.9% of the variance in cannabis self-administration. Specifically, better visual cue discrimination and poorer set shifting performance each predicted higher rates of cannabis self-administration (p’s = 0.015 and 0.030, respectively). The Positive Valence model was also significant, accounting for 21.1% of the variance in cannabis self-administration, with greater motivation for sucrose reinforcement predicting higher rates of cannabis self-administration (p = 0.008). Finally, the Social Processes and Negative Valence models were not significant predictors of cannabis self-administration.

Conclusions: Conclusions: Our data indicate that basal CORT, circulating AEA content, high motivation for sucrose reinforcement, and greater reliance on visual cue-based strategies were all significant predictors of motivation to self-administer cannabis vapor in adulthood. Thus, these endophenotypes may precede the onset of problematic cannabis use, which could be leveraged to identify individuals with increased susceptibility for developing CUD.

Keywords: cannabis use, Behavioral characterization, Vapor, drug self-administration

Disclosure: Nothing to disclose.

Elizabeth Lippard*, Buchmaier Sophie, Kaitlyn Meek, Raquel Kosted, Vanessa Le, Jorge Almeida, Stephen Strakowski, Kim Fromme

University of Texas at Austin, Austin, Texas, United States

Background: Given the high genetic load of bipolar disorder and alcohol use disorder (AUD), familial factors likely contribute to their co-occurrence. There is also a positive familial co-aggregation of bipolar disorder and AUD with additive effects suggested. Familial risk for AUD coincides with altered responses to alcohol, with differences suggested to confer risk for AUD. It is not known if familial risk for bipolar disorder relates to altered sensitivity to alcohol or if risk for bipolar disorder interacts with familial risk for alcohol use problems to modify alcohol expectancies, drinking motives, or subjective response to alcohol.

Methods: 62 young adults (22 with a parent with bipolar disorder, 74% female) between 21 and 26 years of age completed clinical phenotyping, assessment of alcohol expectancies and drinking motives, with family history of bipolar disorder investigated. Alcohol problems were assessed using the Short Michigan Alcoholism Screening Test (SMAST) with offspring filling out the SMAST for both parents. Following standard alcohol administration procedures in a simulated bar lab, participants completed measures of physiological and subjective responses after consuming alcohol and placebo beverages (within-subject, counter-balanced) while wearing transdermal alcohol monitors. Time to peak breath alcohol concentration subjective response measurement during the placebo session was yoked to average time of subjective response collection during the alcohol session. A maximum SMAST score in either parent was generated and used as a continuous variable to model parental bipolar risk group (yes/no)-by-max SMAST (in either parent) interactions on alcohol expectancies, drinking motives, and transdermal alcohol concentration. Additionally, bipolar parental risk-by-max SMAST score interactions were investigated with subjective response to alcohol and placebo beverage condition repeated within subject dependent variables. Relations with alcohol use and total Alcohol Use Disorders Identification Test (AUDIT) scores were explored.

Results: A bipolar parental risk group-by-max SMAST interaction was observed for alcohol expectancies on the Anticipated Effects of Alcohol Scale (AEAS) positive valence/positive arousal subscale (p = 0.008). Max SMAST was positively related to anticipated stimulating effects of alcohol in the bipolar parental risk subgroup, but not in those without parental risk for bipolar disorder. Parental alcohol problems was also related to “liking” how one felt during the alcohol session in individuals with parental risk for bipolar disorder, but not in those without parental risk for bipolar disorder (bipolar parental risk group-by-max SMAST-by-beverage condition interaction: p = 0.007). While experience of stimulating effects of alcohol did not differ between groups, a bipolar parental risk group-by-stimulating effects of alcohol interaction on “liking” was observed (p = 0.03), with a greater positive slope observed between stimulating experience of intoxication with “liking” how an individual felt in the bipolar parental risk group. “Liking” how one felt during the alcohol session was positively related to greater past month alcohol use (r = 0.7, p = 0.001) and total AUDIT score (r = 0.4, p = 0.04) in individuals with parental risk for bipolar disorder. Main effects of both parental risk groups were also observed. Young adults without parental risk for bipolar disorder showed an increase in anxiolytic effects during the alcohol condition, compared to placebo, but this increase was not observed in the parental risk for bipolar disorder subgroup (parental risk group-by-condition interaction: p = 0.04). Parental risk for alcohol problems related to drinking motives (p < 0.01) and differences in transdermal alcohol concentration during alcohol beverage session (p = 0.02).

Conclusions: Findings from this study (registered at ClinicalTrials.Gov; NCT04716036) suggest parental risk for bipolar disorder-by-parental risk for alcohol problems interactions on alcohol expectancies, drinking behavior, and alcohol use problems. Parental risk for bipolar disorder and parental risk for alcohol problems on their own related to distinct differences in subjective response to alcohol and drinking motives respectively. Findings emphasize complex familial etiologic risk factors for alcohol misuse.

Keywords: Subjective Response, Bipolar Disorder, Expectancy, Alcohol consumption, familial risk

Disclosure: Nothing to disclose.

Maryam Sorkhou, Molly Zhang, Darby Lowe, Tony George*

University of Toronto, Toronto, Canada

Background: Cannabis Use Disorder (CUD) has been associated with anxiety, insomnia and depression, with frequent cannabis users having a higher prevalence of anxiety disorders. However, there is limited literature assessing the prospective effects of extended cannabis abstinence on symptoms of anxiety and sleep. We conducted a preliminary analysis of 28-days of cannabis abstinence from an on-going study (N = 100) in participants with co-morbid CUD and major depressive disorder (MDD) to investigate the effects of cannabis abstinence on anxiety and sleep.

Methods: Participants (N = 37, 22 abstainers, 15 non-abstainers, biochemically-verified with urine THC-COOH) with co-morbid MDD and CUD underwent a 28-day cannabis abstinence paradigm, including weekly behavioural support sessions. Timeline Follow-Back (TLFB) for cannabis grams per day (GPD) and THC-COOH urine analysis was done weekly to assess cannabis abstinence status. Beck Anxiety Inventory (BAI) and Pittsburgh Sleep Quality Index (PSQI) were administered weekly.

Results: There were significant Group x Time interactions on BAI total anxiety scores and on cognitive and somatic subscales after 28-days of cannabis abstinence in abstainer versus non-abstainer groups, with modest reductions in BAI total anxiety scores in abstainer (n = 22) versus non-abstainer (n = 15) groups. However, there were no changes in PSQI total sleep scores following 28-days of cannabis abstinence.

Conclusions: This preliminary analysis of extended cannabis abstinence in people with co-morbid CUD and MDD suggests improvements in anxiety but not sleep quality. A more detailed analysis of on-going study data will be presented. Our results may have important implications for understanding and treating cannabis use, anxiety and sleep problems in people with major depression.

Supported in part by a Canada Graduate Studies Doctoral (CGS-D) Award to Ms. Sorkou, a CGS-M Award to Ms. Zhang, and CIHR operating grant (PJT-190053) to Dr. George.

Keywords: Anxiety, Cannabis, Sleep, Major Depression

Disclosure: Nothing to disclose.

Christopher Driskill, Reza Arezoomandan, Lily Vu, Sophia Jalilvand, Frank Salazar, Neha Suji, Sanjana Tata, Neissa Molin, Laney Waydick, Sven Kroener*

The University of Texas at Dallas, Richardson, Texas, United States

Background: Exposure to drug-associated cues can trigger cravings and relapse in individuals with substance use disorders. In extinction learning, new associations are formed that compete with these triggers to inhibit behavioral responses. However, extinction training alone is often insufficient to prevent relapse. Therefore, enhancing extinction processes to prevent relapse is clinically important. Vagus nerve stimulation (VNS) induces the release of several neuromodulators that influence cortical plasticity, which can be harnessed to improve learning and memory. Pairing extinction training with VNS facilitates extinction and reduces cue-induced reinstatement of drug-seeking behavior in cocaine self-administering rats. These effects are associated with altered cellular activity in a network involving the medial prefrontal cortex (mPFC), the basolateral amygdala (BLA), and the nucleus accumbens (NAc). The mPFC receives inputs from the amygdala, thalamus, hippocampus, and other limbic structures, enabling it to integrate information about salience, value, and contextual cues related to both appetitive and aversive outcomes. Within the mPFC, the prelimbic (PL) and infralimbic (IL) subregions play largely opposing roles in regulating conditioned responses to rewarding and aversive stimuli. The PL is involved in the expression of conditioned responses and drives reinstatement through its projection to the NAc core. In contrast, the IL is crucial for extinguishing conditioned responses, inhibiting cocaine-seeking through its projections to the NAc shell and/or to the PL. In this study, we further investigated the network mechanisms through which VNS reduces drug-seeking behavior. In Experiment 1, we examined VNS-induced changes in afferent connections to the mPFC from 3 areas critical for appetitive associative learning, memory retrieval, and extinction: the BLA, the ventral hippocampus (vHPC), and the paraventricular nucleus of the thalamus (PVT). In Experiment 2 we determined VNS-induced changes in synaptic plasticity in afferents from the BLA to the IL and the projection from the IL to the NAc shell. We hypothesized that VNS increases activity in pathways that support the expression of extinction memories.

Methods: In Experiment 1 we infused a retrograde AAV into either the IL- or PL to express GFP in mPFC-projecting cells in regions activated by reinstatement (as indicated by cFos activity). Rats self-administered cocaine for 15 days then underwent 10 days of extinction training paired with VNS or sham stimulation, followed by a cue-induced reinstatement session. Rats were sacrificed after reinstatement, and brain slices were stained for cFos. We quantified the number of cFos+ cells in the BLA, vHPC, and PVT, as well as the co-localization of cFos with GFP+ projection cells. In Experiment 2 we performed in-vivo recordings of evoked local field potentials in 2 pathways (BLA-IL; IL-NAc shell) in anesthetized animals either after the last day of cocaine self-administration (or yoked-saline infusions), or after the reinstatement session. Synaptic plasticity was induced by 3 bursts of 100 pulses at 50Hz (20s intervals).

Results: Pairing VNS with extinction reduced drug-seeking during cue-induced reinstatement. In Experiment 1, we found increased cFos activity in cells projecting from the PVT to both the IL and PL. In the projections from the vHPC, VNS significantly decreased cFos activity in the projection to the IL but did not significantly affect projections to the PL. In the BLA, we found an unexpected decrease in cFos activity in cells projecting to the IL and increased cFos colocalization in cells projecting to the PL (all paired t-tests; p < 0.05). Projections from both the vHPC and the BLA also project strongly to mPFC interneurons. Therefore, we examined cFos activity in parvalbumin+ interneurons (PV + ) in the mPFC. cFos activity in PV+ was increased in the PL, but decreased in the IL (paired t-tests, p < 0.05), consistent with the idea that projections from the vHPC and BLA suppress drug-seeking by increasing inhibition in the PL, while allowing relatively more activation of the IL. In Experiment 2, we observed increased synaptic plasticity in the BLA-IL pathway following cocaine self-administration, but no change (yoked-saline) or LTD (sham VNS) in the IL-NAc shell pathway. VNS modulated both pathways in an opposing fashion, inducing LTP in the IL-NAc shell pathway (2-way ANOVA; F(3, 32) = 95.01, P < 0.0001) and reducing LTP in the BLA-IL projection (2-way ANOVA; F(3, 32) = 147.0, P < 0.0001) following reinstatement.

Conclusions: VNS can reduce drug-seeking by reducing the effects of inputs (from the BLA or vHPC) that drive drug-seeking, while strengthening outputs (IL-NAc shell) needed for the expression of extinction memories.

Keywords: vagus nerve stimulation, prefrontal cortex, extinction, reinstatement, synaptic plasticity

Disclosure: Nothing to disclose.

Yuefeng Huang*, Eduardo R. Butelman, Ahmet O. Ceceli, Greg Kronberg, Sarah G. King, Natalie McClain, Yui Ying Wong, Maggie Boros, Rachel Drury, Nelly Alia-Klein, Rita Z. Goldstein

Icahn School of Medicine At Mount Sinai, New York, New York, United States

Background: Women can develop addiction symptoms more rapidly and severely than men. Also, women smokers show striatal hyperreactivity to smoking cues in the follicular (F) phase (elevated estradiol [E]) and greater success in smoking cessation in the luteal (L) phase (elevated progesterone [P]). However, men are more vulnerable to stimulant and opioid overdose deaths, together suggesting a complex sex- and hormonal-based pattern of risk and resilience in addiction. Since women are greatly underrepresented in relevant neuroimaging research, it is imperative to shed light on the neurobiology underlying drug cue reactivity and its emotional regulation in people with addiction. These efforts could potentially guide translational efforts to develop sex- and menstrual phase-specific treatment options for addiction.

Methods: Forty-nine men with heroin use disorder (HUD, age=41.96 ± 9.71) and 32 women with HUD or cocaine use disorder (CUD, N = 16 each, age=38.85 ± 9.84) performed an fMRI task with drug (heroin and cocaine for HUD and CUD, respectively), food, or neutral passive picture viewing, drug reappraisal, and food savoring conditions. Sixteen women (3 HUD and 13 CUD) were scanned twice (F vs. L phase). Analyses were conducted in FSL 6.0 via voxel-wise analyses with cluster-defining threshold of Z > 3.1, corrected to a cluster-extent threshold of p < 0.05. In addition to whole-brain analyses, we used an independent anatomical mask (including the entire frontal cortex and striatum) for a restricted search in our regions of interest guided by our previous studies. Sex differences analyses were conducted with separate one-tailed t-test comparing men and women on each contrast of interest. Paired t-test was used for examining menstrual cycle effects within the women. For correlation analyses, we first conducted within-subject subtraction of covariates related to drug cravings and serum hormonal levels and subject-level statistical mapping for each contrast of interest between the menstrual cycle phases (Δ) before running correlations between Δ brain activation and Δ covariates.

Results: Sex differences were revealed such that women > men in drug cue reactivity in the medial prefrontal cortex (PFC) and women < men in drug reappraisal activity in the frontal eye field (FEF)/dorsolateral PFC (dlPFC). The higher the dlPFC drug reappraisal activity, the lower was the cue-induced drug craving in men. Menstrual cycle effects were also detected such that the F > L phase in drug cue reactivity in the FEF/dlPFC and the L > F phase in drug reappraisal in the anterior PFC/orbitofrontal cortex and insula. Correlations showed that the higher the F > L phase drug cue reactivity in the ventromedial PFC and inferior frontal gyrus (IFG), the more the respective serum E levels, and task-induced and cue-induced drug craving; opposite correlations with serum E levels were found for drug reappraisal activity in the ventromedial and anterior PFC (with similar findings also observed in the caudate and putamen). In contrast, the higher the L > F phase drug reappraisal activity in the dlPFC and IFG, the more the respective serum P/E ratio, respectively.

Conclusions: We found evidence for hyper drug cue reactivity in women compared to men, and more drug cue reactivity during the F > L phase, as driven by E and positively correlated with Δ drug craving in the medial PFC, consistent with E’s role in craving/drug-seeking behaviors. Compared to men, the lateral PFC drug reappraisal (cognitive control) activity was reduced. The same activity was lower in the F than the L phase and recovered by P, consistent with its protective effects in addiction. These results could inform the development of precisely timed and hormonally informed treatment procedures for women with addiction.

Keywords: sex differences, Menstrual Cycle, substance use disorder, Emotional regulation, Human Neuroimaging

Disclosure: Nothing to disclose.

Brittany Kuhn*, Stephen Walterhouse, Eric Dereschewitz, Arkobrato Gupta, Leah Solberg Woods, Dongjun Chung, Peter Kalivas

Medical University of South Carolina, Charleston, South Carolina, United States

Background: Due to the rise of opioid use disorder (OUD) worldwide it is imperative to disentangle the neurobiological mechanisms mediating two opposing states: OUD vulnerability and resiliency. Using a novel preclinical rodent model of OUD that captures the multi-symptomatic diagnosis and complex multidimensional interactions between symptoms conferring OUD vulnerability versus resiliency, we have shown distinct behavioral and neurobiological profiles associated with each phenotype. Furthermore, we have shown marked sex differences within phenotypes for these measures. Genome-wide association study analysis indicates both resiliency and vulnerability to OUD are heritable states, and furthermore identified genetic variants associated with neuroplasticity for addiction traits. These findings prompted further investigation into mechanisms of synaptic plasticity underlying vulnerability and resiliency within regions of a series circuit critical for cued seeking behaviors: pre-limbic cortex (PrL), nucleus accumbens core (NAcc) and ventral pallidum (VP).

Methods: Male and female heterogeneous stock rats (n = 72/sex), a line capturing genetic and behavioral heterogeneity present in humans, were used in these studies. Rats underwent training to collect various measures of heroin taking, extinction and seeking behavior. Using a non-linear network based clustering approach, rats were characterized as OUD vulnerable, intermediate or resilient. Rats were sacrificed following additional extinction training or after extinction training and a second cued reinstatement test, whereupon markers of plasticity were assessed. Differences in density of perineuronal nets (PNN), a component of the extracellular matrix whereby degradation is associated with experience-dependent learning, and microglial reactivity, a mediator of glutamatergic signaling were assessed in the PrL, NAcc and the VP. Furthermore, dendritic spine morphology was analyzed in the NAcc. We hypothesized that OUD vulnerability would be associated with a potentiation in NAcc spine morphology measures, attenuated PNN density, and augmented levels of reactive of microglial. However, resiliency would likely counter these anticipated effects.

Results: Contrary to expectations, no heroin cue-induced differences in NAcc dendritic spine density (Vulnerable, p = 0.48; Resilient, p = 0.52) or spine head diameter (Vulnerable, p = 0.19; Resilient, p = 0.97) were observed in either phenotype compared to saline control. However, vulnerable extinction animals showed a significant reduction in spine density compared to saline controls (p = 0.03), an effect driven by females. Current analyses are dissecting NAcc cell-specific differences in dendritic spine morphology for phenotypes to better elucidate the mechanisms this effect may be operating through. Differences in PNN density within the PrL and VP are also evident. In the PrL, PNN density is augmented following extinction training and then attenuated to baseline levels following cued reinstatement for all phenotypes. The same trend exists in the VP for vulnerable and intermediate rats; however, in resilient rats PNN density remains high following cued reinstatement, suggesting a possible mechanism of promoting resiliency. Additional data collection and analyses are underway to increase statistical power of these findings. We also confirmed prior work and showed no PNNs are present within the rat NAcc. Lastly, current results show an increase in microglial reactivity within the NAcc following a cued reinstatement task in all OUD phenotypes. Additional microglial morphology analyses within the NAcc, as well as the PrL and VP, are currently underway.

Conclusions: By using a model that captures individual variation in OUD susceptibility akin to humans, we are able to disentangle the behavioral, genetic and neuroplastic modulators of OUD vulnerability and resiliency. Here we show phenotypic and test-dependent neuroplastic adaptations in a series circuit regarded both clinically and preclinically as a main contributor to OUD relapse.

Keywords: microglia, Perineuronal nets, Dendritic spines, Synaptic Plasticity, Opioid addiction

Disclosure: Nothing to disclose.

Christa Corley*, Haley McManus, Kelly Dunn, Heather Barton, Cassandra Gipson

University of Kentucky, Lexington, Kentucky, United States

Background: Fentanyl is a lipophilic μ opioid receptor agonist used both clinically and illicitly. Its lipophilic properties allow for it to be sequestered and accumulate in peripheral tissues, most notably in the stomach. Clinical studies have shown that sequestered fentanyl in the stomach is released and reabsorbed via the intestine causing a secondary plasma peak of fentanyl in the blood. In surgical settings, cimetidine (a histamine-2 receptor antagonist) is an FDA-approved gastric acid reducer that is also given prophylactically to reduce secondary peaks of fentanyl. However, no studies have evaluated cimetidine as a fentanyl withdrawal treatment after chronic use. We hypothesize that cimetidine given after fentanyl self-administration (SA) of fentanyl will reduce fentanyl reabsorption and reverse fentanyl-induced gastrointestinal dysfunction, thus could be a promising clinical therapeutic to reduce fentanyl withdrawal severity.

Methods: 17 female Long Evans rats underwent jugular vein catheter surgery followed by 15 sessions of 2 h fentanyl SA (3.2 µg/kg/infusion). Rats were then randomly assigned to receive one intravenous infusion of vehicle (saline) or cimetidine (5 mg/kg or 10 mg/kg). Following the final day of SA, rats underwent observational testing for somatic signs of spontaneous withdrawal at 0, 24, and 48 h post-SA. Sections of the small intestine were also collected, fixed, and stained with hematoxylin and eosin (H and E) to measure and analyze villi height and crypt depth. We focused on the proximal portion of the intestine which equates to the duodenum. Data were analyzed via appropriate t test or ANOVA with Tukey’s post-hoc comparisons.

Results: Although preliminary, cimetidine dose-dependently reduced somatic signs of withdrawal at the 24h timepoint with the 10 mg/kg dose having the greatest effect (main effect of time (F1.76,22.88 = 10.17; p < 0.05) and trending group x time interaction (F4,26 = 2.370; p = 0.078). Preliminary intestinal morphology results indicate that both doses of cimetidine treatment may increase villi length, although data collection remains ongoing.

Conclusions: Taken together, these preclinical results indicate that intravenously delivered cimetidine may attenuate fentanyl withdrawal severity and reverse fentanyl-induced gastrointestinal dysregulations. Thus, cimetidine may be an easily implemented and efficacious therapeutic to reduce fentanyl withdrawal severity in individuals with opioid use disorder undergoing fentanyl detoxification. Clinical evaluations of cimetidine as a fentanyl withdrawal therapeutic may be warranted.

Keywords: substance use disorder, opioid withdrawal, opioid use

Disclosure: Nothing to disclose.

Lara Ray*, Lindsay Meredith, Craig Enders, Erica Grodin, Steven Nieto, Wave-Ananda Baskerville, Suzanna Donato, Steven Shoptaw, Artha Gillis, Michael Irwin, Karen Miotto

University of California - Los Angeles, Los Angeles, California, United States

Background: The neuroimmune system represents a promising target for the development of novel medications for alcohol use disorder (AUD). Ibudilast (IBUD) is a neuroimmune modulator which selectively inhibits phosphodiesterases (PDE)-3, -4, -10, and -11, and macrophage migration inhibitory factor (MIF). Based on preclinical and clinical early efficacy testing for AUD, this study tested the hypothesis that treatment with ibudilast would be associated with significant reductions in drinking during a 12-week clinical trial compared to placebo. Analyses considered moderators of medication effects: sex, baseline inflammation (i.e., C-reactive protein levels), and medication compliance.

Methods: This double-blind, randomized, single site clinical trial tested the efficacy of ibudilast (50 mg/twice daily), compared to placebo, on alcohol use over the course of a 12-week treatment period. The primary outcome measure was percent heavy drinking days. Participants were 102 (41 females) treatment-seekers with moderate or severe AUD and trial retention was 87.25% at 12-week follow-up.

Results: There was no significant difference between IBUD versus placebo on the primary outcome of percent heavy drinking days (β = 0.06, SE = 0.08, t = 0.74, p = 0.46). There was a significant CRP × Medication × Time 1 interaction (β = 0.26, SE = 0.13, t = 2.03, p < 0.05) and a significant CRP × Medication × Time 2 interaction (β = -0.28, SE = 0.14, t = 1.96, p = 0.05). The rate of PHDD decrease was only superior in the IBUD condition, versus PLAC, during the maintenance time period (weeks 2-12) and the opposite was true during the initial change period (baseline to 2-week follow-up). Analyses accounting for compliance, found that individuals in the IBUD condition who were compliant had lower PHDD compared to individuals in the IBUD condition who were non-compliant. Secondary drinking outcomes revealed significant Sex × Medication interactions on drinks per day (β = -2.48, SE = 1.07, t = -2.32, p < 0.05) and percent days abstinent (Sex × Medication interaction, β = 0.26, SE = 0.14, t = 1.91, p = 0.056), such that IBUD was superior to PLAC in females, compared to male participants.

Conclusions: here was no support for ibudilast, over placebo, for the treatment of AUD in this clinical trial. However, the benefits of ibudilast, when present, appear to be dependent on sex and baseline levels of inflammation. As novel treatments for AUD are developed for novel molecular targets, their effects should be interpreted in the context of mechanisms and moderators of efficacy. Medication compliant and placebo effect remain problematic in 12-week clinical trials for AUD.

Keywords: Alcohol, pharmacotherapy, neuroimmune mechanisms, Sex-specific effects

Disclosure: Nothing to disclose.

Dylan Kirsch*, Kaitlin McManus, Erica Grodin, Steven Nieto, Lara Ray

University of California, Los Angeles, Los Angeles, California, United States

Background: Alcohol cue-induced craving is a central feature of alcohol use disorder (AUD). In individuals with an AUD, exposure to alcohol-related cues—such as the sight, taste, smell of alcohol—is thought to elicit craving. This conditional response, known as cue-reactivity, contributes to the maintenance and recurrence of AUD, and thus, has been a key target for AUD treatment. Researchers have developed alcohol cue exposure paradigms, which involve systematically exposing participants to alcohol-related and control cues while assessing their subjective urge to consume alcohol. These paradigms are widely used in experimental laboratory research to study the role of cue-reactivity in the course of AUD and as a screening tool for AUD medication development. Despite the widespread use of cue-reactivity testing in AUD research, only a subset of individuals with AUD are actually cue-reactive in the laboratory. Failure to consider individual variation in subjective cue-reactivity could hinder its research utility, particularly in the context AUD medication development. Advancing cue-reactivity as an outcome measure in experimental laboratory trials therefore requires careful consideration of individual differences in cue-reactivity.

To that end, a promising avenue involves leveraging data-driven methods, including machine learning, to harness the diversity of samples and identify predictive individual variations. The present study tested the utility of several data-driven machine learning and multivariate models to predict cue-reactivity among individuals with AUD using demographic and clinical variables as predictors.

Methods: Individuals with current AUD (N = 128; 77M/51F) completed a standard alcohol cue-exposure paradigm in the laboratory and a battery of demographic and clinical measures. Participants were categorized as cue-reactive (CR + ) or non-reactive (CR-) based on an a priori definition of subjective cue-reactivity. Specifically, participants were considered CR+ if they had an AUQ score that was > 50% greater during the alcohol, versus water, condition. Participants who did not meet this criterion were classified as CR-. We employed three machine learning models—random forest, lasso regression, ridge regression—and one traditional multivariate statistical model—stepwise regression—to predict individuals as CR + /CR- using 30 demographic and clinical predictors. We then used k-fold cross-validation (k = 10) to evaluate the predictive performance of each model. We plotted receiving operator characteristic curves and calculated area under the curve (AUC). Additional performance characteristics included accuracy, sensitivity, and specificity.

Results: Thirty-nine percent of participants were categorized as CR + . The CR+ group reported greater alcohol urge during the alcohol (versus water) condition (t = 14.6, p < 0.001). The CR- group did not differ in alcohol urge between the water and alcohol conditions (t = -0.5, p = 0.6), thereby supporting that the CR groups differed on the outcomes in which they were dichotomized. During the water condition, alcohol urge was significantly lower in the CR + , compared with the CR-, group (t = -5.1, p < 0.001), but the groups did not differ in alcohol urge during the alcohol condition (t = 1.6, p = 0.1).

The CR+ group was comprised of a greater proportion of individuals who identify as White, exhibited less severe AUD symptomology, less alcohol-related obsessive behaviors/thoughts, less tonic craving, less cigarette smoking and milder anxiety compared with the CR- group.

Stepwise regression (AUC = 0.72) outperformed all machine-learning models in predicting cue-reactivity (lasso, AUC = 0.60; ridge, AUC = 0.62; random forest, AUC = 0.51). Predictor variables retained in the stepwise regression model were related to race, cigarette smoking, obsessive alcohol-related thoughts/behaviors, and acute withdrawal symptoms. The resulting stepwise regression model suggested that individuals who identify as non-White and experience greater obsessive alcohol-related thoughts/behaviors are less likely to be CR+ (p’s > 0.017).

Conclusions: Stepwise regression was the most effective model in predicting alcohol cue-reactivity, outperforming random forest and regression regularization techniques (lasso, ridge). Results suggest that individuals who self-identify as non-White, and those who have greater alcohol-related obsessive thoughts/behaviors are less likely to be CR + ; however, the predictive accuracy of stepwise regression was modest at 69%. In other words, the model outperformed classification by chance (i.e., 50%), yet did not reach the target accuracy of > 80%. Despite the modest predictive accuracy of demographic and clinical variables in these models, there were significant between-group differences in demographic and clinical variables. Specifically, the CR+ group was comprised of a greater proportion of individuals who identify as White, exhibited less severe AUD symptomology (including less tonic craving/obsessive thoughts), less cigarette smoking and milder anxiety compared with the CR- group. While machine learning did not emerge as a robust prediction model of cue-reactivity, this study underscores the need to acknowledge individual variability in cue-reactivity in AUD research. Given the demographic and clinical differences between CR groups, it is critical to develop methods capable of leveraging these factors to predict individual variations in cue-reactivity.

Keywords: Alcohol, machine learning, cue-induced craving, alcohol use disorder, Cue Reactivity

Disclosure: Nothing to disclose.

Constance Guille*, Nathaniel Baker, Brian Neelon, Rachel Tomko, Lisa Nunn, Aimee McRae-Clark, Kathleen Brady

Medical University of South Carolina, Charleston, South Carolina, United States

Background: Opioid Use Disorders (OUDs) and drug overdose deaths are increasing at alarmingly high rates. Stress and dysregulation in noradrenergic systems appear to play an important role in the pathophysiology of substance use disorders and relapse to drug use, particularly for women. Alpha-2 adrenergic agonist medications effectively decrease noradrenergic activity and have demonstrated benefit in preventing relapse to substance use and decreasing stress-reactivity and craving in cocaine- and nicotine-dependent women, compared to men. Alpha-2 adrenergic agonists, such as lofexidine may help decrease stress reactivity in individuals with OUDs and reduce stress and craving, but gender differences have not been systematically explored. The objective of this study is to conduct a randomized trial examining gender differences in stress and craving among adult males and females with OUD taking methadone or buprenorphine (MOUD) and randomly assigned to lofexidine or placebo. We hypothesize that lofexidine, as compared to placebo, will attenuate subjective stress and drug craving, and findings will be more pronounced among women, compared to men.

Methods: Following informed consent, baseline assessments, and 7–10 days of abstinence (measured by urine drug screens), adults aged 18-60 years old with OUD on a stable dose of MOUD for 2 weeks were randomized (1:1) to receive lofexidine (titrated to 0.72mg/TID) or placebo for a 32-day period. At the end of this period while continuing study medication, the Trier Social Stress Test [TSST] and an opioid specific drug-cue stressor, Imaginary Scripts [IS] were administered in a counterbalanced order on consecutive days. Multilevel longitudinal GLMMs using all post-TSST and post-IS data were used to assess the hypothesis that lofexidine will attenuate subjective stress and craving response to the TSST and IS tasks. Participant sex and the interaction between sex and treatment assignment were added to the model to examine the differential effects of lofexidine on stress and craving across female and male subjects, while controlling for baseline levels of outcome variables. The study was powered to find a sex differential treatment effect size of Cohen’s d ~ .8 with a sample size of 34 participants per group, per sex with 20% attrition during treatment (n = 136 total).

Results: Craving:

• TSST Craving: Participants assigned to lofexidine had lower craving at the pre-TSST baseline (p = 0.02) that persisted following the TSST, but no post-TSST treatment effects were noted after adjusting for baseline levels (p = 0.58). No sex differences were identified, and sex did not modify treatment effect.

• IS Craving: Participants assigned to lofexidine had lower reports of craving at pre-cue baseline (p = 0.055), but no post-IS treatment effects were noted after adjusting for baseline levels (p = 0.36). Male participants had significantly lower average craving response following the IS (1.81 (0.20)), compared to female participants (2.55 (0.22); p = 0.01). Although similar at study baseline, females had a more robust initial craving response to the IS compared to males (p < 0.01). This effect in females was consistent across treatment groups (treatment x sex x time, p = 0.87). Sex did not modify treatment effect (p = 0.37).

Stress:

• TSST Stress: No post-TSST treatment effects on stress (p = 0.15) or sex differences (p = 0.30) were identified, and sex did not modify treatment efficacy (p = 0.39). There was no significant treatment by sex interaction (p = 0.69).

• IS Stress: No post-IS treatment effects on stress were noted after adjusting for baseline levels (p = 0.29). Initial post-IS response was greater in the lofexidine group as compared to placebo (p = 0.018). Although stress ratings were similar at baseline for males and females, females had a more robust stress response to the cue than did males (p = 0.008). This effect was seen at 0- and 5-min post-cue but were equivalent to males at the later times (30, 60 min post-cue). Sex did not modify overall treatment efficacy (p = 0.42) and effects in females were consistent across treatment groups (treatment x sex x time, p = 0.36).

Conclusions: Lofexidine did not attenuate drug craving or subjective stress 60 minutes after the TSST or IS stressors. While participants assigned to lofexidine had lower reports of craving in response to the TSST and IS, these effects were attenuated once baseline levels of outcome variables were accounted for. Larger sample sizes may be needed to detect smaller treatment effects of lofexidine on craving. Sex differences in treatment response and efficacy were not identified although analyses may be limited by sample size. Overall, females, compared to males, appear to have a more robust craving and stress response to the IS and suggests that this laboratory task maybe important to include in future studies examining sex differences in OUD. The sample was primarily White which is consistent with disparities in MOUD treatment and requires greater attention to ensuring a more diverse sample in future studies.

Keywords: opioid use disorder, Noradrenergic System, Perceived stress, cue-induced craving, Randomized Double-Blind

Disclosure: Nothing to disclose.

Maricela Martinez*, Giuseppe O. Bravo, Stephen V. Mahler

University of California, Irvine, Irvine, California, United States

Background: Adolescent use of Δ9-tetrahydrocannabinol (THC) is linked to later-life changes in cognition, memory, and susceptibility to drug use. Animal models suggest this could be a causal effect of THC influencing adolescent circuit development, perhaps especially in the prefrontal cortex and interconnected circuits like the amygdala. These regions are involved in cognition, stress, and addiction, and undergo developmental change during adolescence and thus may be sensitive to persistent disruption by THC. Here, we employed our well-characterized adolescent THC exposure model in rats to examine long-term effects on stress-related behaviors and brain activity, and their relationship to addiction-relevant opioid drug seeking.

Methods: We exposed female and male Long-Evans rats to THC (5mg/kg i.p.) or vehicle daily during adolescence, from postnatal day 30 to 43, followed by a washout period in which rats grew to adulthood (PD70 + ). They were then behaviorally tested for their responses to a localized aversive stimulus (a shock-delivering rod, 1.5mA), and in a separate test, to a diffuse aversive stimulus (a foot shock (x3), 0.75mA) to assess responses to these threats and cues predicting them. The same rats were next trained to self-administer the rewarding opioid drug remifentanil (3.2 ug/kg/infusion i.v.), and behavioral economic demand elasticity was quantified in our within-session protocol. We then extracted both “hedonic set point” for remifentanil (reflecting the preferred level of drug intake at very low cost), and motivation to maintain this drug blood level (i.e. sensitivity of intake to price/effort, or demand elasticity). After achieving stable intake patterns in this task, we then asked how stressors (footshock 0.5mA, or the pharmacological stressor yohimbine 2.5mg/kg, i.p.) alter opioid intake patterns in rats with or without adolescent THC history. Finally, we measured locomotor and ultrasonic vocalization responses to yohimbine, as well as c-Fos in prefrontal cortex and amygdala, in both adolescent treatment groups.

Results: Adolescent THC rats, relative to adolescent vehicle-treated controls, showed distinct behavioral and brain responses to stressors that varied by both stress type and sex. Adolescent THC in both sexes led to reduced defensive responses to the localized shock rod, including in a post shock test in which the rod was deactivated and thus served as a shock cue. In contrast, behavioral responses to the more diffuse foot shock threat, and cues predicting it, were instead sex-dependent—with stronger effects in females. Effects of foot shock and pharmacological stressors on remifentanil demand elasticity and intake were also sex-dependent, and were markedly altered in females with a history of adolescent THC. Specifically, yohimbine effects on remifentanil self-administration, locomotion, and brain c-Fos expression depended upon both sex and adolescent treatment history, though neither adolescent treatment nor yohimbine affected vocalizations.

Conclusions: Collectively, these findings demonstrate sex-dependent effects of adolescent THC exposure on the brain and behavior, with pronounced influences on responses to stressful stimuli and cues predicting them. Results will lend insight into the impact of adolescent THC exposure on stress and addiction, with implications for potential long-lasting negative effects of teenage cannabis use in humans.

Keywords: THC, Adolescence, Opioids, Aversive Motivation, Aversive Learning

Disclosure: Nothing to disclose.

Edoardo Tiziani, Francesca Frescura, Luca Ferraro, Robert Schwarcz, Sarah Beggiato*

University of Ferrara, Ferrara, Italy

Background: The use of THC during the adolescence constitutes a risk factor for cognitive impairments in psychiatric disorders in adulthood who begin using cannabis during adolescence (D’Souza et al., 2020). Notably, early exposure to THC causes enduring cognitive dysfunctions and impaired glutamatergic and GABAergic function in the rat prefrontal cortex (PFC; Zamberletti et al., 2014). These alterations are associated with a long-lasting increase in PFC levels of kynurenic acid (KYNA), a metabolite of the kynurenine pathway (KP) of tryptophan degradation, believed to be involved in the pathophysiology of several psychiatric diseases, including schizophrenia (SZ). As the detrimental THC-induced deficits are not seen in all cannabis users, it seems likely that some people have genetic vulnerability linked to THC use (Levine et al., 2017; Hasin, 2018). Mice with a targeted genetic disruption of KMO (i.e. a KP enzyme that indirectly controls the production of KYNA; Giorgini et al., 2013) leading to an increase in KYNA levels compared to wild-type (WT) control mice, might help to clarify the role of PFC KYNA signaling in the dysfunction induced by the exposure to THC during the adolescence.

Methods: Male and female C57BL6/J (a well characterized mouse strain with respect to the behavioral effects of THC; Wise et al., 2011) WT (i.e. Kmo + /+; N = 9-11, 4-6 males-females/group) and HET (i.e. Kmo + /-; N = 9-11, 4-6 males-females/group) were chronically treated with vehicle or ascending doses of THC (2.5 mg/kg; 5 mg/kg; 10 mg/kg i.p.), twice a day from postnatal day (PD) 35 to 45 (see Zamberletti et al., 2014 for protocol validation). Extracellular KYNA and GABA levels have been measured by in vivo microdialysis in the PFC of adult (PND 75) mice. On day 1 after the surgery, after three stable basal values were obtained, the effects of a challenge of L-kynurenine (i.e. the biological precursor of KYNA (10 mg/kg i.p) mimicking a translationally relevant “second hit”, on extracellular KYNA and GABA levels were measured. Data were analyzed by Student’s t-test and/or Two-way ANOVA followed by Tukey’s test for multiple comparisons. Males and females were used.

Results: In vivo microdialysis revealed that basal extracellular PFC KYNA levels were significantly higher (P < 0.05) in female and male HET mice, while basal extracellular PFC GABA levels were significantly higher in female (P < 0.05), but not male, HET mice compared to WT animals. The exposure to THC during the adolescence caused a long-lasting increase of KYNA levels in the PFC of male, but not female, WT mice (P < 0.05), without affecting PFC GABA levels in both sexes. Of note, the THC-induced long-lasting increase in PFC KYNA levels was not observed in male HET mice. As expected, the administration of L-kynurenine (10 mg/kg; i.p), significantly increased (P < 0.01) PFC KYNA levels in female and male WT and HET mice. However, the L-kynurenine-induced increase in PFC KYNA levels was significantly higher (P < 0.05) in WT compared to HET mice and was not affected by adolescent THC treatment. Finally, L-kynurenine administration did not modify PFC GABA levels in vehicle- and THC-treated WT and HET mice, although a trend to a decrease was observed in THC-treated HET mice.

Conclusions: The present study revealed that the use of cannabinoids early in life can induce a malfunction of the KP in male, but not female, WT mice. This is in line with previous studies in male rats. Furthermore, the increase in brain KYNA levels induced by the genetic reduction of KMO levels is associated with a different vulnerability to THC-induced alterations of PFC KYNA/GABA levels. Ongoing molecular studies in our laboratory are carefully investigating possible mechanisms underlying these changes to provide useful information regarding the etiology of adolescent THC-induced risk of neuropsychiatric illnesses in adulthood.

Keywords: Kynurenic acid, GABA, cannabinoids

Disclosure: Nothing to disclose.

Kevin Coffey*, Neethi Belur

University of Washington, Seattle, Washington, United States

Background: The opioid crisis continues to spiral out of control in the United States, fueled by the widespread availability of cheap and potent fentanyl. Current pharmacological treatments for fentanyl use disorder, primarily opioid replacements, have proven insufficient to stem the tide of fentanyl related suffering and deaths. Novel pharmacotherapies are desperately needed, ideally ones that are non-opioid, highly convenient, and produce minimal side effects. One promising class of drugs that meets these criteria are glucagon-like peptide 1 (GLP-1) receptor (GLP-1R) agonists. Endogenous GLP-1 is released in response to food intake, but GLP-1Rs are present in many tissues throughout the body, including brain regions involved in addiction. Early studies have shown GLP-1R agonists can reduce drug seeking, including for alcohol, nicotine, and opioids. However, it’s not clear if they can be administered conveniently (i.e. once weekly), if they reduce can fentanyl intake over a significant period of time, or which brain regions are involved in mediating their anti-addictive effects.

Methods: Male and female rats self-administered (SA) oral fentanyl for five weeks following our recently developed oral self-administration procedure. After SA, rats undergo extinction training (abstinence) followed by cued reinstatement testing. Experimental animals received dulaglutide (0.15mg/kg, s.c.) once weekly, prior to the start of the third week of behavioral sessions. In a subset of animals, GCaMP8f was injected into the LHb along with GRIN lens implants. Individual neurons in the LHb were recorded twice weekly during SA, extinction, and reinstatement. Experimental and control animals (vehicle injections) were compared to determine if once-weekly dulaglutide reduces different components of fentanyl addiction, including intake, escalation, cue-association, or reinstatement. LHb neural dynamics were also compared between groups during different behaviors, including lever-press, cue processing, and during fentanyl consumption.

Results: Male and female control rats both escalate fentanyl seeking across training (p  <  0.01), decrease active lever pressing during extinction (p  <  0.01) and increase active lever pressing again during cued reinstatement (p  <  0.01). LHb activity in control animals is also significantly increased by drug-cues early in training and decreased during fentanyl consumption (ERT, alpha < 0.05). We aim to determine if once-weekly dulaglutide treatment will reduce different components of fentanyl addiction in experimental animals, even after significant SA has occurred. These components include total intake, escalation, cue-association, and LHb modulation.

Conclusions: The United States opioid epidemic is defined by rampant and treatment resistant fentanyl use, and current pharmacological treatments have proven insufficient to stem the tide of fentanyl related suffering and deaths. If the long acting GLP-1R agonist dulaglutide, given once weekly, can reduce fentanyl self-administration (SA) and relapse, even after SA has been well established, it has potential to be rapidly integrated into treatment regimens and to alleviate some the suffering caused by fentanyl use disorder.

Keywords: Fentanyl, GLP-1 receptor agonist, miniscope, Lateral Habenula

Disclosure: Nothing to disclose.

Vincent Brice Owona Ayissi*, Brittany N. Kuhn, Jordan L. Hopkins, Eric Dereschewitz, Peter W. Kalivas

Medical University of South Carolina, Charleston, South Carolina, United States

Background: The ventral pallidum (VP) is a component of the limbic loop responsible for the regulation of motivation, emotions and behavior. Moreover, it connects the nucleus accumbens with downstream motor structures to translate motivation into action. In this study, we used DREADD-based chemogenetic tools to assess the modulatory role of GABA neurons in Cre-mice following heroin self-administration.

Methods: A cre-dependent chemogenetic approach was taken to selectively activate (Gq) or inhibit (Gi) GABA neurons in mice brains. In experiment 1, male and female Cre-IRES (Vgat) mice were trained to self-administer heroin and extinguished for cue-induced reinstatement testing after receiving an anterograde AAV-hSyn-HM3G(q) viral injection into the dlVP. In experiment 2, the same procedure was employed using an inhibitory anterograde AAV-hSyn-HM3G(i) virus. In both experiments, Clozapine-N-oxide (CNO) was used to activate Gq and Gi-DREADDs respectively. Following testing, mice were euthanized, and brain tissues examined for DREADD expression.

Results: Results showed that cell-specific activation (Gq) of dlVP GABA neurons promotes cued-induced reinstatement of heroin seeking (experiment 1), while cell-specific inhibition of GABA neurons (Gi) promotes refraining from heroin consumption (experiment 2).

Conclusions: In summary, this work emphasizes the modulatory role of the dlVP GABA neurons in mediating heroin consumption in mice.

Keywords: ventral pallidum, heroin, chemogenetics, GABA neurons, cue-induced-reinstatement

Disclosure: Nothing to disclose.

Junshi Wang*, Stephanie Caligiuri, Masago Ishikawa, Purva Bali, Lauren Wills, Adam Cato, Paul Kenny

Icahn School of Medicine At Mount Sinai, New York, New York, United States

Background: Alcohol use disorder (AUD) is a leading cause of premature death in the United States. The prevalence of AUD is increasing more prominently in female than male (XX vs. XY) individuals. Females are also at heightened risk of experiencing the adverse health effects of alcohol. Sex-dependent neural mechanisms relevant to AUD are poorly understood. Nicotinic acetylcholine receptors (nAChRs) not only mediate the reinforcing effects of nicotine but also are involved in AUD. The smoking cessation agent and nAChR partial agonist varenicline (Chantix) decreases alcohol intake in humans. Furthermore, allelic variation in the CHRNA5/A3/B4 gene cluster, which encodes α5, α3, and β4 nAChR subunits, respectively, elevates the risk of AUD. α5 nAChRs are highly enriched in the midbrain interpeduncular nucleus (IPN), which receives dense cholinergic innervation from the medial habenula (MHb). α5 nAChRs in the MHb-IPN circuit regulate aversion to acutely administered nicotine and contribute to aversive components of the nicotine withdrawal syndrome. However, the involvement of α5 nAChRs in AUD-related behaviors, including sex-dependence differences in alcohol consumption, remains unknown. Neural activity in the IPN is highly sensitive to the female gonadal hormone progesterone, which is known to influence alcohol consumption. Here, we investigated the role of α5 nAChRs in the MHb-IPN circuit in regulating sex differences in the motivational properties of alcohol and the effects of progesterone on alcohol consumption.

Methods: Behavioral experiments involved adult wild-type and Chrna5 knockout (KO) mice on a C57BL/6J background. Drinking-in-the-dark and operant responding were used to assess ethanol consumption. Single-cell RNA sequencing (scRNA-seq) was performed using the Chromium platform from 10x Genomics and analyzed using ScanPy. Visium spatial transcriptomics (10x Genomics) was analyzed using Seurat. Whole-cell recordings were used to assess neural activity and synaptic transmission. HEK cells stably expressing α4β2α5 nAChRs were used to assess the effects of progesterone on nAChR signaling. CRISPR/Cas9 technology was used for Chrna5 gene cleavage in the IPN.

Results: Spatial transcriptomics and scRNA-seq revealed striking differences in the transcriptional profiles of IPN neurons between female and male mice. Differentially expressed genes (DEGs) in the IPN of females vs. males were enriched in genes known to regulate nAChR signaling. Using slice electrophysiology, we found that ethanol stimulates cholinergic neurons in the MHb and enhances GABA transmission in the IPN, with these effects more prominent in female vs. male mice. Only in females do IPN neurons exhibit a significant and persistent enhancement of GABA transmission in response to prolonged exposure to ethanol. The stimulatory effect of ethanol on IPN GABA signaling was attributed to enhanced presynaptic release, a response that was absent in female Chrna5 knockout mice. Wild-type female mice consumed greater qualities of ethanol than males. This sex difference was abolished in Chrna5 KO mice and by CRISPR-mediated Chrna5 cleavage in the IPN. Finally, we found that progesterone potently inhibited α5 nAChR signaling in cultured cells and abolished ethanol-induced enhancement of GABA transmission in the IPN of female mice. Moreover, progesterone injection induced a robust c-Fos expression in the IPN of female but not male mice and enhanced alcohol consumption in females exclusively. These effects of progesterone on IPN activity and alcohol consumption were abolished in female Chrna5 KO mice and by CRISPR-mediated Chrna5 cleavage in the IPN of female wild-type mice.

Conclusions: We generated transcriptomic, neurophysiological, cellular, and behavioral data that comprehensively characterized the sex-dependent differences in the functioning of IPN cholinergic transmission in the context of alcohol use. Our results identified α5 nAChRs in the IPN as an important regulator of alcohol consumption in females and a novel effector of the stimulatory effect of progesterone on alcohol intake.

Keywords: nicotinic acetylcholine receptors, Alcohol consumption, progesterone, interpeduncular nucleus, sex differences

Disclosure: Nothing to disclose.

Melanie Schwandt*, Vijay Ramchandani, Nancy Diazgranados, David Goldman

National Institute on Alcohol Abuse and Alcoholism, Bethesda, Maryland, United States

Background: Resilience is defined as positive adaption to stress and adversity, and is influenced by psychological, biological, and environmental factors. Psychological resilience refers to the maintenance of mental health in the face of adversity and can be conceptualized as both a capacity (collection of traits, skills and resources that predispose an individual for resilience) and an outcome (absence or reduction of negative consequences on mental health). The Connor-Davidson Resilience Scale (CD-RISC) is a widely used measure of psychological resilience capacity. CD-RISC score has been associated with lower risk for mental health diagnoses, including alcohol use disorder (AUD), and lower problematic drinking. In the current study, we utilized an outcome-based method of assessing resilience - discrepancy-based psychological resilience (DBPR) - to investigate psychological resilience. This method involves regressing a measure of psychological dysfunction on stress/trauma load and extracting the residuals as an indicator of relative resilience (e.g., lower than expected psychological dysfunction as a function of stress/trauma load). The aims of the current study were: 1) to investigate the association of discrepancy-based psychological resilience with risk for problematic drinking and AUD, and 2) to evaluate whether discrepancy-based psychological resilience is associated with potential biological markers of resilience.

Methods: Methods: The study sample included 1274 participants (mean age = 40.5 years; 57.7% male; 46.7% White; 38.3% Black; 49.1% with current AUD) enrolled in the NIAAA Natural History Protocol. Participants were assessed using the Structured Clinical Interview for DSM Disorders, and completed assessments of early and current life stress, and psychological functioning (including measures of internalizing and externalizing behavior). Alcohol-related measures include the Timeline Followback (TLFB), Alcohol Use Disorders Identification Test (AUDIT), and Alcohol Dependence Scale (ADS). Biological markers examined included resting heart rate, c-reactive protein (CRP), and the systemic inflammation response index (SIRI), a peripheral measure of inflammatory response based on number and type of circulating immune cells. Latent measures of stress/trauma load (STL) and psychological distress (PD) were estimated using factor analysis. To generate DBPR scores, PD was linearly regressed on STL, and the residuals were extracted and multiplied by -1 such that positive residuals indicate better than expected psychological functioning, or relative resilience. Linear regression was used to investigate associations of DSBP scores with both alcohol-related measures and biological markers. Covariates included age, sex, race, and household income.

Results: The regression of PD on STL was significant (p < 0.0001, R2 = 0.57). DBPR was moderately correlated with CD-RISC score (r = 0.42, R2 = 0.18, p < 0.0001) in the subsample of participants with both measures (n = 376). Controlling for sociodemographic variables, DBPR was associated with lower AUDIT scores (b = -1.0, partial R2 = 0.15, p < 0.0001) and lower odds of a current AUD diagnosis (OR = 0.83, p < 0.0001). Among individuals with AUD, DBPR was negatively associated with ADS score (b = -0.47, partial R2 = 0.06, p < 0.0001). Controlling for current AUD diagnosis and sociodemographic variables, DBPR was significantly associated with lower heart rate (b = -0.50, partial R2 = 0.02, p < 0.0001) and lower SIRI (b = -0.01, partial R2 = 0.01, p < 0.05), but showed no association with CRP.

Conclusions: Consistent with prior findings utilizing the CD-RISC, discrepancy-based psychological resilience was associated with reduced risk for problematic drinking and AUD, and decreased severity of alcohol dependence. While the CD-RISC is primarily a measure of resilience capacity, DBPR is an outcome-based measure of resilience. Correspondence between these two distinct measures highlights the multidimensional construct of resilience, particularly as it relates to AUD, a clinical context in which resilience is challenged. Discrepancy-based psychological resilience was also associated with a lower heart rate and a minor reduction in the SIRI, which is consistent with findings showing an association between SIRI and risk for depression and anxiety. The current findings add to existing research on potential biomarkers that, when examined along with psychological indicators of resilience, contribute to a better understanding of resilience as an interaction of psychological, biological, and environmental factors.

Keywords: Resilience, Lifetime stress, Alcohol, Biomarkers

Disclosure: Nothing to disclose.

Yasmin Zakiniaeiz*, Nakul Raval, Nabeel Nabulsi, Henry Huang, David Matuskey, Gustavo Angarita-Africano, Robin Bonomi, Sherry McKee, Ansel Hillmer, Kelly Cosgrove

Yale University School of Medicine, New Haven, Connecticut, United States

Background: Chronic alcohol consumption has neurotoxic effects on the brain including neurodegeneration, the loss of neuronal structure and function. Neurodegeneration occurs over years of alcohol use and impairs neurocognition, further precipitating alcohol drinking and driving the addiction cycle. Frontal-striatal, frontal-cerebellar, and hippocampal circuits are heavily implicated in alcohol-induced neurodegeneration in humans; however, little is known about the effects of alcohol on synaptic number and function. Preclinical studies have shown lower cerebellar synaptic density in rodents chronically exposed to alcohol versus control animals but the effects of alcohol on synaptic circuitry in people are not known. The goal of this study was to use positron emission tomography (PET) brain imaging and the radiotracer [11C]UCB-J, which binds to synaptic vesicle glycoprotein 2A (SV2A), a marker of synaptic density, in people with AUD and to examine relationships to alcohol-related clinical correlates. Based on preclinical findings and our prior findings, we hypothesized that individuals with AUD would show lower levels of synaptic density relative to healthy controls and that lower synaptic density levels would be related to poorer cognitive performance.

Methods: Twenty-three people with AUD (mean: 39 years old, 4 drinks/session, 4 drinking days/week for 15 years; 14 females) and 22 age-, sex-, and cigarette smoking status-matched healthy controls participated in one [11C]UCB-J PET scan to measure SV2A availability, a marker of synaptic density. PET scans were acquired for 90 minutes after bolus injection of 541 ± 182 MBq [11C]UCB-J on a High Resolution Research Tomograph with arterial blood sampling acquired throughout to measure the metabolite-corrected input function. The outcome measure, volume of distribution (VT), was estimated regionally with one-tissue compartment model as a measure of SV2A availability. A linear mixed model was conducted with [11C]UCB-J volume of distribution (VT) as the dependent variable, diagnostic group (AUD vs. healthy control) as the between-subject factor, and region-of-interest as a within-subjects repeated factor (cerebellum, hippocampus, frontal cortex, and striatum). These regions were chosen based on previous studies highlighting these areas as important for alcohol-induced neurodegeneration. Post-hoc analyses compared VT values between AUD and control groups by brain region. A subset (n = 33; n = 21 with AUD) of subjects completed the Cogstate computerized cognitive battery. Performance on three cognitive tasks (verbal learning and memory, executive function, and motor function) was compared between the AUD and healthy control group using independent-samples t-tests. Exploratory analyses of relationships between VT and cognitive task performance were conducted using linear regressions.

Results: Preliminary analyses indicate a main effect of diagnostic group such that individuals with AUD had significantly lower levels of SV2A availability than healthy controls in the four brain regions of interest (p = 0.008, Cohen’s d = 0.42). Pairwise comparison between groups revealed that individuals with AUD had significantly lower VT in all four regions (cerebellum p = 0.016; Cohen’s d = 0.38; hippocampus p = 0.003, Cohen’s d = 0.46; striatum p = 0.007, Cohen’s d = 0.42; frontal cortex p = 0.043, Cohen’s d = 0.31) compared to healthy controls. Overall performance on cognitive tasks did not differ between the AUD group and the healthy control group (p > 0.206). However, exploratory analyses revealed that VT values were positively related to cognitive performance in the AUD group only. Specifically, lower hippocampal VT values were significantly related to fewer correct responses on the verbal learning task (p = 0.036, R2 = 0.21).

Conclusions: Consistent with our hypothesis, preliminary analyses showed that SV2A availability was significantly lower in individuals with AUD compared to control subjects in cerebellum, hippocampus, striatum, and frontal cortex. While individuals with AUD did not perform worse on cognitive tasks than their healthy control counterparts, lower hippocampal SV2A availability was related to worse verbal learning in the AUD group. These data suggest that AUD may lead to alcohol-induced neurodegeneration that lowers synaptic number and that lower synaptic number is related to poorer neurocognition. This highlights restoration of synaptic function as a treatment target for AUD.

Keywords: alcohol use disorder, positron emission tomography (PET), synaptic density

Disclosure: Nothing to disclose.

Mia Allen*, Emory Lewis, Michael Nader

Wake Forest School of Medicine, Rural Hall, North Carolina, United States

Background: While the vast majority of people with cocaine use disorders (CUD) also co-use alcohol, most preclinical cocaine research does not include alcohol. Thus, the present study aimed to examine how self-administering ethanol prior to cocaine self-administration (SA) influences vulnerability to cocaine reinforcement when compared to controls with no history of ethanol SA.

Methods: Drug-naïve, female and male cynomolgus macaques living in same-sex stable social groups (n = 4/group) served as subjects and the role of social rank was evaluated in relation to vulnerability to cocaine reinforcement. In this experiment, 7 monkeys (4 dominants and 3 subordinates) were trained to SA ethanol and 13 monkeys (6 dominants and 7 subordinates) served as controls and remained ethanol-naïve throughout the study. The ethanol SA monkeys were trained to consume up to 1.5 g/kg of ethanol in one hour before food-maintained responding under a fixed-ratio (FR) schedule of reinforcement; ethanol SA did not affect food-reinforced responding. Next, both groups underwent cocaine acquisition, in which ascending doses were substituted for food.

Results: The results demonstrated that during initial exposure to cocaine, the subordinate control monkeys were marginally more vulnerable to cocaine reinforcement when compared to dominant control monkeys (p = 0.051). This finding is congruent with previous data from our laboratory and suggests that the chronic social stress experienced by subordinate monkeys increases vulnerability to cocaine SA (Johnson et al., Psychopharmacology, 2024). However, in the ethanol SA monkeys there were no differences between subordinate and dominant monkeys. Comparisons across conditions demonstrated that the subordinate monkeys that self-administered ethanol prior to the session were significantly less vulnerable to cocaine reinforcement when compared to subordinate control monkeys (p > 0.05). No such relationship was seen among the dominant control and ethanol SA monkeys. No sex differences were observed in this study

Conclusions: These results suggest that in subordinate monkeys, ethanol SA before cocaine SA reverses the greater sensitivity to cocaine reinforcement seen in control monkeys when compared to their dominant counterparts. This could be due to the acute anxiolytic effects of ethanol and future studies should be conducted to evaluate whether other anxiolytic compounds are efficacious in reducing cocaine self-administration in populations that experience chronic stress. Moreover, further studies will aim to examine whether ethanol SA influences the potency and reinforcing strength of cocaine (for example, see Allen et al., Drug Alcohol Dep, 2023) and the effectiveness of behavioral and pharmacological interventions aiming to reduce cocaine use.

Supported by F31 DA060614-01 and R01 DA017763-15.

Keywords: Nonhuman Primate Models, polysubstance use, Chronic social stress

Disclosure: Nothing to disclose.

Mehdi Farokhnia*, John Tazare, Claire Pince, Nicolaus Bruns VI, Joshua C Gray, Vincent Lo Re III, David Fiellin, Henry Kranzler, George Koob, Amy Justice, Leandro Vendruscolo, Christopher Rentsch, Lorenzo Leggio

National Institutes of Health, Baltimore, Maryland, United States

Background: Growing evidence from preclinical studies suggest that glucagon-like peptide-1 receptor agonists (GLP-1RAs), approved for treatment of type 2 diabetes mellitus and obesity, may be repurposed to treat alcohol use disorder (AUD). We previously showed that liraglutide and semaglutide, two long-acting GLP-1RAs, reduce alcohol intake in various rodent models. Another method to target the GLP-1 system is to boost endogenous GLP-1 levels by inhibiting dipeptidyl peptidase-4 (DPP-4), the enzyme in charge of GLP-1 degradation. Possible effects of DPP-4 inhibitors (DPP-4Is), also approved for treatment of type 2 diabetes mellitus, on alcohol intake have not been comprehensively explored. Moreover, clinical evidence on the on the potential impact of GLP-1RAs and/or DPP-4Is on alcohol use in humans is scarce.

Methods: We first conducted a set of behavioral pharmacology experiments to test DPP-4I compounds in rodent models of alcohol intake. Specifically, we examined the effects of one peripherally restricted (linagliptin) and one brain penetrant (omarigliptin) DPP-4I on binge-like and dependence-associated alcohol drinking in mice and rats. Glucose challenge tests, with and without alcohol, were also performed to confirm pharmacological effects of these compounds. Next, we sought to investigate the impact of GLP-1RAs and DPP-4Is on alcohol consumption in humans. To do so, we conducted a pharmacoepidemiologic cohort study using electronic health records data from the Department of Veterans Affairs, the largest integrated healthcare system in the United States. Recipients of GLP-1RAs, DPP-4Is, and unexposed comparators were identified and propensity score matched to ensure balance between a wide range of covariates. Severity of alcohol use was assessed via Alcohol Use Disorders Identification Test-Consumption (AUDIT-C); changes in AUDIT-C scores from pre- to post-index assessment were compared between exposure groups, including GLP-1RA recipients, DPP-4I recipients, and unexposed comparators. Differential associations by AUD diagnosis and alcohol use severity at baseline were also examined.

Results: The DPP-4I compounds (linagliptin and omarigliptin), compared to vehicle, had no significant effect on binge-like or dependence-associated alcohol drinking in mice or rats. Both compounds significantly reduced blood glucose levels in the presence and absence of alcohol, confirming target engagement. Consistently with the preclinical data, the human pharmacoepidemiologic study found no differences in alcohol intake between DPP-4I recipients and unexposed comparators. On the other hand, GLP-1RA recipients experienced greater reduction in AUDIT-C scores compared to unexposed comparators [difference-in-difference: 0.09 (0.03, 0.14), p = 0.0025] and DPP-4I recipients [difference-in-difference: 0.11 (0.05, 0.17), p = 0.0002). Effects on drinking were more pronounced among individuals with AUD [GLP-1RA vs. unexposed: 0.51 (0.29, 0.72), p < 0.0001; GLP-1RA vs. DPP-4I: 0.65 (0.43, 0.88), p < 0.0001] and those reporting hazardous drinking at baseline as indicated by AUDIT-C score ≥ 8 [GLP-1RA vs. unexposed: 1.38 (1.07, 1.69), p < 0.0001; GLP-1RA vs. DPP-4I: 1.00 (0.68, 1.33), p < 0.0001].

Conclusions: These convergent findings across rodents and humans, along with prior work, indicate that GLP-1RAs, but not DPP-4Is, reduce alcohol consumption. GLP-1RAs are widely prescribed in clinical practice, have favorable safety profiles, and pending results from randomized controlled trials could be repurposed as a novel pharmacotherapeutic approach for unhealthy alcohol use and AUD.

Keywords: GLP-1, Alcohol, Addiction

Disclosure: Nothing to disclose.

Luca Lazzarato, Megan Bachant, Kerri Budge, Elizabeth Byrnes, Fair Vassoler*

Tufts University, North Grafton, Massachusetts, United States

Background: The United States is in the midst of an opioid addiction crisis, with exposure rates to opioids estimated to be over 3.3% of the population in 2021 alone. In addition to the devastating consequences to the current population, which include addiction, overdose, lost productivity, socioeconomic burdens, and social and familial deterioration, a growing body of evidence indicates that opioid exposure will also affect subsequent generations and contributes to the inheritance of increased susceptibility to develop opioid use disorder. The current set of experiments uses a rat model to determine the impact of adolescent opioid exposure on the expression of the microRNA let-7g in the testes and spermatozoa, because it is known to regulate the mu opioid receptor. We also determine the impact on future generations by examining mu opioid receptor expression and oxycodone self-administration in the offspring.

Methods: Male Sprague Dawley rats were administered increasing doses of morphine (5-25 mg/kg, s.c.) for 10 days during adolescence (P30-39). Rats were then maintained drug free until adulthood (P70-80) at which point they were mated with drug-naïve females. After mating, Sal-F0 and Mor-F0 males were euthanized and the testes and spermatozoa collected. Let-7g was measured via qRT-PCR in these tissues. In addition, oprm1 was measured in the testes in specific cell types (Sertoli, Leydig, spermatogonia, spermatocytes). In the offspring (Sal-F1 and Mor-F1), oprm1 was measured on PND1 from the ventral tegmental area (VTA). In adulthood, one male and one female from each litter were tested for oxycodone self-administration (0.1mg/kg/infusion, 2h sessions).

Results: Results demonstrate increased levels of Let-7g in the sperm (t14 = 3.8; p < 01) and seminiferous tubules (t15 = 2.4; p < 0.05). The almost 8-fold increase in let-7g expression in sperm compared to 1.5-fold increase in let-7g expression in testes may indicate that let-7g in the testes is located within progenitor cells, spermatogonia, or sperm cells and the signal to noise was decreased due to the presence of additional cell types. Moreover, opioid receptor expression on Sertoli and Leydig cells of the testes indicate potential long-term effects of adolescent opioid exposure on cells involved in spermatogenesis. Results from the PND1 brains revealed a significant increase in oprm1 in Mor-F1 animals compared with Sal-F1 animals (p < 0.05). Lastly, male and female Mor-F1 animals self-administered more oxycodone than the respective Sal-F1 control groups. These data revealed a significant main effect of sex F (1, 570) = 39.1, sire F (1, 570) = 106.4, and day F (14, 570) = 13.9. For the males, there was a significant main effect of both sire exposure F (1, 22) = 16.47 and day F (14, 308) = 9.847. For the females, there was a significant main effect of sire exposure F (1, 16) = 4.591, day F (14, 224) = 13.45, and a significant interaction F (14, 224) = 2.04.

Conclusions: This set of data suggests that let-7g in sperm is a persistent mark that may regulate mu opioid receptor expression within the testes. Because it was present in such high levels in the sperm of Mor-F0 animals, we hypothesized that let-7g is involved in the transmission of the molecular and behavioral phenotypes observed in Mor-F1 offspring. Indeed, we measured oprm1 expression from the midbrain of PND1 offspring and saw increased oprm1 transcript in both males and females supporting our hypothesis that let-7g is disrupting expression of the mu opioid receptor protein in F1 offspring, which may underlie the observed increases in oxycodone self-administration during adulthood. Ongoing studies are using transfection of let-7g mimics and inhibitors in spermatozoa, paired with artificial insemination to rescue or recapitulate the offspring phenotype.

Keywords: Epigenetics, microRNAs, Mu opioid receptor, Opioid addiction, transgenerational

Disclosure: Nothing to disclose.

Erica Grodin*, Wave-Ananda Baskerville, Kaitlin McManus, Michael Irwin, Lara Ray

UCLA, Sherman Oaks, California, United States

Background: Insomnia commonly co-occurs with alcohol use disorder (AUD) and predicts poorer outcomes for those with AUD. Insomnia and AUD are individually associated with increases in systemic inflammation. Insomnia and inflammation both serve as risk factors for relapse in AUD. There are notable sex differences in the relationship between sleep and alcohol use and between inflammation and alcohol use. Despite evidence for insomnia and inflammation serving as predictors of relapse and negative outcomes in AUD, little is known about the relationship between sleep disruption and systemic inflammation in individuals with AUD. Based on the associations across inflammation, AUD, and insomnia, this study examined the relationship between self-reported insomnia symptoms and plasma levels of inflammatory cytokines in a sample of treatment-seeking individuals with an AUD.

Methods: This secondary analysis included 101 (61M/40F) individuals with an AUD. Participants were categorized into groups based on their scores on the Insomnia Severity Index: no insomnia (n = 47), subthreshold insomnia (n = 37), clinical insomnia (n = 17). Participants provided blood samples to measure plasma levels of four peripheral markers of inflammation (IL-6, IL-8, TNF-α, and CRP). Inflammatory marker levels were compared between groups through four general linear models. Tukey post-hoc tests were used to conduct pairwise comparisons to identify group differences and correct for multiple comparisons. Bonferroni correction was applied to adjust for multiple comparisons for the four peripheral markers of inflammation (corrected α level: 0.05/4  =  0.0125). Sex and AUD severity were examined as potential moderators through interaction analyses. The relationship between insomnia symptoms and peripheral markers of inflammation was also examined continuously through partial correlations. Following identification of a significant group difference or significant partial correlation, the moderating effect of sex and AUD severity were examined through interaction analyses or through subgroup partial correlations.

Results: There was a significant main effect of insomnia group on log IL-8 levels (F = 6.52, p = 0.002), such that individuals with AUD and clinical insomnia had higher log IL-8 levels compared to both the no insomnia and subthreshold insomnia groups (ps ≤ 0.05). When examined as a continuous variable, there was a significant partial correlation between ISI scores and log IL-8 levels (r = 0.36, p = 0.0003). Sex (F = 5.11, p = 0.008) and AUD severity (F = 3.33, p = 0.04) moderated this relationship, such that men with clinical insomnia and AUD (ps≤0.02) and individuals with severe AUD (ps ≤0.03) had higher log IL-8 levels. The continuous insomnia scores supported these results, such that there was a significant partial correlation between log IL-8 levels and ISI scores in men (r = 0.45, p = 0.0004); however, this was not found in women (r = 0.22, p = 0.20). There was a significant partial correlation between log IL-8 levels and ISI scores in individuals with a severe AUD (r = 0.46, p = 0.0001); however, this was not found in individuals with moderate AUD (r = -0.04, p = 0.83). There were no significant effects of insomnia on IL-6, TNF-α, or CRP levels (ps > 0.56).

Conclusions: This study provides evidence for a specific elevation in IL-8 levels in individuals with AUD and self-reported clinical insomnia, with no support for elevations in other markers of inflammation (IL-6, TNF-α, CRP). Importantly, this relationship was moderated by sex and AUD severity, indicating that those with clinical insomnia and severe AUD or male sex are most vulnerable to the inflammatory consequences of AUD. Overall, this study highlights the nuanced relationship between inflammation and insomnia symptoms in AUD and suggests that IL-8 may be a plausible treatment-sensitive biomarker that can capture the shared mechanisms linking insomnia symptoms and AUD. Moreover, this study indicates that tailored treatments for individuals with co-occurring insomnia and AUD, such as neuroimmune and/or orexin antagonist pharmacotherapies, are warranted.

Keywords: alcohol use disorder, insomnia, inflammation, chemokine, sex differences

Disclosure: Nothing to disclose.

Ming-Fen Ho*, Cheng Zhang, Brandon Coombes, Joanna Biernacka, Michelle Skime, Paul Croarkin, Tyler Oesterle, Victor Karpyak, Hu Li, Richard Weinshilboum

Mayo Clinic, Rochester, Minnesota, United States

Background: Alcohol use disorder (AUD) is the most prevalent substance use disorder worldwide. Excessive alcohol use can impact bodily systems, raising the risk of liver disease, neurodegenerative disease, mental disorders, and more. The Mayo Clinic Center for Individualized Treatment of Alcohol Dependence study recruited 442 patients with AUD. This unique study cohort contains comprehensive clinical information, including alcohol consumption and blood-based multi-omics data. Previously, we identified potential inflammatory markers associated with acamprosate treatment outcomes utilizing multi-omics data. Chronic alcohol consumption is a risk factor for many chronic diseases and conditions. The present study was designed to identify inflammatory markers associated with alcohol consumption, which might provide insight into genetic underpinnings together with their implications for alcohol-associated diseases.

Methods: The Mayo Clinic Center for Individualized Treatment of Alcohol Dependence study recruited 442 patients with AUD. We used timeline follow back (TFLB) to collect alcohol use information within 90 days before blood collection. Plasma samples (n = 410) were available for proteomics assays using the OLINK “Explore Inflammation” panel. Liver enzymes, including gamma-glutamyl transferase (GGT), aspartate aminotransferase (AST), and alanine aminotransferase (ALT) were also measured. Genome-wide associations between genetic variants and plasma TREM2 levels were examined using linear regression, adjusted for sex, age, and study site.

Results: Among the AUD patients we recruited for this study, approximately two-thirds of the patients were men, with an average age of 41.99 ± 11.74 years at the time of consent. The average age of onset for AUD was 29.00 ± 11.91 years, indicating a long history of drinking among the patients. We found that plasma TREM2 was the most significant plasma protein marker positively correlated with recent alcohol consumption metrics, including the total number of drinks (r: 0.312, p: 1.15E-10), number of drinking days (r: 0.29, p: 2.45E-09), and number of heavy drinking days (r: 0.30, p: 1.03E-09) as determined by TFLB. Additionally, GGT, a marker for recent heavy alcohol use, also showed a positive correlation with plasma TREM2 concentration (r: 0.36, p: 1.52E-11). In line with this observation, plasma TREM2 levels were positively correlated with the liver enzymes AST (r: 0.32, p: 4.95E-08) and ALT (r: 0.26, p: 1.13E-05), suggesting a link between plasma TREM2 levels, liver function, and alcohol consumption.

The relationship between alcohol use and TREM2 expression remains unclear. Thus, we applied our “proteomics-informed GWAS” approach to identify potential genes and genetic variants influencing plasma TREM2 levels. The GWAS for plasma TREM2 concentrations revealed a genome-wide significant signal (rs140498820, p: 2.09E-09) on chromosome 19 which mapped to ADGRE4P, a pseudogene. Notably, rs7232, a missense variant in MS4A6A, emerged as one of the top signals (p: 2.59E-07). The minor allele (A) frequency for rs7232 is 37% in European Americans and is associated with higher plasma TREM2 levels. The rs7232 SNP genotype is linked to MS4A6A mRNA expression in whole blood, with the minor allele (A) associated with lower MS4A6A expression (p: 1.44E-12) based on the GTEx database. This mutation substitutes threonine for serine and is predicted to be “possibly damaging” with a score of 0.827 (sensitivity: 0.84; specificity: 0.93) by PolyPhen-2 software (prediction of functional effects of human non-synonymous single nucleotide polymorphisms).

Furthermore, patients with the AA genotype of rs7232 reported more alcohol consumption in the past month (p: 0.0059). We then stratified our data by sex and found that the AA genotype of rs7232 was associated with increased alcohol consumption, i.e. total drinks and number of drinking days in the past month among male subjects.

It has been reported that levels of TREM2 in the blood could serve as a potential circulating marker of non-alcoholic fatty liver disease. Our findings suggest that plasma TREM2 levels are positively correlated with recent alcohol use and with liver enzymes, including GGT, AST, and ALT in patients with AUD. Furthermore, TREM2 is a known risk gene for Alzheimer’s disease (AD). It is important to note that the rs7232 SNP has been previously associated with TREM2 concentrations in both plasma and cerebrospinal fluid (CSF). Pietzner et al. reported in Science in 2021 that rs7232 was linked with plasma TREM2 concentrations (p: 3.0E-190, n = 10708), and this SNP has also been associated with AD risk. Given the association of rs7232 with TREM2 concentrations and Alzheimer’s disease risk, our findings highlight the complex interplay between TREM2 expression, alcohol use, and alcohol-related disease, warranting further investigation.

Conclusions: In conclusion, our study identified plasma TREM2 levels correlated with recent alcohol consumption. Our GWAS analysis highlighted the rs7232 SNP, a missense variant in the MS4A6A gene, which might be one of the key genetic factors influencing plasma TREM2 levels, with the minor allele (A) associated with higher plasma TREM2 levels and lower MS4A6A expression. Additionally, individuals with the AA genotype of rs7232 exhibited higher alcohol use in men. Future research will be required to further explore the complex interplay between genetic variants in TREM2, alcohol use, and alcohol-associated diseases.

Keywords: Proteomics, Multi-omics, Alcohol consumption, biomaker

Disclosure: Nothing to disclose.

Maggie Sweitzer*, Dana Rubenstein, Paolo Mannelli, Francis Keefe, F. Joseph McClernon

Duke University Medical Center, Durham, North Carolina, United States

Background: Tobacco smoking and chronic pain are highly comorbid conditions and have been proposed to influence each other through a reciprocal positive feedback loop, in which pain motivates smoking, and smoking worsens pain over time. Nicotine administration has been shown in both human and animal laboratory studies to have short-term analgesic effects, and smoking is often cited as a way to cope with pain. However, analgesic effects are likely to decrease as tolerance develops, and long-term nicotine use is associated with increased pain severity (i.e., hyperalgesia). In addition, initial smoking abstinence (12-24 hrs) is associated with further increases in sensitivity to painful stimuli and decreased pain tolerance, which could serve as a barrier to smoking cessation. However, prior studies have primarily examined provoked pain responses or increases in spontaneous reports of pain among people reporting no pain after smoking as usual. As such, it is unclear whether smoking abstinence increases pain intensity in people with chronic pain. It is also unclear how longer periods of abstinence from nicotine (e.g., > 3-4 weeks) may impact pain. One study found that withdrawal-induced hyperalgesia may persist for at least 2 weeks, but other evidence hints that pain may decrease with continued abstinence. Here, we present results from a preliminary study of PWS who have chronic back pain examining effects of a) 24 hours smoking withdrawal on ratings of pain; and b) randomization to 4 weeks of smoking very low nicotine content (VLNC) versus normal nicotine content (NNC) cigarettes on weekly ratings of pain. We hypothesized that pain intensity and interference would increase during abstinence relative to smoking as usual. We further hypothesized that any withdrawal-related increases in pain during week 1 of VLNC use would be mild and transient, resulting in no group differences after 4 weeks.

Methods: Thirty-five participants with chronic back pain (mean age = 50.1, 63% female, 29% black, 71% white) who reported smoking at least 10 cigarettes per day (CPD) attended a screening visit followed by an abstinence test, in which ratings of pain intensity and interference, and cigarette craving and withdrawal were assessed after 24 hours of biochemically-verified smoking abstinence. They then smoked their usual brand of cigarettes for a week before attending the first of 5 weekly laboratory visits (V1 to V5). At the conclusion of V1, which served as a smoking as usual comparison to the abstinence session and a baseline for subsequent weekly visits, participants were randomized (double-blinded) to smoke either VLNC or NNC Spectrum investigational cigarettes for the next 4 weeks. Study cigarette adherence was incentivized and biochemically verified. Severity of nicotine dependence, withdrawal symptoms, and ratings of pain intensity and interference were assessed at each weekly visit. Paired t-tests were used to compare ratings of pain intensity and interference during the initial abstinence test versus baseline. To examine effects of VLNCs versus NNCs on pain ratings over time, mixed models with random intercepts were used with week, cigarette group, and their interaction entered as primary predictors, controlling for sex and baseline pain intensity and nicotine dependence.

Results: Symptoms of craving and withdrawal were significantly increased during abstinence compared with baseline (p < 0.005 and p < 0.001, respectively), but there was no difference in pain intensity during abstinence (mean = 4.2) versus baseline (mean= 4.6) or in pain interference (mean = 3.4 during abstinence vs 3.8 at baseline). Similarly, there was no increase in pain, craving, or withdrawal during the first week of study cigarette use relative to baseline for either group. When examining changes across the 5 weekly visits, significant group by week interactions was observed for pain intensity, (p < 0.01) and nicotine dependence (p < 0.001). Pain intensity decreased across weeks for the VLNC group (p = 0.001), with no change in the NNC group. Pain interference similarly decreased for the VLNC group (p < 0.01) but not the NNC group, but the interaction term was not significant. Nicotine dependence decreased in the VLNC group (p < 0.001), but increased in the NNC group (p < 0.01).

Conclusions: Contrary to expectations, we found no evidence of withdrawal-related increases in pain intensity or interference either as a function of 24 hours smoking abstinence or during initial transition to VLNCs. Instead, reducing nicotine use by switching to VLNCs for four weeks resulted in a small but significant decrease in ratings of pain intensity. Previous work has shown that momentary ratings of pain increase prior to smoking and decrease after smoking, which is likely to contribute to negative reinforcement and the belief that smoking helps to cope with pain. However, our results suggest that reducing nicotine use with VLNCs may help to quickly disrupt this cycle, without imposing undue withdrawal-related burden. There is a critical need for effective smoking cessation interventions in patients with chronic pain, and a brief transition to VLNCs may serve as an intermediate step for patients preparing to quit. Further research is warranted to examine the effects of VLNCs on pain and smoking cessation outcomes in a larger sample over a longer period of follow-up.

Keywords: chronic pain, nicotine, tobacco use

Disclosure: Nothing to disclose.

Renata Marchette*, Emma Frye, Lyndsay Hastings, Adriana Gregory-Flores, Aniah Matthews, Lucy Ward, Hugo Tejeda, Leandro Vendrusculo, George Koob

Neurobiology of Addiction Section, NIDA IRP, Baltimore, Maryland, United States

Background: Chronic opioid use leads to hyperalgesia, increased pain sensitivity, and lower pain tolerance during withdrawal, which is hypothesized to contribute to continued opioid intake and escalation of use. Hyperalgesia is mediated by the descending inhibitory system and its reciprocal connections with cortical and limbic brain structures. The periaqueductal grey matter (PAG) is a midbrain structure part of the descending inhibitory system and integrates negative emotions with autonomic and neuroendocrine responses. The dynorphin /κ-opioid receptor (DYN/KOR) system is highly expressed in the PAG but its role in the neurobiology of hyperalgesia remains to be elucidated.

Methods: To further understand the role of DYN on the emergence of addiction-like behaviors, we gave DYN-Cre mice escalating doses of fentanyl over 4 days and tested for hyperalgesia 8 days after the last injection. We then trained the mice on fentanyl vapor self-administration with extended access, which produces addiction-like opioid seeking. To further characterize the anatomical distribution of the DYN/KOR system in the PAG and its functional involvement in the mediation of hyperalgesia, we used a combination of transgenic animals, molecular techniques, and cell-type specific retrograde tracing.

Results: We found that mice that were more vulnerable to hyperalgesia during protracted fentanyl abstinence showed higher fentanyl intake (F2, 26 = 4.443, p = 0.022) and higher sensitivity to capsaicin-punished drug-seeking (F2, 26 = 4.979, p = 0.015) without changes in motivation (F2, 26 = 1.514, p = 0.24) or on a second hyperalgesia test (F2, 26 = 1.607, p = 0.22). Using RNAscope we found that the highest density of DYN and KOR-positive cells was in the PAG’s ventrolateral portion (F4,42 = 68.18, p < 0.0001). We also found that the majority of the DYN- and KOR-positive cells are glutamatergic: 44.2% of DYN-positive cells were also positive for VGLUT2 while 19.2% were positive for VGAT; 50.9% of the KOR-positive cells were VGLUT2 positive and 29.1% VGAT-positive. Combining a DYN-Cre transgenic line and viral vectors we determined the inputs and outputs of DYN in the ventrolateral PAG. Additionally, using the DYN-Cre mouse line we chemogenetically inhibited the ventrolateral PAG during fentanyl abstinence and in preliminary results, did not observe changes in hyperalgesia (F2, 18 = 0.6369, p = 0.54).

Conclusions: Collectively, our data suggests that hyperalgesia is a predictor of addiction-like behaviors in male and female mice and that fentanyl withdrawal-induced hyperalgesia may be modulated by DYN in the ventrolateral PAG.

Keywords: Dynorphin, hyperalgesia, Addiction-like behaviors, Individual vulnerability, periaqueductal grey (PAG)

Disclosure: Nothing to disclose.

Omid Kardan*, Alexander Weigard, Lora Cope, Meghan Martz, Mike Angstadt, Katherine McCurry, Cleanthis Michael, Jillian Hardee, Luke Hyde, Chandra Sripada, Mary Heitzeg

University of Michigan, Ann Arbor, Michigan, United States

Background: Early substance use initiation (SUI) places youth at substantially higher risk for later substance use disorders. Furthermore, adolescence is a critical period for the maturation of brain networks, the pace and magnitude of which are susceptible to environmental influences and may shape risk for SUI. It is therefore important to consider longitudinal brain changes and their environmental correlates when attempting to predict adolescent SUI.

Methods: We examined whether patterns of functional brain connectivity during rest (rsFC), measured longitudinally in pre-and-early adolescence, can predict future SUI. In an independent sub-sample, we also tested whether these patterns are associated with key environmental factors, specifically neighborhood pollution and socioeconomic dimensions. We utilized data from the Adolescent Brain Cognitive Development (ABCD) Study® at baseline (Y0; ages 9-10) and year-two-follow-up (Y2; ages 11-12). The first study involved a total of 461 participants belonging to two groups. The first group comprised participants who had initiated use of any drugs or alcohol (only full doses were considered not single puff or sip) during the Y3-Y4 study periods (ages 12-14) but not earlier (N = 233; SUI group). The second group comprised participants who did not initiate substance use and were matched with the SUI group in terms of age, biological sex (both sexes were included in the study), race/ethnicity, household income, and parental education (N = 228; control group). The second study utilized the remainder of participants with suitable MRI who were not included in the two groups used in the first study and had not initiated substance use at least by the end of Y2 (N = 2854).

Results: In Study 1, multivariate Partial Least Squares (PLS) analysis showed that whole-brain rsFC prior to SUI during 9-10 and 11-12 years of age successfully differentiated the prospective SUI and control groups (p < 0.001, R2 = 0.139, σXY = 0.515, N = 461). This rsFC signature was expressed more at older ages in both groups, suggesting a pattern of accelerated maturation in the SUI group in the years prior to SUI. The rsFC pattern was relatively wide-spread, but most strongly expressed as higher connectivity within the cingulo-parietal and between the two components of the somato-motor network, and lower connectivity within the cingulo-opercular network, between the cingulo-opercular network and the subcortical and the auditory networks, between auditory and subcortical networks, and between the cingulo-parietal and the default mode networks. In Study 2, we found that, adjusted for family socioeconomic factors, expression of this rsFC pattern was associated with higher pollution (standardized adjusted β = 0.10, t = 4.16, p < 0.001, N = 2,854), but not neighborhood disadvantage.

Conclusions: Brain functional connectivity patterns in early adolescence that are linked to accelerated maturation and environmental exposures can predict future SUI in youth.

Keywords: substance use, adolescence, functional brain connectivity, longitudinal, fMRI

Disclosure: Nothing to disclose.

Scott Moeller*, Sameera Abeykoon, Pari Dhayagude, Benjamin Varnas, Jodi Weinstein, Greg Perlman, Roberto Gil, Stephen Fleming, Anissa Abi-Dargham

Renaissance School of Medicine at Stony Brook University, Stony Brook, New York, United States

Background: Individuals with drug addiction show impaired self-awareness of ongoing behavior. These deficits describe problems with metacognition, referring to the ability to monitor and evaluate one’s own cognition and behavior. In the cognitive neuroscience literature, task-based metacognition is operationalized as the degree to which participants’ objective task performance is positively correlated with higher self-confidence in that performance, linked to functional and structural integrity of the anterior prefrontal cortex (PFC) and perhaps especially the rostrolateral PFC (rlPFC). Here, we examined these mechanisms for the first time in opioid use disorder (OUD), using an fMRI metacognition task previously validated in healthy controls (HC). We hypothesized that, compared with HC, OUD participants would show (A) worse metacognition and (B) aberrant confidence-related signals in rlPFC; and that (C) the extent of these behavioral and neural abnormalities in OUD participants would correlate with the extent of illicit opioid use.

Methods: Community samples of OUD (N = 27) and HC (N = 29) participants completed our fMRI metacognition task (3T Prisma, multiband acquisition). On each trial, participants first categorized an image with varying amounts of overlaid white noise as either a face or a house (simple perceptual decision) and rated their confidence regarding whether that judgment was correct (self-evaluation of the perceptual decision). In the “Report” condition, participants rated their confidence on a 6-point scale. In the “Follow” condition, the program specified how participants should respond on the same 6-point scale (active control condition with similar visual and motor demands). Task behavior was analyzed using a linear mixed model (LMM), with trials nested within participants, to predict confidence levels trial-by-trial, with the following predictors: Correctness (i.e., whether the Face/House judgment was correct, per trial: Yes, No), Diagnosis (OUD, HC), and the Correctness x Diagnosis interaction. The fMRI data were modeled using a hemodynamic response function convolved with a boxcar function spanning the time of the confidence rating. The boxcar was separated into two regressors: Report and Follow trials, respectively. The Report regressor was then parametrically modulated by participants’ confidence ratings trial-by-trial, creating first-level contrasts in SPM12. This first-level contrast was then compared between the groups at the second-level. Both ROI analyses of the rlPFC and whole-brain analyses were conducted. In a complementary approach, we also examined overall activation differences between the groups on the first-level contrast of Report > Follow. Finally, we tested correlations between the task readouts and drug use in OUD. Race was controlled in the analyses.

Results: The LMM showed a main effect of Correctness [χ2(1) = 868.86, p < 0.001; higher confidence on correct trials], and a Correctness x Diagnosis interaction [χ2(1) = 21.11, p < 0.001]: OUD participants showed lower levels of confidence in their performance than HC on correct trials [χ2(1) = 3.87, p = 0.049] but not on incorrect trials (p = 0.49). In fMRI ROI analyses, all participants showed a negative trial-by-trial correlation between confidence and brain activity in the bilateral rlPFC [one-sample t(55) > 2.11, p < 0.039]. Activity in these ROIs also differed between the groups, such that the correlation between confidence and activity was more negative in OUD than HC [F(1,50) > 4.35 p < 0.042]. The behavioral and rlPFC ROI measurements were positively correlated in all participants (r = 0.27, p = 0.044). In examining the Report > Follow contrast, the task activated a network of brain regions highly consistent with prior metacognition work, and uniquely showed a case-control difference in a large ventral anterior cluster encompassing the orbitofrontal cortex extending into the striatum (OUD > HC). The behavioral and imaging group effects were especially pronounced among OUD participants who used illicit opioids in the last 3 months, indicating an effect of drug use recency (though not severity).

Conclusions: We hypothesized and found that individuals with OUD had worse metacognition than HC, reflecting a poorer trial-by-trial mapping between task accuracy and confidence in that accuracy. Although we did not initially expect underconfidence in OUD, it is possible that our current treatment-seeking OUD sample may have been experiencing residual withdrawal, stress, and dysphoria (i.e., hyperkatifeia), and many patients have low confidence that they can manage such symptoms. The imaging data revealed case-control differences in the trial-by-trial correlation between activity and confidence in the bilateral rlPFC; these neural signals were further correlated with task behavior and exacerbated among OUD participants who more recently used illicit opioids. Together, these results suggest that confidence-PFC signals, which are critical for effective metacognition as shown in basic research, are disrupted in OUD – particularly in those who recently used opioids – possibly providing a neural basis for the behavioral impairment. Future studies in OUD will need to confirm underconfidence, its neural mechanisms, and potential modulation by recent drug use. It will also be important to examine relationships of metacognition with other laboratory tasks and measures of real-world functioning, such as quality-of-life.

Keywords: Functional MRI (fMRI), Metacognition, Cognitive Neuroscience, clinical neuroimaging research, opioid use disorder

Disclosure: Nothing to disclose.

Kasey Brida*, Emily Jorgensen, Robert Philliips, Catherine Newman, Jennifer Tuscher, Emily Morring, Morgan Zipperly, Lara Ianov, Kelsey Montgomery, Madhavi Tippani, Thomas Hyde, Kristen Maynard, Keri Martinowich, Jeremy Day

The University of Alabama at Birmingham, Birmingham, Alabama, United States

Background: Reelin is a large, secreted glycoprotein with a well-characterized role in brain development and links to numerous neuropsychiatric disorders. While Reelin is abundant in the adult striatum, Reelin’s functional role in this brain region remains poorly characterized. Using a recently generated cellular atlas of the rat nucleus accumbens (NAc) following cocaine experience, we identified Reln mRNA as a marker of cocaine-responsive Drd1+ medium spiny neurons (MSNs). Here, we sought to define Reelin’s role in striatal functions, and to determine its contribution to cocaine behavioral response

Methods: We used RNAscope to validate the finding from single-nucleus RNA sequencing (snRNAseq) to validate that Reln marks cocaine-sensitive cells following acute cocaine or saline (20 mg/kg intraperitoneal) and further characterized Reln expression patterns in the human (n = 2) and rat nucleus accumbens (n = 4M/4F) (NAc). We designed a CRISPR sgRNA targeting the Reln promoter to enable repression of Reln transcription with CRISPR interference (CRISPRi), using a non-targeting lacZ sgRNA as a control. Constructs were delivered steroetaxically to the NAc via lentiviruses, one expressing the sgRNA and one expressing the catalytically inactive spCas9 fused to the transcriptional repressor KRAB-MeCP2. Using this CRISPRi strategy, we assessed transcriptional perturbations of Reln knockdown in the NAc using snRNAseq (n = 2M/2F lacZ, 2M/2F Reln). We followed these findings up with whole cell patch clamp electrophysiology (n = 17/6 (cells/animals) lacZ, 11/5 Reln), cocaine locomotor sensitization (n = 4M/4F lacZ, 8 = 4M/F Reln), cocaine conditioned place preference (n = 6M/8F lacZ, 6M/4F Reln), and cocaine intravenous self-administration (n = 8M/8F lacZ, 8M/F Reln).

Results: RNAscope findings mirrored that of snRNAseq, showing robust Fos response only in Reln + /Drd1+ cells following acute cocaine without changes in Reln mRNA. Characterization of Reln expression demonstrates enrichment in D1-MSNs over D2-MSNs in both rat and human NAc. Single-nucleus RNA-sequencing of the NAc following Reln knockdown reveals changes in the expression of genes important for maintaining calcium homeostasis and indicate loss of cocaine-sensitive Drd1-MSNs (pseudo time score distribution, Kolmogorov-Smirnov test p = 0.0094). To understand Reelin’s role in regulating activity in the NAc, we used whole-cell patch clamp following Reln knockdown. While we saw no changes in passive membrane properties or action potential properties, cells lacking Reln exhibited decreased intrinsic excitability and an inability to maintain sustained firing (input-output curve, p < 0.001, m2-way ANOVA with Sidak multiple comparisons). To assess if Reln serves as a passive marker of cocaine-sensitive cells or facilitates cocaine response, we performed locomotor sensitization and found that Reln knockdown int eh NAc did not affect initial locomotor response to cocaine but abolished sensitization upon subsequent exposure (2-way ANOVA, p = 0.046 gRNA x time). To interrogate Reelin’s role in rewarding aspects of cocaine, we utilized the conditioned place preference paradigm following targeted Reln knockdown in the NAc. Compared to non-targeting gRNA controls, Reln knockdown animals had no preference for the cocaine-paired chamber (2-way ANOVA, main effect gRNA p = 0.014). We further assessed cocaine motivation using intravenous self-administration and found that Reln knockdown dampened this behavior (infusions across days, 2-way ANOVA, gRNA x time, p = 0.0345).

Conclusions: Together, these results reveal a key role for Reelin in striatal function and cocaine reward. Ongoing studies are dissecting Reelin’s role in regulating MSN excitability and other aspects of cocaine-related cellular and behavioral adaptations.

Keywords: cocaine, Neuronal ensembles, Reward, activity-dependent regulation, Reelin

Disclosure: Nothing to disclose.

Tonisha Kearney-Ramos*, Xuejun Hao, Alexandre Guerin, Margaret Haney, Gillinder Bedi

Columbia University Medical Center, New York, New York, United States

Background: The neural mechanisms driving drug-related choices in humans remain poorly understood. In healthy populations, choices for other reinforcers are guided by subjective values (SV; how individuals value a given reinforcer) encoded in a valuation network that includes the ventromedial prefrontal cortex (vmPFC), nucleus accumbens (NAcc), and dorsal posterior cingulate (dPCC). We characterized the neural encoding of SV for smoked cocaine (COC) compared to a non-drug reinforcer, i.e. palatable snack food, in people who smoke cocaine. Our hypotheses were that (1) SV signals for both COC and snacks would be encoded in the canonical valuation network, and that (2) SV signals for COC would be stronger than those for snacks.

Methods: Nine non-treatment-seeking people who smoke cocaine (at least twice per month) participated in a within-subject protocol involving 2 fMRI tasks in counter-balanced order, where they made choices between (1) COC vs. money; and (2) palatable snack foods vs. money. The SV signal was operationalized as the neural correlates of the strength of preference for COC/snack choices on a 5-point scale, using whole-brain univariate analysis with parametric modulation.

Results: Positive SV signals for COC, but not snacks, were observed in the mPFC/dACC, dPCC, and vPCC. SV encoding in all three regions were stronger for COC than for snack foods.

Conclusions: This study mirrors our previous findings that regular cannabis users show neural valuation for cannabis more than snacks. These results suggest that abnormal SV for drug and non-drug reinforcers may be a key feature of problematic drug use. Understanding the relative SV for drug vs. non-drug reinforcers could help improve interventions like contingency management which require patients to be sufficiently motivated by non-drug rewards in order to shift their drug use behavior.

Keywords: cocaine, Reward-based decision-making, food reward, fMRI, Value-based Decision-Making

Disclosure: Nothing to disclose.

Cristina Rivera Quiles, Katie McGrath, Olivia Dodson, Star Fernandez, Barbara Juarez, Michelle Mazei-Robison*

Michigan State University, East Lansing, Michigan, United States

Background: Opioid dependence and addiction are a major health and economic burden, but our limited understanding of the underlying neurobiology limits better interventions. Alteration in the activity and output of dopamine (DA) neurons in the ventral tegmental area (VTA) is known to contribute to drug effects, but the mechanisms underlying these changes remain relatively unexplored. We used translating ribosome affinity purification to identify gene expression changes in VTA DA neurons following chronic morphine and found that Neuromedin S (NMS) is enriched in VTA DA neurons, and its expression is robustly increased by morphine. However, whether all VTA DA neurons express NMS, and their potential functional impact has yet to be determined.

Methods: Adult male and female NMS-Cre mice and wild-type littermates were used. Cre-dependent viral vectors (AAV-DIO-mCherry, AAV-DIO-hM3Dq-mCherry, AAV-DIO-hM4Di-mCherry) were stereotaxically injected into the VTA to allow for DREADD-mediated activation (Dq) or inhibition (Di) of VTA-NMS neurons. Behavioral analyses were completed two weeks after surgery (n = 14-32 mice/group). Locomotor activity was assessed following saline (d1), saline + Clozapine-N-oxide (CNO, 0.3mg/kg, ip, d2-d3), and morphine (15mg/kg) + CNO (d4-d8). A morphine + CNO challenge was done 1 week following d8. For morphine conditioned place preference (CPP) studies, mice underwent a 20 min. pre-test followed by 4 pairing days with 45 min. morning (saline) and afternoon (CNO, 0.3 mg/kg and morphine, 15 mg/kg) sessions and a 20 min. post-test two days after the last pairing. Immunohistochemistry was completed to validate viral expression and specific targeting of the VTA. Repeated-measures two-way ANOVAs were used to determine significant differences (p < 0.05) in behavior.

Results: We found that a subset of VTA DA neurons express NMS ( < 5%) and that they exhibit diverse projection targets, including the NAc. Using DREADDs, we found that activation (Dq) or inhibition (Di) of VTA-NMS neurons did not affect general locomotor activity or elicit CNO-conditioned place preference or aversion. However, we find that both male and female Dq mice exhibit increased morphine-induced locomotor activity (mixed-effects model, day: F(2.885,163.4) = 31.39, p < 0.0001, virus: F(1,61) = 10.12, p = 0.0023, day x virus interaction: F(8453) = 4.991, p < 0.0001; Šídák’s multiple comparison post-hoc test Con vs NMS-Gq: day 7 p = 0.005, day 8 = 0.011). In contrast, Di mice show an opposite effect, as morphine-elicited locomotor activity is decreased (Two-way repeated measures ANOVA, day: F(2.438, 26.8) = 35.30, p < 0.0001, virus: F(1,52) = 4.16, p = 0.046, day x virus interaction: F(8416) = 1.89, p = 0.06). Given that alteration of VTA NMS neuronal activity was sufficient to alter morphine locomotor activity and sensitization, we next sought to determine whether it could also impact morphine CPP. As expected, mice showed a preference for the morphine-paired chamber during the post-test, however, VTA NMS neuron activation did not affect CPP compared to controls (two-way repeated-measures ANOVA: test date: F(1,23) = 5.63, p = 0.026; virus: F(1,23) = 0.80, p = 0.38; test date x virus interaction: F(1,23) = 0.002, p = 0.96). In contrast, inhibition of VTA NMS neurons significantly blunted morphine CPP (two-way repeated-measures ANOVA: test date: F(1,23) = 5.64, p = 0.026; virus: F(1,23) = 4.55, p = 0.044; test x virus interaction: F(1,23) = 4.53, p = 0.044 followed by Fisher’s LSD post-hoc test: Con pre-test vs. post-test p = 0.0035, Con post-test vs. NMS-Gi post-test p = 0.0045). Modulation of VTA NMS neuronal activity did not alter cocaine-elicited behavior, suggesting that activity of this subpopulation of VTA neurons may be particularly important for opioid responses.

Conclusions: Here we show that VTA NMS-expressing neurons play an important role in morphine-elicited behavior. However, it is not clear whether NMS release from these cells is critical for behavioral results. To test this, we have designed a Cre-dependent CRISPR/Cas9 viral vector to knockout NMS expression in VTA DA neurons (via infusion into DAT-Cre mice) as well as floxed-NMS transgenic mouse line. We are currently validating these models and will use them to determine whether NMS expression in VTA neurons is necessary for morphine-induced locomotor activity and morphine conditioned place preference. Overall, our work is the first to implicate NMS in drug behaviors and to identify this peptide as a marker of a functionally relevant neuronal sub-population in the VTA.

Keywords: Ventral Tegmental Area (VTA), morphine, Dopamine, conditioned place preference, neuropeptides

Disclosure: Nothing to disclose.

Lucas Albrechet-Souza*, Linh Ha, Tiffany Wills

LSU Health Sciences Center, New Orleans, Louisiana, United States

Background: Adolescent alcohol use is a strong predictor of developing alcohol use disorders (AUD) later in life. Adolescence is also a critical period for the onset of affective disorders, which can contribute to problematic drinking behaviors and relapse, particularly in females. One proposed mechanism underlying AUD is negative reinforcement, where drinking temporarily alleviates the negative emotional states experienced during abstinence. Additionally, heightened pain sensitivity may contribute to AUD, as individuals with a history of chronic alcohol use often report more severe pain and a greater tendency to use alcohol for pain management compared to social drinkers. Previous research in our laboratory has shown that chemogenetic stimulation of the dorsolateral bed nucleus of the stria terminalis (BNST) promotes negative affect-like behavior in adult female mice with a history of adolescent intermittent ethanol (AIE) vapor exposure. This behavioral phenotype was associated with increased activation of the parabrachial nucleus (PBN), a brain region involved in threat assessment and pain-like behaviors. AIE also induces long-lasting mechanical allodynia, as measured by the von Frey test. Building on these findings, the present study investigated whether BNST-PBN projections modulate negative affect and pain-like behavior in adult females with an AIE history.

Methods: The AIE protocol consisted of exposing mice to two 4-day cycles of 16-h ethanol vapor exposure per day, separated by 3 days of no exposure, from postnatal day (PND) 27 to 38. This protocol resulted in blood ethanol concentrations of 200-250 mg/dL. Approximately 10 days after the final ethanol exposure, AAV5-CaMKIIa-hM3D(Gq)-mCherry or AAV5-CaMKIIa-mCherry viruses were administered into the dorsolateral BNST, and the mice were left undisturbed until adulthood. At PND 80, the mice received an i.p. injection of CNO (3 mg/kg) 1h before the novelty-induced hypophagia (NIH) test. Following this, the mice received another CNO injection and were euthanized 1h later for c-fos in situ hybridization in sections containing the PBN. Images were acquired using a ZEISS AxioScan.Z1 slide scanner and analyzed with QuPath software. For retrograde labeling, cholera toxin B was bilaterally injected into the PBN, and the mice were euthanized 7 days later. Brain images containing the BNST were acquired using the slide scanner. A separate cohort of female mice received intra-dorsolateral BNST AAV8-hSyn-DIO-hM3D(Gq)-mCherry or AAV8-hSyn-DIO-mCherry viruses and intra-PBN retrograde AAV-Cre virus 10 days after the final ethanol exposure, and the mice were left undisturbed until adulthood. At PND 95, these mice received an i.p. injection of CNO 1h before the NIH test and another CNO injection 1h before the von Frey test at PND 100. All procedures were approved by the Institutional Animal Care and Use Committee of the Louisiana State University Health Sciences Center and were in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Results: The data show that chemogenetic stimulation of the dorsolateral BNST increased the latency to consume an appetitive reinforcer only in female mice with a history of AIE (n = 7-10/group; 2-way ANOVA: alcohol exposure x virus interaction = 8.25, p = 0.0073), but not in female controls or males of either group. This behavioral change was associated with an increase in c-fos mRNA expression in both the medial (2-way ANOVA: alcohol exposure x virus interaction = 9.82, p = 0.0037) and lateral PBN neurons (2-way ANOVA: alcohol exposure x virus interaction = 4.31, p = 0.0459). The retrograde tracer confirmed that the BNST sends direct projections to the PBN, and chemogenetic stimulation of these projections increased the latency to consume the appetitive reinforcer in the NIH test (n = 7-10/group; 2-way ANOVA: main alcohol exposure effect = 7.23, p = 0.0113; t-test: alcohol/hM3D(Gq) group x alcohol/mCherry group = 2.32, p = 0.035). Additionally, adolescent ethanol exposure induced long-lasting mechanical allodynia in adult females, which was effectively reversed by the stimulation of BNST-PBN projections (3-way repeated ANOVA: alcohol exposure x virus x sessions interactions = 18.98, p = 0.0001).

Conclusions: The PBN plays a crucial role in maintaining homeostasis under stressful or threatening circumstances to ensure survival. Our findings suggest that a history of adolescent alcohol exposure induces sex-dependent changes in the BNST-PBN circuitry. Specifically, adult females with a history of AIE may perceive novel and mildly anxiogenic environments as highly threatening, mobilizing more resources to face or flee the perceived stressor. Supporting this hypothesis, chemogenetic stimulation of BNST-PBN projections significantly increased the latency to consume an appetitive reinforcer in the NIH test and reduced mechanical allodynia in females with AIE history. These results indicate that the BNST-PBN pathway modulates both negative affect and pain sensitivity, potentially facilitating avoidance and escape responses in the context of injury or other threats.

Keywords: adolescent alcohol exposure, Negative Affect, Pain, bed nucleus of stria terminalis, parabrachial nucleus

Disclosure: Nothing to disclose.

William Stoops*, Thomas Shellenberg, Sean Regnier, David Cox, Reuben Adatorwovor, Lon Hays, Danielle Anderson, Joshua Lile, Joy Schmitz, Jennifer Havens, Suzanne Segerstrom

University of Kentucky, Lexington, Kentucky, United States

Background: Reducing problem behaviors is common in medical practice wherein at-risk individuals are counseled to make moderate, sustainable changes in lifestyle, yet for most substance use disorders, complete abstinence is the only accepted treatment endpoint. Limited prospective research has evaluated the health benefits produced by reduced drug use. This study sought to determine how reduced cocaine use impacted markers of immune function. We hypothesized that reduced cocaine use would translate to improvement in these markers.

Methods: Treatment-seeking participants with Cocaine Use Disorder (N = 127) were enrolled in a 12-week clinical trial that used contingency management to reduce cocaine use as verified by benzoylecgonine [BE] negative urine results. Participants were randomly assigned 1:1:1 to High-Value Reinforcers for cocaine abstinence (n = 44), Low-Value Reinforcers for cocaine abstinence (n = 41) or Non-Contingent Control (n = 42). At baseline and 6-week intervals during the trial, blood was drawn and assayed using ELISA for biomarkers of immune health (i.e., Interleukin [IL]-6; IL-10; C Reactive Protein; TNF-alpha; Granulocyte-Colony Stimulating Factor; Chemokine Ligand 5 [CCL5]). Delayed type hypersensitivity [DTH] to candida yeast was measured by induration and erythema at the same time points. Time- and group-varying percent of BE-negative urine samples were analyzed as the primary predictor of these markers using generalized linear models.

Results: There was a statistically significant main effect of group assignment (F = 4.36; p = 0.013) on the provision of BE-negative urine samples. Participants in the High-Value group (Estimated Marginal Mean [EMM] = 46% negative) were significantly more likely to provide a BE-negative urine sample throughout the course of the study compared to controls (EMM = 24% negative; p = 0.012; Odds Ratio [OR] = 2.72). Participants in the Low-Value group (EMM = 23% negative) were not more likely to provide a BE-negative urine sample compared to Controls (p = 0.874; OR = 0.926). In the High-Value group, erythema and IL-6 increased while IL-10 and CCL5 decreased (Chi-Square values > 4.6; p values < 0.05), suggesting change in immune response for participants who significantly reduced their use but did not attain complete abstinence.

Conclusions: As has been demonstrated in numerous studies, contingency management reduced cocaine use, with the highest magnitude reinforcers promoting significantly greater reductions in use relative to lower value or control conditions. Immune changes as a function of group were largely observed in the High-Value group, the group that experienced the greatest reduction in cocaine use. Specifically, antiviral and anti-inflammatory regulatory cytokines were lower, whereas a marker of systemic inflammation and induced erythema were higher, in that group relative to Control. Although the full implications of these changes cannot be understood given the relative paucity of mechanistic information about immune function in people with Cocaine Use Disorder, these results support the idea that reducing cocaine use increases immune activation, possibly reflecting rebound from cocaine-related immunosuppression. These data add to a growing body of literature suggesting that cocaine use reduction is a viable treatment target with potentially biopsychosocial health benefits.

Keywords: cocaine, Clinical trial, Immune

Disclosure: Anebulo: Consultant (Self). Myosin: Consultant (Self).

Ethan Anderson*, Evgeny Tsvetkov, Makoto Taniguchi, Antonieta Lavin, Christopher Cowan

Louisiana State University, Baton Rouge, Louisiana, United States

Background: Chronic opioid use leads to long-lasting increases in drug-seeking behavior; however, the causal molecular and cellular mechanisms responsible are not fully understood. Self-administration of addictive substances like heroin can couple the rewarding/euphoric effects of the drug with drug-associated cues, and opioid cue reactivity contributes to relapse vulnerability in abstinent individuals recovering from an opioid use disorder (OUD). Opioids are reported to alter the intrinsic excitability of medium spiny neurons (MSNs) in the nucleus accumbens (NAc), a key brain reward region linked to drug seeking, but how opioids altered NAc MSN neuronal excitability and the impact of altered MSN excitability on relapse-like opioid seeking remain unclear.

Methods: First, rats underwent heroin or saline self-administration. NAc tissues were then harvested bilaterally immediately after the 12th saline or heroin self-administration session and processed for immunoblotting with an anti-SCN1b antibody. We next infused a shRNA expression viral vector that reduces SCN1b levels (AAV-shSCN1b) or a control virus bilaterally into the NAc of male rats using stereotaxic surgery. Three weeks later we measured intrinsic excitability and synaptic transmission using acute-slice electrophysiology. Finally, in a separate experiment, we again infused AAV-shSCN1b or a control virus into the NAc and allowed the rats to self-administer heroin for at least 12 days. Following a 7-day abstinence period, we measured context-associated heroin seeking during extinction conditions for at least 6 days until they reached a criterion of < 25 presses per day for 2 consecutive days. Next, we measured cue-induced reinstatement behavior. All experimental protocols in animal studies were approved by the Medical University of South Carolina’s Institutional Animal Care and Use Committee and were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Results: Here we discovered that heroin self-administration reduced NAc protein levels of the voltage-gated sodium channel auxiliary subunit SCN1b. Viral-mediated reduction of SCN1b increased the intrinsic excitability of NAc MSNs, but without altering glutamatergic and GABAergic synaptic transmission. While reducing NAc SCN1b levels had no effect on acquisition of heroin self-administration or extinction learning, we observed a significant increase in cue-reinstated heroin seeking.

Conclusions: Our findings suggest that heroin-induced reduction of SCN1b in the NAc might promote the strength of heroin-cue associations and relapse-like heroin seeking, and that enhancing SCN1b actions might mitigate opioid cue reactivity in individuals with OUD and attempting to remain abstinent.

Keywords: Heroin, intrinsic excitability, accumbens

Disclosure: NeuroEpigenix, LLC: Founder (Self).

Andrea Maxwell*, Eric Rawls, Anna Zilverstand

University of Minnesota, Minneapolis, Minnesota, United States

Background: Current data on gender differences in drinking behavior posit that men more often drink as a form of positive reinforcement (e.g., in social enhancement), while women more often drink as a form of negative reinforcement (e.g., drinking to cope). Furthermore, recent work supports a gender-specific causal effect of low social support on problematic alcohol use in women only. Converging evidence has also linked increased alcohol use to an increased response of the orbitofrontal cortex (OFC) to stressful cues in women but to reward-related stimuli in men. Together, this work suggests that stress and social relationships may drive problematic alcohol drinking behavior differently between men and women; however, the interaction between these factors, and the role of the OFC in these relationships, has yet to be tested. Here, we test the relationship between stress, social relationship quality (SRQ), OFC reactivity and gender on alcohol use in a sample with problematic alcohol drinking.

Methods: We used a subsample of the publicly-available Human Connectome Project data consisting of men and women who endorsed at least one DSM-IV symptom of alcohol dependence or abuse (N = 488; 41.60% women). Using PROCESS v4.3 in R, we tested a moderated moderation model (i.e., three-way interaction), of the effect of SRQ on drinks per drinking day (DPDD), with self-reported binary gender and perceived stress as moderators. We first tested a model including averaged ‘Global’ SRQ as the focal predictor, with follow-up analyses testing subdomains (i.e., social support, companionship, and perceived distress) and subscales (e.g., loneliness, friendship) of SRQ as the focal predictor, accounting for multiple comparisons. We then applied a second moderated moderation model separately in men and women similarly testing the effect of SRQ on DPDD, but with OFC reactivity during an angry/fearful faces task and perceived stress as the moderators. In all models, age, race, ethnicity, and income were included as covariates, and heteroskedasticity-consistent standard error (HC3) and 95% bias-corrected bootstrapping with 5000 iterations were used.

Results: The overall model using averaged ‘Global’ SRQ was significant F(7.18, 476), p < 0.001, R2 = 0.14), and the ‘Global’ SRQ X Stress X Gender interaction was significant (b = -0.25, t(476) = -2.45, p = 0.02). Conditional effects indicated that in men with moderate to high stress, increased SRQ was associated with increased DPDD. This effect was driven by the subdomain ‘Companionship’ (b = -0.37, t(474) = -2.50, p = 0.01) and the subscale ‘Friendship’ (b = -0.20, t(471) = -2.4, p = 0.02), such that higher levels of friendship were linked to DPDD in men with moderate to high stress. Interestingly, we found that while in both men and women increased friendship was associated with increased DPDD, in men, this effect was strengthened with increasing stress levels, while in women, this effect was weakened with increasing stress levels. Additionally, there was a gender-specific effect for loneliness (b = -0.17, t(471) = -2.04, p = 0.04), such that in men with high stress levels, lower loneliness was associated with increased DPDD, although this did not survive Bonferroni correction. The overall model in men including ‘Loneliness’ as the focal predictor and OFC reactivity and stress as moderators was significant F(3.40, 226), p < 0.00, R2 = 0.17), as was the Loneliness X Stress X OFC reactivity interaction (b = 0.12; t(226) = 1.99, p = 0.048). Conditional effects indicated that in men with moderate to high stress and moderate to high OFC reactivity, lower loneliness was associated with increased DPDD.

Conclusions: We found that while increased SRQ, specifically higher friendship, is associated with increased alcohol drinking behavior in both men and women with problematic alcohol use, increasing stress levels exacerbate this relationship in men but dampen the relationship in women. This may be driven by a tendency of men under stress to engage in drinking with friends, therefore strengthening the association between friendship and alcohol use, while women under stress are less likely to drink within a social context, therefore weakening the impact of friendship on alcohol use. Furthermore, we found that in men with low loneliness and moderate to high OFC reactivity, increased stress was associated with increased DPDD, suggesting that increased OFC reactivity during a social task may be a biomarker underlying the behavioral phenotype of social drinking in men under stress. Together, these findings suggest that (1) men with problematic alcohol use drink within a social context to reduce stress (because the association between friendship and drinking strengthened with increasing stress), suggesting that social relationships do not reduce drinking to cope in men, and (2) in women, drinking to cope was buffered by social relationships (because drinking increased with stress only in women with low friendship levels). Accordingly, Alcohol Use Disorder treatment should use gender-specific approaches. Treatments for men may entail developing skills to mitigate drinking to cope in a social context, while those for women may address developing supportive friendships resilient to stress.

Keywords: Stress, Social Support, Gender differences, alcohol use disorder

Disclosure: Nothing to disclose.

Mark Namba*, Julia Wiafe-Jackson, Joshua Jackson, Jacqueline Barker

Drexel University College of Medicine, Department of Pharmacology and Physiology, Philadelphia, Pennsylvania, United States

Background: Substance use and Human Immunodeficiency Virus (HIV) are persistent public health problems throughout the world. People living with HIV (PLWH) use drugs at higher rates than the general population and drug use facilitates the pathophysiology of HIV. Using a novel rodent model of chronic HIV infection – EcoHIV –, we recently demonstrated that HIV promotes cocaine seeking and dysregulates extinction of cocaine conditioned place preference, consistent with findings from other rodent HIV models. However, it is not known whether chronic HIV infection promotes drug seeking by enhancing reward motivation or via other neurocognitive process such as impaired inhibitory learning. Further, the impact of HIV on the neural substrates underlying reward learning and subsequent addiction-related outcomes is not well understood. As HIV is known to impact astrocytes, we hypothesized that astrocytes were dysregulated in the medial prefrontal cortex (mPFC) – a key regulator of inhibitory learning – following HIV infection. Thus, the present study aimed to close these knowledge gaps by examining the impact of DREADD-modulation of cortical astrocytes on extinction learning in the EcoHIV model.

Methods: In Experiment 1, male and female C57BL/6J mice (N = 48) that were inoculated with EcoHIV (300-500 ng, i.p.) or 1X PBS (“sham”) were trained to self-administer a 10% sucrose solution for 13 days. Training began on a fixed ratio 1 (FR1) schedule of reinforcement and progressed across test sessions to variable ratio (VR) 2, VR3, and VR5 schedules. Each sucrose reinforcer (20 uL) was paired with a compound light and tone cue. Following self-administration, mice underwent extinction training for 3 weeks, where lever presses no longer delivered sucrose or cues. After extinction training, mice were tested for cue-induced reinstatement of sucrose seeking. Mice then retrained on sucrose self-administration until responding was stable, followed by assessment on a progressive ratio (PR) test in which the required number of lever presses to earn sucrose increased by 4 with each reinforcer delivery. One day after PR testing, mice were sacrificed via transcardial perfusion and fixed brains were harvested for Sox9 and Iba1 immunofluorescence. In Experiment 2, male and female mice (N = 40) received intracranial microinfusions of a Gq-DREADD or control AAV vector into the infralimbic cortex (IL), driven by a truncated GFAP promoter, prior to EcoHIV- or sham-inoculation. Following recovery, mice were trained to self-administer sucrose as in Experiment 1, followed by 5 days of extinction training, where mice received an injection of deschloroclozapine (DCZ; 0.10 mg/kg, i.p.) 30 mins prior to each extinction session. Mice were then sacrificed 60 mins after the last extinction session via transcardial perfusion. All data was analyzed with GraphPad Prism v10 with a significance threshold of alpha = 0.05.

Results: In Experiment 1, EcoHIV did not impact acquisition of sucrose self-administration. In contrast, EcoHIV infection impaired the ability to extinguish sucrose seeking. During extinction training, male and female mice with EcoHIV infection exhibited increased active lever presses across test sessions relative to sham mice (rmANOVA, p < 0.05 for main effects of infection, session, and lever). No significant effect of EcoHIV was observed on cue-induced reinstatement of sucrose seeking. Lastly, no effect of EcoHIV infection was observed among male mice for progressive ratio responding for sucrose, while EcoHIV-infected female mice showed significantly lower reinforcers earned (unpaired t-test, p < 0.05), suggesting that EcoHIV infection may sex-specifically decrease, rather than increase, the reinforcing efficacy of sucrose. In Experiment 2, male and female mice from all groups exhibited similar sucrose self-administration as was observed in Experiment 1. In male mice, infection and Gq-DREADD stimulation did not impact active lever presses during extinction. However, in female mice, Gq-DREADD stimulation impaired extinction learning among sham mice, but improved extinction learning among EcoHIV-infected mice (significant Session x DREADD x Infection interaction; rmANOVA, p < 0.05). Analyses to quantify mPFC and nucleus accumbens (NAc) Iba1 and Sox9 expression, which are microglia- and astrocyte-specific markers, are ongoing.

Conclusions: Together, our results indicate that EcoHIV infection selectively impairs inhibitory learning among males and females, although the role of cortical astrocytes in regulating this process are sex-specific. Contrary to prior work demonstrating enhanced cocaine seeking in EcoHIV-infected mice, we observed EcoHIV little to no effect on the reinforcing efficacy of sucrose among males and a decrease among females. Rather, we observed impaired extinction learning across both sexes. Moreover, Gq-DREADD stimulation resulted in impaired extinction learning among females, which converges on other studies demonstrating enhanced drug seeking in response to Gq-DREADD stimulation of mPFC astrocytes. Importantly, the deleterious effect of Gq-DREADD stimulation in mPFC astrocytes on extinction learning is reversed among EcoHIV-infected female mice, suggesting that sex may be a critically important factor that determines the mechanisms underlying PFC-dependent reward learning impairments induced by HIV infection. These results offer crucial insight into the development of targeted therapies to treat HAND and SUDs among PLWH.

Keywords: HIV, HIV associated neurocognitive disorder, Extinction learning, Inhibitory control, Addiction comorbidity

Disclosure: Nothing to disclose.

Rui Zhang*, Peter Manza, Dardo Tomasi, Weizheng Yan, Sukru Demiral, Ehsan Shokri Kojori, Melanie Schwandt, Natasha Giddens, Gene-Jack Wang, Nora Volkow

National Institute on Alcohol Abuse and Alcoholism, Bethesda, Maryland, United States

Background: Disruptions in sleep and circadian rhythm are highly prevalent in opioid use disorder (OUD) and are a barrier to successful treatment and recovery. Yet little objective data are available to show whether these disruptions and associated changes in brain function persist among individuals in OUD treatment with opioid agonist therapy. If disruptions remain present despite OUD treatment, this information will yield new targets for adjunctive therapy.

Methods: Here we investigated different aspects of rest-activity rhythms (RAR), a proxy measure for the endogenous circadian rhythm, among OUD participants treated with methadone or buprenorphine (MOUD + , n = 33), contrasting with both healthy controls (HC, n = 31) and OUD participants who remained abstinent without medications (MOUD-, n = 9). To measure RAR, we used one-week actigraphy data. Additionally, we examined how RAR was associated with light exposure, which is a strong modulator of the circadian clock, and brain state dynamics during resting-state fMRI. Both sexes are included

Results: MOUD+ exhibited greater sleep-wake irregularity than MOUD- or HC but did not differ in other RAR components including phase timing, physical activity, duration and restfulness of sleep (F2,70 = 6.24, p = 0.003, pBH = 0.012, Eta-squared = 0.151, post-hoc Tukey HSD: all p < 0.048). Among OUD participants, greater sleep irregularity was associated with longer heroin use history (r26 = 0.449, p = 0.017) and lower daytime light exposure (r42 = -0.544, p < 0.001). Compared to HC, OUD exhibited lower fractional occupancy (percentage of occurrence) in a default mode network dominated brain state (t52 = 2.23, p = 0.030, pBH = 0.150, Cohen’s d = 0.614), with those experiencing more pronounced sleep-wake irregularities displaying exacerbated impairments (r23 = -0.550, p = 0.007).

Conclusions: The current study suggests that interventions to increase light exposure may have benefit as an adjunct treatment to help regulate RAR and brain functional network organization in OUD participants.

Keywords: Sleep inconsistency, opioid use disorder, Resting-state fMRI, circadian rhythms, Light Therapy

Disclosure: Nothing to disclose.

Chase Carter, Daniel Christoffel*

UNC Chapel Hill, Chapel Hill, North Carolina, United States

Background: Appropriate feeding behavior is essential to survival and well-being. Both negative and positive reinforcement signals in the brain and body motivate food seeking. The striatum, particularly the nucleus accumbens (NAc), critical regulates motivated behavior and reinforcement learning. While multiple studies demonstrate its important role in regulating homeostatic and hedonic feeding, overarching mechanistic explanations are lacking. Creating a cohesive framework requires consideration of numerous parameters governing eating in distinct situations that relate to the organism’s internal state and experiential history.

Methods: High fat intake: Mice were grouped housed with ad libitum access to standard chow and water. For exposure, individual mice were placed into the same transparent cage without bedding with a 20 min habituation period prior to high fat access. Weight matched mice were randomly assigned to each experimental condition. A single, pre-weighed high fat pellet was provided to the mice in their own cage daily for 20 min at the same time each day for a given experiment. Intake of the high fat diet within that period was measured. All mice with appropriate expression and targeting were included in data presentation and analyses.

Optogenetics: Transgenic mice expressing Cre under a D1 promoter (D1-Cre) underwent intracranial injections of AAVs encoding either channelrhodopsin variant H134R (ChR2) or eYFP control virus into the NAc bilaterally (bregma coordinates: angle 10°, anteroposterior 1.6, mediolateral ±1.5, dorsoventral −4.4) at a rate of 0.1 μl/min. Needles were removed 5 min after infusions were complete. Fiber optic implants were constructed in-house (NA 0.48) and implanted 0.2mm above viral injection site. Blue light stimulation was delivered using a high-powered LED (Prismatix) at 5 or 20 Hz (5 ms pulse width; 3-5 mW at fiber tip) throughout the experimental session. All mice were habituated to tethering in the morning for the 2 days prior to experimental manipulations.

Whole-brain clearing and cFos mapping: Paraformaldehyde-fixed samples were treated with SHIELD reagents from LifeCanvas Technologies to preserve their endogenous fluorescence, protein antigenicity, and nucleic acids. These samples were then immunolabeled. Next, an automated process, performed by LifeCanvas Technologies, registered the samples to the Allen Mouse Brain Atlas. Subsequently, c-FOS segmentation was carried out using a custom convolutional neural network created with Tensorflow (by Google). Finally, based on the previously calculated Atlas Registration, each cell location was projected onto the Allen Brain Atlas to determine c-FOS cell coverage for each brain region. Regional counts were extracted from the raw data and analyzed using one-way ANOVA with custom-made R scripts.

Results: We find that modulation of D1 MSN activity had opposing effects on high fat intake (measured in cal/g) at 5 vs 20 Hz stimulation. Interestingly, while 5 Hz stimulation only affected intake, 20 Hz stimulation altered high fat intake as well as locomotion. Further whole brain analysis revealed region specific differences in cFos levels between 5 Hz, 20 Hz and eYFP controls, suggesting that differential patterns of activity have unique effects on NAc centered networks.

Conclusions: Frequency dependent modulation of D1 MSN has opposing effects on high fat intake. Higher frequency stimulation altered both reward-related consumption as assayed by high fat intake and locomotor behavior, whereas 5 Hz stimulation only affected consumption. These findings suggest that degree of activation can have unique behaviorally effects. Previous work that suggests higher degrees of D1 MSN stimulation induce the release of neurotransmitters other than GABA, such as adenosine and substance P, pointing to a molecular mechanism underlying these findings. Further investigations will be prioritized to gain insight into how hedonic feeding is uniquely regulated by activity dependent release of neurotransmitters from D1 MSNs.

Keywords: Natural rewards, Binge eating, Nucleus Accumbens

Disclosure: Nothing to disclose.

Michael Lutter*

Private Practice, Plano, Texas, United States

Background: Anorexia nervosa (AN) is a severe mental illness characterized by fear or avoidance of eating resulting in abnormally low body weight. While many medical disorders can also affect appetite and body weight, it can often be difficult identify these cases in part because patients with anorexia nervosa often deny intentional restriction of calorie intake. Furthermore, malnourishment can also alter laboratory findings making it difficult to test for many medical conditions that mimic anorexia nervosa. The price of whole exome sequencing has now dropped to the point that it is accessible to patients willing to pay for the service. Whole exome sequencing offers several advantages to identify medical disorders that may affect the treatment of patients diagnosed with anorexia nervosa including being an unbiased approach and not being affected by nutritional status.

Methods: This case series reports the outcomes of 150 consecutive female and male patients with anorexia nervosa who underwent whole exome sequencing. Genetic sequencing data were analyzed for known disease causing mutations as well as rare mutations (MAF < 0.1) predicted to be damaging (CADD phred > 20). Potential medical disorders were confirmed by appropriate diagnostic testing procedures.

Results: Rare medical disorders that mimic symptoms of anorexia nervosa were found in six out of 150 cases (4%). The most common category observed was for inherited metabolic disorders (citrin deficiency, Refsum disease, short-chain acyl-CoA dehydrogenase deficiency). Additional diagnoses include cystic fibrosis, porphyria, and delayed sleep phase disorder. Most disorders presented with mild symptoms that were not detected in infancy.

Conclusions: Whole exome sequencing identified rare medical disorders that affected the treatment of 4% of patients with anorexia nervosa. A broader study of patients is warranted to determine if whole exome sequencing is a cost effective strategy to identify patients with undiagnosed medical disorders that mimic symptoms of anorexia nervosa.

Keywords: anorexia nervosa, Whole exome sequencing, Eating disorders

Disclosure: Nothing to disclose.

Caitlin Lloyd*, Eileen Hartnett, Jessica Ojeda, Xinwei Han, Karin Foerde, Jonathan Posner, Joanna Steinglass

Columbia University Irving Medical Center and New York State Psychiatric Institute, New York, New York, United States

Background: Anorexia nervosa (AN) is a serious eating disorder with one of the highest mortality rates of all psychiatric illnesses. The core behavioral feature of AN is persistent restriction of food intake, which shows little improvement following treatment to restore weight to within a healthy range. Emerging research suggests that imbalance between ventral frontostriatal circuits, which regulate goal-directed behavior, and dorsal frontostriatal systems, associated with habitual behavior, may contribute to the restrictive eating of AN. Patients report restriction to take on habit-like qualities, and restrictive food choices are associated with dorsal striatum activation in AN. At the same time, individuals with AN show poorer performance on tasks engaging neural circuits underlying goal-directed behavior, and reduced functional connectivity within these circuits, compared to HC. Functional disturbances are often grounded in brain structure. While gray matter is reduced brain-wide at acute stages of AN (i.e., when individuals are seriously underweight), prior studies have identified relative enlargement of dorsal frontostriatal regions. Whether structural alterations in AN resolve with full weight-restoration (achievement of a minimally healthy weight) is unclear. Yet, sustained structural abnormalities could contribute to disturbances in eating behavior that continue following inpatient treatment for AN. The aims of this study were to examine 1) structural alterations within frontostriatal circuits among underweight individuals with AN; 2) whether alterations resolve with full weight restoration. The main hypotheses were 1) underweight patients with AN compared to HC would show reduced gray matter in ventral frontostriatal regions, and relatively increased gray matter within dorsal frontostriatal regions; 2) differences would be present following full renourishment of patients.

Methods: Structural T1-weighted scans were obtained from 66 females with AN and 66 age-matched healthy controls across three studies. Participants were aged between 14 and 40 years. Participants with AN were receiving inpatient treatment for weight restoration; the Time 1 scan occurred within two weeks of admission and the Time 2 scan occurred after full weight restoration (body mass index between 19 and 21 kg/m2). For HC, the Time 2 scan occurred approximately 8 weeks after Time 1, to match the average duration between scans of the AN group.

Brain images were segmented into the 84 cortical and subcortical regions of the Desikan-Killany atlas using Freesurfer. Data were harmonized to remove scanner-related variance. Group differences in regional thickness and volume within ventral and dorsal frontostriatal regions of interest (ROIs; ventral: lateral orbitofrontal cortex, medial orbitofrontal cortex, nucleus accumbens; dorsal: rostral middle prefrontal cortex, caudate, putamen) were examined using linear regression models adjusted for total gray matter volume (to account for global brain atrophy in AN) and age. Exploratory whole-brain analyses identified differences present between patients and HC outside of ROIs. These analyses also included age and total gray matter volume covariates, and corrected for multiple comparisons (false detection rate correction applied). Group comparisons were conducted at Time 1 and Time 2; cohen’s d indexes effect size.

Results: At Time 1, relative to HC, patients had reduced cortical thickness within left ventral frontal ROIs (left lateral orbitofrontal cortex, d = -0.50, p < 0.01; left medial orbitofrontal cortex, d = -0.41, p = 0.03), as well as bilateral dorsal frontal ROIs (left rostral middle frontal cortex, d = -0.52, p < 0.01; right rostral middle frontal cortex, d = -0.43, p = 0.02). Patients also had greater volume within the left dorsal striatum versus HC (left caudate: d = 0.38, p = 0.04). Following weight restoration (Time 2), thickness of dorsal frontal regions was reduced among patients compared to HC (left rostral middle frontal cortex, d = -0.35, p = 0.04; right rostral middle frontal cortex, d = -0.43, p = 0.02). Brain-wide analyses identified particular reductions in temporal and parietal cortical thickness among patients with AN relative to HC at Time 1 (d > 0.71, padj < 0.05); reduced thickness of the right middle temporal cortex was observed among patients versus HC at Time 2 (d = -0.64, padj=0.04).

Conclusions: This is one of the largest structural neuroimaging investigations of fully weight-restored patients with AN, allowing for understanding the neural alterations associated with AN that do not resolve with renourishment. Consistent with the hypothesized imbalance between circuits related to goal-directed behavior and habits in AN, underweight patients had reduced thickness of ventral frontal regions, and relatively enhanced thickness of dorsal striatal regions, compared to HC. In contrast to hypotheses, these disturbances resolved with weight restoration. Furthermore, when accounting for global brain atrophy, dorsal frontal regions were reduced in thickness among underweight patients relative to HC. This disturbance among patients was observed following weight restoration, and may contribute to continued eating pathology after inpatient treatment for AN. Brain-wide analyses identified reduced thickness of temporal regions among patients compared to HC, before and after treatment, which should similarly be explored for relevance to illness behavior.

Keywords: anorexia nervosa, Structural MRI, Longitudinal MRI

Disclosure: Nothing to disclose.

Nhien Nguyen, Adam Klie, Huaze Gao, Hannah Carter, Pei-an (Betty) Shih*

University of California, San Diego, La Jolla, California, United States

Background: Anorexia nervosa (AN) is an eating disorder characterized by extreme restriction of calorie intake, avoidance of high-fat foods, and high relapse and mortality rates. Recent polygenic risk score estimates using loci from AN genome-wide association studies have reported significant genetic correlations of AN with lipid and metabolic phenotypes, leading to a conceptualization of AN as a “Metabo-psychiatric” disorder. Although polygenic risk scores revealed an increased probability of metabolic disorders for AN, they do not account for the contribution of the pathological eating behavior of AN to this connection. This study employed genome-wide transcriptomics to identify differentially expressed genes (DEGs) and associated biological pathways between AN and controls under two metabolic states (fasting and postprandial timepoints) to better clarify the genetic underpinning of AN as a metabo-psychiatric disorder.

Methods: A high-fat meal-challenge protocol was administered to 18 AN patients (age: 27 ± 7, BMI: 21.2 ± 3.6) and 9 healthy control women (age: 29 ± 9, BMI: 21.2 ± 2.2). Study participants’ peripheral blood mononuclear cell samples were collected at both fasting timepoint and postprandial timepoint (2 hours after completion of the study meal). The Nextflow RNA-seq pipeline was used to process RNA-sequencing data. Sequenced reads were trimmed, filtered, and aligned to the genome using Trim Galore!, SortMeRNA, and STAR. Gene level quantification was performed using Salmon, and length-scaled counts were used for downstream analysis. DESeq2 was used for the DEG analysis. DEGs were used to conduct pathway enrichment analysis in the Database for Annotation, Visualization and Integrated Discovery (DAVID) platform. Significant DEG-associated pathways were extracted from the Kyoto Encyclopedia of Genes and Genomes (KEGG) pathway database. Statistical significance was set at a p < 0.05 threshold.

Results: At the fasting timepoint, 852 DEGs were identified between AN and controls. These DEGs were overrepresented in 5 KEGG pathways: glycerophospholipid metabolism, metabolic pathways, nitrogen metabolism, sphingolipid signaling pathway, and ECM-receptor interaction. Of the 852 DEGs, 531 were upregulated while 321 were downregulated. The upregulated DEGs were overrepresented in 7 pathways, of which 5 involved metabolism: glycerophospholipid metabolism, metabolic pathways, nitrogen metabolism, terpenoid backbone biosynthesis, metabolic pathways, and glycerolipid metabolism. The downregulated DEGs were overrepresented in 6 pathways: 2 were related to metabolism (alanine, aspartate and glutamate metabolism; ether lipid metabolism); 2 were related to the immune system and infections (herpes simplex virus 1 infection, hepatitis C); while 2 were related to cellular processes and genetic information processing (cell cycle, ribosome biogenesis in eukaryotes).

At the postprandial timepoint, 943 DEGs between AN and controls were identified and were overrepresented in 28 pathways. Most significant of these pathways are related to 3 biological processes: nervous system signaling and neurodegenerative disorders (n = 7), immune responses and inflammation (n = 6), and genetic information processing and signal transduction (n = 5). Of the 943 DEGs, 540 upregulated DEGs were overrepresented in 14 pathways that related to functions of immune responses and inflammation (n = 4), infections (n = 2), and signal transduction (n = 3). The 403 downregulated DEGs were overrepresented in 19 pathways that related to functions of the nervous system (n = 1), neurodegenerative diseases (n = 6), and genetic information processing (n = 4).

Within AN only, consumption of the study meal was linked to 227 upregulated and 206 downregulated DEGs compared to the fasting timepoint. The upregulated DEGs were overrepresented in 5 pathways that related to immune functions and infections (Salmonella infection, measles) and metabolism (other types of O-glycan biosynthesis, mucin-type O-glycan biosynthesis), while the downregulated DEGs were overrepresented in the colorectal cancer pathway. Within healthy controls only, 481 upregulated and 375 downregulated DEGs were observed after the study meal. The upregulated DEGs were overrepresented in 15 pathways that relate to infections (n = 5), immune-related functions (n = 2), and cellular processes (n = 3) while the downregulated DEGs were overrepresented in 6 pathways that primarily relate to cellular processes (n = 2) and environmental adaptation pathways (n = 24).

Conclusions: At the fasting timepoint, DEGs between AN and controls were predominately overrepresented in pathways relating to metabolism including lipid metabolism, supporting the concept of AN as a “metabo-psychiatric” disorder. Two hours after eating, the metabolic pathways were no longer the notable pathways as different pathways emerged. In AN, postprandial upregulated DEGs-associated inflammation and immune response pathways suggest that high-fat meal may trigger an inflammatory cascade in AN, which may explain the well-characterized high-fat food aversion. Postprandial downregulated DEGs in AN revealed the involvement of genes in signaling cascades in the nervous system and neurodegenerative disorders, implying the modulating potential of fat intake in the neurobiology of AN. Our study validates the significance of metabolism in AN pathogenesis and reveals dynamic biological perturbations upon a high-fat meal challenge.

Keywords: Anorexia nervosa, Lipid metabolism, Transcriptomics

Disclosure: Nothing to disclose.

Sasha Gorrell*, Vanessa Brown, Daniel Le Grange, Anne Collins

University of California, San Francisco, San Francisco, California, United States

Background: Anorexia nervosa (AN) is a psychiatric disorder with exceptionally high mortality, significant costs, and for many, a considerably protracted course of illness. The severe and enduring nature of AN is thought to derive from dysregulation in reward processing that may underpin symptoms and create a cycle of chronicity and relapse. Some work shows normative reward processing among those with AN specifically in behavioral paradigms that focus on disorder-related cues. However, other work shows decreased general reward responsivity (e.g., for money cues) as well as elevated compulsivity, which is thought to contribute to more habit-based behavioral responding. To better understand this mixed evidence, a computational framework can be used to describe differences in behavior by distinguishing between model-based (MB) and model-free (MF) learning. MB behavior draws from knowledge of transitions for learning, enabling flexible adaptation of behavior to new input, whereas MF behavior promotes the execution of well-honed behavioral routines without forethought or attention. To date, few studies have investigated the extent to which these two styles of responding may differentiate between those with AN and healthy controls in task paradigms, knowledge that can ultimately inform understanding of reward responsivity – and its potential role in impacting disease course - in this clinical population.

Methods: The sample comprised a sub-set of a larger study of adolescents with AN and healthy controls (n = 45, 25% with AN, 76% female, Mage [SD] = 16.02[1.18]). All participants completed a two-stage Markov decision task (“Spaceship Task”) twice during one study visit; participants were randomized to task-outcome order where they were instructed to work for either a snack (i.e., general cue) or exercise (i.e., disorder specific cue). Multi-level logistic regression models were estimated to evaluate main effects of stage two reward (vs. no reward), transition type (rare or common), and group (AN v. control). We hypothesized that compared to healthy controls, participants with AN would exhibit task performance that was less dependent on state-level behavioral control (MB, goal-oriented) and instead, more based on prior history of reward (MF, automatic).

Results: Contrary to hypotheses, there was no main or interaction effect of group (AN vs. control). For the full sample (not including a grouping variable), there were main effects evidenced for stage two reward (b = 2.48, p < 0.001), and transition type (b = 0.50, p < 0.001). There was also an interaction effect of stage two reward by transition type, such that participants showed model-based learning (b = - 1.66, p < 0.001).

Conclusions: Results indicate that in the current study sample, behavior on a task that is thought to assess MB and MF learning does not distinguish between cases and controls. Further, it appears that in this sample of adolescents, behavior showed both model-free and model-based learning, similar to other normative samples. Although replication is needed given the modest number of adolescents with AN in this sample, the current study suggests that flexible goal-oriented learning in the context of both general and disorder-specific outcomes may not be a specific deficit for adolescents with AN. Further research is warranted to determine how learning styles could be mechanistically involved in the persistent and cognitive emotional aspects of AN and could contribute to longer-term clinical outcomes.

Keywords: Eating disorders, Reward-based decision-making, Behavioral Tasks, anorexia nervosa

Disclosure: Nothing to disclose.

Abanoub Armanious, Youyi Peng, William Welsh, Morgan James*

Rutgers University, Piscataway, New Jersey, United States

Background: The sigma-1 receptor (S1R) system is a promising target for novel therapeutics to treat several psychiatric disorders, including binge eating disorder (BED). Despite this, the development of S1R antagonists for clinical use has remained elusive, in part due to low bioavailability of candidate molecules. We recently designed and validated a lead compound (PW507) that binds S1R at nanomolar concentrations, has excellent blood-brain permeability, and acceptable oral bioavailability. Here, we characterized the efficacy of PW507 in a rodent model of BED. Because BED is observed in persons of diverse body sizes, we compared effects in rats of higher vs. normal weight.

Methods: All experiments were carried out in female Long Evans rats, reflecting disproportionately high rates of BED in women. In Experiment 1, rats were maintained on a high fat diet (45% fat, ad libitum; n = 14) or regular chow (n = 16) for 8w before being given intermittent access (30 min, twice/week, 4w) to sweetened fat (vegetable shortening/10% sucrose; ‘binge’). In Experiment 2, rats underwent the same diet and binge interventions as in Experiment 1. Following the final binge session, rats were trained to lever press for sucrose pellets, before being tested for economic demand for sucrose (n = 16). Separate rats (n = 11) were tested for sucrose responding on low effort (fixed ratio [FR] 1), high effort (FR5), or progressive ratio schedules of reinforcement. Rats in Experiment 1 and 2 were treated with PW507 (0, 5, 10, 15, 20mg/kg; i.p.) 15mins prior to test sessions using a within-subjects design. We also tested the effects of PW507 on locomotor activity in an open field apparatus.

Results: In rats maintained on HFD, PW507 dose-dependently decreased binge-like intake of sweetened fat (ANOVA, p < 0.0001), with the lowest effective dose being 10mg/kg (Holm-Sidak test, p = 0.0059). In HFD rats, PW507 dose-dependently increased demand elasticity for sucrose (ANOVA, p = 0.0171; Holm-Sidak test for 20mg/kg dose: p = 0.0286), reduced FR5 responding (p = 0.0029) and break points (p = 0.0156) for sucrose. In chow control rats, there was no effect of PW507 on binge-like intake or operant responding for sucrose at any of the doses tested (all p’s > 0.05). There was no effect of PW507 on overall locomotor activity in HFD rats (ANOVAs; p = 0.9486).

Conclusions: PW507 reduces binge-like eating and motivated responding for palatable foods, preferentially in rats maintained on a high fat diet. These data support the potential utility of PW507 as a novel compound for the management of BED and other disorders associated with food overconsumption, specifically in higher weight populations.

Keywords: Binge Eating Disorder, sigma-1 receptor, Obesity and eating disorders, Behavioral Pharmacology

Disclosure: Johnson and Johnson: Stock / Equity - Publicly Traded Company (Spouse/Partner)

Carina Brown*, Audrey Nunez, Amanda Bischoff-Grethe, Christina Wierenga

University of California - San Diego, San Diego, California, United States

Background: Altered neurocognitive processes, particularly learning and decision-making, are thought to play a role in the pathogenesis of anorexia nervosa (AN). Prior work has shown that individuals with AN demonstrate some differences in feedback learning and value-based decision-making across reward and punishment contexts, which may contribute to or interact with eating pathology. More recent application of reinforcement learning (RL) frameworks and computational modeling to studies in eating disorder populations shows promise as an avenue towards understanding core mechanisms of neurocognition that may differentiate individuals with AN. In the current preliminary study, we examined latent learning and decision-making during a gambling card game using computationally modeled RL algorithms and explored whether parameters of interest elicited from the best-fitting model could accurately distinguish individuals with AN from controls.

Methods: Adults diagnosed with AN (n = 23) and a comparator group of undergraduates (UG; n = 40) completed a probabilistic gambling card game to win or avoid losing money. Participants could either gamble that a card they pick from their deck would have a higher number than the computer’s card or they could decline to gamble. Further, participants’ cards could be drawn from one of three decks, each with differing probabilities of having higher or lower numbers. Behavior was fit to several RL model variations, where parameters were estimated using a hierarchical Bayesian estimation procedure. Parameters from the best fitting model were extracted for analysis. Raw choice behavior and model parameters were compared between groups. To explore whether latent features of learning are predictive of group membership, latent discriminant analysis was conducted with group specified as the criterion variable. Data was split into training (60%) and testing (40%) sets for classification with permutation testing (10,000 iterations) to examine robustness of classification results.

Results: Groups did not differ on raw choice behavior, including probability of gambles for each deck (p > 0.05), or individual RL model parameters, except for a decay parameter controlling likelihood of repeating recently taken actions (p < 0.001) and at a trend-level for a bias parameter controlling likelihood of gambling (p =.09). Specifically, individuals with AN displayed slightly greater tendencies to repeat recently taken actions and avoid gambling compared to UG. Additionally, linear discriminant function analysis that utilized all model parameters displayed high group classification accuracy (92%), with permutation testing showing that this accuracy was robust (p < 0.001).

Conclusions: Despite no substantive differences in raw choice behavior or individual latent learning parameters between individuals with AN and UG, latent RL learning and decision-making, as a whole, differentiated individuals with AN from UG. These results indicate that differences in learning between individuals with AN and UG may be more subtle; they also show the preliminary utility of a multidimensional approach in diagnostic classification using neurocognitive metrics.

Keywords: anorexia nervosa, Reinforcement learning, computational modeling

Disclosure: Nothing to disclose.

Sarah Gibson, Brooks Brodrick, Jayme Palka, Yi Luo, Terry Lohrenz, P. Read Montague, Carrie McAdams*

University of Texas Southwestern Medical Center, Dallas, Texas, United States

Background: Anorexia nervosa (AN) and bulimia nervosa (BN) are both eating disorders associated with a desire to alter one’s body to meet social expectations of an idealized body. We hypothesized that altered neural processing of social stimuli as well as resting state brain connectivity could help to identify features related to the development and maintenance of eating disorders.

Methods: The multi-round trust game was completed in a 3T MRI scanner in 136 participants, 85 recovering from eating disorders including anorexia nervosa (49) and bulimia nervosa (36), as well as 51 comparison women. This game simulates the formation of a social relationship, and brain responses can be assessed both when that relationship improves and when it deteriorates; the affective salience network was evaluated using a region of interest (ROI) analysis focusing on the affective salience network (amygdala, putamen, insula, cingulate). A subset of participants (95) also completed resting state connectivity analyses.

Results: Both AN and BN showed increased amygdala (L F(2133) = 6.92, p = 0.001; R F(2133) = 8.93, p < 0.001) and putamen activations (L F(2133) = 8.45, p < 0.001; R F(2133) = 6.80, p = 0.001) to negative, but not positive reciprocity, relative to comparison. Interestingly, the network analysis found reduced amygdala activations to negative events were associated with an internalized bias that other people cause good things for you (positive personalizing) while increased putamen activations to good events were associated with a bias that other people cause bad things for you (negative personalizing). The resting state connectivity analysis found increased connectivity between the bilateral lingual gyrus and the left putamen and left pallidium across both the AN and BN cohorts relative to comparison. Further, the strength of the connectivity between the pallidum and left lingual gyrus was correlated with negative perfectionism, the body shape questionnaire, and overall acute severity of disordered eating symptoms.

Conclusions: Alterations in the activation and connectivity of these subcortical regions, the pallidium, putamen, and lingual gyrus, may contribute to both elevated body-focused attention, social reactivity, and negative perfectionism seen in AN and BN. Future work should evaluate whether changes to these regions can improve treatments for eating disorders.

Keywords: social cognition, human neuroimaging, multimodal Imaging, eating disorders

Disclosure: Nothing to disclose.

Guanya Li, Yang Hu, Wenchao Zhang, Gang Ji, Jiangpeng Wei, Peter Manza, Nora D. Volkow, Yi Zhang, Gene-Jack Wang*

National Institute on Alcohol Abuse and Alcoholism, Bethesda, Maryland, United States

Background: Laparoscopic sleeve gastrectomy (LSG), one of the most effective procedures for treating obesity, produces sustained weight-loss, reduction in craving for high-calorie food and improvement in psychiatric symptoms including anxiety and depression. Therapeutic benefits of LSG are partly mediated through its actions on the brain. The reduced appetite seen after surgery has been attributed to functional and structural changes in frontal-mesolimbic circuitry that regulate appetite, reward, and incentive motivation. LSG also alters the function of the habenula, a region involved in negative emotional processing. These alterations contribute to improvements in negative emotional states and emotional-related eating behavior. The fat mass and obesity-associated gene (FTO) is a well replicated locus of obesity across different ages and populations. FTO variants carriers (rs9939609 A allele) retained abnormally high neural sensitivity to food cues following LSG, less weight loss, and greater and earlier weight regain two years after LSG. Previous studies suggest that depression enhances the detrimental effect of FTO variants on BMI. However, it remains unclear whether the FTO gene polymorphism affects the time course of mental health symptomatology and habenula function following LSG.

Methods: Fifty-three individuals with obesity were recruited for LSG surgery including 21 carriers with one copy of the rs9939609 A allele (AT) and 32 non-carriers (TT). Functional magnetic resonance imaging and cue-reactivity task with high- (HiCal) and low-caloric (LoCal) food cues were performed at pre-surgery (PreLSG) and 6 and 12 months (PostLSG-6, -12) after surgery to evaluate the impact of rs9939609 A allele on food cue-induced habenula activation and its psychophysiological interaction (PPI) connectivity. A designated clinician rated anxiety and depression using the Hamilton Anxiety/Depression Rating Scale (HAMA/HAMD). In addition, participants completed the Yale Food Addiction Scale (YFAS) evaluation to assess addictive eating behaviors. A two-way repeated measures ANOVA was implemented to model the effects of group (AT, TT) and time (PreLSG, PostLSG-6, -12) on habenula function, HAMA/HAMD, YFAS and weight loss response.

Results: There were significant interaction (group × time) effects on BMI (F = 4.65, P = 0.012). Post-hoc tests showed TT relative to AT group has greater BMI-loss (t = 2.26, P = 0.029) and lower BMI (t = -2.16, P = 0.037) at 12-months after LSG (Figure 1). There were significant interaction effects on HAMD (F = 5.92, P = 0.004), such that the TT group showed decreased HAMD score at PostLSG-6 (t = 4.58, P < 0.001) and PostLSG-12 (t = 5.03, P < 0.001), but no significant change in AT group (Figure 1). There were also positive correlations between HAMD and YFAS scores in the two groups at both pre-surgery and after LSG. LSG also significantly decreased HAMA and YFAS scores in both AT and TT groups with significant time effects (F = 34.63, P < 0.001), but no interactive group effects. The ANOVA showed significant group effects on habenula response to HiCal vs. LoCal food-cues (F = 16.03, P < 0.001), with its greater activation in the AT group positively correlated with HAMD score at PostLSG-12 (r = 0.49, P = 0.022). AT relative to TT group showed greater food cue-related functional connectivity between habenula and caudate (PFWE < 0.05, cluster-level correction). In the AT group, habenula-right caudate connectivity strength was positively correlated with YFAS score at PostLSG-12.

Conclusions: These findings indicate that the effects of FTO rs9939609 A allele on greater habenula activities and connectivity is associated with the evolution of depressive symptoms following LSG, which might negatively impact eating behaviors and weight loss.

Keywords: Human Neuroimaging, depression and obesity, FTO, Bariatric Surgery, habenula

Disclosure: Nothing to disclose.

Hayden Peel*, Nicco Reggente, Michael Strober, Jamie Feusner

Centre for Addiction and Mental Health, Toronto, Canada

Background: People with anorexia nervosa (AN) commonly exhibit elevated anxiety and atypical reward responsiveness. To examine multivariate neural patterns associated with reward and the impact of anxiety on reward, we analyzed fMRI data from a monetary reward task using representational similarity analysis, a multivariate approach that measures trial-by-trial consistency of neural responses.

Methods: Twenty-five adolescent girls with AN and 22 mildly anxious controls lacking any history of AN were presented personalized anxiety-provoking or neutral words before receiving a reward, and neural response patterns in reward regions were analyzed.

Results: Consistent with our preregistered hypothesis, AN participants showed lower representational similarity than controls during neutral-word rewarded trials. Within groups, controls showed significant representational similarity in reward circuit regions including the left nucleus accumbens, left basolateral amygdala, and left medial orbitofrontal cortex, which were not observed in AN. Further, reward-related prefrontal cognitive control areas – left ventrolateral prefrontal cortex and left dorsolateral prefrontal cortex – showed significant representational similarity in both groups, but a larger spatial extent in controls. Contrary to predictions, there were no significant between-group differences for the effects of anxiety-words on reward representational similarity, and representational similarity did not predict longitudinal symptom change over six months.

Conclusions: Overall, the results demonstrate relatively inconsistent trial-by-trial responses to reward receipt in the neutral state in AN compared with controls in both reward circuit and cognitive control regions, but no significant differential effects of anxiety states on reward responses. These results add to dynamic understandings of reward processing in AN that have potential implications for planning and guiding reward-focused interventions.

Keywords: Representational similarity analysis, Functional MRI (fMRI), Reward, anorexia nervosa

Disclosure: Nothing to disclose.

Connor Haggarty*, Christine Rabinak

Wayne State University, Detroit, Michigan, United States

Background: Emotion dysregulation is considered a core component of posttraumatic stress disorder (PTSD). Cognitive reappraisal is one therapeutic emotion regulation strategy that has been widely studied among individuals with mood and anxiety disorders, and numerous differences in brain activation patterns have been shown between individuals with and without PTSD during tasks of cognitive reappraisal. Prior research among healthy subjects and trauma-exposed individuals suggests that an acute, low dose of Δ9-tetrahydrocannabinol (THC; 7.5mg) could attenuate the neurophysiological discrepancies that exist between individuals with and without PTSD during tasks of emotional processing; however, no dose-response studies have been conducted to determine if 7.5mg is the minimum effective dose to facilitate cognitive reappraisal in individuals with PTSD. This study aimed to investigate the effects of an acute low (5mg) and high dose (10mg) of THC on negative affect and brain activation during emotion regulation among individuals with PTSD.

Methods: Using a double-blind design, thirty-seven participants were randomized to receive an acute dose of placebo (n = 12), 5mg (n = 12) THC, or 10mg (n = 13) THC prior completing a well-established emotional regulation task during functional magnetic resonance imaging (fMRI)

Results: Data were analyzed using eigenvalues extracted from anatomical ROIs, overall, THC reduced activity in the superior temporal gyrus (STG; p = 0.004), superior frontal gyrus (SFG; p = 0.007) medial frontal gyrus (MFG; p = 0.002), insula (p = 0.06), and posterior cingulate (p = 0.05) compared to placebo during reappraisal of negative affect. There was also a dose-dependent linear effect with reduced activity for 10mg THC compared to 5mg and 5mg compared to placebo (SFG p = 0.03; MFG p = 0.01; STG p = 0.01). Furthermore, we found no significant effects of drug (p = 0.11) or dose (p = 0.08) on subjective affect ratings following each picture block.

Conclusions: The results of this study suggest that THC modulates brain activity associated with cognitive reappraisal, highlighting its potential as a tool for enhancing emotion regulation strategies. THC may prove to be a beneficial pharmacological adjunct to cognitive reappraisal therapy in the treatment of PTSD.

Keywords: THC, BOLD fMRI signal, Emotional regulation

Disclosure: Nothing to disclose.

Nicolas Murgueitio*, Kathryn Garrisi, Michelle Shipkova, Margaret Sheridan

University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States

Background: Individual differences in autonomic nervous system (ANS) functioning have been associated with risk for psychopathology across the lifespan. For example, increased vagal tone at rest (a marker of parasympathetic function) is linked with decreased internalizing and externalizing symptomatology in young children. Moreover, high skin conductance levels at rest (a measure of sympathetic function) are associated with increased risk for post-traumatic symptoms and other fear-related disorders. One mechanistic explanation for these associations is the link between disrupted ANS at rest and deficits in emotion regulation. However, the neural correlates of this link are currently underexplored. Little data exists linking ANS at rest and neural function supporting emotion regulation in young children. The present study explored associations between resting vagal tone, resting skin conductance levels, and children’s neural activation during an emotion regulation task in a subsample of participants with processed fMRI data.

Methods: Participants (N = 50; Mage = 6.29 years, SD = 0.84; 27 female) were enrolled in the study between 4 and 7 years of age. Our sample is racially and ethnically diverse with 37% self-identified as Black or African American, 15% as other/mixed race, 7% Asian, 1% American Indian/Pacific Islander, 12% Hispanic or Latino, and 37% Non-Hispanic White. During a laboratory visit participants were instructed to sit quietly without moving for 3 minutes, respiratory sinus arrhythmia (RSA) was assessed using noninvasive ECG recording, and skin conductance levels (SCL) were assessed using electrodes in the participant’s nondominant index and middle fingers. Trained research assistants visually inspected and edited automatic R-peak detection of ECG data (for RSA), and peaks in skin conductance (for SCL). In a separate neuroimaging visit, we employed a commonly used fMRI paradigm to assess emotional reactivity and regulation modified for use in early childhood with the same participants. On each trial during this task children were: 1) told reappraisal stories by an adult prior to viewing a negative picture, mimicking strategies regularly used by caregivers to help children regulate their emotions; or 2) told what a picture would show them and then instructed to passively view negative or neutral pictures. Children rated feeling positive or negative after each trial. Links between resting RSA, SCL, and activation during the task using whole-brain analyses were tested using FSL with cluster-level corrections (voxel-level correction z = 3.1). We examined associations between resting ANS and BOLD responses as children passively viewed negative (vs. neutral) stimuli (i.e., emotional reactivity); and as children viewed pictures directly following a cognitive reappraisal story vs. a simple description of a negative picture (i.e., emotion regulation). Both sexes were used in the study and used as covariates in the analysis.

Results: Preliminary results show that children with higher resting RSA showed heightened recruitment of the fusiform and parahippocampal gyrus when viewing negative compared to neutral images. Additionally, when scaffolded for reappraisal, children with higher SCL showed increased activation in the middle temporal gyrus and frontal pole. Children’s feelings after each trial did not differ based on RSA or SCL.

Conclusions: While children with different levels of ANS activity respond to negative stimuli and perform cognitive reappraisal similarly, the neural regions they recruit to do so are different. We observed that children with higher resting vagal tone (higher RSA), an index commonly associated with positive emotion regulation, also showed increased activity in regions associated with visual processing when viewing negative vs. neutral images. It may be that the enhanced visual processing observed for these children reflects their ability to situate emotions, naturally controlling reactivity. In contrast, children with higher resting SCL, associated with dysregulated emotional responses and psychopathology, showed increased activity in regions commonly recruited in the service of cognitive reappraisal. Suggesting that resting increases in sympathetic activity require compensatory recruitment of regions associated with regulation during the explicit reappraisal task condition. Taken together, these findings suggest that as expected, resting ANS predicts neural processes associated with emotion dysregulation in young children. Future work will be conducted using our entire sample with neuroimaging data (N = 217).

Keywords: Emotional dysregulation, emotional reactivity, autonomic nervous system, fMRI, Brain development

Disclosure: Nothing to disclose.

Matthew Sloan*, Kelly Xiao, Anthony Ngoy, Nikki Bozinoff, Leslie Buckley, Katia Gouveia, Sean Ferkul, Emily Simpkin, Victor Tang

Centre for Addiction and Mental Health, Toronto, Canada

Background: Emotion regulation refers to the ability to change the way one is feeling about a given situation or the way one expresses or acts on emotions. Alcohol use is associated with reduced use of adaptive emotional regulation strategies and individuals with alcohol use disorder (AUD) have impaired emotion regulation compared to controls. However, there are limited data on whether AUD treatments are associated with improvements in emotion regulation and which specific AUD treatments are most likely to influence emotion regulation.

Methods: Emotion regulation was measured using the Difficulties in Emotion Regulation Scale (DERS) before and after three different treatments for AUD. The first treatment (N = 22) consisted of 12 weeks of weekly group cognitive behavioral therapy (CBT) for AUD combined with optional anticraving medication. The second treatment (N = 33) was a 4-week virtual intensive outpatient program (IOP) for AUD in which patients were treated with CBT for AUD, a dialectical behavioral therapy skills group, a health and wellness group, medical management, and remote breathalyzer monitoring. The third treatment (N = 23) was 3-days of symptom triggered withdrawal management delivered via telemedicine followed by 4 weeks of optional psychotherapy. In each treatment program, emotion regulation was measured at baseline and following treatment. We examined mean DERS scores at baseline and following treatment and conducted paired t-tests examining DERS scores pre- and post-treatment and mixed effects models covarying for age and gender.

Results: Emotion regulation significantly improved from baseline to end of treatment following 12 weeks of weekly CBT for AUD (DERS baseline = 104.2, DERS follow up = 83.8, t(21) = 4.5, p < 0.001) and the 4-week IOP (DERS baseline = 95.3, DERS post-treatment = 79.3, t(32) = 4.04, p < 0.001) but not following alcohol withdrawal management (DERS baseline = 96.2, DERS post-treatment = 86.1, t(22) = 1.75, p = 0.094). There was a trend towards improved emotion regulation in individuals who remained abstinent following alcohol withdrawal management (DERS baseline = 87.1, DERS post-treatment = 69.9, t(12) = 2.06, p = 0.062) but not among those who relapsed to alcohol use (DERS baseline = 108.0, DERS post-treatment = 107.2, t(9) = 0.12, p = 0.911).

Conclusions: A range of AUD treatments are associated with improvements in emotion regulation.

Keywords: Emotional regulation, alcohol use disorder, Alcohol Use Disorder - Treatment

Disclosure: Merck: Stock / Equity - Privately Held Company (Self).

Aziz Elbasheir*, Nathaniel Harnett, Ruth Lanius, Alfonsina Guelfo, Rachel Bond, Travis Fulton, Timothy Ely, Timothy McDermott, Vishwadeep Ahluwalia, Maya Karkare, Vasiliki Michopoulos, Bekh Bradley, Abigail Powers, Greg Siegle, Negar Fani

Emory University School of Medicine, Atlanta, Georgia, United States

Background: Racial discrimination (RD) is a chronic stressor in the lives of many Black Americans. RD disrupts a variety of cognitive and emotional processes, which can have downstream effects on behavior. RD can also lead to feelings of disconnection from the body, disrupting interoceptive awareness. Greater engagement in maladaptive health-related, self-regulatory strategies including alcohol use and emotional eating are potential consequences. Neurobiological pathways that underserve emotion regulation and interoceptive awareness may be affected by RD and subsequently affect regulatory behaviors. Particularly, the amygdala and insula, part of the larger salience network, are critical to emotion regulation. The insula supports interoceptive awareness in the presence of emotions and projects related information to the amygdala which is critical for adaptive emotional responses. Furthermore, this neurocircuit has been associated with greater emotion dysregulation, particularly in the presence of aversive stimuli as well as maladaptive regulatory behaviors including heightened alcohol use and eating dysregulation. Previous research has also shown associations between higher self-reported RD and brain pathways associated with attention to emotion and interoceptive processes. Greater frequency of RD has been associated with greater activation of the ventromedial prefrontal cortex in the presence of emotional stimuli as well as disruptions in insula connectivity at rest, particularly with areas associated with visual attention (e.g., middle occipital cortex), attentional control (dorsolateral PFC) and self-referential processing such as the precuneus. Additionally, RD has been associated with greater activation and connectivity of the amygdala. However, to date, no studies have examined associations between RD and insula-amygdala network connectivity during emotion regulation and how potential variability in connectivity may relate to emotion dysregulation and subsequent regulatory behaviors.

Methods: Seventy-two adult (age 22-61 years) Black American women who were recruited from a long-standing trauma study (n = 54) and trauma intervention study (n = 18) completed measures of RD using the Experiences of Discrimination Scale. Subsets of our sample had data on emotion dysregulation [difficulties in emotion regulation scale, DERS (n = 68)], lifetime alcohol use (AUDIT) (n = 61) and eating dysregulation/food addiction [Yale Food Addiction Scale, YFAS (n = 40)]. Participants completed an affective Stroop (AS) task that involved exposure to threat-relevant and neutral distractor stimuli during fMRI. Generalized psychophysiological interaction analysis was conducted to examine associations of RD with seed-to-voxel functional connectivity (bilateral insula and amygdala seed regions) during attention to threat-relevant AS distractor images. A voxel-wise uncorrected height threshold of p < 0.001 and a cluster-forming threshold of p < 0.05 with false discovery rate (FDR) correction was used to determine significance. Scanner type and age were included as covariates in our model.

Results: More frequent RD associated with lower insula connectivity with a prefrontal cortex (PFC) cluster (MNI coordinates x =−32; y = 50; z =-4; k = 158; t(70) = 4.78; FDR-corrected p < 0.001) after controlling for scanner-type and age; RD was not associated with any significant patterns of amygdala connectivity. Extracted insula-PFC connectivity values were negatively associated with DERS total score (r = -0.29 p = 0.02). Follow up analysis with DERS subscales showed that lesser insula-mPFC connectivity corresponded with more limited emotion regulation strategies (r = -0.31, p = 0.01). Insula-PFC connectivity negatively associated with AUDIT lifetime alcohol use (r = -0.28 p = 0.03) and YFAS subscales including withdrawal (r = -0.48 p = 0.002) tolerance (r = -0.52 p < 0.001) and eating despite consequences (r = -0.34 p = 0.04).

Conclusions: Findings indicate that RD linked to lesser functional connectivity between the insula and mPFC, a pathway that is essential to emotion regulation and interoceptive awareness, during performance of an emotion regulation task; this disrupted connectivity associated with more emotion dysregulation, particularly difficulties with strategies, as well as more dysregulated eating behaviors and lifetime alcohol use. These data show a possible mechanism through which RD may negatively affect self-regulatory behaviors such as food and alcohol consumption, creating vulnerabilities for health problems such as obesity and substance use disorders.

Keywords: Emotional dysregulation, Insula Connectivity, Racial minorities, Alcohol consumption, eating behavior

Disclosure: Nothing to disclose.

Romain Durand-de Cuttoli, Alexandra Fink, Austin Baggetta, Giorgio Coricelli, Helen Mayberg, A. David Redish, James Murrough, Xiaosi Gu, Ignacio Saez, Laurel Morris, Jonathan Depierro, Brian Sweis*

Icahn School of Medicine At Mount Sinai, New York, New York, United States

Background: Regret is a poorly understood emotion that may contribute to nearly every mental illness. Regret describes a form of counterfactual thinking where one recognizes alternative decisions could have led to better outcomes. Despite being widely accepted that regret can be detrimental to emotional well-being, no description appears in the DSM nor is pathognomonic for any disorder. Further, little is known about what aspects of regret if any carry utility worth preserving to restore healthy emotional processing and adaptive coping, even if evoking cognitive dissonance. Although psychologists, economists, and neuroscientists have been working toward understanding regret, this has historically occurred outside of a unified framework without a shared lexicon rooted in underlying neurobiology. Currently, there are limited clinical tools that move beyond plain language to describe regret.

Methods: We propose the concept of a Neuroeconomic Regret Inventory (NRI) inspired by cross-species research efforts to resolve attributes of regret into discretely measurable computational units. The NRI characterizes multiple, orthogonal dimensions of regret. Question items examine cognitive domains derived from neuroeconomic principles, including aspects of reinforcement learning, foraging theory, and temporal discounting. Here, we collected data from 350 subjects online via the Prolific platform.

Results: We found subjects could complete the 115-item NRI survey in approximately 15 min. We found that overall, subjects ranked regret related to relationships as the most important life category compared to finance, health, career, and legal decisions. Interestingly, relationship regret-related decisions was the only category that interacted with sex and age. Across the 115 NRI items, questions elicited a wide distribution of responses that scored with varying direction and magnitude. This included within each of the 6 major themes of question items: general questions, regret recognition and registration, feeling and affect, mental operations, reactions and responses, and lastly, anticipation avoidance and learning.

Conclusions: Our vision is that this tool could provide improved neuroeconomic language to be leveraged in multiple settings, e.g., structured interviews to guide psychotherapy strategies and inform computational models of task-based behavior and physiology. By enhancing the diagnostic nosology of psychiatric disorders through a description of one’s decision narrative, we can develop more effective treatments based on a richer understanding of the psychological mechanisms mediating the perception and influence of one’s prior actions.

Keywords: neuroeconomics, Decision Making, Internet survey, emotion, Translational research

Disclosure: Nothing to disclose.

Joao De Aquino*, Gabriel Costa, Julio Nunes, Justin Hudak, Madeleine Odette, Eric Garland

Yale University School of Medicine, West Haven, Connecticut, United States

Background: With evolving legislation and shifting attitudes towards cannabis in the United States, individuals receiving long-term opioid therapy (LTOT) for chronic pain are increasingly using cannabis, potentially to manage affective aspects of pain and thereby reduce opioid consumption. However, the role of emotion regulation capacity in this population’s patterns of cannabis use has not been systematically investigated. This study examined whether deficits in emotion regulation capacity, an endophenotype assessed via psychophysiological measures, is associated with more frequent cannabis use among individuals with receiving LTOT for chronic pain.

Methods: A total of 117 participants (mean age = 50.78 years; 66.38% female; 88.79% White) receiving LTOT for chronic pain were recruited. Cannabis use frequency was quantified by days of use over 90 days. Emotion regulation capacity was indexed using skin conductance response (SCR) and corrugator electromyography (fEMG) during a task requiring passive viewing or cognitive reappraisal of negative stimuli. Pain severity was assessed using the Brief Pain Inventory. With 80% power and an alpha level of 0.05, a moderate effect size (0.47) can be detected in the relationship between emotion regulation capacity and cannabis use frequency.

Results: Larger reductions in SCR (β = 0.012, p < 0.001) and corrugator fEMG (β = 3.896, p = 0.020) difference scores, signifying greater emotion regulation capacity, were associated with fewer days of cannabis use. Pain severity showed a positive, but not statistical significant, association with cannabis use frequency (β = 0.036, p = 0.0615).

Conclusions: These findings suggest that impaired emotion regulation capacity might be a more critical endophenotype associated with frequent cannabis use than pain severity, and lend support to the notion that cannabis use might primarily modulate affect-related pain processes in this population. These results highlight the potential for targeted interventions improving emotion regulation capacity to address heavy cannabis use patterns among chronic pain patients receiving LTOT.

Keywords: cannabis, opioids, Emotional regulation, Pain, psychophysiology

Disclosure: Nothing to disclose.

Anahita Bassir Nia*, Ansel Hillmer, Ardavan Mohammad Aghaei, Brain Pittman, Henry Huang, Nabeel Nabulsi, Deepak Cyril D’Souza

Yale University School of Medicine, New Haven, Connecticut, United States

Background: There is an urgent need to develop novel medications in opioid use disorder treatment to curb the current unprecedented opioid crisis; targeting the endocannabinoid (eCB) system has been identified as a potential therapeutic approach. Extensive preclinical research demonstrated the eCB system abnormalities in animal models of OUD. However, the evidence from human studies is sparse. We present preliminary findings from a Positron Emission Tomography (PET) imaging study to assess in vivo CB1R availability in individuals with OUD, for the first time.

Methods: CB1R availability as indexed by the [11C]OMAR volume of distribution (VT) was measured in individuals with OUD who were on stable opioid agonist treatment and healthy controls with no other psychiatric or substance use disorders (except for tobacco) using High-Resolution Research Tomography (HRRT) PET imaging and the highly specific CB1R radiotracer [11C]OMAR. During the scan, arterial blood was sampled to compute the metabolite-corrected input function. [11C]OMAR VT was estimated with multilinear analysis (t*=30). Based on previous research on OUD neurocircuitry, five Regions of Interest (ROIs) were selected: amygdala, anterior cingulate (ACC), frontal, hippocampus, and insula. ANCOVA models were used to assess the effects, with ROI as the dependent variable and group, sex, and their interaction as predictors.

Results: Eleven (9 males) participants with OUD were compared to 11 healthy sex- and age-matched individuals. In the OUD group, eight participants were receiving methadone with a mean (SD) daily dose of 83.21 (30.58) mg, and three individuals were on buprenorphine treatment with a mean (SD) daily dose of 22.67 (2.31) mg. There was no significant difference in body mass index and history of trauma exposure between the two groups.

Restricting the analysis to males, OUD participants showed lower CB1R availability. Posthoc pairwise contrasts of adjusted means difference percentage (MD) showed a medium to high effect size group difference between males with OUD and HCs in the hippocampus (MD:-14%, d = 1.92, p = 0.07) and insula (MD:-12 %, d = 1.92, p = 0.07). Similar patterns were observed in other ROIs, including amygdala (-8%, d = 1.1, p = 0.30), ACC (MD:-11%, d = 1.61m p = 0.12), frontal (MD:-8 %, d = 1.38, p = 0.18). In contrast, female participants with OUD had higher CB1R availability compared to matched healthy controls, though the sample size was too small for statistical comparisons.

Conclusions: Our results suggest lower levels of CB1R availability in OUD-related ROIs in males with OUD (on agonist therapy) compared with healthy controls. Opposite effects in the small number of females is consistent with reported sex differences in the eCB system and call for a larger sample size study to definitively investigate sex differences. The lower availability of CB1R in individuals with OUD may have significant clinical implications in targeting the eCB system in OUD treatment.

Keywords: Endocannabinoids, Opioid addiction, cannabinoid receptor type 1

Disclosure: Synendos: Advisory Board (Self).

Alex Mabou Tagne*, Yannick Fotio, Daniele Piomelli

University of California Irvine, Irvine, California, United States

Background: Acute pancreatitis (AP) is a prevalent gastrointestinal disorder characterized by severe abdominal pain. The current reliance on opioid analgesics for pain management in AP introduces risks such as abuse, dependence, and possibly exacerbated disease outcomes. Emerging evidence suggests that the endogenous lipid-signaling molecule anandamide at CB1 receptors may contribute to pain modulation in AP. However, the precise mechanisms by which anandamide is mobilized and regulated in AP is not fully understood.

Methods: AP was induced in C57/Bl6 mice of both sexes by caerulein or L-arginine injections, following established protocols. Pharmacological interventions were conducted before AP induction and pain-related behaviors (exploratory activity in the open field and abdominal nocifensive responses to mechanical stimulation) were assessed at various time points as outlined above. We quantified anandamide levels and expression of Faah in pancreatic tissues from mice with AP, using LC-MS/MS and RT qPCR, respectively.

Results: We found that mutant mice lacking FAAH, the enzyme that degrades anandamide, experience reduced abdominal pain following caerulein or L-arginine injection (s). This effect cannot be attributed to developmental compensation because it was phenocopied by administration of the FAAH inhibitor, URB597. Such effects depend on FAAH blockade in the periphery of the body since they are prevented, at least in part, by the peripherally restricted CB1 antagonist AM6545. Supporting this, systemic delivery of the peripherally restricted FAAH inhibitor, URB937, which selectively interrupts FAAH’s actions outside the brain and spinal cord, produced profound analgesia in mice with AP. Of note, the development of AP pain in the L-arginine model of AP was accompanied by enhanced mobilization of anandamide and decreased transcription of FAAH in pancreatic samples from mice with AP.

Conclusions: Collectively, our findings support a role for FAAH-regulated anandamide signaling in AP pain modulation and identify FAAH as a potential target for highly efficacious opiate-sparing medications.

Keywords: Anandamide, FAAH, Pain, nociception

Disclosure: Nothing to disclose.

Catharine Mielnik*, Brandon Oliver, Stefan Vislavski, Andrew Villa, Natalie Zlebnik, Ruth Ross

University of Toronto, Toronto, Canada

Background: Several serious, debilitating, and lifelong conditions, including psychosis in schizophrenia, mania in bipolar disorder, and attention deficit hyperactivity disorder are believed to be related to dysregulation in dopamine (DA) signalling (DA pathologies). Furthermore, the endocannabinoid system is known to be dysregulated in DA pathologies: (1) enzymes in the biosynthesis pathway of 2-arachidnoylglycerol (2-AG), a major endocannabinoid neurotransmitter, were shown to be altered in schizophrenia; (2) DAGL (2-AG synthesis enzyme) levels are decreased in patients with first episode psychosis; (3) MAGL (primary 2-AG metabolism enzyme) expression levels are significantly lower in patients with schizophrenia; and (4) elevated 2-AG was observed in individuals at high risk of psychosis. Despite the mixed trends, the elevation of 2-AG is potentially beneficial in certain contexts, with clinical trials of MAGL inhibitors (MAGLi) currently underway for post-traumatic stress disorder and Tourette syndrome. However, this is worrisome, as our previous studies (unpublished) suggest that increasing 2-AG might be detrimental in hyperDA pathologies.

We have found, through behavioural pharmacology combined with molecular, lipidomic, and PET imaging studies, that there is profound brain-region specific remodeling of 2-AG in hyperDA pathologies (unpublished); highlighting hitherto unrecognized potential detrimental effects of MAGLi. This also revealed a potential novel therapeutic approach to ameliorate hyperDA pathologies with indirect modulation of DA by reducing 2-AG synthesis via DAGL inhibition (DAGLi). However, it was unknown how manipulating 2-AG altered DA dynamics throughout the striatum in pharmacological models of hyperDA. To answer this question, we assessed DA dynamics using fiber photometry, manipulating both DA and 2-AG levels pharmacologically in two brain regions, the dorsal lateral striatum (DLS) and the nucleus accumbens (NAc), in freely-moving C57Bl/6J mice.

Methods: Using a pharmacological model of hyperDAD, we measured changes in acute amphetamine-induced DA release following the modulation of 2-AG (increase – MAGLi, MJN110: 5mg/kg i.p. 120-min pre-treat; decrease – DAGLi, DO34: 30 mg/kg i.p. 120-min pre-treat) in both male and female adult C57Bl/6J mice. Using the fluorescent GRABDA sensor to record DA transients with millisecond resolution, we simultaneously measured both DLS and NAc DA release during open-field exploration. Briefly, adult mice were anesthetized with isoflurane and GRABDA2m AAV was injected (500nl total/hemisphere) in both the DLS and NAc via unilateral injection of AAV9-hSyn-GrabDA-4.4(2m) (gift from Yulong Li, Addgene viral prep #140553-AAV988) at the following coordinates: antero-posterior [NAc]1.1 mm, [DLS] -0.4 mm; medio-lateral [NAc] 1 mm, [DLS] 2.5 mm; dorsoventral [NAc] -3.8 mm, [DLS] 3.2 mm. GRABDA sensor was infused in the target brain areas, and immediately after, an optical fiber was embedded in the same target brain area, 0.1mm above the previously mentioned coordinates. Optical fibres were secured with resin cement. Mice were allowed to recover for 4-weeks while awaiting viral expression. Fiber photometry of DA signals was conducted during an open field test (2-hour assay/recording) while simultaneously modulating 2-AG levels pharmacologically. Samples were collected at 1017 Hz (1017 samples/sec).

Results: Pre-treatment with a MAGLi potentiated amphetamine-induced DA release, increasing the area under the curve (AUC) and the max peak reached in a time-course of DA signal, over 120-min recording. This was true for both the DLS (AUC: p < 0.0001; peak: p = 0.0126) and the NAc (AUC: p < 0.0001; peak: p = 0.0012). Interestingly, DA transient peak width and peak prominence were unchanged following an increase in 2-AG in both the DLS and NAc. Preliminarily, we see influences of sex on the magnitude of potentiation in both brain regions. Conversely, DAGLi dampened amphetamine-induced DA release, decreasing the AUC and max peak. These DA dynamics mirrored the behavioural effects we previously showed, showing correlations between simultaneous open-field and fiber photometry recordings; MAGLi potentiated amphetamine-induced hyperlocomotion, while DAGLi dampened it.

Conclusions: The present study confirms that modulation of DA release occurs following pharmacological modulation of 2-AG in vivo; increase in 2-AG potentiates high DA states, while decreasing 2-AG dampens them. This data confirms our hypothesis that increased 2-AG is detrimental in hyperDA pathologies and provides mechanistic insight into our previously established behavioural and lipidomic effects in high DA states; potentiated hyperactivity, increased reward association, dysregulated habituation to acoustic startle, dysregulated lipidomics with compensations in place to mitigate high 2-AG, and compensatory downregulation of CB1 and MAGL globally. Furthermore, it confirms 2-AG as a novel therapeutic target to ameliorate hyperDA pathologies, reducing 2-AG synthesis via DAGLi, decreasing 2-AG levels, decreasing DA release. This is immeasurable progress in the treatment of hyperDA pathologies, a clinical field that has not evolved since the discovery of antipsychotics.

Keywords: 2-AG, Dopamine, In vivo fiber photometry, Antipsychotic, Schizophrenia (SCZ)

Disclosure: Nothing to disclose.

Natalie Zlebnik*, Brandon Oliver

University of California, Riverside, Riverside, California, United States

Background: Mesolimbic dopamine is responsible for reward-based learning. Specifically, activation of dopamine neurons facilitates the association of cues with the value of related rewards and is responsible for flexibly updating cue-reward learning when conditions are altered. Endocannabinoids are key modulators of dopamine neuron activation and enable dynamic changes in reward-driven dopamine release. Therefore, endocannabinoids, such as 2-AG, may facilitate changes to cue-reward associations during reversal learning when conditions are altered. Here we employ in vivo fiber photometry to monitor dopamine release across the dorsal and ventral striatum at distinct stages of discrimination and reversal learning following treatment with a monoacylglycerol lipase inhibitor, which prevents the degradation of 2-AG.

Methods: Male and female mice (n = 60) were first trained to discriminate between two levers of differing reinforcement probabilities (80% vs. 20%); this was followed by a reversal phase in which the reinforcement probabilities were inversed across levers. During the reversal session, mice were pre-treated with a monoacylglycerol lipase (MAGL) inhibitor JZL-184 (0, 8, 18 mg/kg, ip) to increase synaptic levels of 2-AG. DA was recorded during early and late acquisition, and early, mid, and late reversal sessions. Performance was analyzed to assess mean differences across various metrics on the first day of the reversal phase. Analysis of error probabilities (win-stay and lose-shift) were also conducted. Dopamine release was compared between- and within-subjects across sessions to determine differences in phasic dopamine release following different trial outcomes. Finally, to evidence a causal role, optogenetic stimulation was employed to probe the function of striatal dopamine release during task-relevant epochs.

Results: Results demonstrate significant impairment in behavioral flexibility and dopamine release following elevation of 2-AG. Errors (% correct choices) were dose-dependently increased by JZL-184 administration (p < 0.05 vs. vehicle). Dynamic changes in dopaminergic encoding of correct and incorrect choices was attenuated but could be restored with time-locked optogenetic stimulation of dopamine release.

Conclusions: These results suggest disruption of endocannabinoid signaling hinders reversal learning performance and dysregulates associated mesolimbic dopamine release. These findings give critical insight into the role of the endocannabinoid system in flexible reward-based learning and may have significant implications for the use of cannabinoids for recreational or therapeutic purposes.

Keywords: Ventral Striatum, Dorsal striatum, Endocannabinoids, Dopamine, Cognitive / behavioral flexibility

Disclosure: Nothing to disclose.

Mark Schmidt*, Michel Koole, Jan de Hoon, Lennert Steukers, Anna Katrin Szardenings, Chunfang A Xia, Hartmuth Kolb, Wayne Drevets, Sofie Celen, Guy Bormans, Koen Van Laere

Janssen Research and Development, LLC, Beerse, Belgium

Background: The endocannabinoid (eCB) system is comprised of the G protein-coupled receptors cannabinoid receptor type 1 (CB1R) and type 2 (CB2R), their endogenous ligands – 2-arachidonoylglycerol (2-AG) and N-arachidonoylethanolamine (anandamide), and the enzymes responsible for the synthesis and degradation of these molecules. The eCB is an important modulator of glutamatergic, GABA-ergic, and dopaminergic neurotransmission, among others. 2-AG acts as a potent endogenous agonist of CB1R and CB2R and, like fatty acid amides, is produced ‘on demand’ in response to cellular activity through hydrolysis of specific membrane phospholipids. Monoacylglycerol lipase (MGLL) is the principal enzyme for clearance of 2-AG.

A site-specific PET ligand for the MGLL enzyme could allow new insights on this important component of the eCB in various neuropsychiatric disorders, provide evidence of target engagement and assist in clinical dose selection for compounds developed to inhibit MGLL. Inhibition needs to be sufficient to reduce clearance of 2AG while balancing against the risk of tolerance or desensitization of the CB1R.

Methods: JNJ-70099731 is a potent and selective ligand for the MGLL active site for studying target engagement by inhibitors, with no obvious off-target binding (Cerep panel). [18F]-JNJ70099731 has been extensively evaluated in preclinical models as a candidate PET ligand. Binding was highly selective, high affinity, and specific in vitro. [18F] JNJ-70099731 autoradiography on rat brain section was in good agreement with immunochemistry staining of MGLL on an adjacent section. PET studies in rodents demonstrated reversibility and specificity of the tracer in KO/WT mice and in blocking experiments in rats. In cynomolgus monkeys, a high brain uptake and biodistribution pattern of [18F]-JNJ-70099731 was observed with adequate region differentiation. Binding of [18F]-JNJ-70099731 was found to be blocked in a dose-dependent manner by pretreatment with an intravenous (IV) selective MGLL inhibitor.

We report here the clinical qualification of the ligand in healthy adult males including dosimetry (Part A), brain distribution (Part B), and test-retest (TRT) variability (Part C).

Part A: A biodistribution and dosimetry study was conducted in 3 healthy males (median age 26, range 20-40 years), with74 MBq in the first volunteer, to determine any unexpected distribution of the tracer in human, and 185 MBq for the next two (1). Participants underwent 10 consecutive whole-body PET/CT scans (Siemens Biograph) for up to 5 hours post injection. Delineation of source organs allowed estimation of the normalized cumulated activity and calculation of the corresponding organ doses and effective dose (ED) with Organ Level INternal Dose Assessment (OLINDAv2.1).

Part B: Brain [18F]-JNJ-70099731 PET/CT of 210 minutes (3 segments with 2 breaks of 20 minutes) was obtained after an IV injection of 185 MBq of [18F]-JNJ-70099731 in 5 healthy male participants (median age 28, range 22-34 years). Long scans were needed due to the slow, irreversible brain kinetics of [18F]-JNJ-70099731 that had been observed in NHP studies and confirmed in human. Arterial blood sampling and metabolite analysis of[18F]-JNJ-70099731 was performed to generate the metabolite-corrected arterial input function.

Part C: TRT performance was evaluated by measurement of regional brain kinetics [18F]-JNJ-70099731 for 3 healthy males (median age 28 years range 27-31) from 2 PET scans obtained at least 1 week apart. Participants from Part B could also enroll in Part C.

Results: Part A: The average (±SD) effective dose (ED) value was 23.4 ± 1.3 µSv/MBq. Based on this, the anticipated ED for an injected activity of 185 MBq is 4.33 mSv. The 3 highest organ absorbed doses were found in the spleen (125 µGy/MBq), lungs (55 µGy/MBq) and upper large intestine (55 µGy/MBq). Part B: Good tracer uptake in the brain was observed with peak cortical SUV that ranged from 1.6 to 3.5 across subjects. Dynamic PET scanning with processing was performed estimating Ki and lambda k3 as quantitative endpoints using both a Patlak and two-tissue compartmental modeling approach. Plasma Free Fraction (PFF) was low (< 1 %), as also predicted from preclinical in vivo studies, and consistent with tracer binding to MGLL in peripheral tissues. The impact of specific activity (SA) on quantitation was explored with lower SA injections used for 2 participants. Lower SA did not affect PFF substantially. For the TRT part C, both Ki and lambda k3 demonstrated TRT variability around 10%.

Conclusions: PET imaging with [18F]-JNJ-70099731 resulted in expected radiation burden, with a good brain signal and a brain distribution consistent with the predicted distribution of MGLL. TRT variability is well suited for future PET dose occupancy studies. The slow clearance of [18F]-JNJ-70099731 from the brain and binding of the tracer to MGLL in the periphery may warrant consideration for conducting such studies under steady state conditions.

Reference:

1: Suggested pathway to assess radiation safety of 18F-labeled PET tracers for first-in-human studies. Zanotti-Fregonara P, Lammertsma AA, Innis RB. Eur J Nucl Med Mol Imaging. 2013 Oct;40(11):1781-3. PMID: 23868334.

Keywords: Monoacylglycerol lipase, PET, 2AG

Disclosure: Johnson and Johnson: Employee (Self).

Jeffrey Strawn*, Jeffrey Mills, Heidi Schroeder, Hilary Marusak

University of Cincinnati College of Medicine, Cincinnati, Ohio, United States

Background: Brain endocannabinoids, such as anandamide (AEA), and 2-arachidonoylglycerol (2-AG), play a crucial role in modulating stress responses, anxiety, and depressive symptoms. These bioactive lipids regulate synaptic transmission and neurotransmitter release by binding to cannabinoid type 1 (CB1) receptors, which are primarily located in the central nervous system. Emerging research indicates that brain endocannabinoids are altered in rodent models of anxiety and depression. Additionally, escitalopram and other selective serotonin reuptake inhibitors (SSRIs), commonly used to treat these disorders, can elevate levels of 2-AG. A cross-sectional study in adults with depression found that those treated with SSRIs had elevated 2-AG concentrations. These findings support the idea that the endocannabinoid system plays a significant role in the mechanism of action of effective treatments. This is particularly important for youth with anxiety disorders, as the most common psychopharmacologic treatments, SSRIs, only produce improvement in 3 of 5 patients.

Methods: We conducted an 8-week, double-blind, placebo-controlled trial in adolescents aged 12-17 years with generalized anxiety disorder (GAD). Patients were randomized (1:1) to receive either escitalopram (n = 26) or placebo (n = 25) (NCT02818751). Written informed consent and assent were provided by a parent/legal guardian and patients, respectively. Patients were required to have a Pediatric Anxiety Rating Scale (PARS) score of ≥15 at screening and baseline visits and a Clinical Global Impression of Severity (CGI-S) score of ≥4 to be eligible. Escitalopram was initiated at 5 mg daily for two days, then 10 mg daily for seven days, and then 15 mg daily. At the week 4 and 6 visits, escitalopram could be optionally titrated to 20 mg daily. Efficacy outcomes included changes from baseline in CGI-S and CGI-improvement (CGI-I) scores. Efficacy assessments were administered at weeks 1, 2, 4, 6, and 8, or at early termination. Plasma samples for endocannabinoid assays were available from 24 patients randomized to placebo and 16 patients randomized to escitalopram. At the endpoint, samples were obtained from 20 patients who received placebo and 15 who received escitalopram. Concentrations of AEA and 2-AG were determined via liquid chromatography with tandem mass spectrometry. For continuous outcomes, regression models were employed to determine the changes in CGI-I and CGI-S scores over the course of the trial in relation to changes in AEA and 2-AG concentrations.

Results: Patients were aged 14.7 ± 1.6 (80% female) and had a mean BMI of 22.67 ± 5.3 kg/m2. In escitalopram-treated patients, the increases in 2-AG from baseline to endpoint were significantly associated with improvements in both CGI-S (β = -28.69 ± 11.97, CI: -5.22 to -52.15, p = 0.034) and CGI-I (beta = -32.98 ± 14.08, CI: -5.39 to -60.58, p = 0.037). In contrast, there was no significant association in patients who received placebo (CGI-S: β = 4.65 ± 13.85, CI: 31.79 to -22.48, p = 0.7407; CGI-I β = 0.996 ± 11.45, CI: 23.44 to -21.44, p = 0.932). There was no statistically significant relationship between changes in AEA and treatment response. Baseline 2-AG and AEA concentrations were not significantly associated with changes in CGI-I or CGI-S in either the escitalopram-treated group or the placebo group. Additionally, baseline 2-AG or AEA concentrations were not associated with baseline PARS score (2-AG: p = 0.918; AEA: p = 0.937).

Conclusions: In adolescents with GAD, treatment escitalopram—but not placebo—response was associated with statistically significant changes in 2-AG but not AEA over the course of 8 weeks. The specificity of escitalopram’s effect on 2-AG aligns with evidence that AEA and 2-AG serve distinct functions as “tonic” (sustained/continuous) and “phasic” (short-term) regulators of emotional responses. These findings replicate and extend data in depressed adults receiving SSRI treatment, as well as from lower animal models of antidepressants to SSRI-treated adolescents. This is significant because anxiety disorders emerge during adolescence, a period marked by dynamic fluctuations in endocannabinoid signaling. The results suggest that changes in endocannabinoid signaling may play a role in the mechanism of action of clinically effective antidepressants. Additionally, these findings underscore the importance of the endocannabinoid system in the pathophysiology of anxiety disorders in youth, stimulating further research and intellectual engagement in this area.

Keywords: Anxiety, SSRI, escitalopram, Endocannabinoids

Disclosure: Alkermes, Cerevel, Boehringer Ingelheim, Otsuka: Consultant) (Self). Genomind: Advisory Board (Self)., Springer Publishing, Cambridge University Press, UpToDate: Royalties (Self). Neuroscience Education Institute: Honoraria (Self). Myriad: Contracted Research (Self).

Ana Deutsch*, Abby Lunge, Connor Haggarty, Matthew Hill, Harriet de Wit, Leah Mayo

Hotchkiss Brain Institute, University of Calgary, Calgary, Canada

Background: The endocannabinoid (eCB) system plays a critical role in stress and reward, both of which are impacted by drug use. For instance, chronic stimulant users have higher peripheral levels of the eCB ligand 2-Arachidonoylglycerol (2-AG), as compared to recreational or non-users, while abstinent users have decreased 2-AG and increased levels of the eCB ligand anandamide (AEA). However, little is known about the acute effects of stimulants on eCB levels in humans.

Methods: Using a within-subject, double-blind design, we assessed the acute effects of methamphetamine (20mg), MDMA (100mg), and placebo on plasma eCB levels in healthy human participants (N = 22) during three separate sessions. Blood samples were collected at the peak of drug effects, approximately 165 min post-drug administration, to assess AEA and 2-AG levels. Subjective drug effects were measured with the Drug Effects Questionnaire prior to, and at four time points after (+60 min, +90 min, +180 min, +240 min) drug administration. Participants also provided Visual Analogue Scale ratings of feeling insightful, sociable, confident, lonely, playful, dizzy, friendly, restless, and anxious at the same time points.

Results: Repeated-measures general lineal models revealed significantly lower 2-AG plasma levels in the methamphetamine condition, compared to placebo (F(1,21) = 4.508, p = 0.046). No significant differences in 2-AG levels were found between the MDMA and placebo conditions, nor between AEA levels in any condition. Interestingly, in the placebo condition, disliking the ‘drug effects’ was correlated with higher AEA levels (rs(20) = 0.649, p = 0.001).

Conclusions: Administration of the stimulant methamphetamine was associated with lower levels of the eCB ligand 2-AG. This contrasts with previous studies in chronic stimulant users, which find significantly higher levels of 2-AG, together suggesting that the eCB response to stimulants may change with repeated use. Moreover, ratings of disliking the ‘drug’ effects in the placebo condition were correlated with higher levels of the eCB ligand AEA, suggesting a possible relationship between AEA levels and expectations of subjective drug effects.

Keywords: Endocannabinoids, MDMA, Methamphetamine, Stimulants, endocannabinoid system

Disclosure: Nothing to disclose.

Ryan Wyatt*, Brian Lord, Leslie Nguyen, Stephanie Fitz, Andrei Molosh, Wayne Drevets, Michael Ameriks, Pascal Bonaventure

Janssen Research and Development, San Diego, California, United States

Background: Activation of the endocannabinoid system in the central nervous system (CNS) modulates neural activity and affects an array of CNS functions such as mood, cognition, pain and appetite. The endocannabinoid system is comprised of the G-protein coupled receptors cannabinoid receptor type 1 (CB1R) and cannabinoid receptor type 2 (CB2R), their endogenous ligands 2-arachidonoylglycerol (2-AG) and N-arachidonoylethanolamine (AEA or anandamide), as well as the enzymes responsible for the synthesis and degradation of these molecules. In the brain, 2-AG is synthesized ‘on-demand’ in an activity-dependent manner in postsynaptic neurons and retrogradely binds to and activates pre-synaptic CB1R. Following CB1R binding, 2-AG is rapidly degraded into arachidonic acid and glycerol by the serine hydrolase monoacylglycerol lipase (MGLL). MGLL is thus a critical point of regulation of both endocannabinoid and eicosanoid signaling pathways in the brain and may provide novel therapeutic opportunities for treatment of mood disorders, including anxiety and post-traumatic stress disorders (PTSD). Here we assessed the effectiveness of the reversible MGLL inhibitor JNJ-69095897 in modulating behavioral and physiological responses in two preclinical models of panic/anxiety and conditioned fear.

Methods: All studies were performed in male Sprague-Dawley rats weighing approximately 300-400 g. Dose-dependent MGLL occupancy by JNJ-69095897 (0.1-10 mg/kg) was assessed via ex vivo binding autoradiography of a tritiated MGLL radioligand in rat brain hippocampal tissue sections collected 1 hour after JNJ-69095897 oral dosing. Group sizes of 3 rats per dose were used.

An acute hypercapnia model in which rats were transiently exposed to 5 minutes of 20% carbon dioxide (CO2)/normoxic gas was used as a model of panic/anxiety. For these studies, rats were pre-implanted with telemetric probes for measurement of mean arterial blood pressure, heart rate and core body temperature. Following the hypercarbic challenge, rats were assessed via a social interaction (SI) test to measure anxiety-like responses.

In separate studies, an auditory fear conditioning model was employed in which a tone (conditioned stimulus, CS) was paired with a brief footshock, and elicited freezing behaviors were measured. Rats were trained with 5 such conditioning trials. Consolidation, extinction and retrieval/recall of fear memories were assessed at various timepoints after conditioning trials via measurement of freezing behavior in response to CS alone.

For behavioral studies, JNJ-69095897 (0.3, 1 and 3 mg/kg) was formulated for p.o. dosing in 20% hydroxypropyl-b-cyclodextrin and administered at various timepoints. Group sizes of 10-13 animals per treatment group were used. Telemetric and freezing behavior data were analyzed using a 2-way repeated measures ANOVA with Fisher’s LSD post hoc test. SI data were analyzed using a 1-way ANOVA with Dunnet’s post hoc test. Data were graphed as the mean +/- SEM. The significance level was set at p < 0.05.

Results: Oral administration of JNJ-69095897 led to a dose-dependent increase in MGLL occupancy (ED50: 0.86 mg/kg; 0.58-1.16, 95% CI) as well as in measured blood and brain 2-AG concentrations.

Rats exposed to an acute 20% CO2/normoxic gas challenge exhibited reduced interaction time in the SI test as compared to control rats exposed to normal air. Acute pre-treatment with JNJ-69095897 significantly attenuated these deficits in a dose-dependent manner (p < 0.005), with both 1 mg/kg and 3 mg/kg dosing groups exhibiting increased (p < 0.005) SI times versus vehicle-treated control animals. Additionally, JNJ-69095897 pre-treatment dose-dependently attenuated CO2-induced decreases in heart rate (p = 0.002) and core body temperature (p < 0.0001), as well as increases in blood pressure (p = 0.008). Efficacy in this model was maintained following 5 days of daily 1 mg/kg JNJ-69095897 dosing.

In the auditory fear conditioning animal model, JNJ-69095897 was efficacious at all doses tested in reducing CS-elicited freezing behaviors when administered 1 hour prior to fear extinction trials (p < 0.01). Moreover, these effects were maintained when fear memory recalls were assessed 1 and 8 days after dosing (p < 0.0001). Co-administration of the CB1R inverse agonist AM251 largely blocked the effects of JNJ-69095897 in this model.

Conclusions: We were able to demonstrate efficacy of the reversible MGLL inhibitor JNJ-69095897 in attenuating behavioral and physiological measures indicative of an increased anxiety- or fear-like states in two preclinical models of panic/anxiety and conditioned fear. Effective doses also led to measurable increases in brain MGLL occupancy and/or 2-AG levels. In the hypercapnia-induced panic/anxiety model, this efficacy was maintained following several days of repeated administration, suggesting a lack of CB1R-mediated desensitization. The results obtained in this study are consistent with previous reports demonstrating the role of enhanced endocannabinoid signaling in mediating resiliency in preclinical stress models.

Keywords: endocannabinoid, CB1 receptor, Anxiety and PTSD, Animal Models, Preclinical pharmacology

Disclosure: Janssen Research and Development LLC: Employee (Self). Janssen Research and Development LLC, Stock / Equity - Publicly Traded Company (Self).

Sophia Allen*, Victoria Francois

Penn State College of Medicine, Hershey, Pennsylvania, United States

Background: Cigarettes are used by over 34 million U.S. adults and cause more than 480,000 deaths annually. The U.S. Food and Drug Administration announced a proposed product standard to ban menthol cigarettes. This ban may cause an increase in utilization of quitlines; however few studies have been conducted among Black adults who use quitline services to ask about the challenges they experience during a quit attempt. To develop and refine a semi-structured interview guide for those who want to quit, we evaluated questions with Pennsylvania (PA) Quitline enrollees.

Methods: In this qualitative and mixed methods study, individuals who enrolled in PA Quitline services were invited to complete a screener. Eligible participants were called to confirm their interest and scheduled to meet via Zoom for an interview. Consent was obtained and compensation was provided. The guide is being used in a larger study of 64 participants. It included a script for the interviewer with consent to record the conversation (audio only), and questions divided into six categories: 1) past quit attempts, 2) perspectives on use of the PA Quitline and its text messaging service, 3) use of technology, 4) interactions with health care provider, 5) experiences with other addictive substances, and 6) suggestions for improvement of the PA Quitline. The study was reviewed and approved by our Institutional Review Board.

Results: In May 2024, the guide was tested with 10 participants who were primarily Black adults (80%), identified as female (70%), completed some college (40%), and had a mean age of 50 years (SD: 15.8, range 20-65). While some participants needed assistance with connecting to the researcher via Zoom, once connected they were willing to share challenges with quitting. After assessing the format of the guide, we made changes to ask participants quantitative questions about smoking behaviors first and subsequently asked more in-depth (open-ended) questions.

Conclusions: Few studies have developed and tested an interview guide to collect the challenges and responses of adults who smoke and want to quit. This study has implications for the design of smoking cessation interventions and demonstrated the feasibility to reach Black adults.

Keywords: Smoking Cessation, Menthol, African Americans, Health Equity

Disclosure: Nothing to disclose.

Amir Krivoy, Dana Arnheim, Arad Dotan, Johnatan Nissan, Linda Levi, Noa Menkes-Caspi, Sharon Taub, Ayala Sophia Magidovich, Jinyoung Park, Jonathan Rabinowitz*, Michael Davidson, John M. Davis, Mark Weiser

Bar Ilan University, Raanana, Israel

Background: Several studies have shown that individuals with schizophrenia are at reduced risk for prostate cancer. This study assesses the risk of prostate cancer in males with schizophrenia untreated or treated with prolactin-elevating and prolactin sparing antipsychotics, and in males not diagnosed with schizophrenia treated with antipsychotics

Methods: Data were extracted from a large Israeli HMO (n = 5.4 million). Subjects were males with schizophrenia aged 40 or older, untreated (n = 4286) or treated (n = 6456) with antipsychotics, and individuals treated with antipsychotic medications who do not suffer from schizophrenia (n = 2902). Controls (1:3, n = 39701) were individuals with no psychiatric diagnosis matched for age and socioeconomic status; analyses controlled for obesity and cigarette smoking. Antipsychotic exposure was defined as at least one year of purchasing medications between 2005 to 2009, incidence of prostate cancer was recorded from 2010 to 2021.

Results: Compared to controls, risk for prostate cancer in males with schizophrenia treated with antipsychotics was significantly lower (aHR=0·532, CI: 0.409-0.6934, p < 0.0001), this risk was driven by prolactin-elevating antipsychotics (aHR=0.521, CI: 0.364-0.747, p < 0.0001). Risk for prostate cancer was non-statistically significantly decreased in patients treated with prolactin-sparing antipsychotics (aHR=0.828, CI: 0.476-1.44, p = 0.504), patients with schizophrenia not treated with antipsychotics (aHR=0.769, CI: 0.574-1.032, p = 0.080), and patients treated with antipsychotics not affected by schizophrenia (aHR=0.778, CI: 0.569-1.063, p = 0.115).

Conclusions: Patients with schizophrenia treated with antipsychotics, specifically prolactin-elevating antipsychotics have a 50% reduced risk for prostate cancer, probably caused by increased prolactin which in turn decreases testosterone blood levels.

Keywords: Schizophrenia (SCZ), cancer, Antipsychotics, prolactin

Disclosure: Minerva: Consultant (Self). Neumarker AI: Consultant (Self) Genetika: Consultant (Self) Gamida: Consultant (Self), Clearmind Medicine: Consultant (Self).

Gonzalo Martinez-Ales*, Teresa Lopez-Cuadrado

Icahn School of Medicine At Mount Sinai, New York, New York, United States

Background: Bipolar disorder (BD) hospitalization rates in children and adolescents vary greatly across place and over time. There are no population-based studies on youth BD hospitalizations in Spain.

Methods: We identified all patients aged 10-19 hospitalized due to BD in Spain between 2000-2021, examined their demographic and clinical characteristics, and assessed temporal trends in hospitalizations – overall and stratified by age and presence of additional psychiatric comorbidity. We used Joinpoint regressions to identify inflection points and quantify whole-period and annual percentage changes (APCs) in trends.

Results: Of 4770 BD hospitalizations in 10-19 year-olds between 2000-2021 (average annual rate: 4.8 per 100,000), over half indicated an additional psychiatric comorbidity, most frequently substance abuse (62.2%), mostly due to cannabis (72.4%). During the study period, admissions increased two-fold with an inflection point: Rates increased annually only between 2000-2008, for APCs 34.0%(95%CI: 20.0%,71.1%) among 10-14 year-olds, 10.3%(6.4%,14.3%) among 15-19 year-olds, and 15.5%(11.5%,22.7%) among patients with additional psychiatric comorbidity. Between 2009-2021, rates decreased moderately among 10-14 year-olds – APC: -8.3%(-14.1%,-4.4%), and slightly among 15-19 year-olds without additional psychiatric comorbidity – APC: -2.6(-5.7,-1.0), remaining largely stable among 15-19-year-olds overall.

Conclusions: Recent trends in hospitalization due to BD in 10-19 year-olds in Spain indicate salient increases in the early 2000s - especially among (i) patients aged 10-14 (decreasing moderately after 2009 among 10-14 year-olds and plateauing among 15-19-year-olds) and (ii) patients with additional psychiatric comorbidity (i.e., cannabis use disorder). These findings suggest links with recent changes in clinical practices for children and recent trends in substance use among Spanish youth.

Keywords: epidemiology, Bipolar Disorder, Adolescent

Disclosure: Nothing to disclose.

Leah Rubin, Katelyn Jenkins, Ana Soule, Courtney Harrington, Mark Yoon, Yong Du, Martin Pomper, Vidyulata Kamath, Jennifer Coughlin*

University of Texas Southwestern Medical Center, Dallas, Texas, United States

Background: The colony stimulating factor 1 receptor (CSF1R) is a tyrosine kinase expressed on the cell surface of microglia and macrophages. While microglial proliferation in disease is marked by high density of the CSF1R, postmortem human brain tissue suggest that the CSF1R is downregulated in expression by microglia in select conditions such as Alzheimer’s disease, while other immune-related proteins are increased in their expression. Building on the hypothesis of downregulation of this homeostatic microglial protein in disorders marked by cognitive impairment and microglial response, we hypothesized that lower [11C]CPPC binding to the CSF1R would associate with worse cognitive performance in individuals with treatment-resistant Major Depressive Disorder.

Methods: In this cross-sectional study, 23 TR-MDD (78% female, 74% White, mean age=32.4) and 18 healthy controls (HC; 61% female, 56% White, mean age=35.2) completed [11C]CPPC PET. [11C]CPPC total distribution volume (VT), representative of CSF1R availability, was calculated using Logan graphical analysis with metabolite-corrected arterial input function in two regions of interest, the anterior cingulate cortex (ACC) and prefrontal cortex (PFC). A mixed-factor analysis of covariance was used to examine group differences in [11C]CPPC VT, with brain region (ACC, PFC) as the within-subjects factor, group (MDD-TR, control) as the between-subjects factor, and body mass index (BMI) as a covariate. Within TR-MDD, the relationship between VT, symptoms (depression, anxiety), and cognition was assessed using partial correlations. BMI was a covariate.

Results: TR-MDD and HC were similar in age, sex, and race. However, TR-MDD had a higher BMI than HC (26.4 vs. 23,5, P = 0.02). There were no differences in age-adjusted brain volumes in these regions between the two groups, F (1, 38) = -0.99, P = 0.32. Across both ACC and PFC, lower estimated [11C]CPPC VT was observed in TR-MDD (mean ± SE = 14.90 ± 0.0.54) compared to healthy controls (mean ± SE = 16.57 ± 0.0.44) after adjusting for BMI, F(1, 38) = 9.71, P = 0.003. There were no association between [11C]CPPC VT in either region and HAM-D or HAM-A. With respect to cognition, lower [11C]CPPC VT in each region (ACC, PFC) was associated with poorer attention/working memory (ACC: r = 0.48, P = 0.02; PFC: r = 0.47, P = 0.03). [11C]CPPC VT in these regions was not associated with psychomotor speed, declarative memory, language, or motor function.

Conclusions: [11C]CPPC PET in TR-MDD and controls suggests that the lower availability of the CSFR in ACC or PFC in TR-MDD may be linked with worse attention/working memory in this condition.

Keywords: microglia, [11C]CPPC, treatment-resistant Major Depressive Disorder

Disclosure: D and D Pharmatech: Patent (Spouse/Partner).

Rebecca Howell*, Mikhail Pletnikov, Joseph Cheer, Samir Haj-Dahmane

University at Buffalo, Buffalo, New York, United States

Background: The spontaneous activity of ventral tegmental area (VTA) dopamine neurons is tightly regulated by their excitatory and inhibitory inputs. Consequently, alterations of excitation/inhibition (E/I) balance will have profound impact on the overall activity of dopamine neurons, central dopamine transmission and dopamine-gated behaviors, including motivation, locomotion and learning. Over the last decade, there was a growing recognition that astrocytes, a key component of the tripartite synapse, play an important role in neuronal computation by regulating the E/I ratio, yet, so far, very little is known about how these brain cells can modulate synaptic physiology in the VTA.

Astrocytes are the most abundant glial cell in the central nervous system, performing many functions to maintain brain homeostasis including controlling the blood brain barrier, regulating ion and neurotransmitter transport, forming glial scars. In addition, by expressing a constellation of neurotransmitter receptors, including a variety of G protein coupled receptors (GPCRs), astrocytes act as sensors of neuronal activity, and as such, fine tune synaptic transmission and plasticity. Canonically, astrocytic GPCRs signal through the PLC/IP3 pathway and promote increased intracellular calcium upon activation. The increased intracellular calcium promotes the release of an array of gliotransmitters, including glutamate, D-serine and ATP. The roles of gliotransmission in controlling synaptic function will vary in each brain region depending on the gliotransmitters and the receptors involved. Generally, ATP released from astrocytes is metabolized by extracellular ectonucleotidase into adenosine, an endogenous agonist of adenosine receptors, the activation of which will gate the strength of central synapses. Here, our goal is to unravel the mechanisms by which astrocytes fine tune synaptic function of VTA dopamine neurons.

Methods: Using VTA-containing brain slices from male and female C57BL/6 mice (6 to 12 weeks old), we performed patch clamp electrophysiology experiments in dopamine neurons to discern whether astrocytes play a role in modulating the intrinsic and synaptic properties. Slices were either designated as control or treated with the gliotoxin, fluoroacetate (1mM, incubated for 1 hour) in order to inactivate astrocytes.

Results: Acute inactivation of astrocytes with fluoroacetate had no effect on the spontaneous firing activity or intrinsic properties of dopamine neurons. We evoked glutamatergic and GABAergic currents to assess the E/I ratio and found a drastic decrease in E/I ratio following fluoroacetate treatment (p < 0.001). No changes were detected in glutamatergic currents, however there was a large increase in the amplitude of GABA currents following fluoroacetate treatment. Analysis of the coefficient of variation (CV) of GABA-mediated synaptic currents revealed that inactivation of astrocytes profoundly reduced the CV (p < 0.005) indicating that the decrease in E/I ratio was likely due to an increase in the probability of GABA release.

To determine the mechanism by which astrocytic inactivation potentiated the release probability of GABA, we tested whether an impairment in ATP/adenosine mediated gliotransmission is involved. We first confirmed the presence of A1 adenosine receptors (A1R) on GABAergic synapses at VTA dopamine neurons. A1R are often located presynaptically and signal through the Gi/o pathway and inhibit neurotransmitter release. Bath application of the A1R agonist, CPA, significantly reduced the frequency and amplitudes of spontaneous and evoked GABA currents, respectively. Since astrocytes are known to be spontaneously active in slices, with a high baseline frequency of intracellular calcium events, we then explored whether astrocyte mediated adenosine signaling at GABA synapses is tonic in the VTA. Perfusion of slices with A1R antagonist, CPT, produced a small increase in frequency of spontaneous GABAergic currents (p < 0.005) that was occluded in slices pretreated with the gliotoxin. Based on this, we replicated the effect of the gliotoxin on E/I ratio by perfusing slices with CPT, suggesting that GABA terminals are under control of astrocytic mediated adenosine transmission in the VTA.

Conclusions: The results of the present study reveal that astrocytic gliotransmission plays a large role in regulating the balance of excitation and inhibition of dopamine neurons in the VTA. The effect is largely mediated by modulation of GABAergic synapses on dopamine neurons. Astrocytes are the source of a tonic adenosine signal on GABAergic synapses on dopamine neurons, inhibiting GABA release through presynaptic A1R. This suggests that astrocytes perhaps play an underappreciated role in regulating synaptic plasticity of VTA dopamine neurons and thus dopamine-dependent behaviors.

Keywords: Dopaminergic neurons, astrocytes, Ventral Tegmental Area (VTA), synaptic physiology, slice electrophysiology

Disclosure: Nothing to disclose.

Alberto Corona*, Masago Ishikawa, Victor Mathis, Lauren Wills, Junshi Wang, Paul Kenny

Icahn School of Medicine At Mount Sinai, New York, New York, United States

Background: Stress increases the propensity of lateral habenular (LHb) neurons to fire in phasic bursts of activity. LHb burst-firing modifies monoaminergic neurotransmission in corticolimbic brain sites, which contributes to stress-related negative behavioral states. The mechanisms underlying this stress-induced maladaptive cellular plasticity in the LHb are unclear. Stress enhances norepinephrinergic (NE) transmission throughout the brain, which orchestrates physiological and behavioral adaptations to stress. Microglia express β2 adrenergic receptors (β2ARs) and are directly regulated by NE. Whether stress-induced NE transmission in microgliaβ2AR contributes to LHb neuron burst-firing has not been investigated. Here, we will test the hypothesis that stress engages NE signaling microgliaβ2AR in the LHb, which promotes burst-firing of LHb neurons to precipitate stress-related behavioral deficits.

Methods: Single-cell RNA sequencing (scRNA-seq) was performed using the 10x Genomics Chromium and Illumina NovaSeq 6000 platforms. NE transmission in the LHb was monitored using a genetically encoded sensor (GRABNE) and fiber photometry. Microglia were depleted for electrophysiology and fiber photometry recordings by permitting mice to consume chow containing the colony-stimulating factor-1 receptor (CSF-1R) antagonist PLX3397. Intrinsic activity patterns of LHb neurons were characterized using whole-cell current-clamp recordings. Calcium activity in microglia was monitored on brain slices using a fluorescent microscope. All experimental protocols in animal studies were approved by the Institutional Animal Care and Use Committee and were conducted per the National Institutes of Health Guide for the Care and Use of Laboratory Animals. Male and female mice were included in these studies and no significant sex differences have been identified.

Results: Using whole-cell recordings in LHb-containing brain slices, we found that depletion of microglia from the brains of mice precipitated a stress-like increase in the propensity of LHb neurons to fire in phasic bursts of activity (n = 125 cells/7mice control; n = 98 cells/5mice treated with PLX3397; t-test **p < 0.01). To confirm microglia regulation of LHb burst-firing in vivo, I used fiber photometry in activity of GCaMP6s-expressing LHb neurons in mice with or without microglia depletion. Aversive foot shocks evoked calcium (Ca2 + ) time-locked responses in LHb neurons from both groups, but responses persisted significantly longer in the microglia-depleted mice (10 vs. 3 sec, respectively) (n = 8 control and n = 7 PLX3397; Two-way ANOVA with diet and time as factors; day1 **p = 0.0078; day2 **p = 0.0011). Using scRNA-seq, we found that β2AR gene transcripts in LHb were expressed almost exclusively by microglia (n = 6655 cells for stress group and n = 14271 for control group). This observation was confirmed using RNAscope in situ hybridization. Using the GRABNE sensor and in vivo fiber photometry, I found that stressors known to reliably promote burst-firing of LHb neurons (foot-shocks and physical restraint) triggered time-locked changes in NE transmission in the LHb of mice over 3 consecutive days (One-way ANOVA; F = 6.777 ****p < 0.0001 followed by Bonferroni’s multiple comparison test day1 = **p = 0.0012; day2 = **p = 0.0056; day3 = *0.0148). Using LHb-containing brain slices and fluorescence Ca2+ imaging, we found that ATP evoked Ca2+ transients in GCaMP-expressing LHb microglia (n = 9 slices; 3 mice) that were blocked after P2ry12 antagonist PSB0739 application (n = 5 slices/3 mice; paired t-test *p = 0.0106). Crucially, NE almost completely abolished this ATP-evoked response in microglia (n = 9 slices/3 mice; paired t-test ***p = 0.0001). Currently, we are investigating the effects of conditionally deleting β2ARs from LHb microglia on firing patterns of LHb neurons and the expression of stress-related behaviors.

Conclusions: Our data suggest that stress enhances NE transmission in the LHb, which acts on microglia through β2ARs. In turn, these stress-sensing microglia regulate adaptations in the activity patterns of LHb neurons that drive stress-induced behavioral abnormalities. We hypothesize that microglia increase the propensity of LHb neurons to engage in burst firing by shifting from a homeostatic to a pro-inflammatory profile. Since microglia in “surveillance” mode regulate synaptic strength and neuronal activity, a shift from surveillance to pro-inflammatory mode likely disrupts LHb microglia-neuron interactions. The fact that β2AR expression in the LHb is restricted to resident microglia is important because β2ARs play crucial roles in coordinating adaptations to stress, which suggests that microglia are critical intermediaries of these actions in the LHb.

Keywords: microglia, norepinephrine, Lateral Habenula

Disclosure: Nothing to disclose.

Hua Chai*, Vijaya Pandey, Matthias Ollivier, Sara L Gutierrez, Oishi Bhattacharya, James A. Wohlschlegl, Baljit Khakh

David Geffen School of Medicine at UCLA, Los Angeles, California, United States

Background: The striatum is a part of the basal ganglia and serves essential functions in motor learning and behavioral habits. From human studies, we expect that age-associated molecular changes in the brain are greater in glia compared to neurons, including in the striatum. Astrocytes are the most abundant type of glia in the mammalian brain. These morphologically complex cells are involved in extracellular milieu homeostasis, circuit function, and the blood-brain barrier. We know from published work that astrocytes demonstrate shifts in transcriptomic profiles during aging. However, gene expression changes do not necessarily translate to changes at the protein level or demonstrate subsequent impact on cellular function. Therefore, we sought to determine how striatal astrocytes change in aging at the proteomic and functional levels and relate these changes to gene expression.

Methods: We employ several approaches in our comparison of striatal astrocytes from young adult (~2 months; 2M) and aged mice (~18 months; 18M). All experiments used both male and female mice. To measure astrocyte intracellular calcium signals, we used local injection of adeno-associated viruses (AAVs) to express genetically encoded calcium indicator GCaMP8m to study calcium signals in acute slices from 18M and 2M mice. We also performed whole-cell patch clamp electrophysiology to examine membrane properties of striatal astrocytes at both ages. To generate proteomes, we used local injection of AAVs to express BioID2 or GFP control specifically in cytosolic and near-membrane compartments of astrocytes in dorsal striatum of 18M mice. In vivo biotinylated proteins were isolated using streptavidin-based pulldown. Proteins were then digested into peptides and analyzed using tandem mass spectrometry. To generate transcriptomes, we used Ribotag mice crossed with Aldh1l1-Cre/ERT2 mice to drive expression of Rpl22HA specifically in astrocytes then immunoprecipitated these HA-tagged ribosomes for extraction and sequencing of astrocyte mRNA.

Results: Astrocytes display intracellular calcium signals that affect neural circuit function and behavioral phenotypes. As we observed changes in striatal-mediated behaviors in aged mice like those reported in the literature, we examined astrocyte intracellular calcium signals in acute slices. Interestingly, 18M astrocytes exhibited significantly diminished intracellular calcium responses during bath application of phenylephrine to activate alpha1 adrenoceptors (91% decrease in response, N = 13-19, Mann Whitney p = 0.0011), a significant change in one of the most consistent calcium responses observed in astrocytes.

Another core feature of astrocytes is regulation of extracellular K+ homeostasis through inward-rectifying K+ channels. We found that 18M astrocytes displayed smaller inward-rectifying potassium currents with a mean difference of 5.1 ± 1.8 nS (N = 13-14, t test p = 0.0085), suggesting age-related changes in another well-established astrocyte function.

To understand the molecular mechanisms responsible for the observed functional changes, we sought to identify the cytosolic and near-membrane subproteomes in astrocytes using the BioID2 approach. In the cytosolic compartment of 18M astrocytes, we identified 687 proteins (N = 3, FDR < 0.05) with 71% not found in previously published data from our laboratory in 2M mice. In the near-membrane compartment, we observed a similar pattern. As tandem mass spectrometry identifies the most abundant proteins in each sample, the results suggest major shifts in the relative expression of proteins in both astrocyte compartments.

As expected, based on prior work demonstrating non-linearity between mRNA and protein expression levels, there were many 18M striatal astrocyte molecules whose relative expression at the protein level did not correlate with their relative gene expression in our transcriptomic data when analyzed using rank-rank hypergeometric overlap analyses. Thus, our 18M subproteomes provide important insight into the molecular changes in astrocytes that cannot be determined by gene expression studies alone. Using the transcriptomic and proteomic data generated in this study, we found differences in potassium channels and molecules related to calcium signaling that potentially explain the age-related changes observed in our functional studies.

Conclusions: These initial results demonstrate that astrocytes exhibit major changes in molecular profiles with aging. The proteomic and transcriptomic data are powerful resources to help understand molecular differences that underlie age-related functional changes seen in astrocytes, such as the changes we observed in K+ homeostasis and alpha adrenoreceptor-mediated calcium signaling. Further studies are ongoing to understand the connection between protein expression and functional changes with age in other astrocyte compartments. We anticipate that the results from this work will generate new hypotheses to understand the molecular changes that underlie aging, which is of relevance to understand age-related vulnerabilities to neuropsychiatric illness.

Keywords: astrocytes, brain aging, Proteomics, calcium imaging, Transcriptomics

Disclosure: Nothing to disclose.

Kateryna Murlanova*, Ksenia Novototskaya-Vlasova, Shovgi Huseynov, Olga Pletnikova, Henry Withers, Yuto Hasegawa, Samir Haj-Dahmane, Roh-Yu Shen, Urs Meyer, Atsushi Kamiya, Mikhail Pletnikov

State University of New York at Buffalo, Buffalo, NY, USA, Buffalo, New York, United States

Background: Cannabis exposure during adolescence has been linked to several long-term negative consequences. However, the molecular, cellular, and circuit underpinnings that could explain why only some individuals are vulnerable to the adverse effects of cannabis remain poorly studied. While genetic risk factors, sex, doses, variable potency, and different times of exposures are important factors contributing to highly heterogeneous outcomes, in utero exposure to environmental adversities (e.g., maternal immune activation [MIA]) could also play a critical role in exacerbating the long-term negative effects of adolescent cannabis use. As both MIA and cannabis can activate neuroinflammatory signaling in astrocytes, we hypothesized that convergence of MIA and adolescent cannabis on astrocytic neuroinflammatory signaling will produce lasting neurochemical and behavioral abnormalities in adult mice.

Methods: MIA was induced by a viral mimetic, polyriboinosinic-polyribocytidilic acid [Poly(I:C), 5 mg/kg; IP] to dams (CD1 mice, GD 12.5). After weaning, male and female offspring received THC (8 mg/kg; SC) or vehicle during adolescence (PND 30–51). Three weeks after THC, mice were tested in a series of behavioral tests. A separate cohort was used for in vivo brain microdialysis to assess extracellular levels of dopamine (DA), glutamate (GLU), and GABA in the dorsal striatum (DS). Ex vivo electrophysiological recordings were performed to assess gliotransmission and glutamate synaptic transmission in DS. A separate cohort of mice were sacrificed three weeks after THC treatment for RNAseq of isolated astrocytes. The behavioral tests included 12-18 mice per group, the neurochemical and RNAseq experiments included 5 mice per group. Data was analyzed using two-way ANOVA followed by Bonferroni post-hoc test. FDR correction was used for analysis of gene expression changes from RNAseq data.

Results: MIA x THC interaction significantly and synergistically increased exploratory behaviors in the hole board test, open field, elevated plus maze, and synergistically elevated amphetamine-induced (2 mg/kg, IP) hyperactivity in mice. MIA x THC interaction synergistically increased extracellular levels of DA and GLU in DS. MIA x THC interaction did not alter the frequency of glutamate-mediated miniature excitatory synaptic currents but increased the frequency of slow inward currents, suggesting increased astrocytic, but not neuronal, GLU release. Transcriptomic profiling of striatal astrocytes revealed MIA-induced priming of pro-inflammatory signaling in astrocytes that was further up-regulated in mice exposed to MIA and THC.

Conclusions: MIA and adolescent THC synergistically activate inflammatory signaling in astrocytes to lead to excessive release of glutamate, affecting striatal DA neurotransmission and DA-dependent behaviors in adult mice.

Keywords: Adolescent cannabis, Maternal immune activation, Dopamine, Gliotransmission, Astrocytes

Disclosure: Nothing to disclose.

Ashley Holloway*, Khukheper Awakoaiye, Aarushi Mittal, Jenny Klein, Hao Li, Reesha Patel

Feinberg School of Medicine, Northwestern University, Chicago, Chicago, Illinois, United States

Background: Appropriate approach and avoidance behaviors are critical for survival—they underlie the abilities of organisms to seek out food and mates, and avoid life-threatening situations. Approach and avoidance behaviors are dysregulated in many psychiatric disorders, ranging from substance use disorders to anxiety and major depressive disorders. Thus, understanding the neurobiological mechanisms underlying approach and avoidance is crucial for the development of therapeutics to treat psychiatric disorders. The ventral tegmental area (VTA) is a cellularly heterogeneous brain region implicated in appetitive and aversive conditioning, and thus approach and avoidance behaviors. Numerous studies have pointed to the roles of VTA neurons in governing reward-guided behaviors and avoidance. However, neurons are not the only cell type within the VTA capable of driving behavior. Throughout the brain, including within the VTA, astrocytes play key roles in supporting neuronal metabolism, regulating levels of extracellular neurotransmitters via reuptake, and tuning synaptic transmission via release of gliotransmitters. Recent studies using chemogenetic and optogenetic approaches have revealed that astrocytes within the VTA can regulate local neuronal activity and modulate approach and avoidance. Evidence for both neuronal and astrocytic involvement in the regulation of approach and avoidance begs the question: do the two different cell types encode distinct information under appetitive and aversive conditions? To address this question, we used fiber photometry to simultaneously characterize the activity of VTA neurons and astrocytes in the same animals during Pavlovian conditioning to sucrose reward and footshock.

Methods: We used dual-color fiber photometry in combination with viruses encoding an astrocyte-specific green fluorescent calcium indicator (AAV5-gfaABC1D-cyto-GCaMP6f) and a neuron-specific red fluorescent calcium indicator (AAV9-hSyn-jRGECO1a) to track calcium activity of ventral tegmental area (VTA) astrocytes and neurons simultaneously in awake, behaving mice of both sexes (n = 5 mice per sex). To assess the activity of VTA astrocytes and neurons during appetitive and aversive tasks, we recorded astrocytic and neuronal activity using fiber photometry as mice were trained on a Pavlovian sucrose reward task and a Pavlovian fear conditioning task.

Results: In the Pavlovian reward task, we found that both astrocyte and neuronal activity reflected port entry outcomes in males, with larger peaks in both cell types observed during rewarded port entries compared to unrewarded port entries (Astrocytes: Two-way ANOVA, main effect of event type, p = 0.0033; Neurons: Two-way ANOVA, main effect of event type, p = 0.0158). In females, only astrocyte activity reflected port entry outcomes, with astrocyte peaks being significantly higher for rewarded port entries than unrewarded port entries (Mixed effects model, main effect of event type, p = 0.0094). We also found that astrocytic activity, but not neuronal activity, during rewarded port entries positively correlated with the number of rewarded port entries in female mice (linear regression between astrocyte area-under-the-curve (AUC) and number of rewarded port entries, r2 = 0.8529, p = 0.0251), but not male mice. Furthermore, we found that astrocyte activity during rewarded port entries was negatively correlated with the latency of females to enter the reward port after tone presentation (linear regression between astrocyte AUC and latency from tone to reward, r2 = 0.9376, p = 0.0068). These results suggest that VTA astrocyte activity reflects motivational vigor for rewards in female mice, and that astrocyte activity is outcome-dependent in both sexes. In the Pavlovian fear conditioning task, we found that VTA neuronal and astrocytic activity was positively correlated with the number of movement bouts initiated in response to the shock in female (linear regression between number of movement bouts and neuron AUC [r2 = 0.9219, p = 0.0095] and astrocyte AUC [r2 = 0.8023, p = 0.0398]), but not male, mice. These results suggest that VTA neuron and astrocyte activity reflect motivational vigor to avoid shock in females, but not males.

Conclusions: These findings suggest that VTA astrocyte activity reflects motivation to acquire rewards during an appetitive task in females, and that both VTA neuronal and astrocyte activity reflect avoidance motivation during an aversive task in females. Ongoing studies are investigating the causal relationship between astrocyte activity and motivation during appetitive and aversive conditions. Overall, our studies indicate that while both VTA astrocytes and neurons respond to appetitive and aversive stimuli, astrocyte activity reflects motivation independent of valence in female mice. Thus, astrocytes are a behaviorally-relevant cell population that may be a potential candidate for therapeutic approaches to ameliorate deficits in approach and avoidance behaviors.

Keywords: Astrocytes, Aversive Motivation, Reward Motivation, Ventral tegmental area, In vivo fiber photometry

Disclosure: Nothing to disclose.

Maura Stoehr, Elise Martin, Jason Xue, Joy Babalola, Niki Li, Matt Kern, Madeline Winters, Djerdjaj Anthony, Staci Bilbo, Caroline Smith*

Boston College, Chestnut Hill, Massachusetts, United States

Background: Gestational exposures to both air pollution and maternal stress are associated with an increased likelihood of autism spectrum disorder. Importantly, these exposures are not equitably distributed. Rather they converge on already vulnerable populations. However, the neurobiological mechanisms by which these exposures interact to alter social functioning remains unclear. In our previously published work, we found that combined prenatal exposure to diesel exhaust particles (DEP) and maternal stress (MS) induces male-specific social behavior deficits in a mouse model. We further found that these behavior deficits were accompanied by changes in the composition of the gut microbiome and changes in the gene expression and function of microglia, the resident immune cells of the brain. Social behavior deficits were prevented by manipulation of the gut microbiome at birth. Oxytocin and vasopressin are neuropeptides that are essential to the regulation of social behavior and sensitive to signals from the gut microbiome. Here we tested the hypothesis that oxytocin and vasopressin systems in the gut and brain may be key substrates on which microglia and microbiota act to alter social behavior.

Methods: Mouse dams were exposed to either DEP/MS or control (VEH/CON) conditions during gestation. Offspring brains and intestines were collected during early adulthood (N = 8/10 per group) and a combination of immunofluorescence and gene expression analyses were used to assess oxytocin and vasopressin cell number/fiber density, microglial number and morphology, and oxytocin receptor (OTR) and vasopressin V1a receptor expression (V1aR). A tdTom-OTR report mouse line was used to assess microglial engulfment of synapses on OTR-expressing neurons (Oxtr1xPA-tdTom, Jax: 037580). Stereotaxic surgery was performed to microinject neutrophil inhibitory factor (NIF) locally into the lateral septum in order to block microglial synaptic pruning and social behavior assays were performed during adolescence (N = 8/group). Both males and females were included in all experiments. 2-way ANOVA’s with sex and treatment as factors or t-tests (treatment) were used to analyze all data.

Results: DEP/MS tended to increase oxytocin but not vasopressin cell number in the paraventricular nucleus of the hypothalamus (OXT: treatment: F(1,28) = 3.5, p = 0.07; AVP: F(1,19) = 0.4, p = 0.49). In the lateral septum, a socially-relevant target region of hypothalamic oxytocin and vasopressin neurons, microglial density is reduced following DEP/MS exposure, but this effect is prevented by cross-fostering on the day of birth (t = 2.31(1,11), p = 0.04). We are currently assessing whether DEP/MS induces changes in microglial pruning of OTRs within the lateral septum and whether this synaptic pruning is responsible for changes in social behavior. In the small intestine, DEP/MS alters V1aR and tended to alter OTR expression (V1aR: treatment: F(1,24) = 8.95, p = 0.063; OTR (trend): treatment F(1,22) = 3.4, p = 0.075, suggesting a novel role for these neuropeptides in the mechanisms by which the microbiome influences behavior.

Conclusions: These findings suggest that both oxytocin and vasopressin neuropeptide systems in the brain and/or gut are altered in offspring following prenatal exposure to DEP/MS. Furthermore, the lateral septum may be an important region within the neural networks that support social behavior where microglial synaptic pruning alters oxytocin receptor expression in response to cues from the gut microbiome. Finally, while a connection between the oxytocin system and the gut microbiome is well established, our evidence for changes in the vasopressin system suggests a novel role for vasopressin signaling in the gut in these processes.

Keywords: microglia engulfment, vasopressin and oxytocin, Gut Microbiome, Social Behavior, Autism

Disclosure: Nothing to disclose.

Jonathan VanRyzin*, Kathryn Reissner

University of North Carolina at Chapel Hill, Chapel Hill, North Carolina, United States

Background: Adolescence is a dynamic developmental period marked by brain circuit refinement. Social stress during adolescence can influence the development of maladaptive behaviors and increase the risk of later-life psychiatric disorders. However, the mechanisms by which this happens remain unknown. Because astrocytes and microglia both promote circuit refinement during development and regulate neural function at maturity, and are subject to dysregulation by stress, we hypothesized that adolescent social isolation affects later-life behavior by impairing glial function within the nucleus accumbens (NAc), a brain region responsible for integrating social, reward, and stress information.

Methods: We isolated male and female Long Evans rats (or group-housed controls, n = 6 per group) from weaning until adulthood (P21-P56). From P57-P58, rats underwent social behavior testing with a novel age- and sex-matched conspecific in social interaction test and 3-chamber preference test. Following behavior, brains were collected for immunohistological analysis of astrocytes and microglia using GFAP and Iba1, respectively. Data were analyzed using Welch’s t test or repeated measures ANOVA, when appropriate.

Results: Socially-isolated rats spent more time interacting with a novel conspecific compared to group-housed controls (p = 0.007) in the social interaction test, while there were no differences in social preference index between groups (p = 0.732). We found GFAP+ astrocytes within the NAc of socially-isolated rats had decreased structural complexity (p < 0.001) as assessed by Sholl analysis of GFAP branching patterns, and a corresponding increase in the percentage of phagocytic Iba1 + /CD68+ microglia (p = 0.011). Moreover, phagocytic microglia had significantly more inclusions that co-labeled with the astrocyte marker GLT-1 (p = 0.004).

Conclusions: Our results demonstrate that adolescent social isolation stress induces aberrant phagocytosis of astrocytes and reduced astrocyte complexity within the NAc, and increased social interaction. We hypothesize that microglia pruning of astrocytes provides a putative mechanism by which isolation stress drives astrocyte atrophy and behavioral dysregulation. Our findings highlight the sensitivity of the neuroimmune system to developmental stressors and propose a novel model of microglia/astrocyte pruning by which neuropsychiatric disorders may arise.

Keywords: Astrocyte, microglia, Social Behavior, Adolescent, social isolation

Disclosure: Nothing to disclose.

Mounira Banasr*, Jessie Muir, Rosemary Bagot, Etienne Sibille, Yashika Bansal

Centre for Addiction and Mental Health, University of Toronto, Toronto, Canada

Background: Astroglia loss and decreased expression of specific markers expressed by GFAP (glial fibrillary acidic protein)-astroglia have been found in key brain regions such as the prefrontal cortex (PFC) in MDD patients and rodent chronic stress models. In this study, we examined the consequences of chronic stress on GFAP-astroglia cell activity and neuronal activity and linked the observed changes to behavioral deficits.

Methods: Mice co-infused in the PFC with an AAV5-GFAP-GCAMP and AAV2-Syn1-RCAMP (viruses driving expression of green or red calcium sensors in astrocytes or neurons, respectively) were implanted with a fiberoptic cannula and behaviorally assessed weekly. Half of the mice were subjected to chronic unpredictable stress for 5 weeks (n = 6 per group). Every week PFC cell activity changes also were recorded in astrocytes and neurons before, during and after an acute immobilization stress challenge. Relationship between cell activity changes in neurons and astrocytes as well as behavioral outcomes in tests measuring anxiety and anhedonia were assessed using correlation analysis.

Results: We found that while PFC neuronal activity in response to an acute stress challenge is acerbated in animals subjected to chronic unpredictable stress (p < 0.01), PFC astroglia cell activity was progressively reduced (n = 6/group, p < 0.05). Correlations between the changes in neuronal and astroglial activity were significant in control mice and early weeks of chronic stress, showing tight coupling between both cell types. However, astroglial and neuronal activity were not correlated on week 4 and 5 (p > 0.05). Using astroglial and neuronal activity R coefficient as an index of neuron/astroglia cross-talk integrity, we found that impaired cross-talk may contribute to anhedonia like behavior.

Conclusions: Our results demonstrate that chronic stress impaired cortical GFAP-astroglia cell activity and neuron/astroglia cross-talk. These effects coincide with the development of anhedonia-like behavior. Altogether, our work suggests a critical role of astroglia in the expression and the treatment of key symptoms of MDD.

Keywords: Astrocyte, Anhedonia, Chronic stress, Depression, Behavior

Disclosure: Nothing to disclose.

Haley Vecchiarelli*, Hayley Thorpe, Sophia Loewen, Colin Murray, Colby Sandberg, Luke Rainer-Pope, Hakan Kayir, Crisel Vital, Jibran Khokhar, Marie-Eve Tremblay

University of Victoria, Victoria, Canada

Background: Microglia are the brain’s resident immune cells—they actively maintain brain health via their many physiological roles (e.g., support for neurogenesis, gliogenesis, myelination and synapse formation; phagocytic removal of weak synapses and toxic debris, as well as the release of pro-/anti-inflammatory mediators and trophic factors) across the lifespan. They possess cannabinoid receptors and respond to cannabinoids, such as the major phytocannabinoids in cannabis plants, including delta-9-tetrahydrocannabinol (THC) and cannabidiol (CBD). Many studies have shown that THC and CBD modulate inflammation mediated by microglia, as activity of cannabinoid receptors generally down-regulates pro-inflammatory cytokines and stimulates anti-inflammatory cytokines. This is highly dependent on the duration and route of administration (e.g., injection or in vitro) as well as the compounds administered (i.e., isolated THC and/or CBD). However, the effects of a highly translationally relevant type of exposure, inhalation, on the physiological functions of microglia that could modulate the brain inflammatory balance, but also promote neuroprotection and neural circuit integrity, remain unexplored, particularly across sexes.

Methods: Whole cannabis plant was administered to young adult, male and female, C57BL/6J mice for 15 min (one 15 second puff every 5 min; 3 puffs total; 0.15 g flower/puff). Four groups were utilized, mice that received control air vapor, and mice that were exposed to either: high CBD/low THC [CBD], high THC/low CBD [THC], or balanced THC/CBD [Balanced] cannabis chemovars. Brains were isolated 30 min post-cannabis administration onset, when THC levels peak in the brain. We stained the tissue with antibodies against IBA1 (microglia and macrophages) and TMEM119 (more specific for microglia). We looked at IBA1+ cell density, nearest neighbor distance and spacing index (changes in number and distribution), in brain regions important for cognition, memory, and emotional regulation, specifically focusing on the prefrontal cortex, hippocampus and amygdala. We also investigated IBA1 and TMEM119 colocalization, as well as IBA1+ cell morphology.

Results: There is an effect of brain region and sex on IBA1+ cell density and distribution. Within the prefrontal cortex, in males, there was a reduction in the nearest neighbour distance of IBA1+ cells in the prelimbic region with exposure to all cannabis chemovars—indicating a difference in microglial distribution in this brain region. However, in males, there were no alterations of microglial cell density and distribution in the infralimbic or anterior cingulate cortices. Additionally, there were no differences of microglial density and distribution in the prefrontal cortex in females, indicating sex differences in the response of microglia to acute cannabis exposure in this region. Further, our preliminary data indicates no differences in microglial density distribution in the basolateral amygdala or throughout the dorsal hippocampus dentate gyrus in males, with ongoing work analyzing microglial density and distribution in the ventral hippocampus and in these regions in females. Additionally, we have found that in the prelimbic area, there was a change in microglial morphology. Specifically, in males we found that exposure to the CBD chemovar altered microglial arbor perimeter and circularity compared to control air and the Balanced chemovar, indicating a potential reduction in surveillance. Furthermore, in males, exposure to the Balanced chemovar compared to control air led to reduced microglial fractal dimension and increased microglial lacunarity, indicating less complex morphology and altered shape, respectively. However, in females, our preliminary morphological analysis indicated no effect of cannabis exposure on microglial morphological parameters in the prelimbic cortex.

Conclusions: Our preliminary data indicates that acute cannabis exposure modifies microglial distribution and morphology in the prelimbic cortex, particularly in males. We will next use scanning electron microscopy to investigate potential changes in microglial organelles and interactions with parenchymal elements. This work will lay the foundation for understanding how vaporized cannabis exposure alters microglial form and function. Future work will focus on the duration of these changes as well as the outcomes following chronic exposure.

Keywords: microglia, cannabis, microscopy, electron microscopy, prefrontal cortex

Disclosure: Nothing to disclose.

Michael Patton*, Trisha Vaidyanathan, Julia Dziabis, Zachary Schettewi, Joanna Kang, Staci Bilbo

Duke University, Durham, North Carolina, United States

Background: The serotonin system regulates a multitude of behaviors, and its dysfunction is evident across numerous neurological disorders. Current pharmacotherapies targeting this diffuse transmitter system generate unwanted side-effects, warranting more selective circuit-specific therapies. Serotonin neurons are one of the first neurological systems to arrive in the developing brain and proper circuit development is dependent on optimal brain serotonin levels. For example, perinatal manipulations of brain serotonin dramatically disrupt proper innervation into terminal fields resulting in lasting circuit dysfunction in adulthood. Therefore, studying mechanistically how serotonin circuits develop will provide critical insight into the etiology of serotonin dysfunction seen in neurological disorders.

Methods: 4-week old C57JB6 female mice were placed on a high saturated fat diet (45% kcal from fat) or a control diet (10% kcal from fat) for 6-week prior to mating and remained on diet until offspring were weaned at P28. Both male and female offspring were used in the following experiments to examine microglia interactions with serotonin circuit development.

Immunohistochemistry was performed on C57JB6 mice (P8 n = 5 males/5 females; P14 n = 4 males/4 females; P21 n = 5 males/5 females; P60 n = 18 males/ 22 females) targeting SERT and synaptophysin to measure serotonin synapse number. To measure microglia 5-HT phagocytosis, brain slices were incubated in primary antibodies targeting IBA1,5HT, and CD68. Confocal images were acquired with an Olympus Fluoview 3000 confocal microscope using a 60x oil immersion objective (1.3NA) and images were analyzed with Synbot analysis or IMARIS.

To measure serotonin release, mice (n = 12 males/ 12 females) were injected with the GRAB5-HT3.0 bilaterally into the NAc. And ex vivo 2-photon images were acquired using a Bruker Ultima 2P plus with a 20x immersion objective. 5-HT fibers were electrically stimulated with a bipolar concentric electrode positioned near the anterior commissure of the NAc delivering a 4V pulse at 4,8,12,20Hz for 3s.

To measure specific 5-HT circuit activity, mice (n = 24 males/ 24 females) were injected with a retrograde AAV expressing FLPo cre-dependently into either of OFC or NAc. An additional virus expressing a FLPo-dependent mCherry was injected into the DRN. Loose-seal cell-attached and current-clamp recordings were made with 4-6 MΩ resistant patch electrodes filled with potassium gluconate.

To quantify 5-HT receptor expression in microglia across brain regions, OFC and NAc were microdissected from p14 mice (N = 10 males/ 10 females) and microglia were isolated using magnetic CD11 beads. RNA was isolated using TRIzol, reverse transcribed to cDNA, and qPCR was run for 5-HT receptors 1A,1B,2A,2B,2C and 18S for a loading control.

To measure reversal learning behavior, adult P60-90 Mice (n = 12 males/ 12 females) food restricted mice learned to lever press on an FR1 schedule for one of two levers for a food reward. This reward contingency was then reversed to measure cognitive flexibility.

To measure voluntary ethanol consumption, male mice (n = 17) underwent a 1 week Drinking in the Dark paradigm with ad libitum access to food and a bottle containing 20% ethanol for 2-4h.

Comparison of two groups was performed with a student’s t-test, comparison of multiple groups was performed with ANOVA with multiple comparison correction used when necessary.

Results: 5-HT synapse density peaks at P14 in the NAc (F(2,25) = 1.557; p < 0.05),but peaks at P21 in the OFC (F(2,25) = 3.801; p < 0.05). Microglia phagocytosis of 5-HT also peaks at P14 in the NAc (F(2,14) = 1.290; p < 0.05). P14 microglia exhibit greater 5HT1B expression in the OFC (t(18) = 4.162; p < 0.05), but greater 5HT2C expression in NAc (t(18) = 3.651; p < 0.05).

Adult mHFD male offspring exhibit greater 5-HT synapses in the NAc, but not OFC, that is corrected by tryptophan supplement (F(2,30) = 4.374; p < 0.05). Adult female mHFD offspring exhibit fewer synapses in the NAc, but not the OFC, that is corrected by tryptophan supplement (F(2,36) = 14.81; p < 0.05). Adult male, but not female mHFD offspring exhibit greater 5-HT release into the NAc (males F(1,10) = 13.87; females F(1,6) = 024). 5-HT neurons projecting to the NAc exhibit decreased tonic activity in male but not female mHFD offspring (males t(10) = 4.175; females t(5) = 0.174). Adult male, but not female mHFD offspring exhibit faster acquisition rate of lever pressing behavior in an FR1 operant task (males t(13) = 2.693; females t(6) = 1; p < 0.05). However, both male and female offspring exhibit no difference in reversal learning (males t(12) = 0.7224; females t(6) = 1.434). Additionally male mHFD offspring voluntarily consume more 20% ethanol compared to control offspring (t(16) = 2.420; p < 0.05)

Conclusions: Here we demonstrate that mHFD disrupts serotonin circuit development in a sex specific manner. Additionally we demonstrate that not all serotonin circuits are sensitive to mHFD. We suggest that microglia heterogeneity promotes region-specific changes in serotonin innervation. Future investigation of microglia 5-HT receptor signaling will determine how mHFD selectively disrupts serotonin innervation to the NAc, but not the OFC, and how this impacts reward processing behavior in adulthood.

Keywords: microglia, Serotonin, dorsal raphe, Reward, Circuit Development

Disclosure: Nothing to disclose.

Daniel Weinberger, Joo Heon Shin, DongSan Kim*

Lieber Institute for Brain Development, Baltimore, Maryland, United States

Background: People of African ancestry have the richest source of genetic variation but are underrepresented in both genetic and genomic studies, despite a higher burden of many diseases, including brain-related disorders. To date, there have been no studies that have focused primarily on postmortem brain tissue from individuals of African ancestry. This new dataset will span genomic, transcriptomic and proteomic data, and will be the largest publicly available dataset to date. Distinct brain regions and individual cell populations will be explored across the human lifespan in African American samples of major neuropsychiatric disorders and neurotypical controls (no known conditions). These data will provide critical insight into current neuropsychiatric disease mechanisms and accelerate efforts towards the ultimate goal of developing new therapeutics and prevention strategies not just for African Americans, but for all people.

Methods: To address the disparity, we conducted 30x Whole Genome Sequencing (WGS) on 450 postmortem brain samples from African-Americans, including 140 neurotypical controls and 310 patients with schizophrenia, bipolar disorder, major depression, Alzheimer’s disease, Parkinson’s disease, and PTSD. From these samples, we collected RNA profiles from the dorsolateral prefrontal cortex (DLPFC) and hippocampus using bulk RNA sequencing (targeting ~80M paired reads) using Illumina TruSeq Stranded Total RNA with Ribo-Zero Gold HMR Kit. Additionally, we plan to perform single-cell RNA sequencing (snRNA-seq) on a subset of neurotypical control samples lifespan early to old (N = 75) across the DLPFC, hippocampus, and caudate nucleus. We’ve finished first batch of single-nuclei RNA sequencing from 32 donors (three regions per donor, total of N = 96) using 10x Genomics Chromium Single Cell 3’ Gene Expression technology (second batch from remaining 43 donors is currently being performed so that we’ll have enriched power).

Results: Following stringent data quality control of WGS, we identified 49,401,666 genetic variations, of which 18,614,848 are common (minor allele frequency > 0.01), including 15,386851 SNPs and 3,227997 indels. Differential gene expression analysis with genetic ancestry revealed more genes whose expression levels are associated with local ancestry than global ancestry. Differential gene expression analysis with disease states revealed 146 genes associated with Alzheimer’s disease in the hippocampus and 54 genes associated with schizophrenia in the DLPFC. In both brain regions, we detected significant expression quantitative trait loci (eQTLs) for nearly every gene (FDR < 0.05).

For snRNA sequencing, we initially generated noise-removed count matrices, resulting in 135,455, 147,588, and 123,122 nuclei for the DLPFC, caudate, and hippocampus, respectively. After further filtering for potential doublets, nuclei with high proportions of mitochondrial genes, or unstructured clustering, we finalized counts of 129,612, 139,567, and 117,162 nuclei, respectively. The number of nuclei averaged 4258 per brain sample, with an interquartile range (IQR) of 3110 to 5334. We annotated 24 subclasses (Non-neuronal: Oligodendrocyte, OPC, Astrocyte, Endothelial cell, VLMC and Microglial-PVM, excitatory neurons: L2/3 IT, L4 IT, L5 IT, L6 IT, L6 CT, L6 IT Car3, L5 ET, L5/6 NP, L6b, inhibitory neurons: Sst, Sst Chodl, Pvalb, Chandelier, Pax6, Lamp5 Lhx6, Lamp5, Sncg and Vip) using the human MTG class as a reference dataset.

We performed a linear mixed model analysis for each gene to investigate sources of gene expression variation and found that most genes exhibit variability depending on cell type classification and brain regions. Additionally, we identified some genes whose expression levels are correlated with global and local ancestry scores, even after controlling for other potential factors, suggesting that these genes may reflect or contribute to ancestry-specific traits.

To investigate this further, we performed eQTL analysis and identified numerous significant subclass-specific eQTLs (FDR < 0.2), underscoring the importance of cellular context in gene regulation, despite limited statistical power. On average, 91 eGenes were identified per analysis. However, the number of identified eGenes varied considerably across brain regions and subclasses (IQR = 3.5 to 150), with most eGenes being highly subclass-specific. For instance, we identified 200, 314, and 249 significant eGenes for oligodendrocytes in the DLPFC, caudate, and hippocampus, respectively. However, only 17 eGenes were consistent across all three brain regions. Colocalization of eQTLs with GWAS summary statistics detected novel risk genes for several brain disorders not identified using eQTLs derived from European-American samples.

Conclusions: Our findings highlight the value of including African ancestry samples in genetic studies, which may provide novel insights into the genetic architecture of brain disorders.

Keywords: RNA Sequencing, Genomics, Postmortem Brain Tissue

Disclosure: Nothing to disclose.

Todd Lencz*, Upasana Bhattacharyya, Jibin John, Max Lam

Zucker Hillside Hospital, Glen Oaks, New York, United States

Background: It has long been proposed that brain imaging measures may serve as “endophenotypes” or “intermediate phenotypes,” bridging molecular genetic variation and key clinical phenotypes such as schizophrenia (SCZ), bipolar disorder (BIP), and major depressive disorder (MDD). However, it is only very recently that large-scale, multimodal neuroimaging genetics data has become available. The UK Biobank has released data on approximately 4000 imaging-derived phenotypes (IDPs) in ~40,000 individuals, but these IDPs demonstrate high degrees of collinearity and can be difficult to interpret. Utilizing a novel clustering method that exploits the genetic covariance structure of the IDPs, the present study not only addresses the challenge of data abundance in the UK Biobank but also aims to uncover latent phenotypic representations, offering a more direct insight into the underlying biology of the brain and its associations with psychopathology at the genetic level.

Methods: First, neuroimaging IDPs were screened such that only traits with non-zero heritability were retained to reduce the likelihood of clustering noisy variables, leaving 1786 IDPs remaining. Utilizing the unsupervised approach we have previously published (Lam et al. 2022, Nat Comms, PMID: 36369282), we first employed Uniform Manifold Approximation and Projection for Dimension Reduction (UMAP) on the resulting 1786*1786 genetic correlation matrix. We then performed clustering, utilizing density-based spatial clustering of applications with noise (DBSCAN) across the first 3 UMAP dimensions, which resulted in 13 distinct clusters of neuroimaging phenotypes, each containing dozens to hundreds of IDP GWAS. Having identified their cluster membership, we then combined the respective IDP summary statistics (sumstats) into meta-analysis utilizing gPCA, to derive a single set of sumstats representing each IDP cluster.

Results: The clustering of neuroimaging IDPs mapped to identifiable brain features. Notably, these 13 IDP clusters were not simply defined by imaging modality (structural, functional, etc.), but revealed critical circuits (e.g., dorsal vs ventral processing streams) and regions (e.g., medial temporal lobe) across imaging modalities, suggesting that these clusters represent biologically meaningful neurocircuits. Genetic correlation analyses using Linkage Disequilibrium Score Regression (LDSC) of the 13 clusters identified novel associations with psychopathology, including the following Bonferroni-corrected significant results: SCZ risk with cluster 13 (resting state functional connectivity, Rg = -0.15, p = 0.0007); BIP risk with cluster 3 (subcortical volume, Rg = -0.11, p = 0.0009); and risk for MDD with cluster 7 (temporal lobe volume, Rg = -0.11, p = 0.0022). Nominally significant correlations were also observed between clusters representing subcortical measures and schizophrenia susceptibility.

Conclusions: Our multivariate cluster analytic approach successfully identified biologically interpretable imaging derived phenotypes in large-scale GWAS data from the UK Biobank. These phenotypes demonstrated specific patterns of genetic correlation with psychopathology, emphasizing a role for genetically-influenced, smaller subcortical volumes in susceptibility to schizophrenia and bipolar disorder. Temporal lobe volume was specifically related to MDD susceptibility.

Keywords: Imaging-genetics, GWAS, endophenotypes

Disclosure: Nothing to disclose.

Daniel Mueller*, Xiaoyu Men, Samar Elsheikh, Daniel M. Blumberger, Jordan Karp, James Kennedy, Charles Reynolds, Eric Lenze, Benoit Mulsant

University of Toronto, Toronto, Canada

Background: Older adults are more susceptible to antidepressant-induced side effects due to higher comorbidity, cognitive decline, and polypharmacy. Onset of side effects is also associated with antidepressant discontinuation, which might increase the risk of relapse or recurrence of depression. Therefore, it is important to investigate genomic factors underlying antidepressant-induced side effects. Here, we used polygenic risk score (PRS) analyses to evaluate the shared genetic architecture between venlafaxine-induced side effects and other disorders of relevance.

Methods: We analyzed genetic and clinical data from participants enrolled in the Incomplete Response in Late-Life Depression: Getting to Remission study (IRL-GRey, NCT00892047) phase 1, where participants received venlafaxine for 12 weeks, with dosage up to 300mg/day (Lenze et al., 2015). Side effects were assessed at the end of phase 1 using the 46-item Udvalg for Kliniske Undersøgelser (UKU) rating scale. The presence of side effects was identified by a two-point increase in UKU scores compared to baseline measurements. Additionally, according to UKU categorization, adverse effects were categorized as psychic (10 items), neurological (9 items), autonomic (11 items), and other (16 items). PRSice-2 was utilized to construct the PRSs on our target sample (IRL-GRey) using 14 summary statistics from the PGC, UK Biobank, IGAP, and MEGASTROKE consortium. We tested the association between the total 14 PRSs and the presence of at least one adverse effect. Furthermore, the presence of any psychic side effects was tested for their associations with six PRSs for psychiatric disorders (i.e., depression, bipolar disorder, schizophrenia, and antidepressant treatment response), while the presence of neurological side effects was analyzed with three PRSs for Alzheimer’s disease. Logistic regression was utilized to test for these associations adjusting for age, sex, and first three ancestry principal components. Due to the limited sample size in other populations, our PRS analyses were restricted to the European-ancestry subsample, and only discovery cohorts with predominantly European ancestry were included. The results were corrected for multiple testing using a stringent Bonferroni correction with adjusted ɑ = 0.05/(14 + 6 + 3) = 0.0022.

Results: A total of 297 individuals were included in the analyses. Overall, higher PRSs for all stroke (OR = 1.43 [1.11, 1.85], p = 0.006, empirical-p = 0.03), ischemic stroke (OR = 1.44 [1.10, 1.88], p = 0.008, empirical-p = 0.04), and small vessel stroke (OR = 1.41 [1.10, 1.82], p = 0.007, empirical-p = 0.03), were nominally associated with a higher likelihood of experiencing at least one side effect after venlafaxine treatment in the IRL-GRey sample. However, none of these nominally significant associations survived multiple testing corrections before or after permutation.

Conclusions: Our findings indicate nominal genetic associations between venlafaxine-related side effects and PRSs for stroke in older adults. Notably, our previous analyses indicated PRS for stroke also to be associated with non-remission in the same sample (Marshe et al., 2021). By identifying individuals at higher risk of side effects based on their genetic profile, treatment plans could be tailored to minimize side effects and optimize treatment. Further research is needed to validate these preliminary findings in larger cohorts and diverse populations of older adults.

Keywords: pharmacogenetics, Depression, Venlafaxine, older adults, Genomics

Disclosure: Nothing to disclose.

Hayley Thorpe, Anne Canella, Hakan Kayir, Malik Talhat, Ahmad Hassan, Yu Gu, Maryam Khan, Ron Johnson, Boyer W Winters, Sandra Sanchez-Roige, Abraham Palmer, Jibran Khokhar*

Western University, London, Canada

Background: Genetic variation in Cell Adhesion Molecule 2 (CADM2) is among the strongest and most reliable genetic predictors of lifetime cannabis use as well as cannabis use frequency. CADM2 is thought to affect reward circuitry and neurotransmitter systems acted on by the primary psychoactive constituent of cannabis, Δ9-tetrahydrocannabinol (THC). However, CADM2’s relationship with cannabis use may be proximal to its other genetic associations that are also predictive of substance experimentation (e.g., impulsivity and risk-taking behavior). Whether CADM2 variation predicts cannabis use vulnerability by affecting the pharmacological actions of cannabis and THC has yet to be established. Given that CADM2 associations with cannabis use are more robust than other commonly used substances, we hypothesized that deletion of the mouse Cadm2 ortholog would reduce cannabis and THC intake, as well as alter pharmacological and neural responses to THC.

Methods: Using male and female Cadm2 mutant knockout (KO), heterozygous (HT), and wildtype (WT) mice, we explored Cadm2 associations with cannabis and/or THC drug preference, cognition, pharmacology, and prelimbic neuron activity. Voluntary intake of THC and commercial cannabis oil (n  =  11-18 mice/genotype/experiment) was assessed with a two-edible choice preference test, in which drug dough intake was recorded across consecutive days and escalating doses (0.2-2.0mg THC/g of dough). Drug- and genotype-dependent differences in cognition were evaluated with the 5-choice serial reaction time task (5CSRTT; executive functioning) and prepulse inhibition test (PPI; sensorimotor gating) 30 minutes after administering 2mg/kg, i.p. of THC (n  =  12-16 mice/genotype). Blood serum levels of THC and its bioactive metabolite 11-OH-THC were measured 30 or 120 minutes after delivering 10mg/kg, i.p. of THC (n  =  12-13 mice/genotype/time) using liquid chromatography-tandem mass spectrometry. Pharmacodynamic response to 3 or 10mg/kg, i.p. of THC was assessed with the cannabinoid tetrad assay following acute and chronic THC exposure (n  =  12-16 mice/genotype/dose). Finally, whole-cell patch clamping was used to determine the electrophysiological properties of layer 5 pyramidal WT and KO neurons in prelimbic cortical slices at baseline (n  =  49-56 neurons/genotype). We also investigated changes to spontaneous excitatory or inhibitory postsynaptic currents (sEPSC/sIPSC; n  =  18-27 neurons/genotype/recording) following bath with 1µM THC. All parametric data were analyzed using one-way or repeated measures ANOVA; nonparametric data were analyzed using Kruskal-Wallis H test or aligned rank transformed ANOVA where appropriate.

Results: Genotype significantly affected preference for THC (F(2,32)  =  4.85, p  <  0.05) and cannabis oil-containing edibles (F(2,34)  =  4.65, p  <  0.05), with KO mice showing lower preference for both drugs in comparison to WT and HT littermates. THC challenge before the 5CSRTT did not modify impulsivity or attentional performance across any genotype, and genotype did not differentially affect serum levels of THC or 11-OH-THC at either time. Cadm2 genotype did not affect acoustic startle under drug-free conditions. However, we observed a main effect of genotype on prepulse inhibition (F2,34.02 = 6.61, p = 3.77E-03) suggesting that Cadm2 KO mice show impaired sensorimotor gating compared to WT (p = 8.41E-03) and HT (p = 0.01) littermates. When administered 2mg/kg THC 30 minutes prior to testing, a similar genotype effect was observed (F2,34.00 = 5.79, p = 6.87E-03) though PPI was only lower in KO mice compared to WT (p = 0.01)m and HT mice no longer differed from KO mice. Baseline locomotion (F(2,76)  =  37.90, p  <  0.001) and thermal pain tolerance (F(2,77)  =  4.64, p  <  0.05) varied by genotype such that KO mice were hyperlocomotive and had reduced thermal nociception compared to WT littermates. In the cannabinoid tetrad assay, KO mice showed sensitization to the antinociceptive properties of THC (3mg/kg: F(2,32)  =  6.23, p  <  0.01), had no hypothermic response to THC (10mg/kg: F(2,36)  =  13.52, p  <  0.001), and showed exacerbated hyperlocomotive response to THC (3mg/kg: F(2,34)  =  50.92, p  <  0.001; 10mg/kg: F(2,35)  =  63.69, p  <  0.001). Finally, whole-cell patch clamp electrophysiology revealed greater membrane resistance (F(1,99)  =  6.71, p  =  0.01) and lower resting membrane potential (F(1101)  =  3.99, p  <  0.05) in KO mice compared to WT. In neurons from male KO mice, after hyperpolarization amplitudes (F(2.12,209.55)  =  7.23, p  =  7.24E-04) and firing frequency (F(1.76,177.60)  =  3.61, p  =  0.03) at lower stimulation intensities were attenuated compared to WT neurons. There were no genotype differences in sEPSC/sIPSC changes after THC bath application.

Conclusions: These data implicate Cadm2 in cannabinoid preference and consumption, pharmacodynamic response to THC, and neuronal activity in a brain area relevant to executive functioning and drug intake behavior. The reported Cadm2-dependent differences in THC pharmacodynamics could suggest these mice are uniquely susceptible the effects of THC and cannabis, thus explaining reductions in voluntary drug intake. This preclinical evidence supports the validity of human genetic associations between CADM2 polymorphisms and cannabis use risk, and points to a dual role for this gene in mediating both cognitive features predictive of substance use and subjective cannabis response.

Keywords: Cell adhesion molecule, cannabis use, delta9-tetrahydrocannabinol, Genetic variability

Disclosure: Nothing to disclose.

Natasha Petry*, Joel Van Heukelom

North Dakota State University, Fargo, North Dakota, United States

Background: Sanford Health (SH), the largest rural health system in the United States, is headquartered in Sioux Falls, SD, and has 48 medical centers and 211 clinic locations including major hubs in North Dakota and Minnesota as well as clinic locations in other states including Iowa, Montana, and Nebraska.

In an effort to provide personalized treatment (or: precision medicine), SH offers pharmacogenetic (PGx) testing based on recommendations elaborated by expert groups (e.g., Clinical Pharmacogenetics Implementation Consortium) or regulatory agencies (e.g., the FDA). At present, recommendations mainly relate to the CYP2D6 and CYP2C19 genes as these genes are highly variable, affecting serum level concentrations which in turn are associated with response and adverse effects to many antidepressants. The aim of this study is to quantify how many visits in the PGx clinic are tied to psychiatric diagnoses, and as a secondary aim, for which specific diagnoses patients were referred or self-referred themselves.

There are robust research plans for clinic data. This abstract examines an initial look at psychiatric related PGx clinic visits.

Methods: Patients from a four state region in the upper Midwest can schedule in-person or video visits with the PGx clinic where they are seen by a clinical PGx pharmacist and genetics nurse practitioner. Pre-test and post-test visits are offered. Pre-test visits generally offer internal testing at SH, but external tests may be ordered pending patient’s clinical and financial situation. All PGx results can be reviewed at post-test visits. Visits may occur days after receiving results and in some cases even years after pending patient and provider preferences and referral requests. After a visit, a note is placed in the patient’s chart and sent to the referring provider with recommendations. This PGx clinic is a subset of the overall PGx service and has previously been described. (1) The standard process for internal PGx results at SH includes a review by a pharmacist who writes a chart note with findings (> 30,000 to date). Only a small subset of patients are seen in the PGx clinic, either by provider or self-referral.

During patient’s first visit to the PGx clinic they are offered to participate in clinic-based research and can either consent to ongoing review of their medical record data and willingness to participate in surveys. Medical records were abstracted for patients who consented to research from the onset of the PGx clinic in April 2022 through July 2024. This research was reviewed and approved by the SH IRB.

Results: Total patients at the time of data collection was 261. Out of the 261 identified patients, 207 ( ~ 79.3%) had psychiatric related diagnosis codes at the time of visiting the clinic. Only 65 had one psychiatric-related diagnosis code (anxiety only, depression only, “anxiety and depression” or “depression and anxiety” coded together as one, or other). Of the remaining patients, 142 were linked to multiple psychiatric-related diagnosis codes with 139 linked to anxiety, 140 linked to depression, in addition to five other psychiatric-related disease state codes.

Approximately 75% of patients seen in clinic (n = 155) had both anxiety and depression coded as a part of the visit. Nearly all patients seen in clinic with a psychiatric-related diagnosis (99.5%) had either anxiety and/or depression linked to the visit.

Conclusions: The PGx clinic is a newer service line at SH, and thus it is important to evaluate target patient populations. Notably, the majority of patients seen in SH’s PGx clinic have a psychiatric-related diagnosis. Primary care providers are often responsible for patient’s mental health care in addition to other conditions. A large majority of patients with behavioral health needs visit their primary care provider at least annually. (2) Primary care providers often seek assistance managing mood disorders and PGx results and refer to the PGx clinic and psychiatry. Psychiatry has a long wait time in SH’s region. Psychiatric medications often require interpretation of multiple metabolism pathways adding to complexity. This clinic can provide additional expertise to patients and support for providers.

However, this project bears some limitations: Only patients who consented to PGx clinic research were included. Some patients who did not consent to be included in the study but who were seen in clinic did have psychiatric diagnoses. Moreover, data is limited to what was available upon data abstraction. Genotype information (allowing to analyze the portion of non-normal metabolizers) was not available through this round of data abstraction but is planned for future analysis.

Given the large percentage of patients seen for psychiatric-related diagnosis codes in PGx clinic, this suggests a need within the health system for support caring for patients with psychiatric related diagnoses. PGx clinic staff should continue to stay current on psychiatric medications and recommendations. Further research should examine CYP2D6, CYP2C19, and CYP2B6 metabolizer results for patients seen in clinic, health outcomes, and economic analyses related to PGx clinic visits.

References:

1. Van Heukelom J, et al. Evolution of pharmacogenomic services and implementation of a multi-state pharmacogenomics clinic across a large rural healthcare system. Front Pharmacol. 2023 Nov.

2. Benefits of integration of behavioral health. Primary Care Collaborative. Accessed July 26, 2024. https://www.pcpcc.org/content/benefits-integration-behavioral-health.

Keywords: pharmacogenetics, Clinical practice, depression and/or anxiety, Pharmacogenomic-guided treatment recommendations, education

Disclosure: Nothing to disclose.

Craig Erickson*, Lauren Schmitt, Sebastian Pilito, Kyle Cullion, Paul Horn, Matilyn Shanahan, Christina Gross

Cincinnati Children’s Hospital Medical Center, Cincinnati, Ohio, United States

Background: Fragile X Syndrome (FXS) is causes by a CGG repeat expansion in the promoter region of the fragile X messenger ribonucleoprotein (FMR1) gene. The expansion, > 200 CGG triplet repeats, results in gene methylation and subsequent silencing of fragile X messenger ribonucleoprotein (FMRP) expression. FXS clinical testing with Southern Blot and PCR typically notes a CGG repeat count summarized as > 200 and methylation of the gene that is described as unmethylated, partially methylated or fully methylation with the large majority of patients with > CGG having Southern Blot and PCR reported full methylation. We have discovered that half of males with FXS with clinical testing noting full methylation express trace or very low levels of FMRP in their peripheral blood. We pursued novel techniques to best describe gene methylation and repeat count to determine what molecular factors not captured by clinical testing are responsible for discovered variation in FMRP expression in patients previously thought to have complete FMR1 gene silencing.

Methods: We completed Cas9-assisted FMR1 gene isolation and Oxford Nanopore (ONT) long read sequencing on a sample of 28 individuals with FXS (23 males, of whom 11 are described as fully methylated full mutation on Southern Blot and PCR). ONT output was then analyzed using METEORE programs Nanopolish pipeline to detect CpG methylation and WarpSTR to detect number of trinucleotide repeats for each participant. We additionally had ONT repeat count and methylation data on 18 subjects who additionally had FMRP levels determined using our published blood spot-based human peripheral blood FMRP assay.

Results: Within this initial pilot sample, we demonstrated methylation load and CGG repeat count positively correlate with each other (r = 0.63, p < 0.001). In the sample of individuals who have both ONT and FMRP data (n = 18), both methylation load (r = -0.59, p = 0.008) and repeat count (r = -0.59, p = 0.01) each negatively correlated with FMRP. Thus, we demonstrate the feasibility of our approaches as well as initial proof of concept that variation in methylation and repeat count corresponds to variation in FMRP expression. In fact, when examining only males with FXS, FMRP was strongly correlated with methylation load (r = -0.61, p = 0.01) and repeat count (r = -0.62, p = 0.01). We also have critically found that methylation load and CGG repeat count varies even among individuals diagnosed as “full mutation, fully methylated” by standard diagnostic techniques. For example, CGG repeat count and methylation load are marginally increased in FM-FM with no FMRP expression compared to FM-FM with trace FMRP expression (p’s < 0.07). Together, our initial preliminary findings highlight both the limited resolution of current diagnostic methods and the promise ONT offers.

Conclusions: Although our initial studies indicate that FMRP accounts for roughly half of the variance in IQ in FXS, FMRP is expected to be less stable compared to methylation load and CGG repeat count, suggesting gene methylation and/or repeat count may be better suited for predicting disease severity and outcomes. In our preliminary FXS sample who completed ONT and IQ testing (n = 20), we found a significant relationship between higher methylation load and lower IQ (r = -0.63, p = 0.003). Following the exciting preliminary results from our pilot sample, we have more recently started to successfully implement Methyartist to detect CpG methylation and Hidden-Markov Model STR (HMMSTR) copy-number caller optimized for sequence-based targeted sequencing data for more accurate data analysis.

Keywords: fragile X syndrome, CpG Methylation, Molecular Genetics

Disclosure: Nothing to disclose.

Jose Pardo*, Ambreen Kanwal, Sadaf Naz

University of Minnesota, Minneapolis, Minnesota, United States

Background: Psychosis is a cardinal feature of schizophrenia, and schizophrenia is the most common psychotic disorder. Only a few Mendelian genes (monogenic, large effects on phenotype, rare) have been associated with typical psychosis or schizophrenia in the absence of comorbidity (e.g., dysmorphia, intellectual disability, autism). A common approach in medical genetics finds such genes by genome or exome sequencing of individual consanguineous families showing apparent recessive inheritance, autozygosity mapping, and filtering out candidates unlikely to prove causal. However, this approach has seen limited application to psychiatric genetics. We aimed to identify single variants of large effects in psychiatric disorders.

Methods: Exome sequencing of all available members of individual consanguineous multiplex families was performed (Kanwal et al. J Psychiatry Neurosci 2022 and Genes 2023) followed by genome sequencing for one of the families. Autozygosity mapping was used to link genomic regions of interest to phenotypes within individual pedigrees. Data were filtered using unbiased a priori criteria (e.g., rare < 1%; damaged protein predicted by software; conserved amino acid; absence as homozygous in large number of individuals in datasets; potential splicing variants; no CNV).

Results: These variants occurred in USP53, RGS3, and IL1RAPL1. USP53 is a newly identified synaptic protein; RGS3 is implicated in sensory behavior in C. elegans, while IL1RAPL1 variants cause nonsyndromic X-linked intellectual disability sometimes accompanied by psychiatric manifestations.

Conclusions: Mendelian genes associated with common-appearing phenotypes in psychiatric disease may arise in multiplex consanguineous families consistent with recessive transmission. WGS, autozygosity mapping, and unbiased filtering of variants unlikely to be causal produced one variant of interest per pedigree. Studies are underway to define potential mechanisms of pathogenicity and causality.

Keywords: DNA, whole-genome, sequencing, schizophrenia, Identical by descent, ubiquitination, RGS3, IL1RAPL1

Disclosure: Nothing to disclose.

Laramie Duncan*, Tayden Li, Madeleine Salem, Will LI, Leili Mortazavi, Naghmeh Shargh, Hazal Senturk, Sam Vesuna, Jong Yoon, Gordon Wang, Jacob Ballon, Longzhi Tan, Scott Pruett, Brian Knutson, Karl Deisseroth, Will Giardino, Katy Werwath

Stanford University, Stanford, California, United States

Background: Despite the success of genome-wide association studies (GWAS) of schizophrenia and other psychiatric disorders, biological interpretation of GWAS results in psychiatry has proven difficult. Fortunately, technological advances have enabled comprehensive surveys of human brain cell types using RNA sequencing of single cells and single nuclei (snRNAseq). Combining these two landmark resources: 1) genome-wide data from GWAS and 2) brain-wide data from snRNAseq, it is now possible to infer causal cell types for psychiatric disorders and other complex brain phenotypes.

Methods: Using the MAGMA gene property analysis, we tested 461 human brain cell types to determine which cell types preferentially express genes found to be associated with specific phenotypes in human GWAS.

Results: This analysis successfully identified true positive cell types (e.g., hypothalamic TAC3, KISS1 neurons for hot flashes) and true negatives, along with many novel discoveries about cell types for psychiatric disorders. The top cell type for schizophrenia in this “unbiased” analysis (i.e., brain-wide and genome-wide) was a cortical somatostatin interneuron type (cell type #239, p = 4.3x10-17). Nine other cell types from cortical regions (including prefrontal and retrosplenial cortex) and subcortical regions (amygdala, thalamus, and hippocampus) were identified as significant independent cell types for schizophrenia. Comparing across phenotypes, amygdala neuronal associations were particularly noteworthy for PTSD, brainstem neurons for sleep, and non-neuronal associations for multiple sclerosis (T-cells) and Alzheimer’s disease (microglia).

Conclusions: Taken together, these results suggest that cellular profiles may help to classify psychiatric disorders in a data-driven, biological manner, and cellular associations may accelerate progress toward improved therapeutics for psychiatric disorders.

Keywords: GWAS, RNAseq, Schizophrenia (SCZ), PTSD, hot flash

Disclosure: Nothing to disclose.

Katelyn Oliver*, Alyssa Roeckner, Cecilia Hinojosa, Justin Santos, Linzie Taylor, Kristina Dahlgren, Helena Zeleke, Amy Murphy, Colin Johnson, Dasani DelRosario, Timothy Ely, Rebecca Hinrichs, Natalie Merrill, Kim Wallen, Vasiliki Michopoulos, Marisa Young, Andrea Braden, Jennifer Stevens

Emory University, Atlanta, Georgia, United States

Background: People with PTSD tend to show a heightened bias toward remembering negative events. Interestingly, studies within non-clinical populations show that women and men demonstrate natural differences in emotional and episodic memory. Rodent literature suggests that these sex differences may be partially attributed to the influence of sex hormones like estradiol (E2) on key emotional memory-related brain regions such as the hippocampal complex and amygdala. However, studies of the effects of E2 on human memory are sparse. As trauma-related disorders are more prevalent in women than men, it is important to understand how fluctuations in sex hormones like E2 may be impacting the neural mechanisms of threat-associated memories in women. Here, we used a categorical conditioning paradigm with women to study the impacts of E2 on threat-enhanced declarative memory processes.

Methods: N = 45 naturally cycling Black and African American women (ages 18-35) tracked their menstrual cycles using urine ovulation tests. During the early-luteal phase of the menstrual cycle, participants received a transdermal E2 or placebo patch in a randomized double-blind, cross-over design. The following day, blood samples were taken to assess serum estradiol, and participants underwent a categorical threat conditioning task during fMRI. Post-scan, participants completed a recognition test to assess their memory of the images seen during the scan. Recognition for previously threat-associated (CS + ) and non-threatening (CS-) images was scored by subtracting the correct recognition of images from the fMRI task with similar but novel images (Hits – False Alarms). CS+ and CS- recognition scores were then subtracted to observe threat enhancement on memory (CS + > CS-). The next month, participants completed the same experimental procedures under the opposite patch assignment. Mixed-effects models were used to evaluate E2 * [CS + > CS- recognition memory] interactions with ROI activity in the amygdala, hippocampus, and cortical components of the medial temporal lobe memory system (parahippocampal, perirhinal, and entorhinal cortices; PHC, PRC, ERC respectively).

Results: Participants learned the cue-contingency associations for the CS+ and CS- images (contingency awareness, CS + = 0.61, CS- = 0.26, t(107) = 9.94, p < 0.001). Threat conditioning predicted deeper encoding of the CS cues, with greater recognition for CS+ images compared to CS- (hits-false alarms for CS+ vs CS-, t(86) = 2.05, p = 0.043, mean CS + = 0.57 [SD = 0.21], mean CS- = 0.57 [SD = 0.21]). Additionally, transdermal E2 patches successfully increased circulating E2 levels (t(54) = 3.98, p = 0.002, mean E2 = 208.3 pg/mL [SD = 81.4], mean Placebo= 126.8 pg/mL [SD = 72.3]). CS + > CS- recognition scores were not impacted by E2 supplementation (p = 0.28, mean Memory-E2 = 0.037 [SD = 0.2], mean Memory-Placebo= 0.045 [SD = 0.15]). In the placebo condition, greater bilateral ERC and right PHC activation during CS + > CS- encoding was associated with threat-related enhancement of recognition in the later memory test, whereas during E2, activity in these regions had no impact on threat-related enhancement of recognition (E2*MemoryCS + > CS-, rPHC p = 0.045, ERC p = 0.002).

Conclusions: While E2 supplementation did not impact recognition behaviorally, increased hippocampal complex activity during threat encoding predicted an increase in threat recognition during placebo only. This indicates that E2 may dampen neural function in regions related to the encoding of threatening episodic memories. While prior literature has been mixed, some rodent studies report that greater E2 may lessen fear acquisition. Likewise, human studies on declarative memory suggest that greater E2 may protect against stress induced memory deficits. Perhaps E2 can help reduce the deeper decelarative encoding of threat-conditioned cues by altering neural patterns related to threat encoding. This has important implications for understanding the functional mechanisms trauma-related disorders in women, and may help inform new treatment options.

Keywords: Threat Conditioning, Estradiol, Episodic Memory, Functional MRI (fMRI), Hippocampal Complex

Disclosure: Nothing to disclose.

Zhenfu Wen, B. Isabel Moallem, Noor Nasar, Hannah McManus, Claudia Becker, Nicolina Bruno, Anna Chemplayil, Zakiya Atiyah, Isabella Mason, Mary-Rose Silvas, Joe Borrelli, Mohammed Milad*

University of Texas-Health Sciences Center at Houston, Houston, Texas, United States

Background: Exaggerated fear and avoidance characterize anxiety and fear-related disorders. While fear processing in the human brain is widely studied, less research exists on the neural mechanisms of active avoidance. Here we used an associative learning paradigm to study the neural correlates of active avoidance and how they intersect with those underlying threat conditioning.

Methods: Participants (n = 59) completed a two-day threat conditioning and active avoidance paradigm (CAAP), composed of four phases. Phase 1 is threat conditioning, in which two colored stimuli were followed by electric shock (CS + ), another colored stimulus was followed by no shock (CS-). Phase 2 is avoidance conditioning, in which participants were allowed to press a button at the beginning of each CS presentation. This action could possibly prevent the shock delivery; button pressing could effectively cancel the shock at the end of one CS+ (CS + EA), but not the other (CS + UA). Participants then undergo extinction learning, which is phase 3. During this phase, CS + EA and CS- were presented without shock and without button pressing option. Phase 4 took place 24 hours later, which is the extinction memory retrieval phase. During this phase, participants were provided the option to press a button while CS + EA, CS + UA, and CS- were presented without shock. Skin conductance response (SCR) and brain activations during each phase were examined with a focus on CS + EA vs. CS + UA. Previously established decoders for threat and safety detection (Wen et al., 2024) were used to examine the neural representations of conditioned threat across phases. Replication analyses were also conducted on an independent cohort (n = 32) that underwent similar experimental settings but using a different scanner and at a different institution.

Results: Analyses of phase 1 data were conducted but are not included here for lack of novelty (threat conditioning has been extensively analyzed and published). Avoidance learning engaged brain regions including the striatum, amygdala, and ventromedial prefrontal cortex (CS + EA vs. CS + UA, small-volume correction, voxel-level p < 0.005, group-level FWE-corrected p < 0.05). In late avoidance learning, neural representation of conditioned threat was lower during CS + EA processing compared to CS + UA processing (p = 0.008). In day 2, participants exhibited lower threat expression to CS + EA compared to CS + UA, as indexed by SCR (p = 0.052) and neural patterns (p = 0.004). Activation patterns during avoidance learning, generalization of the decoders, and SCR patterns were all replicated using the independent cohort. Results from the independent cohort: decreased threat expression to CS + EA compared to CS + UA in late avoidance learning (p = 0.025, one-tail t-test) and memory retrieval (p = 0.031, one-tail t-test).

Conclusions: Our results show that successful avoidance learning leads to activation of neural regions associated with safety and reward. Furthermore, neural representation of threat is dynamically modified during avoidance learning, and are specific to stimuli that were associated with successful avoidance behaviors. Results from our independent cohort show that these findings are generalizable and robust.

Keywords: Fear conditioning and extinction, threat detection, avoidance learning

Disclosure: Nothing to disclose.

Ga In Shin, Megan Quarmley, Vishnu Murty, Johanna Jarcho*

Temple University, Philadelphia, Pennsylvania, United States

Background: To adaptively navigate our social world, it is important to store memories of prior social interactions to inform our approach to future social interactions. A large body of behavioral research shows that memories can be distorted by schemas. We recently demonstrated that a strong positivity bias for social feedback that may help buffer against social anxiety (Johnston et al., 2023). However, the neural systems underlying accurate and distorted social memories are poorly mapped. Isolating the putative circuits underlying the positivity bias may be particularly important given its association with beneficial psychosocial outcomes. Previous research shows that the anterior hippocampus (aHPC) is involved in successfully encoding social feedback, while the ventromedial prefrontal cortex (vmPFC) is implicated in sustaining social schemas. Thus, the positivity bias be an outgrowth of greater engagement of aHPC during the encoding of positive, rather than negative, feedback. However, aHPC-vmPFC connectivity during encoding may also contribute to schema-based distortions in recall. In this study, we examined the relation between accuracy of memory and engagement of aHCP and vmPFC during the encoding of positive and negative social feedback, and the extent to which aHCP-vmPFC functional connectivity predicts the positivity bias.

Methods: Participants (N = 38; 63% female; 21.50 □ 4.10 years) underwent fMRI while completing the Recall After Feedback Task (RAFT). They first completed the feedback encoding phase of the RAFT. For each trial, participants selected which of two purported peers liked or disliked them based on the photo that had been sent to them. They then received positive or negative feedback from the selected peer. After scanning, participants completed the surprise cued response phase of the RAFT. For each purported peer they recalled receiving feedback from, participants indicated whether the feedback was positive or negative. Proportion of memory errors towards positive versus negative feedback was calculated for each participant. An independent sample t-test of bias scores confirmed the presence of a positivity bias. For individual level fMRI analyses, each trial was parametrically modulated based on whether subsequent memory for feedback was accurate (+1) or inaccurate (-1). Two types of group level analyses were performed. First, we tested the extent to which univariate activity in aHCP and vmPFC during the encoding of positive or negative feedback differentially predicted subsequent accuracy of recall. Independent sample t-tests were performed on parameter estimates extracted from each region for positive versus negative feedback. Significant results were probed with follow-up t-tests for positive and negative feedback. Second, we tested the extent to which aHPC-vmPFC connectivity during the encoding of positive or negative feedback predicted accurate compared to inaccurate memory. To accomplish this, we performed psychophysiological interaction (PPI) analysis utilizing a parametric modulation approach for accurate (+1) and inaccurate (-1) memory, and again subjected parameter estimates to independent sample t-tests.

Results: The RAFT successfully elicited a positivity bias such that participants were more likely to erroneously recall more instances of positive than negative feedback (t(37) = 6.19, p < 0.001). Univariate analysis revealed accuracy for social memory was differentially predicted by engagement of aHPC during the encoding of positive compared to negative feedback (t(37) = 2.93; p = 0.005). Specifically, aHPC activity predicted subsequent memory for positive (t(37) = 4.0, p < 0.001) but not negative feedback (t(37) = -0.371, p = 0.71). No such relation was seen in the vmPFC (p > .10). PPI analyses demonstrated that greater aHPC-vmPFC connectivity resulted in greater distortions in memory for social feedback (p < 0.05), regardless of whether encoding happened during positive or negative feedback.

Conclusions: These results align with models suggesting that engagement of aHPC, not vmPFC, during encoding facilitates accurate recall of social feedback. However, we expand on this work by demonstrating specificity of effects in the context of positive, rather than negative feedback. Moreover, we show that functional connectivity with vmPFC can distort otherwise accurate memories encoded by aHPC, which also promotes a positivity bias. Results enhance our understanding of the nuanced relations between brain regions involved in social cognition and memory formation, and suggest potential brain-based targets for interventions aimed at mitigating memory distortions that often occur disorders such as social anxiety.

Keywords: Memory Bias, Social Feedback, Anterior Hippocampus, vmPFC

Disclosure: Nothing to disclose.

Ashley Keiser*, Tri Dong, Hassan Shaikh, Scott La Tour, Dina Matheos, Eniko Kramár, Chris Butler, Siwei Chen, Joy Beardwood, Agatha Augustynski, Ameer Al-Shammari, Yasaman Alaghband, Vanessa Alizo Vera, Nicole Berchtold, Sharmin Shanur, Pierre Baldi, Carl Cotman, Marcelo Wood

Arizona State University, Mesa, Arizona, United States

Background: Our recent studies show that 2 weeks of consistent exercise creates and maintains a molecular memory window for cognitive benefits in mice. A brief 2-day exercise session can re-engage these benefits, re-facilitate long-term potentiation, and enable learning under subthreshold conditions. During these exercise conditions, the type I receptor for the TGFβ family, Acvr1c, is elevated in the hippocampus during consolidation. In sedentary mice, Acvr1c levels decline with age and in an Alzheimer’s Disease (AD) 5xFAD mouse model. Here, we probe the role of ACVR1C in long-term memory formation and begin to define an epigenetic mechanism involving Acvr1c regulation responsible for maintaining cognitive benefits underlying this molecular memory window. We also assess the impact of consistent vs. intermittent exercise on the duration of cognitive benefits and hippocampal regulation of Acvr1c.

Methods: We investigated the role of ACVR1C in hippocampus-dependent long-term memory formation in sedentary male mice using intra-hippocampal delivery of AAV1-ACVR1C constructs that either enhance or disrupt function. Sedentary mice were trained using either a subthreshold (3 min) or standard (10 min) Object Location Memory (OLM) task, and memory was tested the following day. Additionally, we evaluated whether enhancing ACVR1C could improve memory in 18-month-old wildtype and 12-month-old 5xFAD male mice using virus-mediated overexpression of wildtype ACVR1C. To assess epigenetic regulation, we performed chromatin immunoprecipitation (ChIP-qPCR) on the hippocampus of exercising male mice during consolidation, focusing on Acvr1c and Bdnf IV genes. We also compared the effects of consistent (14 days) vs. intermittent (14 days over 7 weeks) exercise, termed the “weekend warrior” group, on duration of cognitive function and hippocampal Acvr1c regulation in male mice. Mice were trained with a subthreshold (3 min) OLM task, and memory was tested the following day.

Results: In sedentary mice, we demonstrate that disrupted ACVR1C function under adequate learning conditions in adults impairs memory (t (17) = 4.65, P = 0.0002). Conversely, overexpression of Acvr1c enables learning under inadequate training conditions in adults (t (18) = 3.303, P = 0.004) and ameliorates cognitive impairment in aging (18 mo. C57: (t (12) = 2.350, P = 0.036)) and 12 mo. 5xFAD male mice: (t (21) = 2.287, P = 0.032). Within the context of exercise, we find that exercise, in any amount, alleviates epigenetic repression at the Acvr1c (F (5,49) = 9.377, P < 0.0001) and Bdnf IV (F (5,53) = 13.90, P < 0.0001) promoters during consolidation in a persistent manner. Mice engaging in consistent and intermittent (weekend warrior) exercise yield similar cognitive benefits as displayed by a higher object location discrimination index compared to the 2-day exercise control group (F (5,39) = 8.91, P < 0.0001), and no significant differences in the DI score between mice in the continuous vs intermittent exercise group (P = 0.9999). However, engagement in intermittent weekend warrior exercise results in cognitive benefits that are long lasting following the cessation of exercise as observed by similar DI scores in mice following 7-day exercise cessation vs mice receiving no cessation (P = 0.9998). Engagement in consistent exercise results in significantly diminished long-term memory performance relative to the mice receiving no cessation (P = 0.0457). Further, hippocampal elevation of Acvr1c (P = 0.0245) and Bdnf IV (P = 0.0222) persists for longer duration following exercise cessation with weekend warrior exercise.

Conclusions: Our findings highlight ACVR1C as a critical bidirectional regulator of long-term memory. The molecular memory window created by consistent exercise involves elevated Acvr1c, which supports cognitive benefits. This elevation is sustained even in intermittent exercise patterns, suggesting a lasting epigenetic regulation of memory-relevant genes like Acvr1c and Bdnf IV. Thus, strategies targeting ACVR1C and its epigenetic mechanisms could offer therapeutic potential for age-related cognitive decline and conditions like Alzheimer’s Disease, both within and outside the context of exercise.

Keywords: Memory, Hippocampus, exercise, Epigenetics, Alzheimer’s Disease

Disclosure: Nothing to disclose.

Naoki Yamamoto, Hisayuki Osanai, Mark Henkemeyer, Sachie Ogawa, Takashi Kitamura*

The University of Texas Southwestern Medical Center, Dallas, Texas, United States

Background: Cortical modules are thought to mediate mammalian brain function. In the medial entorhinal cortex (MEC), grid cells represent an animal’s location and are organized into distinct modules, called grid modules that ensure unique spatial representation. However, the neurobiological mechanisms of grid modules remain unknown. In layer II of MEC, Calbindin-D28K+ (CalB) cells are arranged in patches on a triangular lattice, while Reelin+ cells are distributed between CalB+ patches and form grid modules.

Methods: We examined the roles of Eph/ephrin bidirectional signaling on CalB patches structure as well as grid modules by utilizing genetic modification combined with cell-type specific in vivo calcium imaging from the layer II of medial entorhinal cortex. We also examine the roles of CalB patches on grid modules as well as spatial navigation.

Results: First, our study revealed that EphrinB2, a mediator of cell-contact-mediated neuronal migration signal, is crucial for the development of CalB+ patches in the MEC (p < 0.01). Second, both loss of EphrinB2 signaling and selective ablation of CalB+ cells in the MEC impaired the discreteness of grid modules (p < 0.01), which caused a large error in spatial representation (p < 0.01). Third, CalB+ cells are synchronized within a patch but not across patches and provide the speed information of animal’s running to grid modules. However, the activity was impaired in EphrinB2 mutant mice. These results suggest that multiple CalB+ patches, mediated by EphrinB2 signaling, differentially provide the self-motion signal to locally connected Reelin+ grid cells to facilitate the discreteness of grid modules. Furthermore, our study extends to the functional implications in spatial navigation. We found that mice lacking CalB+ cells in adult MEC showed impaired spatial navigation, particularly path integration (p < 0.01). Since CalB+ cells provide the speed information of animal’s running to grid modules in the MEC, the behavioral impairment suggests that CalB+ patches act as a pre-configured path integrator, essential for integrating self-motion signals into allocentric spatial memory.

Conclusions: These results suggest that Eph/ephrinB2-mediated CalB+ patches differentially provide their unique activity to local Reelin+ cells that facilitate the discreteness of grid modules, which would contribute to spatial navigation.

Keywords: Spatial Navigation, Medial Entorhinal Cortex, Grid Cell, Hippocampal Function

Disclosure: Nothing to disclose.

Ashley Hayden, Katie Brandel-Ankrapp, Edward Pietryk, Emily Leptich, Paul Merlau, Priyadharshini Vijayakumar, Hsiao-Tuan Chao, Jill Rosenfeld, Rachel Arey*

Baylor College of Medicine, Houston, Texas, United States

Background: Cognitive processes, such as learning and memory, require the precise control of mRNA translation and new protein synthesis. One class of proteins that regulates mRNA translation are RNA binding proteins, the dysfunction of which are becoming increasingly associated with neurological disorders. However, despite their biological significance, the role of many RNA binding proteins in neuronal phenotypes such as cognition remains unknown. Using the model organism C. elegans, we sought to identify conserved RNA binding proteins that exhibit neuronal expression, but remain functionally uncharacterized with regards to nervous system phenotypes, particularly complex behavior. Cross-referencing publicly available C. elegans, mammalian, and human datasets, we found that the C. elegans CEY family, and their mammalian orthologs in the YBX family (specifically YBX1 and YBX3), are abundant throughout the nervous system across species. Mammalian YBX family members are abundant in the hippocampus, and highly expressed in C. elegans neurons known to control associative behaviors. In non-neuronal cell types, YBXs regulate polysome and ribonucleoprotein complex formation, a function conserved with the worm CEY proteins. Based on their expression patterns and established role in RNA and translational regulation, we hypothesize that CEY/YBX proteins are involved in associative learning and memory.

Methods: C. elegans maintenance: Wild-type and mutant animals were maintained under standard laboratory conditions and fed the E. Coli strain OP50 ad libitum. Synchronized populations for behavior assays were generated by standard hypochlorite treatment.

RNAi Treatment: Standard RNAi by feeding was performed to achieve gene knockdown. To knock down genes specifically in adulthood, animals were fed RNAi at the L4 larval stage, after terminal nervous system differentiation. To achieve RNA-knockdown selectively in neurons, experiments were performed in a transgenic C. elegans strain that expresses a double stranded RNA transporter exclusively in neurons.

Behavior Assays: Standard positive olfactory association assays were performed. These assays pair the neutral odorant butanone with food (E. coli) so that animals form a positive butanone association. Learning and memory were assayed as a training-dependent increase in preference for butanone as measured by population chemotaxis assays (~100 animals per assay) to obtain a chemotaxis index. Memory performance was calculated by Performance Index (Chemotaxis_Index(trained) - Chemotaxis_Index(naive/untrained). For all behavioral assays, 10-15 replicates were used, and either one- or two-way ANOVA followed by Bonferroni post-hoc tests were performed.

Protein Alignment: C. elegans and mammalian proteins were aligned with MultiAlin, and domains annotated according to SMART domain prediction and previous literature.

Genetic Database Analysis: Publicly available DECIPHER, gnomAD, and Baylor Genetics databases were mined for variants in human Ybx genes. Variants were manually curated to identify de novo mutations in Ybx genes in patients using the UGSC Genome Browser. De novo mutations were assessed for pathogenicity using CADD, REVEL, and GERP to assign scores. PolyPhen2 was used to predict the tolerability of specific variants.

Results: We find that of the C. elegans CEY family members, CEY-1, is most closely related to mammalian YBXs, and is more closely related to mammalian YBXs than other CEY family members. We find that C. elegans cey-1 reduction- or loss-of-function mutants exhibit deficits in associative learning and memory (p  <  0.05), and that nervous-system specific rescue or CEY-1 restores their ability to learn and remember (p < 0.05) Conversely, overexpression of neuronal CEY-1/YBX enhances associative memory (p  <  0.05), suggesting that it is indeed a memory-promoting molecule.

We next examined whether dysfunction of human YBXs are potentially associated with neurological symptoms in patient populations. Using a combination of publicly available, as well as institute specific human variant datasets, we found that the majority of patients with copy number losses in any YBX have severe neurological symptoms. Interestingly, the most common symptom among patients is intellectual disability, mirroring our behavioral findings in C. elegans. We have also identified single nucleotide variants in YBX3 in individuals with neurological symptoms. We identified one predicted deleterious YBX3 variant of unknown significance, p.Asn127Tyr, in two individuals with neurological symptoms. Introducing this variant into endogenous cey-1 locus caused memory deficits in the worm (p < 0.05). We further generated two humanized worm lines expressing human YBX3 or YBX1 at the cey-1 locus to test evolutionary conservation of YBXs in memory and the potential functional significance of the p.Asn127Tyr variant. Both YBX1/3 can functionally replace cey-1 (p < 0.05), and introduction of p.Asn127Tyr into the humanized YBX3 locus caused memory deficits (p < 0.05).

Conclusions: In summary, we have uncovered a new role for the CEY/YBX RNA-binding proteins in the nervous system as regulators of memory and cognition that is conserved across species. Our studies in C. elegans can inform the molecular underpinnings of any potential YBX -related neurological symptoms in human patients and underscore the importance of neuronal RNA binding proteins in cognition.

Keywords: Memory and Learning, RNA binding protein, Molecular Genetics, Caenorhabditis elegans

Disclosure: Nothing to disclose.

Joshua Taliaferro*, Lorenzo Posani, Julia Greenwald, Sean Lim, Josephine McGowan, Elizabeth Pekarskaya, Clay Lacefield, Stefano Fusi, Christoph Kellendonk

Columbia University, New York State Psychiatric Institute, New York, New York, United States

Background: Working memory is the cognitive capacity for temporarily holding information in mind for processing or use, and supports most of our actions and behaviors. Despite a rich history of WM investigation, the characterization of the neural basis of WM is still emerging. Although activity in the rodent prefrontal cortex (PFC) is necessary for spatial working memory utilization, mnemonic PFC representations of spatial information during the delay are not consistently observed in canonical rodent tasks. We hypothesized that increasing task optionality might reveal mnemonic PFC representations of spatial information.

Methods: Adult mice (n = 9) were trained in a novel spatial working memory task in an automated, 8-arm radial arm maze with a large, open center. Cellular-resolution calcium activity in PFC neurons was then recorded during task performance.

Results: Delay-phase activity of PFC neurons indeed contained mnemonic representations of spatial information, at both single-cell and population levels, in this task. These mnemonic representations were retrospective rather than prospective, and—surprisingly—were more evident on trials where the animals used a nonmatch-to-sample behavioral strategy (“error” trials) than when they used a match-to-sample behavioral strategy (“correct” trials).

Conclusions: In the setting of a freely-moving rodent spatial working memory task, PFC mnemonic representations preferentially support executive function-dependent deviation from an entrained behavioral strategy, rather than retrospective encoding more generally.

Keywords: working memory, prefrontal cortex, Behavior, neural populations, calcium imaging

Disclosure: Nothing to disclose.

Ryan Lingg, Dalton Hinz, Shane Johnson, Timothy Skog, Manuela Lizarazu, Sara Romig-Martin, Ryan LaLumiere, Nandakumar Narayanan, Jason Radley*

Dept. Psychological and Brain Sciences, Iowa Neuroscience Institute, University of Iowa, Iowa City, Iowa, United States

Background: Generalizing aversive memories helps organisms avoid danger, whereas discriminating between dissimilar situations promotes opportunistic behaviors. The amygdala and hippocampus are important for encoding and retrieval of threat, contextual memories, and discrimination. However, less is known regarding the mechanisms that shape memory specificity in the aftermath of an aversive experience.

Methods: In the present study, we characterized neural processing in the prefrontal-to-bed nuclei of the stria terminalis (BST) pathway during the immediate aftermath of an aversive event. Using a combination of anatomical, endocrinological, and functional approaches, we evaluated the involvement of the prefrontal–BST pathway in aversive memory consolidation using the inhibitory avoidance discrimination learning task in male rats developed by Roozendaal and colleagues. Ancillary studies assessed whether the involvement of this pathway was directly mnemonic rather than associated with its influence over stress hormones.

Results: Our data revealed activity changes along the prefrontal–BST pathway after inhibitory avoidance discrimination learning that distinguish safe from aversive contexts, as well as a bi-directional role of this pathway in modulating the consolidation of memory specificity while leaving memory strength unchanged. Further studies showed that the necessity of prefrontal–BST pathway modulation of fear generalization occurs independently from increases in stress hormonal activation.

Conclusions: These findings illustrate a novel contribution of the prefrontal–BST pathway in long-term behavioral expression of memory specificity and may provide insight into how circuit dysregulation may underlie fear over-generalization in stress-related psychiatric illnesses.

Keywords: Memory Consolidation and Extinction, Medial Prefrontal Cortex, bed nucleus of the stria terminalis, discrimination learning, stress hormones

Disclosure: Nothing to disclose.

Maryam Hasantash*, Yifei Li, Christoph Anacker

Columbia University, New York, New York, United States

Background: Cognitive flexibility is an executive function vital for adapting thinking and behavior to changing environments. Deficits in cognitive flexibility can lead to perseveration, rumination, and habitual behavior, which can worsen emotional distress and contribute to psychiatric disorders such as depression, anxiety, PTSD, and OCD. Understanding the neurobiological mechanisms underlying cognitive flexibility is thus crucial for improving treatment strategies. The hippocampus is important for learning, memory, and stress regulation, and consists of the dorsal region involved in spatial navigation and the ventral region associated with stress and goal-directed behavior. Human studies link hippocampal gray matter volume with perseveration in cognitive flexibility, and rodent studies highlight its role in response inhibition and behavioral flexibility. However, how the dorsal and ventral hippocampus each contribute to different aspects of cognitive flexibility is not known. Here, we used chemogenetic inhibition of the dorsal hippocampus (dCA1) and ventral hippocampus (vCA1) regions during reversal learning and set-shifting tasks to determine their differential contributions to cognitive flexibility. We also performed simultaneous in vivo calcium (Ca2 + ) imaging of neuronal activity in the dCA1 and vCA1 regions of freely moving mice to explore their distinct neural response profiles during different phases of learning before and after a rule change.

Methods: To investigate how vCA1 and dCA1 contribute to cognitive flexibility, we bilaterally injected separate cohorts of mice with inhibitory hM4Di DREADD receptors (AAV9.hSyn.hM4D(Gi).mCherry) in either vCA1 or dCA1. Control mice received AAV9.hSyn.mCherry virus. We then exposed mice to two complementary tasks of cognitive flexibility 4 weeks after virus injection. In the Y-Maze, food-restricted mice were trained to find a reward in one of two arms on learning days 1-6. The reward location was then switched to the other arm and mice had to find the new reward location during reversal learning days 7-12. Mice were also tested in a 2-choice digging task. They initially learned to associate that one of two bowls was marked with cinnamon scent and contained reward, while a garlic-scented bowl did not contain reward. On the set-shifting day, one of the two digging media texture instead of scent predicted reward. A trials-to-criterion measure (8/10 correct trials) was used to assess perseveration. Clozapine-N-oxide (CNO, 5mg/kg, i.p.) was administered 30 minutes before the first trial on each testing day. Two-way repeated measures ANOVA was used to assess inhibition effects in the Y-Maze. Trails-to-criterion in the Digging Task between control and inhibition groups were analyzed using Two-tailed student’s t-test.

We also investigated the neural signals underlying vCA1 and dCA1’s involvement in reversal learning using in vivo Ca2+ imaging. Here, we injected GCaMP6f into vCA1 and dCA1 and used fiber photometry to image vCA1 and dCA1 activity changes in the Y-Maze and in the Digging Task. Two-way ANOVA was used to assess activity changes in dCA1 vs vCA1 in the Y-Maze. Unpaired student’s t-test was used to compare neural responses in the Digging Task on learning and reversal learning days.

Results: Exp. 1: Inhibiting vCA1 during rule acquisition in the Y Maze did not affect learning (days 1-6). However, when vCA1 was inhibited during the reversal phase (days 7-12), mice made more perseverative errors compared to controls without vCA1 inhibition (***p = 0.0003, n = 12). Similarly, vCA1 inhibition in the Digging Task did not affect initial rule acquisition (p = 0.3448) but increased the number of trials needed for mice to learn the rule reversal (**p = 0.005) and during the set shifting (**p = 0.0024).

In contrast to vCA1, inhibiting dCA1 during rule acquisition in the Y Maze reduced learning accuracy compared to control mice (**p = 0.0017, n = 7-9) but did not impact reversal learning (p = 0.26). In the Digging Task, dCA1 inhibition had no effect on trials-to-criterion across tasks (acquisition: p = 0.712, reversal: p = 0.618, set shifting: p = 0.6103).

Exp. 2: To understand the neural activity changes underlying vCA1 and dCA1’s functional involvement in learning vs reversal learning, we recorded in vivo Ca2+ activity signals in both regions from the same mice. Ca2+ activity in dCA1 was increased during the initial learning phase in the Y-Maze when mice made either correct (***p = 0.0005) or incorrect arm choices (**p = 0.0016). No changes in dCA1 activity were observed during correct or incorrect arm choices in the reversal phase. Ca2+ activity in vCA1 was unaltered during the initial learning phase, but increased specifically when mice entered the correct arm of the Y-Maze during reversal learning (on reversal day 1: ***p = 0.0002, n = 7).

Similar to our findings in the Y-Maze, Ca2+ activity in vCA1 was significantly increased in the Digging Task on the reversal day (**p = 0.0070, n = 6) and on the set-shifting day (**p = 0.0026) when mice investigated the correct bowl. There were no differences between dCA1 and vCA1 activity during initial rule acquisition.

Conclusions: Our results highlight a novel division of labor within the hippocampus, where the dorsal subregion is crucial for the acquisition and storage of new information, and the ventral subregion is essential for the flexible adjustment of behavioral responses when action-outcome contingencies change.

Keywords: cognitive flexibility, reversal learning, Dorsal Hippocampus, ventral hippocampus, in vivo calcium imaging

Disclosure: Nothing to disclose.

Jordan Mak, Lisa Gazdzinski, Xinyi Lin, John Sled, Brian Nieman, Anne Wheeler*

Hospital for Sick Children, Toronto, Canada

Background: Forgetting occurs when patterns of neural activity present at memory encoding cannot be reliably reactivated. The precise timing of activity in neural circuits is regulated by white matter properties such as the distribution and amount of myelin surrounding axons. Traditionally viewed as a stable structure, myelin is now known to undergo dynamic remodeling throughout life. Living in enriched environments has been shown to promote white matter plasticity, which may disrupt previously established activity patterns associated with specific memories and facilitate forgetting.

Methods: The impact of white matter plasticity on contextual fear memory was assessed by housing mice in enriched cages after a single conditioning session where the mice received three 0.75mA foot shocks. Enriched cages consisted of larger, double-decker cages containing a running wheel, and a reconfigurable maze rearranged weekly. Food and water were at opposite ends of the maze to force mice to relearn how to maneuver through each maze arrangement. After living in an enriched environment for four weeks, mice were returned to the conditioning context and freezing was measured as an index of memory. Ex vivo mouse brains were scanned on a 7T Agilent MRI system with a diffusion-weighted acquisition. Microstructural properties of white matter tracts defined by a segmented atlas were assessed with fractional anisotropy (FA), a measure of white matter integrity derived from diffusion tensor imaging. Effects of enrichment on freezing, white matter FA, and associations between freezing and FA were assessed with linear mixed effects models.

Results: Mice housed in enriched environments demonstrated forgetting as measured by reduced time spent freezing during the fear conditioning test compared to mice housed in standard cages (β = -51%, p = 2x10-6). Enriched housing was associated with altered white matter microstructure, where FA was increased in the corpus callosum, fimbria, and anterior commissure compared to mice housed in standard cages (10% FDR). In mice housed in enriched environments, increased FA was associated with reduced freezing in the corpus callosum and fimbria (p < 0.05), suggesting that in these regions, white matter plasticity induced by enrichment contributes to forgetting.

Conclusions: This study identifies a new mechanism that contributes to the forgetting of a fear memory in mice. The results suggest that plasticity occurring in white matter tracts important for contextual fear memory contributes to enrichment-induced forgetting. Knowledge of brain mechanisms that promote forgetting of traumatic memories may identify treatment targets and biomarkers for post-traumatic stress disorder.

Keywords: fear memory, forgetting, white matter, enrichment, diffusion weighted MRI

Disclosure: Nothing to disclose.

Khairunisa Mohamad Ibrahim*, Dominika Burek, Olayinka Idowu, Azra Zec, William Post, Alice Zheng, Nicolas Massaly, Jose Moron-Concepcion

Washington University in St. Louis, St. Louis, Missouri, United States

Background: Drug of abuse such as opioids targets learning and memory circuitry to form long-lasting associations between the drug and the environments where it’s consumed. This leads to cravings and relapse into addiction. We have previously shown that administering morphine in specific contexts led to an increased basal synaptic transmission but disrupted long-term potentiation (LTP) in the dorsal hippocampus (dHPC) (Portugal et al., 2014; Xia et al., 2011). This impaired hippocampal LTP may contribute to relapse by making it difficult to unlearn associations that drive cravings. We found that morphine-induced context-dependent sensitization not only had impaired LTP but also an enhanced SK2 channel-mediated negative feedback on NMDA receptor (Fakira et al., 2014). These calcium-gated potassium channels (SK2) are crucial for contextual learning, regulating neuronal excitability and synaptic adaptability (Hammond et al., 2006). Yet, their role in modulating morphine-seeking behavior remains unexplored.

Methods: To assess this, we use a drug-paired memory-based paradigm, conditioned place preference (CPP) and intracranially infused either SK2 channel blocker (Ler-Dab7) or activator (NS309) into the dHPC of C57/BL6 male mice (n = 5-12). Prior to conditioning, baseline preference for either compartment was determined in a 30-minute pretest. Mice were conditioned for a total of eight sessions, two per day, separated by four hours. During conditioning, mice are injected with either morphine (15 mg/kg) or saline and confined to one compartment for 45 minutes. After conditioning, preference for either compartment was determined in a 30-minute post-test. All statistical analyses were performed using Prism software using either two-tailed unpaired Student’s t-test, one-sample t-test, two-way or three-way repeated measure ANOVAs followed by post-hoc tests to compare individual groups.

Results: We found no change in SK2 gene expression or protein levels in the hippocampus post-morphine CPP (MorCPP) test with polymerase chain reaction and western blots, respectively. However, blocking the SK2 channel with Lei-Dab7 before post-conditioning test attenuates the retrieval of MorCPP memory in a dose-dependent manner. Interestingly, blocking SK2 channel before morphine conditioning trials strengthens the preference for the morphine-paired compartment. Preliminary results with SK2 channel activator (NS309) suggest that infusion either before post-conditioning or post-test blocks preference for the morphine-paired compartment.

Conclusions: These findings suggest that the SK2 channels play a differential role during the formation and retrieval of drug-context memory.

Keywords: Memory and Learning, Drug Relapse, conditioned place preference, Dorsal Hippocampus, morphine

Disclosure: Nothing to disclose.

Maria Diehl*, Shannon Ruble, Cassandra Kramer, Lexe West

Kansas State University, Manhattan, Kansas, United States

Background: To maintain survival, individuals must effectively assess danger in the environment to remain safe from harm. When observing others responding to a potential threat, a wealth of information can be gained from socially transmitted cues without directly experiencing harm. Preclinical rodent studies have revealed key brain structures necessary for observational fear learning, but the behavioral and neural mechanisms promoting observational avoidance learning remain understudied. Here, we modified the platform-mediated active avoidance (PMA) task, a decision-based, clinically-relevant rodent model of avoidance, to allow rats to learn PMA by witnessing a demonstrator avoid a tone-signaled footshock by stepping onto a safe platform at the cost of lever-pressing for a sucrose reward.

Methods: Male and female Sprague Dawley rats were bred in-house and same-sex housed in pairs or trios in standard cages. Demonstrator rats underwent PMA for 10 days, in which 9 presentations of a 30 sec pure tone (4 kHz, 75 dB) co-terminated with a 2 sec footshock (0.4 mA) while lever-pressing for sucrose pellets on a VI-30 schedule of reinforcement and could avoid shock by stepping onto a platform, located opposite of the lever and food dish. Observer rats witnessed same-sex Demonstrators undergo PMA during an early (Day 1) or late (Day 10) training session. Observers viewed Demonstrators through a perforated acrylic barrier and could hear the tone, but did not have their own lever, food dish, or platform. 24 hr after the Observation Session, Observers were placed in the same side of the Demonstrator and underwent a Direct PMA Session with the same contingencies of the Demonstrator (9 tone-shock trials, VI-30 schedule of reinforcement, etc.). A subset of observers experienced shocked prior to observing one of the timepoints of PMA training (n = 10-12 per group) and another subset were naïve to shock (n = 18-20 per group). We used cFos immunohistochemistry to quantify neural activation during the direct PMA session in shock-exposed Observer groups. cFos density was measured in the anterior cingulate cortex (ACC), a region critical for observational learning. Time on platform, number of shocks avoided, and freezing were quantified and correlated with cFos density in Observers that witnessed early and late time points of PMA training. A Generalized Linear Mixed Model was used to assess the percentage of time spent on the platform and freezing during the tone during the direct PMA test session. Trial (9 tone presentations per Session), Session (Day 1 or Day 10), and prior shock experience were used as fixed effects, and individual rat identity and sex were used as a random effects. For cFos analysis, we used a one-way ANOVA to compare cFos density in ACC across the different Observer groups.

Results: We validated a new rat model of observational active avoidance. Day 1 Observers with prior shock experience spent significantly more time on the platform (p < 0.05) and avoided significantly more shocks (p < 0.05) compared to Day 1 Observers without prior shock experience. In contrast, Day 10 Observers, regardless of shock experience, showed similar levels of time on platform and number of shocks avoided. Day 1 Observers, regardless of shock experience, exhibited greater tone-induced freezing, compared to Day 10 Observers (p < 0.05). We next asked whether avoidance was correlated between the Observer and their Demonstrator. Day 1 Observers with shock experience showed a positive correlation with their Demonstrator’s time on platform, whereas Day 1 Observers with no shock experience showed a negative correlation with their Demonstrator’s time on platform. Interestingly, we found the opposite effect in Day 10 Observers: those with shock experience showed a positive correlation with their Demonstrator’s time on platform, whereas those without shock experience showed a negative correlation with their Demonstrator’s time on platform. There were no significant sex differences for avoidance or freezing during the tone for Day 1 Observers (p = 0.999, p = 0.752, respectively), or Day 10 Observers (p = 0.981, p = 0.999, respectively). When assessing neural activation, cFos density in the ACC was greater in D10 Observers compared to D1 and D2-9 Observers (p = 0.009), suggesting that ACC is recruited during observation of late (but not early) phases of PMA training when both the Observer and Demonstrator have positively correlated avoidance behavior.

Conclusions: Observers witnessing early stages (Day 1) of PMA utilize information related to fear-based strategies such as freezing and escape responses to the shock that depend on having prior shock experience. During early Observational PMA, prior shock experience was associated with following the Demonstrator’s behavior, suggesting that the Observer might better learn the PMA task contingencies when they have prior knowledge about the aversive shock. In contrast, Observers witnessing late stages (Day 10) of PMA utilize information related to avoidance-based strategies such as platform mounting that do not specifically require knowledge of shock information. Current results suggest ACC is recruited when Observers witness late stages of PMA training, and additional analyses are underway to assess neural activation in established PMA circuits, including the prelimbic cortex, basolateral amygdala, and ventral striatum. These findings will advance our understanding of the prefrontal circuits necessary to guide and regulate observational learning of active avoidance.

Keywords: fear, cFos, Social Behavior, prefrontal-amygdala-connectivity, sex differences

Disclosure: Nothing to disclose.

Virginia Rodriguez, William Smith-Peters, Sachin Patel, Luis Rosas-Vidal*

Northwestern University, Chicago, Illinois, United States

Background: More than half of the US population has lived through at least one traumatic experience in their lifetime. While many will be able to overcome trauma, a subset of individuals will develop trauma-related and anxiety disorders. Research has shown that traumatic experiences disrupt circuits required to appropriately express fear (i.e., threat) memories and anxiety and lead to erroneous maladaptive behaviors in otherwise safe contexts. The bed nucleus of stria terminalis (BNST) is one of the major structures involved in mediating fear memory, apprehension, and anxiety. Retrograde tracing experiments have shown that the BNST receives inputs from both the basomedial nucleus of the amygdala (BMA), which has been implicated as an inhibitory structure for anxiety and fear responses, and the infralimbic subdivision of medial prefrontal cortex (IL), which has been demonstrated to reduce fear. Furthermore, the IL to BNST projection has been shown to suppress fear of uncertain threats. However, the interaction of these circuits in innate and learned defensive behaviors remains poorly understood.

Methods: Male and female mice were injected with AAVs encoding ChR2 under control of the CaMKII promoter in the BMA and optic fibers were implanted above adBNST. Following 6 weeks of recovery and time for viral expression, mice were exposed to elevated plus maze, fear conditioning, open field, and predator odor assays while blue light was delivered to BMA-BNST terminals through the optic fiber at 10 Hz with 10 ms pulse width and and 10 mW of power.

Results: Optogenetic activation of BMA-adBNST during the elevated plus-maze (EPM), fear conditioning (FC), and predator odor exposure increases exploration of open arms, reduces conditioned freezing, and innate freezing respectively (p < 0.05). Interestingly, Optogenetic activation of IL-adBNST during the elevated plus-maze (EPM), fear conditioning (FC), and predator odor exposure decreases exploration of open arms while reducing conditioned and innate freezing (p < 0.05).

Conclusions: Our results suggesting that BMA-adBNST suppress both learned and innate defensive behaviors are in line with the current literature showing that BMA itself reduces fear and anxiety behaviors. Surprisingly, our IL-BNST results reveal a dissociation were this projection decreases learned and innate fear-related behaviors while increasing innate avoidance behavior.

Keywords: Fear regulation, BNST, circuit optogenetics

Disclosure: Nothing to disclose.

Vincent Costa*, Kathryn Rothenhoefer, McKenna Stocker

Emory University, Atlanta, Georgia, United States

Background: Managing the explore-exploit tradeoff requires decision-makers to explore options with unknown consequences rather than exploit options with known outcomes. While explore-exploit decision-making is typically studied in the context of maximizing gains, it is understudied in the context of minimizing losses. Projections from the amygdala to the nucleus accumbens invigorate conditioned responses to acquire rewards but also enhance conditioned avoidance responses. These findings suggest that amygdala inputs to the nucleus accumbens play a general role in exploiting previously learned information. However, the computational role of this pathway in managing explore-exploit tradeoffs has not been examined in computationally tractable decision-making tasks. Moreover, no direct, within-subject comparisons have been made to determine whether this pathway is preferentially involved in exploiting information about gains versus losses.

Methods: We developed a three-arm bandit task to induce explore-exploit tradeoffs by introducing novel choice options associated with the gain or loss of virtual tokens. These virtual tokens served as secondary reinforcers, which were exchanged for primary juice rewards based on task performance. Using tokens as secondary reinforcers allowed us to include aversive conditions where macaques could lose tokens, providing a means to examine exploratory decision-making in different valence contexts (n = 2 males). We employed pathway-specific chemogenetics to evaluate the effects of inhibiting amygdala inputs to the nucleus accumbens on exploratory decision-making. A retrograde viral vector (retro.AAV2-CRE-eGFP) was injected into the nucleus accumbens to transfect the axon terminals of amygdala neurons projecting to this region and to express CRE in the cell bodies. Subsequently, a second viral vector (AAV2-hM4Di-mCherry) was injected to express an inhibitory Cre-dependent chemogenetic receptor in the same amygdala neurons. We then assessed the impact of inhibiting this glutamatergic amygdala pathway by administering the chemogenetic actuator deschloroclozapine (DCZ; n = 10 sessions per dose) at two different doses (0.1 mg/kg or 0.3 mg/kg). In additional sessions we left this pathway unaffected through the administration of saline (n = 10 sessions). We analyzed the monkeys’ reinforcement learning and novelty-seeking behaviors using hierarchical linear models, with individual sessions within each monkey specified as random effects and drug dose as a fixed effect.

Results: We found that inhibiting amygdala inputs to the nucleus accumbens impaired the monkeys’ ability to learn and discriminate between cues associated with different sizes of gains in their token endowment. This effect was dose-dependent (t = 2.7, p < 0.01). In contrast, inhibiting this amygdala pathway had no effect on the monkeys’ ability to learn to avoid cues associated with different sizes of losses in their token endowment (p = 0.303). We also found that chemogenetic inhibition of amygdala inputs to the nucleus accumbens reduced the monkeys’ exploration of novel choice options in favor of exploiting familiar options they had already learned. This occurred regardless of whether the novel cue was introduced in an appetitive (familiar cues guaranteed to lead to a gain in tokens) or aversive (familiar cues guaranteed to lead to a loss of tokens) decision context. Moreover, the overall reduction in novelty exploration was dose-dependent (F = 4.74, p = 0.014).

Conclusions: These results indicate that, in nonhuman primates, glutamatergic amygdala projections to the nucleus accumbens form an inherently appetitive neural circuit that mediates the exploration of novel opportunities when the value of novelty seeking is uncertain.

Keywords: Amygdala, Nucleus Accumbens, Nonhuman Primates, Decision Making, Computational Reinforcement Learning Model

Disclosure: Nothing to disclose.

Lucas Trambaiolli*, Suzanne Haber

McLean Hospital, Harvard Medical School, Belmont, Massachusetts, United States

Background: The ventral surface of the frontal cortex (VSFC) and the ventral striatum (VS) compose a small network crucial for reward processing and decision-making. These two structures are interconnected by “parallel patches” (Haber et al. 1995 J. Neurosci.): the medial VSFC (mVSFC, areas 14M and 14O) projects to the medial VS (mVS, ventromedial caudate), the central VSFC (cVSFC, 11 and 13) to the central VS (cVS, shell and core of the nucleus accumbens), and the lateral VSFC (lVSFC, 12L and 12O) to the lateral VS (lVS, ventrolateral putamen). These structures also receive strong connections from limbic and cognitive networks (Carmichael and Price 1995 J. Comp. Neurol., Averbeck et al. 2014 J. Neurosci.). Herein, we hypothesize that these connections are organized in two ways: i) inputs from limbic reward regions to the mediolateral subdivisions of the VSFC and VS follow similar topographies; ii) the VSFC and VS receive different inputs from cognitive and premotor regions, with the VSFC having a higher integrative role compared to the VS.

Methods: Using dark-field microscopy, we analyzed anterograde data from injections in the limbic reward system, including the anterior cingulate cortex (ACC), posterior cingulate cortex (PCC), hippocampus, and amygdala, to map the spatial distribution of projections to the VSFC and VS. Four injections were in areas 24 and 32 of the ACC, four in areas 23 and 31 of the PCC, one in the hippocampus (including the subiculum and parahippocampal gyrus), and two in the basolateral nucleus of the amygdala. We also used retrograde injections in different VS and VSFC sub-locations to evaluate the quantitative strength of afferent connections. In the VS, we placed one injection in the mVS, two in the cVS, and one in the lVS. In the VSFC, we had two injections in the mVSFC, three in the cVSFC, and four in the lVSFC. Under bright field microscopy, we quantified labeled cells in the frontal cortex for each retrograde injection. We then computed the strength of inputs as the percentage of cells from each frontal region divided by the total number of labeled cells.

Results: Anterograde injections showed that limbic inputs to the VSFC and VS followed the same spatial distribution. The ACC and BLA projected to all the VSFC and VS subdivisions. The PCC projected mainly to the mVSFC and mVS, and the subiculum projections extended through central and medial subdivisions of the VS and VSFC. Retrograde injections in the VS confirmed the “parallel patches” of projections from the VSFC to the VS (i.e., mVSFC projecting to the mVS, cVSFC to the cVS, and lVSFC to the lVS). Moreover, connectivity strengths confirmed the general distribution of inputs from the ACC to both structures. Differences in inputs to the VSFC and VS are mainly related to dorsal frontal cortical inputs, with cVSFC and lVSFC receiving stronger projections from the dorsal prefrontal cortex (dPFC, areas 9M, 9L, and 46D) than the VS counterpart.

Conclusions: Both VSFC and VS receive inputs from the limbic reward system, with their medial portions being connected with all the areas included in our analysis (ACC, PCC, BLA, and Subiculum). Additionally, the VSFC receives stronger inputs from the dPFC. The dPFC is a region highly involved in cognitive control, whose projections may support decision-making related to reward information in the VSFC. These reward-informed decisions may be rapidly passed to the VS through the VSFC-VS direct connections. Then, the VS may use its inputs to plan and trigger adequate subsequent actions. This circuit delineation is relevant for new treatments in psychiatry. For example, VSFC and VS are commonly identified as biomarkers for obsessive-compulsive disorder (OCD), with the latter being a promising target for deep-brain stimulation intervention (Haber et al. 2020 Biol. Psychiatry). While this may be an excellent choice to improve behavioral symptoms, the stimulation of the VSFC may be a better choice to treat cognitive inflexibility in OCD.

Keywords: limbic system, prefrontal cortex, Ventral Striatum

Disclosure: Nothing to disclose.

Xiyu Zhu*, Lara Hagopian, Kira Wallquist, Vikaas Sohal

UCSF, San Francisco, California, United States

Background: Parvalbumin-expressing (PV) inhibitory neurons in the medial prefrontal cortex (PFC) are essential for cognitive flexibility. Our recent work in mice identified an inhibitory long-range callosal PV (ccPV) projection that can persistently regulate cognitive flexibility in rule-shift (RS) learning task. In this study, we aimed to further characterize PFC ccPV synapses and to identify circuit loci with maladaptive synaptic plasticity associated with persistent cognitive inflexibility. In addition, we conducted proof-of-concept experiments in Syngap1 heterozygous mice demonstrating the therapeutic potential of gamma frequency stimulation of this circuit.

Methods: In PV-Cre mice we injected AAVs into one PFC hemisphere to drive Cre-dependent axonal expression of ChR2 and eNpHR contralaterally. Moreover, we injected AAVrg-tdt into the contralateral medial dorsal thalamus (MD) to retrogradely label PFC-MD projectors, which are known to preferentially receive callosal PV inputs. Similarly, PFC-striatum and Cortico-cortical PFC neurons were labelled in separate sets of animals. Next, we implanted an optic fiber in the contralateral PFC and waited a minimum of four weeks for viral expression. We performed rule-shifting tasks during optogenetic inhibition and/or 40Hz gamma-frequency stimulation of ccPV terminals, followed by whole-cell patch-clamp recordings of the contralateral PFC pyramidal neurons 24-72 hours after the test. Using similar labelling strategies, we expressed ChR2 in the callosal projections of fast spiking neurons or specifically PV neurons and tested the therapeutic effect of 40Hz stimulation in Syngap1 heterozygous (Syngap1 + /- or Syngap1 + /-:PV-cre) mice using behaviors and slice electrophysiology.

Results: Our preliminary data showed that ccPV synapses were differentially responsive in learning different rules, displaying differential callosal inhibition after the tasks. Additionally, optogenetic inhibition of ccPV fibers caused RS learning deficits and increased perseveration, confirming previous findings. Syngap1 heterozygous mice also displayed marked RS learning deficits. Prior ccPV terminal inhibition during the RS task persistently decreased optically evoked inhibitory postsynaptic current (oIPSC) in the ccPV synapses onto both MD projecting neurons, without altering other synaptic strengths or action potential firing. Moreover, we observed decreased synaptic connectivity in non-MD projecting pyramidal cell-types, and paired-pulse ratio recordings indicated that inhibiting ccPV axons disrupted synaptic plasticity. Syngap1 heterozygous mice also exhibited similar behavioral and electrophysiological effects described above. Both the impairments induced by optogenetic inhibition or Syngap1 heterozygosity could be rescued by optogenetically stimulating the ccPV axon terminals at 40Hz.

Conclusions: Our preliminary dataset suggests that prior optogenetic inhibition of PFC ccPV is sufficient to lead to persistent alterations in ccPV synaptic functions, supporting a role for long-term plasticity. Similar pathophysiological changes were observed in Syngap1 mutants, which can be rescued by gamma frequency stimulation of the same projection. Together, current findings suggest PFC ccPV synapse might be a vulnerable circuit locus implicated in neurodevelopmental disorders, including SYNGAP1-related disorders and perhaps schizophrenia. Future work will determine whether callosal PV maladaptive synaptic plasticity represents a conserved circuit mechanism affected by risk genes associated with schizophrenia and autism spectrum disorders.

Keywords: GABAergic interneurons, Medial Prefrontal Cortex, Neurodevelopmental Disorders

Disclosure: Nothing to disclose.

Zahra Farahbakhsh*, Zev Jarrett, Jose Zepeda, Lindsay Cain, Brad Grueter, Cody Siciliano

Vanderbilt University, Nashville, Tennessee, United States

Background: Descending glutamatergic fibers connecting the medial prefrontal cortex (mPFC) to the dorsal periaqueductal gray (dPAG) bidirectionally encode stimulus valence and have been implicated in compulsive alcohol drinking in both preclinical and clinical models. For example, activity patterns in mPFC-dPAG neurons during initiation of alcohol use predict later development of punishment-resistant drinking in mice. Likewise, in clinical populations network-level connectivity between PFC and dPAG confers greater risk of developing problematic drinking behaviors. While there is strong evidence that the mPFC-dPAG projection is critical for a range of motivated behaviors, very little is known as to the physiological properties of these neurons and their downstream connectivity in the dPAG. Paradoxically, the mPFC-dPAG projection population has been identified as exclusively glutamatergic, but previous work demonstrated that stimulation of mPFC inputs inhibited activity in the dPAG. Here, we sought to investigate the mechanisms by which activity of mPFC projections modulate activity in the dPAG to control of approach and avoidance behaviors. Based on previous work suggesting co-release of a neuromodulator may be involved in mPFC-dPAG signaling and given the high levels of opioid receptor expression in the dPAG, we hypothesized involvement of an endogenous opioid system.

Methods: All experiments were performed in male and female adult CL57BL/6J mice (n = 2 – 5 per experiment). To determine the opioid receptor system most likely involved in this circuit, we datamined single cell RNA-sequencing of projection-identified mPFC neurons whereby multiple anatomically-defined populations had been labeled and isolated prior to sequencing (accession number: GSE161936). For whole-cell patch-clamp electrophysiology experiments, recordings were conducted in dorsal and lateral columns of the dPAG in acute ex vivo slices. Spontaneous excitatory post synaptic currents (sEPSCs) were recorded at baseline and after bath application of 1 µM U50,488, a selective kappa opioid receptor agonist. To interrogate mPFC inputs, chanelrhodopsin-2 was expressed in mPFC neurons and presynaptic terminals were photostimulated in slices before and after bath application of a norBNI, a kappa-opioid receptor antagonist. To examine in vivo kappa opioid receptor activity, we expressed the kappa opioid receptor biosensor kLight1.3 in dPAG and monitored bulk fluorescence activity via fiber photometry during a range of motivated behaviors and sensory testing, including exploratory behavior in novel and familiar environments, auditory tuning curves, sucrose consumption, and sensory reinforcement. For all pairwise comparisons between two conditions or groups, we utilized t-tests and comparisons across three or more variables were made using one- or two-way ANOVAs. All tests were two-sided and alpha was set to 0.05.

Results: Single cell sequencing data analysis revealed that mPFC-dPAG projections neurons were enriched in prodynorphin mRNA, and expressed higher levels than any of the other projection populations tested (Kruskal-Wallis ANOVA, statistic = 109, p < 0.0001), including mPFC projections to the contralateral PFC (Z94, 440 = 7.88, padj < 0.0001), the nucleus accumbens (Z94, 93 = 4.67, padj < 0.0001), the amygdala (Z94, 290 = 2.63, padj < 0.043), and the dorsal striatum (Z94, 129 = 5.21, padj < 0.0001). Of mPFC-dPAG cells, 33% expressed detectable levels of prodynorphin, a greater proportion than those expressing precursors to the other endogenous opioid ligands, proenkephalin (χ2 = 6.383, p = 0.0115) and proopiomelanocortin (χ2 = 7.37, p = 0.007). No mPFC-dPAG cells expressed OPRK1 mRNA, the kappa opioid receptor (KOR) gene, suggesting that any potential release of dynorphin from mPFC terminals in the dPAG would not have autoregulatory actions and instead is likely to signal through heteroreceptor systems. In support of the hypothesis that mPFC inputs modulate dPAG signaling through dynorphin release, we also found that KOR activation reduced the frequency of sEPSCs (paired t-test, t4 = 3.39, p = 0.028) with no effect on amplitude (paired t-test, t4 = 1.63, p = 0.18) or excitability (two-way repeated measures ANOVA, current: F(1.652, 13.21) = 16.65, p = 0.0004, U50,488: F(1, 8) = 0, p = 0.99) of dPAG neurons, which is congruent with the effect of photostimulation of mPFC terminals in the dPAG. In vivo kLight1.3 recordings showed preliminary evidence for novelty-induced increases in spontaneous event frequency, putatively reflective of dynorphin release events, which decreased in frequency but increased in amplitude over the course of habituation.

Conclusions: The combination of prodynorphin enrichment in mPFC-dPAG projections and the ability of KOR activation to replicate effects of mPFC input stimulation in the dPAG support our hypothesis that neurons from the mPFC regulate dPAG activity through release of dynorphin and subsequent signaling through kappa opioid heteroreceptors, which are expressed throughout the PAG on post-synaptic cells and axonal inputs. Fiber photometry experiments implicate novelty encoding by dynorphin release events in the dPAG and the close relationship between novelty, salience, and approach/avoidance may modulate decision-making processes across many contexts. Together, this work implicates dynorphin signaling from mPFC-dPAG neurons as a potential mechanism by which the circuit exerts control over motivated behaviors.

Keywords: kappa opioid receptor, periaqueductal grey (PAG), Medial Prefrontal Cortex (mPFC)

Disclosure: Nothing to disclose.

Kenneth Amaya*, Yingchu He, Jamie Maguire

Tufts University School of Medicine, Boston, Massachusetts, United States

Background: The binary assignment of positive or negative valence to cues, environments, and outcomes, termed valence processing, is crucial for behavioral approach or avoidance, and thus survival. A critical brain region for valence processing is the basolateral amygdala (BLA), as both appetitive and aversive learning are dependent on BLA function. Recent advances have highlighted the role of interneurons in mediating BLA oscillatory states associated with both positive and negative valence processing, such as fear expression and reward seeking. Despite their documented ability to govern BLA states and relevant behaviors, the involvement of interneurons in ensemble recruitment and information routing to distinct circuits remains unknown.

Methods: To address this, we conducted a series of experiments aimed at interrogating the ability of BLA interneurons to recruit distinct valenced neuronal ensembles in the BLA and downstream circuits (BLA to bed nucleus of the stria terminalis, BNST; BLA to nucleus accumbens, NAc). We first demonstrated that fear and safety engage distinct BLA projection populations using male and female C57BL6J mice that received retro-mCherry into either BNST (N = 8) or NAc (N = 9). After fear conditioning or extinction, tissue was collected for cFos quantification. Next, we assessed the impact of BLA interneuron-driven oscillatory state entrainment on BLA neuron activation (cFos) and projection population engagement (using a retrograde virus label). Based on prior literature, interneurons were stimulated using channelrhodopsin (Dlx-promoted) at either 4 Hz (N = 12), 8 Hz (N = 12), or not at all (control, N = 12). We conclude by assessing brain-wide network engagement after BLA interneuron stimulation via whole brain cFos expression (N = 8 per frequency). Statistical analyses for these tests include linear mixed models, generalized linear mixed models, t-tests, and Wilcoxon rank-sum tests, where appropriate. Both sexes were included in these studies.

Results: We observe that our manipulations (behavioral, optogenetic) alter valenced network engagement. Behaviorally, we observe selective engagement of BLA to BNST projection neurons compared to BLA to NAc projection engagement after fear conditioning (p < 0.001). We also observe selective engagement of BLA to NAc projection neurons after fear extinction, though we are still processing an additional cohort so statistical analyses have yet to be completed here. Optogenetically, we entrained the BLA to specific oscillatory states and report that distinct rhythmic interneuron stimulations engaged distinct downstream networks, with the 8 Hz BLA interneuron stimulation preferentially engaging NAc neurons and both frequencies engaging the BNST. Brain-wide cFos quantification revealed that distinct networks were engaged by BLA interneuron stimulation.

Conclusions: Our results provide a framework to understand BLA valence processing as a product of microcircuit governance over ensemble engagement, oscillatory states, and network activation. Broadly, we aim to emphasize the importance of interneurons in the valence computation landscape and frame valence processing as a function of microcircuit engagement of systems-level networks.

Keywords: amygdala, oscillations, valence, interneurons

Disclosure: Nothing to disclose.

Scott Urban, Mary Vanhart, Aisha Mesco, Kurt Sarner, Eliza Rider, Jasper Heinsbroek*

The University of Alabama At Birmingham, Birmingham, Alabama, United States

Background: The ventral pallidum (VP) is a central node in the ventral basal ganglia that regulates motivated behavior and reward processing. Although the VP has traditionally been considered an inhibitory (GABAergic) relay nucleus, recent findings show complex cellular heterogeneity within this structure. Work from our lab and others has implicated GABAergic VP (VP-GABA) projection neurons in motivation and reward seeking and glutamatergic VP (VP-Glu) neurons in aversion processing. The VP also comprises multiple other subclasses of cells, including fast-spiking neurons that produce parvalbumin (VP-PV), and regular spiking neurons that produce the opioid neuropeptide pro-enkephalin (VP-Penk). Here we investigate the contributions of these cell-types in the regulation of reward seeking under increasing-effort requirements using progressive ratio (PR) tests wherein we measured the maximum effort an animal is willing to exert (break point) to obtain a sucrose reward.

Methods: Transgenic male and female Cre-driver mice (Vgat-Cre, Penk-Cre, PV-Cre, Vglut2-Cre; n = 8-10 per group) were surgically injected with cell-specific chemogenetic (hM4D) or optogenetic (BiPOLES) actuators and implanted with optic fibers above the VP. Mice were trained to nose poke for sucrose rewards (15 mg) on a fixed ratio schedule of reinforcement (FR1, 8d; FR5, 5d). Afterwards mice underwent PR training and testing. For chemogenetic studies, mice were injected 30 minutes prior to testing with the compound J60 (i.p.). To investigate the temporally specific effects of optogenetic manipulations, BiPOLES-injected mice received 2s photostimulation (630 nm, 40 Hz, 10 ms) or photoinhibition (450 nm) during the beginning of each trial, during reward delivery or during reward retrieval. Paired t-tests were used for all statistical comparisons. All experimental procedures followed the guidelines outlined in the Guide for the Care and Use of Laboratory Animals and were approved by the University of Colorado and University of Alabama at Birmingham Institutional Animal Care and Use Committee.

Results: Chemogenetic inhibition of VP-GABA neurons reduced break points (p < 0.01), and inhibition of VP-Glu neurons increased (p < 0.05) break points for sucrose. By contrast, chemogenetic inhibition of VP-Penk or VP-PV did not consistently alter break points. Optogenetic manipulations revealed increased break points for sucrose when stimulating VP-GABA (p < 0.01), VP-Penk (p < 0.01) or VP-PV neurons (p < 0.05) during reward delivery or reward retrieval. By contrast, only the inhibition of VP-GABA neurons during reward delivery reduced break points for sucrose (p < 0.05). Surprisingly, neither stimulation, nor inhibition of VP-Glu neurons produced any changes in break points.

Conclusions: These results reveal a complex temporal and cell-type specific regulation of motivation for rewards by population of VP neurons. Specifically, the surprising divergence between continuous chemogenetic and temporally specific VP-Glu manipulations warrants further investigation. To this end, ongoing studies in our lab are examining the activity of populations of VP neurons around specific epochs in the PR test using calcium imaging. Ultimately, a better understanding of the cell-specific contributions of VP neurons to motivated behavior will be crucial for developing novel treatments for psychiatric disorders characterized by aberrant motivation such as major depressive disorder and substance use disorders.

Keywords: Ventral Pallidum, Progressive ratio testing, optogenetics, chemogenetics, Reward self-administration

Disclosure: Delix, Consultant, Spouse/Partner

Ronald Kim*, Lorna Role, David Talmage

National Institutes of Health, Bethesda, Maryland, United States

Background: The ability to approach rewarding stimuli and avoid aversive stimuli is critical for survival. The ventral pallidum (VP) contains two distinct subpopulations of cholinergic neurons that differentially encode valence and play opposing roles in approach vs. avoidance behaviors. The primary projection target of VP cholinergic neurons is the basolateral amygdala (BLA). The BLA plays a crucial role in many affective processes, including proper valence encoding of salient stimuli leading to approach vs. avoidance behaviors. However, it is unknown if valence-encoding VP cholinergic neurons form functional connections with valence-encoding BLA neurons to mediate motivation to approach or avoid a stimulus. The goal of the present experiment is to examine the circuit between VP cholinergic neurons and the BLA and its role in innate approach and avoidance behaviors.

Methods: To determine if VP cholinergic neurons work in conjunction with BLA neurons to mediate approach and/or avoidance behaviors, we utilized an appetitive odor (2-phenylethanol), which normally elicits approach behavior, and an aversive odor (mountain lion urine) which typically incites avoidance behavior. To manipulate VP cholinergic signaling in the BLA, ChAT-cre mice were injected with a cre-dependent, red-shifted channelrhodopsin (AAV-FLEX-ChrimsonR) or red-shifted halorhodopsin (AAV-FLEX-Jaws) in the VP. Mice were simultaneously injected with a genetically encoded calcium indicator (AAV-CaMKII-GCaMP6F) and implanted with a GRIN lens in the BLA. We employed single-cell calcium imaging of BLA neurons during presentation of each odor, with and without optogenetic stimulation of VP cholinergic terminals in the BLA. We examined (1) the number of appetitive vs. aversive BLA neurons (i.e., the number of BLA neurons activated exclusively be each odor, and (2) the calcium activity of appetitive and aversive BLA neurons. In the same mice, innate approach and avoidance behaviors in response to each odor was tested with concurrent optogenetic stimulation of VP cholinergic terminals in the BLA.

Results: Prior to optogenetic stimulation of VP cholinergic terminals in the BLA, we first identified an approximately equal number of BLA neurons exclusively activated by each odor (~43% of BLA neurons activated only by the appetitive odor and ~44% of BLA neurons activated only by the aversive odor).

Enhancing VP to BLA cholinergic signaling (with optogenetic excitation) during odor presentation did not change the number of identified appetitive vs. aversive BLA neurons. However, optogenetic excitation of VP cholinergic terminals in the BLA increased the calcium activity of aversive BLA neurons, but not appetitive BLA neurons. When measuring approach or avoidance behavior in response to each odor, optogenetic excitation of VP cholinergic terminals in the BLA abolished approach to the appetitive odor but did not affect avoidance of the aversive odor.

Inhibiting VP to BLA cholinergic signaling (with optogenetic inhibition) during odor delivery altered the number of identified appetitive vs. aversive BLA neurons. With optogenetic inhibition of VP cholinergic terminals, a greater number of BLA neurons were classified as appetitive BLA neurons. Optogenetic inhibition of VP cholinergic terminals in the BLA did not result in any changes in the calcium activity of appetitive or aversive BLA neurons. When measuring approach or avoidance behavior in response to each odor, optogenetic inhibition of VP cholinergic terminals in the BLA enhanced approach to the appetitive odor but did not affect avoidance of the aversive odor.

Conclusions: The results from these studies indicate distinct populations of BLA neurons are activated by an appetitive vs. aversive odor. Increasing acetylcholine in the BLA increased the calcium activity of aversive BLA neurons, but not appetitive BLA neurons. This increase in the calcium response of aversive BLA neurons eliminated approach behavior in response to appetitive odor exposure. Decreasing acetylcholine in the BLA increased the number of identified appetitive BLA neurons. This increase in the number of appetitive BLA neurons was correlated with an exaggerated approach behavior in response to the appetitive odor. In summary, increasing VP to BLA cholinergic signaling abolished approach behavior, whereas inhibition leads to intensified approach behavior. Optogenetic stimulation of VP cholinergic terminals did not affect avoidance behavior. These results indicate VP cholinergic signaling in the BLA can bidirectionally mediate gain control over appetitive responses but is limited in gain control over aversive responses.

Keywords: Ventral Pallidum, cholinergic function, basolateral amygdala, valence, Approach/Avoidance

Disclosure: Nothing to disclose.

Ran Qiao, Michelle Doyle, Benjamin Johnson, Giordano de Guglielmo, Abraham Palmer, Marsida Kallupi*

UCSD, La Jolla, California, United States

Background: Cocaine Use Disorder (CUD) poses a significant public health challenge, driven by a complex interplay of genetic, environmental, and individual factors, with limited treatment options available. While advances in understanding the genetic basis of addiction vulnerability have been made, a critical gap persists in elucidating the neurobiological mechanisms that underlie individual differences in addiction susceptibility, particularly concerning pre-existing cellular differences.

Methods: This investigation employs an innovative approach known as RATTACA (Rapid Assessment of Trait-Associated Cellular Activity), which harnesses comprehensive genotype and phenotype data to enable precise polygenic trait prediction in naïve rats. By utilizing this method, we can effectively select rats anticipated to exhibit high and low cocaine addiction-like behaviors prior to any cocaine exposure. The study focuses on characterizing and comparing basal GABAergic transmission within the central amygdala (CeA) of these selected groups. Furthermore, we explore the differential modulation of these transmission pathways by cocaine through ex vivo application, followed by detailed electrophysiological recordings to assess the functional impact of cocaine on GABAergic and glutamatergic signaling.

Results: Preliminary findings indicate that rats predicted to exhibit high cocaine addiction-like behaviors demonstrate significantly elevated tonic GABA release compared to their low-predicted counterparts. Notably, cocaine administration resulted in a reduction of GABA release exclusively in the high-predicted susceptibility group.

Conclusions: These results highlight the utility of RATTACA in investigating pre-existing individual differences at the cellular level, which may elucidate the neurobiological underpinnings of addiction susceptibility. The observed differences in GABAergic transmission suggest a potential target for therapeutic intervention in CUD and other substance use disorders. This study not only enhances our understanding of the role of GABAergic transmission in addiction vulnerability but also promotes the exploration of novel therapeutic avenues.

Keywords: RATTACA, Cocaine Use Disorder, GABAergic transmission, central amygdala, addiction susceptibility

Disclosure: Nothing to disclose.

Sean Piantadosi*, Madison Martin, Veronica Porubsky, Avi Matarasso, Heidi Neuman, Selena Schattauer, Sarah Thai, Tammy Nguyen, Larry Zweifel, Michael Bruchas

University of Washington, Seattle, Washington, United States

Background: Acute stress and threat produce physiological anxiety thought to facilitate planning and allow an organism to adapt its behavior for future exploration of the environment. This serves as an adaptive mechanism that allows anxiety-like behavior and avoidance to be tuned and selected. However, in many mental health disorders this homeostatic behavioral response becomes maladaptive and dysfunctional, leading to excessive anxiety in scenarios where it is unwarranted or undesirable. Anxiety disorders, which are typified by this maladaptive response, are the most common mental illness in the U.S., affecting roughly 20% of the adult population. Clinically, we know that neuromodulators such as norepinephrine (NE) play pivotal roles in long-term outcomes following stress exposure. Despite this, the mechanism by which monoamine neuromodulatory signals regulate circuit activity associated with anxiety-like behavior remains poorly understood. One key modulatory system well positioned to mediate these behaviors is the locus coeruleus noradrenergic system (LC-NE), as it is one of the first engaged following stressful events and stimuli and activation of the LC and its projections to the basolateral amygdala (BLA) is anxiogenic. However, we know very little about how this critical neuromodulatory LC-BLA circuit generates anxiety-like behavior at the network, circuit, cell type, transmitter, and receptor level.

Methods: We conducted two-photon calcium imaging of LC-NE neurons using an endoscopic prism lens while mice (n = 8, 5 male, 3 female) were exposed to the potent predator stressor 2MT. Next, we measured NE release in the BLA using a NE sensor (GRABNE2m) in response to a 2MT and during activation of LC terminals in the BLA at a tonic (5hz) frequency using combined optogenetics and fiber photometry (n = 6, 6 male). We then performed freely moving calcium imaging of BLA neurons in classic anxiety-like behavior assays while manipulating LC terminals optogenetically (ChrimsonR) at the stress-induced tonic frequency (Dbh-cre mice; n = 8, 4 male, 4 female, 825 total neurons). Systemic pharmacology and rigorous single-cell tracking was used to discern which adrenergic receptors (ARs) mediate the anxiogenic effect of LC-BLA stimulation. Support vector classifiers were trained to predict whether mice were in an anxiogenic context based solely on BLA population activity. Graph theory approaches were developed to test whether and how tonic LC terminal activation altered the correlated activity of BLA neurons as might be expected of a gain control signal. Finally, CRISPR-SaCas9 knockdown of the gene encoding the gene encoding the β2-AR within excitatory BLA neurons was performed and mice were tested in a stress-induced anxiety-like behavior assay (n = 8 Adrb2-CRISPR, 4 control).

Results: LC-NE neuron activity was found to be highly synchronous when mice were presented with the predator odor 2MT. Relative to pre-odor baseline, mean fluorescence of LC neurons was significantly elevated during 2MT exposure (t(99) = 10.2, p < 0.0001). 2MT produced an increase in fluorescence in 89% of LC neurons recorded. Spike deconvolution found that underlying the changes in single-neuron fluorescence was an increase in tonic spiking of LC neurons during predator odor exposure (t(37) = 3.4, p = 0.002). We next examined whether this stress-induced tonic LC-NE neuron activation was reflected in norepinephrine release downstream in the BLA. Closely mirroring what we observed at the cell bodies, a robust and sustained increase in NE release was detected in BLA following 2MT exposure (RM-ANOVA, Sidak’s (t(3) = 6.7, p = 0.013)). We next identified a precise frequency (LC-BLA5hz) of LC terminal activation that produced similar levels of NE release in BLA relative to the release produced following 2MT exposure, as measured by GRABNE2m signal (RM-ANOVA, Sidak’s (t(3) = 4.7, p = 0.04)). With the ability to produce stress-like NE release, we then combined optogenetics and freely moving microendoscopy to record BLA neuron activity while manipulating LC terminals in tests of conflict anxiety. We found that tonic LC-BLA activation was anxiogenic (RM-ANOVA, p < 0.05; post-hoc t(7) = 3.41,p < 0.05) and improved the ability of BLA population activity to classify anxiogenic location (t(6) = 2.95,p < 0.05). We found that LC-BLA5hz stimulation enhanced correlated activity of BLA neurons (t(824) = 2.1, p = 0.03) and dramatically enhanced the clustering coefficient of groups of BLA neurons (KS test, D = 0.19, p < 0.0001), suggesting a change in BLA gain. These effects were blocked by the β-AR antagonist propranolol (p > 0.05). CRISPR-SaCas9 knockdown of Adrb2 in Vglut1+ BLA neurons resulted in a roughly 40% knockdown. Relative to controls, Adrb2-CRISPR mice spent more time exploring (t(10) = 2.8, p = 0.02), produced fewer fecal boli (t(10) = 2.5, p = 0.03), and made more escape attempts (t(10) = 1.5, p = 0.16) during 2MT exposure.

Conclusions: Neuromodulation, and specifically the neuromodulator norepinephrine, has long been suggested to serve as a gain control signal capable of shifting the activity of large populations of downstream neurons. Here using a combination of behavioral, optical, and molecular approaches we show that sustained release of norepinephrine following stress or optogenetic activation produces lasting alterations in BLA network activity, consistent with this theory, and identify the potential receptor (β2-AR) that might be necessary for this type of network shift.

Keywords: Anxiety and stress, basolateral amygdala, in vivo calcium imaging, adrenergic receptor, CRISPR

Disclosure: Nothing to disclose.

Gabriela Lopez*, Louis Van Camp, Ryan Kovaleski, Oscar Moreno-Ramos, Rajeshwar Awatramani, Julia Cox, Talia Lerner

Feinberg School of Medicine, Northwestern University, Chicago, Chicago, Illinois, United States

Background: To avoid danger, animals must learn about cues in their environment that predict danger and act in response. Previous work investigating dopamine release in the nucleus accumbens core (Core) and ventromedial shell (vmShell) implicate the mesoaccumbal circuit in avoidance learning, but interpretations are limited by several factors. First, although several studies have recorded Core and vmShell dopamine responses to unavoidable aversive stimuli, the contribution of these responses to shaping avoidance behavior as learning evolves is unclear. Second, reports of dopamine responses to unavoidable aversive stimuli have found inconsistent results, which we hypothesize are due to dopamine neuron heterogeneity. Our studies address these issues by recording Core and vmShell dopamine dynamics across an active avoidance learning task.

Methods: To determine how nucleus accumbens (NAc) dopamine signals contribute to active avoidance learning, we expressed the fluorescent dopamine sensor dLight1.3b in the NAc core (Core) or ventromedial shell (vmShell) and implanted an optical fiber for in vivo fiber photometry recordings during freely moving behavior. After recovery from surgery, we trained mice to avoid a footshock punishment (0.4 mA) by moving to the opposite side of a 2-chamber apparatus during a five-second warning cue. Movement to the opposite side of the chamber within 5s of cue start resulted in cessation of the cue, indicating safety, and avoidance of the shock (“avoid”). If the mouse did not move to the opposite side of the chamber within 5s, shock punishment began and continued until the mouse moved to the opposite side of the chamber (“escape”). Mice were tested on this task for 7 days, 30 trials per day. Finally, mice were subjected to an inescapable shock task where a 5s light and sound cue predicted 5s of 0.4mA shock that they could not avoid or escape. Mice were tested on this task for 1 session only consisting of 10 trials.

Results: All mice learned to avoid ≥85% of shocks within 7 days of active avoidance training, with individual variability in the speed of learning. Core dopamine showed negative-going responses (“dips”) to the warning cue that strengthen with learning and are especially evident during expert performance. vmShell dopamine signals showed an oppositely signed and more complex pattern of change that were present during early learning but not sustained during expert performance.

Conclusions: Our work supports a model for NAc Core and vmShell dopamine function in which these two regions not only encode aversion oppositely but also guide avoidance through different computational principles. Further understanding this computational heterogeneity in dopamine function, and its behavioral consequences, can provide avenues to understand how dopamine subpopulations work in concert to provide animals with tools to avoid danger in their environment.

Keywords: Active Avoidance, Dopamine, negative reinforcement, neural heterogeneity, Aversive Learning

Disclosure: Nothing to disclose.

Adam Gordon-Fennell*, Madelyn Critz, Madelyn Hjort, Barbara Benowitz, Garret Stuber

University of Washington, Seattle, Washington, United States

Background: The lateral hypothalamus (LHA) and dopamine release throughout the striatum regulate unique components of learning and motivation, but how these two structures are coupled and signal during consumption remains unsettled. Multiple studies have revealed that LHA GABA neurons and LHA glutamate neurons enhance and suppress dopamine release in the nucleus accumbens core, respectively. Subsequent studies have found that LHA glutamate cells can enhance or suppress dopamine release in specific subregions of the nucleus accumbens shell. Beyond the ventral striatum, tracing studies suggest that LHA neurons form synapses on subpopulations of midbrain dopamine neurons defined by projections to all subregions of the striatum. Furthermore, extensive research into reinforcement learning and motivation have determined dopamine dynamics during consumption of fixed reward volumes, but there has been scant investigation into how the dopamine system signals during periods where animals must shape their consumption response following the detection and valuation of solutions.

Methods: To model consummatory behavior, we employed a head-fixed multiple solution brief access taste task that I developed to elicit a range of consumption responses to a spectrum of rewarding solutions (0-30% sucrose) and aversive solutions (0-1.5M NaCl). To measure the activity of LHA GABA and glutamate cells simultaneously, we employed dual-color fiber photometry in double transgenic mice (VGlut2-Cre:VGat-Flp with fDIO-GCaAMP6s and FLEX-jRCaMP1b; n = 12). To measure striatum-wide dopamine dynamics during consumption, we employed multi-site fiber photometry paired with GRAB-DA2m to record dopamine release in up to 6 striatal sites simultaneously including multiple regions such as the ventral striatum, dorsal medial striatum, dorsolateral striatum, and the tail of the striatum (n = 18). To assess the functional connection between the LHA and striatal dopamine release, we used optogenetics (FLEX-Chrimson for stimulation, DIO-HR for inhibition, or DIO-mCherry for control) in transgenic mice (VGlut2-Cre or VGat-Cre) to regulate the activity of LHA neurons while simultaneously recording dopamine release across the striatum (n = 5-7 / group). To assess the components of the task that dopamine represents, we employed a generalized linear model to measure the unique contribution of each task component to the recorded signal. Summary statistics were computed using repeated measures ANOVA followed by a pair-wise Tukey honest significant difference post-hoc test or a Friedman test followed by a pair-wise Wilcox post-hoc test. All experiments were performed in male and female mice.

Results: LHA GABA and glutamate neurons differentially encode consumption of rewarding and aversive solutions (solution x subpopulation interaction: F4,56 = 12.34, P < 0.001 [rewarding], F4,56 = 16.70 P < 0.001 [aversive]). LHA glutamate neurons do not scale when consuming rewarding solutions (F4,56 = 0.96, P = 0.44), but show activity that is negatively correlated with value when consuming aversive solutions (F4,52 = 11.9, P < 0.001). Conversely, LHA GABA neurons show activity that is positively correlated with value when consuming both rewarding and aversive solutions. We found that dopamine release scales with consumption of rewarding and aversive solutions in every region of the striatum (P < 0.001), except the rostral core during consumption of rewarding solutions (P = 0.49). We found that LHA GABA and glutamate neurons have different functional coupling to dopamine release throughout the striatum (Interaction F5,45 = 47.79, P < 0.001 [vgat vs control]; F5,45 = 6.98, P < 0.001 [vgat vs control]). Using general linear models to separate the contribution of licking and other behavioral components, we find widespread encoding of licking across the striatum, and a range of contributions of solution value and solution value history to dopamine across the striatum (F5,80 = 11.06, P < 0.001) with a greater contribution in ventral striatum and less contribution in dorsal striatum.

Conclusions: These studies reveal that the LHA GABA and glutamate neurons have the capacity to regulate dopamine release across a previously unappreciated expanse of the striatum and display opposing scaling of activity during consumption of aversive solutions that coincides with a greater scaling of dopamine responses across the striatum. Furthermore, dopamine release in the striatum shows widespread encoding of licking but a spectrum of encoding of solution value and history. These results suggest that the LHA may have a key function in shaping dopamine signals during feeding and could be a potential intervention target for treating eating disorders.

Keywords: feeding behavior, Striatal Dopamine signaling, Lateral hypothalamus, Neural Reward Circuitry

Disclosure: Nothing to disclose.

Hannah Yoder, Skylar Anthony, Suganya Subramanian, Maria Jose Olvera Caltzontzin, Zhe Zhao, Sebastien Bullich, Saskia Stenzel, Darielle Lewis-Sanders, Stefano Berto, Sarah Stern*

Max Planck Institute, Jupiter, New York, United States

Background: The insular cortex is a multisensory brain region which is involved in many behaviors, including pain, addiction, social behavior, learning and memory, and feeding/metabolism. To date, the gene expression patterns of the insular cortex have not been systematically studied, but it has generally been assumed to be similar to other cortical areas. However, our data indicates that the insular cortex may have some key transcriptomic differences from other cortical areas, and that these insula-specific genes may have important behavioral functions. To probe this, we conducted single cell sequencing of the insular cortex and compared it to somatosensory cortex. We then investigated the function of two insula-specific genes in complex feeding behaviors.

Methods: Single nuclei sequencing was conducted for both insular cortex and somatosensory cortex (n = 9 for both) using the 10x Genomics platform. Resulting cDNA libraries and fragment sizes were quantified by Tapestation (Agilent) for quality control analysis and sequenced (2x300 bp, paired-end reads configuration) at UF Scripps Genomics Core using the NextSeq 2000 (Illumina). Final 10X snRNA-seq cDNA Libraries were sequenced at a minimum of 20,000 read pairs/cell. For individual cell-types, cre-driver lines were obtained (Nos1-Cre and LepR-Cre). Behaviors conducted included homeostatic food intake, conditioned overconsumption, conditioned taste aversion (CTA), fixed-ratio and progressive ratio operant behavior, real-time place preference, and memory and anxiety tasks (n = 7-12). Viral tracing was conducted using retrograde (monosynaptic rabies or retro-AAV) and anterograde (AAV1 or terminal tracing) virus (n = 4/experiment). Groups were compared with t-tests, 1 or 2-way ANOVA’s when appropriate.

Results: Single-nuclei sequencing revealed a conserved set of clusters in the insula and somatosensory cortices. Using an integrated dataset, we then identified differentially expressed genes between the two brain regions. We then followed up on two cell-type specific populations that are differentially expressed, namely Nos1 and LepR. Using a Nos1-cre mouse, we found that these neurons likely encode the exteroceptive, salient properties of food. When inhibited, food intake is only decreased in tasks in which learning is involved (conditioned overconsumption and CTA), while homeostatic feeding behaviors are unaffected. Short-term activation leads to acutely increased food intake, whereas long term activation leads to long term increases in food intake. Ongoing calcium imaging studies suggest that Nos1 neurons encode the salience of food to promote learned associations. In contrast, we find that LepR signals interoceptive signals to the insular cortex. Infusions of leptin into the insula decrease food intake, body weight and motivation, whereas activation of leptin receptors mediates operant feeding responses and promotes avoidance behaviors.

Conclusions: This insular cortex has a distinct transcriptomic makeup compared to somatosensory cortex. Probing the function of genes that are insula-specific reveals important insights into the functions of the insular cortex vis a vis complex motivated behaviors.

Keywords: Single nucleus RNA sequencing, insular cortex, Nos1, Leptin, food intake

Disclosure: Nothing to disclose.

Aubrey Moe*, Scott Blain, Aravind Kalathil, Mike Angstadt, Vaibhav Diwadkar, Stephan Taylor, Ivy Tso

The Ohio State University, Columbus, Ohio, United States

Background: Individuals with schizophrenia have prominent impairments in social cognition that contribute to poor functional outcomes. Gaze processing – a basic building block of social cognition involving the ability to accurately perceive human eye gaze direction – is similarly abnormal in schizophrenia and associated with social functioning. Though gaze processing has known neural underpinnings in cerebral social-processing brain regions, emerging data suggest that the posterior cerebellum plays a key but often neglected role in social-cognitive function. Elucidating the bi-directional cerebellum-cerebrum dynamics underlying social cognition is critical to identifying a novel treatment target.

Methods: 72 adult participants (males and females ages 18-55; n = 39 individuals with schizophrenia, n = 33 non-clinical controls) completed an eye gaze processing task during functional magnetic resonance imaging. One cerebellum (Crus II) and three cerebrum (inferior parietal lobule, anterior insula, and fusiform gyrus) regions were identified as nodes for dynamic causal modeling (DCM).

Results: Model comparison revealed extreme evidence that gaze processing (relative to general face processing) modulates cerebellar outputs to social-processing nodes in the cerebrum (Bayes Factor =1.63 x 104). Relative to non-clinical controls, individuals with schizophrenia showed less upregulation of cerebellar outputs during gaze processing (t(68) = 2.853, p = 0.006) and this decreased upregulation was associated with social dysfunction (ρ = -0.254, p = 0.032).

Conclusions: People with schizophrenia have altered cerebellum-cerebrum dynamics during gaze processing, with our results providing compelling evidence that gaze-related modulation of cerebellum-cerebrum connectivity is linked to social functioning. Given the accessibility of the cerebellum from the scalp for non-invasive brain stimulation relative to many cerebral social-cognitive regions, our findings have critical implications for translational interventions.

Keywords: Dynamic Causal Modeling, Schizophrenia (SCZ), Social Cognition, Cerebellum, Neural Networks

Disclosure: Nothing to disclose.

Elizabeth Crummy*, Brittany Chamberlain, Matthew Geramita, Susanne Ahmari

University of Pittsburgh, Pittsburgh, Pennsylvania, United States

Background: Obsessive-Compulsive Disorder (OCD), a psychiatric disorder impacting 2-3% of the population, is characterized by distressing obsessions and compulsive actions or thoughts performed to alleviate anxiety and avoid perceived threats. A core feature of OCD is avoidant behaviors, which are exhibited by 60% of patients, but the role of avoidance in the development and persistence of OCD is unknown. The bed nucleus of the stria terminalis (BNST) is known to regulate anxiety, and deep brain stimulation (DBS) in the BNST of patients with OCD reduces symptom severity, making it a promising target for treatment intervention. However, how the BNST regulates active avoidance behaviors is not well understood. We therefore used one-photon calcium imaging during a lever-press avoidance task to understand if and how the BNST encodes activate avoidance. We hypothesized that BNST neurons are recruited and encode avoidance responses following acquisition of a conditioned active avoidance behavior.

Methods: Male and female C57BL6 mice (n = 13; 9M, 4F) were injected with AAV9-syn-GCaMP6M (A/P: +0.26, M/L: +0.85, D/V = -3.9, -4.0) and implanted with an integrated GRIN lens and baseplate (D/V = -3.8) in the BNST for in vivo calcium imaging recordings during an operant active avoidance task. Mice were trained to lever press to avoid impending footshocks across 7 active avoidance sessions (50 trials/session), with acquisition criteria of 80% of trials avoided on the final session. Calcium recordings were preprocessed, down-sampled to 10 frames per second, and motion corrected. Cells were identified using constrained nonnegative matrix factorization. Modulation of BNST neurons during event windows of -2s to +2s was examined for cue onsets, avoid presses, escape presses, and shocks. Cells were classified as modulated if their trial-averaged, z-scored activity was 2 standard deviations above a shuffled trace within -0.5s to +1s of event onsets. To determine if there was an increase in BNST neuron modulation following task acquisition, the proportion of trials excited, inhibited, or unmodulated for each event was compared between day 1 and day 7 with unpaired t-tests, and average event traces between day 1 and day 7 for excited and inhibited populations were compared with two-way ANOVAs. To determine if the reliability in event modulation increased with well-learned avoidance, cell responses around each event window were considered modulated on that trial if the trace was ±1 standard deviation from the shuffle. The average number of modulated trials was compared between modulated and unmodulated neurons, as was the response composition of session excited or inhibited neurons in their proportion of trials excited, inhibited, or unmodulated for training day 1 vs. day 7.

Results: Lever press-modulated neurons emerged within the first session and were largely selective to avoid-only or escape-only response types (89.1% avoid-responsive vs. 10.8% avoid + escape responsive). The average proportion of cells inhibited during avoid presses increased from 5.1% to 12.4% following acquisition (day 1 vs. day 7: t(11) = 2.34, p = 0.039), while the excited proportions did not change (9.8% D1 and 8.5% D7 on average: t(11) = 0.37, ns). Cue modulation did not change with acquisition (4.2% D1 v 6.1% D7 excited: t(12) = 0.76, ns; 2.6% D1 v 3.4% D7 inhibited: t(12) = 0.77, ns). While the proportion of neurons inhibited increased with acquisition, average amplitude during avoid events was not significantly different between day 1 and day 7 (day 1 vs. day 7 event time x session: F (15, 1050) = 2.192, p = 0.0054; Sidak’s multiple comparisons: not significant). When examining whether modulation was consistent across trials, cells identified as avoid modulated across the session were modulated for 35-37% of trials on day 1 and day 7, respectively, whereas cells that were not modulated over the session were classified as modulated for 25-27% of trials on day 1 and day 7 (avoid modulated vs. non-modulated cell reliability unpaired t-test day 1: t(314) = 4.84, p < 0.0001; day 7: t(467) = 8.33, p < 0.0001). However, while the overall proportion of inhibited cells increased from day 1 vs. day 7, no significant differences were found in the reliability of their modulation across these sessions (two-way ANOVA main effect of session: F (1, 210) = 0.0001827).

Conclusions: Our data suggest that BNST neurons are modulated following learned defensive actions which are specific to avoid responses, but not threat cues. Specifically, there is an increased proportion of inhibited BNST populations during an avoid response. However, cell modulation does not become more consistent with avoidance acquisition, suggesting that the magnitude or stability of the response is not driving conditioned avoidance. Future work will examine whether the BNST is encoding uncertainty of encountering threats, and whether stimulation of BNST neurons at the time of lever extension (beginning of opportunities to make avoid responses) reduces avoid responses in well-learned mice. These findings help elucidate mechanisms for development of avoidant behaviors, and may inform how these behaviors become maladaptive in OCD.

Keywords: Active Avoidance, bed nucleus of the stria terminalis, in vivo calcium imaging, OCD, negative reinforcement

Disclosure: Nothing to disclose.

Seungwon Choi*

The University of Texas Southwestern Medical Center, Dallas, Texas, United States

Background: Pain is initiated by the activation of nociceptors that innervate the skin and internal organs. Nociceptive signals are propagated into the spinal cord and then transmitted to the brain by spinal cord projection neurons (PNs). These spinal PNs are attractive therapeutic targets for pain treatment because nociceptive signals emanating from the periphery are channeled through these spinal cord output neurons en route to the brain to produce pain sensations. Spinoparabrachial (SPB) neurons, a major population of spinal PNs that innervate the lateral parabrachial nucleus of the pons, represent an ideal neuronal population for developing new approaches to treat pain because they convey touch and pain information to higher brain centers that control the affective aspects (i.e., emotional “feelings”) of touch and pain. Traditional view of pain pathways has been focused on the spinal projection neurons expressing TACR1, the Substance P receptor. However, antagonizing TACR1 has turned out not efficacious in treating pain in human patients, suggesting there might be additional, independent spinal output pain circuits. Previously, we identified a novel, largely distinct subset of spinal PNs that express another neuropeptide receptor, GPR83, and demonstrated that these spinal PNs also transmit noxious signals to the brain in mice.

Methods: To further define the contribution of each SPB subdivision to pain sensation and associated behavioral responses to noxious stimuli, we first examined the effects of acute silencing of Tacr1+ and Gpr83+ SPB neurons, individually or simultaneously, on noxious heat-induced nocifensive behaviors using a newly generated dual recombinase-dependent SYP1-miniSOG mouse line that enables selective, light-dependent silencing of synaptic transmission. In addition, to determine if neuropeptide signaling mediated by either TACR1, GPR83 or both in the spinal cord is required for pain transmission, we also examined acute and neuropathic pain behaviors following spinal cord-specific deletions of the Tacr1 and Gpr83 genes using mouse lines that harbor conditional alleles of Tacr1 and Gpr83 in conjunction with spinal cord specific Cre lines. Furthermore, we also examined Tacr1 and Gpr83 gene expression in the human spinal cord using RNAscope analysis. While we continue to collect more numbers, we have collected data from 5 – 10 animals (both males and females) for each genotype/manipulation. Multiple group comparisons have been assessed using one-way, two-way, or repeated-measures analysis of variance ANOVA tests, followed by post hoc tests.

Results: Our preliminary results suggest that simultaneous silencing of both Tacr1+ and Gpr83+ SPB neurons significantly reduces (~50%) the noxious heat-induced escape locomotion. We are currently analyzing the single silencing effect to test whether simultaneous silencing of the two major SPB populations is more effective in alleviating acute pain than silencing one population. Similarly, spinal cord-specific double KO, but not single KO of Tacr1 and Gpr83 genes, significantly reduced (~20 - 30%) neuropathic pain (mechanical and cold allodynia). In addition, our preliminary results suggest that Gpr83 is more abundantly expressed than Tacr1 in the superficial dorsal horn of the human spinal cord.

Conclusions: Overall, our preliminary results suggest that neuropeptide signaling mediated by TACR1 and GPR83 in the spinal cord plays redundant roles in mediating neuropathic pain. Likewise, Tacr1+ and Gpr83+ SPB neurons may form functionally redundant pain pathways. Furthermore, abundant expression of Gpr83 in the human spinal cord suggests that this newly identified population of spinal neurons may serve as a novel therapeutic target for treating neuropathic pain in humans.

Keywords: affective touch, chronic pain, Somatosensory processing, parabrachial nucleus, Spinal Cord

Disclosure: Nothing to disclose.

Michelle Sequeira*, Kathryn Stachowicz, Lisette Bahena, Shannon Gourley

Emory University, Atlanta, Georgia, United States

Background: In an ever-changing environment, organisms must be able to integrate new learning into established routines to behave flexibly. This process often requires the encoding and retrieval of new memories. The orbitofrontal cortex (OFC) integrates new learning into established knowledge and appears to maintain memories for novel outcome expectancies. Here, we tested whether the ventrolateral subregion (VLO) is necessary for memory encoding and retrieval, and whether inputs from the ventral hippocampus (vHC) contribute to this process and related neurosequelae in the VLO. Finally, given that the VLO and vHC are stress-sensitive regions, we tested whether glucocorticoid receptor activity is involved in action flexibility.

Methods: To quantify action flexibility, mice were trained to generate two distinct responses for food, then one behavior ceased to be reinforced. This triggers new memory encoding and its later retrieval when mice choose future actions (to favor the reinforced behavior). Chemogenetic strategies were used to investigate the role of excitatory neurons in the VLO and vHC in memory encoding and retrieval. Thy1-driven YFP allowed for the visualization of layer V excitatory neurons in the VLO. The glucocorticoid receptor antagonist, mifepristone, was used to assess the role of glucocorticoid signaling.

Results: The VLO is necessary for both memory encoding and retrieval. vHC-to-VLO projections appear necessary, but not sufficient, for optimal encoding, and projection activity impacts learning-related dendritic spine plasticity in the VLO. Finally, glucocorticoid receptor activity is necessary for memory retrieval, though interestingly, only in female mice.

Conclusions: The VLO is necessary for organisms to encode and retrieve outcome expectancies to behave flexibly. Inputs from the vHC appear to optimize memory formation and learning-related dendritic spine plasticity, potentially providing contextual or emotive content that reinforces memory. Glucocorticoid receptor signaling is necessary for memory retrieval, though only in female mice, which will be further investigated in the future.

Keywords: orbitofrontal cortex, ventral hippocampus, learning and memory, glucocorticoid, sex differences

Disclosure: Nothing to disclose.

Raajaram Gowrishankar*, Madelyn Gray, Abigail Elerding, Sofia Shirley, Josie Van Tilburg, Khalid Abrera, Katherine Motovilov, Sean Piantadosi, Adam Gordon-Fennell, Chunyang Dong, Lin Tian, Garret Stuber, Michael Bruchas

University of Washington, Seattle, Washington, United States

Background: Endogenous opioids, signaling via G protein-coupled receptors, are uniquely poised to regulate neuronal activity and behavior. Indeed, decades of literature have implicated aberrant opioid signaling in multiple neuropsychiatric disorders, including substance use disorder (SUD). However, they have largely only been used as markers to identify neurons and the few studies ascribing neuropeptide contributions to behavior lack spatiotemporal precision. Moreover, while studies have reported on the activity of neuropeptide-expressing neurons during behavior, the synergy between neuronal activity, neuropeptide release and the resulting encoding of behavior is lacking. The endogenous opioid dynorphin is highly enriched in the dorsal striatum, known to be critical for regulating reward-seeking behavior. However, the locus, the precise timescale, or functional role of endogenous dynorphin signaling via the kappa opioid receptor on goal-directed behavior is unknown.

Methods: All studies were conducted in equal numbers of male and female mice in accordance with NIH guidelines, approved by the IACUC at the University of Washington. No sex-dependent effects were observed. Below were the methods used:

1. In-vivo 2-photon imaging of DMS dyn neurons during head-fixed operant behavior (D1-tdTom, n = 4 mice).

2. In-vivo fiber photometry during operant behavior of dyn release in the DMS (KOR-cre, n = 6 mice).

3. Operant behavior with conditional dyn deletion (Ctrl or pdyn-cKO, n = 12 mice).

4. Time-locked optogenetic manipulations of DMS dyn neurons in-vivo during operant behavior for activation using ChR2 (Ctrl or pdyn-cre, n = 9 mice).

5. Time-locked optogenetic BLA stimulation and dyn release during operant behavior (KOR-cre, n = 4 mice).

6. Ex-vivo viral tracing and in situ hybridization of BLA projections to DMS (WT, n = 3 mice, 6 slices Ai14, n = 3 mice, 6 slices; vglut1-cre, n = 2 mice, 6 slices; KOR-cre, n = 2 mice, 6 slices).

7. Ex-vivo electrophysiology of DMS neurons during optogenetic activation of BLA terminals (D1-tdtmto, n = 5 mice, 16 cells); KOR activation via bath application (U69,593 – 1uM) at BLA terminals during optogenetic activation (D1-tdtmto, n = 4 mice, 9-16 cells).

8. In-vivo fiber photometry during operant behavior of BLA terminals in the DMS multiplexed with conditional dyn deletion (WT or pdyn-cKO, n = 8 mice) or dyn neuron stimulation (pdyn-cre, n = 4 mice).

Results: Here, we report that dynorphin neuron activity and subsequent dynorphin release are essential in the learning and maintenance of reward-seeking behavior. We find that as animals learn to perform actions to seek rewards (n = 4 mice, operant behavior, t test; p < 0.05, t(3) = 3.4), significant DMS dynorphin neuronal ensembles evolve dynamically to encode the anticipation of reward, as measured by in vivo two photon imaging calcium imaging. Remarkably, we observe dynorphin release at strikingly fast timescales, specifically during the anticipation of reward (n = 6 mice, activity z-score, unpaired t test; p < 0.0001 – early vs. trained). Upon conditional deletion of the precursor for dynorphin in the DMS, we find a reduction in the learning and maintenance of goal-directed behavior, without impacting innate preference for the reward (n = 12 mice, Simple Linear Regression; p < 0.005, R2 = 0.48). Conversely, when we stimulate dynorphin release only during the anticipation of reward via closed-loop optogenetics, we demonstrate an enhancement in reward-seeking, without being inherently reinforcing (n = 8 mice, paired t test, p < 0.005, t(7) = 10.55 – early vs. trained). Furthermore, we show that axon terminals from the basolateral amygdala activate DMS dynorphin neurons (via holographic optogenetic stimulation in vivo two photon imaging calcium imaging), and subsequent dynorphin release (by multiplexing optogenetics and in vivo biosensor imaging). Importantly, we find that DMS dyn, or BLA KOR deletion, significantly diminishes BLA-DMS activity dynamics to action and outcome (n = 7 mice, activity z-score, unpaired t test; p < 0.0005 – WT vs. dyncKO), and negatively impacted the learning and maintenance of goal-directed behavior (n = 7 mice, operant behavior, 2 Way ANOVA; p < 0.0005 – day x genotype, F(4,40) = 8.99 - WT vs. dyncKO).

Conclusions: Altogether, we reveal that retrograde dynorphin transmission from the DMS onto KOR at basolateral amygdala terminals dynamically shapes the neural activity for reward-seeking behavior. Collectively, our findings isolate a causal role for endogenous opioid release and a novel mechanism whereby subsequent GPCR activity at excitatory terminals shape fundamental reward-seeking behaviors, providing a framework for how they may go awry in reward-seeking for misused substances and SUDs.

Keywords: reward-seeking behavior, Dynorphin, kappa opioid receptor

Disclosure: Nothing to disclose.

Vedrana Cvetkovska, Mariana Alonso, Yiu-Chung Tse, Emily Cha, Eshaan S. Iyer, Delong Zhou, Katrina Khosravi, Joelle Lopez, Rosemary Bagot*

McGill University, Montreal, Canada

Background: Stress is a major risk factor for depression, yet some individuals remain resilient to stress. Both environmental and genetic factors contribute to the dysregulation of neural circuits in depression and genes that regulate the formation, maintenance, and plasticity of synapses are of particular interest as potential molecular hubs for regulation of circuit function.

We previously identified Sidekick-1 (Sdk1) as an affective circuit hub gene in a network regulating stress resilience and overexpression of Sdk1 in the prefrontal cortex (PFC) increased resilience to stress in male mice. Separately, several genome-wide association studies in humans found SDK1 variants associated with depression, suggesting an evolutionarily conserved role in regulation of emotional behavior.

Sdk1 is a cell surface molecule implicated in circuit formation in the developing retina. However, little is known about cell-specific expression patterns of Sdk1 in the adult brain and its function within affective circuits. Here we probe Sdk1 cell-type and layer-specific expression in the PFC, conservation between mice and humans, its modulation by stress, and effects on reward learning in both male and female mice.

Methods: female mice. Fluorescent in-situ hybridization qauntified Sdk1 levels and patterns of expression in the PFC of control, susceptible and resilient mice (n = 10-16. Probes for Slc17a7 (vGlut1) and Slc32a1 (vGAT) identified excitatory and inhibitory neurons, respectively. PFC layers 1, 2/3, 5 and 6 were designated by distance from the midline. scRNAseq datasets (Allen Brain Map) from mouse and human whole cortex were used to identify cell-type specific expression of Sdk1 and de novo clustering on sub-regions was performed using Seurat. Projection patterns of cells expressing Sdk1 were identified by injecting injecting cre-dependent ChR2 into PFC of transgenic mice and expressing cre in Sdk1+ cells followed by fluorescent microscopy (n = 3) and patch-clamp electrophysiology (n = 8-12). The activity of Sdk1+ cells was manipulated by injecting cre-dependent DREADD constructs in PFC of Sd1-Cre transgenic mice (n = 0. Reward learning was assessed using a probabilistic reversal learning task (n = 9-12).

Results: We found that Sdk1 is subpopulations of vGlut1-positive neurons and vGAT-positive neurons with the highest level of Sdk1 expression in deep layer 6 and moderate expression in layers 2/3 and 5. Patterns of cell-type and layer-specific Sdk1 expression were conserved in scRNAseq analysis of human PFC. Stress modulated Sdk1 expression in the PFC in a sex-specific manner. Sdk1+ cells project thalamic regions and synapse on other Sdk1+ (p < 0.05 vs. Sdk1- cells). Cell-type specific activity modulation of Sdk1+ cells in the PFC modulated reward learning in the probabilistic reversal learning task (p < 0.05).

Conclusions: In this study we characterized the prefrontal expression of Sdk1, a novel regulator of stress resilience. We found that Sdk1 expression is modulated by stress and is expressed in subsets of both excitatory and inhibitory neurons, with a significant enrichment in deep layers of the PFC and conserved expression between mice and humans. We identify thalamic nuclei as the primary projection target of Sdk1+ PFC neurons. Manipulating activity of these neurons in the probabilistic reversal learning task revealed effects on reward learning suggesting relevance to anhedonia. Future work will further examine the circuit-specific effects of Sdk1 expression in behavioral adaptation to stress. As a cell surface receptor, Sdk1 is a potential target for pharmacological intervention. Understanding how Sdk1 confers resilience may lead to development of new mechanistically-informed treatments for depression.

Keywords: Medial Prefrontal Cortex (mPFC), stress resilience and susceptibility, Cell- and Circuit-selectivity, Reward

Disclosure: Nothing to disclose.

Matthew Geramita*, Susanne Ahmari, Eric Yttri

University of Pittsburgh School of Medicine, Pittsburgh, Pennsylvania, United States

Background: Determining the best possible action in an uncertain situation is often challenging, and organisms frequently need extra time to deliberate. This pause in behavior in response to uncertainty – also known as hesitation – commonly occurs in many aspects of daily life, yet its neural circuits are poorly understood.

Methods: Mice (n = 32; 15 female) expressing either channelrhodopsin-2 (ChR2) in either direct or indirect pathway medium spiny neurons (dMSNs or iMSNs) were trained to perform an auditory conditioning task with three cue that predicted reward on 0%, 50% or 100% of trials. To manipulate activity, dMSNs or iMSNs were stimulated during the cue period. In a separate cohort of animals (n = 10; 4 female), MSN activity was recorded using Neuropixel probes, and single units were identified using Kilosort and Phy software.

Results: Using an auditory association task, we show that mice hesitate – that is they pause anticipatory licking exclusively in response to cues associated with an uncertain reward. We then discovered that hesitation is mediated by a specific population of dorsomedial striatum (DMS) indirect pathway neurons. We performed cell-type specific in vivo electrophysiology and found a population of indirect pathway spiny projection neurons (iSPNs) in the DMS that signal uncertain tone-reward associations. Additionally, the degree of hesitation correlated with trial-to-trial changes in the activity of these uncertainty-responsive iSPNs. Finally, we showed that increasing iSPN activity increased hesitation, while decreasing iSPN activity decreased hesitation. Importantly, neither DMS dSPNs, nor dSPNs or iSPNs in a separate striatal region influenced hesitation, suggesting a specific role for DMS iSPNs in hesitation.

Conclusions: Taken together, we show that the indirect pathway of the basal ganglia plays a critical role in hesitation, a poorly understood but important behavior in both health and disease, and that separate neural circuits underly action cancellation and the pausing involved in cognitive tasks like hesitation.

Keywords: uncertainty, Indirect Pathway, in vivo electrophysiology

Disclosure: Nothing to disclose.

Jingyi Chen*, Leandra Mangieri, Sophia Mar, Sean Piantadosi, David Marcus, Phoenix Davis, Benjamin Land, Michael Bruchas

University of Washington, Seattle, Washington, United States

Background: Stress has been shown to promote the development and persistence of binge eating behaviors. However, the neural circuit mechanisms for stress‐induced binge‐eating behaviors are largely unreported. The endogenous dynorphin (dyn)/kappa opioid receptor (KOR) opioid neuropeptide system has been well established to be a crucial mediator of and anhedonic components of stress. Here, we aimed to dissect the basis of dynorphinergic control of stress‐induced binge‐like eating behavior.

Methods: To examine how acute stressors alter mouse feeding behavior, we developed an approach where mice are subjected to either a 15‐minute forced swim stressor or in the empty bucket/home cage as control. After a 15‐minute recovery period, the mice were presented with both regular chow and previously familiarized HPD pellets (high sugar, high fat), and their intakes were recorded for 1 hour. We record the Claustrum signal cell activities using an endoscope. We also record the neural activities in the Claustrum subpopulations using fiber photometry with either calcium indicator (GCaMP) or genetically encoded dynorphin biosensor (Klight) combined with pharmacology manipulation. Then, we combine optogenetic stimulation and fiber photometry to study the changes in Claustrum neurons while stimulating the Insulin cortex (Pdyn + ) terminals.

Results: We first established a mouse behavioral model for stress‐induced binge‐like eating behaviors. We found that mice exposed to stress increased their food intake of familiar palatable food (high fat, high sugar, HPD) compared to non‐stressed mice. Following a brain‐wide analysis, we isolated robust cFos‐positive cells in the Claustrum (CLA), a subcortical structure with highly abundant KOR expression followed by stress‐induced binge‐eating behavior. We report that KOR signaling in CLA is necessary for this elevated stress‐induced binge eating behavior using local pharmacology and local deletion of KOR. In vivo calcium recordings using fiber photometry revealed a disinhibition circuit structure in the CLA during the initiation of HPD feeding bouts. We further established the dynamics of endogenous dynorphinergic control of this behavior using a genetically encoded dynorphin biosensor, Klight. Combined with 1‐photon single‐cell calcium imaging, we report significant heterogeneity with the CLA population during stress‐induced binge eating, and such behavior attenuates local dynorphin tone. Furthermore, we isolate the anterior Insular cortex (aIC) as the potential source of endogenous dynorphin afferents in the CLA.

Conclusions: By characterizing neural circuits and peptidergic mechanisms within the CLA, we uncover a pathway that implicates endogenous opioid regulation stress‐induced binge eating. These results suggest that there is a likely dynamic, binary role for dyn/KOR modulatory signaling in the CLA and implicate dynorphin acting to shape CLA output via actions on both the excitatory KOR‐ KOR-expressing neurons and inhibitory KOR‐expressing neurons within the region. Altogether, we have demonstrated a dyn/KOR dependent disinhibition circuit between aIC(Pdyn) ‐CLA(vGAT) ‐CLA(CaMKII) neurons that are critical for stress-induced binge eating behavior.

Keywords: Dynorphin, claustrum, kappa opioid receptors, binge-like eating

Disclosure: Nothing to disclose.

Shinwoo Kang, Minsu Yang, Minryung Song, Sang Wan Lee, Doo-Sup Choi*

Mayo Clinic College of Medicine and Science, Rochester, Minnesota, United States

Background: The external globus pallidus (GPe) integrates GABAergic signaling from the dorsal striatum, including caudate-like dorsomedial striatum (DMS) and putamen-like dorsolateral striatum (DLS). The main output of GPe is the subthalamic nucleus (STN). Notably, GPe contains abundant astrocytes, which may orchestrate different GABAergic neurons. However, whether GPe astrocyte-neuron interactions are critical for reward-seeking and reversal (switching the active hole) during operant conditioning remains unknown.

Methods: To quantify GPe astrocytes and neuronal activities, we employed DeepLabCut, a machine-learning-based video-tracking software, and in vivo calcium imaging with dual-channel fiber photometry. We then utilized the chemogenetic approach to determine whether inhibition of PV (GPe-- > STN) neurons alters reversal reward-seeking behaviors. We examined whether optogenetic activation of iMSN (DLS-- > GPe) neurons exhibited similar effects. Also, using optogenetics, we investigated the role of GluSTN→GPe neurons (AAV-DIO-ChR2 x the STN of Vglut2-Cre mice) in reversal reward-seeking behavior. Finally, we examined the PV (GPe-- > STN) circuit using Cre-dependent caspase-induced ablation.

Results: We found that GPe astrocytes encode, maintaining continuous reward-seeking. In contrast, parvalbumin PV (GPe-- > STN) neurons (retrograde jRGECO in the STN) are active during the reversal reward-seeking behavior. Interestingly, chemogenetic inhibition of PV (GPe-- > STN) neurons impairs the reversal reward-seeking behavior without changing the continuous reward-seeking, indicating that PV (GPe-- > STN) neurons are critical for adaptive reward-seeking behavior. We consistently found that the dampening of pan-GPe neuronal activities shows a similar deficit of reversal reward-seeking when optogenetically activating the iMSN (DLS-- > GPe) neurons (DIO-ChR2 x A2a-Cre in the DLS). Furthermore, chemogenetic activation of STN pan-neurons increases reversal reward-seeking behavior. Next, we found that activation of Glu (STN-- > GPe) neurons (AAV-DIO-ChR2 x the STN of Vglut2-Cre mice) accentuates reversal reward-seeking behavior. Furthermore, caspase-induced ablation of PV (GPe-- > STN) neurons impairs the reversal of reward-seeking and re-exploration, a similar behavior phenotype of indecisiveness.

Conclusions: Our findings provide a novel role of astrocytes and neurons in indirect basal ganglia (DLS-GPe-STN) circuits for adaptive reward-seeking and learning, which are related to maladaptive habits, such as addiction and obsessive-compulsive disorder.

Keywords: Globus Pallidus, fiber photometry, astrocyte-neuron interaction, Operant behavior, opto- chemogenetics

Disclosure: Nothing to disclose.

Serena Mares Malta*, Tamiris Sabrina Rodrigues, Matheus Henrique Silva, Fernanda Mascarenhas, Lucas Matos Martins Bernardes, Letícia Leandro Batista, Rafael Bernardes Ferreira, Alexandre Souza Marquez, Ana Carolina Santos, Foued Salmen Spindola, Carlos Ueira-Vieira, Ana Mendes-Silva

Federal University of Uberlândia, Uberlândia, Brazil

Background: Alzheimer’s disease (AD) is a progressive neurodegenerative disease and the most common form of dementia in the elderly. The amyloidogenic hypothesis, which describes the extracellular accumulation of amyloid beta peptides (Aβ), is the main theory explaining this phenomenon. Model organisms such as Drosophila melanogaster have been used to better understand this disease and its potential treatments. Previous studies from our group have shown that Brazilian kefir, both in natura and in its peptide and metabolic fractions, can modulate neurodegenerative processes (e.g.,vacuolar lesion) in a Drosophila model expressing human amyloid precursor protein (APP) and beta-site APP cleaving enzyme (BACE). Here in this study, we tested the effects of peptide and metabolic fractions derived from Brazilian kefir in a neurodegenerative phenotype using the D. melanogaster strain engineered to express the human Aβ42. Moreover, we evaluate the dynamic of Aβ42 aggregation in vitro with and without adding the peptide and metabolic fractions and how these fractions influence neuronal viability.

Methods: The effects of the < 10kDa peptide fraction and metabolic fractions derived from Brazilian kefir were evaluated in a D. melanogaster strain engineered to express the human Aβ42 peptide specifically in the eye, driven by the GMR-GAL4 system. Parameters evaluated included ommatidial organization, vacuole area, retinal thickness, and relative levels of Aβ peptides. In vitro, we used dynamic light scattering to evaluate the dynamics of Aβ aggregation and fibril formation in the presence and absence of the < 10kDa and Ethyl Acetate (EtOAc) fractions. In addition, human neuron cell lines (SH-SY5Y) were incubated with Aβ peptides, then the < 10kDa and EtOAc fractions were added, and cell viability was calculated using the AlamarBlue assay. All cell culture experiments were performed in three biological and technical replicates. Data analysis was performed using two-tailed ANOVA and unpaired t-test with GraphPad Prism.

Results: The results showed that the treatment with the peptide fraction < 10kDa [0.25 mg/mL] resulted in an increase in retinal thickness (p = 0.005) and a significant reduction in vacuole area and amyloid content (p = 0.009 and p = 0.008, respectively) in all comparisons with the untreated control group. Among the metabolic fractions, the EtOAc [0.25mg/mL] fraction stands out for its ability to promote a reduction in vacuole area (p = 0.023). Both fractions, < 10kDa and EtOAc at a concentration of 0.25mg/mL, were able to alter the in vitro aggregation dynamics of Aβ peptide, suggesting an interaction with it. In support of this possible interaction, SH-SY5Y cells incubated with < 10kDa peptide fraction exhibited a significant 15% increase in cell viability (p = 0.0017). In contrast, incubation with EtOAc did not show a significant increase in cellular viability (p = 0.26), compared to the control group.

Conclusions: The results of this study indicate that peptide and metabolite fractions derived from Brazilian kefir have significant therapeutic potential against the neurodegenerative effects associated with the Aβ42 peptide in Drosophila melanogaster. The < 10kDa peptide fraction showed a robust neuroprotective effect as evidenced by an increase in retinal thickness and a reduction in vacuole area and amyloid levels. In addition, the EtOAc fraction also showed the ability to reduce tissue degeneration, and both were able to modify the aggregation dynamics of Aβ peptide in vitro. These results suggest that kefir components may interfere with the Aβ aggregation process and modulate its toxic effects, which was further supported by the increase in cell viability in human neuroblastoma cells. Therefore, kefir fractions emerge as potential candidates for the development of innovative therapeutic strategies in the treatment of Alzheimer’s disease. Further research is needed to elucidate the molecular mechanisms underlying the interaction between kefir fractions and Aβ peptides. This would be a critical step towards potential clinical applications.

Keywords: Alzheimer disease, Drosphila melanogaster, Kefir fractions

Disclosure: Nothing to disclose.

Ethan Burstein*, P. Markus Dey, Sanjeev Pathak

ACADIA Pharmaceuticals Inc., San Diego, California, United States

Background: Pimavanserin (PIM) is a potent and selective 5-HT2A inverse agonist that is FDA approved for Parkinson’s disease psychosis. PIM carries a QT prolongation warning that limits the dose, and has a long (57 h) half-life, delaying attainment of steady state therapeutic levels. Therefore, we sought a compound that builds upon learnings from PIM, but with an improved profile, including reduced QT prolongation risk and a shorter t1/2, culminating in the identification of ACP-204.

Methods: Receptor Selection and Amplification Technology (RSAT) functional assays were performed as described previously (Vanover et al., 2006). Phosphatidyl inositol (IP1) and radioligand binding assays were conducted at Eurofins. hERG (human ether-à-go-go-related gene), L-type calcium (Cav1.2) and late sodium (Nav1.5) manual patch clamp assays were conducted at Charles River Laboratories (CRL, Cleveland, OH). hERG was run GLP. Rodent pharmacology experiments were conducted at Behavioral Pharma (La Jolla, CA) in accordance with the NIH Guidelines for the Care and Use of Laboratory Animals. DOI head twitch assays were carried out with male C57BL/6 mice and male Sprague-Dawley rats, using 2.5 mg/kg DOI given intraperitoneally (IP) and vehicle (0.9% saline) or ACP-204 given subcutaneously (SC) or orally (PO). MK-801 (dizocilpine) locomotor assays were carried out with Male BALB/c mice. Mice received vehicle, ACP-204 or M100907 SC 30 min prior to MK-801 (0.4 mg/kg, IP). Locomotor activity was measured by an automated infrared photobeam monitoring apparatus (Kinder Scientific®, San Diego). Positron emission tomography (PET) studies were carried out at Invicro (New Haven, CT), using a male Rhesus monkey (16 kg, 12 y) and [18F]Altanserin. Venous blood samples were collected at pre-dose and 15 min, 30 min, 60 min, 90 min, and 120 min post-dose for determination of ACP-204 plasma concentrations. PET images were acquired over 120 min and processed using PMOD 3.802 (PMOD Technologies, Zurich, Switzerland). The outcome measure was binding potential (BPND) using the cerebellum as the non-displaceable reference region. Occupancy was computed as % Occupancy = [1 - (Post- ACP-204 BPND/Baseline BPND)] x 100. 5-HT2A receptor occupancy and ACP-204 plasma concentration were related with a single binding site model with Hill slope = 1. Rat and Cynomolgus monkey pharmacokinetic (PK) studies were carried out at Wuxi using male (230-260 g) and female (200-230 g) Sprague-Dawley rats and male Cynomolgus monkeys (2 – 5 kg, ≥ 2 years old). Plasma and CSF concentration vs time data was analyzed by non-compartmental approaches using Phoenix WinNonlin 6.3 software. Rat and Cynomolgus monkey repeat-dose toxicity studies, monkey cardiovascular safety, and rat neurobehavioral, respiratory and genotoxicity studies were performed GLP at CRL.

Results: ACP-204 had 0.3 to 0.5 nM potency as an antagonist/inverse agonist at 5-HT2A in RSAT and IP1 assays, and 0.14 nM affinity in radioligand binding. Compared to PIM, ACP-204 was 3 to 9-fold more potent in the functional assays, and similar affinity in binding. ACP-204 had 16 to 37 nM potency as an antagonist/inverse agonist at 5-HT2C, 1.9 nM binding affinity for 5-HT2C, no appreciable functional activity against 19 other monoamine targets, and < 50% inhibition at 5 µM against 56 additional targets in radioligand binding. ACP-204 potently suppressed DOI-induced head twitch in mice and rats with minimum effective doses (MEDs) of 0.1 and 0.3 mg/kg respectively, and nearly complete suppression at 0.3 and 1 mg/kg respectively. ACP-204 was orally active. ACP-204 suppressed MK-801 induced locomotor activity in mice with an MED of 0.1 mg/kg and with comparable efficacy to M100,907. PET studies in Rhesus monkeys using [18F]altanserin demonstrated ACP-204 occupied central 5-HT2A receptors with an ED50 of 0.002 mg/kg and an EC50 of 0.33 ng/ml. ACP-204 inhibited hERG, L-type calcium current (hCav1.2), and sodium NaV1.5 ion channels with an IC50s of 1.8, 3.5, and > 10 µM, respectively, compared to 0.2, 1.2 and 1.2 µM for PIM. Rat PK showed that ACP-204 has high oral bioavailability, moderate clearance (Cl), and a volume of distribution (Vd) considerably higher than total body water. Cynomolgous monkey PK showed that total drug exposure and Cl was similar, but Vd and t1/2 was ~2-fold higher for PIM than for ACP-204, and drug exposure in CSF was ~5-fold higher for ACP-204 than PIM. Using this data, and available clinical data for PIM, the t1/2 of ACP-204 in humans was estimated to be 14.7 to 21.7 hours. Rat 6-month and monkey 9-month toxicity studies demonstrated no evidence of phospholipidosis, with safety margins of 30 to 47-fold estimated by comparing drug exposures at the maximum effective doses in rat DOI HTW to the no-observed-adverse-effect levels (NOAELs) in the toxicity studies.

Conclusions: Compared to PIM, ACP-204 is more potent at 5-HT2A, with significantly lower potency at hERG and other ion channels involved in regulating heart rhythm. PET showed ACP-204 potently occupies central 5-HT2A receptors in Rhesus monkeys. Monkey PK showed substantially more free ACP-204 enters the CNS compared to PIM, and predicts QD dosing for ACP-204 with a shorter t1/2. The nonclinical safety studies provided ample safety margins to projected therapeutic doses. ACP-204 is currently being tested clinically in Alzheimer’s disease psychosis at 30 and 60 mg doses.

Keywords: 5-HT2A and 5-HT2C receptor pharmacology, Pharmacokinetics, Receptor Occupancy

Disclosure: Acadia Pharmaceuticals: Employee (Self).

Leslie Nucifora*, Laura Gomez-Isaza, Brian Lee, Christopher Ross, Russell Margolis, Gwenn Smith, Julia Kofler, Robert Sweet, Frederick Nucifora

Johns Hopkins University School of Medicine, Batimore, Maryland, United States

Background: Alzheimer’s disease (AD) is a neurodegenerative disorder characterized by protein aggregation and cognitive decline. Many patients with AD experience neuropsychiatric symptoms, including hallucinations and delusions, also prominently observed in schizophrenia (SZC), that are associated with poor clinical outcomes, in part due to the lack of safe and effective treatments. Evidence also suggests that female patients with AD have a more insidious onset of illness and greater neuropsychiatric diagnoses. The neurobiological basis of psychotic symptoms in AD is not well understood. Understanding the mechanisms of psychotic symptoms in SCZ, may provide an improved understanding of psychosis in AD (AD + P).

An observation that is a logical focus of study for AD + P comes from our recent work in SCZ, which showed increased protein aggregation in a subset of SCZ patients using postmortem brain. Intriguingly, this data suggests that protein aggregation may be more predominant in female SCZ patients. A preliminary comparison of the insoluble proteins in this SCZ subtype identified in our study with proteins identified in a similar AD proteomic insolubility study demonstrated proteomic overlap. Furthermore, preliminary evidence that correlates aggregation in SCZ to cognitive deficits is shown. These results suggest a common proteomic signature in AD and SCZ that may be related to cognitive deficits and psychosis. Therefore, we hypothesize that the overlapping proteomic changes associated with protein aggregation in schizophrenia and Alzheimer’s disease will implicate genes and proteins, and thus specific cellular processes, associated with neuropsychiatric symptoms and psychosis observed in Alzheimer’s disease. This work may provide evidence that the mechanisms responsible for aggregation in schizophrenia contribute to the neuropsychiatric symptoms that occur in Alzheimer’s disease, particularly those observed in women.

Methods: AD + P and AD - P postmortem brains were obtained from the University of Pittsburgh Medical Campus Alzheimer’s Disease Research Center (ADRC) and matched control brains were obtained by the JHUSOM Alzheimer’s Disease Research Center (ADRC). A brain fractionation protocol was used to obtain a purified insoluble protein pellet from a 10% homogenate brain sample. The insoluble protein fraction has been analyzed by LC-MS/MS on an Orbitrap XL™ Hybrid Ion Trap-Orbitrap Mass Spectrometer (Thermo Scientific) using a label free proteomic approach. Proteome Discoverer 2.2. Beta versions has been used to perform all data processing and finalized in 2.20.388. SequestHT was used for the identification of peptides and Percolator has been used to assess the false discovery rate (FDR) at the peptide spectral match (PSM) level and filtered to a 1% estimate cutoff. MaxQuant, MetaMOrpheus, MSFragger and Proteome Discoverer was used to analyze the raw proteomic data to maximize the number of proteins identified and determine significant protein enrichment. Gene Ontology Enrichment Analysis, Ingenuity Pathway Analysis (Ingenuity Systems; Redwood City, CA, USA), and weighted correlation network analysis has been used to analyze the enriched protein data to determine most relevant proteins and biological processes, and networks associated with protein enrichment.

Results: We have obtained promising preliminary data through a comparative proteomic investigation of the insoluble proteins in a SCZ subtype, characterized by protein aggregation, with proteins identified in a previous similar proteomic insolubility study in AD, that suggests overlapping proteins and pathways between the two neuropsychiatric disorders. These results suggest a common proteomic signature in AD and SCZ.

Conclusions: By elucidating cellular mechanisms of protein aggregation observed in SCZ, a disease characterized by psychosis and cognitive decline, and the wealth of proteomic and genomic data related to AD, the cellular abnormalities leading to psychosis associated with AD can be better understood. Considering that aggregation associated with SCZ is more prevalent in women, this study may provide insight into the underlying mechanisms implicated in the psychosis observed in females with AD. Understanding of etiology and pathogenesis may ultimately result in the identification of new therapeutic targets for the treatment of AD + P.

Keywords: psychosis, Alzheimer’s Disease, Postmortem Brain Tissue, Protein Aggregation, Proteomics

Disclosure: Nothing to disclose.

Sandhya Kortagere*, Sanjay Das, Binod Nepal, Alexandra Hicks, Sergi Ferre

Drexel University Medical College, Philadelphia, Pennsylvania, United States

Background: Loss of dopaminergic neurons in substantia nigra and depletion of dopamine in the striatum are hallmark pathological features of Parkinson’s disease (PD). Dopamine (DA) replacement is still the gold standard therapy for treating the motor symptoms of PD. However, long term treatment with DA agonists such as Levodopa lead to loss of efficacy (tolerance) and painful abnormal involuntary movements called dyskinesia (or Levodopa induced dyskinesia – LID). While several hypotheses have been tested to understand the underlying pathobiology of LID, there are no drugs available to treat LID. Dopamine D3 receptors (D3Rs) have a unique distribution in the basal ganglia on both direct and indirect pathways and are colocalized on D1R and D2R expressing neurons respectively. Because D3Rs can modulate both pathways, it can be considered as therapeutic targets to treat LID in PD.

Methods: All animal experimental procedures were approved by Institutional Animal Care and Use Committee of Drexel University. Adult male and female Sprague Dawley rats (n = 46/sex) were unilaterally injected with 6-OHDA (4μg in 2μl) into the medial forebrain bundle. Two weeks post surgery, animals were treated with Apomorphine (0.5mg/kg; s.c) and recorded for one hour in a cylindrical chamber and those that had > 60 contralateral rotations with minimum ipsilateral rotations were considered as exhibiting PD symptoms (called PD rats). PD rats were then injected with either saline (n = 8/sex), L-dopa (with carbidopa and ascorbic acid) (6mg/kg; 1x/day for 3 weeks; n = 40/sex) and tested for abnormal involuntary movements (AIMs). The L-dopa treated group was then further grouped into Saline (n = 8/sex) or S-SK609 (n = 8/dose/sex; 0.5, 1, 2, 4 mg/kg) and treated for 3 weeks. Post behavioral assessment, rats were euthanized, and the brain tissue was rapidly dissected, and the striatum was isolated and stored in RNA later for biochemical experiments. One- and two-way ANOVA with repeated measures was used to analyze the data.

Results: Results from the study showed that S-SK609 alone did not produce any AIMs at any of the doses tested in male or female PD rats. L-dopa (6mg/kg) produced significantly higher AIMs compared to saline treated PD rats. S-SK609 significantly reduced AIMs produced by L-dopa at 2 (*p = 0.02) and 4mg/kg dose (**p = 0.008) in male rats and at 4mg/kg dose (***P = 0.0005) in female rats. As hypothesized, the expression of D1R, D3R and β-arrestin2 were significantly upregulated in L-dopa treated animals and was reduced with S-SK609 treatment. Further, both GluA1 (*p = 0.017) and GluA2 (**p = 0.0081) subunit expression was reduced in S-SK609 treated animals which correlated with the downstream GSK-3βpS9 levels. RNA sequencing of striatal tissue from S-SK609 and L-dopa treated groups showed changes in gene expression levels of G-protein signaling proteins and interesting effect of sex in the differentially expressed gene set.

Conclusions: S-SK609 is a G-protein biased signaling agonist of D3R, which has unique properties that it does not recruit β-arrestin for signaling and hence does not lead to long term desensitization of D3Rs. Instead, D3Rs activated by S-SK609 undergo time dependent internalization. Based on these atypical pharmacological properties of S-SK609, we hypothesized that S-SK609 in combination with L-dopa can reduce the hyperactivation of the direct pathway neurons to reduce AIMs in LID. Results from the study demonstrate that S-SK609 under chronic treatment conditions does not produce dyskinesia side effects on its own and when co-administered with L-dopa blocks AIMs associated with LID. Biochemical analysis of striatal tissue post treatment shows S-SK609 normalized the expression of D1R and D3R receptors and their downstream signaling partners. These preliminary results are high encouraging to develop S-SK609 as a therapeutic for LID in PD.

Keywords: Dopamine (D2, D3) receptors, L-dopa induced dyskinesia, Parkinson’s disease, Biased signaling, RNA Sequencing

Disclosure: PolyCore Therapeutics: Founder (Self).

Mala Ananth*, Ronald Kim, David Talmage, Lorna Role

National Institutes of Health, Bethesda, Maryland, United States

Background: It is becoming increasingly clear that the complex molecular processes that precede neurodegenerative disease occur decades before diagnosis. As such, identifying factors that signal the onset of these processes is of the utmost importance. Research finds that structural and functional changes observed in the entorhinal cortex (EC) are amongst the earliest signs of cognitive aging and neurodegenerative disease. In fact, EC-related deficits in humans have been found to precede, and even predict, future cognitive impairment. While previously regarded as a late stage phenomenon in pathological aging, we have found that alterations to cholinergic synaptic integrity of the EC, and corresponding loss of EC function, are in fact an early element of EC dysfunction in a mouse model with genetic susceptibility to AD. Whether this molecular process can be identified in natural aging remains unclear. Here we find that mice at 12-months of age (middle age for a mouse) already display early signs of cholinergic synaptic dysfunction in the EC. We explore the consequences of this dysfunction in these older mice and identify the unique gene expression profiles of those earliest affected (vulnerable) cholinergic neurons.

Methods: First, to assess structural changes to the cholinergic EC, we used confocal imaging to compare cholinergic terminal density in the EC in young (3 month) and older (12 month) animals. Next, using the displaced object recognition behavioral assay, we evaluated the consequence of age on EC function. To better understand mechanisms underlying age-related dysfunction of the cholinergic EC, we used activity dependent markers to determine whether engagement of EC-projecting cholinergic neurons during behavioral tasks was impaired in aged animals. Finally, we utilize single nucleus RNAseq to evaluate gene expression profiles of these older cholinergic neurons using a genetic labeling strategy coupled with fluorescence assisted cell sorting (FACS) to isolate cholinergic nuclei.

Results: We found that, already at 12-months of age, cholinergic terminal density was lower in the EC as compared to young mice. Notably, this reduction was not seen in other cortical areas. Using the DOR task as a quantifiable measure of EC function, we found that these older mice already display a loss of preference for the displaced object during the DOR task unlike their younger counterparts. This poor performance was related to impaired engagement of cholinergic neurons and EC neurons during the DOR task, as assessed by cFos immunoreactivity. Using single nucleus RNAseq, we evaluated the gene expression profiles of cholinergic neurons in 12-month aged animals. We found specific clusters of cholinergic neurons that were under-represented in aged animals thereby identifying the gene expression profiles of putative ‘vulnerable’ cholinergic neurons.

Conclusions: Our findings demonstrate that impaired activation of cholinergic neurons and compromised cholinergic input to the EC occur far sooner than previously expected, and onset in middle age, as early as 12-months, in mice! These processes are intimately related to the decline of EC function, regarded as an early sign of cognitive aging. Using single nucleus RNAseq we gained insight into the effects of age on the cholinergic neuron transcriptome, identifying the molecular determinants of vulnerable cholinergic neurons. Taken together, this dataset provides an essential foundation for future studies to be able to target vulnerable cholinergic neuron populations and support cognition throughout lifespan.

Keywords: Cholinergic system, Aging, entorhinal cortex, Cognition, RNAseq

Disclosure: Nothing to disclose.

David Horovitz, Melanie Tieman, Anna Sikora, Michael Pitre, Joseph McQuail*

University of South Carolina, Columbia, South Carolina, United States

Background: N-methyl-D-aspartate receptors (NMDARs) play a crucial role in learning and memory and are implicated in age-related cognitive decline and Alzheimer’s disease (AD) pathology. Synaptic NMDAR activation supports synaptic plasticity and neuroprotection via ERK signaling, while extra-synaptic activation contributes to excitotoxicity and neuronal damage. Aging is associated with a decline in D-serine levels, an essential synaptic NMDAR co-agonist, which may compromise synaptic function and contribute to cognitive decline. Recent studies have highlighted D-amino acid oxidase (DAAO) inhibition as a promising strategy to enhance synaptic NMDAR signaling. DAAO degrades D-serine in the brain; inhibition with specific inhibitors (DAAOIs) prevents D-serine breakdown, thereby enhancing its availability at synaptic sites. DAAOIs have shown efficacy in mitigating AD-related neuropathological changes, including reducing amyloid-beta deposition and neuroinflammation, while preserving synaptic integrity and neuronal function. However, no studies have examined the impact of DAAOIs on normal age-related memory changes.

Methods: Aged (22-month-old) male and female F344 rats were screened for spatial learning deficits using the Morris water maze, compared to young adults (4-month-old). Rats with spatial learning deficits underwent further evaluation using a delayed match-to-place (DMTP) task. Aged rats received varying doses of 3-Methylpyrazole-5-carboxylic acid (3-MPCA), a DAAOI that prevents D-serine degradation. Post-treatment, rats were re-tested on the DMTP task, and memory performance and ERK phosphorylation in the prefrontal cortex and hippocampus were assessed.

Results: 3-MPCA treatment dose-dependently enhanced memory performance in aged rats and increased ERK phosphorylation in memory-associated brain regions, indicating enhanced synaptic NMDAR activity. These findings support the hypothesis that augmenting synaptic D-serine availability via DAAO inhibition rescues age-related memory deficits by bolstering neuroplasticity-associated NMDAR signaling. Additionally, 3-MPCA-treated rats showed a significant positive correlation in ERK phosphorylation levels between the prefrontal cortex and hippocampus, suggesting optimized brain network function.

Conclusions: Prolonging D-serine availability via DAAO inhibition emerges as a promising therapeutic approach to mitigate age-related memory decline and potentially delay the onset or progression of neurodegenerative disorders like Alzheimer’s disease. Future studies should investigate DAAOI effects on D-serine levels and explore biomarkers of drug efficacy to facilitate the development of preventative applications against memory loss and dementia.

Keywords: NMDA Receptor, D-Amino Acid Oxidase Inhibitor, D-serine, Memory, Biology of Aging

Disclosure: Nothing to disclose.

Dara Ghahremani, Prabha Siddarth, Thomas Nichols, Susan Bookheimer, Helen Lavretsky*

UCLA Semel Institute for Neuroscience and Human Behavior, Los Angeles, California, United States

Background: Physical activity can impact cognition and brain structure in older adults, potentially acting as a buffer against cognitive decline. We aimed to examine the relationship between physical activity, cognition, and brain volume in a cross-sectional study of a large sample of older adults (N = 1107; mean age: 62 (SD = 15), 36-98 years) from the Human Connectome Project – Aging (HCP-A) and the Adult Aging Brain Connectome (AABC) studies. We primarily focused on subcortical brain structures important for learning and memory, including subfields of the hippocampus.

Methods: Participants completed the International Physical Activity Questionnaire (IPAQ) and provided high resolution whole brain structural MRIs [whole brain T1-weighted MPRAGE (voxel size: 0.8mm isotropic) and hippocampus-focused T2-weighed Turbo Spin Echo (voxel size: 0.3906 x 0.3906 x 2 mm) images] as part of the HCP-A and AABC projects. Brain volume and thickness was computed using Freesurfer (v. 7.4.1). Physical activity was measured in units of the metabolic equivalent of task (MET). We used general linear models to examine relationships between IPAQ and regional volumes, with age, sex, ethnicity, study site, and intra-cranial volume as covariates. Age was treated as a factor with 2 levels (< 65 and ≥65 years). Measures of cognition included short-term memory recall (using RAVLT) and loadings from a factor analysis of several cognitive measures resulting in 3 factors: “memory”, “fluid IQ” (e.g., executive function), and “crystalized IQ”.

Results: We found significant age-by-physical activity interaction effects on both cognition and brain volume. Specifically, physical activity was positively associated with immediate recall, “memory”, “fluidIQ”, and hippocampal volume in adults ≥65 years, but not in adults < 65 years. Probing hippocampal subfields determined that the effect was driven mostly by volume in the anterior portions (right dentate gyrus/CA3 head). Cortical regions showing this interaction included the occipital pole and inferior frontal gyrus.

Conclusions: In older adults only, physical activity has a significant impact on memory, executive function, and brain structure, with greater physical activity associated with better cognition and larger hippocampal volume (primarily in anterior hippocampal subfields, which are associated with encoding of higher-order information) and thickness in specific cortical regions. Results from this large sample underscore the importance of physical activity for brain health in older adults as a potential buffer against cognitive decline.

Keywords: Aging; Cognition; Stress; Acetylcholinesterase; Splice Variants, Human Neuroimaging, Cognitive performance, hippocampus subfields, Physical exercise

Disclosure: Nothing to disclose.

Hari Deva Muthu Balasubramanian*, Jesu Arockiaraj

University of Pittsburgh, Pittsburgh, Pennsylvania, United States

Background: Alzheimer’s disease (AD) presents a complex interplay of neuropathological features, including the accumulation of beta-amyloid plaques, neurofibrillary tangles, oxidative stress, and synaptic dysfunction. These factors contribute to the progressive cognitive decline seen in patients. Current therapeutic strategies, primarily acetylcholinesterase (AChE) inhibitors, target individual aspects of this multifaceted disease, offering limited efficacy and primarily symptomatic relief. This study, however, proposes a unique approach that simultaneously targets multiple pathological pathways, potentially enhancing treatment efficacy. Our hypothesis is centered on a novel hydroxyl-chalcone molecule, designed with structure-guided insights indicating that specific molecular features—such as hydroxyl groups and a furan moiety—enhance biological activity, including blood-brain barrier permeability, AChE inhibition, and antioxidant properties. This multi-targeted approach aims to comprehensively reduce amyloid burden, mitigate oxidative damage, and improve neuronal survival, offering a potential breakthrough in therapeutic strategies for AD.

Methods: The in-house library of chalcone molecules was virtually screened against acetylcholinesterase (AChE) inhibition, and hydroxyl-chalcone was synthesized. This was followed by in vivo testing in a scopolamine-induced zebrafish model of AD. Zebrafish (N = 96) of both sexes, aged 6-8 months, were randomly divided into control and treated groups to assess the in vivo efficacy of the hydroxyl-chalcone. The hydroxyl-chalcone molecule was administered intraperitoneally to the zebrafish, who were then subjected to behavioral assessments for cognitive function, including memory retention and anxiety-like behavior, using novel tank and maze-based tests. Biochemical assays were performed to quantify oxidative stress and AChE activity, while histological staining was used to observe amyloid plaque formation and neuronal integrity. Statistical analysis involved ANOVA and post-hoc tests, with a significance threshold set at P < 0.05 and effect sizes calculated to determine the practical significance of the findings. Data were represented as mean ± standard deviation (SD).

Results: The hydroxyl-chalcone demonstrated potent AChE inhibition (IC50 = 5.7 ± 0.13 µM; DF = 18, R2 = 0.982), suggesting a high potential for reducing acetylcholine breakdown. Treated zebrafish demonstrated a significant improvement in memory retention. They decreased anxiety-like behavior, with significant improvements in latency to reach the target zone (F [5, 42] = 2.719, P < 0.0001) and time spent in the novel object zone (F [5, 42] = 1.035, P < 0.0001). The hydroxyl-chalcone also markedly reduced oxidative stress as evidenced by decreased levels of malondialdehyde (F [5, 18] = 1.432, P < 0.0001) and increased antioxidant enzyme activities (F [5, 18] = 0.466, P < 0.0001). Histopathological analysis showed a substantial reduction in amyloid plaques and protection against neuronal degeneration in the brains of treated zebrafish.

Conclusions: Our findings underscore the potential of the hydroxyl-chalcone as a comprehensive therapeutic agent for Alzheimer’s disease. This novel molecule significantly mitigates neurodegenerative processes and enhances cognitive functions in a zebrafish model. The study supports the molecule’s ability to address multiple AD-related pathological targets, suggesting a promising avenue for therapeutic development. Future studies should focus on elucidating the molecular mechanisms underlying its effects and validating its efficacy in mammalian models. These investigations could pave the way for clinical trials, aiming to develop a more effective, multi-targeted treatment strategy for AD.

Keywords: neurodegeneration, Alzheimer’s dementia, Cognitive Enhancement, zebrafish

Disclosure: Nothing to disclose.

Jason Russell, Alexander Conley, Brian Boyd, Rachel Schlossberg, Allison Stranick, Adam Rosenberg, Dann Martin Martin, Lealani Acosta, Michael Rafii, Sepideh Shokouhi, Paul Newhouse*

Vanderbilt University Medical Center, Nashville, Tennessee, United States

Background: Life expectancy for adults with Down Syndrome (DS) has dramatically increased, with many individuals now living to the sixth decade of life. The amyloid precursor protein is found on the triplicated chromosome 21, producing a profoundly increased lifetime risk for the development of Alzheimer’s disease (AD) in adults with DS. Declining cholinergic integrity plays a role in the cognitive deficits observed in late-onset AD. Understanding the fate of the cholinergic system in adults with DS as AD pathology develops will provide important data to understand the potential utility of cholinergic targeting compounds in adults with DS and how disease-modifying therapies may impact cholinergic functioning. [18F]Fluoroethoxy-benzovesamicol ([18F]-FEOBV) is a ligand that binds the vesicular acetylcholine transporter, labeling cholinergic terminals. The present study assesses the relationship between cholinergic basal forebrain volume or [18F]-FEOBV uptake, AD pathology, and cognitive performance in adults with DS.

Methods: 188 individuals from the Alzheimer’s Biomarker Consortium-Down Syndrome (ABC-DS) cohort (NIA U01AG051406; U01AG051412; U19AG068054) were assessed cross-sectionally at baseline. Basal forebrain volumes were calculated from T1 MRIs using the ScLimbic cholinergic basal forebrain, FreeSurfer, and DnSeg Nucleus Basalis of Meynert pipelines. Individuals were recruited as part of the Trial Ready Cohort-Down Syndrome (TRC-DS; NIA R33AG066543) sub-study (25-55 years old) or de novo (18-24 years old (R21AG075643-01; AARG-21-850839) where they underwent MRI imaging (n = 16), [18F]-FEOBV cholinergic PET imaging (n = 16), [11C]-PiB amyloid PET imaging (n = 15) and [18F]-MK6240 tau PET imaging (n = 8). [18F]-FEOBV and [18F]-MK6240 PET images were RBV partial volume corrected with PetSurfer, [18F]-FEOBV images were normalized to the eroded supratentorial white matter in a voxelwise manner, then smoothed and transformed to MNI-152 space using ANTsPy. For [18F]-MK4620, SUVRs in Braak stage 1/2, 3/4, and 5/6 regions were calculated in PetSurfer using the eroded cerebellar grey matter as a reference region. Centiloids (global amyloid accumulation) were calculated from the [11C]-PiB PET images using the standard pipeline in PMOD Neuro. A study-specific grey matter mask was applied, and generalized linear models were performed in nilearn (python) with voxelwise [18F]-FEOBV uptake as the dependent variable and age, centiloids, and SUVR in Braak stage regions as independent variables, p < 0.005 with a cluster size of 50 was used to determine significance. Participants from both studies underwent numerous cognitive assessments. Linear regressions were performed between cognitive performance and regional [18F]-FEOBV SUVR or basal forebrain volume correcting for intellectual disability.

Results: In the ABC-DS cohort, age and global amyloid accumulation were negatively associated with Ch1-Ch4 (p < 0.005) and Ch4 (p < 0.05) cholinergic basal forebrain volumes. The Down Syndrome Mental State Exam (DSMSE) total score was positively associated with basal forebrain volume (p < 0.05), and total free recall on the cued recall task was positively associated with left basal forebrain volume (p = 0.0027). In adults with DS, increasing age in the TRC-DS sub-study was associated with decreased [18F]-FEOBV uptake in the anterior cingulate, parietal, limbic, and temporal cortices. Amyloid centiloid score and Tau deposition in Braak stage 1/2 were associated with decreased [18F]-FEOBV uptake in the frontal cortex, while Tau deposition in Braak stage 3/4 was associated with reduced [18F]-FEOBV uptake in the frontal and parietal cortices and thalamus. Braak stage 5/6 SUVRs were not assessed due to limited tau deposition in these regions in this population. [18F]-FEOBV uptake in the right amygdala, right rostral anterior cingulate cortices, and left transverse temporal cortex were positively associated with DSMSE total score (p < 0.05). In the cued recall task, [18F]-FEOBV uptake in the left cuneus and right entorhinal cortexes and the isthmus of the right cingulate cortex was negatively associated with total free recall (p < 0.05).

Conclusions: In the ABC-DS cohort, increased age and amyloid accumulation were associated with decreased cholinergic integrity as measured by basal forebrain volume, and decreased cholinergic integrity predicted reduced DSMSE and cued recall task performance. Assessing regional [18F]-FEOBV uptake can indicate where cholinergic terminal loss is occurring. In the TRC-DS sub-study, age-associated loss was observed in the cingulate, parietal, limbic, and temporal cortices, while in contrast, amyloid and tau-related cholinergic decreases are found in the frontal and parietal cortices and thalamus. Decreased regional [18F]-FEOBV uptake predicted reduced performance on the DSMSE but increased [18F]-FEOBV uptake predicted decreased free recall in the cued recall task. These preliminary results suggest that cholinergic systems are vulnerable to AD pathology in adults with DS and that AD pathology-associated cholinergic loss in this population may differ from aging-related effects.

Keywords: Acetylcholine, Alzheimer’s Disease, Down Syndrome, PET Imaging

Disclosure: Eisai: Advisory Board (Self). Phillip Morris International: Advisory Board (Self).

Christian Ramos Jiménez*, Kankana N Aji, Chris Hung-Hsiao, Robert Hopewell, Carolin Jaworski, Gassan Massarweh, Jean-Paul Soucy, Vladimir Grouza, Daniel Chartrand, Ralf Schirrmacher, Alexey Kostikov, Alexander Thiel, Pablo Rusjan

Douglas Mental Health Research Institute, Montreal, Canada

Background: Tropomyosin receptor kinases (Trks) are transmembrane proteins belong to the receptor tyrosine kinases (RTKs) family. TrkA, TrkB and TrkC are the major isoforms encoded by NTRK1, NTRK2 and NTRK3 genes respectively and are involved in development neurobiology, growth, neuronal survival, and cellular differentiation. Evidence has demonstrated alterations in the Trk signaling results in several neurodegenerative and psychiatric diseases. [18F]TRACK, an optimized fluorine-18-labeled has been evaluated in the first-in-human study, showing high selectivity for the TrkB/C and a moderate brain uptake. In this work we present the first PET evaluation of [18F]TRACK in healthy human brain.

Methods: Six healthy volunteers (male=2, female=4; age: 20-61 years) were scanned in a High Resolution Research saTomograph (HRRT) for 90min after an I.V. bolus injection of [18F]TRACK (0.19-1.78ug, 2.25-6.4mCi). Arterial blood mples, collected discretely and continuously during the scan, were used to create a curve of unmetabolized parent compound in plasma. One and 2 tissue compartment model (2TCM) were evaluated for a subset of regions of interest (ROIs): amygdala, putamen, thalamus, pons, cerebellum, striatum, white matter (WM), temporal and frontal cortex.

Results: Following an intravenous bolus injection of [18F]TRACK, activity was distributed in all the brain tissues. Time-activity curves (TACs) peaked within the first 6 min with standardized uptake values ranging from 1.96 ± 0.54 (putamen) to 1.14 ± 0.27 (pons). The activity washout was faster in the cortical and subcortical regions and slower in the pons. Analysis of plasma samples using High performance liquid chromatography (HPLC) showed a moderate metabolism rate (the parent compound accounts for ~60% for activity in plasma after 90min).

The radiometabolites were more polar than the parent compound (logD value: 4.23 ± 0.1). On the subset of ROIs studied, the 2TCM fits the TACs better than the 1TCM based on the Akaike Information Criterium (AIC) (Fig 1). Regional total distribution volume (VT) presented good identifiability (average standard error ~3%). The tissue/plasma ratio reaches a plateau at ca. 60min for the gray matter ROIs, but not for the white matter areas. The regional gray matter rank order based on VT (mL/cm3) was thalamus (2.58 ± 0.33), putamen (2.53 ± 0.31), cerebellum (2.33 ± 0.33), temporal cortex (2.31 ± 0.26), striatum (2.18 ± 0.26), frontal cortex (2.33 ± 0.26), and amygdala (2.17 ± 0.24). VT provided by the Logan graphical method (LGM) correlates very well with those estimated using 2TCM (r2 = 0.9).

Upon intravenous bolus injection, [18F]TRACK crosses the blood-brain barrier (BBB) with a low rate of brain entry (K1 < 0.1 ml/cm3/min). The TACs peaked within the first 6 minutes post-injection, reaching activities between 1 and 2 SUV depending on the brain region. Analysis of plasma samples using HPLC revealed a moderate metabolism rate, with the parent compound accounting for approximately 60% of activity in plasma after 90 minutes. Radiometabolites are more polar than the parent compound, therefore less likely to cross the BBB. The tissue/plasma ratio reached a plateau around 60 minutes for gray matter regions but not for white matter areas (not shown).

The regional rank order based on VT values revealed differences in [18F]TRACK distribution among various brain regions. White matter (WM) show greater VT than gray matter ROIs. The thalamus showed the highest VT within the gray matter ROIs. The slow washout of activity in WM and high VT could be related to nonspecific binding, as the radioligand has high lipophilicity. VT values exhibited good identifiability. VT estimates with the LGM correlated well with those from the 2TCM.

Conclusions: The two-tissue compartment model (2TCM) provided a better fit to the TACs compared to the 1TCM, as indicated by the Akaike Information Criterion (AIC) in the cerebellum, frontal, temporal, thalamus, pons, and WM, while not in amygdala, caudate, striatum, and putamen. LGM could be a better alternative to quantify this radioligand as it is independent from the number of compartments.

Keywords: F-18 PET Imaging, Neurodegenerative Disease, Neuropsychiatric Disorders, Pharmacokinetics

Disclosure: Nothing to disclose.

Louis Kolling*, Michael Chimenti, Ruixiang Wang, Catherine Marcinkiewcz

The University of Iowa, Iowa City, Iowa, United States

Background: Previously, we found that transgenic mice expressing human tau (htau mice) present with depressive and anxiety-like behaviors consistent with prodromal symptoms of early AD. In 5-HT (serotonergic) neurons of the dorsal raphe nucleus (DRN), we found evidence of tau pathology that was not uniform in presentation. As such, we sought to identify any underlying genetic components associated with susceptibility or resistance of DRN neurons to AD pathology.

Methods: We used the Visium Spatial Gene Expression platform to identify transcriptomic changes across the DRN in htau vs C57BL6/J mice aged 16 weeks (4 male each), and to identify distinct spatial subregions of the DRN. We then used RNAscope and immunofluorescence to validate findings of primary interest in htau +/-, htau -/-, C57, and rTG4510 mice (5 male each).

Results: We find that the DRN of htau mice differentially expresses AD-related genes, including those related to kinase binding and ion channel activity. We further find that computational subclustering of the DRN is consistent with previous circuitry-driven characterizations, allowing for spatial bounding of distinct subregions within the DRN. Of these, we find the dorsolateral DRN is preferentially impacted by 5-HT neuron loss (*p = 0.0248) and development of tau pathology (**p = 0.0070), which coincide with increased expression of the long noncoding RNA Map2k3os (log2FC = 1.70, P = 0.0286).

Conclusions: Overall, we find that tau-mediated dysfunction of the DRN may be mediated by the preferential susceptibility of the dorsolateral DRN to tau pathology. Differential expression of map2k3os lncRNA may affect tau phosphorylation in this model, and warrants further investigation.

Keywords: tau, spatial transcriptiomics, intracranial EEG, Preclinical Alzheimer’s Disease

Disclosure: Nothing to disclose.

Osama Abulseoud*, Asmaa Yehia

Mayo Clinic Arizona, Scottsdale, Arizona, United States

Background: Post-COVID-19 syndrome presents an untreatable public health crisis of great magnitude. In order to achieve effective treatment options, we need to understand the pathophysiology of this complex medical condition. Current animal models of post-COVID syndrome that utilize the SARS-CoV-2 virus are limited by the high fatality rates and the absence of longitudinal data. Similarly, the SARS-CoV-2 spike protein models are also limited by the use of the direct brain inoculation route of administration rather than the natural intranasal (IN) route of infection. This study aimed to examine whether IN SARA-CoV-2 spike protein administration to K18-hACE2 transgenic mice could model post-COVID longitudinal phenotypical and pathological changes.

Methods: Male K18-hACE2 transgenic mice received IN SARS-CoV-2 spike protein [S1, (600 µg/kg) vs. saline: total n = 72] at 8-weeks of age. Behavioral tasks [Open Field (OF), Novel Object Recognition (NOR), Forced Swim Test (FST), and cued Fear Conditioning (FC)] were performed at 2-, 6-, and 12-weeks post S1 (n = 24/time point) followed by euthanasia and brain collection for histopathological examination [right hemisphere for 5µ midline sagittal sections and H and E staining (n = 24/time point)]. Left hemisphere was dissected to collect hippocampus (n = 8/time point) and prefrontal cortex (n = 8/time point) for transmission electron microscopy examination. Other brain tissues [cerebellum, prefrontal cortex, striatum and hippocampus, (n = 8/time point) were collected and stored at -80 °C for future molecular examination.

Results: Compared to IN saline control, IN S1 administration was not associated with changes in body weight or locomotor activity (OF) at any time point. Mice spent significantly less time investigating the novel object at 2- and 6-wk (P = 0.03 and P = 0.02 respectively). We observed marked hypervigilance demonstrated as increased freezing time in the cued FC at 2- and 6-wk (P = 0.04, and P = 0.02 respectively), and mal-adaptive response to stress as detected in FST by decreased immobility time (P = 0.007 both). Upon dissection, we observed few cases of cortical and brain stem hemorrhage at 12-wks. In addition, we found leptomeningeal and perivascular inflammatory infiltrates, tissue edema and neuronal clumping. At the later timepoints, we found granulomatous necrotizing vasculitis with evident endothelial hyperplasia, swollen pericytes and perivascular edema. Astrocytic histopathological examination revealed dense astrocytic infiltration in the subventricular zone and frequent Alzheimer type II astrocytes. Ultrastructural examination of astrocytes showed marked electron-lucent cytoplasm and nucleus, loss of organelles, extensive vacuolation and swelling. We found dark microglia with electron dense cytoplasm and heavy chromatin condensation with extensive ramifications surrounding large number of synapses and frequent lipofuscin deposition. We observed neuronal ER distress shown as ER whorls and neuronal degenerative changes in the form of crenated nuclear and plasma membrane with filament-like processes, extensively electron-dense cytoplasm, and dilated ER-Golgi network, with several patterns of mitochondrial changes include fusion, swelling with disrupted cristae, and ruptured mitochondrial membrane. Some neurons showed nuclear abnormalities in the form of multilobed nuclei, binucleation, and ruffled and disintegrated nuclear membranes. Furthermore, oligodendrocytes exhibited enlarged nuclei with chromatin condensation and a thin cytoplasmic rim with few organelles. Myelinated axons showed swollen mitochondria, loss of cytoplasm, vacuolation and disrupted myelin layers. Sac-like structures filled with large numbers of mitochondria and lysosomes where frequently observed suggestive of dystrophic neurites. Dendrites showed loss of cytoplasm, and vacuolation filled with debris-like membranous structures.

Conclusions: Our results suggest that the spike protein alone is enough to create persistent phenotypic and pathological changes reminiscent of post-COVID syndrome if administered through the natural route of viral infection. Furthermore, the spike protein-induced cognitive and neuropathological changes are suggestive of accelerated aging and neurodegeneration. Further studies are urgently needed to investigate the underlying molecular mechanisms and identify potential therapeutic targets.

Keywords: COVID-19, Animal Models, neurodegeneration, accelerated aging

Disclosure: Nothing to disclose.

Ana Paula Costa*, Sang Joon Joon, Chang-Le Chen, Jinghang Li, Noah Schweitzer, Chang Hyung Hong, Hyun Woong Roh, Yong Hyuk Cho, Bumhee Park, Na-Rae Kim, Jin Wook Choi, So Young Moon, Seong Hye Choi, Akiko Mizuno, Shaolin Yang, Howard J. Aizenstein, Minjie Wu

University of Pittsburgh, Pittsburgh, Pennsylvania, United States

Background: High-density lipoprotein (HDL) is recognized for its protective role in cardiovascular health and is increasingly being studied for its potential to contribute to maintaining cerebrovascular integrity—a critical factor in preventing cognitive decline and neurodegeneration. Studies reporting HDL may help prevent the development of white matter hyperintensities (WMH) by enhancing endothelial function and reducing inflammation, thereby preserving white matter integrity and reducing the risk of cognitive decline. Additionally, the glymphatic system, a waste clearance pathway in the brain, is vital for removing metabolic waste products, including amyloid-beta, which accumulates in neurodegenerative conditions like Alzheimer’s disease (AD). Diffusion MRI-based techniques such as DTI-ALPS can quantify perivascular interstitial fluid dynamics, yet their links with other early AD markers remain underexplored. Therefore, identifying factors that preserve or enhance cerebrovascular integrity, such as lipid profiles, could be essential in mitigating the impact of cerebrovascular damage on cognitive outcomes in these vulnerable populations.

Methods: The analysis is based on a large South Korea cohort of predominantly cognitively impaired participants, as part of the ongoing Biobank Innovations for Chronic Cerebrovascular Disease With ALZheimer’s Disease Study (BICWALZS) and the Centre for Convergence Research of Neurological Disorders. At time of this analysis, BICWALZS had recruited 713 participants from 6 sites. We included participants recruited at memory clinics and underwent a 3T MRI scan baseline (N = 478). We used a pretrained deep learning model to quantify white matter hyperintensities (WMH) on T2-weighted fluid-attenuated inversion recovery (T2w FLAIR) images and estimated the DTI-ALPS index from diffusion tensor imaging (DTI) data. An algorithm was developed to refine DTI-ALPS by targeting the periventricular space as the target area and calculating transverse tensor ratios as the primary measure. Linear regression analysis was applied to investigate the association between plasma cholesterol levels (HDL, low density lipoprotein (LDL), and Total Cholesterol (TC); mg/dL) and cerebrovascular integrity (e.g.: augmented DTI-ALPS and WMH).

Results: The study utilized data from 478 participants (152 males and 326 females) with a mean age 72.7 years and 73.5 for males and females, respectively. Among the participants, 120 males and 268 females were diagnosed with SCD or MCI, while 32 males and 58 females were diagnosed with dementia. The linear regression analyses revealed a significant negative association between WMH and HDL levels (p = 0.005), but not significant associations between WMH and LDL (p = 0.845) or TC (p = 0.556) levels. The results indicating that higher HDL levels are associated with lower WMH. This relationship was robust even after adjusting for sex, age, and BMI. In the unadjusted model, HDL levels was a significant predictor of augmented DTI-ALPS (p = 0.003), suggesting that higher HDL levels are associated with changes in diffusion tensor imaging metrics of the brain. When controlling for age and sex, this association remained marginally significant (p = 0.055). Stratifying by sex indicated that the effect of HDL on augmented DTI-ALPS was primarily driven by females (p = 0.042). No significant associations were found between augmented DTI-ALPS and TC or LDL levels (p = 0.286 and p = 0.580, respectively).

Conclusions: The results revealed a significant negative association between HDL levels and WMH, with higher HDL linked to lower WMH. Additionally, after adjusted for age and sex, higher HDL showed a marginally significant association with reduced augmented DTI-ALPS. These findings suggest that plasma HDL levels may influence brain health through various mechanisms, including effects on neuroinflammation, cholesterol metabolism, and oxidative stress. Further research is needed to clarify these biological pathways and explore lipid-targeted interventions for potential therapeutic benefits.

Keywords: High-density lipoprotein, Diffusion tensor imaging, DTI-ALPS, White matter hyperintensities, Cerebrovascular integrity

Disclosure: Nothing to disclose.

Fernanda Mascarenhas, Ana Carolina Santos, Tamiris Sabrina Rodrigues, Serena Malta, Lucas Bernardes, Matheus Henrique Silva, Renata Graciele Zanon, Carlos Ueira-Vieira, Ana Mendes-Silva*

University of Saskatchewan College of Medicine, Saskatoon, Canada

Background: Alzheimer’s disease (AD) is a progressive neurodegenerative disorder and the most common cause of dementia in older adults. The amyloidogenic hypothesis, the leading theory explaining this condition, describes the extracellular buildup of amyloid beta (Aβ) peptides. Kefir, a fermented dairy product, has gained prominence in the literature due to its health benefits, including modulating the intestinal microbiota. This modulation is particularly relevant as the gut-brain axis plays a critical role in influencing neuronal survival and inflammatory processes. In this study, we evaluate the effect of four kefir peptides with high homology to the human Aβ 1-42 peptide domain on neuron viability using an in vitro model of Aβ aggregation.

Methods: Four best candidate kefir peptides (PW, M11, M20 and M25) based on their Aβ42-like similitude score (> 80%) were selected from our previous in silico findings and synthesized. The peptides were then diluted in 3 different doses for testing (10 µM, 100 µM and 1000 µM), filtered (0.22µm) for sterilization, and stored at 4˚C until the moment of treatment. The human neuron cell line SH-SY5Y was cultured in Dulbecco’s modified Eagle’s medium (DMEM) supplemented with 10% fetal bovine serum and 1% antibiotic. The cells were seeded in a 96-well plate, with 5000 cells per well in 100μL of DMEM, and kept in an incubator for 24 hours for adherence. Following this period, we employed two different experimental approaches to assess the effect of kefir peptides on cellular viability before and after Aβ senile plaque formation. In the first approach, the amyloid beta peptide [886 µM] was diluted to 1 µM, sterilized by filtration (0.22 µm), and then homogenized with kefir peptides (Aβ+treatment) in a 1:1 ratio. This mixture was incubated at 37°C for 48 hours before being added to the cells, which were then incubated for an additional 48 hours, totaling 96 hours. In the second approach, the amyloid beta peptide was added directly to the cells and incubated for 48 hours to promote the formation of senile plaques, also totaling 96 hours. After this incubation period, the cells were treated with the most promising doses of kefir peptides, which were maintained for a further 24 hours. Cell viability assessed using the AlamarBlue assay (Invitrogen®). Cells were incubated at 37°C with 10% (10 µl) resazurin for approximately 3 hours. A positive control group of untreated cells and a negative control group of dead cells in DMEM with 0.2% Triton-X were included in all analyses. All cell culture experiments were performed in three biological and technical replicates. Statistical analysis was conducted using a Student’s t-test to determine differences between cell groups, with all analyses performed using GraphPad Prism v9.0. A p-value of ≤ 0.05 was considered statistically significant.

Results: Among the three peptide concentrations evaluated, 100 µM yielded the most promising results for all peptides. Notably, three out of the four peptides tested significantly increased neuron viability when administered simultaneously with amyloid beta before senile plaque formation, compared to controls. Specifically, the M11 and M20 peptides resulted in an 11% increase in viability (p = 0.004 and p = 0.0007, respectively), while the M25 peptide achieved an 18% increase (p = 0.003). In contrast, the PW peptide did not show a significant effect (p = 0.33). Regarding neuron viability after senile plaque formation, the PW peptide led to a 14% increase (p = 0.03), the M20 peptide achieved a 15% increase (p = 0.01), and the M11 peptide demonstrated a 6% increase (p = 0.09). The M25 peptide resulted in an 11% increase (p = 0.14), though this was not statistically significant.

Conclusions: In conclusion, the kefir peptides M20 and M25 significantly enhanced neuron viability both before and after senile plaque formation, whereas the M11 peptide was effective only before plaque formation and the PW peptide only after. These preliminary findings suggest that kefir peptides may offer neuroprotective benefits in the context of Alzheimer’s disease. Further research is needed to elucidate whether these effects are due to inhibition of amyloid beta peptide aggregation or degradation of existing senile plaques. Future studies should incorporate additional assays, such as the thioflavin T assay for plaque detection, single-cell sequencing for detailed cellular insights, and electrophysiological analyses to assess functional outcomes.

Keywords: Alzheimer’s Disease, beta-amyloid peptide 1-42, kefir peptide, SH-SY5Y, in vitro cellular model

Disclosure: Nothing to disclose.

Ethan Burstein, Brian Raether, Xiaoshu Feng, Bryan Dirks, Mona Darwish, Peter Zhang, Sanjeev Pathak*

ACADIA Pharmaceuticals, Inc., Princeton, New Jersey, United States

Background: ACP-204 is a potent and selective antagonist and inverse agonist of serotonin 2A (5-HT2A), with moderate activity at 5-HT2C receptors and no appreciable interactions at any other targets, including dopaminergic, histaminergic, adrenergic, muscarinic, or other serotonergic receptor subtypes. In comparison to pimavanserin (PIM), an FDA-approved therapy for Parkinson’s disease psychosis, ACP-204 was designed to have an improved pharmacological profile, including a markedly reduced risk for QT prolongation. ACP-204 has been tested and well tolerated in single doses up to 180 mg in healthy adults and multiple doses up to 60 mg and 130 mg in healthy elderly and adult patients, respectively, with no observed QT prolongation and a substantially shorter half-life (ACP-204: 18 hours; PIM: 57 hours). In this study, we compared the 5-HT2A receptor occupancy of ACP-204 with that of PIM using fluorine-18 [18F] altanserin, a radiopharmaceutical with high specificity and selectivity to 5-HT2A receptors. The findings of this study were used to inform the selection of target doses for ongoing clinical trials evaluating the efficacy of ACP-204 in Alzheimer’s disease psychosis.

Methods: A total of 16 healthy adult males (aged 18-55 years) with a body mass index between 18 to 32 kg/m2 were analyzed in this phase 1, open-label, single-center study. Prior to study drug administration, all patients completed a baseline [18F] altanserin positron emission tomography (PET) scan. During all PET imaging sessions, [18F] altanserin was administered intravenously over 2.5 hours, and imaging acquisition was performed over the last 30 minutes. Venous blood samples for tracer metabolite analysis were collected at pre-injection of [18F] altanserin and at approximately 2.0, 2.25, and 2.5 hours after initiating the [18F] altanserin infusion. Patients were given a single oral dose of 20 mg ACP-204 (n = 3), 10 mg ACP-204 (n = 7), or 34 mg PIM (n = 6) in a fasted state and completed 2 postdose [18F] altanserin PET scans and blood samplings at 4 (Tmax), 26, 50, and 74 hours after ACP-204 administration or 6 (Tmax), 54, 168, and 288 hours after PIM administration. Sampling time points were selected based on the Tmax and t½ of each drug and aimed to cover a period of 5 consecutive half-lives. Patients were discharged after completion of all samplings and returned 2 ± 1 day(s) later for end of treatment assessments (EOTs) followed by safety assessments 7 ( ± 2) days after EOTs. The primary endpoint, defined as the percentage of central 5-HT2A receptor occupancy, was calculated regionally using the percentage reduction in binding potential (BPP, BPND) at each postdose PET scan from baseline. Global receptor occupancy was calculated from a Lassen plot. [18F] Altanserin PET scans and blood sample data were quantitatively analyzed to assess the receptor occupancy as a function of ACP-204 or PIM plasma concentrations for each postdose PET scan. Additionally, the relationship between ACP-204 and PIM plasma concentrations and the 5-HT2A receptor occupancy was determined by fitting the global receptor occupancy data to a single specific binding site model with a Hill slope fixed at 1 and a fixed maximum at 100% according to the following equation:

Occupancy= Emax [Cpost-dose / (Cpost-dose + EC50)] where Cpost-dose was the ACP-204 or PIM plasma concentration at the time of imaging. Safety was assessed via the collection and review of treatment-emergent adverse events (TEAEs), physical examination, vital signs, electrocardiograms (ECGs), and laboratory test parameters. Formal statistical calculations of sample size were not performed.

Results: The mean age across cohorts was 40.9 years; most patients were Black/African American (56%) or White (31%). Mean (±SD) global 5-HT2A occupancies following 20 mg of ACP-204 were 103% (±15%) and 102% (±13%) at 4 and 26 hours post-dose, respectively, and following 10 mg of ACP-204, were 95% (±4%), 86% (±3%), 80% (±2%), and 62% (±8%) at 4, 26, 50, and 74 hours post-dose, respectively. Mean (±SD) global 5-HT2A occupancies for PIM were 91% (±4%), 90% (±1%), 86% (±6%), and 50% (±13%) at 6, 54, 168, and 288 hours post-dose, respectively. Based on the occupancies and plasma concentrations of drug at each of the scanning time points, the EC50 values for ACP-204 and PIM were calculated to be 0.37 ng/mL (95% confidence interval [CI]: 0.31 to 0.44 ng/mL) and 0.55 ng/mL (95% CI: 0.40 to 0.75 ng/mL), respectively. Only 1 TEAE was observed across all cohorts but was determined to be unrelated to treatment. No serious TEAEs or clinically meaningful changes in physical examination, vital signs, or ECG results were reported.

Conclusions: Single doses of 10 and 20 mg of ACP-204 were safe and generally well tolerated in healthy adult males. The administration of ACP-204 resulted in marked, dose- and plasma concentration-dependent occupancy of central 5-HT2A receptors. This potency was similar to or slightly > that of PIM and was well predicted by a single specific binding site model.

Keywords: Dementia-related psychosis, 5-HT2A and 5-HT2C receptor pharmacology, Alzheimer’s Disease

Disclosure: ACADIA Pharmaceuticals: Employee (Self).

Ronald Cowan*, Todd Monroe, Michelle Failla

University of Tennessee Health Science Center (UTHSC), Memphis, Tennessee, United States

Background: Given the high prevalence of pain and dementia with increasing age, understanding alterations in the pain experience and the association of pain with the development of cognitive decline in Alzheimer’s Disease (AD) and related dementias (ADRD) is a critical topic in public health. Altered pain processing has been reported across a range of pain self-report, psychophysical, and neurobiological measures in ADRD. Chronic pain is a risk factor for accelerated brain volume loss, cognitive decline, and ADRD. While there are mixed reports and some contradictory findings, overall, people with ADRD seem less sensitive to pain, may have more affective distress to pain, have structural and functional alterations in brain regions mediating pain, show increased brain activation in response to some painful stimuli, and may receive less analgesic medication compared to those without ADRD. The threshold for pain detection and reporting might be lower in ADRD, while affective responses to pain are augmented and some brain imaging measures suggest increased neural activity in brain regions mediating the pain response. Better characterization of pain and risk factors for cognitive decline in ADRD is essential to gain specificity for our understanding of the neurobiological mechanisms linking pain to dementia, and to target interventions that might prevent or slow the pace of cognitive decline. Patients with cancer have high rates of pain that negatively impacts quality of life and there may be differential treatment of pain in people with AD vs. those with cancer. Experimental psychophysical pain testing predicts clinical pain outside the laboratory and can be used to systematically examine pain responses across groups and conditions. We report preliminary results from an ongoing psychophysical study of pain in AD, cancer, and cancer with AD.

Methods: Participants were recruited using various media and from physician practices at three sites. Diagnoses were based on clinical diagnosis and DSM-5 criteria for AD and on clinical diagnoses for cancer. Using quantitative sensory testing psychophysics, we tested self-reported thermal and pressure pain threshold, unpleasantness, and temporal summation of pain. Pain stimuli were delivered to the thenar eminence of the hand using a heating probe (Medoc Q-Sense thermal) controlled by a pseudo-randomized computerized protocol for thermal stimuli, and a hand-held pressure algometer (Medoc Algo-Med computerized pressure algometer). Pain percepts were just noticeable pain (JNP), mild pain (MP), and moderate pain (MoP).

We recorded the temperature at which participants reported each pain percept and the unpleasantness associated with that percept. To control for differences in threshold, we examined the intensity and unpleasantness associated with fixed thermal stimuli at JNP = 34, MP = 39, and MoP=44°C. The study was designed to enroll 264 participants in 3 equal-sized groups (AD, cancer, cancer and AD) and had 82% power to detect the small effect sizes we had observed in prior work. We report interim findings from 66 participants (31 AD; 35 cancer) recruited via media and physician practices at 3 sites. There were insufficient numbers of participants to permit an analysis of people with both cancer and AD. We present results from all 66 participants meeting entry requirements with an effective detectable effect size of d = 0.74 (two-sided Wilcoxon-Mann-Whitney test; unequal group sizes of 31 and 35, respectively; 80% power, type I error alpha = 0.05). Using multivariable regression, we present unadjusted and age-adjusted differences.

Results: There were no statistically significant differences between the groups for the threshold temperature to produce JNP, MP, or MoP or for the unpleasantness ratings at each percept. For fixed thermal temperatures, people with AD rated JNP more intense and unpleasant vs. those with cancer (p = 0.037 unadjusted and 0.049 age-adjusted, and p = 0.016 and 0.021 adjusted, respectively). None of the differences for the intensity and unpleasantness ratings for MP and MoP were statistically significant. For pressure pain, people with AD detected JNP and MP at, respectively, on average 54 and 55 kPA higher pressure, but these differences were not statistically significant after age-adjustment (JNP: p = 0.011 and 0.072 adjusted; MP: p = 0.041 and 0.052 adjusted; MoP not significant, p = 0.147). The differential responses to thermal and pressure pain in people with AD and those with cancer were less pronounced when adjusting for age (average age 75.5 vs 68.9 years in AD and cancer, respectively, p < 0.001).

Conclusions: People with AD and those with cancer have differential responses to experimentally delivered thermal and pressure pain, with implications for quality of life and cognitive decline. Chronic pain is a risk factor for altered brain structure and function that predisposes to cognitive decline. A better characterization of pain mechanisms and risk factors for cognitive decline in AD and related dementias is essential to gain specificity for our understanding of the neurobiological mechanisms linking pain to dementia, and to target interventions that might prevent or slow the pace of cognitive decline.

Keywords: Pain sensitivity, Alzheimer’s Disease, psychophysics, cancer

Disclosure: Galen Mental Health: Advisory Board (Self)

Elise Webber, Lainey Toennies, Grace Stutzmann, Holly Hunsberger*

The Chicago Medical School At Rosalind Franklin University, North Chicago, Illinois, United States

Background: Nearly 7 million people currently live with Alzheimer’s disease (AD), and this number is expected to rise to 13 million by 2050. Despite several FDA-approved treatments for AD, no known cure or drug can stop disease progression. In addition, many studies focused solely on the removal of late-stage AD hallmarks, such as amyloid beta (Aβ) plaques, do not fully ameliorate symptoms such as neuropsychiatric symptoms and cognitive decline, emphasizing the need to develop treatments administered earlier in disease progression. Emerging studies highlight a potential beneficial role of psilocybin in cognition, memory, and behavioral regulation in psychiatric diseases, though little is known about its impact on neurodegenerative diseases such as AD and underlying changes to synaptic plasticity. We hypothesize that a single administration of psilocybin can improve markers of synaptic plasticity and behavioral manifestations in a mouse model of AD.

Methods: 6-8 month-old control and APP/PS1 mice (male = 8-10, female = 5-10) were injected i.p. with psilocybin (3mg/kg) 1 week before either behavior testing or electrophysiology. To test avoidance and exploratory behaviors, mice were administered an open field and zero maze. Fear memory was measured using 3-shock contextual fear conditioning. Freezing behavior was used as a proxy for memory. A two-way ANOVA (drug x Tg) was run within each sex for all behaviors with RMANOVA run across time in the fear conditioning test.

For electrophysiology, acute brain slices (400 um) were prepared in oxygenated artificial cerebrospinal fluid. A stimulating electrode was placed along Schaffer collaterals at CA3, recording electrode placed along Schaffer collaterals at CA1. The Input/Output curve assessed basal synaptic excitability, Paired pulse ratio assessed vesicle release properties upon short-interval paired stimulation, and Long-term potentiation assessed increased excitatory post-synaptic potential increase post-high frequency tetanic stimulation. Whole-cell patch clamp was then performed using Ryanodine receptor evoked calcium release (Caffeine 10mM) and measuring spontaneous miniature excitatory postsynaptic currents (sEPSCs) and inhibitory postsynaptic currents (sIPSCs).

Results: We found increased exploratory behavior in female control and AD mice treated with psilocybin compared to saline in the zero maze. Psilocybin treatment increased distance traveled and mobility (F (1, 12) = 20.76, p = 0.0007). Psilocybin-treated male control mice exhibited increased time spent in the open areas of the zero maze compared to saline-treated mice (F (1, 17) = 5.930, p = 0.0262), indicating less anxiety-like behavior. While female mice, regardless of drug, exhibited similar freezing levels, male control and AD mice treated with psilocybin showed enhanced freezing compared to saline mice (F (1, 17) = 5.798, p = 0.0277). This suggests that fear memory retention is improved with psilocybin treatment in male mice. Electrophysiology revealed decreased synaptic hyperexcitability in the hippocampus of male AD mice as shown by I/O curves and Paired Pulse Ratio. Patch clamp also revealed decreased evoked calcium release (t-Test, p < 0.0001) and increased sIPSCs. There was no significant difference in number, Ri, threshold, amplitude and resting membrane potentials evoked by a single 200pA current step in current clamp mode. Hyperexcitability has been observed in human AD patients and rodent models of AD prior to cognitive decline. Therefore, decreasing this hyperexcitability could be a beneficial treatment option in the early stages of AD.

Conclusions: We observe beneficial effects of psilocybin in AD male mice after only 1 injection. Psilocybin or drugs that mimic psilocybin could represent novel drugs for treating the neuropsychiatric symptoms associated with Alzheimer’s disease and, in turn, cognitive decline. This would give alternative options to those suffering as current medications come with negative side effects and do not work for all patients. However, we did not observe any rescue of memory in female AD mice, which means future studies will need to address different doses and why this is not as effective in female mice. We will also need to determine how long these effects last and whether the psychedelic side effects are necessary.

Keywords: Psychedelics, Alzheimer’s disease, Synaptic plasticity, Hippocampus, Sex differences

Disclosure: Ataraxia LLC: Contracted Research (Self).

Emily Hagan, Pravda Quinones-Labernik, Danielle Preuschl, Charlotte Tesar, Anthony Lange, Lynn Teesch, Serena Gumusoglu, Donna Santillan, Mark Santilan, Sarah Ferri*

University of Iowa Carver College of Medicine, Iowa City, Iowa, United States

Background: Autism Spectrum Disorder (ASD) is a complex, heterogeneous neurodevelopmental disorder that affects 1 in 36 children and 4 times as many males as females. Females are likely underdiagnosed due to sex differences in presentation, the neurobiological mechanisms of which are not known. Testosterone levels in utero are one of the first developmental differences that may confer sex-sensitive vulnerability to deficits and they play a critical role in permanent organization of the male brain. Several complications of pregnancy associated with increased risk of ASD in offspring, including maternal obesity, inc