Addiction Biology

Pavlovian conditioned stimuli can compromise goal-directed behavior, particularly when they conflict with the required instrumental response. Previous research linked an increased impact of such Pavlovian-to-instrumental transfer (PIT) interference to risky drinking and alcohol dependence. However, it remained unclear whether the observed behavioral and neural alterations generalize to the substantially larger population of non-treatment-seeking individuals with predominantly mild-to-moderate alcohol use disorder (AUD) and to other substances like nicotine. We therefore studied 200 participants from the general population who were stratified by AUD and smoking status before completing a single-lever PIT task during functional imaging. PIT interference was measured as the difference in error rates between incongruent PIT trials, characterized by conflicts of Pavlovian and instrumental stimuli, and congruent trials. Associations with AUD and smoking status, as well as dimensional associations of behavioral PIT interference and neural responses to PIT conflicts with substance-related measures, were examined. As expected, participants with AUD showed greater PIT interference and reduced conflict-related activation in the lateral prefrontal, dorsomedial prefrontal, and anterior cingulate cortex than participants without AUD. The magnitude of these alterations also correlated with higher AUD severity and alcohol consumption, while being independent of smoking status and other smoking-related measures. These findings suggest that enhanced PIT interference generalizes to the broader AUD population but may not extend to other substances like nicotine. They further implicate a reduced recruitment of prefrontal control networks during PIT conflicts as a potential mechanism underlying enhanced PIT interference in AUD and its association with clinically relevant outcome measures.

Genetics shape cocaine behavior and abuse liability, but the underlying neurophysiological effects remain understudied. In male C57BL/6J, B6129SF2/J, and DO/J mice, we report conserved and divergent cocaine induced neuroplasticity in the accumbens shell following 5 days of cocaine and 10 to 14 days of abstinence.

Background: Deficits in inhibitory control are considered a central feature of alcohol use disorder (AUD), yet behavioral and neuroimaging findings have been inconsistent. One potential explanation is the marked heterogeneity of AUD, which encompasses multiple symptom dimensions that may differ in their underlying neurocognitive mechanisms. In this study, we investigated whether behavioral and neural markers of inhibitory control are associated with specific DSM-5 AUD symptom domains. Methods: A total of 589 participants (432 with non-treatment seeking AUD, 157 controls) completed a Stop-Signal Task (SST) during fMRI. Associations between DSM-5 symptom domains (impaired control over alcohol use, risky use, social problems, physical dependence) and SST performance, inhibition-related brain activation, and alcohol consumption were examined. Real-life drinking behavior was additionally assessed over 12 months using ecological momentary assessment (EMA). Results: Greater impaired control symptom severity was associated with poorer performance on go trials and diminished post-error slowing during SST. At the neural level, higher impaired control symptom severity was associated with reduced activation across fronto-insular, parietal, cingulate, supplementary motor, and cerebellar regions. In contrast, other symptom domains showed little to no associations with inhibition-related behavioral or neural measures. Higher impaired control symptom severity was associated with greater alcohol consumption across questionnaire-based and EMA-based assessments. Conclusions: In this large cohort, altered behavioral performance and recruitment of inhibition-related brain regions in the SST was selectively linked to impaired control symptoms. These findings highlight the importance of symptom-based approaches for understanding heterogeneity in addiction and may inform the development of more targeted interventions.

In the KARE trial, 96 adults with severe alcohol use disorder, abstinent following detoxification, were randomised in a 2x2 factorial design to 3 weekly infusions of ketamine (0.8 mg/kg) or saline, each combined with mindfulness-based psychological support (PS) or alcohol psychoeducation (PE). We re-expressed the trial data on endpoints acceptable to regulators and compared the most and least intensive arms. Across pooled weeks 17-24, ketamine plus PS gave 78.5% versus 57.8% days abstinent (difference, 20.7 percentage points; P = .032); adjusted for baseline, the difference at weeks 21-24 was 26.9 percentage points (P = .011). At weeks 21-24, 50.0% versus 26.1% had no heavy drinking days (P = .127). Relative reductions in heavy drinking days were 55% in KARE, against 25%, -17%, 46% and 35% reported for naltrexone, exenatide, semaglutide and pemvidutide. These post hoc differences, measured 4 to 6 months after a 4 week course ended, are of magnitude similar to agents requiring continuous administration.

Objective Existing genetic studies of nicotine consumption have largely focused on cigarette smoking. Different modes of nicotine use have varying use patterns and user characteristics that may reflect distinct environmental and genetic influences. This study aims to identify the shared and specific genetic risks for five nicotine use modes. Methods We performed a common variant multi-ancestral genome-wide association study (cross-ancestry N = 304,213) of five types of nicotine use in the All of Us Research Program: lifetime use of electronic nicotine (e-nicotine) products, cigars, tobacco in a hookah, smokeless tobacco, or at least 100 cigarettes. Results Nicotine users, regardless of mode, were more likely to be male and of younger age (except for cigarette use), and all modes of use were positively associated with tobacco use disorder diagnoses. GWAS meta-analyses of cigarette, e-nicotine, cigar, and smokeless tobacco use identified 87, 4, 1, and 1 genome-wide risk loci, respectively (none for hookah), including previously identified genes (e.g., CHRNA4 for cigarettes). Cigarette, e-nicotine, and smokeless tobacco use showed strong genetic correlations (rg=0.82-0.92), while cigar and hookah use were strongly genetically correlated (rg=0.87). Cross-trait correlations diverged: cigarette, e-nicotine, and smokeless tobacco use showed the strongest positive correlations with psychiatric disorders, negative urgency, and neuroticism, whereas cigar and hookah use showed null-to-weak correlations with psychiatric disorders but positive correlations with sensation-seeking and openness. Conclusions Our findings reveal important phenotypic and genetic distinctions across modes of nicotine use, particularly in relation to personality traits such as impulsivity and to socioeconomic context.

Genome-wide association studies (GWAS) have identified numerous loci associated with substance use disorders (SUDs). To connect common genetic variation to cell type- and region-specific gene expression and chromatin accessibility, we examined the enrichment of genetic liability for several SUDs (alcohol, cannabis, tobacco, and opioid use disorders) and substance consumption phenotypes (drinks per week, cannabis ever-use, and cigarettes per day) in available single-nucleus RNAseq data spanning 10 brain regions and snATACseq data spanning 42 brain regions. Significant cell-type enrichment after multiple testing corrections was observed only for drinks per week and tobacco use disorder in the snRNAseq data and for cannabis use disorder and problematic alcohol use in snATACseq data. Across both transcriptomic and chromatin accessibility annotations, enrichment patterns were concentrated in excitatory neuronal populations, particularly upper layer intratelencephalic neurons, amygdala excitatory neurons, and regionally resolved striatal medium spiny neuron subtypes. Drinks per week, problematic alcohol use, and cannabis use disorder showed enrichment values that were significantly correlated within cell-type across snRNAseq and snATACseq, suggesting convergent biological signals across independent single-nucleus modalities. These findings implicate chromatin-mediated expression differences in specific excitatory neuronal populations as potential mediators of SUD genetic liability and highlight the value of integrating transcriptomic and chromatin accessibility data for characterising the neurobiology of addiction.

Background: Pavlovian-to-instrumental transfer (PIT) paradigms assess how environmental cues influence instrumental behaviour. Our previous research linked PIT to alcohol use disorder (AUD) and high-risk drinking using monetary rewards during learning. To better capture alcohol-related cue effects, we developed a PIT paradigm delivering trial-by-trial alcohol and juice rewards alongside monetary rewards, to examine alcohol-specific and general PIT effects in parallel. Methods: Seventy-five participants with AUD and ninety-five controls completed the task during functional magnetic resonance imaging. Alcohol-specific PIT was defined as increased alcohol choices during presentation of alcohol-associated versus juice-associated cues. General PIT was defined as increased response vigor during gain- versus loss-associated cues. Behavioural and neural associations with AUD status and AUDIT scores from baseline through one-year follow-up were examined. Results: Participants with AUD showed stronger alcohol-specific PIT effect than controls. Alcohol-specific PIT was positively associated with baseline AUDIT scores and with AUDIT scores across the one-year follow-up. Alcohol cues elicited stronger anterior insula responses, and region-of-interest analysis showed greater left amygdala responses to alcohol versus juice cues in AUD compared with controls. In contrast, general PIT showed a negative association with AUDIT but no association with AUD status. In the general PIT, no AUD-related neural differences were observed. Conclusions: These findings suggest that alcohol-specific PIT captures a clinically meaningful mechanism of cue-driven alcohol seeking that is distinct from generalized motivational transfer, supporting its potential utility as a mechanistic marker in AUD.

Gambling disorder (GD) is a behavioural addiction that may help identify addiction-related neural features without the direct neurobiological effects of a primary substance of dependence. We examined regional grey matter volume (GMV) and resting-state functional connectivity (rsFC) in the same well-characterised sample. Eighteen men with GD and 21 matched healthy controls underwent high-resolution structural and resting-state functional MRI. GMV was quantified across 214 cortical and subcortical regions, and seed-based rsFC analyses focused on striatal subdivisions and mesocorticolimbic regions. Group differences were evaluated using permutation testing and cluster-corrected mixed-effects modelling. GD was associated with lower GMV in the ventromedial prefrontal cortex, orbitofrontal regions and other cortical and subcortical areas, alongside higher GMV in a subset of limbic and default-mode regions. Participants with GD also showed lower connectivity between the limbic striatum and the hippocampus, thalamus and putamen. In exploratory analyses, somatomotor connectivity was positively associated with gambling severity (Problem Gambling Severity Index: Spearman's rho = 0.71, p = 0.003, false-discovery-rate-adjusted q = 0.016). Structural and functional findings overlapped spatially in regions associated with valuation, memory, reward and habit formation, but regional GMV did not mediate group differences in rsFC. These findings are broadly consistent with corticostriatal models of GD and identify candidate circuit-level differences for independent replication. Larger, more diverse and longitudinal samples are required to establish their reproducibility, temporal direction and clinical relevance.

A single nucleotide polymorphism (rs2304297) in the 3' untranslated region (UTR) of CHRNA6, encoding the alpha6 nicotinic acetylcholine receptor subunit, is associated with adolescent smoking behavior. Humanized rats carrying the 'C' or 'G' allele show sex- and genotype-dependent responses to adolescent nicotine, yet the variant does not alter the protein sequence and the mechanism has remained unknown. We examined the molecular consequences of rs2304297 in vivo in the adolescent ventral tegmental area (VTA). Chrna6 transcript abundance did not differ by genotype, nor did transcript stability or polyadenylation site selection. The variant was instead associated with a transcriptional signature spanning myelin maintenance (Enpp6), lipid metabolism (Retsat), and genomic repair (Polq), and with elevated basal lipid peroxidation in alpha6GG animals. Computational folding predicted greater conformational flexibility of the 'G' allele, and spatial colocalization of Chrna6 with miRNA-365-5p was reduced in alpha6GG animals of both sexes. Coordinated expression of the two transcripts, however, followed a sex-by-genotype pattern, absent in alpha6CC females and alpha6GG males. These same groups showed elevated tyrosine hydroxylase expression and dopaminergic neuron number, and expansion of the Chrna6 co-expression network following nicotine. We propose that rs2304297 acts through two distinct layers: genotype determines spatial engagement of miRNA-365-5p with the transcript, while sex determines whether that engagement produces functional regulation. These findings identify a post-transcriptional mechanism for a non-coding nicotine addiction risk variant and demonstrate that its molecular effect can be uniform while its consequences are not.

N6-methyladenosine (m6A) is the most abundant internal mRNA modification, dynamically regulating post-transcriptional gene expression. Although m6A dysregulation has been implicated in addiction neurobiology, comprehensive characterization of m6A alterations in alcohol use disorder (AUD) remains unexplored. We profiled m6A methylomes across eight reward- and motor-related brain regions in postmortem tissue from 12 AUD subjects and 12 matched controls (192 samples total) using MazF-mediated enzymatic cleavage and microarray-based m6A site mapping. Region-specific differential analysis identified 1,403 differentially methylated m6A sites (P < 0.05, |fold change| [&ge;] 1.5): amygdala (n = 198), caudate nucleus (n = 91), cerebellum (n = 201), hippocampus (n = 89), nucleus accumbens (n = 276), prefrontal cortex (n = 362), putamen (n = 142), and ventral tegmental area (n = 241). Host transcripts were enriched for functional categories spanning intercellular communication, immune and inflammatory responses, RNA and protein synthesis, energy metabolism, and neurodegeneration. Gene set enrichment analysis revealed a coordinated inverse relationship between m6A methylation status and transcript abundance: hypermethylated transcripts were enriched among downregulated genes and hypomethylated transcripts among upregulated genes, consistent with m6A-mediated transcript destabilization. Cross-validation with Allen Human Brain Atlas region-specific expression profiles demonstrated that differential methylation patterns were primarily AUD-driven rather than secondary to regional transcriptional heterogeneity. These findings position m6A epitranscriptomics as a novel molecular signature of AUD and suggest that m6A regulatory machinery may serve as a therapeutic target. Future studies with larger validation cohorts and mechanistic interrogation are essential to establish causal relationships and facilitate translation to drug development.

Following cocaine self-administration, male and female rats received daily semaglutide during a 30-day abstinence period before cue- and cocaine-induced reinstatement tests. Semaglutide attenuated cue-induced reinstatement and cocaine-induced stereotypy, although operant responding was elevated following cocaine priming. These findings support further investigation of semaglutide as a treatment for cocaine use disorder.

Background Cocaine use disorder (CUD) is estimated to be 50% heritable, yet few genetic variants have been robustly associated with risk. During January 2021 to June 2024, cocaine was involved in 30% of U.S. overdose deaths. Identifying genetic contributors to CUD could inform diagnostics and treatments. The missing heritability may stem from small sample sizes, models that inadequately account for small-effect common variants, and the exclusion of large-effect structural variants like copy number changes. Epistasis likely contributes but remains challenging to detect genome-wide. Results We focused on epistasis and haplotypes rather than individual SNPs by applying the custom correlation coefficient in BlocBuster to construct allele networks from eQTL SNPs. Testing these regulatory haplotype networks for association with cocaine use in 7,826 individuals with opioid use disorder identified 12 significant associations across two ancestries. For one haplotype, raw genotyping array intensity data revealed a copy number increase at the CYP2D6 locus in individuals of European ancestry. In a subcohort of individuals who co-use cocaine and heroin (i.e., speedball), two haplotypes were also significantly associated with HIV infection. One of these involved GPR75, whose ligand, CCL5 (RANTES), is shared with CCR5, a coreceptor for HIV entry. Conclusions Our findings, integrated with long-read human pan-genomes, support a model in which non-functional copies of CYP2D6 disrupt dopaminergic, glutamatergic, and kynurenine pathways in the brain, contributing to cocaine use severity. The association of additional haplotypes with HIV status in speedball users highlights the genetic contribution to comorbid risk in high-exposure populations.

Alcohol misuse is a leading cause of preventable death worldwide. Chronic alcohol is associated with disrupted circadian rhythms, yet molecular mechanisms linking circadian rhythm dysregulation and alcohol consumption are poorly understood. Current FDA-approved treatments for alcohol use disorder (AUD) do not target molecular rhythms or sleep-wake cycles. Mammalian circadian rhythms are regulated by transcription-translation feedback loops that regulate expression of 'clock genes' (e.g., Arntl encoding for BMAL1). Both human and rodent studies demonstrate associations between clock gene variants and changes in reward-seeking behavior. Evidence suggests astrocytes, non-neuronal brain cells with cell-autonomous rhythms, may regulate both circadian rhythms and reward. In the nucleus accumbens (NAc), a region responsible for modulating alcohol- and reward-related behavior, over 43% of the astrocyte transcriptome is expressed rhythmically. However, no studies to date have investigated roles of NAc astrocyte rhythmicity in regulating alcohol drinking. We used AAV8-Gfap-Cre to functionally ablate molecular rhythms in NAc astrocytes of BMAL1 floxed mice. Two-bottle choice (2BC), drinking-in-the-light (DIL), and drinking-in-the-dark (DID) assessed alcohol drinking. Behavioral assays included locomotor response to novelty, sucrose preference, and social interaction. Disrupting molecular clock function in NAc astrocytes increased binge-like drinking during both DIL and DID paradigms (d = 1.44), but not with any other drinking paradigm or behavior. This study carves out a unique role for astrocytes in controlling the temporal organization of reward circuitry and susceptibility to binge-drinking. Future studies will investigate clock-controlled astrocyte mechanisms, such as glutamate uptake and ATP release, that may underlie binge-like drinking behavior.

Fentanyl-related overdose deaths now commonly involve non-injection routes, yet preclinical opioid self-administration is modeled predominantly intravenously. Here we establish an oral fentanyl self-administration procedure in male and female inbred C57BL/6 and outbred CD1 mice that measures volitional intake, cue-driven seeking, extinction, and relapse. Mice self-administered oral fentanyl (70 {micro}g/mL) on a fixed-ratio 1 schedule across fifteen 3-hour sessions, followed by ten extinction sessions and a cued reinstatement test. A separate cohort underwent between-session dose thresholding across a quarter-log series from 222 to 22 {micro}g/mL. Seventy-five percent of mice acquired self-administration, with similar rates across genetic background and sex. Responding increased as fentanyl concentration fell, indicating dose-sensitivity toward a preferred drug level. C57BL/6 mice escalated intake and lever pressing across sessions, responded persistently early in extinction before declining, and reinstated pressing to a conditioned cue. CD1 mice consumed high levels from the outset with limited escalation and showed neither extinction nor cued reinstatement of pressing, but shortened their reward-port approach latency when cues returned. This shows that lever presses alone would have misclassified them as weakly conditioned. A composite severity score summing seven components of fentanyl-use risk varied continuously rather than splitting into high and low groups, even among inbred mice. Sex differences were largely confined to C57BL/6 mice, in which females showed stronger cue association and higher severity scores than males. These results reveal separable escalation-prone and relapse-prone phenotypes that track genetic background. Protocols, hardware specifications, and analysis code are openly available, lowering the barrier to adopting oral fentanyl self-administration.

Alcohol use disorders are heritable, and large-scale genome-wide association studies have recently identified novel genetic loci that are correlated with alcohol consumption and problematic alcohol use. One such gene is Slc39a8, which encodes the ion transporter ZIP8, responsible for cellular import of zinc and other metals. Slc39a8 is highly pleiotropic, and a common missense variant has been associated with decreased problematic alcohol use. Here, we used Slc39a8 heterozygous (HET) knockout mice and their wild type (WT) littermates to characterize the effects of reduced gene function on two models of voluntary alcohol consumption (drinking in the dark and intermittent access two-bottle choice), and anxiety- and depression-like behavior. We found that HET mice consumed less alcohol than WT in the intermittent access two-bottle choice procedure. No evidence was found for genotypic differences in baseline locomotor activity, anxiety-like, or depression-like behavior in the open field and forced swim tests. Slc39a8 mRNA expression and alcohol metabolism experiments confirmed decreased brain expression of Slc39a8 in HET mice and unchanged alcohol metabolism. Basal and post-alcohol measurement of zinc levels showed that HET mice after ethanol exposure had lower zinc concentrations in the liver and higher concentrations in blood than WT animals. Analysis of the spatial distribution of multiple trace metals in the brains of naive Slc39a8 HET and WT mice using Laser Ablation Inductively Coupled Plasma Mass Spectrometry (LA-ICP-MS) showed region-specific genotypic differences in levels of zinc and iron. These results provide evidence for a direct role of Slc39a8 in ethanol consummatory behavior that does not appear to be due to metabolism, affective state, or locomotor activity. Moreover, we confirm that this animal model shows tissue- and brain region-specific alterations in zinc levels, highlighting potential mechanisms for future study.

N-methyl-D-aspartate receptor (NMDAR) neurotransmission plays a central role in the neurophysiological and behavioral adaptations produced by chronic alcohol (EtOH) exposure and withdrawal. Although alcohol-induced changes in canonical NMDAR subtypes have been extensively studied, the mechanisms by which prolonged alcohol exposure remodels NMDAR function remain incompletely understood. To investigate the impact of chronic voluntary alcohol consumption on glutamatergic signaling, adult C57BL/6J mice were exposed to 20% EtOH using the intermittent two-bottle choice (I2BC) approach for 8 weeks and NMDAR function was examined in ventral tegmental area (VTA) dopamine (DA) neurons using whole-cell electrophysiology. Chronic EtOH exposure reduced NMDA-evoked currents, an effect that became more pronounced during withdrawal. Withdrawal also reduced NMDAR outward current rectification, indicating that chronic alcohol exposure altered the biophysical properties of NMDAR-mediated signaling. These unexpected observations led us to investigate the contribution of GluN3A-containing NMDARs. Pharmacological analyses revealed a significant increase in the proportion of VTA DA neurons expressing functional GluN3A-containing receptors following chronic EtOH exposure, with all tested neurons expressing these receptors during withdrawal. Glycine activation directly depolarized VTA DA neurons and increased action potential firing, demonstrating that the recruited receptors function as excitatory NMDARs. Moreover, the alcohol-induced changes in NMDAR currents, receptor pharmacology, and glycine responsiveness were absent in GluN3A null mice. Together, these findings identify functional recruitment of GluN3A-containing NMDARs as a previously unrecognized neuroadaptation through which chronic alcohol remodels glutamatergic signaling in the adult VTA.

4-hydroxytamoxifen (4-OHT) is commonly used to facilitate activity-dependent recombination in neurons. Here we show that one dose of 4-OHT devalues alcohol, contexts, and tastants, indicating that experimental designs using 4-OHT should control for its aversiveness. Further, reactivation of 4-OHT-responsive neurons causes conditioned taste avoidance. An activity screen coupled with graph theoretical analysis identifies candidate brain areas mediating these effects. We suggest these areas could be therapeutic targets for treating addiction.

Background: Pavlovian-to-instrumental transfer (PIT) describes how conditioned cues shape instrumentally learned behavior and has been linked to cue-driven alcohol seeking and relapse in alcohol use disorder (AUD). Acute psychosocial stress is an established relapse risk factor that may amplify cue-driven motivational control. Whether acute stress modulates PIT behaviorally and neurally, and whether this differs in AUD, remains unclear. Methods: Within the TRR265 study, 109 participants (62 AUD, 46 controls (CON)) completed a within-subject, randomized, two-day PIT paradigm under Trier Social Stress Test and placebo conditions, with repeated salivary cortisol sampling across sessions. A subsample (n = 62; 34 AUD, 28 CON) underwent fMRI during the transfer phase. Results: A robust PIT effect emerged across groups, and acute stress significantly enhanced cue-driven response vigor (p = .034), with no AUD-specific behavioral effect. Exploratorily, cortisol responders showed stronger stress-related PIT enhancement than non-responders (p = .022). Neurally, PIT engaged the right amygdala during placebo. Acute stress induced less downregulation of PIT-related left amygdala activity in AUD than in CON (pFWE = .019), with no group difference during placebo. Conclusions: Acute psychosocial stress potentiated the motivational influence of non-drug-related conditioned cues on behavior irrespective of AUD status, identifying PIT as a candidate mechanism through which stress may promote alcohol-seeking behavior. Despite comparable behavioral effects, AUD was associated with blunted stress-induced downregulation of PIT-related amygdala activity, suggesting altered neural regulation of cue-driven processing whose relevance for future alcohol consumption and clinical outcome remains to be established.

The lateral habenula (LHb) encodes aversive states and negative affect, positioning it as a candidate region for the negative reinforcement that drives alcohol withdrawal. However, little is known about how chronic ethanol exposure affects LHb neuronal function and glial biology during withdrawal. Here, we used chronic intermittent ethanol (CIE) vapor exposure, a well-established model of alcohol dependence that reliably produces somatic and affective signs of withdrawal, to examine LHb physiology and astrocytic markers during acute withdrawal in male and female rats. Whole-cell and cell-attached recordings revealed that withdrawal reduced evoked and spontaneous firing in LHb neurons, with rebound firing following a crossover pattern between males and females. Despite these excitability changes, the overall distribution of firing phenotypes was unchanged, suggesting a shift in gain rather than a reorganization of cell types. Immunofluorescence revealed increased Sox9+ and GFAP labeling in the LHb during withdrawal at the same time point when electrophysiology experiments uncovered impaired astrocytic regulation of glutamate clearance. Together, these findings reveal that withdrawal from chronic ethanol exposure produces neuronal and glial adaptations in the LHb, pointing to impaired glutamate regulation as a candidate mechanism relevant to the negative affective state of alcohol withdrawal. These findings position the LHb as a potential node linking astrocyte-neuron dynamics to withdrawal symptoms and relapse vulnerability in alcohol use disorder.

RationaleThe opioid epidemic continues to be driven by synthetic opioids, particularly fentanyl, yet the long-term behavioral manifestations of withdrawal remain poorly characterized. Women exhibit unique vulnerabilities to opioid use disorder, including greater susceptibility to withdrawal-related symptoms and relapse, suggesting that defining the behavioral adaptations that emerge during prolonged fentanyl abstinence may identify mechanisms underlying female-specific relapse vulnerability. Thus, we used an oral fentanyl drinking model to longitudinally examine nociceptive, affective, and exploratory behaviors across chronic exposure and abstinence in female mice. MethodsAdult female C57BL/6J mice underwent a modified 5-week Drinking-in-the-Dark paradigm with 4-hour daily access to fentanyl (30 {micro}g/mL) or water. Somatic withdrawal was assessed following naloxone-precipitated withdrawal. Thermal and mechanical nociception, sucrose preference, exploratory behavior, acoustic startle, nest building, and locomotor activity were evaluated during fentanyl exposure and throughout one month of abstinence. ResultsFemale mice exhibited robust naloxone-precipitated somatic withdrawal, confirming physiological effects of opioid drinking. Chronic fentanyl exposure produced transient thermal hyperalgesia during weeks 2-4 of drinking that resolved by the final week, but re-emerged after 30 days of abstinence. After 30 days of abstinence, we also found increased mechanical sensitivity. During early abstinence, fentanyl-exposed mice exhibited increased sucrose consumption and greater exploration of anxiogenic environments without alterations in general locomotion. Increased exploratory behavior persisted into prolonged abstinence and was partially rescued in mice who previously received naloxone. Additionally, fDID mice exhibited impaired nesting behavior, while prior naloxone-precipitated withdrawal improved nest-building performance. ConclusionsWe demonstrate that chronic opioid exposure, precipitated withdrawal trials, and duration of abstinence interact to shape protracted withdrawal phenotypes in female mice that extend well beyond drug exposure. These results provide insight into persistent withdrawal symptoms that may shape relapse vulnerability using a novel translationally relevant framework.