Altered molecular signaling pathways in the hippocampus of rhesus monkeys following chronic alcohol use Authors: Tanya Pareek, John M Vergis, Xiaolu Zhang, Donna M Platt, Kathleen A Grant, Robert McCullumsmith, Barbara Gisabella, Sinead M O’Donovan, Harry Pantazopoulos Published: 2026 ABSTRACT AbstractContext-induced relapse is a significant factor limiting recovery from alcohol use disorder (AUD). However, the molecular processes in the hippocampus, critical for contextual memory, impacted by chronic alcohol use, remain poorly understood. We used a non-human primate model to test the hypothesis that chronic alcohol use impacts hippocampal molecular pathways that may serve as therapeutic targets for memory processing in chronic alcohol use. We conducted RNAseq profiling on hippocampal samples from adult male rhesus monkeys with chronic alcohol use (n = 7) and controls (n = 5). We identified 2,575 differentially expressed genes in subjects with chronic alcohol use, including genes implicated in genome-wide association studies of alcohol dependence, such as GLP2R and GABBR2. Downregulated pathways included chemical synaptic transmission, trans-synaptic signaling, and neuron development, and upregulated pathways involved mitochondrial function. Targeted pathway analysis highlighted downregulation of synaptic signaling and upregulation of mitochondrial processes. Leading-edge gene analysis revealed downregulated genes involved in synaptic signaling and upregulated genes involved in mitochondrial processes. Drug repurposing analysis identified several potential therapeutic targets, including epidermal growth factor receptor inhibitors and L-type calcium channel blockers. Our results provide critical insights into molecular pathways underlying hippocampal pathology in chronic alcohol use and offer potential novel therapeutic targets.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-026-51816-0.Subject terms: Drug discovery, Neuroscience Introduction According to the 2023 National Survey of Drug Use and Health survey, approximately 28.9 million individuals in the United States were affected by an alcohol use disorder (AUD). Although available pharmacotherapies for AUD exist, high rates of relapse persist due to various complex factors1. Moreover, these treatments are effective only for certain populations2, highlighting the need to develop effective therapeutic strategies and novel pharmacotherapies to reduce relapse. There is a critical need to identify the molecular signaling pathways involved in chronic alcohol use and relapse in order to identify therapeutic targets for novel pharmacotherapies. Drug relapse can be induced by a multitude of factors, including contextual cues and stress3,4, which contribute to the formation and strengthening of drug-associated contextual memories. Furthermore, subjects with chronic alcohol use display enhanced memory for contexts associated with alcohol5. The hippocampus is at the center of neurocircuitry involved in reward memory processing6, including contextual information associated with reward memories7–9. Specifically, accumulating evidence suggests that hippocampal subfield CA1 integrates dopamine reward signaling with contextual memories and regulates reward expectation and extinction10,11. For example, exposure to alcohol can enhance future alcohol intake by altering contextual and cue hippocampal processes12, highlighting the importance of acute changes (i.e., enhancement of cue, context, and reward processes) underlying learning processes following alcohol exposure. However, long-term exposure to alcohol can impair learning and memory by disrupting the hippocampal processes13,14, and may contribute to generalized reward cues promoting relapse15–18. While acute and chronic alcohol exposure may cause differential effects in the hippocampus, such as enhanced contextual alcohol reward memory compared with general memory impairments, it is clear that hippocampal circuits play a critical role in relapse. Therefore, improving our understanding of molecular factors in the hippocampus involved in chronic alcohol use may provide potential targets for the development of effective therapeutic approaches for AUD. Although pharmacotherapies are available for alcohol use disorder (AUD), effective treatments for context-induced relapse have not been identified. Furthermore, there is a lack of information regarding hippocampal alterations in chronic alcohol use in the primate hippocampus, limiting the development of novel therapeutic strategies. Current information from a small number of human studies is complicated by confounding factors inherent to human postmortem cohorts, such as comorbidities and medication exposure19,20. We used a nonhuman primate model to examine the effects of chronic alcohol use on molecular pathways in the primate hippocampus, without the confounding factors typically present in human postmortem studies (Fig. 1). Our bioinformatic approach also identified potential therapeutic targets that may reverse the gene expression signatures identified in subjects with chronic alcohol use. Study Overview. Twelve male rhesus monkeys from the MATRR cohort 5 were given an open-access period with either ethanol and water (n = 7) or 10% maltose dextrin and water (n = 5) for 12 months. RNA-seq processing was performed at the UMMC Molecular Genomics Core. Furthermore, bioinformatic analyses were performed, including GSEA pathway analysis, Enrichr pathway analysis, and iLINCS perturbagen analysis. Results Differential gene expression analysis We identified 2,575 differentially expressed genes (DEGs) in the hippocampus of monkeys with chronic alcohol use with p-value < 0.05, and 566 DEGs meeting the criteria for p-adjusted significance (Fig. 2, Table S2). These include several genes implicated in GWAS studies of alcohol dependence (e.g., GLP2R, GABBR2)21. Downregulated genes included GLP2R, GABBR2, genes involved in synaptic signaling (e.g., GRIN2A, GABRB2, KCNA1), genes involved in circadian rhythms and sleep (e.g., PER2, GSK3B, and ADORA1; Figure S1, Table S2), and extracellular matrix molecules (e.g., NCAN, TNR, SPOCK1, CTSB, ADAM22 and ADAM23; Figure S1, Table S2). Upregulated genes included genes involved in metabolism (e.g., NDUFA1, ND3, COX5B, ATP5E, ATP5L) ribosomal processes (e.g., RPL19, RPS29, RPL34, RPL21) and oxidative stress (e.g., NUDT1, ROMO1, NDUFB2, NDUFB1). Volcano Plot of differentially expressed genes (DEG) in male monkeys with a history of chronic alcohol use. Differentially expressed genes in the hippocampus of monkeys with chronic alcohol use include several downregulated genes (e.g., GLP2R, GRIN2A) and upregulated genes (NDUFA1, RPS29). X-axis represents Log2 Fold Change (i.e., 1 or -1 signifies a 2-fold increase or decrease in gene expression, respectively). Y-axis represents -log10(p-value). Blue and red illustrate significantly downregulated and upregulated genes (padj < 0.05), respectively. Full transcriptome pathway analysis Gene Set Enrichment Analysis (GSEA) pathway analysis identified 117 significantly altered pathways between the ethanol and control groups. Upregulated pathways (57 total) included ribosomal process, metabolic processes, and immune response (Fig. 3A, Table S3). In comparison, downregulated (60 total) pathways include regulation of trans-synaptic signaling, modulation of chemical synaptic transmission, synapse organization, postsynaptic density, dendritic spine, and regulation of synapse organization (Fig. 3A, Table S4). The top 20 upregulated pathways include pathways related to cellular protein synthesis and energy metabolism (e.g., mitochondrial function; Fig. 3B). Downregulated pathways include those involved in neuronal development, activity, and synaptic signaling (Fig. 3B). Pathway analysis. (A) Full transcriptome pathway analysis (GSEA) of hippocampal pathways of monkeys with chronic alcohol use. The top 20 upregulated and downregulated pathways (padj < 0.05) were plotted against the normalized enrichment score on the X-axis. This score considers the overrepresentation of a gene set in a ranked list of all genes between two conditions (i.e., EtOH vs. control). A positive NES indicates significant upregulation at the top of the list, a negative NES signifies significant downregulation at the bottom, while an NES close to 0 suggests no enrichment. Larger absolute NES values indicate stronger enrichment. (B) Targeted pathway analysis of hippocampal pathways of monkeys with chronic alcohol use. The top 20 upregulated and downregulated pathways (padj < 0.05) are presented. The X-axis represents the log combined score (i.e., logCS). The combined score (i.e., CS) ranks enriched pathways by multiplying the log of the enrichment analysis’s p-value (statistical significance) by the Z-score (degree of enrichment). A high CS denotes significant and robust enrichment unlikely to be due to random chance, whereas a low CS indicates weak enrichment or randomness. (C) A Venn diagram of pathways unique to either GSEA or Enrichr, or shared between the two, is shown, along with a list of the 21 shared pathways. Pathway analysis using Enrichr Enrichr pathway analysis focused on the top 10% and bottom 10% DEGs. The combined score (CS) values represent the likelihood that enriched pathways are significant and unlikely due to chance. Enrichr analysis detected 81 significantly altered pathways, with 15 upregulated and 66 downregulated, including calcium ion-regulated exocytosis of neurotransmitter and the cytoskeleton of the presynaptic active zone. Further, ribosomal pathways show higher expression in monkeys following chronic alcohol use, whereas synaptic signaling-related pathways show lower expression (Fig. 3B, Tables S7 and S8), highlighting possible disruptions in neuronal communication. The top 20 upregulated pathways largely centered on the fundamental processes of protein synthesis, ribosome assembly, and antimicrobial humoral response (Fig. 3B). Downregulated pathways included those involved in neuronal structure, synaptic organization, and neurotransmission (e.g., calcium ion-regulated exocytosis of neurotransmitters, axonal fasciculation, and the postsynaptic density) (Fig. 3B). Overlaps between GSEA and Enrichr pathways Overlap between the GSEA and Enrichr analyses, representing the most consistent changes identified across both analyses, identified 21 shared pathways. Shared upregulated pathways included the small ribosomal subunit, cytoplasmic translation, and the antimicrobial humoral response (Fig. 3C). Shared downregulated pathways included postsynaptic density, voltage-gated potassium channel activity, and chemical synaptic transmission (Fig. 3C). Leading-edge gene analysis identified transcripts that appear most frequently across the top and bottom pathways (i.e., their expression levels increase or decrease in the chronic alcohol group compared to controls, respectively). The top 20 common leading-edge genes derived from our GSEA analysis (Fig. 4, Tables S5 and S6) include statistically significant genes involved in inflammatory response (e.g., CCL2, CCL3, CCL11, C5 and IL6) and genes involved in mitochondrial function such as mitochondrial respiration, ATP production, and oxidative phosphorylation (e.g., NDFUFA2, NDFUFA5, NDFUFA8, NDUFB7, NDUFS4, MT-CYB, COX4I1, COXB5). Leading-edge genes in downregulated pathways included genes involved in synaptic transmission (e.g., NGLN1, GRIN2A, GRIN2B, DLG4, SHANK3), L-type voltage-gated calcium channel signaling (e.g., CACNA1C, CACNA2D1, and CACNA1G), and extracellular matrix molecules (NGLN1, NRXN1, TNR, and ADAM22) are downregulated following chronic alcohol use, highlighting potential dysregulation. Leading Edge Gene Analysis. Top 20 leading-edge genes (A) and bottom 20 leading-edge genes (B) in the hippocampus of monkeys with chronic alcohol use. X-axis represents the number of pathways evaluated. Drug Repurposing analysis using iLINCS Signature-based connectivity analysis utilizing the Library of Integrated Network-based Signatures (LINCS) database was used to identify chemical perturbagens that reverse (discordant) or simulate (concordant) the transcriptional signatures we identified in the hippocampus of monkeys with chronic alcohol use (Fig. 5). We identified 324 unique concordant (i.e., perturbagens that may induce disease-associated gene expression changes) and 373 discordant (putative therapeutic) mechanisms of action (MoAs). Clustering analysis of iLINCS pathways (Supplementary Fig. 2) identified key biological processes that are dysregulated in ethanol-treated subjects. Major clusters include positive regulation of protein modification processes, protein phosphorylation-mediated amino acid binding, regulation of cell differentiation, chemical synaptic transmission, and cytokine-mediated signaling pathways. Several discordant chemical perturbagen signatures representing potential therapeutic compounds were identified, including perturbagens associated with these pathways such as AKT Inhibitor III, GSK-3 Inhibitor II, the serotonin/norepinephrine transporter inhibitor and dopamine antagonist Amoxapine, the L-type Ca channel antagonist Felodipine, and several P13K inhibitors (Fig. 5A, Table S10). The top 20 discordant mechanisms of action (MoAs), based on the number of chemical perturbagen signatures, identified included VEGFR inhibitors, PDFGR inhibitors, NFkB inhibitors, serotonin receptor antagonists, dopamine receptor antagonists, calcium channel blockers, and NADH dehydrogenase inhibitors (Fig. 5B, Table S12). Further, the top 20 concordant pathways included GABA-gated chloride ion channel activity, GABA-A receptor activity, and benzodiazepine receptor activity, as expected, as well as 3’,5’-cyclic-AMP phosphodiesterase activity, ionotropic glutamate receptor activity, and high voltage-gated calcium channel activity (Fig. 5C, Table S15). In comparison, the top 20 discordant pathways included several pathways involved in ion channel and receptor-mediated signaling, including GABA-gated chloride channel activity, adenylate cyclase-inhibiting serotonin receptor signaling, and extracellular ligand-gated monoatomic ion channel activity, as well as pathways involved in energy metabolism and mitochondrial function, including mitochondrial electron transport, mitochondrial respiratory chain complex I and NADH dehydrogenase activity (Fig. 5D, Table S16). In addition, 454 unique concordant and 497 unique discordant gene targets were identified (Tables S13 and S14). iLINCS Drug Repurposing Analysis. iLINCS drug repurposing analysis organized as top discordant (blue; < − 0.4) and concordant (red; > 0.4) chemical perturbagens (A). Here, X-axis represents the concordance (positive correlation coefficient) and discordance (negative correlation coefficient) scores. (B) illustrates the top 20 discordant mechanisms of action (MOA), where the x-axis represents the number of signatures associated with a MOA. The 20 top and bottom pathways (padj < 0.05) associated with concordant (C) and discordant (D) gene targets. CS represents the combined score in order to rank enriched pathways. Discussion Although chronic alcohol use contributes to several hippocampal-dependent processes, including memory impairment and context-induced relapse, information regarding the molecular processes underlying contextual memory alterations in chronic alcohol use is currently limited, particularly in regard to humans and non-human primates. Furthermore, the information from the currently available small number of human studies is complicated by confounding factors inherent in human postmortem cohorts, such as comorbidities and medication exposure19,20. We used a nonhuman primate model to examine the effects of chronic alcohol use on molecular pathways of the primate hippocampus without the challenges of confounding factors present in human postmortem studies in order to identify effects specific to chronic alcohol exposure. Therefore, our dataset can also serve as a resource for interpreting potential confounding effects in current and future human postmortem studies. Our study, to our knowledge, represents the first molecular characterization of hippocampal alterations in non-human primates with chronic alcohol use. Downregulated pathways involved in synaptic transmission and synaptic signaling, along with upregulated pathways involved in mitochondrial processes and energy metabolism, may contribute to memory alterations inherent in chronic alcohol use and AUD. Furthermore, our findings identify several potential therapeutic candidates for context-induced relapse that warrant exploration in future functional and behavioral pharmacology studies. Several synaptic signaling processes, including chemical synaptic transmission, dendrites, presynapse, and regulation of synapse organization, were among the most consistently identified downregulated pathways (Fig. 3A), indicating broad synaptic deficits in the hippocampus following chronic alcohol use. Downregulation of synaptic signaling pathways identified by targeted pathway analysis, including the presynaptic active zone cytoskeleton, dendrite morphogenesis, and GABAergic synapse pathways, provides further support for broad synaptic deficits (Fig. 3B). Furthermore, downregulated leading-edge genes included genes in synaptic regulation, such as NLGN1 and GRIN2A, GRIN2B, and GRIN1 (Fig. 4, Tables S5 and S6). Many of the downregulated synaptic genes and pathways are prevalent in the dendrites and axons where they are locally translated during synaptic processing {Glock, 2021 #77012}, suggesting that they may reflect broader reductions of the neuropil. Reports that alcohol exposure decreases hippocampal dendritic complexity and neuropil support this hypothesis {Sircar, 2023 #77013}{Lescaudron, 1989 #77014}{McMullen, 1984 #77015}{King, 1988 #77016}. Downregulation of synaptic signaling pathways, including several glutamate receptors, indicates impaired hippocampal neurotransmission and synaptic plasticity. Glutamatergic NMDA receptors, including GRIN2A, GRIN2B, and GRIN1, are essential for hippocampal learning22,23. Downregulated genes included several involved in synaptic regulation, such as GLP2R and GABBR2, which are implicated in GWAS studies of AUD21. Hippocampal decrease of GLP2R expression is associated with deficits in spatial memory processing, and upregulation of GLP2R enhances hippocampal neurogenesis and synaptic strength of hippocampal neurons24, suggesting that the observed decrease of GLP2R mRNA may contribute to hippocampal synaptic deficits with chronic alcohol use, and people with a GLP2R genetic predisposition may be more susceptible to such alterations. Broad decreases of synaptic signaling pathways may underlie memory deficits inherent with chronic alcohol use. Synaptic signaling decreases are in line with memory impairment reported in subjects with chronic alcohol use16,18 as well as with preclinical studies reporting impaired spatial and contextual memory following alcohol exposure25–27. In addition to synaptic signaling, we detected downregulation of pathways involved in chemical synaptic transmission, neuronal projection development (e.g., axonogenesis and dendrite morphogenesis), and calcium ion transport, indicating a loss of neuronal connectivity. Such impairments in connectivity and plasticity have been implicated in the cognitive deficits observed in individuals with AUD28–30. Cell culture and rodent model studies of alcohol exposure, however, indicate complex alterations in synaptic marker proteins. Chronic (4 days) of ethanol exposure results in increased NR1(GRIN1) and NR2B (GRIN2) presynaptic clustering, accompanied by enhanced NMDA currents in rat hippocampal cultured neurons {Carpenter-Hyland, 2004 #77033}. Similarly, chronic intermittent alcohol exposure in cultured hippocampal rat neurons results in increased NR2B postsynaptic clustering accompanied by dendritic spine increases {Carpenter-Hyland, 2006 #77034} In comparison, protein analysis of hippocampal samples from animals following 7 days of withdrawal demonstrated inceased expression of NR2A and NR2B {Nelson, 2005 #77035} Changes in hippocampal protein levels of these receptor subunits have also been shown to be age dependent in a chronic ethanol exposure rat model {Pian, 2010 #77036}. Proteomic analysis of GluN2B synaptic proteins from mice following withdrawal from chronic intermittent alcohol exposure revealed complex interactions of GluN2B with other synaptic proteins {Wills, 2017 #77037}. These changes included dissociation of Glun2B receptors from postsynaptic scaffolding proteins, along with increased interactions with metabotropic glutamate receptors, Arc, and Homer1 {Wills, 2017 #77037}. Collectively, these studies indicate that synaptic protein-level alterations depend on age and time course and occur at the levels of localization and complex formation. Although transcriptomic data may not always capture the complexity of protein changes, our results provide a broad overview of alterations in the nonhuman primate hippocampus following chronic alcohol use that may serve as targets for protein localization and functional analyses. Sleep and circadian rhythm disturbances are commonly reported in people with chronic alcohol use31,32. We observed the downregulation of genes involved in sleep and circadian rhythm regulation, including PER2, GSK3B and ADORA1 (see Figure S1, Table S2). Downregulation of the core clock gene PER2 is consistent with previous studies showing that mice with a PER2 deletion exhibit increased alcohol intake33. Furthermore, mutations in clock genes, including PER2, have been associated with alcohol dependence34. Similarly, ADORA1 is critically involved in sleep regulation35, with mice lacking ADORA1 expression displaying decreased sleep drive. Decreased hippocampal ADORA1 expression in monkeys with chronic alcohol use may contribute to insomnia associated with chronic alcohol use36. Our previous study, using hippocampal samples from the same rhesus monkey cohort as this study, identified increased densities of perineuronal nets (PNNs) in monkeys with chronic alcohol use37. PNNs are extracellular matrix structures that form primarily around subsets of inhibitory interneurons, stabilizing synapses, including synapses involved in reward learning, and protecting neurons from oxidative stress38. We observed several extracellular matrix genes downregulated in the hippocampus of monkeys with chronic alcohol use, including the endogenous proteases ADAM11, ADAMTS8, ADAMTS19, MMP16, and CTSB (see Figure S1, Table S2), suggesting decreased PNN degradation in these monkeys. Increased PNNs, alongside downregulated synaptic signaling pathways observed in the same subjects, may reflect broad decreases in synaptic signaling in the presence of PNNs stabilizing alcohol reward-associated synapses, possibly contributing to enhanced memory for alcohol-associated contexts reported in subjects with AUD5. Furthermore, PNNs in monkeys with chronic alcohol use may protect neurons from oxidative stress39, as suggested by upregulated pathways involved in mitochondrial processes (Fig. 3A). In comparison, GSEA analysis identified fewer upregulated pathways than downregulated pathways (280 upregulated vs. 623 downregulated pathways) and largely consisted of mitochondrial processes, including protein synthesis, energy metabolism, oxidative phosphorylation, and cellular respiration (Fig. 3A, Tables S3 and S4). The top 20 leading-edge genes were all involved in mitochondrial and NADH metabolic processes. Furthermore, targeted pathway analysis identified more selective metabolic processes as the top upregulated pathways, including mitochondrial electron transport, RNA metabolic processes and NADH dehydrogenase activity. Upregulated pathways involved in protein synthesis and mitochondrial function may represent enhanced metabolic processes and increased oxidative stress induced by alcohol use40,41. Enhanced metabolic stress may reflect neurotoxicity from oxidative stress and inflammation reported in the brain following chronic alcohol use42–44. Our iLINCS perturbagen analysis identified putative therapeutic candidate signatures and pathways encompassing dopamine and serotonin receptor antagonists, VEGFR/KIT inhibitors, PDFGR inhibitors, mitochondrial complex I inhibitors, and L-type calcium channel blockers (Fig. 5). Several compounds are predicted to partially reverse alcohol-induced changes in hippocampal gene expression, representing potential pharmacological targets. These included pharmacological compounds such as AKT Inhibitor VIII, Amoxapine, GW842166X, and Felodipine (Fig. 5A). AKT activation has been implicated in prior alcohol use studies demonstrating that AKT activation in the nucleus accumbens contributes to alcohol self-administration and binge drinking, which are attenuated by AKT inhibition45. Also, AKT has been implicated as part of a signaling pathway with PI3K and GSK3beta contributing to withdrawal effects from chronic alcohol exposure, and PI3K inhibition can alleviate these effects46. Several PI3K inhibitors were identified as discordant chemical perturbagens in our analysis, including PF-04691502 and TG100-115 (Fig. 5B). However, these studies were not focused on the hippocampus; thus, further studies examining PI3K in the hippocampus following chronic alcohol exposure are needed to confirm the potential role of PI3K alterations in this region. Along the same line, the potent GSK3beta inhibitor GSK-3 Inhibitor II was identified as a top discordant chemical perturbagen (Fig. 5B). In addition to targeting the PI3K-AKT-GSK3beta pathway, GSK3beta is a core molecular clock gene that may contribute to the alterations in molecular circadian clock genes we observed (Figure S1). Further, Amoxapine, a serotonin/norepinephrine transporter inhibitor and dopamine receptor antagonist, may be repurposed to address dysregulated dopaminergic signaling reported with chronic alcohol use47,48. Dopamine signaling to the hippocampus is critically involved in reward processing and relapse49. Hippocampal activity increases during reward anticipation, together with increased connectivity with the ventral tegmental area, indicating that dopamine modulation of the hippocampus by the VTA enhances hippocampal encoding during reward learning50. Furthermore, the hippocampus is critically involved in encoding predictability and uncertainty, key factors involved in reward processing51–53. GW842166X is a cannabinoid CB2 receptor agonist (Fig. 5B). Several preclinical studies report that the CB2 receptor is involved in regulating the rewarding effects of alcohol54,55. CB2 receptor knockout mice demonstrate increased alcohol preference and consumption, and genetic polymorphisms for the CB2 gene have been associated with AUD54. Furthermore, CB2 receptor knockout mice display increased proinflammatory markers in the hippocampus in response to alcohol exposure, suggesting that CB2 agonists may also reduce the inflammatory effects of alcohol on this brain region55. Felodipine is a mineralcorticoid receptor antagonist and L-type calcium channel (LTCC) blocker (Fig. 5B). Preclinical studies report increased L-type calcium channel expression and amplitude of LTCC currents in the hippocampus of alcohol-dependent rats56. Furthermore, cue-induced alcohol reinstatement in these animals was attenuated by central administration of the LTCC antagonist verapamil56, providing support for the use of these blockers as potential therapeutic agents. Our observed high voltage-gated calcium channel activity in the concordant pathway (Table S15) provides additional support for LTCC blockers as a potential therapeutic strategy. In addition, felodipine has been reported to attenuate neuroinflammation57, suggesting it may counteract multiple effects of alcohol on the hippocampus. It is important to note that this iLINCS analysis represents an initial step in identifying candidate therapeutic targets. As this analysis is based entirely on transcriptional gene expression signatures from cell lines, several limitations are discussed below. Future studies are needed to confirm and expand on candidate therapeutic targets. Overall, these findings highlight several targets for future investigations into pharmacological agents to alleviate hippocampal pathology associated with chronic alcohol use. Finally, converging evidence between the full transcriptome GSEA and targeted Enrichr pathway analyses highlighted shared upregulated pathways related to small ribosomal unit, cytoplasmic translation, and antimicrobial humoral response (Figs. 3C and S3) and shared downregulated pathways involved in postsynaptic density, voltage-gated potassium channel activity, and chemical synaptic transmission (Fig. 3C and S3). The top shared pathways between these pods further support decreased synaptic signaling and upregulated gene expression pathways that promote neuroinflammatory responses and cell survival in response to enhanced mitochondrial stress. Impaired synaptic signaling and metabolic stress may underlie memory impairment and enhanced memory for contexts associated with alcohol5 reported in subjects with chronic alcohol use16,18. Technical limitations Although our study identifies several gene expression pathways and potential pharmacotherapy targets, our analysis was limited to bulk RNA-seq. As transcriptomic changes do not consistently result in corresponding protein changes, due in part to potential post-translational modifications, future studies examining protein expression of molecular targets of interest are needed to confirm functional implications. Furthermore, future studies examining cell-type specificity would provide critical insight into the neurocircuitry underlying our findings. Our iLINCS perturbagen analysis represents a first step toward identifying drug-repurposing candidates. However, this analysis was limited to gene expression signatures for compounds and cell types currently available in this database, and represents cell line samples rather than hippocampal samples. Therefore, future studies using expanded databases with more compounds and incorporating hippocampal tissue samples may identify additional and more selective therapeutic targets for follow-up behavioral and longitudinal studies. Lastly, our studies are conducted on samples from male rhesus monkeys and may not necessarily reflect changes in female animals. Future studies incorporating samples from female animals are needed to clarify potential sex differences. In summary, we identified a number of gene expression pathways altered in the primate hippocampus by chronic alcohol, including impaired synaptic signaling and metabolic stress that may contribute to memory impairment and enhanced memory for contexts associated with chronic alcohol use5,16,18, along with several potential therapeutic drug targets for this hippocampal pathology. Future studies evaluating cell type specificity and the effectiveness of potential therapeutic compounds (e.g., AKT Inhibitor II, Amoxapine, PI3K inhibitors, and Felodipine) may provide insight into therapeutic targets for hippocampal dysfunction in subjects with chronic alcohol use. Materials and methods Subjects and tissue collection Nonhuman primate samples were obtained exclusively from Monkey Alcohol Tissue Research Resource (MATRR) Cohort 5 (www.matrr.com), see Table S1 for demographic and experimental information. Fresh frozen hippocampal samples containing mid-body hippocampus from adult, male rhesus monkeys (Macaca mulatta) with a history of chronic, oral alcohol use (n = 7; age = 7 yo) or no alcohol (n = 5; age = 5–8 yo) were collected following the 12-month daily access to alcohol and water without imposed abstinence (i.e., within 2–4 h of the prior 22 h self-administration session)58. All animal-use procedures were approved by the Oregon National Primate Research Center’s Institutional Animal Care and Use Committee and were conducted in accordance with the National Research Council’s Guide for Care and Use of Laboratory Animals (8th edition, 2011). Monkeys were experimentally naïve at the onset of alcohol induction and followed a standard operating procedure of alcohol self-administration59. Briefly, open-access (i.e., 22 h/day) self-administration sessions were conducted daily for 12 months, with average ethanol intake ranging from 2.55 to 3.78 g/kg/day. Cohort 5 monkeys used were defined as binge drinkers (BD; n = 1), heavy drinkers (HD; n = 2), or very heavy drinkers (VHD; n = 4). Control monkeys’ self-administration sessions were identical to the alcohol oral self-administration sessions. However, the control monkeys had access to water and 10% maltose dextrin solution. RNA sequencing Hippocampal samples were removed from slides with sterile scalpels and processed for RNAseq profiling. RNA isolation, library preparation, and next-generation sequencing were performed by the Molecular and Genomics Core Facility at the University of Mississippi Medical Center, as described previously60. Total RNA was extracted from tissue samples using the Invitrogen PureLink RNA Mini kit with Trizol (Life Technologies; Carlsbad, CA, USA) according to the manufacturer’s protocol. The quality control of the total RNA was evaluated using the Qiagen QIAxcel Advanced System and the Qubit Fluorometer for quality and concentration measurements, respectively. Library preparation employed the TruSeq Stranded Total RNA LT Sample Prep Kit from Illumina (San Diego, CA, USA), per manufacturer’s protocol using up to 1 ug of RNA per sample. Libraries were index-tagged and pooled for multiplexing and sequenced on the Illumina NextSeq 500 platform with paired-end reads (2 × 75 bp) using the Illumina 150-cycle High-Output reagent kit. Sequence reads were aligned to the MacaM genome with the basespace application RNA-Seq Alignment (Version: 2.0.1 [workflow version 3.19.1.12 + master]) that conducted both splice-aware genome alignment with STAR alignment (version 2.6.1a61, and transcriptome quantification with Salmon (version 0.11.262, . RNAseq data are available at NCBI GEO GSE297855. R scripts for the bioinformatics workflow are available at: https://zenodo.org/badge/latestdoi/642681935. Differential gene expression analysis The DESeq2 R package was used to assess the differential expression of all genes in comparisons between the control and ethanol groups following the default workflow described in63. Prior to differential expression testing, genes were filtered to remove lowly expressed features. Genes where fewer than 50% of samples had a normalized count ≥ 1 were filtered out from downstream analysis. Size‑factor estimation, dispersion estimation, and fitting of gene‑wise negative binomial generalized linear models were performed using the design formula ~ condition to assess differential expression between the two groups. Results were reported with both raw p-values and p-values adjusted for multiple testing using the Benjamini–Hochberg false discovery rate (FDR) procedure. Full transcriptome pathway analysis Gene Set Enrichment Analysis (GSEA) with the full set of genes was performed using the fgsea R package (version 1.16.0) against human enrichment map gene sets downloaded from (https://baderlab.org/EM_Genesets/). sign(logFC) * (-log10(p-value)) obtained from DESeq2 output were used as a gene ranking metric. Human genes were converted to rhesus monkey homologs as needed. GSEA first ranked the genes based on a measure of each gene’s differential expression, and then an enrichment score was generated (i.e., degree of overrepresentation of genes in a gene set at the top or bottom of the list of all ranked genes between ethanol and control conditions64. Unless otherwise specified, significantly altered pathways (padj < 0.05) are represented by normalized enrichment score (NES). A positive NES and a negative NES represent a significant upregulation or downregulation, respectively. Leading-edge (LE) genes, i.e., genes most consistently represented across the top pathways in the GSEA analysis, were identified using the leading-edge gene analysis.LE genes can be interpreted as the core of a gene set that accounts for the enrichment signal64. The overlap between multiple leading-edge subsets was analyzed. Targeted pathway analysis using Enrichr To complement GSEA analysis, which determines if pathways (sets of functionally related genes) are coordinately up- or downregulated even if individual genes within the pathway are not strongly differentially expressed, targeted pathway analysis was conducted using Enrichr (R package version 3.0) to identify the most strongly dysregulated pathways based on a predefined threshold: (top and bottom 10% genesabsolute log2FC). The Enrichr sub-tools gene ontology (GO) databases, GO biological process, GO cellular component, and GO molecular function, were used in the analysis. A combined score (CS), calculated by multiplying p-values with z-scores {Chen, 2013 #77010}, was generated to identify pathways with a high CS (i.e., a significant and robust enrichment unlikely due to randomness) or low CS (i.e., a weak enrichment may be due to randomness) that are significantly (padj < 0.05) up or downregulated65. Overlap between GSEA and Enrichr pathways To obtain a more complete insight into the molecular changes associated with AUD in the hippocampus, shared GO pathways (padj < 0.05) between full transcriptome pathway analysis (GSEA) and targeted pathway analysis (Enrichr) were identified using the R script available at https://zenodo.org/badge/latestdoi/642681935. GSEA analysis uses all genes ranked by expression, and pathway analysis with Enrichr uses DEGs based on a threshold (top and bottom 10% L2FC). iLINCS analysis To identify perturbagens altering gene expression, the Library of Integrated Network-Based Cellular Signatures (LINCS) (https://www.ilincs.org/ilincs/) was used to identify concordant (i.e., causative) and discordant (i.e., treatment-related) gene signatures, mechanisms of action (MOAs), and pathways. LINCS is a National Institute of Health initiative that aims to create a comprehensive network of molecular reactions in response to environmental and internal stressors66. LINCS is a transcriptomic database that uses the L1000 assay, a gene expression array of 978 “hub” genes, to generate gene signatures. The L1000 genes are a reduced representation of the transcriptome, accounting for ~ 82% of its information content67. The database contains hundreds of thousands of gene signatures generated in human cell lines treated with chemical perturbagens (i.e., drugs) or following knockout or overexpression of individual genes. The log2 fold-change (log2FC) and raw p-value for the L1000 genes were extracted from the DEG analysis and submitted to the iLINCS portal to obtain a list of chemical perturbagens altering gene expression. The reported score is the Pearson correlation coefficient between the disease signature and the precomputed iLINCS drug signatures. The chemical perturbagens with discordance scores < -0.4 and concordance scores > 0.4 were retained. Chemical perturbagens were clustered by MOA categories inquired from L1000 FWD68, DrugBank database69, and the Broad Institute. Targeted pathway analysis of iLINCS concordant and discordant gene targets was conducted. Pathways were clustered using Pathway Analysis Visualization with Embedding Representations (PAVER) to group similar pathways and identify functionally related clusters {Ryan, 2024 #77011}. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 2 Author contributions T.P. contributed to data collection, J.M.V., X.Z., and S.O. analyzed the data, T.P., H.P., S.O., B.G, D.M.P., K.A.G. and R.M. contributed to data interpretation, T.P., J.M.V., S.O. and H.P. wrote the manuscript, B.G., D.M.P. , K.A.G. and R.M. contributed to manuscript preparation, H.P., T.P.,S.O. and B.G. designed the studies. Funding This research was supported by NIMH R01-MH125833 and the Baszucki Brain Research Foundation to HP; Joe W. and Dorothy Dorsett Brown Foundation Neuroscience Initiative: AWD-001292, BG P20-GM144041 to BG; NIAAA R24-AA019431 to KAG and R01-AA029023 to DMP. Data availability RNAseq data are available at NCBI GEO GSE297855. Declarations Competing interests The authors declare no competing interests. Footnotes Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. Tanya Pareek and John M. Vergis contributed equally to this work. Contributor Information Sinead M. O’Donovan, Email: Sinead.ODonovan@ul.ie Harry Pantazopoulos, Email: cpantazopoulos@umc.edu. References Associated Data Supplementary Materials Supplementary Material 1 Supplementary Material 2 Data Availability Statement RNAseq data are available at NCBI GEO GSE297855.