Skip to main content
NCBI home page
Wiley Open Access Collection logoLink to Wiley Open Access Collection
. 2026 Sep 14;50(9):e70413. doi: 10.1111/acer.70413

Transcriptome Analysis of Hippocampus After Weekly Binge Alcohol Exposure and Voluntary Exercise in Rats

Erika A Mehrhoff 1, Theo Andre 2, Marissa A Ehringer 1, J Leigh Leasure 2,3,✉
PMCID: PMC13575474  PMID: 42740391

ABSTRACT

Background

Binge alcohol exposure is associated with neurodegeneration in corticolimbic regions in both humans and animal models. We have previously shown that exercise limits hippocampal damage and neuroimmune activation in rats. Here we conducted a transcriptome analysis in order to elucidate genetic mechanisms underlying the protective influence of exercise on the binged hippocampus.

Methods

Male and female rats were gavaged once weekly for 5 weeks with 5 g/kg ethanol or isocaloric control solution. Half of the animals had access to running wheels 4 days/week. All animals were euthanized immediately following the final exercise session/c. 18–20 h after the last binge exposure. The hippocampus was microdissected and RNA extracted and sequenced.

Results

Transcription of 806 genes was altered by exercise, but only seven by binge alcohol, most of which were transcription factors. Only one gene, Zfp62, showed evidence for an interaction effect. Direct comparisons within exercised groups showed that exercise altered gene expression only in control (not binge) animals. Gene ontology (GO) analysis identified pathways associated with receptor signaling, namely protein kinase signaling, as differentially impacted by exercise. Deconvolution methods were used to explore gene expression changes in specific cell populations. This analysis indicated that expression differences found in bulk RNA‐Seq were not cell‐type specific.

Conclusions

Our results suggest that five weekly treatments of binge alcohol minimally alter the hippocampal transcriptome assessed 20 h after the last binge. However, binge ethanol exposure markedly blunted exercise effects on the transcriptome, suggesting future study of this interaction is merited. Deconvolution showed similar patterns, upholding its utility as a supplemental analysis for bulk sequencing studies.

Keywords: bulk RNA‐Seq, deconvolution, wheel running

Exercise massively impacts same‐day gene expression and may suppress binge effects on the transcriptome. Binge alcohol did not exert numerous effects on next‐day gene expression, although it likely delays exercise effects on the transcriptome.

graphic file with name ACER-50-0-g005.webp

1. Introduction

Although defined as consuming 4+ alcoholic drinks in 2 h for women (5+ for men) (NIAAA 2004), binge drinking may involve far more drinks in a single episode (Patrick and Azar 2018). In both humans and in animal models, excessive alcohol consumption is associated with degeneration of corticolimbic brain regions, including the hippocampus and frontal cortex (Kelso et al. 2011; Kubota et al. 2001; Wilson et al. 2017). Alcohol is one of the world's most commonly consumed drugs, and its effects on the brain are impacted by other lifestyle variables, such as smoking, social interaction, and exercise (Bittner et al. 2019). Exercise is of particular interest, as it is known to influence key alcohol outcomes in humans and animal models, including intake (Ehringer et al. 2009; Gallego et al. 2015; Leasure et al. 2015; Ozburn et al. 2008), hangover severity (Neighbors et al. 2024), and alcohol‐induced brain damage (Booher et al. 2020; West, Najjar, and Leasure 2019).

Known mechanisms of damage to alcohol‐sensitive regions such as the hippocampus include increased oxidative stress, inflammation, and decreased neurotrophic support, all of which contribute to decreased neurogenesis (Crews and Nixon 2009). Exercise achieves the opposite of these effects (Crews et al. 2004; Maynard et al. 2018). Gene expression analyses have elucidated genes and pathways underlying alcohol damage and those underlying brain exercise benefits. In the hippocampus, alcohol exposure alters expression of genes associated with oxidative stress, as well as cellular plasticity and neurogenesis (Choi et al. 2020; McClintick et al. 2018). Wheel running in rodents has been shown to increase expression of genes involved in neurotrophic factor production, protein trafficking, synaptic structure, and immune response in the intact rat hippocampus (Tong et al. 2001).

The genetics underlying the combined effects of alcohol and exercise have also been examined, with a focus on reward pathway regions, in an effort to understand how exercise reduces alcohol intake in rodents (Darlington et al. 2014, 2016). A comprehensive profile of the brain health pathways affected negatively by alcohol and favorably by exercise, however, has not yet been performed. In the present study, we conducted a transcriptome analysis of the binged hippocampus, with or without voluntary exercise, in an effort to pinpoint genetic mechanisms underlying the restorative effect of exercise on binge alcohol damage. We used a rat model in which animals received a weekly dose of binge alcohol, which we have previously shown to be associated with hippocampal cell loss and neuroimmune activation (West et al. 2021; West, Wooden, et al. 2019). In addition, we have shown that voluntary wheel running ameliorates both of these effects (West, Wooden, et al. 2019), so in the present study, animals had access to exercise. Of interest were genes uniquely impacted by the combination of binge exposure and exercise.

2. Materials and Methods

2.1. Animals, Binge Exposure and Exercise Access

Male and female Long‐Evans rats were bred in our colony and raised in a vivarium on a reversed light–dark cycle (9 AM off/9 PM on). At 21 days of age, animals were weaned and housed in single‐sex groups of 2–3. After weaning all rats were acclimated to experimenters through routine gentle handling (5 min per week until 8 weeks of age, then 5 min daily until experiments began) and provided access to food and water ad libitum. All experimental procedures were performed according to the Guide for the Care and Use of Laboratory Animals of the National Institutes of health and approved by the Institutional Animal Care and Use Committee at the University of Houston (UH IACUC). At 9 weeks of age (body weight females 155–240; males 250–370) rats were randomized into four groups: controls that were sedentary (n = 12), binged animals that were sedentary (n = 10), controls that had exercise access (n = 11) and binged animals with exercise access (n = 10). Rats in the binge groups administered alcohol via oral gavage once weekly for 5 weeks with 5 g/kg ethanol or isocaloric control solution (Figure 1A), as we have previously described (West et al. 2021; West, Wooden, et al. 2019). Animals in the exercise groups had running wheel access 4 days/week, beginning early in the dark cycle, as previously described (West, Wooden, et al. 2019). The day after the fifth dose of alcohol (or control solution), animals in the exercise groups had wheel access as usual. All animals were euthanized immediately following their final exercise session and/or ~20 h after the last binge exposure.

FIGURE 1.

FIGURE 1

Open in a new tab

The timeline of weekly experimental events is shown in panel A. For 4 weeks, animals were gavaged with 5.0 g/kg ethanol or control solution in the last half of the light/dark cycle. The following 4 days/week, animals in the exercised conditions had access to wheels for 4 h. On the fifth week, animals were gavaged as usual. The following day, exercised animals had wheel access as usual, then all animals were sacrificed immediately after exercise and/or 16–20 h post binge. Panel B depicts our analysis strategy. Read counts were used for differential expression analysis, WGCNA, and an exploratory deconvolution analysis. For the latter, publicly available single‐cell data were also used.

2.2. RNA‐Seq

Brains were rapidly extracted and hippocampi dissected. The entire hippocampus was dissected out, laid flat and sufficient lengths of the dorsal and ventral termini removed to yield a total sample weight of 25–30 mg. Samples were snap frozen and stored at −80°C until further processing. Total RNA was isolated from the medial region, using the Qiagen RNeasy Mini Kit. The RNA samples were sent to Novogene for mRNA sequencing using Illumina NovaSeq 6000 with paired‐end 150 base pair reads, which yielded fastq files with ~20 m reads. Four samples did not pass Novogene's initial quality control for integrity of mRNA, and so were not sequenced. The NF‐core pipeline was used to process raw read data into expected counts nf‐core/rnaseq v3.12.0 (P. A. Ewels et al. 2020). The pipeline was executed with Nextflow v23.04.2 (di Tommaso et al. 2017). A short summary is as follows. The fastq files were merged, and strandedness was inferred through Salmon v1.10.1 (Patro et al. 2017). Quality control reports were generated through FastQC (http://www.bioinformatics.babraham.ac.uk/projects/fastqc/), and unique molecular identifiers (UMIs) were extracted using UMI‐Tools (Smith et al. 2017). Files were trimmed for low quality reads and adapters using trim galore (https://www.bioinformatics.babraham.ac.uk/projects/trim_galore/), and BBsplit/sortmeRNA (Kopylova et al. 2012) was used to remove contaminants and ribosomal RNA, respectively. Reads were aligned to the Ensembl mRatBN7.2 reference genome using STAR (Dobin et al. 2013), and subsequently quantified using RSEM (Li and Dewey 2011). SAMtools was used to sort and index these alignments (Danecek et al. 2021), and the previously extracted UMIs were used to deduplicate the reads. Transcripts were subjected to extensive quality control using RSeQC, DESeq2, and MultiQC (Ewels et al. 2016; Love et al. 2014; Wang et al. 2012). The MultiQC file was used to quality check the sample data following processing. A DESeq2 principal component plot and heat map showing sample similarity were used to identify five samples as outliers (see Figure S1). Following this exclusion, the groups were as follows: exercise/binge (n = 8; 3F/5M), exercise/control (n = 10; 5F/5M), sedentary/binge (n = 9; 4F/5M), and sedentary/control (n = 12; 6F/6M).

2.3. Differential Expression Analysis

The overall analysis strategy for the transcriptome data is shown in Figure 1B. Differential expression analysis was performed using the DEseq2 R package v3.1 (Love et al. 2014). Specifically, a likelihood ratio test (LRT) was used, with exercise, alcohol, and their interaction as the model. Male and female rats were included; however, the study was not powered to identify sex differences, therefore sex was not included in the model. Genes with fewer than 20 reads across 50% of samples were excluded, bringing the total gene count to 13,385 from 30,562. The resulting DEGs were then evaluated for function using gene ontology classification (see below). In addition, contrasts were pulled from the DESeq2 model to stratify by group and examine DEGs, thus allowing for more direct comparisons between groups. The contraster function was used to identify genes differentially expressed between exercised and sedentary animals only within the control group, exercised and sedentary animals only within the binge group, binged and control animals only within the exercised group, and binged and control animals only within the sedentary group (Ekiz 2024). Resulting DEGs were also classified using gene ontology.

2.4. Weighted Gene Coexpression Network Analysis

Data were then analyzed using weighted gene coexpression network analysis, or WGCNA. The WCGNA package creates modules of genes that are coexpressed that can be analyzed for effects of exercise and alcohol (Langfelder and Horvath 2008). A soft power threshold of 16 was used. Following module creation, linear models were run using eigengene values of each module, and then empirical Bayes was used to smooth standard error. A model assessing the main effects of exercise and alcohol, as well as their interaction, was used. The genes in each significant module were investigated for gene ontology.

2.5. Gene Ontology

The PANTHER classification system was used for determining gene ontology (GO) of the DESeq2 and WGCNA gene results (Thomas et al. 2022). The statistical overrepresentation test was used to identify specific gene ontology terms and pathways for which the differentially expressed genes and WGCNA module genes are involved. This analysis utilizes Fisher's exact test and a cutoff of FDR < 0.05 was used. Various annotations sets were used to analyze the gene lists, including GO biological process complete, GO cellular process complete, GO molecular process complete, and PANTHER pathways. Gene lists from DE analysis and WGCNA modules were analyzed with each of the PANTHER categories.

2.6. Deconvolution Analysis

While the use of bulk RNA sequencing does not allow for the direct extraction of cell type specific information, deconvolution analysis can be used to estimate the underlying cell types within a bulk sample by leveraging single cell reference data. There are various tools available to conduct this type of analysis, and one of the most widely used is CIBERSORTx (Newman et al. 2019), which allows for both the imputation of cell type fractions and group (cell) level expression. This software has also been demonstrated to be superior to other single‐cell deconvolution methods like MuSiC and dtangle, providing the best deconvolution of brain cell subtypes (Sutton et al. 2022). CIBERSORTx requires a reference file of single‐cell RNAseq data to compute these cell type estimates and works best when each sample cell type present in the sample is also present in this reference file (Sutton et al. 2022). Thus, an ideal dataset should be from the rodent hippocampus. One dataset was identified that fulfilled these criteria, from Wang and colleagues (Wang et al. 2024). However, this dataset only included 2 control animals that could be used as a reference. Given that the gold standard reference dataset, the Allen Brain Atlas, was assembled using 530 animals (Yao et al. 2021), the dataset from the Allen Brain Atlas was selected for reference. Since data for the Norway rat is not currently available, we used the Orthogene R package (Schilder and Skene 2022) to convert a single‐cell Mus Musculus dataset to its Rattus Norvegicus orthologs. We elected to use the SMART‐seq version of the dataset, as it provides greater read depth compared to 10× (Yao et al. 2021).

The Allen reference dataset was filtered to include only hippocampal cells and associated cell types, following the methods detailed on STAR (Marquez‐Galera et al. 2022). This decreased the number of cells from 73,363 to 8027 and retained the cells and subsequent genes designated by Allen as neurons (pyramidal neurons, interneurons, and granule cells), endothelial cells, microglia, astrocytes, and oligodendrocytes. The Allen dataset started with 42,055 genes, which were then converted to their rat orthologs using the Orthogene package, which resulted in 15,641 genes. The genes were then converted to ENSEMBL IDs to match the bulk data; this resulted in 15,583 genes. Genes that did not have at least one read across all cells were removed, which resulted in 15,425 genes. The count data were normalized to counts per million using convertCounts from the DGEobj.utils package. The bulk data were also normalized using the same method to allow comparisons between the two.

CIBERSORTx then was used to estimate cell type–specific information from bulk data. First, a signature matrix was created using the reference dataset to identify genes that could discriminate each cell type. The default settings were used to create this file. Then the bulk RNA data were uploaded, and the cell type fractions imputed. For this step, batch correction and relative mode were enabled, and 1000 permutations were used. This fraction analysis was used to determine if cell type proportions differed between groups, which could potentially confound the gene expression analysis and obscure true differentially expressed genes. This is of particular interest because binge alcohol exposure is known to differentially affect certain cell types (Ji et al. 2018). After this, CIBERSORTx High‐Res mode was used to create an imputed expression matrix for each cell type. This expression matrix is rendered in counts per million, so the original bulk library sizes were used to convert these CPM counts to simple counts for downstream analysis.

3. Results

3.1. Differential Expression Analysis

Following raw read processing and quality control measures, the filtered expected count data were analyzed using DESeq2 to identify differentially expressed genes. Likelihood ratio tests were used to analyze the main effects of exercise, alcohol, and the interaction between exercise and alcohol, on gene expression. Animals in the exercised groups averaged 2.71 km per exercise session. Exercise had a substantial impact on gene expression, with 806 genes differentially expressed between exercise and sedentary animals at a BH adjusted p‐value < 0.1 (see Figure 2A). Of these 806, 258 genes (32%) were downregulated in exercised animals and 548 genes (68%) were upregulated in exercised animals (see Supporting Information). The 806 exercise DEGs were analyzed for overrepresentation of gene ontology (GO) terms using the PANTHER database. The categories of terms included molecular GO terms, biological GO terms, cellular GO terms, and pathways. Among these categories, exercise‐related DEGs showed significant enrichment (FDR < 0.05) only in the molecular function and biological process GO categories. The molecular GO terms identified were related to molecular transducer activity, signaling receptor activity, and transmembrane signaling receptor activity. The biological GO terms identified were related to cellular response to endogenous stimulus, cell surface receptor signaling, transmembrane receptor protein tyrosine kinase signaling, enzyme‐linked receptor protein signaling, and response to endogenous stimulus (see Figure 2B). A complete list of GO terms can be found in Table S1.

FIGURE 2.

FIGURE 2

Open in a new tab

Differentially expressed genes and associated pathways. (A) Many more DEGs were found in the exercise condition, compared to the binge condition. Only 1 gene, zfp62, was found to be changed in the binged, exercised animals. (B) The top GO terms for the DEGs identified with a main effect exercise. PANTHER identified GO terms at padj. < 0.1 for the 806 genes that were differentially expressed between exercised and sedentary animals at padj. < 0.05. (C) A volcano plot of the 806 exercise DEG.

The average BEC of animals in the binged groups was 115 ± 9 mg/dL. There were seven genes that were differentially expressed between binged and control animals at a BH adjusted p‐value < 0.1. Of these seven, three genes were downregulated (Neurod6, Pou3f1, and Myh2) in binged animals, and four genes were upregulated (Meis1, Slc39a12, Zfp62, and Unc13c) in binged animals (see Table 1). There was only one gene, Zfp62 (zinc finger protein 62; a transcription factor), that showed differential expression based on the interaction between exercise and alcohol animals at a BH adjusted p‐value < 0.1 (p = 0.073). Contrasts of the DESeq2 object show a significant difference in Zfp62 expression between binge and control animals within the sedentary group (p = 0.047).

TABLE 1.

DEGs identified with a main effect of binge.

Gene symbol Gene name Function
Meis1 Meis homeobox 1 Transcription factor
Slc39a12 Solute carrier family 39 member 12 Zinc transporter
Zfp62 Zinc finger protein 62 Transcription factor
Unc13c Unc‐13 homolog C Synaptic vesicle priming; neuroprotection
Neurod6 Neuronal differentiation 6 Transcription factor
Pou3f1 POU class 3 homeobox 1 Transcription factor
Myh2 Mysoin heavy chain 2 Skeletal muscle contraction; unknown role in brain
Open in a new tab

3.2. Contrasts From Differential Expression Analysis

An exploratory analysis with contrasts from the DESeq2 model was used to identify genes differentially expressed between each group comparison: exercised and sedentary animals only within the control group (22 animals), exercised and sedentary animals only within the binge group (17 animals), binged and control animals only within the exercised group (18 animals), and binged and control animals only within the sedentary group (21 animals). At a BH p adj. < 0.1, 1279 genes were differentially expressed between exercised and sedentary animals in the control group, however, in the binge group only one gene, Tipin, was differentially expressed due to exercise (Figure 3). There were 22 DEGs identified between binge and control animals in the sedentary group, however, in the exercise group, there were 0 DEGs due to binge. The 1279 DEGs due exercise in the control group had biological GO terms related to developmental processes, cell differentiation, and response to growth factor, and the gonadotropin‐releasing hormone receptor PANTHER pathway.

FIGURE 3.

FIGURE 3

Open in a new tab

Contrasts from DESeq suggest that binge exposure suppressed exercise effects on the transcriptome and vice versa. As shown in panel A, 1279 DEGs were detected when comparing sedentary and exercised animals within the control group at padj. < 0.1. However, only one gene was detected when comparing sedentary and exercised animals within the binge group. There were only 22 DEGs identified between control and binged animals, but none in exercised animals, suggesting that exercise suppresses binge effects on the transcriptome panel B.

3.3. Weighted Gene Co‐Expression Network Analysis (WGCNA)

WGCNA with a soft power threshold of 16 resulted in 44 modules. Linear models using eigengene values resulted in significant modules for the main effect of exercise but not the main effect of alcohol or the interaction effect. Significant modules were only found at a more liberal p‐value threshold of 0.15 (Table S2). The 10 modules significantly associated with exercise included nine that were upregulated in exercised animals (modules 2, 5, 11, 19, 21, 30, 36, 37, and 38) and one that was downregulated in exercised animals (module 1). Gene lists from each module were analyzed for overrepresentation of gene ontology (GO) terms using the PANTHER database. Only modules 1 and 21 had significant GO terms. Module 1 was downregulated in exercised animals and had genes related to translation/ribosomes, transcription factor activity, RNA metabolism/processing, metabolic/biosynthetic process, and respiratory electron transport. Module 21 was upregulated in exercised animals and had genes related to nervous system development/neuron differentiation, signaling receptor activity/transmembrane receptor protein tyrosine kinase activity.

3.4. Exploratory Deconvolution Analysis

CIBERSORTx was used to create a signature matrix with the single‐cell Allen mouse reference data (Figure S2). The signature matrix was then used to estimate cell type proportions of the neurons (pyramidal, interneuron, granule), astrocytes, microglia, oligodendrocytes, and endothelial cells (see Figure 4). Two‐way ANOVAs using exercise, alcohol, and their interaction were run within each cell type to compare the proportions between groups. There were no significant effects of exercise, alcohol, or their interaction on the estimated proportions of neurons, astrocytes, microglia, oligodendrocytes, and endothelial cells. The signature matrix was also used to estimate sample‐level gene expression for each of the five cell types. Endothelial cells and astrocytes did not return any confident predictions, so no further analysis was done on these cell types. NAs and 1 s in the data represent lack of confident gene prediction, therefore, they were filtered out before analysis, leaving neurons with 4876 genes, microglia with 7581 genes, and oligodendrocytes with 10,883 genes. Following conversion to counts from CPM, DESeq2 was run within each cell type. Neurons had no differentially expressed genes for exercise, alcohol, or their interaction. This is likely due to the low number of genes predicted with confidence in the neuronal cells. Microglia had 13 DEGs between exercise and sedentary animals, three DEGs between binge and control animals, and none for the interaction between exercise and alcohol at padj < 0.1. All 13 of the microglia DEGs were found in the initial bulk DE analysis, which includes Slc7a1, Mgst3l1, Amdhd2, Ino80b, Chrac1, Wdr45, Trmu, Rnasek, Mtx1, Dus1l, Abracl, Paip1, and Samd4b. The three alcohol DEGs identified in microglia were not found in the initial bulk DE analysis, which includes Mgst3l1, Nrgn, and Abracl. Oligodendrocytes had 197 DEGs between exercise and sedentary animals, no DEGs between binge and control animals, and three for the interaction between exercise and alcohol at padj < 0.1. Of the 197 exercise DEGs, 167 of them were found in the initial bulk DE analysis exercise effect. Results from this analysis can be found in Table 2. The large number of exercise‐related DEGs found within oligodendrocytes allowed for the use of PANTHER for GO. The GO biological process pathway identified the GO term “cellular response to endogenous stimulus” at FDR = 0.076. PANTHER protein class identified “RNA metabolism” at FDR = 0.071. Both of these GO findings were also identified in the bulk expression analysis. The three interaction genes identified in oligodendrocytes, Taf1, Klhl24, and Zfp62, were all found in the initial bulk DE analysis exercise effect, but only Zfp62 was found in the initial bulk alcohol effect.

FIGURE 4.

FIGURE 4

Open in a new tab

Results of the deconvolution analysis. Cell type proportions in each experimental group were estimated using CIBERSORTx. Cell types are ordered based on the average proportion of cells, with the highest proportion on the bottom, and below is shown the exact percentages of each cell type per group. There were no statistically significant differences between groups for any cell type.

TABLE 2.

Deconvolution analysis of differential expression within each cell type.

Microglia Endothelial cells Oligodendrocytes Astrocytes Neurons
Reference cells 178 197 231 969 6452
Genes used for DESeq 7581 NA 10,883 NA 4876
Exercise DEG 13 NA 197 NA 0
Binge DEG 3 NA 0 NA 0
Interaction DEG 0 NA 3 NA 0
Open in a new tab

4. Discussion

The initial goal of this study was to identify genes involved in exercise mitigation of binge‐induced hippocampal damage. Thus, our original focus was genes differentially expressed in animals exposed to both binge alcohol and exercise. Initial differential expression analyses, however, identified only one such gene, Zfp62. In addition, it was immediately clear that while exercise exerted a tremendous effect on gene expression, with 806 DEGs, binge exposure had a comparatively modest effect, with only seven DEGs. We therefore shifted our focus from individual genes to the independent impacts of binge exposure and exercise on the transcriptome, as well as potential interactions.

Our initial interpretation of the disproportionate effect of exercise was that it has a powerful influence on the transcriptome that is not altered by binge alcohol exposure the previous day. An exploratory analysis with contrasts from the DESeq2 model, however, presents a different picture. Within controls (no binge exposure), there were 1279 genes differentially expressed between exercised and sedentary animals (Figure 3). Within binged animals, however, there was only one gene affected by exercise. This suggests that exposure to binge alcohol the previous day suppressed exercise effects on the transcriptome. The opposite was also the case, such that prior exercise exposure suppressed binge effects on the transcriptome. Binge exposure changed the expression of 22 genes in sedentary animals, but 0 genes in exercised animals. Overall, our initial hypothesis that binged, exercised animals would have a unique suite of DEGs was not supported. Rather, it seems that both binge and exercise are capable of suppressing each other's transcriptomic effects, at least temporarily.

We know from prior studies that exercise restores the binge‐damaged hippocampus and other alcohol‐vulnerable corticolimbic structures (Maynard et al. 2018; West, Najjar, et al. 2019; West, Wooden, et al. 2019). Therefore, suppressive effects of binge alcohol on exercise‐driven gene expression must be short‐lived. Alcohol is known to exert numerous effects on noncoding RNAs (Denham et al. 2022), suggesting a possible mechanism by which exercise‐driven gene expression could be delayed. Moreover, some of the binge DEGs are transcription factors (see Table 1), so they may have downstream effects that temporarily suppress exercise DEG. In addition, in the studies where we have found that exercise restores the brain after binge damage, animals had exercise access over several days following binge exposure, so exercise later after binge exposure likely exerts full benefits on the hippocampal transcriptome. In addition, given that binge did not change gene expression in exercised animals, exercise may prevent some binge effects on DEGs.

Six of the seven binge DEGs (Table 1) are directly related to known mechanisms of binge alcohol‐induced hippocampal degeneration, including oxidative stress and inflammation (Crews and Nixon 2009). Moreover, four of the DEGs code for transcription factors—Meis1, Zfp62, Neurod6, and Pou31f, all of which regulate pathways involved in cell proliferation and differentiation (Kittke et al. 2024), hippocampal aging (Su et al. 2023), cellular differentiation, regeneration, and survival (Uittenbogaard et al. 2010), and generation of neural stem cells (Zhu et al. 2014), respectively. Of the remaining three, Slc39a12 and Unc13c have well‐defined functions related to cell survival and neuroprotection. Slc39a12 (ZIP12) is a zinc transporter. Zinc, and its efficient transport across the membrane, is crucial for cell function, including neurite outgrowth and defense against oxidative stress (Davis et al. 2021). Unc13c (Munc13‐3) is involved in synaptic vesicle priming and neurotransmission, with a putative neuroprotective role in Alzheimer's disease (Ansari et al. 2023). The last DEG, Myh2, functions in skeletal muscle contraction, with low levels of expression and unclear function in the brain (Protein Atlas, https://v20.proteinatlas.org/ENSG00000125414‐MYH2/brain).

Given the large effect of exercise on the transcriptome, gene ontology analysis was performed for the exercise DEGs to better understand their functions and associated pathways. The 806 DEGs identified from the main effect of exercise had molecular and biological GO terms related to receptor signaling, such as transmembrane receptor protein tyrosine kinase signaling. The 1279 exercise DEGs identified in the control (nonbinged) animals were also analyzed. These had biological GO terms related to developmental processes, cell differentiation, and response to growth factor, and the gonadotropin‐releasing hormone receptor PANTHER pathway. Between the 806 exercise DEGs and 1279 control exercise DEGs, there are 659 DEGs overlapping. The 659 DEGs resulted in GO terms similar to those identified in the original DEG lists, which related to cell differentiation and development, receptor signaling, and gonadotropin‐releasing hormone receptor PANTHER pathway. There is evidence that some receptor signaling, such as receptor tyrosine kinase signaling, is affected by exercise. For example, the BDNF–NTRK2 signaling pathway (Brain‐Derived Neurotrophic Factor Neurotrophic and Tyrosine Kinase Receptor Type 2) is a receptor tyrosine kinase signaling pathway, which has been shown to play a critical role in promoting hippocampal plasticity in rats following exercise (Ding et al. 2011). Ntrk2 and Bdnf were found to be significantly upregulated in exercised animals, with the contrasts revealing this effect was driven by the exercised control animals (Ntrk2, p = 0.0094; Bdnf, p = 0.048). Bdnf and Ntrk2 both appeared on the list of genes classified as receptor protein tyrosine kinase signaling in PANTHER. Ddr2, Axl, and Egfr are all receptor protein‐tyrosine kinases that all appear on the same GO list and show the same pattern of expression as Ntrk2 and Bdnf. Further, regarding the cell differentiation and development GO terms, there is evidence that voluntary running activates quiescent neural stem cells, which leads to proliferation of neural stem cells and intermediate progenitor cells and enhances differentiation (Gao et al. 2023). In addition, growth factors, specifically neurotrophins, are upregulated following exercise and promote neuronal development (Gao et al. 2023). This includes nerve growth factor, insulin‐like growth factor 1, vascular endothelial growth factor, and the hormone adiponectin. Last, the gonadotropin‐releasing hormone receptor (GnRHR) is a part of the hypothalamic–pituitary‐gonadal axis. Exercise‐trained rats had lower GnRH, leutinizing hormone, and follicle stimulating hormone, and higher testosterone in the hypothalamus than sedentary rats, suggesting a modulatory effect of exercise on this pathway (Khairy and Salama 2023).

Gene network analysis identified modules of genes that were significantly impacted by exercise. Of the 10 modules that were identified, two modules had genes with an overrepresentation of GO terms, modules 1 and 21. Some of the GO findings overlap with the exercise results from the differential expression analysis, such as signaling receptor activity/transmembrane receptor protein tyrosine kinase activity. Tyrosine Kinase Receptor Type 2 (Ntrk2) was identified in module 21. In addition, GO terms from module 1 and 21 related to transcription and translation, such as transcription factor activity, RNA metabolism/processing, and translation/ribosomes. There is evidence that exercise impacts both transcription and translation in the brain, predominately by upregulating expression of genes related to neuronal plasticity, learning, as well as memory (Fernandes et al. 2017). For example, exercise increases production of BDNF which leads to neuroprotective effects (Ding et al. 2011). As previously discussed, Ntrk2 and Bdnf were found to be significantly upregulated in control animals when exercised (Ntrk2, p = 0.0094; Bdnf, p = 0.048). Module GO terms associated with nervous system development and neuron differentiation were also identified. Exercise is known to impact these processes by promoting neurogenesis and synaptic plasticity, as well as increasing neurotrophic factors (Consorti et al. 2021). Last, metabolic/biosynthetic processes and respiratory electron transport were identified as module GO terms. Exercise is known to improve mitochondrial function by increasing electron transport chain activity and efficiency (Granata et al. 2018). Some or all of these effects may override binge‐induced degenerative processes.

The present study made use of bulk samples, which do not yield cell type‐specific gene expression information. Since single‐cell transcriptomics remain cost‐prohibitive for many researchers, deconvolution has been used successfully in order to obtain cell type‐specific information from bulk‐sequenced samples (Donovan et al. 2020). In the present study, CIBERSORTx was used to estimate cell type proportions in each sample as well as sample‐specific gene expression. The estimated cell type proportions did not significantly differ between exercise and alcohol groups. Differential expression analysis within each cell type identified a small number of DEGs within microglia and oligodendrocytes, and a similar pattern of expression as that seen in the bulk analysis, with exercise showing the biggest effect. This suggests that cell type specific expression differences between groups were not obscured by the use of bulk RNA‐Seq. In addition, the one exercise by alcohol interaction DEG, Zfp62, that was identified in the original bulk data was also identified within oligodendrocyte‐specific differential analysis. However, this may be due to oligodendrocytes having the most genes with estimated expression (allowing for increased power) and does not necessarily mean Zfp62 expression differences are limited to oligodendrocytes. Limitations of this exploratory deconvolution analysis include prediction confidence of CIBERSORTx, the number of cells available for analysis, and the use of a mouse reference dataset instead of rat. Despite the limitations, however, deconvolution of bulk RNAseq data appears to be a useful tool that warrants further exploration.

A limitation to this study is that following outlier sample removal, the experimental groups did not remain even. This disparity could potentially reduce power and impact differential expression analysis, although the high n per group (8–12) should minimize this. In addition, the present study assessed only protein‐coding RNAs. Future studies should examine noncoding RNAs, in order to develop a fuller picture of the transcriptomic effects of binge alcohol, exercise and their combination. Another possible limitation of this study concerns the timing of tissue collection following alcohol exposure versus exercise. Animals were sacrificed shortly after their final exercise session, but approximately 18 h after the last binge exposure. Therefore, the substantially greater number of DEGs observed in the exercise group compared to the alcohol group may be partially attributable to differences in sacrifice timing rather than purely to differences in the biological impact of each condition. In addition, the amount of exposure differed between the treatments, with a total of five binge exposures versus 17 for exercise, which may also have contributed to the disparity observed in gene expression. Future studies could shorten postexposure intervals or include multiple sacrifice time points to disentangle acute versus sustained transcriptional effects of alcohol and exercise as well as those potentially due to relative amount of exposure.

The present study expands our understanding of the hippocampal effects of binge alcohol, as most extant studies focus on chronic exposure. We did find a study of gene expression changes 4 h after ethanol injection that showed changes in 217 genes in the mouse hippocampus (Baker et al. 2017). In the present study, we found only seven genes altered by binge, but our timepoint was much later, approximately 18 h. It may be that binge exposure simply does not have a lingering effect on protein‐coding RNAs. We note, however, that we assessed gene expression after the fifth weekly binge exposure. It is possible that many more gene expression changes would have been seen after the initial weekly binge exposure (versus the fifth). A future direction is a direct comparison of gene expression changes after a single binge dose versus multiple (five or more). This will help to elucidate whether adaptations indicative of tissue‐level tolerance are occurring across repeated exposures. Overall, our results indicate that binge alcohol does not have a lasting direct effect on the hippocampal transcriptome, although it may modulate the transcriptomic effects of next‐day exercise.

Funding

This work was supported by National Institutes of Health, R01 AA025380, T32 DA015937.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Data S1: Supporting Information.

ACER-50-0-s001.docx (14.6KB, docx)

Data S2: Supporting Information.

ACER-50-0-s002.zip (96.3MB, zip)

Data S3: Supporting Information.

ACER-50-0-s005.xlsx (216KB, xlsx)

Figure S1: MultiQC data were used for quality control for all samples. (A) A principal components plot depicts sample similarity based on expression profiles. (B) A heatmap of overall expression in the samples also provides information on sample similarity. Both plots indicate that 5 samples are outliers, and they were therefore removed before further analyses.

ACER-50-0-s003.docx (419KB, docx)

Figure S2: A heatmap depicting the signature matrix created with Allen Smart‐Seq reference data and CIBERSORTx. The signature matrix identifies genes with expression levels that discriminate between cell types. The matrix was used to estimate cell type proportions and gene expression analysis in the bulk dataset.

ACER-50-0-s004.docx (103.9KB, docx)

Table S1: The 806 exercise DEGs were analyzed for overrepresentation of gene ontology (GO) terms using the PANTHER database, and the full list of GO terms is shown. Also shown are the full lists of GO terms associated with the genes in modules 1 and 21, which were identified by WGCNA of the main effect of exercise.

ACER-50-0-s006.xlsx (73.7KB, xlsx)

Table S2: Details of the exercise WGCNA module. A linear model using exercise identified 9 modules at a more liberal cutoff of padj. < 0.15. A positive fold change indicates an overall upregulation in exercised animals. The hub genes are the genes within each module with the highest connectivity.

ACER-50-0-s007.docx (216.2KB, docx)

Data Availability Statement

The data that support the findings of this study are openly available in GEO at https://www.ncbi.nlm.nih.gov/geo/, reference number GSE330623.

References

  1. Ansari, U. , Chen V., Sedighi R., et al. 2023. “Role of the UNC13 Family in Human Diseases: A Literature Review.” AIMS Neuroscience 10, no. 4: 388–400. 10.3934/Neuroscience.2023029. [DOI] [PMC free article] [PubMed] [Google Scholar]
  2. Baker, J. A. , Li J., Zhou D., et al. 2017. “Analyses of Differentially Expressed Genes After Exposure to Acute Stress, Acute Ethanol, or a Combination of Both in Mice.” Alcohol 58: 139–151. 10.1016/j.alcohol.2016.08.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  3. Bittner, N. , Jockwitz C., Muhleisen T. W., et al. 2019. “Combining Lifestyle Risks to Disentangle Brain Structure and Functional Connectivity Differences in Older Adults.” Nature Communications 10, no. 1: 621. 10.1038/s41467-019-08500-x. [DOI] [PMC free article] [PubMed] [Google Scholar]
  4. Booher, W. C. , Reyes Martinez G. J., and Ehringer M. A.. 2020. “Behavioral and Neuronal Interactions Between Exercise and Alcohol: Sex and Genetic Differences.” Genes, Brain, and Behavior 19, no. 3: e12632. 10.1111/gbb.12632. [DOI] [PubMed] [Google Scholar]
  5. Choi, M. R. , Han J. S., Chai Y. G., Jin Y. B., Lee S. R., and Kim D. J.. 2020. “Gene Expression Profiling in the Hippocampus of Adolescent Rats After Chronic Alcohol Administration.” Basic & Clinical Pharmacology & Toxicology 126, no. 4: 389–398. 10.1111/bcpt.13342. [DOI] [PubMed] [Google Scholar]
  6. Consorti, A. , di Marco I., and Sansevero G.. 2021. “Physical Exercise Modulates Brain Physiology Through a Network of Long‐ and Short‐Range Cellular Interactions.” Frontiers in Molecular Neuroscience 14: 710303. 10.3389/fnmol.2021.710303. [DOI] [PMC free article] [PubMed] [Google Scholar]
  7. Crews, F. T. , and Nixon K.. 2009. “Mechanisms of Neurodegeneration and Regeneration in Alcoholism.” Alcohol and Alcoholism 44, no. 2: 115–127. 10.1093/alcalc/agn079. [DOI] [PMC free article] [PubMed] [Google Scholar]
  8. Crews, F. T. , Nixon K., and Wilkie M. E.. 2004. “Exercise Reverses Ethanol Inhibition of Neural Stem Cell Proliferation.” Alcohol 33, no. 1: 63–71. 10.1016/j.alcohol.2004.04.005. [DOI] [PubMed] [Google Scholar]
  9. Danecek, P. , Bonfield J. K., Liddle J., et al. 2021. “Twelve Years of SAMtools and BCFtools.” GigaScience 10, no. 2: giab008. 10.1093/gigascience/giab008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  10. Darlington, T. M. , McCarthy R. D., Cox R. J., and Ehringer M. A.. 2014. “Mesolimbic Transcriptional Response to Hedonic Substitution of Voluntary Exercise and Voluntary Ethanol Consumption.” Behavioural Brain Research 259: 313–320. 10.1016/j.bbr.2013.11.011. [DOI] [PMC free article] [PubMed] [Google Scholar]
  11. Darlington, T. M. , McCarthy R. D., Cox R. J., Miyamoto‐Ditmon J., Gallego X., and Ehringer M. A.. 2016. “Voluntary Wheel Running Reduces Voluntary Consumption of Ethanol in Mice: Identification of Candidate Genes Through Striatal Gene Expression Profiling.” Genes, Brain, and Behavior 15, no. 5: 474–490. 10.1111/gbb.12294. [DOI] [PMC free article] [PubMed] [Google Scholar]
  12. Davis, D. N. , Strong M. D., Chambers E., et al. 2021. “A Role for Zinc Transporter Gene SLC39A12 in the Nervous System and Beyond.” Gene 799: 145824. 10.1016/j.gene.2021.145824. [DOI] [PMC free article] [PubMed] [Google Scholar]
  13. Denham, A. N. , Drake J., Gavrilov M., Taylor Z. N., Bacanu S. A., and Vladimirov V. I.. 2022. “Long Non‐Coding RNAs: The New Frontier Into Understanding the Etiology of Alcohol Use Disorder.” Non‐Coding RNA 8, no. 4: 59. 10.3390/ncrna8040059. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. di Tommaso, P. , Chatzou M., Floden E. W., Barja P. P., Palumbo E., and Notredame C.. 2017. “Nextflow Enables Reproducible Computational Workflows.” Nature Biotechnology 35, no. 4: 316–319. 10.1038/nbt.3820. [DOI] [PubMed] [Google Scholar]
  15. Ding, Q. , Ying Z., and Gomez‐Pinilla F.. 2011. “Exercise Influences Hippocampal Plasticity by Modulating Brain‐Derived Neurotrophic Factor Processing.” Neuroscience 192: 773–780. 10.1016/j.neuroscience.2011.06.032. [DOI] [PMC free article] [PubMed] [Google Scholar]
  16. Dobin, A. , Davis C. A., Schlesinger F., et al. 2013. “STAR: ultrafast universal RNA‐seq aligner.” Bioinformatics 29, no. 1: 15–21. 10.1093/bioinformatics/bts635. [DOI] [PMC free article] [PubMed] [Google Scholar]
  17. Donovan, M. K. R. , D'Antonio‐Chronowska A., D'Antonio M., and Frazer K. A.. 2020. “Cellular Deconvolution of GTEx Tissues Powers Discovery of Disease and Cell‐Type Associated Regulatory Variants.” Nature Communications 11, no. 1: 955. 10.1038/s41467-020-14561-0. [DOI] [PMC free article] [PubMed] [Google Scholar]
  18. Ehringer, M. A. , Hoft N. R., and Zunhammer M.. 2009. “Reduced Alcohol Consumption in Mice With Access to a Running Wheel.” Alcohol 43, no. 6: 443–452. 10.1016/j.alcohol.2009.06.003. [DOI] [PubMed] [Google Scholar]
  19. Ekiz, A. 2024. “A guide to designs and contrasts in DESeq2.” https://www.atakanekiz.com/technical/a‐guide‐to‐designs‐and‐contrasts‐in‐DESeq2/.
  20. Ewels, P. , Magnusson M., Lundin S., and Kaller M.. 2016. “MultiQC: Summarize Analysis Results for Multiple Tools and Samples in a Single Report.” Bioinformatics 32, no. 19: 3047–3048. 10.1093/bioinformatics/btw354. [DOI] [PMC free article] [PubMed] [Google Scholar]
  21. Ewels, P. A. , Peltzer A., Fillinger S., et al. 2020. “The Nf‐Core Framework for Community‐Curated Bioinformatics Pipelines.” Nature Biotechnology 38, no. 3: 276–278. 10.1038/s41587-020-0439-x. [DOI] [PubMed] [Google Scholar]
  22. Fernandes, J. , Arida R. M., and Gomez‐Pinilla F.. 2017. “Physical Exercise as an Epigenetic Modulator of Brain Plasticity and Cognition.” Neuroscience and Biobehavioral Reviews 80: 443–456. 10.1016/j.neubiorev.2017.06.012. [DOI] [PMC free article] [PubMed] [Google Scholar]
  23. Gallego, X. , Cox R. J., Funk E., Foster R. A., and Ehringer M. A.. 2015. “Voluntary Exercise Decreases Ethanol Preference and Consumption in C57BL/6 Adolescent Mice: Sex Differences and Hippocampal BDNF Expression.” Physiology & Behavior 138: 28–36. 10.1016/j.physbeh.2014.10.008. [DOI] [PMC free article] [PubMed] [Google Scholar]
  24. Gao, Y. , Syed M., and Zhao X. Y.. 2023. “Mechanisms Underlying the Effect of Voluntary Running on Adult Hippocampal Neurogenesis.” Hippocampus 33, no. 4: 373–390. 10.1002/hipo.23520. [DOI] [PMC free article] [PubMed] [Google Scholar]
  25. Granata, C. , Jamnick N. A., and Bishop D. J.. 2018. “Training‐Induced Changes in Mitochondrial Content and Respiratory Function in Human Skeletal Muscle.” Sports Medicine 48, no. 8: 1809–1828. 10.1007/s40279-018-0936-y. [DOI] [PubMed] [Google Scholar]
  26. Ji, Z. , Yuan L., Lu X., Ding H., Luo J., and Ke Z. J.. 2018. “Binge Alcohol Exposure Causes Neurobehavioral Deficits and GSK3beta Activation in the Hippocampus of Adolescent Rats.” Scientific Reports 8, no. 1: 3088. 10.1038/s41598-018-21341-w. [DOI] [PMC free article] [PubMed] [Google Scholar]
  27. Kelso, M. L. , Liput D. J., Eaves D. W., and Nixon K.. 2011. “Upregulated Vimentin Suggests New Areas of Neurodegeneration in a Model of an Alcohol Use Disorder.” Neuroscience 197: 381–393. 10.1016/j.neuroscience.2011.09.019. [DOI] [PMC free article] [PubMed] [Google Scholar]
  28. Khairy, E. Y. , and Salama O. A.. 2023. “Effect of Exercise on the Hypothalamic‐Pituitary‐Gonadal Axis in a Rat Model of Alzheimer's Disease.” Scientific Reports 13, no. 1: 14300. 10.1038/s41598-023-41415-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  29. Kittke, V. , Zhao C., Lam D. D., et al. 2024. “RLS‐Associated MEIS Transcription Factors Control Distinct Processes in Human Neural Stem Cells.” Scientific Reports 14, no. 1: 28986. 10.1038/s41598-024-80266-9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  30. Kopylova, E. , Noe L., and Touzet H.. 2012. “SortMeRNA: Fast and Accurate Filtering of Ribosomal RNAs in Metatranscriptomic Data.” Bioinformatics 28, no. 24: 3211–3217. 10.1093/bioinformatics/bts611. [DOI] [PubMed] [Google Scholar]
  31. Kubota, M. , Nakazaki S., Hirai S., Saeki N., Yamaura A., and Kusaka T.. 2001. “Alcohol Consumption and Frontal Lobe Shrinkage: Study of 1432 Non‐Alcoholic Subjects.” Journal of Neurology, Neurosurgery, and Psychiatry 71, no. 1: 104–106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  32. Langfelder, P. , and Horvath S.. 2008. “WGCNA: An R Package for Weighted Correlation Network Analysis.” BMC Bioinformatics 9: 559. 10.1186/1471-2105-9-559. [DOI] [PMC free article] [PubMed] [Google Scholar]
  33. Leasure, J. L. , Neighbors C., Henderson C. E., and Young C. M.. 2015. “Exercise and Alcohol Consumption: What we Know, What we Need to Know, and Why It Is Important.” Frontiers in Psychiatry 6: 156. 10.3389/fpsyt.2015.00156. [DOI] [PMC free article] [PubMed] [Google Scholar]
  34. Li, B. , and Dewey C. N.. 2011. “RSEM: Accurate Transcript Quantification From RNA‐Seq Data With or Without a Reference Genome.” BMC Bioinformatics 12: 323. 10.1186/1471-2105-12-323. [DOI] [PMC free article] [PubMed] [Google Scholar]
  35. Love, M. I. , Huber W., and Anders S.. 2014. “Moderated Estimation of Fold Change and Dispersion for RNA‐Seq Data With DESeq2.” Genome Biology 15, no. 12: 550. 10.1186/s13059-014-0550-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  36. Marquez‐Galera, A. , de la Prida L. M., and Lopez‐Atalaya J. P.. 2022. “A Protocol to Extract Cell‐Type‐Specific Signatures From Differentially Expressed Genes in Bulk‐Tissue RNA‐Seq.” STAR Protocols 3, no. 1: 101121. 10.1016/j.xpro.2022.101121. [DOI] [PMC free article] [PubMed] [Google Scholar]
  37. Maynard, M. E. , Barton E. A., Robinson C. R., Wooden J. I., and Leasure J. L.. 2018. “Sex Differences in Hippocampal Damage, Cognitive Impairment, and Trophic Factor Expression in an Animal Model of an Alcohol Use Disorder.” Brain Structure & Function 223, no. 1: 195–210. 10.1007/s00429-017-1482-3. [DOI] [PMC free article] [PubMed] [Google Scholar]
  38. McClintick, J. N. , McBride W. J., Bell R. L., et al. 2018. “Gene Expression Changes in the Ventral Hippocampus and Medial Prefrontal Cortex of Adolescent Alcohol‐Preferring (P) Rats Following Binge‐Like Alcohol Drinking.” Alcohol 68: 37–47. 10.1016/j.alcohol.2017.09.002. [DOI] [PMC free article] [PubMed] [Google Scholar]
  39. Neighbors, C. , Leasure J. L., Shank F., et al. 2024. “Physical Activity as a Moderator of the Association Between Alcohol Consumption and Hangovers.” Addictive Behaviors 159: 108145. 10.1016/j.addbeh.2024.108145. [DOI] [PubMed] [Google Scholar]
  40. Newman, A. M. , Steen C. B., Liu C. L., et al. 2019. “Determining Cell Type Abundance and Expression From Bulk Tissues With Digital Cytometry.” Nature Biotechnology 37, no. 7: 773–782. 10.1038/s41587-019-0114-2. [DOI] [PMC free article] [PubMed] [Google Scholar]
  41. NIAAA . 2004. “Binge drinking defined.” NIAAA Newsletter 3, no. 3: 1. [Google Scholar]
  42. Ozburn, A. R. , Harris R. A., and Blednov Y. A.. 2008. “Wheel Running, Voluntary Ethanol Consumption, and Hedonic Substitution.” Alcohol 42, no. 5: 417–424. 10.1016/j.alcohol.2008.04.006. [DOI] [PMC free article] [PubMed] [Google Scholar]
  43. Patrick, M. E. , and Azar B.. 2018. “High‐Intensity Drinking.” Alcohol Research: Current Reviews 39, no. 1: 49–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Patro, R. , Duggal G., Love M. I., Irizarry R. A., and Kingsford C.. 2017. “Salmon Provides Fast and Bias‐Aware Quantification of Transcript Expression.” Nature Methods 14, no. 4: 417–419. 10.1038/nmeth.4197. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Schilder, B. M. , and Skene N. G.. 2022. “Orthogene: an R package for easy mapping of orthologous genes across hundreds of species.” 10.18129/B9.bioc.orthogene. R package version, Bioconductor. [DOI]
  46. Smith, T. , Heger A., and Sudbery I.. 2017. “UMI‐Tools: Modeling Sequencing Errors in Unique Molecular Identifiers to Improve Quantification Accuracy.” Genome Research 27, no. 3: 491–499. 10.1101/gr.209601.116. [DOI] [PMC free article] [PubMed] [Google Scholar]
  47. Su, X. , Xie L., Li J., Tian X. Y., Lin B., and Chen M. H.. 2023. “Exploring Molecular Signatures Related to the Mechanism of Aging in Different Brain Regions by Integrated Bioinformatics.” Frontiers in Molecular Neuroscience 16: 1133106. 10.3389/fnmol.2023.1133106. [DOI] [PMC free article] [PubMed] [Google Scholar]
  48. Sutton, G. J. , Poppe D., Simmons R. K., et al. 2022. “Comprehensive Evaluation of Deconvolution Methods for Human Brain Gene Expression.” Nature Communications 13, no. 1: 1358. 10.1038/s41467-022-28655-4. [DOI] [PMC free article] [PubMed] [Google Scholar]
  49. Thomas, P. D. , Ebert D., Muruganujan A., Mushayahama T., Albou L. P., and Mi H.. 2022. “PANTHER: Making Genome‐Scale Phylogenetics Accessible to All.” Protein Science 31, no. 1: 8–22. 10.1002/pro.4218. [DOI] [PMC free article] [PubMed] [Google Scholar]
  50. Tong, L. , Shen H., Perreau V. M., Balazs R., and Cotman C. W.. 2001. “Effects of Exercise on Gene‐Expression Profile in the Rat Hippocampus.” Neurobiology of Disease 8, no. 6: 1046–1056. 10.1006/nbdi.2001.0427. [DOI] [PubMed] [Google Scholar]
  51. Uittenbogaard, M. , Baxter K. K., and Chiaramello A.. 2010. “NeuroD6 Genomic Signature Bridging Neuronal Differentiation to Survival via the Molecular Chaperone Network.” Journal of Neuroscience Research 88, no. 1: 33–54. 10.1002/jnr.22182. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Wang, L. , Wang S., and Li W.. 2012. “RSeQC: Quality Control of RNA‐Seq Experiments.” Bioinformatics 28, no. 16: 2184–2185. 10.1093/bioinformatics/bts356. [DOI] [PubMed] [Google Scholar]
  53. Wang, Y. , Ma Y., Zhong Q., Song B., and Liu Q.. 2024. “Transcriptomic Analysis of Rat Brain Response to Alternating Current Electrical Stimulation: Unveiling Insights via Single‐Nucleus RNA Sequencing.” MedComm (2020) 5, no. 4: e514. 10.1002/mco2.514. [DOI] [PMC free article] [PubMed] [Google Scholar]
  54. West, R. K. , Najjar L. Z., and Leasure J. L.. 2019. “Exercise‐Driven Restoration of the Alcohol‐Damaged Brain.” International Review of Neurobiology 147: 219–267. 10.1016/bs.irn.2019.07.003. [DOI] [PubMed] [Google Scholar]
  55. West, R. K. , Rodgers S. P., and Leasure J. L.. 2021. “Neural Perturbations Associated With Recurrent Binge Alcohol in Male and Female Rats.” Alcoholism, Clinical and Experimental Research 45, no. 2: 365–374. 10.1111/acer.14529. [DOI] [PMC free article] [PubMed] [Google Scholar]
  56. West, R. K. , Wooden J. I., Barton E. A., and Leasure J. L.. 2019. “Recurrent Binge Ethanol Is Associated With Significant Loss of Dentate Gyrus Granule Neurons in Female Rats Despite Concomitant Increase in Neurogenesis.” Neuropharmacology 148: 272–283. 10.1016/j.neuropharm.2019.01.016. [DOI] [PubMed] [Google Scholar]
  57. Wilson, S. , Bair J. L., Thomas K. M., and Iacono W. G.. 2017. “Problematic Alcohol Use and Reduced Hippocampal Volume: A Meta‐Analytic Review.” Psychological Medicine 47, no. 13: 2288–2301. 10.1017/S0033291717000721. [DOI] [PMC free article] [PubMed] [Google Scholar]
  58. Yao, Z. , Liu H., Xie F., et al. 2021. “A Transcriptomic and Epigenomic Cell Atlas of the Mouse Primary Motor Cortex.” Nature 598, no. 7879: 103–110. 10.1038/s41586-021-03500-8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  59. Zhu, Q. Q. , Song L., Peng G. D., et al. 2014. “The Transcription Factor Pou3f1 Promotes Neural Fate Commitment via Activation of Neural Lineage Genes and Inhibition of External Signaling Pathways.” eLife 3: e02224. 10.7554/eLife.02224. [DOI] [PMC free article] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Data S1: Supporting Information.

ACER-50-0-s001.docx (14.6KB, docx)

Data S2: Supporting Information.

ACER-50-0-s002.zip (96.3MB, zip)

Data S3: Supporting Information.

ACER-50-0-s005.xlsx (216KB, xlsx)

Figure S1: MultiQC data were used for quality control for all samples. (A) A principal components plot depicts sample similarity based on expression profiles. (B) A heatmap of overall expression in the samples also provides information on sample similarity. Both plots indicate that 5 samples are outliers, and they were therefore removed before further analyses.

ACER-50-0-s003.docx (419KB, docx)

Figure S2: A heatmap depicting the signature matrix created with Allen Smart‐Seq reference data and CIBERSORTx. The signature matrix identifies genes with expression levels that discriminate between cell types. The matrix was used to estimate cell type proportions and gene expression analysis in the bulk dataset.

ACER-50-0-s004.docx (103.9KB, docx)

Table S1: The 806 exercise DEGs were analyzed for overrepresentation of gene ontology (GO) terms using the PANTHER database, and the full list of GO terms is shown. Also shown are the full lists of GO terms associated with the genes in modules 1 and 21, which were identified by WGCNA of the main effect of exercise.

ACER-50-0-s006.xlsx (73.7KB, xlsx)

Table S2: Details of the exercise WGCNA module. A linear model using exercise identified 9 modules at a more liberal cutoff of padj. < 0.15. A positive fold change indicates an overall upregulation in exercised animals. The hub genes are the genes within each module with the highest connectivity.

ACER-50-0-s007.docx (216.2KB, docx)

Data Availability Statement

The data that support the findings of this study are openly available in GEO at https://www.ncbi.nlm.nih.gov/geo/, reference number GSE330623.

Articles from Alcohol, Clinical & Experimental Research are provided here courtesy of Wiley

RESOURCES