Adolescence is a period marked by physical, behavioral, social, and cognitive changes. Cognitive changes during this time include heightened reward sensitivity, sensation seeking, impulsivity, and a diminished self-control to inhibit behaviors, all of which contribute to increased participation in risky behaviors, including the initiation and escalation of alcohol use (Spear, 2000; Casey et al., 2008; Romer et al., 2017; Lees et al., 2020). While alcohol consumption in underage youth has been on the decline over the past few decades, about 15% (5.9 million) of Americans between the ages of 12 and 20 reported current alcohol consumption in 2022 (SAMHSA, 2023). Approximately 8.3% of adolescents ages 12 to 20 (3.2 million teens) did so in the form of binge drinking in the past month, defined as having five or more drinks on one occasion (SAMHSA, 2023). Binge drinking during adolescence can be particularly harmful, as the brain is still undergoing processes such as myelination and synaptic pruning (Spear, 2000). Alcohol use can delay or disrupt this ongoing development, resulting in many structural, cognitive, and behavioral effects (De Bellis et al., 2005; Medina et al., 2008; Bava et al., 2013; Squeglia et al., 2015; Lees et al., 2020). Adolescents respond differently to alcohol than their adult counterparts, showing increased sensitivity to the rewarding aspects of alcohol, and decreased sensitivity to the aversive aspects of alcohol (Spear, 2014). This enables adolescents to increase their alcohol consumption, giving rise to the binge drinking behavior commonly seen among this age group, while subsequently increasing the risk for developing an alcohol use disorder (AUD) later in life (Grant and Dawson, 1997; Spear, 2000).

Human brain imaging studies have shown that consuming alcohol in adolescence can disrupt myelination and cause brain structural changes, finding reductions in prefrontal cortex (PFC) (De Bellis et al., 2005; Medina et al., 2008), white matter (Luciana et al., 2013; Pfefferbaum et al., 2018; El Marroun et al., 2021) and myelin fiber tract volumes and integrity (McQueeny et al., 2009; Bava et al., 2013; Elofson et al., 2013). Binge drinking in teens is associated with thinning of cortical and subcortical regions that play a key role in memory, attention, language, awareness and consciousness (Cservenka and Brumback, 2017). Likewise, teens with an alcohol misuse disorder encounter cognitive deficits in visuospatial functioning or reading recognition, some of which persist after abstinence (Brown and Tapert, 2004). This effect may be dose-dependent, as diminished PFC white matter was negatively correlated with alcohol consumption (De Bellis et al., 2005). The age of onset for binge drinking also affects the susceptibility to these white matter changes. In a study of adult veterans, those who reported binge drinking at an early age (<15 years old, i.e., early onset binge drinkers) had decreased white matter integrity in frontostriatal circuits that mediate inhibitory control compared to late onset binge drinkers (Knoff et al., 2023).

Similar outcomes have been characterized in rodent studies, as adolescent binge ethanol has induced frontal neurodegeneration (Crews et al., 2000), loss of neurogenesis (Crews et al., 2006), decreased myelin expression (Montesinos et al., 2015; Wolstenholme et al., 2017), altered myelin structure (Tavares et al., 2019), and alterations in oligodendrocyte (OLG) differentiation (Guo et al., 2021). Myelin fiber density is decreased (Vargas et al., 2014) and the intralaminar sheath is damaged in young rodents exposed to intermittent ethanol and persists into adulthood (Montesinos et al., 2015). Lowered myelin expression and formation disrupts axonal integrity and can put individuals at risk for continued AUDs (Bava et al., 2009; Harper, 2009). In adult animals, chronic ethanol delays OLG maturation (Chiappelli et al., 1991), reduces the proliferation, differentiation, and survival of premyelinating OLGs, and decreases myelin basic protein (MBP) expression in the mPFC (Richardson et al., 2009; Kim et al., 2015). Some of these changes reverse upon protracted ethanol abstinence (Navarro and Mandyam, 2015). Similarly, in adolescent rodents, mature OLGs and grey matter myelination in the infralimbic and prelimbic PFC decreased shortly after binge ethanol (from postnatal day (PND) 26–52), but increased in ethanol groups in adulthood during abstincence (Liu et al., 2021). In a shorter adolescent binge paradigm (PND 28–37), grey matter myelin in the medial PFC (mPFC) and hippocampus was reduced after ethanol binges, but this did not persist in adult mPFC (Rice et al., 2019). Adult ethanol decreased the number of newly generated mature OLGs and there was reduced expression of HDAC1 in new oligodendrocyte precursor cells (OPCs), suggesting that ethanol alters the epigenetic programming on OLG maturation (Guo et al., 2021). While it is known that alcohol can alter the developmental trajectory of the brain in regards to myelination and structure, the precise trajectory of these changes and the molecular factors underlying their development, however, remain elusive.

Epigenetic regulation of gene expression provides an attractive mechanism for driving changes in myelin structure after binge ethanol. Epigenetic modifications can positively or negatively impact gene transcription, and are dynamic marks that are affected by environmental stimuli, such as stress or drugs of abuse (Jaenisch and Bird, 2003; Liu et al., 2008; Karlić et al., 2010; Hitchcock and Lattal, 2014). There is a growing body of literature that demonstrates epigenetic dysregulation in the brain after alcohol exposure in both humans (Ponomarev et al., 2012) and rodent models (Barnea-Goraly et al., 2005; Pascual et al., 2009, 2012; Warnault et al., 2013; Barbier et al., 2015; Pandey et al., 2015; Montesinos et al., 2016; Wolstenholme et al., 2017; Kyzar et al., 2019; Bohnsack et al., 2022; Brocato and Wolstenholme, 2023) (as reviewed in Brocato and Wolstenholme, 2021), and this dysregulation provides a pathway that could result in the loss of myelin.

OLGs are the main myelinating cell in the brain, and the differentiation of OLGs into their mature, myelinating form is under epigenetic control. OPCs arise from multipotent neuroepithelial progenitor cells in the developing central nervous system (CNS) (Emery and Lu, 2015). OPCs are mobile and highly proliferative, and once they arrive in their final position, they can become OLGs via terminal differentiation (Emery and Lu, 2015). DNA methylation (Jackson et al., 2004), histone acetylation (Cunliffe and Casaccia-Bonnefil, 2006; Ye et al., 2009; Dai et al., 2015b; Zhang et al., 2018) and histone methylation (Sher et al., 2008; He et al., 2017) have all been shown to play a role in OLG differentiation processes, and epigenetic disruption of these processes underlies myelin loss in pathologic states such as multiple sclerosis (Pedre et al., 2011; Huynh et al., 2014) and Rett Syndrome (Nguyen et al., 2013; Sharma et al., 2015). OPCs are electrically responsive cells, and the differentiation of OPCs into mature, myelinating OLGs is characterized by progressive repression of genes associated with electrical responsiveness and activation of genes responsible for producing structural components of myelin (Liu et al., 2015). Tri-methylation of histone 3, lysine 9 (H3K9me3) was found to be essential for this process (Liu et al., 2015), and an increase in H3K9me3 is correlated with an increase in myelin (Chen et al., 2020). Increased H3K9me3 at specific loci is responsible for repression of genes related to membrane excitability, and was found to reduce the electrical activity of OPCs to allow for their differentiation into functional OLGs (Liu et al., 2015).

Alterations in H3K9me2 have been associated with drugs of abuse, such as cocaine (Maze et al., 2010), and H3K9me3 dysregulation was found to be a result of adolescent binge ethanol (Wolstenholme et al., 2017). H3K9me1 and Prdm2, the enzyme responsible for H3K9 mono-methylation, were decreased in the PFC of adult dependent rats (Barbier et al., 2016). Reduction of Prdm2 in the PFC of non-dependent rats increased ethanol drinking even in the presence of quinine adulteration, displaying phenotypes characteristic of alcohol dependence (Barbier et al., 2016). The histone demethylase, Lsd1, was altered by adolescent ethanol exposure and H3K9me2 protein was increased in the amygdala of adults following adolescent ethanol (Kyzar et al., 2016). Previously, our lab used an unbiased transcriptomics approach to investigate the possible mechanisms of reduced myelin associated with adolescent binge ethanol (Wolstenholme et al., 2017). Altered levels of histone methyltransferases specific for the H3K9me3 mark, as well as decreased expression of myelin-related genes were found in the PFC of adolescent mice treated with binge ethanol (Wolstenholme et al., 2017). Together, this data suggests that adolescent binge ethanol negatively impacts oligodendrocytes, and that this deficit may be epigenetically regulated through an oligodendrocyte differentiation pathway. To further investigate ethanol-induced epigenetic changes and learn how they influence myelin, we used a series of chromatin immunoprecipitation and qPCR assays to assess H3K9me3 and gene expression changes in the PFC after adolescent binge ethanol. We conducted chromatin immunoprecipitation coupled to sequencing (ChIP-seq) for H3K9me3 to identify global changes to H3K9me3 in the PFC after adolescent binge ethanol treatment. We also conducted qPCR time courses to measure myelin-related gene expression changes in the PFC. To focus our attention on oligodendrocyte differentiation-specific effects, we used O4+ magnetic cell separation to analyze OPC and OLG lineage cells within the PFC. O4+ cells were assayed in a combined time course and dose response to assess expression changes in genes involved in the transcriptional and epigenetic regulation of myelin and the structural components of myelin. These assays helped to unveil potential genes that may be contributing to adolescent binge ethanol-related myelin loss.

These studies were designed to interrogate the effects of adolescent binge ethanol on the epigenetic regulation of genes in the frontal cortex with the aim of understanding the dysregulation of oligodendrocyte maturation that may lead to myelin and white matter deficits. ChIP-Seq analysis of H3K9me3 in bulk PFC tissue of both sexes identified differential H3K9me3 at genes involved in synaptic transmission and synaptic function. Males also showed differential H3K9me3 at genes involved in ion channels, potassium channels and synapse maturation. Females, however, showed ethanol-induced H3K9me3 binding at genes involved in synapse assembly, neuron differentiation, and deacetylase activity. Our time course to assess myelin-related gene expression at 4 and 24 h after the last ethanol dose replicated our earlier findings that ethanol decreases myelin-related genes largely 24 h after the last dose and this effect appears to be stronger in males.

Interestingly, we saw that ethanol dysregulated the trimethylation of both H3K9 and H3K36 at about 8% of the same loci in males (176 genes), and about 15% of the same gene loci in females (308 genes). This was surprising given the vastly different functions of the H3K36me3 mark, typically associated with active transcription, and the H3K9me3 mark, typically associated with heterochromatin (Barral et al., 2022). It could be possible that these loci showed changes in both histone methylation marks, as they are transitioning from an active to a repressive state or vice versa due to adolescent binge ethanol treatment. Gene ontology analysis was performed on the genes that overlapped between the H3K36 and H3K9me3 ChIP-seq analyses. In the 176 genes that showed differential trimethylation of both H3K36 and H3K9 in males, top gene ontology categories included transmitter-gated channel activity, channel regulator activity, and synapse assembly. In females, 308 genes showed differential trimethylation of both H3K36 and H3K9 due to adolescent binge ethanol; top gene ontology categories included GPI-linked ephrin receptor activity, glutamate receptor activity, and synapse assembly. Of these 176 genes in males and 308 genes in females that overlapped between the ChIP-seq analyses, 56 of those genes showed differential trimethylation of H3K36 and H3K9 in both males and females. These lists of genes can be found in Supplementary Table S8. Again, gene ontology analysis showed that these genes largely had synaptic-related functions. Adolescent ethanol induced changes in H3K9me3 at neurexin1 (Nrxn1) in both males and females, a gene crucial for synaptic assembly (Gomez et al., 2021). This gene was also cryptically transcribed by H3K36me3 in the same tissue (Brocato and Wolstenholme, 2023). One study found that accumulation of H3K9me3 allowed for alternative splicing of Nrxn1, and reducing H3K9me3 at Nrxn1 altered the spliced isoform in the hippocampus, impacting the stability of established memories (Ding et al., 2017). Interestingly, our previous study using the same tissue found that Nrxn1 also showed differential exon use after adolescent binge ethanol, indicating alternative splicing may be occurring at Nrxn1 in the PFC following ethanol exposure (Brocato and Wolstenholme, 2023). Dysregulation of Nrxn1 by adolescent binge ethanol may negatively impact adolescent neurodevelopment in the PFC.

We overlaid our H3K9me3 ChIP data with RNA-seq data that was generated from the same tissue (Brocato and Wolstenholme, 2023) to ask whether there was a pattern between H3K9me3 epigenetic regulation at differentially bound loci and significant differential gene expression due to adolescent ethanol exposure. Given H3K9me3’s repressive role (Kim and Kim, 2012), we expected to find that genes with decreased H3K9me3 due to ethanol would be the same genes where we saw increased gene expression. However, the overlap of our male differentially bound genes (n = 691) and male DEGs (n = 342 genes at p < 0.01) only resulted in 7 genes (Plcl1, Gnpda2, Tnc, Luzp2, Insig1, Magohb, Dab2), and the overlap of our female differentially bound genes (n = 1,266) and female DEGs (n = 376 genes at p < 0.01) only resulted in 15 genes (Tceal6, Ush2a, Ano5, Oprk1, Arhgef6, Cfh, Insig1, Itgb1, Eln, C4b, Car8, Sema3d, Adamts6, Igsf1, Gpr101). These overlaps show there is little correlation between H3K9me3 differentially bound regions and gene expression. In a survey to map the location of H3K9me3 in human CD4+ cells, Rosenfeld et al. (2009), found that H3K9me3 was not enriched at either genes or in heterochromatin. The presence of H3K9me3 at the 3′ end of a gene (as opposed to the gene body or promoter) was associated with decreased expression, but not strongly, suggesting that H3K9me3 may play a role in chromatin organization (Rosenfeld et al., 2009).

We also predicted that ethanol would decrease H3K9me3 at more genes rather than induce H3K9me3, given how ethanol interacts with the methionine metabolism pathway to influence methylation reactions (Brocato and Wolstenholme, 2021). Ethanol is known to cause a decrease in methylation through its inhibition of enzymes responsible for producing S-adenosylmethionine, the sole methyl donor for all methylation reactions in the cell (Brocato and Wolstenholme, 2021). However, we found that in both males and females, a majority of ethanol-induced changes to H3K9me3 led to increased binding of genes in the ethanol exposed mice. These unexpected results may be due to ethanol’s ability to alter H3K9me3-mediated chromatin regulation and gene expression through separate processes that are not necessarily linked. Ethanol can impact chromatin regulation through its metabolites (Mews et al., 2019; Brocato and Wolstenholme, 2021), impacts a number of receptor systems (Abrahao et al., 2017), and can also move intracellularly to impact a number of intracellular proteins (Abrahao et al., 2017). While changes in chromatin accessibility are typically associated with corresponding changes to gene expression, ethanol may not produce such predictable changes given its impact on multiple pathways and cell processes. For example, ethanol may reduce methylation by reducing the pool of S-adenosylmethionine present for methylation reactions to occur, which would reduce levels of H3K9me3. Simultaneously, ethanol may impact certain transcription factors, such as CREB (Yang et al., 1998; Soares-Simi et al., 2013; Zhang et al., 2018) to reduce gene expression levels at specific genes. In this case, even if H3K9me3 is reduced, transcription factors or other important cell signaling molecules could be negatively impacted by ethanol and unable to initiate transcription, even when chromatin is left accessible by the lack of H3K9me3. Further studies are needed to uncover this complicated relationship of chromatin regulation and gene expression in the presence of ethanol.

Initially, we hypothesized that adolescent binge ethanol would induce H3K9me3 occupancy at specific membrane excitability-related genes, given the data presented in Liu et al. (2015). Liu et al. showed that H3K9me3 must be present at specific genes regulating membrane excitability to reduce their gene expression and allow oligodendrocyte precursor cells to mature into myelin-forming oligodendrocytes (Liu et al., 2015). While we did not see H3K9me3 changes in these specific genes (Supplementary Table S3), this could be due to the tissue type used in the Liu et al. (2015) study—while we used adolescent mouse prefrontal cortex tissue, H3K9me3 changes were found by Liu et al. (2015) at membrane excitability related genes in cultured OPCs or OLGs taken from rat cortex (PND 1). Furthermore, it is possible that changes in H3K9me3 at genes directly related to oligodendrocyte differentiation were not detectable in bulk tissue. Performing single-cell ChIP-seq or ChIP-seq in O4+ cells would allow us to gain a better understanding of the genes that H3K9me3 regulates in OPCs and OLGs, and provide evidence that ethanol acts through H3K9me3 to reduce myelin-related gene expression. It is also possible that H3K9me3 may be regulating other membrane excitability related genes in adolescent mice. In our ChIP-seq study, we found that adolescent ethanol altered trimethylation of H3K9 at other ion channels, glutamate receptors, and synaptic proteins in both males and females, including Grin3a, Gria4, Grid2, Grik2, Grm3, Glra3, Kcnh7, Nrxn1, and Cntn5 (Supplementary Table S3). The gene expression values for these genes from our previous RNA-seq study (Brocato and Wolstenholme, 2023) are shown in Supplementary Figure S4. Specifically, Kcnh7, Nrxn1 and Cntn5 have been associated with myelin formation or phenotypes (Boscia et al., 2021; Dauar et al., 2022; Sebastian et al., 2022), and glutamate signaling has been shown to be important to oligodendrocyte development and survival (Kougioumtzidou et al., 2017; Turan et al., 2021). Changes to the regulation of these genes and others involved in synaptic function could impact the progression of oligodendrocyte maturation and play a role in ethanol-induced myelin reduction.

We noted some consistent differences in ethanol’s transcriptional response between the sexes. Our ChIP-Seq data identified some overlaps in genomic regions differentially bound by H3K9me3 in both sexes, but a majority of the genes were differentially bound in one sex or the other. In the qPCR time course studies, we found that many of the myelin structural component genes were mostly altered in the males. Similarly, at PND 65 in our O4+ enriched cells, males with a history of binge ethanol decreased expression of OLG transcriptional regulators and myelin structural components. These sex differences could be due to the developmental differences between males and females—male brains have been shown to mature at a slower rate than females (Drzewiecki et al., 2017). As our ethanol administration is based on postnatal day, and females are further along in brain development at the onset of our dosing paradigm, ethanol may not impact females OLG development as strongly as it affects males.

We replicated our earlier findings in regard to decreased myelin expression (Wolstenholme et al., 2017), showing that adolescent binge ethanol treatment significantly decreased Plp, Mobp, Mag, and Mal in males twenty four hours after the last binge. We have also extended these findings, potentially showing ethanol’s early impact on transcription factors necessary for oligodendrocyte development. The transcription factors, Pdgfra and Olig1, showed a trend towards an ethanol-induced decrease in gene expression which was stronger at 24-h (Figure 3). As Olig1 regulates transcription of major myelin-specific genes including Mag and Plp (Li and Richardson, 2008), and Pdgfra has been shown to be a major determinant of oligodendrocyte differentiation (Zhu et al., 2014), this ethanol-induced early decrease in gene expression could lead to the later significant decrease in gene expression of Plp, Mobp, Mag, and Mal at the 24-h mark. These early reductions in Olig1 and Pdgfra could be indicative of a delay in OLG maturation in ethanol exposed mice. As a result, genes involved in later development of myelin structure (i.e., Plp, Mag, and Mal) could be less abundant due to fewer mature myelin forming OLGs.

In our study to directly investigate the progression of oligodendrocyte development using O4+ enriched cell populations from the mouse frontal pole, we measured gene expression of the same myelin structural components (Mbp, Plp, Mobp, Mag, and Mal) and transcription factors (Pdgfra, Olig1 and Sox10), as well as genes that regulate H3K9me3 (Kdm4c, Kdm4d, Ehmt1 and Ehmt2). We expected to find a dose dependent alteration in these genes during our binge ethanol paradigm. However, we only detected a main effect of age where, in general, a decrease in gene expression for most of these genes was found at later ages. As expected, this maturational decline was also reflected in the input tissue used for the O4+ enrichment studies. We did not detect any changes in gene expression due to increasing ethanol binges, even when the data was normalized for the maturational decline. At PND 65, three weeks following the last ethanol binge, we do note a general trend for ethanol to decrease expression of these genes in male O4+ cells as compared to control males suggesting a lasting effect of ethanol’s transcriptional dysregulation of OLG maturation in O4+ cells in the frontal cortex. This change in OLG transcriptional regulators and myelin structural components was not seen in the input tissue, nor in females (see Figure 4).

There are a few possible reasons we may not have been able to detect robust changes in gene expression in our O4+ enriched cell studies. First, we collected these cells from mouse frontal pole which includes large white matter tracts such as the corpus callosum and anterior commissure as well as grey matter structures (the prefrontal cortex, nucleus accumbens caudate putamen, etc.). These regions have highly heterogeneous population of cells and amounts of white matter and myelin, and all have different OLG maturation rates. Liu et al. (2021) showed that qPCR in bulk tissue was not able to detect changes in myelin related gene expression, even though brain regional effects of myelin proteins were found in adults following adolescent intermittent ethanol (Liu et al., 2021). Second, there is a possibility that the magnetic antibody sorting and tissue processing altered cellular processes such as transcription. This could be a likely possibility, since gene expression changes from the input tissue did not replicate our prior ethanol findings from bulk PFC tissue snap frozen immediately after harvest (in Experiment 2 and Wolstenholme et al., 2017). Third, the choice of O4+ as a marker to collect late OPC progenitor cells to mature myelin-forming OLGs may either not be as selective as desired, or may not enrich for all of these cell types. This possibility is less likely since we do see enrichment of these myelin-related genes as compared to input tissue. Overall, we found that while binge ethanol in adolescence decreased gene expression of myelin related genes in male PFC tissue, our O4+ enrichment studies were unable to detect such changes when using larger brain regions and studies are needed at more discrete brain regional resolution. The trend for a lasting decrease in the transcriptional regulation and expression of myelin structural genes coupled with our H3K9me3 differential binding at genes that regulate synaptic activity suggest that binge ethanol during adolescence can lead to lasting changes in the regulation, development and maturation of OLGs leading to decreased white matter in adulthood. These studies have identified potential target genes and signaling pathways for further study as potential therapeutic strategies for intervention.

Overall, this study unveiled potential genes that may be contributing to adolescent binge ethanol-related myelin loss. As H3K9me3 plays an important role in oligodendrocyte differentiation (Liu et al., 2015), and H3K9-related gene expression was previously found to be decreased in the PFC after adolescent binge ethanol (Wolstenholme et al., 2017), we assessed ethanol-induced changes to this histone mark in the developing adolescent PFC. We further assessed ethanol-induced changes at the transcription level with qPCR time course approaches in O4+ cells to assess oligodendrocyte progenitor and oligodendrocytes specifically. Through this set of experiments, we found that adolescent binge ethanol altered H3K9me3 regulation of synaptic-related genes, and genes specific for glutamate and potassium channels. Given the data presented in Liu et al. (2015), dysregulation of ion channels could impede oligodendrocyte differentiation required for ongoing myelin development in the PFC throughout adolescence. Our gene expression findings support our previous work; that adolescent binge ethanol decreased expression of myelin-related genes in the PFC of male animals, 24 h after the last dose of ethanol. We further expanded this finding by looking at an earlier time point, potentially showing ethanol’s early impact on transcription factors necessary for oligodendrocyte development. This early change in gene expression in transcription factors associated with oligodendrocyte differentiation may lead to the later significant decrease in myelin-related gene expression at Mag, Mal, and Plp at the 24-h mark. The O4+ enrichment time course and dose response studies were inconclusive, but could suggest lasting dysregulation of OL maturation at the transcriptional level. Further experiments will be required to assess the exact role of these genes in ethanol-induced myelin loss, and whether they are regulated by histone marks other than H3K9me3.

The original contributions presented in the study are publicly available. The datasets generated and analyzed for this study have been deposited with the Gene Expression Omnibus resource (GSE220746 with data subsets ChIP-seq: Accession #GSE220742 and RNA-seq: Accession #GSE220745).

The animal study was approved by Virginia Commonwealth University Institutional Animal Care and Use Committee. The study was conducted in accordance with the local legislation and institutional requirements.

EB: Conceptualization, Formal analysis, Investigation, Methodology, Validation, Visualization, Writing – original draft, Writing – review & editing. RE: Investigation, Methodology, Writing – review & editing. AM: Investigation, Visualization, Writing – review & editing. MK: Investigation, Methodology, Validation, Writing – review & editing. GL: Formal analysis, Investigation, Writing – review & editing. JW: Conceptualization, Formal analysis, Investigation, Methodology, Project administration, Resources, Supervision, Validation, Visualization, Writing – review & editing.