Cerebellar Transcriptomic Analysis in a Chronic plus Binge Mouse Model of Alcohol Use Disorder Demonstrates Ethanol-Induced Neuroinflammation and Altered Glial Gene Expression

Alcohol use disorder (AUD) is one of the most common preventable mental health disorders and can result in pathology within the CNS, including the cerebellum. Cerebellar alcohol exposure during adulthood has been associated with disruptions in proper cerebellar function. However, the mechanisms regulating ethanol-induced cerebellar neuropathology are not well understood. High-throughput next generation sequencing was performed to compare control versus ethanol-treated adult C57BL/6J mice in a chronic plus binge model of AUD. Mice were euthanized, cerebella were microdissected, and RNA was isolated and submitted for RNA-sequencing. Down-stream transcriptomic analyses revealed significant changes in gene expression and global biological pathways in control versus ethanol-treated mice that included pathogen-influenced signaling pathways and cellular immune response pathways. Microglial-associated genes showed a decrease in homeostasis-associated transcripts and an increase in transcripts associated with chronic neurodegenerative diseases, while astrocyte-associated genes showed an increase in transcripts associated with acute injury. Oligodendrocyte lineage cell genes showed a decrease in transcripts associated with both immature progenitors as well as myelinating oligodendrocytes. These data provide new insight into the mechanisms by which ethanol induces cerebellar neuropathology and alterations to the immune response in AUD.


Introduction
Excessive alcohol consumption in adolescents and adults has significant societal impacts, with an estimated economic cost of $249 billion in the U.S. alone [1]. Studies have shown that alcohol misuse can lead to low academic achievement, an increased risk of suicide, and a lifetime struggle with addiction [2][3][4]. Furthermore, alcohol use disorder (AUD) is one of the most prevalent mental health disorders, with 15.7 million Americans aged 12 and older diagnosed [5,6], and is associated with many physical and psychiatric comorbidities [7,8]. Despite the known consequences of excess alcohol consumption, 29.7% of men and 22.2% of women were diagnosed with an AUD in 2019 [9]. AUD is associated with pathology to organ systems including the central nervous system (CNS). Animal models of AUD have been developed which simulate the behavioral abnormalities and neuropathologies associated with human AUD, thus allowing researchers to investigate the biological mechanisms associated with AUD [10]. Within the CNS, the cerebellum is

Bioinformatic Analysis
To identify significant differences in mRNA gene expression and global biological pathways associated with alterations of cerebellar genes between the control and ethanol treatment groups, raw RNA-sequence data (NCBI GEO accession GSE222445) were analyzed. RNA-seq reads were quality-checked, trimmed, and aligned to the GRCm39 reference genome (accession: GCA_000001635.9) using the Nextflow RNAseq pipeline, nf-core/rnaseq (version 3.4), available at DOI 10.5281/zenodo.1400710. The resulting gene counts were transformed to Log 2 counts per million (CPM) [24]. Lowly expressed genes were filtered out, and libraries were normalized by trimmed means of M-values [25]. The Limma R package was used to calculate differential expression among genes [26]. Log 2 fold change values were calculated for ethanol compared to control, and genes with an adjusted (adj.) p ≤ 0.05 were considered statistically significant.
Heat map and principal component analysis (PCA) plots were created from the processed differential gene expression data using R statistical software. The R-based Enhanced-Volcano package was used to make the volcano plots [27]. Pathway and network analysis were conducted using the QIAGEN Ingenuity Pathway Analysis (IPA) software (QIAGEN Inc., Valencia, CA, USA, https://digitalinsights.qiagen.com/IPA, accessed on 22 July 2022 ) using the "Core Expression Analysis". IPA analysis parameters were set with the "species" parameter as "mouse", and the "tissues and cell lines" parameter as "cerebellum", with gene cut offs of an adj. p ≤ 0.05 and Log 2 fold change ≥0.5 or ≤−0.5.
To obtain a better understanding of the specific cellular processes and cell types of the cerebellum that are most sensitive to ethanol exposure, we extracted cell type-specific gene lists from publicly available single-cell RNA-seq (scRNA-seq) resources, which have been used previously to deduce the cell composition of bulk RNA-seq tissue [28]. Using this approach, we identified a total of 822 microglia-associated genes from scRNA-seq resources [29][30][31][32][33] (Supplemental Table S1A). We compared this list of microglia-associated genes to the list of genes significantly differentially regulated by ethanol (adj. p ≤ 0.05) in our dataset, which identified 151 microglia-associated genes whose expression was altered by ethanol (Table 1). Table 1. Uncategorized microglia-associated genes dysregulated by ethanol exposure in the cerebellum. Genes were identified by cross-referencing our significantly (adjusted p < 0.05) differentially regulated gene list with the 822 microglia-associated genes extracted from previous studies [identified in [29][30][31][32][33]] (Supplemental Table S1A) using R statistical software, which identified 151 genes associated with microglia. We were able to characterize 23 of the 151 genes as being either homeostatic or neurodegenerative (Table 2), as defined in previous studies [33][34][35][36][37] (Supplemental Table S1B,C).
To further evaluate the effects of ethanol on homeostatic versus neurodegenerative microglial phenotypes, we computed mean z-scores to compare control versus ethanol for the transcripts associated with these phenotypes. Since the goal was to determine relative gene expression changes in our dataset, i.e., to determine whether the genes are up-or down-regulated due to ethanol, the average z-score was computed. We calculated the average z-score across individual genes in our extracted microglia homeostatic and neurodegenerative-associated gene lists, and then averaged these individual gene z-scores within each sample. The average z-score of each sample in the homeostatic and neurodegenerative group was then evaluated using a two-tailed Student's t test, with p ≤ 0.05 being considered statistically significant. R statistical software was used to conduct the Student's t-test as well as construct the average z-score graphs.
Similar to microglia, we utilized scRNA-seq data to compose a list of 309 astrocyteassociated genes (Supplemental Table S2) [37]. From this list we identified 56 astrocyteassociated genes that were differentially expressed in response to ethanol in our current study. We then characterized these transcripts as being associated with an astrocyte pheno-Cells 2023, 12, 745 5 of 20 type common to acute injury, chronic neurodegenerative diseases, or pan-injury (Table 3), the last of which includes genes associated with both acute injury and chronic neurodegenerative disease phenotypes [37]. Table 2. Categorized microglia-associated genes dysregulated by ethanol exposure in the cerebellum. The microglia-associated genes identified in our data set in Table 1 with an adjusted p < 0.05 and Log 2 fold change ≥ 0.25 or ≤ −0.25 were further categorized as being homeostatic or neurodegenerative, as defined by previous studies [identified in [35][36][37]].  Table 3. Categorized astrocyte-associated genes dysregulated by ethanol exposure in the cerebellum. Genes were identified by cross-referencing our significantly (adjusted p < 0.05) differentially regulated gene list with the list of 309 astrocyte-associated genes extracted from a previous study [identified in [38]] (Supplemental Table S2) using R statistical software. The astrocyte-associated genes identified in our dataset were then further categorized as being associated with acute injury, chronic neurodegenerative diseases, or pan-injury, as described in a previous study [38].  To test for statistical significance, the average z-scores of each gene in our extracted acute, chronic, and pan-injury astrocyte-associated gene lists were generated, and these individual gene z-scores were then averaged within each sample in a manner consistent with the microglia described above. The Student's t-test and average z-score graphs were constructed using R statistical software. Due to the small number of chronic neurodegenerative disease astrocyte-associated genes (n = 3), no z-score graph was generated for this group.
For oligodendrocyte lineage-associated genes, we extracted gene lists for oligodendrocyte precursor cells (OPCs) (381 genes), committed oligodendrocyte precursor cells (COPs) (55 genes), newly formed oligodendrocytes (NFOL) (9 genes), myelin-forming oligodendrocytes (MFOL) (347 genes), and mature oligodendrocytes (MOL) (7 genes) from publicly available scRNA-seq studies [29,30,39] (Supplemental Table S3), in a manner consistent with microglia and astrocytes described above, to determine which genes were significantly differentially regulated by ethanol. From these lists, we identified 71 differentially expressed genes associated with OPCs, 12 genes associated with COPs, 2 genes associated with NFOL, 2 genes associated with MOL, and 108 genes associated with MFOL within our significantly differentially regulated dataset (Table 4). Table 4. Categorized oligodendrocyte lineage-associated genes dysregulated by ethanol exposure in the cerebellum. Genes were identified by cross-referencing our significantly (adjusted p < 0.05) differentially regulated gene list with the list of OPC, COP, NFOL, MFOL and MOL-associated genes [identified in [29,30,39]] (Supplemental Table S3  We performed statistical analyses in a manner similar to the microglia and astrocytes above. Briefly, the average z-scores of each gene in our OPC, COP, and MFOL-associated gene lists were generated, and the individual gene z-scores were then averaged between each sample. The Student's t-test and average z-score graphs were constructed using R statistical software. Due to the small number of NFOL and MOL-associated genes differentially regulated by ethanol, z-score graphs were not generated for these groups.

Alcohol-Induced Differential Gene Expression in the Cerebellum
A principal component analysis (PCA) was performed to provide an overview of the transcriptomic changes that occurred in response to ethanol. PCA analysis demonstrated that gene transcripts correlating and anticorrelating to the first and second principal components could differentiate control animals from those exposed to ethanol. ( Figure 1A). Hierarchical clustering analysis of significant genes was conducted using Pearson's correlation, while controlling for false discovery rate adj. p ≤ 0.05 ( Figure 1B). RNA-seq analysis identified 732 genes that were significantly differentially regulated (adj. p ≤ 0.05 and log 2 FC 0.5). Of these 732 genes, 269 were upregulated genes (36.75%) and 463 were downregulated genes (63.25%), ( Figure 1C).

Pathway Analysis of the Alcohol-Induced Differentially Regulated Genes
IPA analysis was performed to determine the specific pathways altered by ethanol in the cerebella of adult mice. The results of the top canonical pathways altered by ethanol exposure included those related to the generation of precursor metabolites and energy, pathogeninfluenced signaling, cellular immune response, degradation/utilization/assimilation, cellular stress and injury, biosynthesis, disease-specific pathways, cardiovascular signaling, nuclear receptor signaling, and ingenuity toxicity list pathways (Figure 2A). A description of the pathway names, p-values, and molecules associated with each significantly altered pathway category is shown in Table 5. The top disease and biological function categories altered by ethanol exposure included nervous system development and function, tissue/cell morphology, cell-to-cell signaling and interaction, cell death and survival, cellular compromise, immune cell trafficking, and inflammatory response [−log(p.val) range = 5.5-2.1] ( Figure 2B). . Ethanol-induced differential gene expression in the cerebellum. Principle component analysis (PCA) of genes contributing to variance between ethanol (E) and control (C) in the cerebellum were analyzed using R statistical software (A). A heatmap and hierarchical clustering dendrogram of relative gene expression across samples was constructed using R statistical software for significantly (adjusted p < 0.05) altered genes. Red indicates positive z-scores (upregulation) and blue indicates negative z-scores (downregulation) (B). The R EnhancedVolcano package was utilized to construct a volcano plot displaying fold change versus adjusted p-value of all detected genes in the cerebellum. 732 of 17,791 total identified transcripts displayed an adjusted p < 0.05 and Log2 fold change ≥ 0.5 or ≤ −0.5, shown in red (C). n = 5 males per treatment group E or C.

Pathway Analysis of the Alcohol-Induced Differentially Regulated Genes
IPA analysis was performed to determine the specific pathways altered by ethanol in the cerebella of adult mice. The results of the top canonical pathways altered by ethanol exposure included those related to the generation of precursor metabolites and energy, Figure 1. Ethanol-induced differential gene expression in the cerebellum. Principle component analysis (PCA) of genes contributing to variance between ethanol (E) and control (C) in the cerebellum were analyzed using R statistical software (A). A heatmap and hierarchical clustering dendrogram of relative gene expression across samples was constructed using R statistical software for significantly (adjusted p < 0.05) altered genes. Red indicates positive z-scores (upregulation) and blue indicates negative z-scores (downregulation) (B). The R EnhancedVolcano package was utilized to construct a volcano plot displaying fold change versus adjusted p-value of all detected genes in the cerebellum. 732 of 17,791 total identified transcripts displayed an adjusted p < 0.05 and Log 2 fold change ≥0.5 or ≤−0.5, shown in red (C). n = 5 males per treatment group E or C.

Alcohol Suppresses Microglia Homeostatic Genes while Increasing the Expression of Microglia Neurodegenerative-Associated Genes
Alcohol has been demonstrated to induce neuroinflammation in both humans and rodents which may include microglial activation, characterized by shortening and thickening of processes, along with the secretion of proinflammatory cytokines and chemokines that may contribute to neuropathology [19,40,41]. We performed hierarchical clustering analysis on homeostatic and neurodegenerative disease microglia-associated genes that were differentially expressed (adj. p ≤ 0.05) in response to ethanol ( Figure 3A). A Student's t-test comparing the average z-scores across all relevant genes indicated that ethanol caused an overall significant downregulation of microglia homeostatic genes (p.val = 3.191 × 10 −6 ) ( Figure 3B, Table 2) and an overall significant upregulation of microglia genes associated with neurodegenerative diseases (p.val = 7.786 × 10 −5 ) ( Figure 3C, Table 2). Collectively, these data suggest that ethanol may alter the microglial phenotype from a homeostatic and protective phenotype to a more activated phenotype observed in neurodegenerative diseases.   The diseases and biological function annotations that correlate to the diseases and biological functions categories, as shown in Figure 2B, are myelination (p.val = 2.88 × 10 −6 ) or demyelination (p.val = 0.0053) of the cerebellum; quantity (p.val = 0.000125) or coupling (p.val = 0.000556) of oligodendrocytes; thickness of myelin sheath (p.val = 0.000556); quantity of cells (p.val = 0.00783); activation of microglia (p.val = 0.00783); permeability of blood-brain barrier (p.val = 0.0236); and astrocytosis of cerebella (p.val = 0.0467), (Table 6). These results suggest that in the cerebellum, ethanol alters biological functions that pertain to alterations in the formation of myelin, along with possible microglia and astrocyte phenotypic changes. Table 6. Tabular descriptions of the disease and biological function categories, including annotation, p-value, and indicated molecules. Qiagen Ingenuity Pathway Analysis (IPA) software was utilized to assess the top diseases and biological functions altered by ethanol exposure using the "cerebellum" selected analysis settings. All analyses were restricted to genes with an adjusted p < 0.05 and Log 2 fold change ≥0.5 or ≤−0.5.

Categories
Disease

Astrocytes Undergo a Phenotypic Switch following Chronic plus Binge-like Alcohol Exposure
Astrocytes are one of the most abundant cell types in the CNS and play a critical role in regulating CNS functions in health and disease by maintaining homeostasis, providing energy to neurons, regulating synapse development and plasticity, modulating blood-brainbarrier integrity, and controlling neurological function and behavior [42][43][44][45][46]. Similarly to microglia, astrocytes play a role in CNS inflammation [47,48], and ethanol has been demonstrated to trigger an immune response in astrocytes [49,50]. In the current study, we performed hierarchical clustering analysis on acute injury, chronic neurodegenerative, and pan-injury astrocyte-associated genes that were differentially expressed (adj. p ≤ 0.05) in response to ethanol ( Figure 4A). A Student's t-test comparing the average z-scores across all relevant genes indicated that ethanol caused an overall significant increase in astrocyte genes related to acute injury (p.val = 7.085 × 10 −5 ) ( Figure 4B, Table 3) and an almost even number of up-and down-regulated genes (12 up vs. 13 down) pertaining to pan-injury (p.val = 0.6266) ( Figure 4C, Table 3). Ethanol only altered the expression of three genes associated with the chronic neurodegenerative disease category (Table 3), thus the effect of ethanol on this small number of genes was not statistically evaluated. These data suggest that alcohol-induced transcriptomic changes in astrocytes are consistent with an acute injury phenotype.

2023, 12, x FOR PEER REVIEW
15 of clustering analysis on homeostatic and neurodegenerative disease microglia-associat genes that were differentially expressed (adj. p ≤ 0.05) in response to ethanol ( Figure 3A A Student's t-test comparing the average z-scores across all relevant genes indicated th ethanol caused an overall significant downregulation of microglia homeostatic gen (p.val = 3.191 × 10 −6 ) ( Figure 3B, Table 2) and an overall significant upregulation of microg genes associated with neurodegenerative diseases (p.val = 7.786 × 10 −5 ) ( Figure 3C, Table  Collectively, these data suggest that ethanol may alter the microglial phenotype from homeostatic and protective phenotype to a more activated phenotype observed in neu degenerative diseases. . Individual genes were z-scored across sa ples, followed by calculation of average z-score for each treatment group which was used for test statistical significance in R with Student's t-test. Quantification by average z-score of homeosta microglia-associated genes (B) and neurodegenerative microglia-associated genes (C). n = 5 ma per treatment group E or C; *** p < 0.001.

Astrocytes Undergo a Phenotypic Switch following Chronic Plus Binge-Like Alcohol Exposure
Astrocytes are one of the most abundant cell types in the CNS and play a critical ro Figure 3. Microglia-associated genes altered by ethanol exposure in the cerebellum. R statistical software was utilized to construct a heatmap and hierarchical clustering dendrogram of relative gene expression across samples for significantly (adjusted p < 0.05) altered and categorized microgliaassociated genes as detailed in Methods. Red indicates positive z-scores (upregulation) and blue indicates negative z-scores (downregulation) (A). Individual genes were z-scored across samples, followed by calculation of average z-score for each treatment group which was used for testing statistical significance in R with Student's t-test. Quantification by average z-score of homeostatic microglia-associated genes (B) and neurodegenerative microglia-associated genes (C). n = 5 males per treatment group E or C; *** p < 0.001.

Oligodendrocyte Lineage Cells Are Depleted upon Chronic plus Binge-like Alcohol Exposure
Ethanol has been demonstrated to alter myelination in adult humans and rodents [51,52]. We performed hierarchical clustering analysis on genes associated with distinct oligodendrocyte lineages (immature and myelinating) whose expression was altered by ethanol ( Figure 5A,B). Evaluation of the effects of ethanol on immature oligodendrocyte lineages indicated that ethanol significantly decreased the expression of genes associated with COPs (p.val = 0.0006784) ( Figure 5C, Table 4), and that ethanol skewed toward decreasing the expression of genes associated with OPCs (p.val = 0.1702) ( Figure 5D, Table 4). For the myelinating oligodendrocyte lineage cells, ethanol significantly decreased the expression of genes associated with MFOLs (p.val = 2.905 × 10 −05 ) ( Figure 5E, Table 4). NFOL and MOL groups only contained two differentially expressed genes; therefore, statistical significance was not evaluated for these categories (Table 4). These results suggest that ethanol effects both immature and myelinating oligodendrocyte lineage cells, which could potentially lead to altered myelination. been demonstrated to trigger an immune response in astrocytes [48,49]. In the current study, we performed hierarchical clustering analysis on acute injury, chronic neurodegenerative, and pan-injury astrocyte-associated genes that were differentially expressed (adj. p ≤ 0.05) in response to ethanol ( Figure 4A). A Student's t-test comparing the average zscores across all relevant genes indicated that ethanol caused an overall significant increase in astrocyte genes related to acute injury (p.val = 7.085 × 10 −05 ) ( Figure 4B, Table 3) and an almost even number of up-and down-regulated genes (12 up vs. 13 down) pertaining to pan-injury (p.val = 0.6266) ( Figure 4C, Table 3). Ethanol only altered the expression of three genes associated with the chronic neurodegenerative disease category (Table  3), thus the effect of ethanol on this small number of genes was not statistically evaluated. These data suggest that alcohol-induced transcriptomic changes in astrocytes are consistent with an acute injury phenotype. . Astrocyte-associated genes altered by ethanol exposure in the cerebellum. R statistical software was utilized to construct a heatmap and hierarchical clustering dendrogram of relative gene expression across samples for significantly (adjusted p < 0.05) altered and categorized astrocyte-associated genes, as detailed in Methods. Red indicates positive z-scores (upregulation) and blue indicates negative z-scores (downregulation) (A). Individual genes were z-scored across samples, followed by calculation of the average z-score for each treatment group, which was used for testing statistical significance in R with Student's t-test. Quantification by average z-score of acute injury astrocyte-associated genes (B) and pan-injury astrocyte-associated genes (C). Due to the small number of chronic neurodegenerative injury astrocyte-associated genes, no z-score graph was generated for this group; however, this group is further characterized in Table 3. n = 5 males per treatment group E or C; *** p < 0.001.

Oligodendrocyte Lineage Cells Are Depleted upon Chronic Plus Binge-Like Alcohol Exposure
Ethanol has been demonstrated to alter myelination in adult humans and rodents [50,51]. We performed hierarchical clustering analysis on genes associated with distinct . Astrocyte-associated genes altered by ethanol exposure in the cerebellum. R statistical software was utilized to construct a heatmap and hierarchical clustering dendrogram of relative gene expression across samples for significantly (adjusted p < 0.05) altered and categorized astrocyteassociated genes, as detailed in Methods. Red indicates positive z-scores (upregulation) and blue indicates negative z-scores (downregulation) (A). Individual genes were z-scored across samples, followed by calculation of the average z-score for each treatment group, which was used for testing statistical significance in R with Student's t-test. Quantification by average z-score of acute injury astrocyte-associated genes (B) and pan-injury astrocyte-associated genes (C). Due to the small number of chronic neurodegenerative injury astrocyte-associated genes, no z-score graph was generated for this group; however, this group is further characterized in Table 3. n = 5 males per treatment group E or C; *** p < 0.001. creasing the expression of genes associated with OPCs (p.val = 0.1702) ( Figure 5D, Table  4). For the myelinating oligodendrocyte lineage cells, ethanol significantly decreased the expression of genes associated with MFOLs (p.val = 2.905 × 10 −05 ) ( Figure 5E, Table 4). NFOL and MOL groups only contained two differentially expressed genes; therefore, statistical significance was not evaluated for these categories (Table 4). These results suggest that ethanol effects both immature and myelinating oligodendrocyte lineage cells, which could potentially lead to altered myelination.

Figure 5.
Alterations in oligodendrocyte lineage-associated genes by ethanol exposure in the cerebellum. R statistical software was utilized to construct a heatmap and hierarchical clustering dendrogram of relative gene expression across samples for significantly (adjusted p < 0.05) altered and categorized oligodendrocyte lineage-associated genes as detailed in Methods : immature oligodendrocyte lineage-associated genes (A) and myelinating oligodendrocyte lineage-associated genes (B) Red indicates positive z-scores (upregulation) and blue indicates negative z-scores (downregulation) (A and B). Individual genes were z-scored across samples, followed by calculation of average z-score for each treatment group, which was used for testing statistical significance in R with Student's t-test. Quantification by average z-score of COP-associated genes (C), OPC-associated genes (D), and MFOL-associated genes in the cerebellum (E). Due to the small number of NFOL and MOLassociated genes, no z-score graph was generated for this group; however, this group is further characterized in Table 4. Abbreviations: OPC, oligodendrocyte precursor cell; COP, committed oligodendrocyte precursor; MFOL, myelin-forming oligodendrocyte; NFOL, newly formed oligodendrocyte; MOL, mature oligodendrocyte. n = 5 males per treatment group E or C; *** p < 0.001. Figure 5. Alterations in oligodendrocyte lineage-associated genes by ethanol exposure in the cerebellum. R statistical software was utilized to construct a heatmap and hierarchical clustering dendrogram of relative gene expression across samples for significantly (adjusted p < 0.05) altered and categorized oligodendrocyte lineage-associated genes as detailed in Methods: immature oligodendrocyte lineage-associated genes (A) and myelinating oligodendrocyte lineage-associated genes (B) Red indicates positive z-scores (upregulation) and blue indicates negative z-scores (downregulation) (A,B). Individual genes were z-scored across samples, followed by calculation of average z-score for each treatment group, which was used for testing statistical significance in R with Student's t-test. Quantification by average z-score of COP-associated genes (C), OPC-associated genes (D), and MFOL-associated genes in the cerebellum (E). Due to the small number of NFOL and MOL-associated genes, no z-score graph was generated for this group; however, this group is further characterized in Table 4. Abbreviations: OPC, oligodendrocyte precursor cell; COP, committed oligodendrocyte precursor; MFOL, myelin-forming oligodendrocyte; NFOL, newly formed oligodendrocyte; MOL, mature oligodendrocyte. n = 5 males per treatment group E or C; *** p < 0.001.

Discussion
Pathway analysis indicated that ethanol had significant effects on immune processes in the cerebella of adult mice. In addition, these analyses suggested that ethanol may alter the phenotype and function of glial cells including microglia, astrocytes, and oligodendrocyte lineage cells. We and others have previously demonstrated that ethanol induces neuroinflammation in adult rodents. However, the amount of neuroinflammation varies depending on the ethanol administration paradigm. For example, acute 4-day ethanol exposure did not alter the expression of pro-inflammatory molecules, although microglial activation was observed [17,53]. Following 10-day ethanol exposure, increased expression of pro-inflammatory molecules was observed, although it was somewhat modest [16,18,19]. Chronic ethanol exposure over a period of 3-5 months resulted in more robust neuroin-flammation [15,49,54,55]. Using a variation of the same model as the current study, in which gene expression in both male and female mice was evaluated in control, ethanol, and ethanol + pioglitazone experimental groups, we have previously demonstrated robust neuroinflammation following chronic plus binge exposure to ethanol in less than one month [22]. This model is similar to an alcoholic liver disease model used previously by the Gao laboratory, in which they showed systemic inflammation and liver injury [20,21]. At this point, it is unclear in our studies whether ethanol induces CNS inflammation directly or indirectly through ethanol induced inflammation outside of the CNS. In order to begin to understand the possible mechanisms by which ethanol induces neuroinflammation in this chronic plus binge model of AUD, we have treated a unique set of male mice for the purpose of RNAseq analysis in the current study. We acknowledge that the use of only male mice is a limitation of the current study. Furthermore, some of the pathways identified in the current study only contain 1 or 2 genes, and some genes are represented in multiple pathways. Thus, we have exercised caution to not overinterpret the results.
We evaluated the transcriptomic data to identify immune-regulated genes whose expression was most strongly induced by ethanol, which included FOSB, CCL2, CCL7, C5AR1, SPP1, CD68, SOCS3, C3AR1, and KLF4. The most highly upregulated gene is FOSB, which encodes a transcription factor that dimerizes with Jun protein to form AP-1 and plays a critical role in alcohol and drug addiction [56]. Alcohol increases the expression of FOSB in the mesocorticolimbic system, which is believed to contribute to alcohol use disorder [57,58]. Furthermore, ethanol was demonstrated to alter synaptic plasticity and epigenetic alterations in the FOSB promoter, resulting in increased FOSB expression in the medial prefrontal cortex in wild-type but not TLR4 deficient mice. Since ethanol is believed to activate TLR4, resulting in downstream immune signaling [59], a role of ethanol-induced neuroinflammation is suggested in these processes. FOSB has also been demonstrated to contribute to excitotoxic microglial activation through regulation of complement C5a receptors in these cells [60]. Interestingly ethanol strongly increased the expression of complement C5AR1 and C3AR1 in our RNA-Seq studies. C5AR1 expression is increased in the liver of patients with alcoholic hepatitis [61], and is believed to contribute to alcoholinduced inflammation and liver injury [62,63]. Additionally, ethanol induces the expression of complement receptors including C3AR1 expression in microglia, resulting in altered phagocytosis [64]. We previously demonstrated that ethanol induces the expression of the chemokine CCL2 or MCP-1 following acute ethanol exposure in adult rodents [65], as well as in animal models of fetal alcohol spectrum disorders (FASD) [66]. It is interesting that in the current study, ethanol induced the expression of CCL2 as well the related chemokine CCL7 or MCP-3 in this chronic plus binge model. It should also be noted that transcriptomic changes were only evaluated at one timepoint, 24 h after the final ethanol exposure. Future studies may wish to evaluate transcriptomic changes at different times following the final ethanol exposure. It is also noteworthy that the other immune-related molecules we identified previously in this model were not indicated in the current study; this may be due to less sensitivity and smaller "n", both of which are limitations that come with RNAseq when compared to quantitative real-time PCR [22].
Microglia are capable of responding to signals, resulting in activation and an altered phenotype. Our IPA analysis indicated that ethanol treatment resulted in microgliosis or microglial activation in the cerebellum. Upon activation, microglia have traditionally been hypothesized to undergo classical activation, resulting in a M1 pro-inflammatory phenotype, or alternative activation, resulting in an M2 anti-inflammatory or protective phenotype [67,68]. However, more recently it has become clear that microglial phenotypes are complex, and cannot be defined or categorized effectively using this simple binary system [69]. One recent nomenclature to distinguish microglial phenotype focuses on homeostatic versus neurodegenerative disease phenotypes. Under homeostatic conditions, microglia have a homeostatic phenotype, described by playing a role in synaptic plasticity and synaptogenesis, trophic support, chemotaxis and immune cell recruitment, and neurogenesis [37]. During insult to the CNS, microglia commonly lose their homeostatic sig-nature and assume a chronic inflammatory signature [70][71][72]. Evaluation of the phenotype of microglia in a variety of neurodegenerative diseases have resulted in the identification of a common neurodegenerative disease-associated microglia phenotype [34,37,71,73]. In the current study, ethanol induced a microglia phenotypic switch in the cerebellum. This phenotypic switch was similar to that observed in neurodegenerative diseases, with a downregulation of homeostatic signature genes and an upregulation of neurodegenerative signature genes.
Astrocytes, like microglia, are capable of functioning in the innate immune response in the CNS. Once astrocytes are activated, commonly referred to as astrogliosis/astrocytosis, they produce cytokines and chemokines, nitric oxide, and other reactive oxygen species as part of an inflammatory response [74], Our IPA analysis indicated that ethanol treatment resulted in "astrocytosis". Astrocytes were classically defined to respond to various stimuli to become reactive A1 astrocytes (neurotoxic or reactive A2 astrocytes) which are protective and neurotrophic [75,76]. However, as with microglia, this binary system of classifying reactive astrocytes appears inadequate to fully define and distinguish astrocyte phenotypes. More recently, Serrano-Pozo and colleagues performed a meta-analysis of mouse transcriptomic studies which resulted in a nomenclature that classified reactive astrocytes as being consistent with acute injury, chronic neurodegeneration, or pan-injury reactive astrocytes which exhibited characteristics of both acute injury and chronic neurodegenerative phenotypes [38]. In the current study, we determined that ethanol induced changes consistent with an acute injury astrocyte phenotype. Interestingly, LPS was previously shown to trigger an acute injury astrocyte phenotype [38]. ethanol has also been shown to activate TLR4 receptors, suggesting that ethanol-mediated neuroinflammation could occur in response to recruitment of TLR4 during alcohol use/abuse [77][78][79]. Therefore, we speculate that in this model of AUD, in the cerebellum, ethanol induces an acute injury astrocytic phenotype through the activation of TLR4, subsequently inducing an immune response.
Oligodendrocytes are responsible for forming a myelin sheath around axons of neurons in the CNS, facilitating the efficient propagation of action potentials [80]. OPCs are produced during embryogenesis, and migrate to their functional location wherein they differentiate into mature myelinating oligodendrocytes. Most myelination occurs at later stages of CNS development but can occur throughout life [81]. Ethanol has profound effects on the developing CNS and is believed to significantly contribute to the pathology associated with FASD, at least in part by altering myelination [82]. Ethanol also alters myelination in adults with AUD [83,84]. Ethanol is highly toxic to oligodendrocyte lineage cells, with OPCs being particularly susceptible [85,86]. Alcohol exposure is known to disrupt OPC differentiation and survival by decreasing the expression of platelet-derived growth factor receptor α (PDGFRα), a molecule crucial for differentiation of OPCs into mature oligodendrocytes [87]. In the current study, we found that adult chronic plus binge-like alcohol exposure depletes the expression of genes associated with both immature oligodendrocyte precursor cells as well as myelinating oligodendrocytes. Future studies are needed to determine the mechanism by which ethanol effects oligodendrocyte lineage cells and myelination in AUD.

Conclusions
The current study demonstrates that ethanol alters the transcriptomic profile in the adult cerebellum in a chronic plus binge model of AUD. The pathways altered by ethanol included those involved in immune response. Ethanol caused a shift in the expression of microglial-associated genes, with a decrease in homeostatic and an increase in chronic neurodegenerative-associated transcripts. Ethanol also increased the expression of astrocyte-associated genes common to acute injury. Finally, ethanol decreased the expression of genes associated with immature oligodendrocyte progenitor cells, as well as myelinating oligodendrocytes. These results provide clues about the mechanisms by which ethanol induces neuroinflammation and altered glial function in AUD.
Author Contributions: All authors had access to the data for the study, made substantial contributions to the manuscript, approved the submitted version of the manuscript, and take responsibility for the accuracy and integrity of the data.

Informed Consent Statement: Not applicable.
Data Availability Statement: The data discussed in this publication have been deposited in NCBI's Gene Expression Omnibus [88] and are accessible through GEO Series accession number GSE222445 (https://www.ncbi.nlm.nih.gov/geo/query/acc.cgi?acc=GSE222445, accessed on 24 February 2023).