Abstract
Chronic excessive alcohol (ethanol) consumption induces neuroadaptations in the brain's reward system, including biochemical and structural abnormalities in white matter that are implicated in addiction phenotypes. Here, we demonstrate that long-term (12 week) voluntary ethanol consumption enhances myelination in the nucleus accumbens (NAc) of female and male adult mice, as evidenced by molecular, ultrastructural, and cellular alterations. Specifically, transmission electron microscopy analysis showed increased myelin thickness in the NAc following long-term ethanol consumption, while axon diameter remained unaffected. These changes were paralleled by increased mRNA transcript levels of key transcription factors essential for oligodendrocyte (OL) differentiation, along with elevated expression of critical myelination-related genes. In addition, diffusion tensor imaging revealed increased connectivity between the NAc and the prefrontal cortex, reflected by a higher number of tracts connecting these regions. We also observed ethanol-induced effects on OL lineage cells, with a reduction in the number of mature OLs after 3 weeks of ethanol consumption, followed by an increase after 6 weeks. These findings suggest that ethanol alters OL development prior to increasing myelination in the NAc. Finally, chronic administration of the promyelination drug clemastine to mice with a history of heavy ethanol consumption further elevated ethanol intake and preference, suggesting that increased myelination may contribute to escalated drinking behavior. Together, these findings suggest that heavy ethanol consumption disrupts OL development, induces enhanced myelination in the NAc, and may drive further ethanol intake, reinforcing addictive behaviors.
Significance Statement
Myelin is crucial for the development, maintenance, and proper functioning of the brain. Here, we provide evidence for the involvement of myelin alterations in alcohol (ethanol)-drinking behaviors. We show that chronic ethanol intake leads to enhanced myelination in the nucleus accumbens of adult mice. Moreover, we demonstrate that increasing myelination in heavily drinking mice leads to an escalation in ethanol intake. Thus, our results suggest that ethanol affects myelination processes, which, in turn, may affect ethanol-drinking patterns. Understanding the impact of ethanol on myelination could enhance our comprehension of alcohol addiction and open new avenues for treatment.
Introduction
Alcohol use disorder (AUD), characterized by the loss of control in limiting alcohol (ethanol) intake (American Psychiatric Association, 2013), is a chronic and relapsing psychiatric disease affecting 5–9% of the population (World Health Organization, 2018). Nevertheless, pharmacological treatment is currently highly limited (Garbutt et al., 1999; Witkiewitz et al., 2019). Long-term excessive ethanol consumption results in brain neuroadaptations, postulated to lead to AUD phenotypes (Vengeliene et al., 2008; Koob, 2013; Ron and Barak, 2016; Abrahao et al., 2017). These alterations occur mainly in the brain reward system, consisting of the mesocorticolimbic pathway that projects from the ventral tegmental area to the nucleus accumbens (NAc), hippocampus, amygdala, and prefrontal cortex (PFC). While most research on brain adaptations has focused on neurons, recent findings suggest that glial cells, and specifically oligodendrocytes (OLs), can also be dynamically regulated and actively participate in nervous system plasticity (Chang et al., 2016; Kaller et al., 2017).
Mature OLs (mOLs), which differentiate from OL progenitor cells (OPCs) in response to various signaling cues (Nishiyama et al., 2009; Boulanger and Messier, 2014), are the primary cells responsible for producing the myelin sheath in the central nervous system (Kuhn et al., 2019). This sheath is essential for neuronal function, providing insulation to axons and ensuring the efficient propagation of action potentials (Saab et al., 2013; Nave and Werner, 2014; Pajevic et al., 2014). Structurally, myelin is composed of tightly packed layers of lipids and key myelination proteins, including proteolipid protein 1 (PLP1), myelin basic protein (MBP), and myelin-associated glycoprotein (Greenfield et al., 1973). These components are crucial for maintaining the myelin layer's compact and multilayered structure (Kister and Kister, 2023), which allows it to preserve axonal integrity and optimize neural transmission.
Alterations in myelination have been associated with alcohol exposure, as abnormal white matter properties were observed in alcohol-dependent patients (Rice and Gu, 2019). For example, alcohol abuse and withdrawal were shown to induce microstructural white matter deficits in cortical areas in diffusion tensor imaging (DTI)/magnetic resonance imaging (MRI) studies (Pfefferbaum et al., 2009; Yeh et al., 2009; Bava et al., 2013; Zorlu et al., 2013; De Santis et al., 2019). In contrast, other MRI studies reported increased white matter in mesolimbic and cortical regions relevant to addiction (De Bellis et al., 2008; Cardenas et al., 2013; Morris et al., 2022). Similar inconsistencies are presented by postmortem studies, with some demonstrating a decline in white matter measurements in AUD patients, compared with controls (de la Monte, 1988; Kril et al., 1997; Lewohl et al., 2001; Liu et al., 2006; Miguel-Hidalgo et al., 2017; de la Monte et al., 2018), and others indicating no such alterations or even increases in white matter measurements (Mayfield et al., 2002; Liu et al., 2004; Lewohl et al., 2005).
Animal model studies have shown that ethanol exposure reduces myelin density and alters the integrity and composition of the myelin sheath in several brain regions, including the corpus callosum, hippocampus, and PFC (Alfonso-Loeches et al., 2012; Pascual et al., 2014; Vargas et al., 2014; Montesinos et al., 2015, 2017; Samantaray et al., 2015; De Santis et al., 2019; Degiorgis et al., 2022). However, other studies found ethanol to increase the expression of various myelin- and OL-related genes in the PFC, hippocampus, and amygdala (Weng et al., 2009; Pascual et al., 2014; Tong et al., 2015; Somkuwar et al., 2016; Narendra et al., 2022). Finally, hippocampal overexpression of myelin transcription factor 1, which modulates the proliferation of OL lineage cells (Nielsen et al., 2004), decreased ethanol intake in rats (Bahi and Dreyer, 2017), suggesting that this transcription factor negatively regulates ethanol drinking.
These inconsistent findings may result from the use of different ethanol-exposure protocols. Specifically, most of the previous animal studies that reported ethanol-induced disruption of myelination used short-term drinking models [e.g., binge drinking (Treadwell and Singh, 2004; Weng et al., 2009; Pascual et al., 2014; Montesinos et al., 2015, 2017; Pfefferbaum et al., 2015; Wolstenholme et al., 2017; Tavares et al., 2019) or chronic forced exposure to ethanol by inhalation (Vargas et al., 2014; Kim et al., 2015; Samantaray et al., 2015; Somkuwar et al., 2016)] rather than long-term voluntary ethanol consumption, which provides a better model of the human ethanol addiction pattern. Moreover, the effects of ethanol consumption on myelination have only rarely been assessed in the NAc, a main component of the mesolimbic system, which plays a critical role in alcohol addiction (Ron and Barak, 2016; Abrahao et al., 2017). Thus, the present study characterized myelin-related alterations in the NAc induced by long-term voluntary ethanol consumption and tested whether myelin alterations regulate ethanol consumption.
Materials and Methods
Animals
Female and male C57BL/6J (20–25 g at the beginning of the study) were bred at Tel Aviv University and housed at 21–23°C under a 12 h light/dark cycle (lights on at 07:00, lights off at 19:00), with food and water available ad libitum. All experimental protocols conformed to the guidelines of the Institutional Animal Care and Use Committee of Tel Aviv University and NIH guidelines (A5010-01). All efforts were made to minimize the number of animals used and their suffering.
Reagents
Fast SYBR Green Master Mix (catalog #4385618), TRIzol reagent (catalog #15596026), and RevertAid kit (catalog #K1691) were supplied by Thermo Fisher Scientific. Protease inhibitor cocktail 1 (catalog #539131) was purchased from Merck. DNA oligonucleotides, isopropanol (catalog #563935), glutaraldehyde (catalog #G5882-50ml), paraformaldehyde (PFA; catalog #P6148-1kg), sodium cacodylate (catalog #233854), and OsO4 (catalog #251755-5ml) were obtained from Sigma-Aldrich. Isoflurane (catalog #66794-017-25) was obtained from Piramal Critical Care. Ethanol absolute (catalog #052523), purchased from Bio-Lab, was diluted to 20% (v/v) with tap water. 4,6-Diamidino-2-phenylindole (DAPI; catalog #D9542) was purchased from Sigma-Aldrich, anti-GAPDH antibodies (catalog #sc-32233) were purchased from Santa Cruz Biotechnology, anti-H3K9me3 antibodies (catalog #ab8898) and anti-MBP antibodies (catalog #ab123499) were purchased from Abcam, anti-OL transcription factor 2 (OLIG2; catalog #AB-9610) and anti-CC1 antibodies (catalog #OP-80) were purchased from Merck Millipore, anti-Ki-67 antibodies (catalog #PA5-19462) were purchased from Invitrogen, and anti-platelet–derived growth factor α receptor (PDGFRα) antibodies (catalog #14-1401-82) were purchased from Thermo Fisher Scientific. Mounting medium (catalog #E18-18) was purchased from Origene Technologies. Glycid ether (catalog #21045.02) was purchased from Serva. Uranyl acetate (catalog #22400-4), lead citrate (catalog #17800), and formvar/carbon-coated grids (catalog #FCF 200-cu-50) were purchased from Electron Microscopy Sciences. Clemastine (catalog #1453) was purchased from Tocris Bioscience.
Intermittent access to 20% ethanol two-bottle choice (IA2BC)
After a week of habituation to individual housing, mice were trained to consume ethanol according to the IA2BC procedure, as previously described (Carnicella et al., 2014; Even-Chen et al., 2017, 2022; Even-Chen and Barak, 2019; Ziv et al., 2019). Briefly, mice received three 24 h sessions of ad libitum access to a two-bottle choice per week [tap water and 20% (v/v) ethanol] on Sundays, Tuesdays, and Thursdays, with a 24 or 48 h period of ethanol-deprivation between ethanol-drinking sessions. During withdrawal periods, the mice received only water. The position (left or right) of each solution was alternated between sessions to control for side preference. Water and ethanol bottles were weighed before and after each ethanol-drinking session, and consumption levels were normalized to body weight. The training lasted 3–12 weeks before brain extraction or pharmacological manipulation.
Clemastine administration
After 12 weeks of ethanol consumption as described above, clemastine was dissolved in a vehicle consisting of 10% DMSO in PBS and injected to the mice 1 h before each 24 h drinking session at a dose of 10 mg/kg intraperitoneally in a volume of 10 ml/kg for 7 weeks. This dosage was selected based on previous reports investigating the effects of systemic clemastine administration on myelination (Pan et al., 2020; Leyrolle et al., 2022).
Quantitative reverse transcriptase polymerase chain reaction (qRT-PCR)
qRT-PCR was conducted as we previously described (Even-Chen et al., 2017; Zipori et al., 2017; Even-Chen and Barak, 2019; Ziv et al., 2019; Goltseker et al., 2021, 2023). Following brain dissection, tissue samples were immediately snap-frozen in liquid nitrogen and stored at −80°C until use. Frozen tissues were mechanically homogenized in TRIzol reagent, and total RNA was isolated from each sample according to the manufacturer's instructions. mRNA was reverse transcribed to cDNA using a reverse transcription system and RevertAid kit. Plates (96 wells) were prepared for SYBR Green cDNA analysis using Fast SYBR Master Mix. Thermal cycling was initiated by incubation at 95°C for 20 s, followed by 40 cycles of PCR with the following conditions: heating at 95°C for 3 s and then at 60°C for 30 s. Samples were analyzed in duplicate with a real-time PCR system (StepOnePlus, Applied Biosystems) and quantified using the ΔΔCt method, relative to Gapdh internal control gene. Changes in mRNA expression in the experimental groups were calculated as a percentage of the control group. Specific primers for the detection of mRNA expression were diluted to 10 mM in ultrapure water according to the manufacturer's instructions (primer sequences are listed in Table 1).
Primers used for qRT-PCR
Immunohistochemistry
A standard immunohistochemistry procedure was employed to detect the presence of various markers (Barak et al., 2011, 2013, 2019). Briefly, mice were deeply anesthetized with isoflurane and transcardially perfused with a 15 ml ice-cold phosphate–buffered saline (PBS) solution, followed by 15 ml fresh ice-cold 4% PFA in PBS. Brains were dissected and kept in 4% PFA overnight at 4°C and then sectioned into 100-μm-thick slices using a vibratome (Leica Biosystems). Floating brain slices were washed with PBS for 5 min thrice and permeabilized with 1% bovine serum albumin (BSA) in TBS-Tween 20 (TBST; 0.05 M Tris–HCl, 0.15 M NaCl, and 0.1% Tween 20), pH 7.5, for 1 h. An array of primary antibodies diluted in blocking buffer were applied to sections overnight at 4°C at the following dilutions: anti-MBP antibodies (1:500); anti-OLIG2 antibodies (1:500); anti-CC1 antibodies (1:500); anti-PDGFRα antibodies (1:700); anti-H3K9me3 antibodies (1:400); or anti-Ki-67 antibodies (1:1,000). Slices were then washed with TBS for 5 min thrice and incubated for 2 h at room temperature (RT) with Alexa Fluor 488 (green)-labeled goat anti-rabbit, DyLight 550 (red)-labeled goat anti-chicken, goat anti-mouse, or goat anti-rat secondary antibodies. Slices were washed with TBS for 5 min thrice, followed by DAPI (1:10,000 in TBS) staining for 5 min, and an additional three 5 min washes in TBS. Sections were mounted on positively charged slides, and coverslips were added using mounting medium. Images were acquired using a light microscope (IX-83, Olympus), with the experimenter blind to the treatment type. For quantification of MBP staining and cellular properties in the NAc, images were taken at 10× or 20× magnification. Cell numbers and intensity were quantified using the ImageJ program.
Western blot
Brains were dissected from adult mice and homogenized immediately in RIPA buffer containing (in mM) 50 Tris–HCl, 120 NaCl, and 5 EDTA and 0.5% (w/v) sodium deoxycholate, 0.1% (w/v) sodium dodecyl sulfate (SDS), 1% Triton X-100, and protease inhibitor cocktail 1 (diluted 1:100), pH 7.6. Protein concentration was determined using a BCA assay, and equal amounts of protein from each sample were separated by SDS–polyacrylamide gel electrophoresis through a 12% gel. The gel was electrophoretically transferred to a nitrocellulose or PVDF membrane in transfer buffer (25 mM Tris–HCl, 190 mM glycine, and 10% methanol absolute). Membranes were blocked for 1 h in TBST with 5% BSA and blotted with primary antibodies diluted in blocking buffer overnight at 4°C. Following washes with TBST, the membranes were incubated for 1 h at RT with the appropriate horseradish peroxidase-conjugated secondary antibodies (1:5,000, anti-mouse antibodies). After extensive washing with TBST, antibody binding was visualized using enhanced chemiluminescent substrate (GE Healthcare) and captured using the ImageQuant LAS 500 imager system. The optical density of the relevant immunoreactive band was quantified using the ImageLab software (version 6.1, Bio-Rad Laboratories). MBP and GAPDH (the latter used as loading controls) were detected using commercial anti-MBP (1:10,000) and anti-GAPDH (1:10,000) antibodies. Results are expressed as the percentage of control group band intensity.
Transmission electron microscopy (TEM)
Following 12 weeks of training in the IA2BC procedure (with water-drinking controls), mice were deeply anesthetized with isoflurane, transcardially perfused with 15 ml ice-cold PBS, followed by 15 ml fresh ice-cold 2.5% glutaraldehyde + 2% PFA in 0.1 M sodium cacodylate buffer, pH 7.4. Brains were dissected into small pieces (2 mm3 cubes), and tissue samples from the NAc were kept in the fixation solution overnight at 4°C. After several washings with PBS, the tissues were postfixed in 1% OsO4 in PBS for 2 h at 4°C. Dehydration was carried out using a graded ethanol series, followed by embedding in glycid ether. Ultrathin sections (70–100 nm) were mounted on formvar/carbon-coated grids, stained with uranyl acetate and lead citrate, and examined in a JEOL 1400 Plus transmission electron microscope. Images were captured using SIS Megaview III and the iTEM imaging platform (Olympus). These procedures were carried out at the Electron Microscopy Unit, Life Sciences, Tel Aviv University. To quantify the g-ratio, and myelin thickness, the inner and outer diameters of the myelin sheath from fibers randomly selected were analyzed by the GRatio plug-in (http://gratio.efil.de) in ImageJ (https://imagej.nih.gov/ij), allowing for semiautomated analysis. Analysis was performed by an experimenter blinded to the experimental groups.
MRI acquisition
MRI scans were conducted in the Strauss Computational Neuroimaging Center at Tel Aviv University by a Bruker Biospec 7T/30 scanner equipped with a 660 mT/m gradient unit, using a cross-coil configuration of an 86 mm transmissive volume coil and a mouse quadrature coil as a receiver. The MRI scan protocol included structural T2-weighted (T2w) images that were acquired with rapid acquisition with relaxation enhancement (RARE) sequence and DTI acquisition with a diffusion-weighted spin–echo echoplanar imaging (EPI) pulse sequence. T2w acquisition was conducted with the following parameters: TR, 2,500 ms; effective TE, 31.2 ms; RARE Factor 8 with eight repetitions; and 20 coronal slices, 0.6 mm thick (no gaps), with an in-plane resolution of 0.0875 mm2 covering the entire brain and lasting 4:00 min. For DTI acquisition, we used TR/TE, 3,000/20 ms; Δ/δ, 10/2.5 ms; two EPI segments; 30 gradient directions with a b value of 1,000 s/mm2; and three B0 images; 28 axial slices, 0.6 mm thick (no gaps); and in-plane resolution of 0.175 mm2. The DTI acquisition took ∼10 min. Mice were anesthetized throughout the scanning procedure using isoflurane (1.5% in pure oxygen). A heating system was used to maintain the animal's body temperature, and their respiration was monitored and maintained at 30–50 breaths/min using a pneumatic balloon positioned against the animal's chest. Total MRI protocol acquisition took ∼40 min.
DTI analysis and tractography
The DTI dataset was first corrected for head movement and eddy current distortion using the ExploreDTI platform (Leemans et al., 2009) within MATBLAB (MathWorks). To perform the EC/EPI correction, the dataset underwent nonlinear tensor estimation with a correction to the structural T2w image. For the data quality check, we first reviewed the DTI data by visually inspecting the slice images, followed by an inspection of the motion correction parameters. Whole-brain tractography (Basser et al., 2000) was then performed using a seed point resolution similar for all samples (0.875 × 0.875 × 6 mm), a seed FA threshold of 0.05, and a step size of 0.875. The range of the length threshold was set to 50–500 mm to exclude ultrashort/big and spurious fibers, and an angle threshold of max 30° between contiguous voxels was acquired. Then, based on the tractography, we quantified the number of reconstructed tracts connecting the NAc and the PFC. All analyses were conducted using the ExploreDTI program (Leemans et al., 2009).
Blood ethanol concentration (BEC) assay
Blood was collected with a heparinized capillary tube, 900 or 1,800 s after the alcohol injection. Blood samples were centrifuged for 10 min at 7,000 rpm, and serum was extracted. BEC was assessed using the alcohol dehydrogenase (ADH) assay (Weiss et al., 1993; Zapata et al., 2006; Carnicella et al., 2011). Briefly, serum (10 µl) was incubated in 200 µl reaction buffer [0.5 M Tris–HCl buffer; 2.75 µg/ml of ADH and 0.5 mg/ml β-nicotinamide adenine dinucleotide (β-NAD)], pH 8.8, for 30 min at RT. Accumulation of β-NADH was determined by reading the absorbance at 340 nm. BEC was estimated using a standard calibration curve (Zapata et al., 2006; Even-Chen et al., 2022).
Experimental design and statistical analysis
Female and male mice were distributed approximately equally across all experiments, with sex being initially analyzed as a factor. Our analysis did not yield any interaction of sex with other factors measured (p > 0.05). As such, the data were collapsed across this factor. The detailed n per group and female/male distribution are shown in the Extended Data tables.
Effects of long-term ethanol consumption on myelin-related gene expression
We assessed the effects of voluntary ethanol consumption on myelin-related gene expression following 3, 6, or 12 weeks of drinking in the IA2BC drinking protocol. mRNA expression levels of target genes were determined at the end of a 24 h ethanol-withdrawal session. The expression levels in the NAc were normalized to that of Gapdh expression (Even-Chen et al., 2017; Zipori et al., 2017; Ziv et al., 2019; Goltseker et al., 2023), which was found to be unaffected by ethanol, according to the protocols used (Even-Chen et al., 2017). mRNA expression levels of target genes are expressed as a percentage of the corresponding control (water-drinking) group. qRT-PCR data were analyzed by two-way (2 × 3) factorial ANOVA with the between-subject factors of drinking solution (water, ethanol) and drinking duration (3, 6, 12 weeks). LSD post hoc analysis was used where indicated. Sample sizes were chosen based on our previous studies with similar measures (Barak et al., 2013; Even-Chen and Barak, 2019; Ziv et al., 2019; Goltseker et al., 2021, 2023; Even-Chen et al., 2022).
Effects of long-term ethanol consumption on MBP levels
We assessed the effects of long-term (12 week) ethanol consumption in the IA2BC drinking protocol on MBP levels in the NAc by Western blot and immunofluorescence. Brains were dissected at the end of a 24 h ethanol-withdrawal session. MBP levels were then determined by the two methods. In the Western blot experiment, the immunoreactivity of MBP was normalized to that of GAPDH (Barak et al., 2013; Goltseker et al., 2023) and was expressed as a percentage measured in the water-drinking control group. In the immunofluorescence experiment, MBP immunoreactivity was quantified by calculating fluorescence intensity (Hong et al., 2017), expressed as the percentage of that measured with the water-drinking control group. Data were analyzed by independent t tests. Sample sizes were chosen based on our previous studies with similar measures (Barak et al., 2013; Goltseker et al., 2023; Hose et al., 2024).
Effects of long-term ethanol consumption on OL lineage cells
We assessed the effects of voluntary ethanol consumption (3, 6, and 12 weeks in the IA2BC drinking procedure) on the OL lineage cell in the NAc by immunofluorescence. To quantify OL lineage cells, we determined the number of OPCs by staining for PDGFRα, a cell surface tyrosine kinase receptor used as a marker of OPCs (Rivers et al., 2008). To specifically identify mOLs among all OL lineage cells, we costained for OLIG2, a transcription factor used to label all OL lineage cells (from OPCs to mOLs; Fancy et al., 2009), together with anti-CC1 antibodies, which bind to the adenomatous polyposis coli (APC) protein, used as a marker of mOLs (Bhat et al., 1996). The number of OL lineage cells was quantified and expressed as a percentage of those numbers in the control (water-drinking) group. In addition, we assessed the effects of voluntary ethanol consumption (12 weeks in the IA2BC drinking procedure) on proliferation and differentiation properties of OLs in the NAc by immunofluorescence. To quantify the number of proliferating OPCs, we costained for PDGFRα, to mark OPCs, and Ki67 to mark proliferating cells. To quantify differentiation properties of mOLs, we costained for CC1, to mark mOLs, and H3K9me3, a histone modification essential for OL differentiation. Intensity of H3K9me3 within mOLs was calculated by the mean of the gray value divided by the cell area. Data from the immunofluorescence studies were analyzed by two-way (2 × 3) factorial ANOVA with the between-subject factors of drinking solution (water, ethanol) and drinking duration (3, 6, 12 weeks). LSD post hoc analysis was used where indicated. For proliferation and differentiation experiments, data were analyzed by unpaired t test or by Kolmogorov–Smirnov test. Sample sizes were chosen based on our previous studies with similar measures (Barak et al., 2011, 2013, 2015).
TEM experiment
We assessed the effects of ethanol consumption on myelin ultrastructure in the NAc. Specifically, we addressed the effects of voluntary ethanol consumption on myelin ultrastructure after 12 weeks of applying the IA2BC procedure, described above. Axon diameter, myelin thickness, and g-ratio were measured using the GRatio software in ImageJ on at least 300 myelinated axons randomly chosen from each group. Specifically, the axon diameter was calculated by measuring the diameter of each myelinated axon within the inner borders of the myelin sheath, excluding the myelin sheath. Myelin thickness was calculated by subtracting the axon diameter from the fiber diameter and dividing by two, and g-ratio was calculated by measuring the axon diameter divided by the diameter of the axon with its myelin sheath (Goebbels et al., 2010; Bélanger et al., 2011; Della Santa et al., 2021). Thus, the smaller was the g-ratio, the thicker was the myelin sheath layer. Data from the TEM experiment were analyzed by independent t tests or fitted by simple linear regression. Sample sizes were chosen based on previous studies with similar measures (West et al., 2018; Barak et al., 2019).
DTI experiment
To assess the number of tracts between the NAc and the PFC, we conducted an in vivo DTI study, scanning the brains of mice after 12 weeks of voluntary ethanol consumption or only water consumption as controls. We then analyzed the DTI data produced to assess the connectivity modifications induced by long-term ethanol consumption. We used the t test to analyze these data.
Clemastine administration experiment
Mice were trained in the IA2BC drinking procedure for 12 weeks. The effects of clemastine (10 mg/kg, i.p.) administered daily for 7 weeks (while continuing drinking according to the IA2BC procedure) on ethanol consumption and preference, as well as on water intake, were then assessed. Dosages were based on those used in earlier reports (Pan et al., 2020; Leyrolle et al., 2022). Data were analyzed using mixed-model two–way ANOVA with the between-subject factor of treatment (vehicle, clemastine) and a repeated-measure factor of the duration of treatment in weeks. Ethanol consumption and preference were averaged for each week. LSD post hoc analysis was used where indicated. Sample sizes were chosen based on our previous studies with similar measures (Barak et al., 2013; Even-Chen and Barak, 2019; Ziv et al., 2019; Goltseker et al., 2021; Even-Chen et al., 2022).
Results
Long-term excessive ethanol consumption increases myelin-related gene expression in the NAc
We initially sought to determine whether ethanol alters the expression of genes involved in brain myelination in the mesocorticolimbic and nigrostriatal pathways by quantifying mRNA levels. To that end, we trained mice to drink ethanol in the IA2BC procedure for 3, 6, or 12 weeks and examined the mRNA levels of several genes implicated in the differentiation and maturation of OLs in the NAc following a 24 h washout period (Fig. 1A). This drinking procedure yielded very high and escalating ethanol consumption and preference levels (Fig. 1B,C), which generated high blood ethanol concentrations (BEC, 67–120 mg%; average 97 ± 16), positively correlating with ethanol intake levels (r = 0.85; p < 0.05).
Voluntary ethanol consumption increases myelin-related gene expression in the NAc in a time-dependent manner. Amount of alcohol (g/kg/24 h) consumed over 84 d (12 weeks). A, Mice consumed ethanol in the intermittent access to 20% ethanol two-bottle choice procedure for 3, 6, or 12 weeks. Control mice consumed only water. Tissues were collected 24 h after the last drinking session (ethanol group). B, C, Ethanol intake (B) and preference (C) over 84 d. Data are expressed as mean ± SEM intermittent daily averages. D–I, mRNA levels of Sox10 (D), Olig2 (E), Myrf (F), Mbp (G), Plp1 (H), and Mag (I) in the NAc, as determined by qRT-PCR and normalized to the level of Gapdh mRNA. Bar graphs represent mean + SEM. n = 5–8 per group; for more details on n per group and sex distribution, see Extended Data Table 1-1; *p < 0.05; **p < 0.01; relative to corresponding water-drinking controls.
Table 1-1
n per group and sex distribution in experiments depicted in Figure 1. Download Table 1-1, DOCX file.
We first examined the expression of Olig2 and Sox10 in response to ethanol consumption, as these transcription factors are essential for OL differentiation and regulate early myelination and downstream gene expression (Yu et al., 2013; Elbaz and Popko, 2019; Qi et al., 2022). We found that while there was a decrease in Olig2 expression at the 3 week time point, the mRNA levels of both genes were increased following 6 weeks of heavy ethanol consumption, and there was a slight decrease in Sox10 levels following 12 ethanol-drinking weeks (Fig. 1D,E): two-way ANOVA, Sox10, a main effect of drinking duration (F(2,30) = 5.786; p = 0.008) and a drinking solution × duration interaction (F(2,30) = 19.96, p = 0.008), but no main effect of drinking solution (F(1,30) = 0.214; p = 0.647); post hoc, ethanol versus water at 6 weeks (p = 0.020) and 12 weeks (p = 0.040), but not at 3 weeks (p > 0.05); Olig2, a main effect of drinking duration (F(2,31) = 4.7; p = 0.015) and a drinking solution × duration interaction (F(2,31) = 4.7; p = 0.015), but no main effect of drinking solution (F(1,31) = 0.01; p = 0.982); and post hoc, ethanol versus water at 3 weeks (p = 0.043) and 6 weeks (p = 0.030), but not at 12 weeks (p > 0.05).
Among the downstream targets regulated by SOX10 and OLIG2 is the myelin regulatory factor (Myrf; Hornig et al., 2013; Lopez-Anido et al., 2015; Sock and Wegner, 2021), a critical transcription factor involved in the later stages of OL differentiation and myelin production (Emery et al., 2009; Cantone et al., 2019). Given the observed changes in Sox10 and Olig2 expression, we next examined the expression of Myrf in response to ethanol consumption. We found that ethanol consumption led to a time-dependent increase in Myrf expression in the NAc compared with water-drinking controls, which emerged only following 12 weeks of ethanol drinking (Fig. 1F): two-way ANOVA, a main effect of drinking duration (F(2,33) = 6.759; p = 0.003) and a drinking solution × duration interaction (F(2,33) = 6.759; p = 0.003), but no main effect of drinking solution (F(1,33) = 0.965; p = 0.331); and post hoc, ethanol versus water at 12 weeks (p = 0.001), but not at 3 or 6 weeks (p > 0.05).
We then evaluated the effects of ethanol consumption on the expression of genes encoding major proteins involved in myelin formation and maintenance, namely, Mbp (Boggs, 2006), Plp1 (Aggarwal et al., 2011), and Mag (Li et al., 1994; Quarles, 2007). These genes were chosen due to their pivotal roles in myelination processes and because their expression is regulated by the transcription factors analyzed earlier. We found that ethanol consumption caused time-dependent increases in the expression of all three genes in the NAc, which emerged only following 12 weeks of ethanol drinking (Fig. 1G–I): two-way ANOVA, NAc, Mbp, a significant main effect of drinking solution (F(1,34) = 7.33; p = 0.010), a trend to main effect of drinking duration (F(2,34) = 2.954; p = 0.066) and a drinking solution × duration interaction (F(2,34) = 2.954; p = 0.066); Plp1, a main effect of drinking duration (F(2,33) = 5.168; p = 0.011) and a drinking solution × duration interaction (F(2,33) = 5.168; p = 0.011), but no main effect of drinking solution (F(1,33) = 0.784; p = 0.383); Mag, a significant main effect of drinking duration (F(2,35) = 5.852; p = 0.006) and a drinking solution × duration interaction (F(2,35) = 5.852; p = 0.006), but no main effect of drinking solution (F(1,35) = 1.033; p = 0.316); and post hoc, ethanol versus water at 12 weeks, Mbp, p = 0.004; Plp1, p = 0.008; Mag, p = 0.003.
Together, these results indicate a temporal transcriptional pattern in the NAc, with Olig2 and Sox10 upregulated at the 6 week time point, followed by the increased expression of Myrf and the myelination-related genes Plp1, Mbp, and Mag at the 12 week time point.
Long-term excessive ethanol consumption increases MBP protein levels, number of myelinated fibers, and NAc–PFC connectivity
Since chronic ethanol consumption affected the expression of myelin-related gene expression in the NAc, we next determined whether changes in the protein levels of MBP in the NAc occurred after long-term voluntary consumption of high levels of ethanol. Accordingly, mice consumed high levels of ethanol for 12 weeks in the IA2BC procedure, as described above, after which MBP levels were determined by a Western blot. We found that chronic ethanol consumption increased MBP levels in the NAc, compared with water-drinking controls. Furthermore, our immunohistochemistry quantification indicated that the alcohol-drinking group had a higher number of MBP-positive events than water-drinking controls, likely representing myelinated fibers within the NAc, with a similar area per event in both groups (Fig. 2B,C; Western blot; t(16) = 2.841; p = 0.011; Fig. 2D–F; immunohistochemistry; number of MBP events; t(15) = 2.81; p = 0.013; MBP area of events t(15) = 0.35; p = 0.734). These results further indicate that chronic ethanol consumption leads to increases in the levels of MBP in the NAc.
Long-term ethanol consumption increases MBP protein levels in the NAc and connectivity between the NAc and the PFC. Mice consumed ethanol in the intermittent access to 20% ethanol two-bottle choice procedure for 12 weeks. Control mice consumed only water. Tissues were collected 24 h after the last drinking session (ethanol group). A, Schematic representation of the sampling region (Paxinos and Franklin, 2019). Scale bar, 100 µm. B, Quantification of Western blot of MBP levels in the NAc, normalized to GAPDH levels, expressed as a percentage of water-drinking controls. C, Representative blots. See Extended Data Figure 2-1 for the whole blot image. D, Representative immunofluorescence images of MBP staining and DAPI nuclear staining in the NAc. E, Quantification of the number of MBP-positive events in the NAc. F, Mean area of MBP-positive events in the NAc. G, Representative images of tracts between the NAc and the PFC, analyzed from DTI scans. H, Quantification of the number of tracts between the NAc and PFC. Bar graphs represent mean + SEM. n = 8–9 per group for immunofluorescence and Western blot; n = 21–23 per group for DTI; for more details on n per group and sex distribution, see Extended Data Table 2-1. *p < 0.05.
Figure 2-1
Western blot visualization of MBP in the nucleus accumbens (NAc). Mice consumed alcohol in the intermittent access to 20% ethanol (E) 2-bottle choice procedure for 12 weeks. Control mice consumed water only (W). Tissues were collected 24 h after the last drinking session. MBP and the GAPDH loading control were labeled with appropriate antibodies. Download Figure 2-1, TIF file.
Table 1-2
n per group and sex distribution in experiments depicted in Figure 2. Download Table 1-2, DOCX file.
Consistent with these findings, DTI analysis revealed an increased number of tracts connecting the NAc and the PFC in the ethanol group, compared with the water-drinking control group. These results align with the staining data and indicate enhanced connectivity between these regions (Fig. 2G,H; tractography; t(42) = 2.063; p = 0.045).
Long-term excessive ethanol consumption impairs myelin ultrastructure in the NAc
Given how we found that long-term ethanol consumption led to alterations in the levels of myelin-associated mRNA and proteins in the NAc, we next considered whether these molecular changes are also associated with alterations at the level of myelin sheath structure. To this end, we used TEM to analyze the effects of long-term ethanol consumption (12 weeks, as described above) on myelin ultrastructure in the NAc.
As shown in Fig. 3, we found that ethanol consumption caused a significant increase in myelin thickness, without affecting the axon diameter, as also reflected by the decreased g-ratio, a parameter used to assess the extent of axonal myelination. Plotting g-ratio values as a function of axon diameter revealed that reduced g-ratios, indicative of increased myelin thickness, occurred across all axon sizes (Fig. 3B). Thus, long-term ethanol consumption increased the amount of myelin per axon [t tests, significant effects for g-ratio (t(6) = 5.29; p = 0.002) and myelin thickness (t(6) = 4.66; p = 0.004), but not for axon diameter (t(6) = 1.53; p = 0.177)]. Ultrathin sections of the NAc from the water-drinking control group revealed axons surrounded by regular myelin sheaths with preserved compact lamellar structure. In contrast, the ethanol-drinking group presented impaired formations of myelin loops and wide separations within the myelin lamellar structure, displaying focal disorganization of the myelin sheath (Fig. 3A). Together, these results indicate that long-term consumption of high ethanol quantities leads to structurally enhanced myelination in the mouse NAc.
Long-term ethanol consumption affects myelin ultrastructure in the NAc. A, Representative images of myelin ultrastructural in the NAc of ethanol-drinking mice, as compared with water-drinking controls. Scale bars, 500 nm. B, A scatterplot of individual g-ratio versus axon diameter. Approximately 300 myelinated axons were considered per group. The data were fitted by simple linear regression, and slopes and intercepts are indicated. C, Quantification of g-ratio in the NAc. D, Myelin thickness (half of the difference between the axon + myelin diameter and the axon diameter in micrometer). E, Axon diameter. Data show mean + SEM. n = 4 per group (2 males, 2 females); **p < 0.01.
Long-term ethanol consumption affects the development of OL lineage cells
To further address the effects of ethanol consumption on adaptive myelination processes in the NAc, we tested whether the OL cell populations were altered after 3, 6, and 12 weeks of voluntary ethanol drinking in the IA2BC procedure.
First, we determined the number of OPCs by immunostaining PDGFRα, considered a marker of OPCs. We found that 3 weeks of ethanol drinking increased the number of PDGFRα-positive cells in the NAc of the ethanol-drinking group, as compared with water-drinking controls. In contrast, after 6 weeks, ethanol-drinking mice presented fewer PDGFRα-positive cells than did the water-drinking controls, whereas after 12 weeks of ethanol drinking, no differences were observed between the groups (Fig. 4A).
Long-term ethanol consumption affects OL lineage cells in the mouse NAc in a time-dependent manner. Mice consumed ethanol in the intermittent access to 20% ethanol two-bottle choice procedure for 3, 6, or 12 weeks. Control mice consumed only water. Tissues were collected 24 h after the last drinking session (ethanol group). A–C, To identify OPC, total OL and mOL immunohistochemistry was performed using antibodies against PDGFRα+ (A), OLIG2+ (B), the mOLs marker APC (CC1+; C) and the OL lineage marker OLIG2+. D–F, Representative immunofluorescence staining image of anti-PDGFRα (blue; D), anti-OLIG2 (green; E), and anti-OLIG2 + CC1 (merge in yellow; F). Scale bar, 100 µm. Bar graphs represent mean + SEM. n = 7–11 per group; for more details on n per group and sex distribution, see Extended Data Table 4-1. *p < 0.05; **p < 0.01; ***p < 0.0001.
Table 1-4
n per group and sex distribution in experiments depicted in Figure 4. Download Table 1-4, DOCX file.
Next, we coimmunostained for OLIG2, a general marker of OL lineage cells, and for APC (using anti-CC1 antibodies), used as a marker of mOLs. We observed no differences in OLIG2 staining between groups at all drinking time points (Fig. 4B). However, the number of OLIG2 + CC1-positive cells was downregulated in the ethanol group, compared with water-drinking controls at the 3 week time point yet increased after 6 weeks of ethanol drinking. No difference was found between the groups at the 12 week time point (Fig. 4C): two-way ANOVA, PDGFRα, a main effect of drinking duration (F(2,37) = 7.040; p = 0.003) and a significant drinking solution × duration interaction (F(2,37) = 7.040; p = 0.003), but no effect of drinking solution (F(1,37) = 0.055; p = 0.815); post hoc, significant differences between ethanol versus water at the 3 week (p = 0.005) and 6 week time points (p = 0.029), but not at 12 weeks (p = 0.824); OLIG2, no effects of duration of drinking (F(2,37) = 0.057; p = 0.944) or drinking solution (F(1,37) = 0.086; p = 0.771) and no drinking solution × duration interaction (F(2,37) = 0.057; p = 0.944); OLIG2 + CC1, a main effect of duration of drinking (F(2,40) = 13.58; p < 0.0001) and a drinking solution × duration interaction (F(2,40) = 13.58; p < 0.0001), but no effect of drinking solution (F(1,40) = 0.723; p = 0.400); and post hoc, significant differences between ethanol and water at the 3 week (p = 0.009) and 6 week (p < 0.0001) time points, but not at 12 weeks (p = 0.597).
Together, our findings show that 3 weeks of ethanol consumption led to an increase in the number of OPCs and a decrease in the number of mOLs with no change in the total number of OLs, suggesting that ethanol impaired OPC differentiation in the initial stages of drinking. In contrast, 6 weeks of ethanol consumption led to a decrease in the number of OPCs and an increase in the number of mOLs, as compared with water-drinking controls, suggesting that at this stage, ethanol led to enhanced OPC differentiation. Finally, these apparent OL differentiation impairments were not seen after 12 weeks of ethanol drinking. To further investigate the proliferation and differentiation properties of OLs, we analyzed the number of proliferating OPCs and the expression of H3K9me3, a histone modification critical for OL differentiation (Liu et al., 2015), in mOLs at the 12 week time point. To assess proliferation, we quantified the number of Ki67 + PDGFRα-positive cells, where Ki67 is a marker for proliferating cells. We observed no significant difference in the number of Ki67 + PDGFRα-positive cells between the ethanol and water-drinking control groups (Fig. 5A,B; t(10) = 0.633; p = 0.54), consistent with our findings above, which showed no change in the number of OPCs at 12 weeks. However, we detected a trend toward increased H3K9me3 expression in CC1-positive cells in the ethanol group compared with the water-drinking control group (Fig. 5C,D; frequency distribution, Kolmogorov–Smirnov d = 1.66; p = 0.072; t test for total, t(11) = 0.732; p = 0.336). These findings align with, and further support, our observation above (Fig. 4), showing an increased number of mOLs at the 6 week time point, indicating enhanced differentiation after chronic ethanol consumption.
Effects of long-term ethanol consumption on OL proliferation and differentiation properties. Mice consumed ethanol in the intermittent access to 20% ethanol two-bottle choice procedure for 12 weeks. Control mice consumed only water. Tissues were collected 24 h after the last drinking session (ethanol group). Immunohistochemistry was performed to evaluate OPC proliferation and OL differentiation properties. A, Number of Ki67+ PDGFRα+ cells in the NAc, determined by immunohistochemistry, expressed as the percentage of the water-drinking control group's value. B, Representative immunofluorescence images of Ki67+ (green), PDGFRα+ (yellow), and Ki67+ PDGFRα+ merged images, with arrows pointing to PDGFRα+Ki67 + colocalized cells. Scale bar, 50 µm. C, H3K9me3 intensity in CC1 + cells. D, Representative immunofluorescence images of H3K9me3+ (green), CC1+ (yellow), and H3K9me3 intensity in CC1 + cells, arrows pointing to CC1+ H3K9me3 + colocalized cells. Scale bar, 50 µm. Bar graphs represent mean + SEM. n = 6–7 per group; for more details on n per group and sex distribution, see Extended Data Table 5-1.
Table 1-5
n per group and sex distribution in experiments depicted in Figure 5. Download Table 1-5, DOCX file.
Enhancement of myelination by clemastine treatment increases ethanol drinking and preference in long-term ethanol-drinking mice
While we found that chronic ethanol drinking increases myelination, it remained unclear whether the increased myelination contributed to the escalation in ethanol drinking seen. Therefore, we next evaluated whether further enhancement of myelin formation would affect ethanol consumption in mice with a history of ethanol drinking. After training mice for 12 weeks in the IA2BC procedure, mice were treated with clemastine (10 mg/kg), a promyelination drug known to enhance MBP expression and OL differentiation (Liu et al., 2016; Green et al., 2017; Cree et al., 2018; Wang et al., 2018) or with a vehicle. Clemastine was administered daily for 7 weeks (Li et al., 2015; Lee et al., 2021) while continuing the ethanol-drinking procedure (Fig. 6A).
Chronic systemic administration of clemastine elevates ethanol consumption and increases MBP expression and mOL number in the NAc of mice. A, Experimental scheme. Following 12 weeks of ethanol drinking in the intermittent access to 20% ethanol two-bottle choice procedure, daily treatment with clemastine (10 mg/kg, i.p.) or vehicle was initiated for 7 weeks. Mouse ethanol consumption and preference were recorded. NAc tissues were collected 24 h after the last drinking session and processed for Western blot or immunohistochemistry. B, C, Ethanol intake (B) and preference (C) per 24 h session were averaged by week. D, MBP levels in the NAc, as determined by Western blot and normalized to GAPDH levels, expressed as the percentage of the levels determined in the vehicle-treated group. E, Representative Western blot images. See Extended Data Figure 6-1 for the whole blot image. F, G, Identification of total OLs (F) and mOL (G) lineage cells in the NAc, as determined by immunohistochemistry performed using antibodies against the mOLs marker APC (CC1+) and the OL lineage marker OLIG2+. H, Representative immunofluorescence image of OLIG2 + CC1+ cells. Scale bar, 100 µm. Bar graphs represent mean + SEM. n = 12 per group for behavioral data; n = 5–6 for biochemical data; for more details on n per group and sex distribution, see Extended Data Table 6-1; #p = 0.07; *p < 0.05; ***p < 0.001, relative to vehicle-treated controls.
Figure 6-1
Western blot visualization of MBP in the nucleus accumbens (NAc) following clemastine injections. Following 12 weeks of ethanol drinking in the intermittent access to 20% ethanol 2-bottle choice procedure, daily treatment with clemastine (10 mg/kg, i.p.; C) or vehicle (V) was initiated for 7 weeks. NAc tissues were collected 24 h after the last drinking session and processed for Western blot or immunohistochemistry. MBP and the GAPDH loading control were labeled with appropriate antibodies. Download Figure 6-1, TIF file.
Table 1-6
n per group and sex distribution in experiments depicted in Figure 6. Download Table 1-6, DOCX file.
We found that following 5 weeks of daily clemastine treatment, ethanol consumption was increased in the clemastine-treated group, relative to the vehicle-treated group, without altering water intake [Fig. 6B–D; mixed-model ANOVA; ethanol intake, a main effect of duration of treatment (F(6,22) = 8.87; p < 0.0001) and a treatment × duration of treatment interaction (F(6,132) = 4.003; p = 0.001), but no main effect of treatment (F(1,22) = 2.676; p = 0.116); post hoc, clemastine versus vehicle, Week 5, p = 0.051; Week 6, p = 0.012; Week 7, p = 0.024; all other weeks, p > 0.05; ethanol preference, a main effect of duration of treatment (F(6,22) = 3.428; p = 0.013) and a trend toward an effect of treatment (F(1,22) = 3.378; p = 0.079), but no treatment × duration of treatment interaction (F(6,132) = 1.710; p = 0.123); post hoc, clemastine versus vehicle, Week 6, p = 0.012; Week 7, p = 0.042; all other weeks, p > 0.05; water intake, all p’s > 0.05].
Next, we confirmed that clemastine indeed increased MBP expression and OL numbers in the NAc. As expected, we found increased levels of MBP protein in the NAc of the clemastine-treated group, compared with the vehicle-treated controls (t(11) = 2.49; p = 0.03; Fig. 6E,F). Similarly, we found that the number of OLIG2 + CC1-positive cells in the NAc was significantly increased by clemastine treatment (t(8) = 6.25; p = 0.0002) yet without a difference in the number of OLIG2-positive cells (t(8) = 0.102; p = 0.921), confirming the expected increases in mature myelinating OLs (Fig. 6G–I).
Together, these results indicate that promyelinating treatment with clemastine increases ethanol consumption in mice with a history of ethanol drinking. This finding suggests that increased levels of myelin-related factors can enhance ethanol intake and preference.
Discussion
Our findings indicate that long-term (i.e., 12 weeks) voluntary consumption of high ethanol quantities leads to enhanced myelination in the NAc. This includes an observed increase in myelin thickness, which was accompanied by increases in the levels of myelin-related genes and protein and increased connectivity between the NAc and the PFC. In addition, we found time-dependent effects of ethanol drinking on the differentiation and maturation of OPCs, suggesting that ethanol disturbs myelin formation and OL development in a dynamic, time-dependent manner. Finally, we found that the chronic administration of clemastine, which increased the number of mature myelinating OLs and MBP expression in the NAc, enhanced ethanol consumption and preference in mice with a history of voluntary long-term ethanol consumption. Together, our results suggest that long-term voluntary consumption of high levels of ethanol leads to increased myelination in the NAc of adult mice, which in turn maintains and can even enhance ethanol intake, potentially contributing to the escalation of drinking behaviors and the development of an excessive ethanol-drinking phenotype.
While previous rodent studies typically reported ethanol-induced demyelination (Alfonso-Loeches et al., 2012; Montesinos et al., 2015; Samantaray et al., 2015; Tong et al., 2017; Wolstenholme et al., 2017), our findings revealed increased myelination in the NAc after long-term ethanol consumption. This seeming contradiction may be due to differences in study focus and/or methods. Thus, previous studies concentrated on the corpus callosum and cortical regions (Alfonso-Loeches et al., 2012; Vargas et al., 2014; Kim et al., 2015; Samantaray et al., 2015; Somkuwar et al., 2016; Yalcin et al., 2017), whereas the present effort focused on the NAc. Additionally, these earlier studies typically involved shorter ethanol-exposure periods, lasting up to 6 weeks (Pascual et al., 2014; Vargas et al., 2014; Pfefferbaum et al., 2015; Wolstenholme et al., 2017; Tavares et al., 2019), compared with the extended 12 week drinking period followed here. Finally, ethanol-exposure protocols in the different studies varied as well, with the previous studies using forced ethanol exposure, such as by relying on ethanol vapor chambers (Kim et al., 2015; Samantaray et al., 2015; Somkuwar et al., 2016) or injections (Montesinos et al., 2015, 2017), as opposed to our model of voluntary oral consumption.
Here, we used a considerably longer ethanol-drinking protocol, which included intermittent access to 20% ethanol, an approach that generates very high voluntary ethanol consumption levels over 12 weeks (16–20 g/kg/24 h; Fig. 1B,C). This drinking pattern thus resembles alcohol consumption in humans with AUD and, therefore, allows for modeling of brain adaptations induced by heavy ethanol consumption (Becker, 2008; Carnicella et al., 2014; Ron and Barak, 2016; Goltseker et al., 2019). Indeed, we found that the 3 and 6 week drinking periods disrupted normal OL differentiation and did not lead to increases in key myelin-related gene expression, with some genes showing a trend of decreased expression. Thus, our results suggest that although shorter periods of ethanol exposure may induce demyelination processes in several brain regions, long-term exposure to high levels of ethanol induces enhanced myelination in the NAc.
Progressive effects of ethanol on myelin dynamics in the NAc: from initial disruption to enhanced myelination
We show here that 3 weeks of consuming large quantities of ethanol led to disturbed myelin-related processes in the NAc. Specifically, we found increased levels of OPCs and reduced levels of mOLs, compared with water-drinking controls. This suggests impaired OPC differentiation and maturation, potentially indicative of demyelination processes (Zou et al., 2014; Gould and Kim, 2021). Surprisingly, drinking ethanol for an additional 3 weeks (6 week total drinking) resulted in the opposite effect, namely, a decreased number of OPCs and an increased number of mOLs, likely reflecting the enhanced differentiation and maturation of OPCs into myelinating OLs (Fernández-Castañeda et al., 2020). Finally, following a longer drinking period (12 weeks), no differences were observed between ethanol drinkers and water-drinking control group members. One possible explanation is that ethanol consumption delays OL differentiation rather than enhancing it during prolonged exposure. However, our observation of a trend toward increased expression of H3K9me3 within mOLs, coupled with the lack of changes in proliferation properties, suggests that chronic ethanol consumption alters the differentiation process of OLs.
These results are further reinforced by evidence of pronounced myelin-related changes following a longer period of heavy ethanol consumption, i.e., 12 weeks. This period of ethanol consumption led to increased myelin thickness, elevated expression of myelin-related genes and MBP protein levels, all of which suggest enhanced myelination rather than a delay in OL differentiation. Interestingly, we also identified a higher number of MBP-positive events in the NAc, likely reflecting an increased number of myelinated fibers. DTI-mediated tractography further revealed enhanced connectivity between the NAc and the PFC, consistent with the structural changes observed in the NAc. Moreover, TEM analysis showed a significant reduction in the g-ratio of NAc axons in ethanol-exposed mice compared with controls, indicating thicker myelin. It is noteworthy that deviations from the optimal g-ratio (∼0.8) have been associated with reduced myelination efficiency (Chomiak and Hu, 2009), which may have functional implications for neural circuit connectivity.
The processes we observed are tightly regulated by transcription factors critical for OL differentiation and myelination. Chronic ethanol consumption resulted in increased expression of Sox10 and Olig2 at 6 weeks, followed by elevated Myrf expression in the NAc at 12 weeks. These findings suggest that transcription factors regulating key myelination-related genes and supporting mOL integrity (Denic et al., 2011; Li and Richardson, 2016) may become overactive in the NAc after prolonged ethanol consumption. This overactivity likely contributes to enhanced myelination, as evidenced by increased expression of key myelination genes, enhanced connectivity between the NAc and PFC, and thicker myelin in later stages.
Interestingly, Sox10 expression levels were reduced at 12 weeks in ethanol-drinking mice compared with controls. This reduction may have resulted from the increased expression of Myrf, which is known to regulate SOX10 activity (Aprato et al., 2019), although this hypothesis requires further investigation in future studies.
Taken together, our findings imply that the disruptive effect of ethanol on OLs precedes its enhanced myelinating effects in the NAc. Specifically, ethanol initially disrupts OL differentiation and gene regulation in the NAc, potentially leading to compensatory enhanced myelination after prolonged heavy consumption.
Human studies on myelination in AUD patients report conflicting findings. Postmortem analysis showed reduced brain weight due to less white matter volume (Harper and Kril, 1985; de la Monte, 1988; Kril et al., 1997), in contrast to studies showing unchanged myelinated fiber dimensions (Tang et al., 2004; Skuja et al., 2013). Myelin-related gene expression studies also reported different trends, with some indicating downregulation (Lewohl et al., 2000, 2001; Liu et al., 2006; McClintick et al., 2013; Miguel-Hidalgo et al., 2017) and others noting upregulation or no change (Mayfield et al., 2002; Liu et al., 2004; Lewohl et al., 2005). In DTI studies, reduced fractional anisotropy (FA) was recorded in the corpus callosum of AUD patients (Pfefferbaum et al., 2000; Monnig et al., 2014; De Santis et al., 2019), whereas increased FA was noted in adolescent drinkers (De Bellis et al., 2008; Cardenas et al., 2013), along with increased intracortical myelin signals in addiction-related brain regions (Morris et al., 2022).
Interestingly, previous research suggested a potential protective effect of alcohol against multiple sclerosis (MS; Hedström et al., 2014, 2021; Andersen et al., 2019), with alcohol consumers reportedly having a 20% lower risk of developin MS (Hedström et al., 2021) and being less likely to present neurological disabilities (D'hooghe et al., 2012; Diaz-Cruz et al., 2017). However, other studies did not find such associations (Maas and Angulo, 2021; Yuan et al., 2021; Dreyer-Alster et al., 2022), clearly indicating the need for further investigation.
Treatment with the promyelinating drug clemastine increases ethanol consumption
We found that chronic systemic treatment with clemastine increased the expression of mOLs and MBP levels in the NAc and significantly elevated ethanol consumption and preference in mice with a history of ethanol intake. Together with the well-documented promyelinating effects of clemastine (Barak et al., 2019; Pan et al., 2020), our findings suggest that ethanol consumption may be affected by changes in myelination in the NAc. Importantly, as clemastine is used as an antihistaminic drug and has additional pharmacological properties, we cannot rule out the possibility that our observation of increased ethanol consumption may have been mediated via additional, non-myelin-related mechanisms.
Clemastine has previously been shown to increase brain myelination in several brain regions, including the corpus callosum (Barak et al., 2019; Pan et al., 2020; Leyrolle et al., 2022), cortex (Cree et al., 2018; Xie et al., 2021), and striatum (Cree et al., 2018), under different pathological conditions in mice. Moreover, clemastine was shown to decrease the g-ratio in TEM studies (Cree et al., 2018; Barak et al., 2019; Lee et al., 2021), pointing to its effect on myelin thickness. As it is plausible that clemastine increased myelination in additional brain regions, we cannot localize its activity exclusively to the NAc. Nevertheless, together with the converging evidence from this and other studies, it is reasonable to assume that the promyelinating effects of clemastine in the NAc are likely to be involved in its effects on ethanol drinking.
The normal function of most brain regions relies on the normal myelin sheath surrounding axons, which facilitates axon conductivity and information processing (Mount and Monje, 2017). Consequently, abnormal myelination regulation could disrupt the balance of fiber conductivity required for normal brain function, leading to pathological behaviors (Fields, 2008; Lee et al., 2012; Duncan and Radcliff, 2016; Forbes and Gallo, 2017). For example, as increased myelination can support the function of axons and synapses (Hines et al., 2015; Choi et al., 2019; Maas and Angulo, 2021), it is possible that increased myelination in the NAc while chronically drinking ethanol enhances neuronal functions that play a role in the development and maintenance of excessive ethanol consumption. Thus, given how myelination has been shown to positively correlate with learning processes (Fields, 2010; Williamson and Lyons, 2018; Kato and Wake, 2021), the increased myelination we observed could contribute to abnormally persistent learning and memory processes in addiction (Hyman, 2005) and to excessive ethanol consumption.
Taken together, it appears that maintaining normal and optimal levels of myelin is crucial for preventing alcohol-drinking escalation, as either excessive or inadequate myelination can disrupt the delicate balance required for normal neural functioning, as well as healthy brain activity and behavioral outcomes.
In summary, our findings indicate that while initial ethanol consumption may impair myelination, long-term voluntary heavy ethanol consumption leads to pathological enhanced myelination in the NAc, potentially exacerbating ethanol consumption. This enhanced myelination in the mesolimbic system, a phenomenon also observed in humans (Morris et al., 2022), may contribute to AUD symptoms. Therefore, further investigation into ethanol–myelin interactions is essential for understanding the neurobiological mechanisms of AUD and for developing new treatments.
Footnotes
The research was supported by funds from the Israel Science Foundation (ISF) Grant 508/20 and the German Science Foundation (DFG) Grant 857/29-1. S.B. is the Stephen Harper Chair of Translational Neuroscience at the Faculty of Social Sciences, Tel Aviv University.
↵*M.L. and I.F. contributed equally to this study.
The authors declare no competing financial interests.
- Correspondence should be addressed to Segev Barak at barakseg{at}tau.ac.il or Boaz Barak at boazba{at}tauex.tau.ac.il.
