Alcohol‐Induced Early‐Stage Liver Injury Contributes to Increases in Voluntary Alcohol Intake and Cognitive Deficits in Mice

ABSTRACT

Background

Liver–brain communication is increasingly recognized as a potential contributor to vulnerability to alcohol use disorder (AUD). However, the behavioral consequences of early alcohol‐associated liver disease (ALD) remain underexplored. This study aimed to determine whether alcohol‐induced hepatic injury influences subsequent alcohol consumption and cognitive outcomes.

Methods

Female C57BL/6J mice were fed a Lieber–DeCarli (LDC) ethanol diet to induce early‐stage ALD‐like pathology, followed by a voluntary two‐bottle choice (TBC) paradigm measuring alcohol consumption. Behavioral assessments included novel object recognition, T‐maze with spontaneous alternation, and elevated plus maze. Hepatic steatosis was measured by histology and triglyceride quantification; inflammatory gene expression was evaluated by RT‐qPCR, and hippocampal BDNF/TrkB signaling was investigated by western blot.

Results

LDC mice on an ethanol diet developed hepatic steatosis and hepatomegaly compared with controls. These mice also consumed significantly more alcohol during initial TBC testing, with intake positively correlated with the degree of steatosis. Hepatic steatosis was transient, recovering by the end of TBC period with accompanied upregulation of inflammatory markers and evidence suggestive of neutrophilic infiltration in ethanol‐exposed mice. These mice also exhibited impaired spatial working memory without significant differences in long‐term recognition or anxiety‐like behavior in the elevated plus maze. In addition, ethanol exposure was associated with reduced hippocampal expression of BDNF and its receptor TrkB.

Conclusion

These findings suggest that alcohol‐induced early hepatic injury may contribute to increased alcohol consumption and impaired spatial working memory performance paralleling sustained hepatic inflammation and dysfunction in neurotrophic signaling. Further investigation is warranted to elucidate the mechanisms underlying liver–brain communication in the context of AUD and ALD progression.

While the liver–brain axis is well‐characterized in alcohol‐related disorders, the impact of alcohol‐associated liver disease (ALD) on alcohol use disorder (AUD)‐related behaviors remains unclear. This study investigated the effects of early‐stage ALD on alcohol consummatory behavior and cognitive function in C57BL/6 mice. Early ALD induced a transient increase in alcohol consumption that correlated with hepatic steatosis and produced impairments in spatial working memory associated with altered hippocampal BDNF/TrkB signaling.

1. Introduction

According to the National Institute on Alcohol Abuse and Alcoholism (NIAAA), alcohol use disorder (AUD) is a chronic, relapsing disorder defined by problematic patterns of alcohol consumption including heavy and binge drinking (NIAAA ^2025b). As of 2024, nearly 28 million Americans meet the criteria for an AUD (NIAAA ^2025a). Currently, the few pharmacotherapies approved by the US Food and Drug Administration to treat AUD have limited real‐world effectiveness, in part due to low rates of prescriptions by physicians and poor patient adherence (Hodgkin et al. 2024; Reus et al. ²⁰¹⁸; Walker et al. ²⁰¹⁹). In addition, the complex neurobiological basis of AUD further hinders the development of novel medications (Swift and Aston 2015).

Increasing evidence suggests that AUD complexity extends beyond the central nervous system (CNS) and that peripheral pathology may contribute to AUD progression. Among peripheral organs, the liver is uniquely positioned to influence alcohol‐related behaviors, not only as the primary site of ethanol metabolism but also as a major source of metabolic and inflammatory signals capable of altering brain function (Lanquetin et al. 2021). However, existing evidence that liver injury can influence alcohol consumption is limited and somewhat conflicting. The most direct and well‐characterized consequence of liver damage on the CNS is hepatic encephalopathy, a serious condition marked by neurologic and cognitive impairment that occurs in up to 40% of individuals with cirrhosis (Butterworth 2003; Louissaint et al. ²⁰²²). Beyond advanced disease, inflammatory liver injury may also influence drinking behavior. This has been demonstrated in a mouse model of alcohol dependence, where the blockade of IL‐17A, a cytokine commonly upregulated in alcohol‐associated liver disease (ALD) and thought to facilitate alcohol‐associated damage, reduced alcohol‐induced liver and brain inflammation, and suppressed excessive alcohol intake (Xu et al. 2020). Hepatic ethanol metabolism can additionally activate pro‐inflammatory pathways, disrupting AUD‐related glutamatergic signaling in the brain and affecting communications between hepatic and neural pathways (Lee et al. 2025). Conversely, studies of chemically induced liver injury have shown that hepatic damage can impair ethanol metabolism, leading to acetaldehyde accumulation that correlated with reduced voluntary alcohol consumption (Ren et al. 2020). This conflicting evidence highlights a critical gap in understanding how alcohol‐associated liver pathology may contribute to the development or progression of AUD.

Notably, as many as 95% of individuals with AUD exhibit some degree of ALD (Ramkissoon and Shah 2022), underscoring the close pathological interdependence of these two disorders. The earliest and most common manifestation of ALD is steatosis, characterized by the accumulation of fat in hepatocytes and accompanied by organelle dysfunction, apoptosis, and altered lipid signaling (Nguyen et al. 2008). Although steatosis is an initial and reversible stage of ALD, hepatic steatosis is not benign; recent data indicate an annual mortality rate of approximately 6%, driven predominantly by extrahepatic complications of alcohol consumption (Parker et al. 2019). Preclinical models of chronic alcohol consumption demonstrate that early liver damage is associated with increased pro‐inflammatory cytokines in the brain as well as impaired memory and sensorimotor coordination, supporting the importance of the liver–brain axis in alcohol‐induced cognitive deficits (King et al. 2020). Similar findings have been reported in models of metabolic dysfunction–associated steatotic liver disease (MASLD), where diet‐induced steatosis is accompanied by neuroinflammation and cognitive impairment (Kjærgaard et al. 2024). Several small clinical studies similarly show that MASLD is associated with impairments in verbal fluency, executive function, memory, and language (Filipović et al. 2018; Takahashi et al. ²⁰¹⁷; Tuttolomondo et al. ²⁰¹⁸). While MASLD and ALD share pathophysiological hallmarks such as lipid accumulation and inflammation, the direct contribution of ethanol metabolism and toxicity introduces unique mechanisms by which ALD‐associated steatosis may influence the CNS. Despite this, the impact of early‐stage ALD on alcohol‐seeking behavior and alcohol‐related cognition remains largely unexplored.

In this study, we examined how acute‐on‐chronic alcohol pre‐exposure, sufficient to induce liver pathology, influences subsequent voluntary alcohol intake and cognitive function in female C57BL/6J mice. Early‐stage ALD was modeled with the use of the Lieber–DeCarli ethanol diet, an established preclinical model for the study of ALD (Bertola et al. 2013; Shen et al. ²⁰¹⁹; Yin and Lee ²⁰⁰⁸). Subsequent voluntary alcohol consumption was assessed using the two‐bottle choice (TBC) paradigm, an established method in rodents that recapitulates key features of alcohol use disorder (Hou et al. 2024; Hwa et al. ²⁰¹¹). Additional behavioral measures, including cognitive and anxiety‐like tests, were performed to evaluate how ethanol‐induced hepatic injury may alter behavioral phenotypes relevant to AUD. This sequential approach enabled investigation of how early‐stage alcohol‐induced liver injury influences alcohol‐related behaviors, representing the key novel contribution of the study.

2. Materials and Methods

2.1. Animals and Experimental Design

Female C57BL/6J mice were selected for this study because they exhibit higher binge‐like drinking behaviors and consumption rates (Satta et al. 2018), providing a sensitive model for alcohol‐related behaviors. Mice (n = 50; from Jackson Laboratories; Bar Harbor, ME, USA), aged 8 weeks, were individually housed on a 12‐h light/dark cycle with consistently maintained light, temperature (22°C), and humidity (40%–60%). Following vivarium acclimation, mice were introduced to the liquid Lieber–DeCarli diet (LDC; Bio‐Serv, Flemington, NJ, USA). Following one week of acclimation to LDC diet, mice were randomly assigned to continue receiving either the control LDC diet (C) or an isocaloric LDC diet supplemented with ethanol (E). Ethanol concentration was gradually increased to 5% (5E; v/v) over 5 days and maintained for 8 days, following a modified version of the protocol described by Bertola et al. (2013). Freshly prepared liquid diet was provided to the mice each morning. The duration of exposure was reduced due to pronounced weight loss in ethanol‐treated mice approaching 10% of baseline, necessitating early intervention (Figure S1). Over the course of treatments, mice were given 24/7 ad libitum access to their respective diets, and wooden blocks were provided in cages to prevent malocclusion.

Following completion of the LDC paradigm, all mice were transitioned to a TBC paradigm to evaluate voluntary ethanol consumption. Within each dietary group, half of the animals were given access to two bottles containing tap water (water condition), while the other half were provided one containing water and the other 10% (v/v) ethanol (10E condition). This yielded four experimental groups: (1) Control diet + Water (C/H₂O); (2) Control diet + 10E (C/10E); (3) Ethanol diet + Water (E/H₂O); and (4) Ethanol diet + 10E (E/10E).

To minimize withdrawal or taste aversion effects, mice were gradually introduced to TBC over three days. Mice assigned to ethanol exposure received 5% (v/v) ethanol for two consecutive days prior to escalation to 10% ethanol. Conversely, mice undergoing ethanol removal were provided sequential access to 5% ethanol on Day 1, 2.5% ethanol on Day 2, followed by ethanol‐free solutions on Day 3. During TBC procedure, all animals had ad libitum access to standard chow, and bottle positions were alternated daily to control for side preference. Fluid intake (to the nearest 0.2 mL) and food consumption were measured daily over a two‐week period. After completion of TBC, 10E mice were maintained on a 12‐h on/12‐h off ethanol access schedule, and ethanol was removed 12 h prior to behavioral testing to avoid acute intoxication. The study design and timeline are provided in Figure 1.

FIGURE 1.

Experimental design and timeline. Mice underwent a five‐day acclimation to the LDC control diet and were subsequently divided into two experimental groups: LDC control (C) and LDC ethanol (E) diets. This was followed by a TBC paradigm with experimental cohorts assigned as indicated. Behavioral assays including NOR, T‐maze with spontaneous alteration, and EPM were conducted over a 10‐day period. Necropsies were performed at predetermined time points on Days 14, 32, and 40.

Mice were sacrificed at three predetermined time points during the study: 24 h after the completion of LDC, TBC, or behavioral testing. During necropsy, cardiac punctures were performed to collect whole blood, which was centrifuged for 10 min at 10,000 × g in 4°C to collect serum. Liver was collected for immunohistochemistry and mRNA analysis, while hippocampus was collected and stored in −80°C for protein analysis. All animals were handled in accordance with the University of Southern California's Department of Animal Resources Institutional Animal Care and Use Committee (IACUC) guidelines.

2.2. Immunohistochemistry

Formalin‐fixed, paraffin‐embedded liver sections were stained with hematoxylin and eosin (H&E; Sigma‐Aldrich, St. Louis, MO, USA) for morphological analysis. Liver tissues embedded in O.C.T. compound (Fisher HealthCare, Houston, TX, USA; Cat. #4585) were sectioned at 10 μm thickness and stained for lipid visualization using an Oil Red O Staining Kit (Abcam, Cambridge, United Kingdom; Cat. #ab150678). Images were acquired using a Revolve microscope (Discover Echo Inc., San Diego, CA, USA). Image analysis was performed using the ImageJ software (National Institutes of Health, Bethesda, MD, USA) with the Coloc2 plugin in Fiji. The total lipid area (%) was quantified using whole‐image analysis with the Color Threshold function in ImageJ.

2.3. Liver Triglyceride Assay

Hepatic triglyceride (TG) levels were quantified using a commercial Triglyceride Colorimetric Assay Kit (Cayman Chemical, Ann Arbor, MI, USA; Cat. No. 10010303) according to the manufacturer's protocol. Approximately 20 mg of liver tissue was homogenized for TG extraction and analyzed in duplicate. Following enzymatic reaction with assay reagents, absorbance was measured at 540 nm using a microplate reader, and TG concentrations were calculated from a standard curve generated with provided standards.

2.4. Reverse Transcription Quantitative Polymerase Chain Reaction (RT‐qPCR)

Total RNA was extracted from homogenized liver tissue using TRIzol reagent (Thermo Fisher Scientific, Waltham, MA, USA; Cat. #15596018). Following cell lysis, phase separation was carried out using chloroform, and RNA was precipitated with isopropanol. The RNA pellet was washed with 75% ethanol, air‐dried, and resuspended in RNase‐free water. RNA concentration and purity were assessed using a NanoDrop spectrophotometer (Thermo Fisher Scientific, Waltham, MA, USA). Complementary DNA (cDNA) was synthesized from purified RNA using the PrimeScript RT Master Mix (Takara Bio Inc., Shiga, Japan) according to the manufacturer's instructions. Quantitative PCR was then performed on a QuantStudio 12K Flex Real‐Time PCR System (Thermo Fisher Scientific, Waltham, MA, USA) using SYBR Green PCR Master Mix (Thermo Fisher Scientific, Waltham, MA, USA). Primers were designed using Primer‐BLAST (National Center for Biotechnology Information, Bethesda, MD, USA) and are listed in Table 1. Primer efficiencies were validated prior to testing. Gene expression levels were normalized to β‐actin, and relative mRNA expression was calculated using the 2^−ΔΔCt method.

TABLE 1.

List of primers used for reverse transcription quantitative polymerase chain reaction (qPCR) analysis.

Gene	Forward primer	Reverse primer
ActB	GGCTGTATTCCCCTCCATCG	CCAGTTGGTAACAATGCCATG
Ccl2	TTAAAAACCTGGATCGGAACCAA	GCATTAGCTTCAGATTTACGGGT
Cd68	CCCACCTGTCTCTCTCATTTC	GTATTCCACCGCCATGTAGT
Clec4f	CAGATCTGCAGGCGACCAAA	TTCCTCCTGAGAGGTCACCG
Il1b	CTGGTGTGTGACGTTCCCATTA	CCGACAGCACGAGGCTTT
Il6	AACAAAGCCAGAGTCCTTCAG	GGTCTTGGTCCTTAGCCACTC
Il10	CCAGTTTTACCTGGTAGAAGTGATG	TGTCTAGGTCCTGGAGTCCAGCAGACTCAA
Mpo	CGTGTCAAGTGGCTGTGCCTAT	AACCAGCGTACAAAGGCACGGT
Nfkb1	TGGAGGCATGTTCGGTAGTG	CCTGCGTTGGATTTCGTGAC
Tnf	GACCCTCACACTCAGATCATCTTCT	CCTCCACTTGGTGGTTTGCT

2.5. Western Blot Analysis

Total protein extract (10 μg) was separated using sodium dodecyl sulfate‐polyacrylamide gel electrophoresis (SDS‐PAGE) and transferred onto a PVDF membrane (Bio‐Rad Laboratories, Hercules, CA). Following blocking, membranes were incubated overnight at 4°C with primary antibodies against BDNF (1:1000; Invitrogen, Waltham, Massachusetts, USA), TrkB (1:1000; Invitrogen, Waltham, Massachusetts, USA), and β‐Actin (1:1000; Cell Signaling Technology, Danvers, MA, USA). Immunoreactive bands were visualized using a ChemiDoc imaging system (Bio‐Rad Laboratories, Hercules, CA, USA), and band intensity was quantified by relative densitometry using ImageJ (NIH, Bethesda, MD, USA; Fiji distribution). Protein expression levels were normalized to β‐actin.

2.6. Behavioral Assays

Behavioral assays were performed to assess cognitive function and anxiety‐like behavior in mice. Testing occurred over a one‐week period, with novel object recognition (NOR) performed on Days 1–3, the T‐maze spontaneous alternation task on Day 6, and the elevated plus maze (EPM) on Day 7. All behavioral testing was conducted during the light phase of the light–dark cycle beginning at 10:00 (ZT9) each day. All groups were tested at the same circadian time and within the same daily testing window to minimize variability due to time of day. All tests were performed in a dimly lit room following a 1‐h acclimation period. Apparatuses used during testing were cleaned with 70% ethanol between subjects to eliminate olfactory cues. Behavioral sessions were recorded using a mounted camera, and videos were analyzed by an experimenter blinded to group assignments.

2.6.1. Novel Object Recognition

Novel Object Recognition (NOR) (n = 8) was employed to assess long‐term recognition memory across a three‐day period, which included habituation, training, and testing phases. Prior to testing, mice were habituated to a clean, open‐field apparatus and allowed to explore and familiarize themselves with the environment for 10 min. On the following day, the apparatus was set up with two identical objects (LEGO blocks with dimensions of 5 cm ×3 cm × 2 cm), and each mouse was allowed to explore for 10 min. Testing phases began 24 h later. Mice were placed in the apparatus where one familiar object from the training phase was replaced with a novel object: a cylindrical block. Mice were allowed to explore the open field for 5 min. Interaction time with each object was defined as the duration during which the mouse's nose was within 1 cm of the object. Any outliers that failed to reach a minimum time of 20 s of total exploration were excluded from our analysis. A recognition index was then calculated using recorded interaction times for each mouse to assess recognition memory, defined as novel object exploration time divided by total exploration time. A recognition index greater than 0.5 indicated a preference for the novel object and thus intact recognition memory.

2.6.2. T‐Maze With Spontaneous Alternation

T‐Maze with spontaneous alternation procedure (n = 8) was carried out to assess spatial working memory. The maze was constructed of black‐colored Plexiglass and consisted of two short arms (15 × 8 cm) and one long arm (20 × 8 cm), enclosed by 13‐cm high walls. At the start of the test, mice were placed in the longer arm and allowed to freely explore all arms of the maze for 5 min. Mouse entry in each arm of the maze was recorded, with entry defined as both fore and hind paws fully entering the arm. Spontaneous alternation was calculated as the portion of arm choices differing from the previous two choices (alterations) relative to the total number of arm entries.

2.6.3. Elevated Plus Maze

Elevated Plus Maze (EPM) test (n = 8) was conducted to measure anxiety‐like behaviors. The testing apparatus consisted of a four‐armed cross‐shaped maze made of 0.6‐cm‐thick opaque black Plexiglass. Two arms of the maze were enclosed by 20‐cm high walls, while the other two remained open. Each arm measured 25 × 5 cm, and the maze was elevated 50 cm above the ground. A central platform, measuring 5 × 5 cm, connected the arms. Floors of the open and closed arms were white to ensure uniform visual contrast. During testing, each mouse was placed in the center facing an open arm and allowed to explore the maze for 5 min. The number of entries into open arms, closed arms, and center, as well as the time spent in each area, was recorded. Entry into an arm of the maze was defined as the placement of both fore and hind paws fully entering into an arm.

2.7. Statistical Analyses

All statistical analyses were conducted using GraphPad Prism Version 10 (GraphPad Software Inc., La Jolla, CA, USA). For comparisons involving two independent variables (LDC exposure and TBC ethanol access), data were analyzed using two‐way ANOVA. When appropriate, post hoc Šídák multiple comparisons tests were conducted. One‐way ANOVA was conducted for group level analysis. Ethanol intake and preference across weeks were analyzed using two‐way ANOVA with factors of LDC exposure and time. Pearson correlation was used to assess associations between hepatic steatosis and alcohol intake. Two‐tailed Student's t‐tests were used for single comparisons where appropriate. Data are presented as mean ± SEM, and p ≤ 0.05 was considered statistically significant.

3. Results

3.1. Early Lipid Accumulation and Later Inflammatory Response Define Hepatic Effects of Prolonged Ethanol Exposure

To evaluate the impact of ethanol exposure on hepatic lipid accumulation, liver histology was assessed at two time points: immediately following the LDC paradigm and after completion of the TBC paradigm. Mice fed an ethanol‐supplemented LDC diet displayed pronounced hepatic steatosis compared to controls, as evidenced by Hematoxylin & Eosin (H&E) and Oil Red O (ORO) staining (Figure 2A). Semi‐quantitative analysis of lipid‐positive area (% of total tissue) by ORO staining confirmed a significant increase in lipid accumulation in ethanol‐exposed animals (Figure 2B; p = 0.0214). A similar increase was found in TG levels of LDC‐E exposed liver tissues (Figure 2C, p = 0.0049). Additionally, the liver‐to‐body weight ratio was significantly increased in E mice, consistent with alcohol‐induced hepatomegaly (Figure 2D; p = 0.0113).

FIGURE 2.

Ethanol‐induced hepatic steatosis following TBC paradigm. Representative liver sections from the LDC paradigm stained with (A) Hematoxylin & Eosin (H&E) and Oil Red O (ORO) reveal marked steatosis in ethanol treated animals compared to controls. (B) Quantification of percent lipid‐positive area from ORO confirms increased lipid accumulation in the LDC‐E group (*p = 0.0214). (C) Hepatic TG quantification (mg/g) demonstrates greater TG content in LDC‐E group (**p = 0.0049). (D) Liver‐to‐body weight ratios (g/g) were significantly elevated in ethanol‐exposed animals (*p = 0.0113). (E) Representative liver sections from the TBC groups stained with H&E and ORO show minimal steatosis in most groups. (F) Quantification of ORO‐positive area after TBC revealed a significant interaction between prior LDC exposure and TBC ethanol access (two‐way ANOVA: F(1,8) = 40.68, p = 0.0002), as well as a main effect of LDC exposure (F(1,8) = 6.69, p = 0.0323), with no main effect of TBC alone. Post hoc comparisons indicated that E/10E mice retained significantly greater hepatic lipid content (***p = 0.0004). (G) TG content and (H) Liver‐to‐body weight ratios did not differ significantly among groups in the TBC paradigm. Scale bars = 100 μm. Data are shown as mean ± SEM.

Following the TBC paradigm, steatosis was largely resolved across groups, as illustrated in H&E and ORO staining images (Figure 2E). Two‐way ANOVA of ORO‐positive area revealed a significant interaction between prior LDC exposure and TBC ethanol access (F(1,8) = 40.68, p = 0.0002), as well as a main effect of LDC exposure (F(1,8) = 6.69, p = 0.0323), with no main effect of TBC alone. Post hoc comparisons indicated that mice exposed to ethanol during both the LDC and TBC paradigms (E/10E) retained significantly greater hepatic lipid content relative to all other groups (p = 0.0004), whereas lipid levels were comparable among mice lacking continuous ethanol exposure (Figure 2F). Despite these group differences in lipid accumulation, obtained using ORO staining, TG levels and liver‐to‐body weight ratios did not differ significantly across groups following TBC (Figure 2G,H). No steatosis was present in any group following behavioral testing, during which ethanol access was restricted to brief, intermittent sessions (data not shown).

Hepatic expression of cytokine, chemokine, and macrophage‐associated genes was measured by RT‐qPCR at both necropsy time points. Following LDC, transcriptional changes were minimal across groups, with a significant elevation seen only in myeloid marker myeloperoxidase (Mpo; p = 0.0090) (Figure S2). In contrast, completion of the TBC paradigm was associated with a marked induction of inflammatory gene expression, particularly in mice exposed to ethanol across both paradigms.

Two‐way ANOVA revealed a significant interaction between LDC exposure and TBC ethanol access for Tnf expression (F(1,20) = 4.68, p = 0.0427), along with significant main effects of both LDC exposure (F(1,20) = 28.73, p < 0.0001) and TBC ethanol access (F(1,20) = 6.375, p = 0.0201). Post hoc comparisons demonstrated that Tnf expression was significantly elevated in mice exposed to ethanol during both paradigms (E/10E) relative to all other groups (p < 0.0001; Figure 3B). Similarly, Il1b expression was significantly affected by prior LDC exposure (F(1,20) = 6.95, p = 0.0158), with higher expression in ethanol‐exposed groups and the largest values observed in E/10E mice (Figure 3C). While not statistically significant, Il6, Ccl2, and Nfkb1 expression levels were elevated in mice exposed to ethanol during both the LDC and TBC paradigms (E/10E).

FIGURE 3.

Hepatic inflammatory gene expression following LDC and TBC exposure. (A) Heatmaps display relative mRNA expression of cytokines, chemokines, transcription factors, and macrophage‐associated markers in mouse liver measured by RT‐qPCR. Values represent fold changes relative to C/H₂O (set to 1). (B) Tnf expression exhibited a significant LDC × TBC interaction (F(1,20) = 4.68, p = 0.0427), along with significant main effects of LDC exposure (F(1,20) = 28.73, p < 0.0001) and TBC ethanol access (F(1,20) = 6.38, p = 0.0201). Post hoc comparisons revealed significantly higher expression in E/10E mice (****p < 0.0001). (C) Il1b expression showed a significant main effect of LDC exposure (F(1,20) = 6.95, p = 0.0158), with no significant interaction or TBC main effect; *p = 0.0458 when comparing C/10E vs. E/10E via Šídák multiple comparisons tests. (D) Mpo expression demonstrated significant main effects of LDC exposure (F(1,8) = 465.7, p < 0.0001) and TBC ethanol access (F(1,8) = 955.3, p < 0.0001), together with a significant LDC × TBC interaction (F(1,8) = 232.1, p < 0.0001). Post hoc comparisons showed significant differences for E/10E mice (****p < 0.0001). (E) Clec4f expression exhibited a significant main effect of LDC exposure (F(1,20) = 7.44, p = 0.0130), with no significant interaction or TBC main effect; expression was lowest in E/10E mice (post hoc **p = 0.0034). Data are presented as mean ± SEM.

Mpo expression exhibited robust ethanol‐dependent regulation, with significant main effects of LDC exposure (F(1,8) = 465.7, p < 0.0001) and TBC ethanol access (F(1,8) = 955.3, p < 0.0001), together with a robust LDC × TBC interaction (F(1,8) = 232.1, p < 0.0001). Post hoc tests confirmed markedly elevated Mpo expression in E/10E mice relative to all other groups (Figure 3D).

Clec4f, a marker for Kupffer cells, showed a significant main effect of LDC exposure (F(1,20) = 7.44, p = 0.0130), with the lowest expression observed in E/10E mice (p = 0.0034; Figure 3E), consistent with a reduction in resident macrophage marker expression following repeat ethanol exposure. Meanwhile, while Cd68 was modestly reduced, it did not reach statistical significance.

Among other transcripts, Il10 was significantly affected by TBC ethanol access (F(1,8) = 9.94, p = 0.0136), suggesting engagement of compensatory anti‐inflammatory signaling during voluntary ethanol consumption. Full statistical results for all genes are provided in Figure S3.

3.2. Ethanol Pre‐Exposure Increases Voluntary Alcohol Intake

To evaluate the effect of prior ethanol exposure on subsequent alcohol consummatory behavior, mice were subjected to a 15‐day TBC paradigm and daily fluid intake was recorded and analyzed. Due to technical issues, alcohol intake data from Day 12 were excluded from analysis.

Mice pre‐exposed to ethanol during LDC displayed significantly greater ethanol intake across the two‐week period compared to controls (Figure 4A; unpaired t‐test, p = 0.0341). When intake was analyzed by week using two‐way ANOVA with factors of LDC exposure and time, a significant main effect of LDC exposure was revealed (F(1,24) = 5.28, p = 0.0305), indicating that pre‐exposed mice consumed more ethanol overall than control‐diet mice. No significant LDC × time interaction was detected. Post hoc analysis indicated significantly greater cumulative ethanol intake in E/10E mice compared to C/10E mice during Week 1 of TBC (Figure 4B; p = 0.029), with no significant group differences observed during Week 2.

FIGURE 4.

Altered ethanol intake during TBC and its association with hepatic steatosis. (A) E/10E mice consumed significantly more ethanol during the 2‐week TBC paradigm compared to controls (*p = 0.0341). (B) Two‐way ANOVA with factors of LDC exposure and time (week) revealed a significant main effect of LDC exposure (F(1,24) = 5.28, p = 0.0305). Post hoc comparisons indicated significantly higher ethanol intake in E/10E mice during Week 1 (*p = 0.029), with no group differences during Week 2. (C) Ethanol preference across the full TBC procedure did not differ between groups. (D) Two‐way ANOVA of cumulative ethanol preference by week demonstrated a significant main effect of time (F(1,24) = 11.4, p = 0.0025), with increased preference in Week 2 relative to Week 1, but no main effect of LDC exposure or interaction. Data are presented as mean ± SEM.

Ethanol preference was not significantly altered by prior LDC exposure across the voluntary exposure period, with no effect observed when preference was averaged across the TBC procedure (Figure 4C). Two‐way ANOVA of cumulative ethanol preference by week revealed a significant main effect of time (F(1,24) = 11.4, p = 0.0025), indicating increased preference across time regardless of prior LDC exposure. However, no main effect of LDC exposure and no LDC × time interaction were detected (Figure 4D).

To explore whether hepatic steatosis was associated with subsequent drinking behavior at the cohort level, a Pearson correlation analysis was performed using cohort‐averaged daily intake values paired with Oil Red O‐derived % lipid area (representative of the degree of steatosis). Tissue samples collected immediately prior to the start of the two‐bottle choice procedure were paired with average intake from Week 1, while samples collected at the end of TBC were paired with average Week 2 intake. Because steatosis and intake were measured in separate animals, intake values were averaged across cohorts of mice to approximate group‐level consumption patterns. This exploratory analysis revealed a significant positive association between the degree of hepatic steatosis and alcohol intake in Week 1 (Figure S4A; r = 0.8748, p = 0.0025), while Week 2 showed no association (Figure S4B; r = 0.2445, p = 0.6405).

3.3. E/10E Mice Demonstrate Impairments in Spatial Working Memory

To further characterize alcohol‐related behavioral phenotypes, including potential cognitive deficits, a battery of behavioral assays was conducted (Figure 5A). Spatial working memory was assessed using the spontaneous alternation T‐maze, which relies on a rodent's innate tendency to explore novel environments (D'isa et al. ²⁰²¹). Successful alternation between arms reflects intact spatial working memory. One‐way ANOVA revealed that mice exposed to ethanol during both the LDC and TBC paradigms (E/10E) exhibited significantly reduced alternation performance compared to controls (Figure 5B; p = 0.0403), indicative of impaired working memory function. Meanwhile, two‐way ANOVA demonstrated a significant main effect of TBC ethanol access (F(1,28) = 5.51, p = 0.0262), with no main effect of LDC exposure or interaction. Consistent with this pattern, E/10E mice exhibited the lowest alternation performance across groups.

FIGURE 5.

Assessment of cognitive and anxiety‐like behaviors following ethanol exposure. (A) Schematic illustrations of Spontaneous T‐Maze, NOR and EPM apparatuses. (B) Spontaneous alternation performance in the T‐maze. One‐way ANOVA revealed a significant effect of group (*p = 0.0403), with reduced alternation observed in E/10E mice. Two‐way ANOVA demonstrated a significant main effect of TBC ethanol access (F(1,28) = 5.51, p = 0.0262), with no significant main effect of LDC exposure or interaction. (C) Two‐way ANOVA revealed no significant main effects of LDC exposure or TBC ethanol access on recognition index scores from NOR, and no interaction. (D) No significant effects of LDC exposure or TBC ethanol access were observed across groups in percent open arm entries or percent time spent in open arms in the EPM. Data are presented as mean ± SEM.

Based on the natural tendency of mice to explore novel objects, memory can be indexed by their ability to discriminate between a familiar and an unfamiliar object (Bello‐Medina et al. ²⁰¹³). As such, long‐term recognition memory was assessed via the novel object recognition (NOR) task, where memory performance is measured using a recognition index, with values above 0.5 indicating a preference for the novel object. All experimental groups exhibited recognition indices near 0.5. Two‐way ANOVA revealed no significant main effects of LDC exposure or TBC ethanol access, and no LDC × TBC interaction (Figure 5C), indicating that recognition memory was preserved following combined LDC and TBC paradigms. It should be noted that a small number of mice were excluded from analysis due to failure to reach the minimum exploration threshold (C/H₂O: n = 3; C/10E: n = 1; E/H₂O: n = 3; E/10E: n = 3).

Given that anxiety‐like behaviors can confound cognitive outcomes, anxiety was assessed using EPM. No significant effects of LDC exposure or TBC ethanol access were observed across groups on either percent open arm entries or percent time spent in open arms (Figure 5D), indicating that ethanol exposure did not alter anxiety‐like behavior in EPM and is therefore unlikely to account for the observed deficits in spatial working memory, although the contribution of anxiety‐like behaviors cannot be completely excluded due to the use of a single test.

3.4. Western Blot Analysis Reveals Alterations in the Hippocampal BDNF–TrkB Pathway After E/10E Exposure

To attempt a mechanistic explanation underlying the behavioral impairments observed in E/10E mice, western blotting was performed on hippocampal tissue collected at the conclusion of the study (Figure 6A). Specifically, the analysis focused on the effects on brain‐derived neurotrophic factor (BDNF) and its receptor, tyrosine receptor kinase B (TrkB), given their established role in regulating processes including hippocampal synaptic plasticity, learning, and memory that are disrupted by chronic‐alcohol exposure (Gliwińska et al. 2023). BDNF expression exhibited a nonsignificant downward trend in E/10E mice, with no significant main effects of LDC exposure, TBC ethanol access, or their interaction detected by two‐way ANOVA (Figure 6B). TrkB was detected as both full‐length (~150 kDa; TrkB.FL) and truncated (~100 kDa; TrkB.T) isoforms. Two‐way ANOVA revealed a significant main effect of LDC exposure on TrkB.FL expression (F(1,20) = 8.17, p = 0.0097), with reduced TrkB.FL levels observed in mice pre‐exposed to ethanol (Figure 6D). While no significant main effect of TBC ethanol access or LDC × TBC interaction was detected, post hoc comparisons demonstrated significantly lower TrkB.FL expression in E/10E mice compared to C/10E mice (p = 0.0309). Similarly, analysis of the truncated TrkB isoform revealed a significant main effect of LDC exposure (F(1,20) = 8.09, p = 0.0100), with post hoc testing indicating significantly reduced TrkB‐T expression in E/10E mice relative to C/10E mice (p = 0.0064; Figure 6E). Because changes in neurotrophic signaling may occur alongside alterations in glial activation, GFAP expression was also assessed. GFAP exhibited a trend toward increased expression in E/10E mice compared to controls; however, no significant main effects or interactions were detected (Figure 6C).

FIGURE 6.

The effect of ethanol exposure on the BDNF–TrkB signaling axis in the hippocampus. (A) Representative western blot images of BDNF, full‐length TrkB (TrkB.FL), truncated TrkB (TrkB.T), and GFAP, with β‐Actin as the loading control. (B) Two‐way ANOVA of BDNF protein levels normalized to β‐Actin revealed no significant main effects; however, BDNF expression showed a nonsignificant reduction in ethanol‐exposed mice. (C) Similarly, GFAP expression demonstrated no significant main effects or interaction by two‐way ANOVA, although levels appeared increased in ethanol‐exposed groups. (D) Two‐way ANOVA of TrkB.FL demonstrated a significant main effect of LDC exposure (F(1,20) = 8.17, p = 0.0097), with no significant main effect of TBC ethanol access or interaction. Post hoc comparisons indicated reduced TrkB.FL expression in E/10E mice compared to C/10E mice (*p = 0.0309). (E) Two‐way ANOVA of TrkB.T revealed a significant main effect of LDC exposure (F(1,20) = 8.09, p = 0.0100), with post hoc testing demonstrating reduced TrkB.T expression in E/10E mice relative to C/10E mice (**p = 0.0064). Data are presented as mean ± SEM.

4. Discussion

In this study, we investigated whether early hepatic dysfunction, induced by an acute‐on‐chronic ethanol exposure model, influences alcohol‐related behavior and cognitive function. Ethanol exposure via the LDC ethanol diet produced pronounced hepatic steatosis with minimal inflammatory gene induction. Following TBC, inflammatory transcriptional changes were most evident in mice that continued ethanol intake (E/10E), indicating that sustained exposure amplifies hepatic immune activation. Early ethanol pre‐exposure also led to a transient increase in subsequent ethanol self‐administration as observed in Week 1 of TBC. Our initial assessment of the potential link between liver pathology and ethanol intake behavior further demonstrated a positive correlation between hepatic steatosis and voluntary drinking for the first week of TBC in E/10E mice. These mice also exhibited deficits in spatial working memory, while recognition memory and anxiety‐like behaviors in EPM were largely unaffected. At the molecular level, hippocampal BDNF–TrkB signaling was perturbed, suggesting potential mechanisms underlying the observed cognitive impairments. Collectively, these findings indicate that early‐stage alcohol‐induced liver injury can influence both voluntary alcohol consumption behavior and hippocampal‐dependent cognitive function.

The LDC liquid diet containing ethanol is a flagship preclinical model of ALD, extensively validated for its ability to induce hepatic steatosis and initiate hepatic inflammatory responses (Bertola et al. 2013; Shen et al. ²⁰¹⁹; Yin and Lee ²⁰⁰⁸). The extent of steatosis and hepatic inflammation produced by this model can vary considerably depending on several factors, including the duration of ethanol exposure, the ethanol concentration, and whether a secondary insult (e.g., a chemical agent or binge exposure) is applied. Consistent with previous bodies of work, mice receiving the LDC ethanol diet developed pronounced steatosis, as evidenced by higher lipid area percentages on ORO staining, increased TG content, and increased liver‐to‐body weight ratios. Although steatosis was largely resolved following the completion of TBC, ORO staining, which detects both TG and cholesteryl esters, the two major lipid components of hepatic steatosis, demonstrated that E/10E mice retained some level of residual steatosis, potentially due to the slower clearance of cholesterol esters (Jeon and Carr 2020). The observed reduction in steatosis is consistent with reports in LDC‐fed rodents that ethanol withdrawal restores normal lipid metabolism and resolves hepatic fat accumulation within 1–3 weeks of abstinence (Pi et al. 2021) (Kang et al. 2022). Importantly, E/10E mice continued to receive ethanol during TBC, indicating that while partial withdrawal can reduce steatosis, ongoing exposure sustains residual hepatic lipid accumulation.

It is well‐acknowledged that steatosis can lead to lipotoxicity, which in turn contributes to hepatic inflammation (Mooli and Ramakrishnan 2022). Therefore, we next assessed whether these differences in residual lipid burden translated into inflammatory responses. Following ethanol exposure in LDC, pro‐inflammatory gene expression remained largely unchanged, with the notable exception of Mpo, which was significantly elevated in ethanol‐exposed mice. Myeloperoxidase (MPO) is a myeloid cell‐associated enzyme involved in the generation of oxidants contributing to tissue injury during early hepatic inflammation and is commonly used as a marker for neutrophil infiltration (Frangie and Daher 2022; Yin et al. ²⁰²⁵). This selective transcriptional increase may therefore reflect early neutrophil priming in the absence of broad inflammatory activation, particularly given previous reports that 10‐day LDC feeding does not elevate MPO protein levels (Ki et al. 2010). Following TBC, Mpo expression persisted and was further increased in E/10E mice, coinciding with a pro‐inflammatory transcriptional profile including marked upregulation of Tnf, Il1b, Il6, and Ccl2. Il10 was also elevated, potentially reflecting compensatory anti‐inflammatory signaling. Together, these suggest that sequential exposure to LDC and TBC primes hepatic innate immunity, as indicated by increased pro‐inflammatory cytokine expression and transcriptional evidence of potential presence of neutrophils. However, MPO expression alone is insufficient to establish a definitive contribution of neutrophils, highlighting the need for histological validation. Despite this limitation, inflammatory priming of the liver may compromise mechanisms of inflammatory resolution, such that milder secondary insults, like those seen in TBC, elicit sustained and amplified immune responses despite improvements in lipid burden. TNFα, for instance, is a key mediator of early alcohol‐induced liver injury, and its sustained elevation, even after TBC ethanol exposure, indicates ongoing activation of innate immune pathways (Yin et al. 1999). In contrast, markers of resident Kupffer cells (Clec4f and Cd68) were modestly reduced, suggesting that secondary TBC ethanol exposure may selectively impair resident macrophage populations while promoting neutrophil‐driven inflammation. Collectively, these findings highlight a shift in hepatic immune cell dynamics following acute‐on‐chronic ethanol exposure and TBC, in which early neutrophil priming transitions to sustained neutrophilic infiltration, contributing to prolonged inflammatory signaling and residual liver injury.

To investigate the association between liver pathology and alcohol‐related behaviors, we further assessed ethanol intake and preference using a TBC paradigm. Mice pre‐exposed to ethanol during LDC exhibited a significant increase in ethanol intake during the first week of TBC; however, this effect was not sustained during the Week 2 of TBC. Although ethanol preference trended upward during Week 1, this did not reach statistical significance. This pattern of transient increased ethanol intake or preference following pre‐exposure to ethanol aligns with other published findings. For example, in a study from Atzram (2015), male C57BL/6 mice exhibited increased ethanol preference for 1 day after LDC‐induced dependence (Atzram 2015). Another study using female rats chronically exposed to ethanol showed much higher but transient consumption after withdrawal periods (Pitock et al. 2025). Together, these findings support the hypothesis that prior ethanol exposure primes rodents for increased alcohol‐seeking behavior, producing a temporary escalation of ethanol self‐administration.

To gain an initial understanding of the contribution of liver pathology in alcohol consumption behavior, we attempted a correlational analysis between the presence of liver steatosis and alcohol intake. Hepatic lipid burden, quantified as percent lipid area (i.e., fatty liver score), was positively correlated with drinking behavior in E/10E mice during Week 1, when increased ethanol intake was noted. This relationship was absent in Week 2 when parallel reductions in liver steatosis and ethanol intake were observed. These findings suggest a potential cohort‐level temporal association between hepatic pathology and ethanol intake across experimental time points. However, because steatosis and intake were measured in separate animals and intake values were averaged across cohorts, these findings are considered as exploratory and do not establish an individual‐level or causal relationship. Importantly, the present design does not allow dissociation of hepatic signaling from dependence‐related neuroadaptations, as the LDC paradigm involves repeated ethanol exposure sufficient to induce liver pathology while also engaging neurobiological processes associated with ethanol exposure. Thus, the observed increase in intake may reflect combined contributions of peripheral liver‐derived signaling as well as established neurobiological mechanisms of dependence.

Chronic alcohol consumption is known to impair cognitive domains including spatial learning, working memory, and executive function across multiple preclinical models of alcohol exposure (Gao et al. 2024; Guo et al. ²⁰²²; Ma et al. ²⁰²³; Stragier et al. ²⁰¹⁵). In the present study, cognitive outcomes were assessed using T‐maze spontaneous alternation for spatial working memory and novel object recognition (NOR) for long‐term recognition memory. E/10E mice exhibited significant deficits in spontaneous alternation, indicating impaired hippocampal‐dependent spatial working memory (Lalonde 2002). In contrast, NOR performance was preserved, suggestive of intact long‐term recognition memory. These findings contrast with prior work reporting that free‐choice ethanol consumption in C57BL/6J mice for approximately three weeks did not impair spatial learning but produced modest deficits in object recognition (Stragier et al. 2015). These differences likely reflect variations in alcohol exposure paradigm, exposure duration and dose, and the presence or absence of peripheral organ pathology, all of which are documented to influence cognitive outcomes (Charlton and Perry 2022). Anxiety‐like behavior evaluated using EPM was also unchanged across groups, indicating that differences in affective state were unlikely to account for cognitive performance. Together, these findings need to be considered in the context of both central and peripheral influences. Notably, liver pathology has been shown to influence brain function independent of ethanol exposure, as models of metabolic liver disease demonstrate associations between hepatic steatosis, neuroinflammation, and cognitive impairment (Kjærgaard et al. 2024; Takahashi et al. ²⁰¹⁷). While these findings support a potential contribution of liver–brain signaling, cognitive outcomes in the present study may also reflect direct neuroadaptations to sequential LDC and TBC ethanol exposure. Additionally, because brain tissue was collected after completion of behavioral testing, the neurobiological changes observed here may reflect residual or compensatory adaptations rather than real‐time mechanisms directly mediating behavioral outcomes.

To investigate molecular correlates of the observed spatial working memory impairments, we assessed hippocampal BDNF/TrkB signaling, a pathway critical for neuronal survival and synaptic plasticity often impaired with chronic alcohol consumption (Numakawa and Kajihara 2023; Xu et al. ²⁰¹⁵). Notably, TrkB is expressed as full‐length and truncated isoforms, with the TrkB T1 isoform constituting the predominant truncated species and serving as the major astrocytic receptor mediating BDNF responses (Holt et al. 2019; Tessarollo and Yanpallewar ²⁰²²). In E/10E mice, a trend toward decreased BDNF expression was observed along with a significant reduction in both full‐length and truncated TrkB isoforms, indicating broad alterations in BDNF‐driven signaling in both neuronal and astrocytic populations following sequential ethanol exposure. In addition to reducing neurotrophic signaling capacity, loss of the truncated TrkB isoform may also alter BDNF availability and signaling regulation within astrocytic networks. These findings are consistent with clinical evidence demonstrating reduced plasma TrkB and mature BDNF levels in patients with AUD (Shafiee et al. 2023). Although GFAP expression demonstrated only a nonsignificant upward trend, prior studies have associated increased GFAP expression with astrocytic activation and neuroinflammatory responses (Nadler et al. 2025). Therefore, the combined changes in astrocytic TrkB expression and GFAP may suggest altered astrocytic function following ethanol exposure. Together, these suggest altered neurotrophic signaling and astrocytic dysfunction may contribute to impaired hippocampal plasticity and memory deficits following sequential ethanol exposure.

Several limitations of the present study should be acknowledged. First, the relatively short period of alcohol exposure likely produced less robust liver pathology than longer ALD models, which may have limited the magnitude of liver‐to‐brain signaling effects on alcohol‐related cognitive outcomes and immune regulation. Sampling at earlier time points, particularly within the first week of TBC, may have captured transient hepatic and neurobiological changes with greater relevance to behavior. Second, the exclusive use of female C57BL/6J mice, a strain with well‐documented preference for voluntary ethanol consumption, may have influenced intake levels in the TBC paradigm (Lê et al. 1994). Finally, brain pathology was assessed only after the completion of behavioral testing, which may not have captured earlier or more transient neurobiological and behavioral alterations. Future studies incorporating multiple early sampling points would better define the trajectory of liver and brain pathology progression and regression, refining the relationship between hepatic pathology, neurobiological changes, and behavioral outcomes.

Overall, findings from this study suggest that early‐stage alcohol‐associated liver injury can influence alcohol‐seeking behavior and impact cognitive function, underscoring the importance of accounting for liver pathology when considering the neurobiological mechanisms underlying AUD. While the present study was designed to examine how pre‐existing hepatic pathology influences subsequent alcohol consumption, future studies employing reverse or bidirectional paradigms (e.g., voluntary alcohol exposure followed by LDC) will be important to determine whether prior drinking behavior alters hepatic susceptibility to steatosis and to further resolve causal relationships within the liver–brain axis. To our knowledge, this is the first effort to elucidate a relationship between early‐stage alcohol‐induced hepatic steatosis and alcohol consummatory behavior. Although traditional addiction treatments focus primarily on brain‐based targets, the observed associations between liver pathology and brain function highlight the importance of peripheral organ health in AUD. Monitoring liver health may therefore provide a valuable insight into AUD vulnerability, and future studies should investigate whether interventions targeting hepatic health can modulate alcohol consumption patterns and cognitive outcomes, providing insights into more integrated strategies for the prevention or treatment of AUD.

Author Contributions

Conceptualization: Samantha G. Skinner and Liana Asatryan. Methodology: Samantha G. Skinner, Nikhila Kalapatapu, and Liana Asatryan. Investigation: Samantha G. Skinner, Eliana G. Aleman, Julio Rivera, Nikhila Kalapatapu, and Dhruv Kantilal. Formal analysis: Samantha G. Skinner and Eliana G. Aleman. Data curation: Samantha G. Skinner. Writing – original draft: Samantha G. Skinner. Writing – review and editing: Samantha G. Skinner, Eliana G. Aleman, Julio Rivera, Nikhila Kalapatapu, Dhruv Kantilal, Daryl L. Davies, and Liana Asatryan. Visualization: Samantha G. Skinner, Eliana G. Aleman, and Julio Rivera. Supervision: Daryl L. Davies and Liana Asatryan. Funding acquisition: Daryl L. Davies. All authors have read and agreed to the published version of the manuscript.

Funding

This work was supported by the Timothy M. Chan Professorship in Complementary Therapeutics, USC Mann School of Pharmacy and Pharmaceutical Sciences, USC Institute of Addiction Science, and Cai Foundation.

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: Body weights of mice over the course of LDC administration. Data are presented as mean ± SEM.

Figure S2: Hepatic genetic expression after LDC. (A) Heatmap showing relative mRNA expression of cytokines, chemokines, transcription factors, and macrophage‐associated markers in liver tissue collected immediately following the Lieber–DeCarli (LDC) paradigm. Values represent fold change relative to control diet (C, set to 1). Quantification of individual transcripts measured by RT‐qPCR including (B) Tnf, (C), Il1b, (D) Il6, (E) Ccl2, (F) Nfkb1, (G) Il10, (H) Mpo (**p = 0.0090), (I) Clec4f, and (J) Cd68. Data are presented as mean ± SEM.

Figure S3: Hepatic gene expression after TBC. Data were analyzed by two‐way ANOVA with factors of LDC exposure (control vs. ethanol) and TBC ethanol access (H₂O vs. 10E). Significant effects were observed for (A) Tnf, (B) Il1b, (G) Mpo, (H) Clec4f, and (F) Il10, as described in the Results section. (C) Il6, (D) Nfkb1, (E) Ccl2, and (I) Cd68 did not show significant main effects or interactions. Post hoc multiple comparisons were performed where appropriate. Data are presented as mean ± SEM. Statistical significance is indicated by asterisks (*p ≤ 0.05, **p ≤ 0.01, ***p ≤ 0.001, ****p ≤ 0.0001).

Figure S4: Correlation between degree of hepatic steatosis (%) and EtOH intake during Week 1 and Week 2 in C/10E and E/10E groups. (A) Linear regression analysis demonstrated a significant positive correlation between steatosis and EtOH intake during Week 1 (r = 0.8748, p = 0.0225). (B) No significant correlation was observed during Week 2 (r = 0.2445, p = 0.6405). Red symbols represent C/10E animals and blue symbols represent E/10E animals.

Data Availability Statement

The data that support the findings of this study are available from the corresponding author upon reasonable request.