Short‐Term Alcohol Consumption Is Sufficient to Disrupt Microbial Communities and the Intestinal and Circulating Metabolites in a Non‐Human Primate Model

ABSTRACT

The gut microbiome is a complex and dynamic community of microorganisms that play a crucial role in maintaining host health and facilitating inter‐organ communication. The microbiome is altered by the exposome, a phenomenon known as dysbiosis. Alcohol misuse can disrupt the gut microbial community in a dose‐ and duration‐dependent manner. Previous studies showed that chronic alcohol misuse can induce intestinal hyperpermeability and significantly shift microbial communities. However, these data were obtained primarily from cross‐sectional clinical and rodent‐based studies following long‐term chronic alcohol consumption. Clinical studies often utilize samples from patients with alcoholic liver disease and are confounded by variability in factors that can modulate the microbiome and reliance on self‐reported data. The intestinal cellular and microbial composition of rodent models differs significantly from that of humans, limiting the clinical translation of those findings. Therefore, the impact of alcohol misuse on the gut microbial communities and their metabolic implications remains poorly understood. Since rhesus macaques share significant genetic/physiological traits, as well as gut microbial communities, with humans, we assessed the impact of 6 months of voluntary daily alcohol consumption on the gut microbiome and its implications for the circulating and intestinal metabolomes. We report that short‐term drinking induces dynamic changes in the abundance of short‐chain fatty acid producers and increases microbes associated with the negative regulation of inflammation. Finally, altered fatty‐ and amino‐acid metabolite signatures were observed. Taken together, these data enhance our understanding of the longitudinal implications of alcohol use.

This study assessed the impact of short‐term (6 months) voluntary ethanol self‐administration on the composition of the gut microbiome as well as fecal and circulating metabolomes using a non‐human primate model. We demonstrate that short‐term ethanol consumption increased the abundance of microbes associated with inflammation resolution, while disrupting the abundance of short‐chain fatty acid producers. Finally, levels of fatty and amino acid metabolites important for the regulation of inflammatory responses and epigenetic modifications were altered in both fecal and circulating metabolomes.

1. Introduction

Alcohol is one of the most widely used substances across the United States, with ~80% of individuals over the age of 12 reporting alcohol consumption [1]. Alarmingly, 1 in 10 of these individuals will develop alcohol use disorder (AUD) [2]. The effects of alcohol can be observed in multiple organs and organ systems, resulting in increased incidences of liver disease, cancer, cardiovascular disease, and both viral and bacterial infections [3]. Therefore, it is crucial to better understand the mechanisms by which alcohol consumption leads to adverse health consequences. Recent work has suggested that alcohol‐induced changes to the gut microbiome could contribute to the pathophysiology of alcohol‐related organ damage [4].

The microbiome is a complex network of bacteria, viruses, fungi, archaea, and other eukaryotes [5] that is easily modified by aging, medications, behavior, diet, and other environmental factors [6]. Together, these microorganisms comprise “the hidden organ” and exist in different locations across the human body, including but not limited to the respiratory tract, oral cavity, skin, reproductive tract, and gastrointestinal tract [6]. The relationship between the gut microbiome and host is multidirectional, as it has been shown to facilitate communication between immunological, metabolic, neural, and humoral pathways [7]. Disruption in this crosstalk may induce negative health consequences [7].

Alcohol exerts a dose‐ and duration‐dependent effect on the microbiome [8]. It has been reported that individuals with chronic drinking patterns of < 10 drinks/day have a higher abundance of Erysipelotrichaceae and Lachnospiraceae when compared to those who chronically consumed ≥ 10 drinks/day [9]. Studies in rodents allowed to drink ad libitum (10% v/v EtOH/water) also demonstrated dose‐dependent alterations in microbial abundance, with a significant increase in the genus Streptococcus after 12 months of open access [10]. However, microbial communities in rodents significantly differ from those of humans [11]. Additionally, mice require ethanol dosages 2–5 times greater than humans to reach comparable peak blood concentrations, which could induce perturbations in the microbiome not observed in humans [12]. Non‐human primate models have been utilized to overcome some of these limitations, as they have similar gut microbiomes and a natural propensity to consume alcohol [13, 14, 15, 16, 17]. A non‐human primate study demonstrated that a longer duration of alcohol consumption (12.7 years) was associated with an increased abundance of the genus Lactobacillus, Streptococcus, and a decreased abundance of the genus Parabacteroides, Oribacterium, Collinsella, and Desulfovibrio when compared to shorter durations (2.7 years) [16].

Chronic ethanol consumption can also lead to disruptions in the gut barrier [18], allowing for the translocation of microbial products and other molecules into circulation [19]. Indeed, work from our lab and others has shown that there are increased levels of circulating endotoxins, microbial metabolites, and lipids in plasma following 12 months of chronic alcohol consumption (CAC) [17, 18, 20, 21]. These translocated products have the potential to impact the function of myeloid cells, including monocytes [22], by promoting a hyperinflammatory phenotype [23, 24]. Interestingly, the impact of alcohol consumption on monocyte function is dose‐ and duration‐dependent. While chronic heavy drinking leads to heightened inflammatory responses, acute or short‐term alcohol use results in dampened inflammatory responses [25, 26, 27]. These data highlight the need to study longitudinal changes in the microbiome in tandem with metabolic and immune pathways after shorter durations of alcohol consumption. However, these studies are lacking.

Here, we utilize a rhesus macaque model of voluntary self‐administration to investigate the impact of short‐term drinking on the microbiome and its subsequent effects on the fecal and plasma metabolomes after 6 months of voluntary ethanol consumption. Data from this study demonstrate that short‐term drinking is sufficient to alter gut microbial communities, specifically increasing the abundance of microbes that promote the resolution of inflammation and skewing the abundance of short‐chain fatty acid producers. Finally, metabolomics analysis revealed unique fatty acid and amino acid metabolite signatures in plasma and feces after 6 months of alcohol consumption. Taken together, these data help to bridge the gap in understanding the multi‐organ consequences of prolonged alcohol use.

2. Materials and Methods

2.1. Ethics Declaration

All animal procedures were approved by the Oregon Health and Science University (OHSU) West Campus Institutional Animal Care and Use Committee (IACUC) and conducted in accordance with the ONPRC animal care program. This program is fully accredited by AAALAC International and adheres to the regulations and guidelines stipulated by the United States Department of Agriculture (e.g., the Animal Welfare Act and Animal Welfare Regulations), the Guide for the Care and Use of Laboratory Animals, 8th edition (Institute for Laboratory Animal Research), and the Public Health Service Policy on Humane Care and Use of Laboratory Animals. At the study's conclusion, macaques were sedated with ketamine for transport to the necropsy suite. Euthanasia was performed via intravenous administration of sodium pentobarbital after confirmation of deep sedation and absence of palpebral, corneal, and deep pain withdrawal reflexes, prior to tissue harvest [28]. All methods are reported in compliance with ARRIVE guidelines.

2.2. Cohort Description and Sample Collection

The Monkey Alcohol Tissue Research Resources (MATRR; www.matrr.com) repository was utilized to access samples before and after 6 months of ethanol consumption. Samples from 40 macaques across cohorts 14, 16, 18, and 19 were utilized in this study. Plasma and fecal samples were obtained from a rhesus macaque model of voluntary ethanol self‐administration. Macaques were trained to use operant drinking panels via a 4‐month protocol of scheduled‐induced polydipsia that started with water induction for the first month and then increased ethanol concentrations (0.5 g/kg, 1 g/kg, and 1.5 g/kg) every 30 days [14]. Following induction, macaques had continuous access to both 4% ethanol w/v solution and fresh water for 22 h/day for 6 months. Each macaque was housed in a quadrant cage with constant temperature, humidity, and an 11/13 h light/dark cycle. Additionally, macaques were socially paired with another macaque daily for 1–2 h. For sample acquisition, they were trained to present their legs for femoral blood collection, while freshly caught fecal samples were collected from their pans. Collected plasma and fecal samples were stored at −80°C until use to minimize batch effects.

2.3. 16s Amplicon Sequencing and Bioinformatics Analysis

Amplification of the 16 s rRNA gene hypervariable V4 region was done with the 515F/806R PCR primers and analyzed as previously described [29]. A 12‐bp barcode was conjugated to the forward primers [30]. GoTaq master mix (Promega Corporation, Madison, WI, USA) was used to run each reaction in duplicate. The reaction was initiated by incubating at 94°C for 3 min. This was followed by 37 cycles of 94°C for 45 s, 50°C for 1 min, and 72°C for 1 min. There was a final incubation at 72°C for 10 min. The Quant‐iT PicoGreen dsDNA Assay Kits and dsDNA Reagents (Thermo Fisher, Waltham, MA, USA) were used to quantify the amount of DNA in each well. Amplified regions were multiplexed and ~15%–20% PhiX was added to the library and sequenced on an Illumina MiSeq. The QIIME2 analysis pipeline was used to process sequences via demultiplexing and filtering using DADA2 [31]. Sequence variants were aligned via MAFFT and a phylogenetic tree was constructed with FastTree2 [32, 33]. The q2‐feature‐classifier (release 138) against the SILVA database was used to assign taxonomy to sequence variants [34]. Data were rarified to 10 000 sequences per sample. Alpha diversity metrics were generated with QIIME2. Weighted and unweighted UniFrac distances were used to estimate beta diversity [35]. Finally, the Linear Effect Size (LEfSe) algorithm with a linear discriminant analysis score cutoff of 2 identified significant changes in bacterial abundance between groups [36].

2.4. Hydrophobic Interaction Liquid Chromatography (HILIC) LC–MS Analysis at 6 Months

2.4.1. Plasma

Methods were adapted from [37]. Polar metabolites were separated by HILIC using a ZIC‐pHILIC column (2.1 × 150 mm, 5 μm; Millipore Sigma, Burlington, MA, USA) on a Thermo Vanquish UPLC liquid chromatography (LC) system. The binary solvent system consisted of 10 mM ammonium acetate in water, pH 9.8 (mobile phase A) and acetonitrile (mobile phase B). The flow rate was 0.15 mL/min, the column temperature was 25°C, and the injection volume was 10 μL. For gradient separation, mobile phase B was ramped linearly from 90% to 30% from 0 to 19 min, held at 30% from 19 to 24 min, and mobile phase B was held at 90% from 24 to 33 min.

2.4.2. Feces

Methods were adapted from [38]. Feces were ground to powder at liquid nitrogen temperatures in the Spex 6875 Large Freezer/Mill. Polar metabolites were separated from 5 to 20 mg of sample by HILIC using a ZIC‐pHILIC column (2.1 × 150 mm, 5 μm; Millipore Sigma, Burlington, MA, USA) on a Thermo Vanquish UPLC liquid chromatography (LC) system. Mobile phase A was 100% water and B was 95:5 acetonitrile: water with both phases modified with 10 mM ammonium formate and 0.125% formic acid. The flow rate was 0.25 mL/min, column temperature was 25°C, and the injection volume was 5 μL. The LC gradient was 100% mobile phase B from 0 to 2 min, 77% by 3.2 min, 30% by 16.4 min, ramped back up to 100% by 24 min and held at 100% until 28 min for re‐equilibration.

Spectra were acquired in both positive and negative ionization modes on a Thermo Exploris 240 Orbitrap Mass Spectrometer. Each polarity was acquired at a resolving power of 120 000 full‐width at half maximum (FWHM) with the automatic gain control set to 100% with a maximum inject time of 50 ms. The scan range was set to 60–900 Da. Heated electrospray ionization (H‐ESI) source conditions were as follows: sheath gas flow 50, auxiliary gas flow 13, sweep gas flow 3, capillary temperature 263°C, S‐lens RF level 50, auxiliary gas heater temperature 350°C, and needle voltage 3.5 kV and −3.5 kV for positive and negative ionization modes, respectively.

For compound identification, a pooled sample containing equal volumes of all individual samples was acquired using data‐dependent MS2 (ddMS2) in both negative and positive ionization modes. Precursor MS1 scans were acquired using the settings described above. Product ion MS data were acquired with a resolving power of 15 000 FWHM, the AGC target was set to 10 [6] ions, injection time set to auto. For ddMS2, the isolation window was set to 1 Da and fragmented with a stepped collision energies of 20, 30, and 60 normalized collision energy (NCE) units. The minimum AGC target was 8000 with a dynamic exclusion of 3 s. Samples were processed with Compound Discoverer (CD) (version 3.3.3.200, Thermo Scientific, Waltham, MA, USA) using the Untargeted Metabolomics Workflow. Identifications were manually curated to remove features with poor peak quality and features that were highly abundant in blank control samples. Data were analyzed using Metaboanalyst 6.0 [39] for statistical and enrichment analyses. Fecal and plasma metabolites with a log (2) fold change threshold of ±0.585 and a raw p‐value threshold of ≤ 0.05 were considered differentially abundant. A p‐value cutoff of < 0.1 and a minimum of 2 genes were required for pathway enrichments.

2.5. Short Chain Fatty Acid GC/MS Analysis

Frozen fecal powder was suspended in 500 μL water with 5 μg d7‐butyric acid. 100 μL 37% HCl and 100 μL methyl tert‐butyl ether (MTBE) were added in a stepwise manner. Fatty acids were extracted by homogenizing the slurry for 3 min followed by a 30‐min incubation at room temperature. The organic MTBE layer was separated from the aqueous layer by centrifugation and transferred to tubes containing Na₂SO₄ for 10 min at room temperature to absorb excess water. After centrifugation at 14 000×g for 1 min, 20 μL of supernatant was transferred to a low‐recovery glass insert containing 5 mL MTBSTFA (Restek, Bellefonte, PA, USA). Samples were derivatized by incubating for 30 min at 80°C. An Agilent 5977B GC/MS containing an Agilent HP‐5MS column was used to run derivatized samples. The GC/MS had the following settings: inlet set to 250°C with a pressure of 4.9 psi and septum purge flow of 3 mL/min; split inlet mode ratio set to 10:1 and split flow of 7.67 mL/min; initial oven temperature set to 50°C for 1 min, followed by a linear ramp to 300°C at a rate of 11.5°C per minute; full‐scan MS detection range set to 50–550 m/z with a source and quad temperature of 230°C and 150°C, respectively. MassHunter Qualitative Analysis (Agilent, Santa Clara, CA, USA) was used to extract and quantify peak areas for each sample. The Cayman Short‐chain Fatty Acid Mixture 2 (Cayman, Ann Arbor, MI, USA) was used for targeted analysis of acetate, propionate, isobutyrate, n‐butyrate, 2‐methylbutyrate, isovalerate, and n‐valerate.

2.6. Statistical Analysis for Short‐, Medium‐, and Long‐Chain Fatty Acid Analysis

GraphPad Prism software (GraphPad Software Inc., La Jolla, CA, USA) was used to assess statistical significance. Outliers were determined via the ROUT method, and data were tested for normality via the Shapiro–Wilk test. If the Gaussian requirements were met, an unpaired t‐test with Welch's correction was applied. If the Gaussian requirements were not met, comparisons were carried out using a Mann–Whitney test.

3. Results

3.1. Short‐Term Drinking Is Sufficient to Alter Gut Microbial Communities

We performed 16S rRNA sequencing, LC–MS, and GC–MS to understand how alcohol consumption impacts gut microbial communities and their subsequent effect on the metabolome. Principal coordinate analysis (PCoA) of the 16S data revealed a significant difference in the β‐diversity, or microbial diversity between samples, after 6 months of alcohol consumption. When compared to controls, the structure (weighted) (Figure 1A) and community membership (unweighted) (Figure 1B) were significantly different, indicating short‐term drinking drives compositionally distinct gut microbial communities. Moreover, there were significant differences in the composition and abundance of microbes when assessing sex as a biological variable (SABV) (Figure 1A,B). However, we also observed no significant differences in the number of amplicon sequence variants (ASV) (Figure 1C).

FIGURE 1.

Short‐term alcohol use induces opposing changes in SFCA‐producing genera, while increasing genera associated with inflammation regulation. (A) Weighted and (B) unweighted principal coordinate analysis (PCoA) 16 s analysis on fecal samples obtained before and after 6 months of alcohol use. (C) Bar plot depicting amplicon sequence variants (ASV) observed at baseline and after 6 months. Error bars represent SEM. Bubble plot of bacteria significantly more abundant in samples obtained at (D) baseline and (E) after short‐term alcohol use. The size of the bubble represents the relative abundance of the microbe, and the color represents the linear discriminant analysis (LDA). (F) Microbes MAASLin analysis determined to be positively or negatively associated with alcohol intake (g of EtOH/kg of NHP). The size of the bubble represents the adjusted p‐value (q‐value), and the color represents the coefficient. (G) Bar plot depicting metabolic pathways predicted to be associated with bacteria found to be more abundant at baseline or after short‐term alcohol use. For 16 s experiments: N = 22 controls, n = 52 after 6 months of alcohol.

Linear Discriminant Effect Size (LEfSe) was used to assess significant changes in bacterial abundance. To control for temporal changes in the gut microbiome, we compared fecal samples from housing controls collected at baseline with those collected from the same controls after 6 months of maltose self‐administration. We observed an increase in Ruminococcus and Blauntia, concurrent with a decrease in Escherichia shigella and Christensenellaceae (Figure S1). Therefore, these bacteria were removed from subsequent analyses. Data showed that 6 months of alcohol consumption significantly reduces the relative abundance of SCFA producers (e.g., genus Eubacterium, Subdoligranulum, Lactobacillus and family Butyricioccaceae) and known commensals (e.g., genus Tremonema, Christensenellaceae) (Figure 1D). Conversely, we see an increase in the relative abundance of bacteria associated with immunoregulatory functions (e.g., genus Provotella, Lactobacillus, Faecalibacterium, Dorea, Sutterella) [40, 41] and a group of butyrate producers distinct from those noted in baseline samples (e.g., genus Lachnospiraceae, Parabacteroides, Ruminococcus) (Figure 1E).

Using MAASLin2 analysis, we assessed whether the abundance of any bacteria correlated with the average daily amount of ethanol consumed (g of EtOH/kg of NHP). Our data indicated that 7 bacterial genera or families demonstrated a significant dose‐dependent change in abundance (Figure 1F, Figure S1B). Genera Lactobacillus, a natural probiotic that can fortify the intestinal barrier via mucus secretion [42, 43], Terrisporobacter, shown to coincide with elevated oxidative stress [44], and Colidextribacter, associated with gut barrier dysfunction [45], were positively correlated with intake level (Figure 1F, Figure S1). On the other hand, Clostridium, a genus that encompasses probiotics and pathobiont species [45], Catenibacterium, reported to be elevated in individuals with chronic infection [45], Sarcina, associated with an increased risk of bloat/emphysematous gastritis [46], and family Oscillospiraceae, known to possess butyrate‐producing bacteria and often negatively associated with inflammatory diseases [47], were found to be negatively correlated with the amount of alcohol consumed (Figure 1F, Figure S1B). These data show that 6 months of alcohol consumption leads to a decline in the abundance of commensals, a rise in the abundance of bacteria previously linked to inflammation regulation, and a restructuring of the SCFA producing taxa.

We then leveraged Phylogenetic investigation of communities by reconstruction of unobserved states (PICRUSt) [48] to predict putative changes in bacterial metabolic pathways. A group of metabolic pathways involved in fatty acid biosynthesis (e.g., stearate, palmitoleate, mycolate, oleate, palmitate) was predicted to be enhanced with alcohol use (Figure 1G). These data indicate that the alcohol‐related alterations to the intestinal microbiome after short‐term alcohol use could disrupt the production of fatty acids.

3.2. Short‐Term Alcohol Use Alters Amino Acid and Fatty Acid Metabolite Levels

To understand how alcohol‐related changes in microbial abundance influence the fecal metabolome, we carried out untargeted metabolomics with LC–MS on the feces at baseline and after 6 months of alcohol use (Table S1). Using sparse Partial Least Squares Discriminant analysis (sPLSDA) we observed a clear separation in the fecal microbiome at baseline and after 6 months of alcohol use (Figure 2A). Products of amino acid metabolism such as mercaptopyruvic acid, sulfuric acid, and urocanic acid were increased following 6 months of drinking, while tyrosine, N‐methylproline, and Gly‐Pro were decreased (Figure 2B). As such, when these metabolites were queried against databases with well‐annotated metabolic pathways, we observed a predicted decrease in pathways associated with amino acid metabolism and the oxidation of fatty acids (Figure 2C).

FIGURE 2.

Six months of daily heavy drinking downregulates amino acid and fatty acid metabolic processes. (A) sPLSDA of LC–MS analysis of macaque fecal samples from ethanol naive controls and after 6 months. (B) Volcano plot of differentially abundant metabolites in fecal samples after 6 months when compared to controls. (C) Bar plot depicting selected enrichments of differentially abundant metabolites in fecal samples after 6 months (terms with p‐value < 0.1 and > 2 genes). For fecal LC–MS: N = 4 controls, n = 4 after 6 months of alcohol. Bar plot of results from targeted metabolomics (GC–MS) determining (D) SCFA or (E) medium‐ and long‐chain fatty acids identified in stool samples before and after 6 months of daily drinking. Error bars represent SEM. For fecal GC–MS: N = 31 controls, n = 23 after 6 months of alcohol.

To follow up on the changes in SCFA producers noted in the 16 s analysis, we assessed SCFA levels in stool samples collected at baseline and after 6 months of alcohol use via GC/MS and observed only a significant reduction in propionic acid (Figure 2D, Figure S2). Interestingly, we also observed a significant decrease in the abundance of medium and long chain fatty acids, dodecanoic acid, palmitic acid, oleic acid, stearic acid, 10‐nonadecenoic acid, 11‐eicosenoic acid, nonadecanoic acid, behenic acid, and lignoceric acid with a concurrent increase in heptanoic acid following 6 months of daily drinking (Figure 2E).

We then surveyed the circulating metabolome at baseline and after 6 months of daily drinking using LC–MS to understand how alcohol‐induced changes in the fecal metabolome extend into circulation (Table S1). A sPLSDA revealed a clear separation in the plasma metabolome before and after short‐term alcohol use (Figure 3A). Specifically, we observed an increase in the levels of acetyltaurine, β‐hydroxybutyric acid, stachydrine, urate, and kynurenate, which are known to modulate neurological and inflammatory pathways [38, 49, 50, 51, 52] after 6 months. Guanine, guanosine, and glutamic acid were also increased, which are known to modify the epigenetic landscape [53, 54, 55] (Figure 3B). These changes suggested increases in fatty acid oxidation (β‐oxidation of very long chain fatty acids), metabolic processes (Glucose‐Alanine Cycle, Purine Metabolism, Beta‐Alanine Metabolism, and Tryptophan Metabolism), and the electron transport between the cytosol and mitochondria (Malate–Aspartate Shuttle) (Figure 3C).

FIGURE 3.

The metabolome after short‐term alcohol use has elevated levels of metabolites involved in epigenetic remodeling and inflammation regulation. (A) sPLSDA of LC–MS analysis of macaque plasma samples from ethanol naive controls and after 6 months. (B) Volcano plot of differentially abundant metabolites in plasma samples after 6 months compared to controls. (C) Bar plot depicting selected enrichments of differentially abundant metabolites in plasma samples after 6 months (terms with p‐value < 0.1 and > 2 genes). For plasma LC–MS: N = 8 controls, n = 23 after 6 months of alcohol.

4. Discussion

The gut microbiome plays a critical role in regulating host health [7]. Unfortunately, the microbiome can be altered by lifestyle factors, including alcohol consumption [6]. Given the increasing prevalence of alcohol misuse, it is important to understand the dose‐ and duration‐dependent effects of alcohol on the microbiome and its subsequent impact on the immune system. Therefore, in this study, we aim to bridge this gap by utilizing 16S rRNA sequencing to assess the implications of short‐term alcohol consumption on the gut microbiome and mass spectrometry to investigate the subsequent effects on the fecal and circulating metabolome.

We found that short‐term alcohol consumption altered the relative abundance of microbes. Specifically, after 6 months of alcohol use, the abundance of Senegalimassilia, which may help regulate blood pressure [56], is increased. Since blood pressure can be elevated with AUD [57], this may be a compensatory mechanism. Prevotellaceae, a known inhibitor of inflammatory processes through activation of PPARα and reduction in NFκβ activity [58], was also increased after 6 months of alcohol use. Similarly, Faecalibacterium, shown to be negatively correlated with inflammatory bowel disease [41] was more abundant after 6 months of alcohol use. These data align with the hypo‐inflammatory phenotype observed in monocytes isolated from macaques after 6–months of voluntary ethanol consumption [59]. We also observe a change in the abundance of SCFA producers, including the genera Lachnospiraceae, Parabacteroides, and Rumminococcus, concurrent with a decrease in Eubacterium, Subdoligranulum, Lactobacillus, and the family Butyricioccaceae. SCFA are known to facilitate gut barrier integrity [60]. Studies in humans [61] and mice [62] demonstrate that binge drinking episodes are sufficient to disrupt the gut barrier and allow for significantly elevated levels of circulating endotoxins. Therefore, it is probable that short‐term alcohol use also enhances intestinal permeability. Taken together, these data suggest 6 months of alcohol use induces significant reorganization of SCFA‐producing bacteria, possibly the reprogramming that sets in motion the changes in SCFA noted with chronic alcohol use.

When using LC–MS to survey changes in the fecal metabolome, we found that short‐term alcohol use significantly reduced levels of numerous medium‐ and long‐chain fatty acids. This is likely due to oxidative degradation via reactive oxygen species, which is increased with alcohol misuse [63]. Medium‐chain fatty acids also contribute to the activation of the inflammatory cascade through interactions with G protein‐coupled receptor 84 (GPR84) [64]. The role of long‐chain fatty acids in inflammation is less clear, as some exhibit both inflammatory and anti‐inflammatory properties, depending on the cell type and stimuli (e.g., oleic acid); in contrast, others are known to promote inflammatory processes (e.g., palmitic acid) [65]. Increased palmitic acid concentrations in erythrocytes of patients with alcoholic liver disease reduce membrane fluidity, impairing hepatic circulation and metabolism [66, 67]. Reduced palmitic acid levels observed in this study suggest that 6 months of drinking may reduce hepatic inflammation.

Metabolites deoxycholic acid (DCA), products of cysteine metabolism (mercaptopyruvic acid, sulfuric acid), and histidine metabolism (urocanic acid) were increased in the fecal metabolome. Interestingly, urocanic acid and DCA have anti‐inflammatory properties, reduce the production of inflammatory mediators (e.g., IL6 and TNFα), and interfere with the inflammasome. They were also linked to a tolerogenic phenotype in macrophages upon activation [68, 69, 70]. This evidence aligns with the hypo‐inflammatory profile, characterized by reduced cytokine production and increased expression of hypo‐inflammatory genes, that we recently described in monocytes following 6 months of daily heavy drinking [59]. Together, these data suggest that alcohol‐induced changes in the gut metabolome can/may contribute to the regulation of immune responses.

Higher levels of guanine, guanosine, and glutamic acid in circulation, among other metabolites linked with epigenetic and transcriptional regulation [53, 54, 55], were observed in plasma after 6 months of daily heavy drinking. We also observed an increase in β‐hydroxybutyrate, a ketone body known to be synthesized primarily by the liver and released into peripheral tissues [71], where it inhibits histone deacetylases [72], possibly driving the elevated deposition of acetylations observed in the amygdala [73]. Moreover, levels of stachydrine [74], nicotinamide [75], itaconate [76], and uradine [77], which can regulate inflammation, were increased. Increased levels of these metabolites could further explain the reduction in TNFα and IL6 production by monocytes after 6 months of alcohol exposure [59]. Short‐term alcohol use also altered host response to SARS‐CoV‐2 infection by lung‐resident macrophages, increasing BDNF, FGFβ, TNFα, production while decreasing secretion of VEGFA, TGFα, IFNβ, IL‐15, PD‐L1, and CCL2 [78]. Urate, shown herein to be upregulated with short‐term drinking, is linked to enhanced inflammation in the lungs [79] and has antioxidant and neuroprotective properties in the brain micro‐environment [80]. These data suggest that metabolites exert compartment‐specific effects on host inflammation.

While our findings provide valuable insights into alcohol‐mediated changes in gut microbiome and metabolomes, it is important to consider some limitations. Despite being able to assess sex as a biological variable when evaluating the effects of short‐term alcohol consumption on microbial communities, we were not powered to conduct the same comparisons on our metabolomics data. Additionally, we did not assess the expression of tight junction proteins. Future studies should collect longitudinal biopsies to understand the implications of short‐term alcohol use on the expression of ZO‐1 and occludin, known regulators of barrier function. Moreover, inflammatory hallmarks of cardiovascular disease (elevated cholesterol, blood pressure, heart rate, plaque deposition) [81], liver disease (elevated AST, ALT) [82], neural inflammation (increased neuroimmune markers in cerebrospinal fluid and microglia activation via positron emission tomography) [83], and wound healing (decreased wound healing rate) [83] could be assessed alongside microbial and metabolic studies to confirm that changes in these compartments dampen overall inflammation in the macaque model.

In conclusion, our study expands the current understanding of how alcohol longitudinally changes the gut microbial communities and their implications for metabolic profiles. The data herein demonstrate that short‐term alcohol use disrupts levels of genera known to produce SCFA, while the abundance of genera associated with regulating inflammation increased. Metabolic profiles of both gut microbial community and plasma were poised towards inflammation resolution, albeit with differing metabolite signatures. These data are in line with studies from our group and others indicating lower inflammatory responses by monocytes after short‐term ethanol consumption.

Author Contributions

Ilhem Messaoudi, Madison B. Blanton, and Ethan G. Napier conceived and designed the research. Madison B. Blanton, Ethan G. Napier, Katelyn E. Keen, Ethan V. Stuart, Isaac R. Cinco, and Ronald C. Bruntz performed the research and acquired the data. Madison B. Blanton, Ethan G. Napier, Katelyn E. Keen, Ethan V. Stuart, Hami Hemati, and Ilhem Messaoudi analyzed and interpreted the data. Rupak Khadka and Kathleen A. Grant aided in animal selection, sample collection, and supervised animal care. Madison B. Blanton, Ethan G. Napier, and Ilhem Messaoudi drafted the manuscript. All authors were involved in revising the manuscript. Madison B. Blanton and Ethan G. Napier should be considered joint first authors.

Funding

This work was supported by the NIH 1R01AA028735‐01 (IM), U01AA013510 (KG), R24 AA019431 (KG), 1F31AA031600‐01A1 (MB), and T32 AA027488 (EN).

Conflicts of Interest

The authors declare no conflicts of interest.

Supporting information

Figure S1: (A) Bubble plot depicting changes in bacterial abundances associated with 6 months of captivity (n = 8). The size of the bubble represents the relative abundance of the microbe, and the color represents the linear discriminant analysis (LDA). (B) Linear regressions between the abundance of bacterial genera or families and the amount of alcohol consumed (g of EtOH/kg of NHP).

Figure S2: (A) Bar plot of results from targeted metabolomics (GC–MS) determining SCFA levels in fecal samples. Error bars represent SEM.

Table S1: Differentially abundant metabolites.

Blanton M. B., Napier E. G., Keen K. E., et al., “Short‐Term Alcohol Consumption Is Sufficient to Disrupt Microbial Communities and the Intestinal and Circulating Metabolites in a Non‐Human Primate Model,” The FASEB Journal 39, no. 23 (2025): e71280, 10.1096/fj.202503336R.

Data Availability Statement

Additional information and requests for resources should be directed to Dr. Ilhem Messaoudi (ilhem.messaoudi@uky.edu). The datasets supporting the conclusions of this article are available on NCBI's Sequence Read Archive (SRA# PRJNA1276562).