Bacteriophage targeting of gut bacterium attenuates alcoholic liver disease

Abstract

Chronic liver disease due to alcohol-use disorder contributes markedly to the global burden of disease and mortality^1-3. Alcoholic hepatitis is a severe and life-threatening form of alcohol-associated liver disease. The gut microbiota promotes ethanol-induced liver disease in mice⁴, but little is known about the microbial factors that are responsible for this process. Here we identify cytolysin-a two-subunit exotoxin that is secreted by Enterococcus faecalis^5,6-as a cause of hepatocyte death and liver injury. Compared with non-alcoholic individuals or patients with alcohol-use disorder, patients with alcoholic hepatitis have increased faecal numbers of E. faecalis. The presence of cytolysin-positive (cytolytic) E. faecalis correlated with the severity of liver disease and with mortality in patients with alcoholic hepatitis. Using humanized mice that were colonized with bacteria from the faeces of patients with alcoholic hepatitis, we investigated the therapeutic effects of bacteriophages that target cytolytic E. faecalis. We found that these bacteriophages decrease cytolysin in the liver and abolish ethanol-induced liver disease in humanized mice. Our findings link cytolytic E. faecalis with more severe clinical outcomes and increased mortality in patients with alcoholic hepatitis. We show that bacteriophages can specifically target cytolytic E. faecalis, which provides a method for precisely editing the intestinal microbiota. A clinical trial with a larger cohort is required to validate the relevance of our findings in humans, and to test whether this therapeutic approach is effective for patients with alcoholic hepatitis.

Extended Data Figure 1.. Intestinal dysbiosis in patients with alcoholic hepatitis.

(a) 16S rRNA sequencing of fecal samples from controls (n=14), patients with alcohol use disorder (AUD; n=43), or alcoholic hepatitis (n=75). The graph demonstrates the relative abundance of sequence reads in each genus. (b) Bacterial diversity (Shannon-Index and Simpson-Index) and richness (Chao-Richness) was calculated in controls (n=14), patients with AUD (n=43), or alcoholic hepatitis (n=75). (c) E. faecalis in fecal samples from controls (n=25), patients with AUD (n=38), or alcoholic hepatitis (n=82), assessed by qPCR. (d) Percentage of fecal samples positive for E. faecalis in controls (n=25), patients with AUD (n=38), or alcoholic hepatitis (n=82), assessed by qPCR. E. faecalis was detected in feces from 80% of patients with alcoholic hepatitis vs 36% of controls (P<0.001). There was also a significant difference between patients with alcohol use disorder and alcoholic hepatitis (P<0.01). (e) ROC curves and AUC for the comparison of 90-day mortality and cytolysin positivity (red; n=57), MELD score (blue; n=56), ABIC score (yellow; n=57), and DF (green; n=42) in patients with alcoholic hepatitis. (f) E. faecalis in fecal samples from patients with alcoholic hepatitis whose fecal samples were cytolysin positive (n=25) or cytolysin negative (n=54), assessed by qPCR (P=0.8174). (g) 16S rRNA sequencing of fecal samples from patients with alcoholic hepatitis from different centers (France n=9; Mexico n=6; Spain n=5; UK n=11; USA East n=16; USA Midwest n=12; USA West n=16). Principal coordinate analysis (PCoA) based on Jaccard dissimilarity matrices was used to show ß-diversity among groups, at the genus level. Composition of fecal microbiota was significantly different (P<0.01). (h) Percentage of fecal samples positive for cylL_L and cylL_S DNA sequences (cytolysin-positive), in patients with alcoholic hepatitis from different centers (France n=16; Mexico n=6; Spain n=6; UK n=10; USA East n=16; USA Midwest n=13; USA West n=15), assessed by qPCR (P=0.6094). (i) E. faecalis in fecal samples from patients with alcoholic hepatitis from different centers, assessed by qPCR (P=0.5648). (j) Percentage of fecal samples positive for E. faecalis in patients with alcoholic hepatitis from different centers (France n=16; Mexico n=6; Spain n=6; UK n=10; USA East n=16; USA Midwest n=13; USA West n=15), assessed by qPCR (P=0.0529). (k) Percentage of subjects with fecal samples positive for cylL_L and cylL_S DNA sequences (cytolysin positive), in patients with alcoholic hepatitis and cirrhosis (n=30), or without cirrhosis (n=18), assessed by qPCR (P=0.3431). (l) E. faecalis in fecal samples from patients with alcoholic hepatitis and cirrhosis (n=30), or without cirrhosis (n=18), assessed by qPCR (P=0.5736). (m) Percentage of fecal samples positive for E. faecalis in alcoholic hepatitis patients with cirrhosis (n=30), or without cirrhosis (n=18), assessed by qPCR (P=0.2878). Results are expressed as as mean ± s.e.m (c, f, i, l). For the Box and Whisker plots in (b), the box extends from 25^th to 75^th percentiles, with the center line representing the median; for all three groups, the lower whiskers show the minimum values; for the “Controls” group (black), the higher whisker shows the maximum value; for the other two groups, the higher whiskers represent the 75^th percentile plus 1.5 times inter-quartile distance (the distance between the 25^th and 75^th percentiles), all values greater than this are plotted as individual dots. P values are determined by Kruskal-Wallis test (i) with Dunn’s post-hoc test (b, c), two-sided Fisher’s exact test (h, j, k, m) followed by false discovery rate (FDR) procedures (d), two-sided Mann-Whitney-Wilcoxon rank-sum test (f, l), or permutational multivariate analysis of variance (PERMANOVA) (g). The exact group size (n) and P values for each comparison are listed in Supplementary Table 10. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

Extended Data Figure 2.. Cytolytic E. faecalis causes progression of ethanol-induced liver disease in mice.

(a–n) C57BL/6 mice were fed oral isocaloric (control) or chronic–binge ethanol diets and gavaged with vehicle (PBS), a cytolytic E. faecalis strain (FA2–2(pAM714)) (E. faecalis) (5×10⁸ colony forming units (CFUs)), or a non-cytolytic E. faecalis strain (FA2–2(pAM771)) (E. faecalis Δcytolysin) (5×10⁸ CFUs) every third day. (a) Serum levels of ALT. (b) Hepatic triglyceride content. (c) Representative oil red O-stained liver sections. (d–f) Hepatic levels of mRNAs. (g) Kaplan-Meier curve of survival of mice on chronic–binge ethanol diets (day 0, start of ethanol feeding). Mice gavaged with PBS all survived and were not included in figure. A higher proportion of mice (n=15) gavaged with non-cytolytic E. faecalis survived than mice (n=25) gavaged with cytolytic E. faecalis. (h) Proportions of mice positive for cytolysin in liver, measured by qPCR for cylL_S (the gene encoding cytolysin subunit CylL_S”). (i) Proportions of mice positive for E. faecalis in liver, measured by qPCR. About 80% of mice colonized with cytolytic E. faecalis, as well as non-cytolytic E. faecalis, were positive for E. faecalis in their livers. (j) Liver CFUs of Enterococcus in mice on a chronic–binge ethanol diet. (k) Paracellular intestinal permeability was evaluated by measuring fecal albumin content and serum levels of lipopolysaccharide (LPS) by ELISAs. (l) Fecal samples were collected and 16S rRNA genes were sequenced. Principal coordinate analysis based on Jaccard dissimilarity matrices showed no significant differences among mice gavaged with PBS, cytolytic or non-cytolytic E. faecalis following the diets. Compared to control-diet fed mice, mice fed with ethanol diet had significant different fecal microbiomes after gavaging E. faecalis (P<0.05). (m and n) Serum levels of ethanol and hepatic levels of Adh1 and Cyp2e1 mRNAs did not differ significantly among these mice after ethanol feeding. (o) Mice were gavaged with cytolytic or non-cytolytic E. faecalis strains (carrying erythromycin resistance gene; 5×10⁸ CFUs) at time 0, and feces was collected at 0, 8, 24, 48, and 72 hrs. Fecal CFUs of Enterococcus were determined by culturing fecal samples on BBL Enterococcosel Broth agar plate with 50 μg/ml erythromycin. At time 0 and 72, 5/5 and 4/5 mice, respectively, had no detectable erythromycin-resistant Enterococcus in feces. These points are not shown on the graph, but have been included in the calculation of mean ± s.e.m. Scale bar=100 μm. Results are expressed as mean ± s.e.m (a, b, d–f, j, k, m–o). P values among groups of mice fed with control diet or ethanol diet are determined by One-way ANOVA with Tukeýs post-hoc test (a, b, d–f, j, k, m, n), two-sided Log-rank (Mantel-Cox) test (g), two-sided Fisher’s exact test followed by FDR procedures (h and i), or PERMANOVA followed by FDR procedures (l). All results were generated from at least three independent replicates. The exact group size (n) and P values for each comparison are listed in Supplementary Table 10. P values between control-diet fed mice and ethanol-diet fed mice are determined by Two-way ANOVA (k). *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

Extended Data Figure 3.. Transplantation of cytolysin-positive feces increases ethanol-induced liver disease in gnotobiotic mice.

(a–f, h, i) C57BL/6 germ-free mice were colonized with feces from two different cytolysin-positive and two different cytolysin-negative patients with alcoholic hepatitis, and then fed isocaloric (control) or chronic–binge ethanol diets. (a) Percentage of terminal deoxynucleotide transferase-mediated dUTP nick-end labeling (TUNEL) positive hepatic cells. (b) Representative oil red O-stained liver sections. (c and d) Hepatic levels of mRNAs encoding the inflammatory cytokine Cxcl2, and Acta2 (marker of activated hepatic stellate cells). (e) Kaplan-Meier curve of survival of mice on chronic–binge ethanol diets (day 0, start of ethanol feeding) gavaged with feces from cytolysin-positive (n=48 mice) or cytolysin-negative (n=32 mice) patients with alcoholic hepatitis. (f) Fecal samples were collected and 16S rRNA genes were sequenced. The graph shows principal coordinate analysis of fecal microbiomes. No significant difference was observed between mice colonized with feces from cytolysin positive or negative alcoholic hepatitis donors following the control diet. Mice transplanted with feces from cytolysin-positive alcoholic hepatitis patient #2 showed a significantly different microbiota than the other mouse groups following ethanol administration (P<0.01). (g) Percentage of cytolysin-positive E. faecalis in 4 patients with alcoholic hepatitis. Stool samples from the four patients were placed on plates with selective medium and Enterococcus colonies were identified by the production of dark brown or black color generated by hydrolysis of esculin to esculetin that reacts with ferric ammonium citrate. Enterococcus colonies were confirmed to be E. faecalis by qPCR. Cytolysin status of each E. faecalis colony was determined by qPCR. (h) Serum levels of ethanol were comparable among colonized mice after ethanol feeding. (i) Hepatic levels of Adh1 and Cyp2e1 mRNAs did not differ significantly among colonized mice on control or ethanol diets. Scale bar=100 μm. Results are expressed as mean ± s.e.m (a, c, d, h, i). P values are determined by One-way ANOVA with Tukeýs post-hoc test (a, c, d, h, i), two-sided Log-rank (Mantel-Cox) test (e), or PERMANOVA followed by FDR procedures (f). All results were generated from at least three independent replicates. The exact group size (n) and P values for each comparison are listed in Supplementary Table 10. *P<0.05, **P<0.01, ***P<0.001.

Extended Data Figure 4.. Isolation and amplification of bacteriophages against cytolytic E. faecalis isolated from mice.

(a) BHI agar plate showing bacteriophage plaque morphology. Bacteriophage cocktail (100 μl, 10²–10³ PFUs) was mixed with overnight grown E. faecalis culture (100 μl) and then added to BHI broth top agar (0.5% agar) and poured over a BHI plate (1.5% agar). After overnight growth at 37°C, images were captured on an Epson Perfection 4990 Photo scanner. (b) Simplified illustration of different bacteriophage morphology. Siphophages have long, flexible, noncontractile tails (left), myophages have contractile tails (middle), and podophages have short, noncontractile tails (right). (c) Transmission electron microscopy revealed that bacteriophages isolated were all podophages (Efmus1, Efmus2, Efmus3 and Efmus4). Phages specific to E. faecalis strain isolated from mouse feces were named as Efmus with a number (Ef for E. faecalis, mus for mouse, digit for isolation order). (d) Genetic map of phage genomes. The linear maps are based on nucleotide sequences of the phage genomes and predicted open reading frames. The name and length (bp) of each genome are indicated to the left of each phage map. Protein-coding sequences are colored based on functional role categories (see box). Scale bar=50 nm. All results were generated from at least three independent replicates.

Extended Data Figure 5.. Phages reduce translocation of cytolysin to the liver and reduce ethanol-induced liver disease in Atp4a^Sl/Sl mice.

(a–k) Wild-type (WT) and their Atp4a^Sl/Sl littermates were fed oral isocaloric (control) or chronic–binge ethanol diets and gavaged with vehicle (PBS), control phages against C. crescentus (10¹⁰ plaque forming units (PFUs)), or a cocktail of 4 different phages targeting cytolytic E. faecalis (10¹⁰ PFUs) 1 day before ethanol binge. (a) Serum levels of ALT. (b) Hepatic triglyceride content. (c) Representative oil red O-stained liver sections. (d–f) Hepatic levels of mRNAs. (g) Proportions of mice positive for cytolysin in liver, measured by qPCR for cylL_S. (h) Fecal colony forming units (CFUs) of Enterococcus. (i) Fecal samples were collected and 16S rRNA genes were sequenced. Principal coordinate analysis based on Jaccard dissimilarity matrices found no significant difference in fecal microbiota among mice given PBS, control phage, or phages targeting cytolytic E. faecalis in each group. (j and k) Serum levels of ethanol and hepatic levels of Adh1 and Cyp2e1 mRNAs did not differ significantly among colonized mice after ethanol feeding. Scale bar=100 μm. Results are expressed as mean ± s.e.m (a, b, d–f, h, j, k). P values are determined by Two-way ANOVA with Tukeýs post-hoc test (a, b, d–f, h, j, k), two-sided Fisher’s exact test followed by FDR procedures (g), or PERMANOVA followed by FDR procedures (i). All results were generated from at least three independent replicates. The exact group size (n) and P values for each comparison are listed in Supplementary Table 10. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

Extended Data Figure 6.. Isolation and amplification of bacteriophages against cytolytic E. faecalis isolated from patients with alcoholic hepatitis.

(a) BHI agar plates showing bacteriophage plaque morphology. (b) Transmission electron microscopy graphs of myophages Ef2.1 and Ef2.3, stained with phosphotungstic acid showing contracted tails. (c) Genetic map of phage genomes. The linear maps are based on nucleotide sequences of the phage genomes and predicted open reading frames. The name and length (bp) of each genome are indicated to the left of each phage map. Protein-coding sequences are colored based on functional role categories (see box). Sequences encoding tRNA genes are indicated by a cloverleaf structure. Scale bar=50 nm. All results were generated from at least three independent replicates.

Extended Data Figure 7.. Phages that target cytolytic E. faecalis reduce ethanol-induced liver disease in gnotobiotic mice.

(a–h) C57BL/6 germ-free mice were colonized with feces from two different cytolysin-positive patients with alcoholic hepatitis (feces from 1 patient is used in Figure 2). The mice were then fed oral isocaloric (control) or chronic–binge ethanol diets, and gavaged with control phages against C. crescentus (10¹⁰ plaque forming units (PFUs)), or a cocktail of 3 or 4 different phages targeting cytolytic E. faecalis (10¹⁰ PFUs) 1 day before an ethanol binge. (a) Percentage of TUNEL-positive hepatic cells. (b) Representative oil red O-stained liver sections. (c and d) Hepatic levels of mRNAs encoding the inflammatory cytokine Cxcl2, and Acta2 (marker of activated hepatic stellate cells). (e) Fecal colony forming units (CFUs) of Enterococcus. (f) Fecal samples were collected and 16S rRNA genes were sequenced. Principal coordinate analysis based on Jaccard dissimilarity matrices shows no significant differences in fecal microbiota of mice gavaged with control phage and phages targeting cytolytic E. faecalis in each group. (g and h) Serum levels of ethanol and hepatic levels of Adh1 and Cyp2e1 mRNAs did not differ significantly among colonized mice after ethanol feeding. Scale bar=100 μm. Results are expressed as mean ± s.e.m (a, c–e, g, h). P values are determined by Two-way ANOVA with Tukeýs post-hoc test (a, c–e, g, h), or PERMANOVA followed by FDR procedures (f). All results were generated from at least three independent replicates. The exact group size (n) and P values for each comparison are listed in Supplementary Table 10. *P<0.05, ***P<0.001.

Extended Data Figure 8.. Isolation and amplification of bacteriophages against non-cytolytic E. faecalis isolated from patients with alcoholic hepatitis.

(a) BHI agar plates showing bacteriophage plaque morphology. (b) Genetic map of phage genomes. The linear maps are based on nucleotide sequences of the phage genomes and predicted open reading frames. The name and length (bp) of each genome are indicated to the left of each phage map. Protein-coding sequences are colored based on functional role categories (see box). Sequences encoding tRNA genes are indicated by a cloverleaf structure. (c) Phylogenetic tree of Enterococcus bacteriophages. A whole-genome average nucleotide distance tree was constructed for 73 available Enterococcus phage genomes, 54 from GenBank (black letters), 19 from this study (4 phages against cytolysin-positive E. faecalis isolated from mice in blue letters, 7 phages against cytolysin-positive E. faecalis isolated from alcoholic hepatitis patients in pink letters, and 8 phages against cytolysin-negative E. faecalis isolated from alcoholic hepatitis patients in green letters) with Mash using a sketch size of s=5000 and k-mer size of k=12 and GGRaSP (see Methods). Colored branches denote specific phage genera: Sap6virus, P68virus and Spounavirinae. The scale bar represents percent average nucleotide divergence. All results were generated from at least three independent replicates.

Extended Data Figure 9.. Phages that target non-cytolytic E. faecalis do not reduce ethanol-induced liver disease in gnotobiotic mice.

(a–h) C57BL/6 germfree mice were colonized with feces from two different cytolysin-negative patients with alcoholic hepatitis. Transplanted gnotobiotic mice were fed oral isocaloric (control) or chronic–binge ethanol diets and gavaged with control phages against C. crescentus (10¹⁰ plaque forming units (PFUs)), or a cocktail of 4 different phages targeting non-cytolytic E. faecalis (10¹⁰ PFUs) 1 day before an ethanol binge. (a) Percentage of TUNEL-positive hepatic cells. (b) Representative oil red O-stained liver sections. (c and d) Hepatic levels of mRNAs encoding the inflammatory cytokine Cxcl2, and Acta2 (marker of activated hepatic stellate cells). (e) Proportions of mice positive for cytolysin in liver, measured by qPCR for cylL_S. (f) Fecal samples were collected and 16S rRNA genes were sequenced. Principal coordinate analysis based on Jaccard dissimilarity matrices found no significant difference in fecal microbiota among mice gavaged with control phages and phages targeting cytolytic E. faecalis in each group. (g and h) Serum levels of ethanol and hepatic levels of Adh1 and Cyp2e1 mRNAs did not differ significantly among colonized mice after ethanol feeding. Scale bar=100 μm. Results are expressed as mean ± s.e.m (a, c, d, g, h). P values are determined by Two-way ANOVA with Tukeýs post-hoc test (a, c, d, g, h), two-sided Fisher’s exact test followed by FDR procedures (e), or PERMANOVA followed by FDR procedures (f). All results were generated from at least three independent replicates. The exact group size (n) and P values for each comparison are listed in Supplementary Table 10.

Figure 1.. E. faecalis cytolysin associates with mortality in patients with alcoholic hepatitis.

(a) 16S rRNA sequencing of fecal samples from controls (n=14), patients with alcohol use disorder (AUD; n=43), or alcoholic hepatitis (n=75). Principal coordinate analysis (PCoA) based on Jaccard dissimilarity matrices was used to show ß-diversity among groups, at the genus level. Composition of fecal microbiota was significantly different between each group (P<0.01). (b) Percentage of subjects with fecal samples positive for cylL_L and cylL_S DNA sequences (cytolysin-positive), in controls (n=25), patients with AUD (n=38), or alcoholic hepatitis (n=82), assessed by qPCR. Statistically significant differences were detected between controls and alcoholic hepatitis patients (P<0.01), and between patients with alcohol use disorder and alcoholic hepatitis patients (P<0.001). (c) Kaplan-Meier curve of survival of patients with alcoholic hepatitis whose fecal samples were cytolysin-positive (n=25) or cytolysin-negative (n=54) (P<0.0001). (d) Core genome single nucleotide polymorphism (SNP) tree of E. faecalis strains isolated from patients with alcoholic hepatitis (n=93 from 24 patients), showing phylogenetic diversity of cytolysin-positive (red) E. faecalis. Genomically identical isolates from the same patient were combined and are shown as one single dot. Scale bar represents the nucleotide substitutions per SNP site. P values are determined by permutational multivariate analysis of variance (PERMANOVA) followed by false discovery rate (FDR) procedures (a), two-sided Fisher’s exact test followed by FDR procedures (b), or two-sided Log-rank (Mantel-Cox) test (c). The exact group size (n) and P values for each comparison are listed in Supplementary Table 10.

Figure 2.. Transplantation of feces from cytolysin-positive patients with alcoholic hepatitis exacerbates ethanol-induced liver disease in gnotobiotic mice.

(a–g) C57BL/6 germfree mice were colonized with feces from two different cytolysin-positive and two different cytolysin-negative patients with alcoholic hepatitis and subjected to the chronic–binge feeding model. (a) Serum levels of ALT. (b) Hepatic triglyceride content. (c) Representative H&E-stained liver sections. (d–f) Hepatic levels of mRNAs encoding Il1b, Cxcl1, and Col1a1. (g) Proportions of mice positive for cytolysin in liver, measured by qPCR for cylL_S. (h) LDH assay to measure cytotoxicity of hepatocytes isolated from mice fed oral isocaloric (control) diet (5 groups, left) or chronic–binge ethanol diet (5 groups, right) and incubated with vehicle, CylL_S”, CylL_L”, or both cytolysin subunits, at indicated concentrations, without (–) or with (+) ethanol (25mM) for 3 hours. Survival of hepatocytes was determined in 3 independent experiments. Scale bar=100 μm. Results are expressed as mean ± s.e.m (a, b, d–f, h). P values are determined by One-way ANOVA with Tukeýs post-hoc test (a, b, d–f), two-sided Fisher’s exact test followed by FDR procedures (g), or Two-way ANOVA with Tukeýs post-hoc test (h). All results were generated from at least three independent replicates. The exact group size (n) and P values for each comparison are listed in Supplementary Table 10. *P<0.05, **P<0.01, ***P<0.001, ****P<0.0001.

Figure 3.. Phage therapy against cytolytic E. faecalis abolishes ethanol-induced liver disease in gnotobiotic mice.

(a) Transmission electron microscopy revealed that bacteriophages isolated were either siphophages (Ef5.1, Ef5.2, Ef5.3, Ef5.4 and Ef2.2) or myophages (Ef2.1 and Ef2.3). Scale bar=50 nm. (b–h) C57BL/6 germfree mice were colonized with feces from two different cytolysin-positive patients with alcoholic hepatitis (feces from 1 patient also used in Figure 2) and subjected to chronic–binge feeding model, gavaged with control phages against C. crescentus (10¹⁰ PFUs), or a cocktail of 3 or 4 different phages targeting cytolytic E. faecalis (10¹⁰ PFUs) 1 day before an ethanol binge. (b) Serum levels of ALT. (c) Hepatic triglyceride content. (d) Representative H&E-stained liver sections. Scale bar=100 μm. (e–g) Hepatic levels of mRNAs encoding Il1b, Cxcl1 and Col1a1. (h) Proportions of mice positive for cytolysin in liver, measured by qPCR for cylL_S. Results are expressed as mean ± s.e.m (b, c, e–g). P values are determined by Two-way ANOVA with Tukeýs post-hoc test (b, c, e–g) or two-sided Fisher’s exact test followed by FDR procedures (h). All results are generated from at least three independent replicates. The exact group size (n) and P values for each comparison are listed in Supplementary Table 10. *P<0.05, **P<0.01, *** P<0.001.

Figure 4.. Phages that target non-cytolytic E. faecalis do not reduce ethanol-induced liver disease in gnotobiotic mice.

(a) Transmission electron microscopy revealed that bacteriophages isolated were either podophages (Ef6.2, Ef6.3, Ef7.2, Ef7.3 and Ef7.4) or siphophages (Ef6.1, Ef6.4 and Ef7.1). Scale bar=50 nm. (b–h) C57BL/6 germfree mice were colonized with feces from two different cytolysin-negative patients with alcoholic hepatitis and subjected to chronic–binge feeding model, gavaged with control phages against C. crescentus (10¹⁰ PFUs), or a cocktail of 4 different phages targeting non-cytolytic E. faecalis (10¹⁰ PFUs) 1 day before an ethanol binge. (b) Serum levels of ALT. (c) Hepatic triglyceride content. (d) Representative H&E-stained liver sections. Scale bar=100 μm. (e–g) Hepatic levels of mRNAs encoding Il1b, Cxcl1, and Col1a1. (h) Fecal colony forming units (CFUs) of Enterococcus. Results are expressed as mean ± s.e.m (b, c, e–h). P values are determined by Two-way ANOVA with Tukeýs post-hoc test (b, c, e–h). All results were generated from at least three independent replicates. The exact group size (n) and P values for each comparison are listed in Supplementary Table 10. *P<0.05.