Background: Significant sex differences exist in regard to alcohol use disorder (AUD) and alcoholic liver disease (ALD). To date, no studies have examined the brain-gut-microbiome (BGM) axis (which is the relationship between the gut, brain, and the bacteria within the gut) and sex-differences in AUD and ALD.

Aims: 1) Demonstrate baseline sex differences in the microbiome and metatranscriptome of AUD and ALD and correlate those differences to severity, 2) determine if these baseline sex differences predicts abstinence or ALD related outcomes, and 3) show how altering the microbiome can decrease the severity of AUD and ALD in a sexdependent manner. Hypothesis: Our project is aimed to explore the hypothesis that sex-related differences of the BGM axis in AUD and ALD explains the variation in patient severity and outcome by sex, and that alterations of the BGM axis can decrease the severity of AUD and ALD in a sex-dependent manner.

Methods: A pilot randomized placebo (VSL#3 vs placebo) control trial will be performed in patients with AUD and ALD for 6 months. Questionnaire data, clinical labs, serum, and feces for shotgun metagenomics will be collected at baseline, 3-months, and 6-months.

Anticipated Results: Patients with severe AUD/ALD will have more microbes and microbial genes associated with inflammation. These differences will predict outcomes at 6-months and that changes of this baseline microbiome with VSL#3 will lead to more positive outcomes than placebo, with men having greater benefit from VSL#3 than women.

Implications and Future Studies: The discovery of the mechanisms underlying sex-related differences in AUD/ALD is needed for the development of personalized recommendations for prevention and treatment in men and women

The prevalence of AUD and ALD is on the rise and the burden of disease is affecting men and women differently. Research has shown that the prevalence AUD and alcoholic liver disease (ALD) has grown significantly higher for women than for man over the last decade (80% vs 30% for alcohol use disorder and 50% vs 30% for ALD). Research has also found that women, as compared to men, are more likely to succumb to the negative effects of alcohol. Studies have shown there is variability in alcohol absorption, distribution and metabolism between men and women, as a result of hormonal, genetic and environmental factors. While women tend to drink less than men, women who do have AUD are more likely to have ALD, cardiovascular disease, and cancer. Sex differences exist for how women and men consume alcohol but sex differences also exist for how likely they are to abstain, with married women having a higher rate of relapse than married men. But despite recent efforts, the sex disparity in alcohol abuse and alcoholic complications has not been clearly elucidated. There is a bidirectionality between the gut microbiome and AUD and ALD. Heavy alcohol use has been associated with a decrease in intestinal barrier function, increase bacterial translocation, and a shift in the intestinal microbiome to a more inflammatory phenotype. Conversely, recent research have shown probiotics to be beneficial in ALD, and in a recent phase 1 clinical trial, researcher showed that fecal microbial transplant was able to significantly reduce alcohol use disorder in a male population of veterans. Specifically VSL#3, a probiotic mixture, has been shown to be safe and protective against intestinal barrier dysfunction, alcohol-related liver disease, and even re-hospitalization. Therefore, there is mounting data to suggest that the brain-gut-microbiome axis (BGM) may be important in alcohol use disorder and ALD. Yet to date, there is no studies examining the role of the BGM axis in sex differences of alcohol use disorder or ALD. Sex-based differences in the microbiome is related to other metabolic diseases. In both human and animal models, studies have shown that the gut microbiome composition and effects can vary between sexes based on the disease. This group of investigators for example, was able to show a sex-based difference in the BGM axis in relationship to obesity and food addiction. Whether there is a sex difference in the BGM axis that affects disease progression and outcome of AUD and ALD will be the key purpose of this pilot study.

EXPERIMENTAL APPROACH AND TIMELINE The investigators hypothesize that sex-related differences of the BGM axis in AUD and ALD explains the variation in patient outcome by sex, and that alterations of the BGM axis can decrease the severity of alcohol-use disorder and ALD.

Overview of Aim 1: Demonstrate baseline sex differences in the gut microbiome and metatranscriptome of patients with AUD and ALD and correlate those differences to the severity of AUD and ALD. Utilizing fecal collection and shotgun metagenomics, the investigators will identify key bacterial markers and metatranscriptome signatures related to AUD and ALD as compared to healthy controls. The investigators will then determine if these markers are associated with AUD and ALD severity in a sex-related manner. Patients will be recruited from Santa Monica UCLA hospital and Ronald Reagan UCLA Hospital. Potentially eligible patients will be asked by their provider if they can be contacted by the research team. Afterwards, the team will do a preliminary chart review looking at past medical history, problem list, and current inpatient notes to determine if the patient is eligible based on inclusion and exclusion criteria's.

Measurement of Baseline AUD and ALD characteristics. In addition to patient demographic information, the investigators will also collect questionnaire data to assess the severity of their AUD and clinical lab parameters to assess the severity of their ALD. To assess the severity of their AUD, the investigators will use the DSM-V criteria for severity, the AUDIT score, Timeline Followback (TLFB) alcohol assessment, Quality of Life (QOL) scale, and the Alcohol craving questionnaire-Short form (ACQ-SF). To assess the patient's ALD severity, clinical laboratory data will be collected to measure the patients Maddrey's Discriminant Function score, MELD-Na score, Glasgow Alcoholic Hepatitis Score, and baseline transaminases. To assess for other comorbities the team will ask Hospital Anxiety and Depression scale (HADS). DSM-V criteria for severity, the AUDIT score, Timeline Followback (TLFB) alcohol assessment, Quality of Life (QOL) scale, the Alcohol craving questionnaire-Short form (ACQ-SF) and the HADS are questionnaires administered for research purposes. As such the HADS questionnaire will be scored at a later date. Maddrey's Discriminant Function score, MELD-Na score, Glasgow Alcoholic Hepatitis Score, and baseline transaminases are based on clinical labs and medical chart review. Specimen Collection and Microbiome Sequencing. While patients are hospitalized, both ethanol preserved and fresh stool will be collected at baseline along with serum for future analysis similar to our previous works. Stool will be collected in a provided container, the patient will then take a provided para pak vial which contains a spoon and take a small spoon full and place it in the para pak vial which also contains 95% ethanol. These samples will then be stored at -80oC and then processed as a single batch. DNA will be extracted from fecal samples using bead beating in conjunction with the ZymoBiomics DNA Miniprep kit (Zymo Research). DNA extracted from fecal samples will be fragmented and barcoded using a commercial kit (Illumina Nextera Flex). Shotgun libraries will undergo sequencing using an Illumina NovaSeq 6000 (S4 flowcells) with a target depth of 20 million paired-end sequences per sample. Sequence data will then be inputted into MetaPhlAn2 for taxonomic identification of species for compositional analysis equivalent to that performed with 16S rRNA sequence data. The shotgun metagenomics data will also be used to perform strain-level microbiome analysis with the ConStrains algorithm, which identifies single nucleotide polymorphisms in metagenomics sequence data that can be used to track strains across samples. In addition, metagenomics data will be processed in HUMAnN2 for annotation of gene families and identification of the microbial source of these genes. Bacterial gene abundances will be aggregated into pathways based upon KEGG and MetaCyc classifications. Serum processing: Blood will be collected at each study visit (about 3-5 ml). Blood will be centrifuged and serum will be collected. Serum will then be stored at -80C for future use for metabolomics. Deidentified serum samples will be sent to Metabolon for metabolite processing and analysis.

Overview of Aim 2: Determine if baseline sex differences in the gut microbiome and metatranscriptome predicts continued abstinence, recidivism, or ALD-related outcomes. Patients recruited as above will be randomized in a 1:1 fashion to either a placebo or VSL#3 for 6-months. They will take take one VSL#3 pill once a day while they are in the study (112.5 billion CFU per capsule) or a placebo pill with 100 mg of sugar once a day. VSL#3 At 3- months and 6-months patient will return for a visit and the same questionnaires and clinical scores as their baseline visit will be obtained again. Patient will also provide ethanol persevered stool and mail it back to us in a self-addressed paid envelope, and a serum blood sample will be collected at the visit for future analysis . A total of 3-5 ml of blood will be collected at each collection. Prospective outcomes. In addition to the questionnaire and scores listed above measured at 3-months and 6- months, the team will also examine the rate of recidivism, re-hospitalization for alcohol-related causes, all-cause rehospitalization, mortality, and liver transplantation.

Overview Aim 3. Show how altering the gut-microbiome can decrease the severity of AUD and ALD in a sexdependent manner. As mentioned above, patients recruited will be randomized in a 1:1 fashion to either a placebo or VSL#3 for 6-months. They will take one VSL#3 pill once a day (112.5 billion CFU per capsule) or a placebo pill with 100 mg of sugar once a day. VSL#3 was chosen as the probiotic of choice because prior published literature has shown its safety in ALD and animal models have shown it to be effective against inflammation of gut barrier dysfunction in models of AUD and ALD. By power calculation based on previous works on VSL#3, the investigators have determined that about 20 patients will be needed per group to determine a difference (Male-Placebo, Male-VSL#3, Female-Placebo, Female-VSL#3, total=80). Both patient and study personnel will be blinded to the patient's randomization. The study coordinator will be the only person who is not blinded to the patient randomization while the PI and Co-investigators will be blinded throughout the study including the data analysis. Primary Outcomes and Secondary Outcomes. Several previously published trials on VSL#3 have shown no significant adverse events as compared to placebo. Therefore, the primary outcome of the prospective randomized placebo-control study will be sex differences in the rate abstinence and the proportion of resolution of ALD as measured by the Maddrey's Discriminant Function score, MELD-Na score, Glasgow Alcoholic Hepatitis Score, and transaminases. Secondary outcomes will be sex differences regarding microbiome changes, alcohol consumption, QOL, and alcohol craving. Specimen Collection and Microbiome Sequencing. Similar collection and microbiome sequencing will be done as mentioned above. But instead of shotgun metagenomic sequencing, the investigators will perform full 16S sequencing for the samples collected at 6-months to reduce cost. Full 16S sequencing is less expensive but is able to give us similar strain profiles as shotgun metagenomic sequencing.

Inclusion Criteria:

• Adult patients admitted to Santa Monica UCLA hospital or Ronald Reagan UCLA Hospital with a diagnosis of AUD and ALD will be screened and potentially recruited.
• ALD patients will be identified based on the recent American Association of the Study of Liver Disease (AASLD) guidelines: onset of jaundice within 8 weeks, 2) ongoing consumption of ethanol of >40 for women or >60 in men for 6 months or <60 days of abstinence before onset of jaundice, AST>50, AST:ALT>1.5 and both <400 IU/L, total bilirubin >3, or liver biopsy showing histologic features of ALD.
• AUD will be defined based on the current DSM-V definition

Exclusion Criteria:

• To avoid any confounders to the gut microbiome analysis, patients will be excluded if they have been on antibiotics within 3 months or probiotics within 1 month of enrollment
• If patients have a history of inflammatory bowel disease, irritable bowel syndrome, gastrointestinal malignancy, or gastrointestinal surgery.
• If patients are started on antibiotics while in the study, they will be withdrawn from the study to avoid confounders of antibiotic use on microbiome analysis.
• To avoid any cofounders of advanced liver fibrosis on the microbiome analysis, we will also exclude patients with underlying cirrhosis and/or overt hepatic encephalopathy (West Haven criteria >3).
• Patients with acute pancreatitis or history of chronic pancreatitis will also be excluded. The diagnosis of acute pancreatitis and chronic pancreatitis will be ascertained from their history and physical.
• Cirrhosis will be determined by a review of their medical charts and Fib4 calculation. If their medical charts do not mention any history of cirrhosis and their Fib4 score is less than or equal to 1.35 then they will be deemed as being non-cirrhotic. Subjects will be provided their Fib4 score if it is greater than 1.35 and instructed to follow up with their primary care physician for further evaluation if warranted. Fib4 is based on their clinical blood work and age which will be abstracted from their chart and is validated at ruling out advanced liver fibrosis and cirrhosis.

Patient will be recruited from the inpatient ward at RR and SM UCLA hospital. The primary team will identify patients who have alcohol use disorder. The study coordinator will then chart review the patient and determine eligibility based on the inclusion and exclusion criteria.