P- 9 BIOMARKERS OF THE BACTERIAL, VIRAL AND HUMAN GUT TRANSCRIPTOME IN METABOLIC ASSOCIATED FATTY LIVER DISEASE (MAFLD) IN ARGENTINA

Abstract

Introduction and Objectives: The functional dynamics of the gut microbiome and its interactions with the human transcriptome represent a niche for non-invasive biomarkers to risk-stratify MAFLD. This study aimed to identify gut transcriptomic signatures associated with MAFLD in Argentina. Materials and Methods: Stool samples, diet, demographic and clinical data were obtained from 33 biopsy-proven MAFLD patients (12 simple steatosis -SS- and 21 steatohepatitis -SH-) and 19 healthy volunteers (HV). RNA-seq was performed in NovaSeq6000®. Data were analyzed with Maaslin2-v1.2.0, bioBakery-v1.8 and DESeq2-v4.1. Co-expression analysis was performed with Hmisc-v4.7-0. Results: BMI was higher in MAFLD, particularly in SH patients (q=4.49 × 10−6). After adjusting for BMI, differentially expressed genes (DEGs) were found when comparing MAFLD vs. HV and SH vs. SS in bacterial (5 and 13, respectively), viral (112 and 26, respectively) and human (4 and 46, respectively), transcriptomes (q<0.01). Functional profiling of DEGs in MAFLD and SH patients revealed augmented bacterial sulfur and uric acid metabolisms, viral life cycle and viral regulation of the host immune system. Inflammatory regulation, lipid, iron and carbohydrate metabolisms, and response to oxidative stress were enhanced among human DEGs. After comparing transcript abundance, the most active bacterial families were Bactereoidaceae, Rikenellaceae, Oscillospiraceae and Prevotellaceae in all groups. Bifidobacteriaceae expression occurred mostly in HV, while Prevotellaceae prevailed in MAFLD patients. The Firmicutes/Bacteroidetes ratio was higher in MAFLD and SH. Myoviridae, Podoviridae, Siphoviridae and Microviridae were the most transcriptionally active viral families in all groups. Myoviridae and Microviridae showed up-regulated activity in MAFLD (FDR=0.006 for Microviridae) and SH groups (FDR=0.01 and 4.2 × 10−6, respectively), whereas Podoviridae and Siphoviridae were less active in these groups. Significant correlations were observed between the expression of Faecalibacterium phage Mushu, Prevotella copri and the human mucin gene (Figure). Conclusions: We identified specific signatures of the interaction between microbial and human gut transcriptomes that could be useful as non-invasive biomarkers of MAFLD diagnosis and progression.