Abstract:
An imbalance in gut microbiome is strongly linked to liver inflammation disease and hepatocellular carcinoma (HCC) via the gut-liver axis. However, the understanding of how gut microbiota interacts with the host gene expression is still limited. In this study, we aim to investigate the relationship between gut microbiome profile and transcriptomic profile in patients with HCC. In this study, 17 patients with viral-related HCC, 13 non-viral-related HCC, and 10 healthy controls were recruited. We investigated gut microbiome profile from fecal samples using 16S rRNA sequencing and host transcriptomic profile from the peripheral blood mononuclear cells (PBMCs) using RNA sequencing method. Individual datasets were examined and integrated for association analysis between two datasets using bioinformatic tools. Moreover, machine learning has been performed to detect HCC and then identify that bacterial and genes that can be used as diagnostics for HCC. Based on Pearson’s correlation analysis, the interaction of 268 gut microbes and 6,137 genes were performed. We found that 4 genera of bacteria were associated with 18 host genes expression. In these interactions, these bacteria was related to lipopolysaccharide (LPS) production and the functional analysis of those genes was mainly involved in signal transduction and immune regulation. Finally, based on machine learning approach, 4 genera of bacteria including Eubacterium, Eubacterium nodatum group, Lachnospiraceae AC2044 group and Ruminococcus gnavus group were revealed to be diagnostic biomarkers in discriminating non-viral-related HCC from viral-related HCC (AUC = 0.85, Sensitivity = 88%, Specificity = 80% and Accuracy = 86%). However, the performance in differentiate the non-viral and viral-related HCC of host genes were not satisfactory. Our results suggested that alteration of the abundance of specific taxa was associated with specific host gene expression. The modulation of gut microbiota might improve gut homeostasis especially in patients with non-viral-related HCC.
