Unveiling the predictive power of bacterial response-related genes signature in hepatocellular carcinoma: with bioinformatics analyses and experimental approaches

Background: Despite progress in therapeutic strategies, treatment failure in hepatocellular carcinoma (HCC) remains a major challenge, resulting in low survival rates. The presence of bacteria and the host ' s immune response to bacteria can in fl uence the pathogenesis and progression of HCC. We developed a risk model based on bacterial response-related genes (BRGs) using gene sets from molecular signature databases to identify new markers for predicting HCC outcomes and categorizing patients into different risk groups. Methods: The data from the Cancer Genome Atlas (TCGA) portal was retrieved, and differentially expressed BRGs were identi fi ed. Uni- and multivariate Cox regression and least absolute shrinkage and selection operator (LASSO) LASSO analyses were executed to develop the prognostic risk model. Key contributor to the prognostic model was identi fi ed, and the results were tested by using experimental assays in HCC cell lines. Results: Multivariate analysis demonstrated an independent prognostic factor of 12-BRG signature in HCC patients. The low-risk group had better overall survival with signi fi cantly lower tumor mutation burden (TMB). The risk scores were negatively correlated with the presence of tumor-in fi ltrating immune cells. In an effort to fi nd the key contributor of the 12-BRG signature, we found polo like kinase1 (PLK1) had the best accuracy with 1-, 3-, and 5-year AUC of 0.72, 0.66, and 0.65, respectively. Both PLK1 inhibitor Volasertib and the knockdown of the PLK1 gene resulted in diminished viability in HCC cell lines. The combination of PLK1 inhibition with low-dose chemotherapy exhibited an ampli fi ed effect of the treatment. Conclusion: To date, there have been no reports of BRG biomarkers in HCC, and this study represents for the fi rst time that a 12-BRG signature has the potential to predict the survival of HCC.