Comprehensive bioinformatics analysis of a RBM family-based prognostic signature with experiment validation in hepatocellular carcinoma.

Abstract

Although some RBM proteins family members play important roles in hepatocellular carcinoma (HCC) development, their value of prognosis and tumor treatment is not clear. To reveal the expression patterns and clinical significance of RBM family members in HCC, we constructed a RBM family-based prognosis signature. We collected the data of HCC patients from TCGA and ICGC database. The prognostic signature was constructed in TCGA and verified using ICGC cohort. Based on this model, risk score was calculated and patients were divided into high- and low-risk group. Comparison of immune cell infiltration, the response to immunotherapy, and IC50 of chemotherapeutic drugs were employed between different risk subgroups. Besides, CCK-8 and EdU assays were performed to investigate the role of RBM45 in HCC. Among 19 differential expression RBM protein family genes, 7 prognostic genes were picked out. Through LASSO Cox regression, a 4-gene prognostic model was successfully constructed, which included RBM8A, RBM19, RBM28 and RBM45. Results of validation and estimation suggested this model could be applied for prognostic prediction in HCC patients with a well predictive value. Risk score was shown to be an independent predictor and high-risk patients had poor prognosis. High-risk patients had an immunosuppressive tumor microenvironment while patients with low risk could benefit more from ICI therapy and sorafenib treatment. In addition, knockdown of RBM45 inhibited the proliferation of HCC. This prognostic signature based on RBM family had a great value for predicting OS of HCC patients. Low-risk patients were more suitable for receiving immunotherapy and sorafenib treatment. The RBM family members made of the prognostic model might promote the progression of HCC.