Background
Transforming growth factor-beta (TGF-β) signaling is essential in initialization and progression of hepatocellular carcinoma (HCC). Therefore, a treatment targeting TGF-β pathway may be a promising option for HCC control.
Conclusion
We established a novel prognostic prediction method to predict the risk scores of TGF-β genes in HCC prognosis. TGF-β1 is highly expressed in HCC cell lines and tissues, correlates to poor prognosis, and thus can be used as a potential biomarker to predict HCC prognosis. We showed that TGF-β1 may play its roles in HCC prognosis by modulating the immune microenvironment of tumor cells. Our data may shed more light on better understanding the role of TGF-β1 in HCC prognosis.
Methods
First, publicly available RNA-seq datasets and clinical characteristics of 374 HCC patients in The Cancer Genome Atlas (TCGA) database were downloaded. Then, Cox regression analysis and LASSO analysis were used to construct a prognostic model for TGF-β family genes. The area under the curve (AUC) of the risk signature was calculated to evaluate the predictive power of the model. Cox regression analysis was applied to predict whether TGF-β1 can be an independent prognosis factor for HCC. Next, hazard ratio and survival analyses were performed to investigate the correlation between TGF-β1 expression and survival time. Furthermore, differential expression level of TGF-β1 in HCC tissues and cells was determined. In addition, Gene Set Enrichment Analysis (GSEA) identified the top significantly activated and inhibited signal pathways related to high expression of TGF-β1. Finally, the CIBERSORT tool was adopted to correlate the tumor-infiltrating immune cells (TICs) with TGF-β1 expression in HCC cohorts.
Results
Cox regression analysis and LASSO analysis revealed that seven TGF-β family members (including TGF-β1) could be used as prognostic factors for HCC. Interestingly, TGF-β1 was demonstrated to be an independent prognostic factor of HCC. RT-qPCR and immunofluorescence staining confirmed the high expression of TGF-β1 in HCC cell lines and tissues, which is significantly related to pathological classifications, poor prognosis, and short survival time. Finally, GSEA and CIBERSORT analyses suggested that TGF-β1 may interact with various immune cells and influence the prognosis of HCC patients through Tregs and γδ T cells.