Identification of splicing factors signature predicting prognosis risk and the mechanistic roles of novel oncogenes in HNSCC

Head and neck squamous cell carcinoma (HNSCC) is the most frequent subtype of head and neck cancer, generally with a poor prognosis and limited therapeutic options due to its highly heterogeneous malignancy. In this study, we screened functional splicing regulatory RNA binding proteins (RBPs) that were closely related with the prognosis of HNSCC patients and showed significant expression differences between HNSCC tumors and normal tissues. Based on this finding, we chose six candidate genes (HNRNPC, ZCRB1, RBM12B, SF3A2, SF3B3, and SRSF11) to generate a prognostic prediction model and validated the accuracy of the prognostic model for predicting patient survival outcomes. We found that the risk score predicted by our model can serve as an independent prognostic predictor. Notably, HNSCC tumors showing higher expression of SF3B3, HNRNPC, or ZCRB1 possessed higher risk scores in the discovered prediction model. The investigation of the underlying mechanism validated that knockdown of SF3B3, HNRNPC, and ZCRB1 separately induced a substantial impairment of HNSCC cell survival. Conversely, overexpression of each of the three genes promoted tumor cellular proliferation. High throughput RNA sequencing analysis revealed that changes in the expression of SF3B3 and HNRNPC remarkably affected alternative splicing of genes related to cell cycle regulation, whereas the depletion of ZCRB1 contributed to aberrant splicing events involving in DNA damage response. In addition, the prognostic prediction model's risk score was demonstrated to be related with the immune infiltration score. Particularly, SF3B3 has a negative correlation with CD8A expression. Therefore, our findings provide promising prognosis predictors and potential therapeutic targets for better treatment efficacy of HNSCC.