Integrative Analysis and Validation of a Cancer-associated Fibroblasts Senescence-related Signature for Risk Stratification and Therapeutic Prediction in Esophageal Squamous Cell Carcinoma.

Abstract

Cellular senescence is closely associated with cancer development and progression. There is ample evidence that tumor stromal cells, especially cancer-associated fibroblasts (CAFs) undergo senescence in response to various stimuli. However, the possible biological roles and prognostic significance of senescent CAFs in esophageal squamous cell carcinoma (ESCC) remain unexplored. In this study, we found that CAFs exhibited a significantly higher level of cellular senescence than other cell clusters at the single-cell level. Then, we constructed a CAFs senescence-associated risk model with 7 genes (GEM, SLC2A6, CXCL14, STX11, EFHD2, PTX3, and HCK) through Cox regression and LASSO analysis. Kaplan-Meier survival analysis revealed that the risk model was significantly correlated with worse prognosis in training and validation cohorts. Subsequent analysis indicated that the risk model was an independent prognostic factor. In addition, the signature showed a distinct negative correlation with immune cell infiltration and immunotherapy responses. In vitro experiments showed remarkably higher mRNA and protein levels of prognosis-related genes (STX11 and EFHD2) in senescent CAFs than control group, consistent with the bioinformatics analysis results. Moreover, senescent CAFs significantly promoted ESCC cell proliferation and migration as shown by CCK-8 and scratch assays. In conclusion, our study identified a novel CAFs senescence-based classifier that may help predict prognosis of ESCC, and a thorough characterization of the signature could also be helpful in evaluating the response of ESCC to anti-tumor therapies and provide meaningful clinical options for cancer treatment.