Splicing factor HNRNPD and alternative splicing of MAP4K4 are associated with cell apoptosis and immune microenvironment features in Wilms’ tumor

Abstract

Background As the most common malignant renal tumor in children, the progression of Wilms’ tumor is frequently driven by abnormal alternative splicing (AS), cell death imbalance, and an immunosuppressive microenvironment. However, the precise regulatory chain connecting these three critical elements remains largely unexplored. This study aimed to systematically construct and characterize an “AS-cell death-immunity” regulatory network in Wilms’ tumor. Methods We performed a comprehensive in silico analysis using matched paired Wilms’ tumor and adjacent normal RNA-seq data from the GSE138869 cohort. The SUVA algorithm was employed to identify cell death-related regulated alternative splicing events (RASEs). A tripartite regulatory network was constructed via correlation analysis to link these RASEs with upstream differentially expressed splicing factors (DESFs). Immune cell infiltration was quantified using CIBERSORT. Finally, the HNRNPD knockout and FLASH-seq multi-omics dataset (GSE212767) was utilized to computationally validate the predicted regulatory axis. Results Our analysis identified 118 cell death-related host genes undergoing significant alternative splicing in Wilms’ tumor. Network integration highlighted a critical regulatory axis where the overexpressed splicing factor HNRNPD is strongly correlated with an aberrant AS event (clualt5p51764) in the apoptosis-related kinase MAP4K4. Further immune deconvolution demonstrated that both HNRNPD upregulation and the MAP4K4 splicing shift were significantly correlated with increased monocyte infiltration in the tumor microenvironment. Moreover, cross-validation utilizing the GSE212767 dataset confirmed that HNRNPD perturbation directly alters MAP4K4 splicing. Conclusions Our computational framework proposes that the HNRNPD-MAP4K4 splicing axis links apoptotic dysregulation to immune microenvironment remodeling in Wilms’ tumor. These correlative in silico findings provide a robust, hypothesis-generating basis for discovering novel prognostic biomarkers and developing targeted therapeutic strategies directed at the splicing machinery.