[Bioinformatics analysis of programmed cell death ligand 1 co-expression genes and their regulatory network in head...

This study aimed to construct a network of programmed celldeath ligand 1 (PD-L1) co-expression genes and screen potential biomarkers for PD-L1 expression in head and neck squamous cell carcinoma (HNSCC) by bioinformatics analysis. Moreover, the genes and pathways participating in PD-L1 and regulating the tumor immune status were determined.

The HNSCC transcriptomic dataset in TCGA was selected to retrieve gene sets on the cBioPortal platform, where PD-L1 co-expressional genes were acquired. With these genes, GO-BP, KEGG, and string analyses were performed in R clusterProfiler. Cytoscape was used for network analysis and hub gene screening.

A total of 117 co-expression genes were obtained, most of which were enriched in immune regulation and response to viral processes. Node degree analysis indicated that STAT1, IFNG, CXCL10, CCR5, FCGR3A, CXCL9, GBP5, CD86, GZMB, IRF1 were the highest connected genes and functioned as hub genes. Survival analysis of these hub genes resulted in CCR5, CXCL9, and GZMB as the prognostic biomarkers for patients with HNSCC, all of which were involved in immune regulation and their expression levels were related to PD-L1 (Pearson correlation coefficient was 0.30, 0.35, 0.39; P<0.01). High expression levels of these three hub genes were protective factors in patients with HNSCC.

PD-L1 co-expression hub genes are related to immunity, among which CCR5, CXCL9, and GZMB are prognostic markers with the possibility to be involved in programmed cell death protein 1 (PD-1)/PD-L1-induced tumor immune escape. These genes provide new clues to study the mechanism and precision target medicine of PD-1/PD-L1 in HNSCC.