A novel ubiquitin-related genes-based signature demonstrated values in prognostic prediction, immune landscape sculpture and therapeutic options in laryngeal cancer.

Abstract

Background: Laryngeal cancer (LC) is characterized by high mortality and remains challenging in prognostic evaluation and treatment benefits. Ubiquitin-related genes (UbRGs) are widely involved in cancer initiation and progression, but their potential value in LC is unknown.

Methods: RNA-seq and clinical data of LC were obtained from TCGA and GEO. UbRGs that independently influenced the overall survival (OS) of LC patients were screened with differential expression, COX and LASSO regression analyses. A prognostic signature was then established and assessed for its predictive value, stability and applicability using Kaplan-Meier analysis and receiver operating characteristic curves. The nomogram was further generated in combination with the signature and clinical characteristics. Characterization of immune properties and prediction of drug sensitivity were investigated on the signature-based subgroups using a panel of in silico platforms. Verification of gene expression was conducted with Western blot, qRT-PCR and ELISA, ultimately.

Results: PPARG, LCK and LHX1 were identified and employed to construct the UbRGs-based prognostic signature, showing a strong ability to discriminate LC patients with distinct OS in TCGA-LC and GSE65858, and excellent applicability in most clinical conditions. The nomogram showed higher predictive value and net clinical benefit than traditional indicators. As evaluated, the low-risk group had a more activated immune function, higher infiltration of anti-cancer immune cells and stronger expression of immune-promoting cytokines than the high-risk group. Immune properties were also correlated with individual signature genes. PPARG and LHX1 were negatively correlated, whereas LCK positively correlated, with the immuno-promoting microenvironment. Additionally, chemotherapy would be more effective in high-risk patients, while immune checkpoint inhibitors would be more effective in low-risk patients. Finally, dysregulation of the signature genes was confirmed in LC cell lines by Western blot, and PPARG knockdown significantly reduced the expression of the immunosuppressive cytokines IL6, TGFB1, TGFB2 and VEGFC by qRT-PCR and ELISA.

Conclusion: We have developed a UbRGs-based signature for LC prognostic evaluation that is valuable in clinical application, indicative of the immune microenvironment and beneficial for individualized treatment guidance.