SOD2- and NRF2-associated Gene Signature to Predict Radioresistance in Head and Neck Cancer

Abstract

Background: We propose a novel prognostic biomarker-based strategy for increasing the efficacy of radiotherapy (RT) in head and neck squamous cell carcinoma (HNSCC).

Materials and methods: We identified genes associated with superoxide dismutase 2 (SOD2) and nuclear factor erythroid-2-related factor 2 (NRF2) from gene-expression data of The Cancer Genome Atlas (TCGA) by calculating Pearson correlation. Patients were divided into two groups using hierarchical clustering. Colony-formation assay was performed to determine radioresistance in HNSCC cell line CAL27. Pathway analysis was conducted using The Database for Annotation, Visualization and Integrated Discovery (DAVID).

Results: We developed a 49-gene signature with SOD2- and NRF2-associated genes. Using mRNA expression data for the 49-gene signature, we performed hierarchical clustering to stratify patients into two subtypes, subtype A and B. In the TCGA cohort, subgroup A demonstrated a better prognosis than subgroup B in patients who received RT. The signature robustness was evaluated in other independent cohorts. We showed through colony-formation assay that depletion of SOD2 or NRF2 leads to increased radiosensitivity.

Conclusion: We identified and validated a robust gene signature of SOD2- and NRF2-associated genes in HNSCC and confirmed their link to radioresistance using in vitro assay, providing a novel biomarker for the evaluation of HNSCC prognosis.

Keywords: NRF2; SOD2; head and neck cancer; prognosis predictor; radioresistance.