Novel Disulfidptosis-associated Genetic Signatures to Predict Prognosis in Patients with Hepatocellular Carcinoma

Abstract

Abstract Background Hepatocellular carcinoma (HCC) poses significant challenges to prognosis prediction due to its heterogeneity and high recurrence rate. Disulfidoptosis, a unique form of cell death dependent on disulfide aggregation in cells overexpressing SLC7A11 under glucose starvation, distinguishes itself from known programmed cell death. However, the prognostic implications of disulfidoptosis-related genes in HCC require further elucidation. Methods From public databases, we gathered mRNA expression profiles and corresponding clinical data on HCC patients. Utilizing the least absolute shrinkage and selection operator (LASSO) Cox regression model, a four-gene signature was constructed in the TCGA cohort. Validation was performed from the ICGC and GSE14520 cohorts. According to the risk score, TIMER algorithm was used to analyze the infiltration of immune cells in the microenvironment of HCC. Predicted the sorafenib-therapeutic response was conducted based on the Genomics of Drug Sensitivity in Cancer (GDSC). Gene Ontology (GO) and Kyoto Encyclopedia of Genes and Genomes (KEGG) analyses was performed to gain insights into the biological functions of the disulfidptosis model-associated genes. Results Seven known disulfidoptosis-related genes significantly correlated with overall survival (OS) according to univariate Cox regression analysis (P adj. < 0.05). High-risk patients demonstrated a significantly lower OS than low-risk patients (P = 0.002 in the TCGA and P = 0.004 in the ICGC cohort). The risk scores served as independent predictors of OS in both TCGA and ICGC cohorts, according to multivariate Cox regressions (HR > 1, P < 0.001). Gene signature prediction was further validated by receiver operating characteristic (ROC) analysis. Notably, immune cell infiltration and the sorafenib treatment response differed between the two groups. Functional analysis revealed enrichment of mitosis-related pathways. Conclusion our study proposes a novel disulfidoptosis-related gene signature with potential clinical utility for prognostic prediction in HCC. For HCC, targeting disulfidoptosis may be a promising therapeutic option.