Toward precision oncology in LUAD: a prognostic model using single-cell sequencing and WGCNA based on a disulfidptosis relative gene signature.

BACKGROUND

Disulfidptosis, a recently identified mechanism of cell death characterized by intracellular sulfide accumulation, leading to cellular exhaustion. Our objective is to create a prognostic model using a cohort of disulfidptosis-related genes (DRGs) to assess their prognostic value in lung adenocarcinoma (LUAD). This research not only deepens our understanding of the molecular mechanisms underpinning LUAD but also offers promising avenues for new clinical treatment biomarkers and therapeutic targets.

METHODS

We employed various methodologies to assess DRGs in LUAD. Gene expression in single cell RNA sequencing (scRNA-seq) data was assessed using the AUcell algorithm. In the TCGA [LUAD] dataset, disulfidptosis-related enrichment scores were calculated using ssGSEA, and core gene sets were identified through the Weighted Gene Co-expression Network Analysis (WGCNA) algorithm. Differential gene analysis was conducted using the limma package and intersected with core gene sets. Univariate Cox regression analysis revealed genes with significant effects on LUAD prognosis. A prognostic model was developed using LASSO and Cox regression, utilizing median model scores for stratifying patient risk. Kaplan-Meier curves assessed prognostic differences between risk groups. Comprehensive analyses were performed on the tumor microenvironment (TME) and mutational landscape across different risk groups. Immune response characteristics and functional enrichment patterns were further evaluated in these cohorts.

RESULTS

Our study delved into disulfidptosis in LUAD through a series of analyses: scRNA-seq data processing, WGCNA analysis, construction of a prognostic model, evaluation of clinical features and risk, enrichment analysis, mutation landscape assessment, and examination of the tumor microenvironment. We identified core genes related to disulfidptosis and established a prognostic model to classify patients based on risk scores. Notable differences in TME characteristics, immune cell infiltration, mutation landscape, and biological pathway activities were observed between risk groups, shedding new light on LUAD clinical treatment and biomarker discovery. Cell experiments highlighted the significance of KCNK1 in LUAD cells, suggesting its potential as a therapeutic target.

CONCLUSION

A prognostic model centered on DRGs was effectively developed to predict prognosis of LUAD and immunotherapy response. Our initial investigations unveiled KCNK1's oncogenic role in LUAD, identifying it as a potential therapeutic target.