Identification of Disulfidptosis-Related Genes in Ischemic Stroke by Combining Single-Cell Sequencing, Machine Learning Algorithms, and In Vitro Experiments

Abstract

Background

Ischemic stroke (IS) is a severe neurological disorder with a pathogenesis that remains incompletely understood. Recently, a novel form of cell death known as disulfidptosis has garnered significant attention in the field of ischemic stroke research. This study aims to investigate the mechanistic roles of disulfidptosis-related genes (DRGs) in the context of IS and to examine their correlation with immunopathological features.

Methods

To enhance our understanding of the mechanistic underpinnings of disulfidptosis in IS, we initially retrieved the expression profile of peripheral blood from human IS patients from the GEO database. We then utilized a suite of machine learning algorithms, including LASSO, random forest, and SVM-RFE, to identify and validate pivotal genes. Furthermore, we developed a predictive nomogram model, integrating multifactorial logistic regression analysis and calibration curves, to evaluate the risk of IS. For the analysis of single-cell sequencing data, we employed a range of analytical tools, such as "Monocle" and "CellChat," to assess the status of immune cell infiltration and to characterize intercellular communication networks. Additionally, we utilized an oxygen–glucose deprivation (OGD) model to investigate the effects of SLC7A11 overexpression on microglial polarization.

Results

This study successfully identified key genes associated with disulfidptosis and developed a reliable nomogram model using machine learning algorithms to predict the risk of ischemic stroke. Examination of single-cell sequencing data showed a robust correlation between disulfidptosis levels and the infiltration of immune cells. Furthermore, "CellChat" analysis elucidated the intricate characteristics of intercellular communication networks. Notably, the TNF signaling pathway was found to be intimately linked with the disulfidptosis signature in ischemic stroke. In an intriguing finding, the OGD model demonstrated that SLC7A11 expression suppresses M1 polarization while promoting M2 polarization in microglia.

Conclusion

The significance of our findings lies in their potential to shed light on the pathogenesis of ischemic stroke, particularly by underscoring the pivotal role of disulfidptosis-related genes (DRGs). These insights could pave the way for novel therapeutic strategies targeting DRGs to mitigate the impact of ischemic stroke.