Machine learning-driven single-cell and bulk RNA sequencing integration unveils key genes modulating radiosensitivity and tumor progression in esophageal squamous cell carcinoma

Abstract

Radiotherapy resistance remains a significant clinical challenge in oesophageal squamous cell carcinoma (ESCC), often leading to treatment failure and tumour recurrence. Understanding the molecular mechanisms underlying radiotherapy resistance is crucial for improving patient outcomes. In this study, we explored the roles of MAP1LC3B and HSD17B10 in radiotherapy resistance and their potential as therapeutic targets. We employed single-cell RNA sequencing (scRNA-seq), bulk RNA-seq and machine learning to identify key genes and cellular mechanisms associated with response to radiotherapy in ESCC. The scAB algorithm was used to integrate the transcriptomic data and identify key genes associated with radiotherapy. Functional enrichment and immune infiltration analyses were performed, followed by machine learning algorithms (random forest and LASSO) to identify disease-specific signature genes. In vitro and in vivo experiments validated the biological functions of MAP1LC3B and HSD17B10 in radiotherapy resistance and metastasis in ESCC. Our study revealed significant cellular and molecular heterogeneity in ESCC, with distinct subpopulations linked to radiotherapy resistance. We identified two key genes (MAP1LC3B and HSD17B10) that play critical roles in regulating cell survival, immune evasion, and tumour metastasis under radiation stress. Notably, we observed altered immune cell infiltration patterns, suggesting that the tumour microenvironment (TME) may contribute to the observed resistance. Moreover, our findings revealed that MAP1LC3B and HSD17B10 are significantly overexpressed in ESCC tumour tissues and that their high expression are correlated with poor prognosis and radiotherapy resistance. This study provides new insights into the molecular mechanisms of radiotherapy resistance in ESCC. By identifying key genes, we highlight potential biomarkers for predicting treatment outcomes. These findings may inform future therapeutic strategies aimed at overcoming radiotherapy resistance and improving patient prognosis in ESCC.