@article{JGO116936,
author = {Shiyang Zhan and Shurong Huang and Yangqiang Wang and Wei Zheng and Zongda Cai and Jinping Chen},
title = {Single-cell transcriptomics reveals the prognostic and immune infiltration significance of FOLFOX-bevacizumab treatment and lysine crotonylation characteristics in colon cancer},
journal = {Journal of Gastrointestinal Oncology},
volume = {17},
number = {3},
year = {2026},
keywords = {},
abstract = {Background: Colon cancer (CC), a malignant tumor originating from the colonic mucosal epithelium, frequently exhibits limited efficacy in advanced patients undergoing first-line FOLFOX (oxaliplatin, leucovorin, and 5-fluorouracil)-bevacizumab (FOLFOX-Bev) therapy due to resistance. Lysine crotonylation (Kcr), a novel histone post-translational modification, regulates gene transcription and participates in processes such as cell proliferation, metabolic reprogramming, and epithelial-mesenchymal transition, presenting potential research value in CC. However, its impact on FOLFOX-Bev treatment response and immune microenvironment remodeling at the single-cell level remains unclear. Therefore, this study aimed to integrate single-cell transcriptomics with The Cancer Genome Atlas (TCGA) data to systematically characterize the role of Kcr modification in FOLFOX-Bev treatment response and immune microenvironment remodeling in CC, and to construct a Kcr-based prognostic risk model for guiding precision therapy.Methods: This study integrated single-cell transcriptome with TCGA CC data [colon adenocarcinoma (COAD)]. Malignant cells were identified using Inference of Copy Number Variations and were stratified into high/low groups based on area under the curve cell-level enrichment analysis scores calculated for a Kcr-related gene set. Differential analysis and cell communication studies were performed. Sensitive/resistant malignant subpopulations were distinguished based on pre-/post-treatment changes, and their Kcr scores were compared to screen differentially expressed genes (DEGs) between “sensitive & high- Kcr” vs. “drug-resistant & low-Kcr” cells. In TCGA-COAD, Kcr single-sample Gene Set Enrichment Analysis (ssGSEA) scores, weighted gene co-expression network analysis modules, and single-cell DEGs were intersected. Utilizing the intersecting genes, a prognostic risk model was constructed via univariate-Least Absolute Shrinkage and Selection Operator (LASSO)-multivariate Cox regression. Immune infiltration, immunotherapy response, mutational landscape, and pathway enrichment features were assessed between high-/low-risk groups.Results: Focusing on “sensitive & high-Kcr” vs. “drug-resistant & low-Kcr” cells, 387 key genes were identified. Univariate-LASSO-multivariate Cox regression identified four core biomarkers (USP53, UACA, C7orf50, and IDH2) to form the risk prognostic model, which demonstrated good predictive performance in two independent datasets. The two risk groups exhibited remarkably different immune characteristics: the high-risk group showed poorer predicted immunotherapy response and lower mutation rates. Six categories of differential candidate drugs were screened (e.g., Erlotinib_1168, AZD3759_1915, Dasatinib_1079, and Sepantronium.bromide_1941).Conclusions: This study is the first to reveal, at single-cell resolution, that Kcr modification drives FOLFOX-Bev resistance in CC by reshaping the immune microenvironment. The prognostic model, based on key genes (USP53/UACA/C7orf50/IDH2), accurately stratifies patient risk. This model provides novel insights for targeted therapy and immune-combination strategies in CC.},
issn = {2219-679X}, url = {https://jgo.amegroups.org/article/view/116936}
}