背景：结肠癌是世界范围内一种流行且致命的恶性肿瘤，是癌症相关死亡的第三大原因。二硫下垂应激会引发一种独特形式的程序性细胞死亡，称为二硫下垂，其特征是细胞内胱氨酸过度积累。本研究旨在基于与二硫下垂诱导的细胞死亡相关的长非编码 RNA (LncRNA) 建立可靠的生物指标，为结肠腺癌 (COAD) 患者的免疫治疗反应和预后评估提供新的见解。方法：进行单变量Cox比例风险分析和Lasso回归分析，以确定与预后密切相关的差异表达基因。随后，使用多重 Cox 比例风险回归开发了用于预后风险评估的多因素模型。此外，考虑临床病理特征、肿瘤微环境和化疗敏感性，我们对二硫下垂反应相关的LncRNA的特征进行了综合评估。使用定量实时荧光 PCR (qRT-PCR) 验证 COAD 患者预后相关基因的表达水平。此外，通过CCK8测定、伤口愈合实验和transwell实验研究了ZEB1-SA1在结肠癌中的作用。结果：二硫下垂反应相关的 LncRNA 被确定为 COAD 预后的有力预测因子。多因素分析显示，源自这些 LncRNA 的风险评分是 COAD 的独立预后因素。与高风险组患者相比，低风险组患者表现出更高的总生存期（OS）。因此，我们开发的列线图预测模型结合了临床特征和风险评分，表现出了出色的预后功效。体外实验表明ZEB1-SA1促进COAD细胞的增殖和迁移。结论：利用医学大数据和人工智能，我们基于TCGA-COAD队列构建了二硫下垂反应相关LncRNA的预测模型，能够准确预测结肠癌患者的预后。该模型在临床实践中的实施可以促进 COAD 患者的精确分类，识别更有可能对免疫治疗和化疗产生良好反应的特定亚组，并为基于科学证据的 COAD 患者个性化治疗策略的制定提供信息。版权所有 © 2023 Chi 、黄、严、江、张、陈、江、张、张、杨、田。

Background: Colon cancer, a prevalent and deadly malignancy worldwide, ranks as the third leading cause of cancer-related mortality. Disulfidptosis stress triggers a unique form of programmed cell death known as disulfidoptosis, characterized by excessive intracellular cystine accumulation. This study aimed to establish reliable bioindicators based on long non-coding RNAs (LncRNAs) associated with disulfidptosis-induced cell death, providing novel insights into immunotherapeutic response and prognostic assessment in patients with colon adenocarcinoma (COAD). Methods: Univariate Cox proportional hazard analysis and Lasso regression analysis were performed to identify differentially expressed genes strongly associated with prognosis. Subsequently, a multifactorial model for prognostic risk assessment was developed using multiple Cox proportional hazard regression. Furthermore, we conducted comprehensive evaluations of the characteristics of disulfidptosis response-related LncRNAs, considering clinicopathological features, tumor microenvironment, and chemotherapy sensitivity. The expression levels of prognosis-related genes in COAD patients were validated using quantitative real-time fluorescence PCR (qRT-PCR). Additionally, the role of ZEB1-SA1 in colon cancer was investigated through CCK8 assays, wound healing experiment and transwell experiments. Results: disulfidptosis response-related LncRNAs were identified as robust predictors of COAD prognosis. Multifactorial analysis revealed that the risk score derived from these LncRNAs served as an independent prognostic factor for COAD. Patients in the low-risk group exhibited superior overall survival (OS) compared to those in the high-risk group. Accordingly, our developed Nomogram prediction model, integrating clinical characteristics and risk scores, demonstrated excellent prognostic efficacy. In vitro experiments demonstrated that ZEB1-SA1 promoted the proliferation and migration of COAD cells. Conclusion: Leveraging medical big data and artificial intelligence, we constructed a prediction model for disulfidptosis response-related LncRNAs based on the TCGA-COAD cohort, enabling accurate prognostic prediction in colon cancer patients. The implementation of this model in clinical practice can facilitate precise classification of COAD patients, identification of specific subgroups more likely to respond favorably to immunotherapy and chemotherapy, and inform the development of personalized treatment strategies for COAD patients based on scientific evidence.Copyright © 2023 Chi, Huang, Yan, Jiang, Zhang, Chen, Jiang, Zhang, Zhang, Yang and Tian.