背景:TP53突变和缺氧在癌症进展中起重要作用。然而,代谢重编程和由它们介导的肿瘤微环境(TME)异质性仍未完全了解。
方法:获得了32种癌症类型和免疫治疗队列的多组学数据,以全面表征代谢重编程模式和跨癌症类型的TME,并探索免疫治疗候选物。融合多种机器学习方法,建立代谢重编程评估模型,包括套索回归,神经网络,弹性网络,和生存支持向量机(SVM)。进行药物基因组学分析和体外测定以鉴定潜在的治疗药物。
结果:首先,我们在肝细胞癌(HCC)中鉴定了代谢亚型A(缺氧-TP53突变亚型)和代谢亚型B(非缺氧-TP53野生型亚型),并显示代谢亚型A具有"免疫炎症"微环境.接下来,我们建立了代谢重编程的评估模型,与传统的预后指标相比更有效。然后,我们确定了一种潜在的靶向药物,替尼泊苷。最后,我们进行了泛癌症分析,以说明代谢重编程在癌症中的作用,并发现代谢改变(MA)评分与肿瘤突变负荷(TMB)呈正相关。新抗原负荷,以及跨癌症类型的同源重组缺陷(HRD)。同时,我们证明代谢重编程在膀胱癌中介导了潜在的免疫治疗敏感性微环境,并在免疫治疗队列中进行了验证.
结论:缺氧和TP53突变介导的代谢改变与TME调节和肿瘤进展相关。在这项研究中,我们分析了代谢改变在癌症中的作用,并提出了癌症预后和免疫治疗反应性的预测模型.我们还探索了一种潜在的治疗药物,替尼泊苷。
TP53 mutation and hypoxia play an essential role in cancer progression. However, the metabolic reprogramming and tumor microenvironment (TME) heterogeneity mediated by them are still not fully understood.
The multi-omics data of 32 cancer types and immunotherapy cohorts were acquired to comprehensively characterize the metabolic reprogramming pattern and the TME across cancer types and explore immunotherapy candidates. An assessment model for metabolic reprogramming was established by integration of multiple machine learning methods, including lasso regression, neural network, elastic network, and survival support vector machine (SVM). Pharmacogenomics analysis and in vitro assay were conducted to identify potential therapeutic drugs.
First, we identified metabolic subtype A (hypoxia-TP53 mutation subtype) and metabolic subtype B (non-hypoxia-TP53 wildtype subtype) in hepatocellular carcinoma (HCC) and showed that metabolic subtype A had an \"immune inflamed\" microenvironment. Next, we established an assessment model for metabolic reprogramming, which was more effective compared to the traditional prognostic indicators. Then, we identified a potential targeting drug,
teniposide. Finally, we performed the pan-cancer analysis to illustrate the role of metabolic reprogramming in cancer and found that the metabolic alteration (MA) score was positively correlated with tumor mutational burden (TMB), neoantigen load, and homologous recombination deficiency (HRD) across cancer types. Meanwhile, we demonstrated that metabolic reprogramming mediated a potential immunotherapy-sensitive microenvironment in bladder cancer and validated it in an immunotherapy cohort.
Metabolic alteration mediated by hypoxia and TP53 mutation is associated with TME modulation and tumor progression across cancer types. In this study, we analyzed the role of metabolic alteration in cancer and propose a predictive model for cancer prognosis and immunotherapy responsiveness. We also explored a potential therapeutic drug,
teniposide.