非靶向代谢组学揭示宫颈癌发展进程的特征代谢谱研究  

Non-targeted metabolomic profiling reveals characteristic metabolic pro-file associated with development process of cervical cancer

作　　者：翟青枝[1] 马韵之 叶明侠[1] 王铭洋 李阳[3] 李莉[3] 孟元光[1] 李立安[1] ZHAI Qingzhi;MA Yunzhi;YE Mingxia;WANG Mingyang;LI Yang;LI Li;MENG Yuanguang;LI Lian(Department of Obstetrics and Gynecology,Seventh Medical Center,Chinese PLA General Hospital,Beijing 100005,Chi-na;School of Medicine,Nankai University,Tianjin 300071,China;Department of Obstetrics and Gynecology,First Medi-cal Center,Chinese PLA General Hospital,Beijing 100853,China)

机构地区：[1]解放军总医院第七医学中心妇产科,北京100005 [2]南开大学医学院,天津300071 [3]解放军总医院第一医学中心妇产科,北京100853

出　　处：《中国病理生理杂志》2025年第2期230-238,共9页Chinese Journal of Pathophysiology

基　　金：中国人民解放军计生专项基金资助项目(No.22JSZ16,No.24JSZ15);解放军总医院第七医学中心2023年度创新培育基金资助项目(No.qzx-2023-22)。

摘　　要：目的:探讨宫颈病变过程中的代谢谱差异及其用于辅助诊断宫颈癌的潜在临床价值。方法:采用超高效液相色谱串联高分辨率质谱分析技术,对43例宫颈癌患者,34例高级别鳞状上皮内病变患者和43例健康对照人群的宫颈拭子样本进行了非靶向代谢组学分析。在三组代谢谱均有明显特征的基础上,利用主成分分析法确定健康对照,高级别鳞状上皮内病变与宫颈癌之间的代谢差异,并使用Metabo Analyst 5.0对显著差异代谢物进行了KEGG通路富集分析。最后,通过随机森林机器学习构建区分宫颈癌与健康对照,高级别鳞状上皮内病变与健康对照和宫颈癌与高级别鳞状上皮内病变人群的分类预测模型,并通过ROC曲线评估模型效能。结果:健康对照、高级别鳞状上皮内病变和宫颈癌3组经过滤后共得到1543种代谢物,其中包含407个组间差异代谢物。研究发现内源代谢物PGE2同时存在于3组中,且其表达水平随宫颈病变的发展呈现逐步上升状态。差异代谢物富集分析显示宫颈癌具有特异的癌症相关代谢通路,包括三羧酸循环、酪氨酸代谢、色氨酸代谢、戊糖磷酸代谢等。研究基于代谢产物构建了3种可用于辅助诊断高级别鳞状上皮内病变和宫颈癌的预测模型,即全模型、简化模型和PGE2模型。结果表明代谢物具有良好的诊断效能,其中全模型及简化模型均可有效区分宫颈癌与高级别鳞状上皮内病变,宫颈癌与健康人群,以及高级别鳞状上皮内病变与健康人群。前者AUC值分别达到了0.90、0.92和0.84,后者AUC值分别达到了0.81、0.95和0.85。PGE2模型区分宫颈癌与健康对照,高级别鳞状上皮内病变与健康对照的AUC值分别是0.74和0.80。结论:宫颈癌的不同发展进程其代谢谱具有显著差异,且这些代谢物有成为宫颈病变生物标志物的潜在临床价值。AIM:The aim of our study is to investigate the metabolic profile differences during cervical lesion progression and evaluate their potential clinical value in assisting the diagnosis of cervical cancer(CC).METHODS:Ultra-high-performance liquid chromatography coupled with high-resolution mass spectrometry(UHPLC-HRMS)was employed to conduct non-targeted metabolomic analysis of cervical swab samples from 43 CC patients,34 high-grade squamous intraepithelial lesion(HSIL)patients,and 43 healthy controls.Based on the distinct features among the three groups,principal component analysis(PCA)was used to identify the metabolic differences among CC,HSIL and healthy groups.MetaboAnalyst 5.0 was then employed to perform KEGG pathway enrichment analysis on the differential metabo-lites.Finally,random forest machine learning algorithm was used to construct classification prediction models for distin-guishing CC from healthy,HSIL from healthy,and CC from HSIL.The performance of these models was evaluated using receiver operating characteristic(ROC)curve analysis.RESULTS:A total of 1543 metabolites were identified across the healthy,HSIL and CC groups after filtration,with 407 metabolites differing between the groups.The study found that metabolite PGE2 was present in all three groups,with its expression levels progressively increasing with the progression of cervical lesions.Differential metabolite enrichment analysis demonstrated that CC is associated with specific cancer-relat-ed metabolic pathways,including the tricarboxylic acid cycle,tyrosine metabolism,tryptophan metabolism,and the pen-tose phosphate pathways.Additionally,the study developed three prediction models based on metabolic products for diag-nosing HSIL and CC:the full model,the simplified model,and the PGE2 model.The results indicated that metabolites ex-hibited strong diagnostic efficiency.Both the full model and the simplified model effectively distinguished CC from HSIL,CC from healthy,and HSIL from healthy.The AUC values for the full model were 0.90,0.92 an

关 键 词：宫颈癌 高级别鳞状上皮内病变 宫颈拭子 代谢谱 非靶向代谢组学 

分 类 号：R737.33[医药卫生—肿瘤] R363[医药卫生—临床医学] R446