| 1,139 | 0 | 342 |
| 下载次数 | 被引频次 | 阅读次数 |
本研究旨在整合网络毒理学、多种机器学习算法与分子对接技术，系统性筛选并验证与马兜铃酸(aristolochic acids, AAs)暴露相关的肝细胞癌(hepatocellular carcinoma, HCC)核心生物标志物，并揭示其潜在的致癌分子机制。整合多个GEO公共数据集，经批次校正后筛选HCC的差异表达基因(differentially expressed genes, DEGs)。通过ChEMBL、 SwissTargetPrediction及PharmMapper 3个互补数据库预测AAs潜在靶点。应用12种特征选择算法(包括Lasso、 Ridge、 Enet、 Stepglm、 SVM、 glmBoost、 LDA、 plsRglm、 RF、 GBM、 XGBoost和NaiveBayes)构建了113种算法组合，用于筛选最优模型，并对关键靶基因进行特征筛选，鉴定出最优诊断效能的核心基因组合。采用SHAP(shapley additive explanations)分析模型可解释性，并构建人工神经网络(artificial neural network, ANN)模型以验证核心标志物的分类效能。进一步通过CIBERSORT和ssGSEA算法分析核心基因与免疫微环境的关联。最后，采用分子对接技术验证靶点结合能。研究共获得68个AAs与HCC的共同靶基因，并对关键靶点进行基因本体论(gene ontology, GO)功能、基因组百科全书(kyoto encyclopedia of genes and genomes, KEGG)通路富集分析。最优机器学习模型筛选出8个核心基因RND3、HMMR、LY6E、CCNA2、PCK1、PSPH、KIF11、TK1,其在HCC诊断中表现出较强潜力(AUC>0.85)。SHAP分析表明，RND3是模型中最重要的预测特征(SHAP value=0.164)。基于该基因组合构建的ANN模型展现出极高的分类效能(AUC=0.990, 95%CI:0.984～0.995)。免疫浸润分析表明，这些基因在调控免疫微环境中呈现两种相反的功能模式：部分基因(如HMMR、CCNA2)与免疫激活状态正相关，而另一部分(如RND3、PCK1)则与免疫静息或抑制状态正相关。分子对接结果证实了6个核心蛋白与AAs具有良好的结合活性。本研究成功鉴定出以RND3为代表的8个核心基因，可作为诊断AAs相关HCC的潜在工具。同时揭示了AAs可能通过直接结合关键蛋白，进而调控细胞代谢重编程和免疫微环境来诱导HCC的潜在分子机制。这些发现为阐明AAs的肝脏致癌机制提供了具体的分子靶点，并为HCC的早期诊断与精准治疗提供了潜在的生物标志物。后期相关研究应集中于对这些核心靶点进行体外与体内实验功能验证。
Abstract：This study aims to integrate network toxicology, multiple machine learning algorithms, and molecular docking techniques to systematically screen and validate core biomarkers associated with hepatocellular carcinoma(HCC) induced by aristolochic acids(AAs) exposure, and to reveal their potential carcinogenic molecular mechanisms. Multiple GEO public datasets were integrated, and differentially expressed genes(DEGs) associated with HCC were screened after batch effect correction. Potential targets of AAs were predicted using three complementary databases: ChEMBL, SwissTargetPrediction, and PharmMapper. Twelve feature selection algorithms(including Lasso, Ridge, Enet, Stepglm, SVM, glmBoost, LDA, plsRglm, RF, GBM, XGBoost, and NaiveBayes) were used to construct 113 algorithm combinations for selecting the optimal model and performing feature selection on key target genes, identifying the core gene combination with the best diagnostic performance. SHAP(shapley additive explanations) was used for model interpre-tability analysis, and an artificial neural network(ANN) model was constructed to validate the classification efficacy of the core biomarkers. Furthermore, the relationship between the core genes and the immune microenvironment was analyzed using CIBERSORT and ssGSEA algorithms. Finally, molecular docking was employed to verify target binding affinity. A total of 68 common target genes for AAs and HCC were identified, and gene ontology(GO) functional and kyoto encyclopedia of genes and genomes(KEGG) pathway enrichment analyses were performed on the key targets. The optimal machine learning model selected eight core genes: RND3, HMMR, LY6E, CCNA2, PCK1, PSPH, KIF11, and TK1, which showed strong potential for HCC diagnosis(AUC>0.85). SHAP analysis indicated that RND3 was the most important predictive feature(SHAP value=0.164). The ANN model based on this gene combination demonstrated excellent classification performance(AUC=0.990, 95% CI: 0.984～0.995). Immune infiltration analysis revealed that these genes exhibited two opposing functional patterns in regulating the immune microenvironment: some genes(e.g., HMMR, CCNA2) were positively correlated with immune activation, while others(e.g., RND3, PCK1) were positively correlated with immune resting or suppressive states. Molecular docking results confirmed that six core proteins exhibited strong binding activity with AAs. This study successfully identified eight core genes, represented by RND3, as potential diagnostic tools for AAs-related HCC. It also revealed that AAs may induce HCC by directly binding to key proteins, thereby regulating cellular metabolic reprogramming and immune microenvironment alterations. These findings provide specific molecular targets for understanding the hepatocarcinogenic mechanisms of AAs and offer potential biomarkers for early diagnosis and precision treatment of HCC. Future research should focus on the functional validation of these core targets in vitro and in vivo.
陈孟兰，朱正兰，2007.马兜铃属植物的生物活性和药理作用研究进展.时珍国医国药，18(3):702-704.[CHEN M L,ZHU Z L,2007.The advance on the studies of pharmaco-logy of Aristolochia L.Plants.Lishizhen Medicine and Materia Medica Research,18(3):702-704.]
米士丽，2021.9种中药材中马兜铃酸类毒性物质的定性与定量研究，硕士学位论文.南京：南京中医药大学.(MI S L,2021.Qualitative and quantitative study on aristolochic acid toxic substances in 9 Chinese herbal medicines,Thesis for M.S.Nanjing:Nanjing University of Chinese Medicine.)
王芳，2023.马兜铃水提物致大鼠肝肾毒性研究及其体内马兜铃酸类代谢产物的定性分析，硕士学位论文.北京：中国中医科学院.(WANG F,2023.Study on hepatorenal toxicity induced by water extract of Aristolochia in rats and qualitative analysis of aristolochic acid metabolites in vivo,Thesis for M.S.Beijing:China Academy of Chinese Medical Sciences.)
王馨羚，肖琼，付昱，等，2023.CCNA2在泛癌中的表达及其与预后和免疫浸润的关系.中国医科大学学报，52(9):791-797.[WANG X L,XIAO Q,FU Y,et al.,2023.Expression of CCNA2 in pan-cancer and its association with prognosis and immune infiltration.Journal of China Medical University,52(9):791-797.]
贠凯祎，徐志超，宋经元，2019.含马兜铃酸中药及其检测研究进展.中国科学：生命科学，49(3):238-249.[YUN K Y,XU Z C,SONG J Y,2019.Traditional Chinese medicine containing aristolochic acids and their detection.Scientia Sinica(Vitae),49(3):238-249.]
AKINYEMIJU T,ABERA S,AHMED M,et al.,2017.The burden of primary liver cancer and underlying etiologies from 1990 to 2015 at the global,regional,and national level:results from the global burden of disease study 2015.JAMA Oncol.,3(12):1683-1691.
BASBOUS S,DIF L,DANTZER C,et al.,2024.Loss of RND3/RHOE controls entosis through LAMP1 expression in hepatocellular carcinoma.Cell Death Dis.,15(1):46.
CAI Q,ZHU M Y,DUAN J N,et al.,2022.Comprehensive analysis of immune-related prognosis of TK1 in hepatocellular carcinoma.Front.Oncol.,11:786873.
CHEN C J,YANG Y H,LIN M H,et al.,2018.Herbal medicine containing aristolochic acid and the risk of hepatocellular carcinoma in patients with hepatitis B virus infection.Int.J.Cancer,143(7):1578-1587.
CHEN Y,YANG Z K,LIU Y,et al.,2024.Exploring the prognostic impact of triglyceride-glucose index in critically ill patients with first-ever stroke:insights from traditional methods and machine learning-based mortality prediction.Cardiovasc.Diabetol.,23(1):443.
CHEN Y L,ERIKSSON S,CHANG Z F,2010.Regulation and functional contribution of thymidine kinase 1 in repair of DNA damage.J.Biol.Chem.,285(35):27327-27335.
FARASATI FAR B,RABIE D,HEMATI P,et al.,2023.Unresectable hepatocellular carcinoma:a review of new advances with focus on targeted therapy and immunotherapy.Livers,3(1):121-160.
FENG T,WEI D C,LI Q K,et al.,2021.Four novel prognostic genes related to prostate cancer identified using co-expression structure network analysis.Front.Genet.,12:584164.
GAO W T,LU J J,YANG Z T,et al.,2024.Mitotic functions and characters of KIF11 in cancers.Biomolecules,14(4):386.
HAN J Y,XIAN Z,ZHANG Y S,et al.,2019.Systematic overview of aristolochic acids:nephrotoxicity,carcinogenicity,and underlying mechanisms.Front.Pharmacol.,10:648.
HAYASHI N,KOLLER E,FAZLI L,et al.,2008.Effects of Eg5 knockdown on human prostate cancer xenograft growth and chemosensitivity.Prostate,68(12):1283-1295.
HU Z D,JIANG Y,SUN H M,et al.,2021.KIF11 promotes proliferation of hepatocellular carcinoma among patients with liver cancers.BioMed Res.Int.,2021:2676745.
JI Y,GAO L,ZHANG C,et al.,2021.Identification of the hub genes and prognostic indicators of gastric cancer and correlation of indicators with tumor-infiltrating immune cell levels.J.Cancer,12(13):4025-4038.
JIANG A,ZHOU Y,GONG W,et al.,2022.CCNA2 as an immunological biomarker encompassing tumor microenvironment and therapeutic response in multiple cancer types.Oxid.Med.Cell Longev.,2022:5910575.
JIANG M,ZHUANG H R,XIA R,et al.,2017.KIF11 is required for proliferation and self-renewal of docetaxel resistant triple negative breast cancer cells.Oncotarget,8(54):92106-92118.
LI C X,LI X Y,NIU M,et al.,2024.Unveiling correlations between aristolochic acids and liver cancer:spatiotemporal heterogeneity phenomenon.Chin.Med.,19(1):132.
LI L Q,ZHU M J,HUANG H,et al.,2020a.Identification of hub genes in anaplastic thyroid carcinoma:evidence from bioinformatics analysis.Technol.Cancer Res.Treat.,19:1533033820962135.
LI X L,GUO X Q,WANG H R,et al.,2020b.Aristolochic acid-induced genotoxicity and toxicogenomic changes in rodents.World J.Tradit.Chin.Med.,6(1):12-25.
LI Z L,LIN Y,CHENG B Z,et al.,2021.Identification and analysis of potential key genes associated with hepatocellular carcinoma based on integrated bioinformatics methods.Front.Genet.,12:571231.
LU Z N,LUO Q,ZHAO L N,et al.,2020.The mutational features of aristolochic acid-induced mouse and human liver cancers.Hepatology,71(3):929-942.
NG A W T,POON S L,HUANG M N,et al.,2017.Aristolo-chic acids and their derivatives are widely implicated in liver cancers in Taiwan and throughout Asia.Sci.Transl.Med.,9(412):eaan6446.
NTELLAS P,CHAU I,2024.Updates on systemic therapy for hepatocellular carcinoma.Am.Soc.Clin.Oncol.Educ.Book,44:e430028.
PAYSAN L,PIQUET L,SALTEL F,et al.,2016.Rnd3 in cancer:a review of the evidence for tumor promoter or suppressor.Mol.Cancer Res.,14(11):1033-1044.
RUMGAY H,FERLAY J,DE MARTEL C,et al.,2022.Glo-bal,regional and national burden of primary liver cancer by subtype.Eur.J.Cancer,161:108-118.
SUN R F,HE N,ZHANG G Y,et al.,2023.Combined inhibition of KIF11 and KIF15 as an effective therapeutic strategy for gastric cancer.Curr.Cancer Drug Targets,23(4):293-306.
SUNG H,FERLAY J,SIEGEL R L,et al.,2021.Global cancer statistics 2020:GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries.CA Cancer J.Clin.,71(3):209-249.
TANG Y F,ZHANG Y B,WANG C H,et al.,2018.Overexpre-ssion of PCK1 gene antagonizes hepatocellular carcinoma through the activation of gluconeogenesis and suppression of glycolysis pathways.Cell.Physiol.Biochem.,47(1):344-355.
The Cancer Genome Atlas Research Network,2017.Comprehensive and integrative genomic characterization of hepatoce-llular carcinoma.Cell,169(7):1327-1341.e23.
XIANG J,CHEN C,LIU R,et al.,2021.Gluconeogenic enzyme PCK1 deficiency promotes CHK2 O-GlcNAcylation and hepatocellular carcinoma growth upon glucose deprivation.J.Clin.Invest.,131(8):e144703.
XIANG Y Q,ZENG H,LIU X,et al.,2013.Thymidine kinase 1 as a diagnostic tumor marker is of moderate value in cancer patients:a meta-analysis.Biomed.Rep.,1(4):629-637.
YANG H Q,LI J H,NIU Y T,et al.,2025.Interactions between the metabolic reprogramming of liver cancer and tumor microenvironment.Front.Immunol.,16:1494788.
YE J,QIAN Z Z,XUE M,et al.,2019.Aristolochic acid I aggravates renal injury by activating the C3a/C3aR complement system.Toxicol.Lett.,312:118-124.
ZENG H,JI J R,SONG X D,et al.,2020.Stemness related genes revealed by network analysis associated with tumor immune microenvironment and the clinical outcome in lung adenocarcinoma.Front.Genet.,11:549213.
ZHANG M M,LIU H Y,HAN Y M,et al.,2020.On-line enrichment and determination of aristolochic acid in medicinal plants using a MOF-based composite monolith as adsorbent.J.Chromatogr.B Analyt.Technol.Biomed.Life Sci.,1159:122343.
ZHOU J,HE E,SKOG S,2013.The proliferation marker thymidine kinase 1 in clinical use.Mol.Clin.Oncol.,1(1):18-28.
ZHOU Y,CHEN X,LI B S,et al.,2023.KIF11 is a potential prognostic biomarker and therapeutic target for adrenocortical carcinoma.Transl.Androl.Urol.,12(4):594-611.
基本信息:
DOI：10.13417/j.gab.044.001190
中图分类号:R735.7
引用信息:
[1]吴红伟,吴翠红,魏立晓,等.网络毒理学与机器学习整合分析揭示马兜铃酸诱导肝细胞癌的关键生物标志物及其免疫调控机制[J].基因组学与应用生物学,2025,44(11):1190-1208.DOI:10.13417/j.gab.044.001190.
基金信息:
兰州市科技计划项目(2023-ZD-30); 甘肃省药品监督管理局药品监管科学研究项目(2023GSMPA046); 甘肃省中医药高水平重点课题(GZKZ-2024-16)共同资助
2025-10-17
2025-10-17
2025-10-17