基于构建肺癌免疫相关预后模型及网络药理学探讨金福安汤的作用机制

Exploration on the Mechanism of Action of Jinfu’an Decoction Based on the Construction of Immune-Related Prognostic Model for Lung Cancer and Network Pharmacology

目的通过构建免疫相关基因的肺癌临床预后模型及网络药理学,探讨化痰祛瘀法代表方金福安汤在抗肿瘤治疗中的潜在作用机制。方法从公共数据库下载肺腺癌(lung adenocarcinoma,LUAD)患者的RNA表达数据及临床信息,使用加权基因共表达网络分析(weighted gene co-expression network analysis,WGCNA)构建基因之间的关联网络得到与患者预后最相关的基因集,并与免疫相关基因数据库(Immunology Database and Analysis Portal,Immport)的基因取交集,进行单因素COX分析和多因素COX分析筛选出风险基因,构建免疫相关预后模型并绘制动态列线图直观评估患者的预后情况。通过Kaplan-Meier法、受试者工作特征曲线(receiver operator characteristic curve,ROC)以及外部验证集对预后模型进行验证,通过校准曲线评价预测能力。通过GEPIA验证风险基因对患者预后的影响,构建肿瘤-基因-蛋白质作用网络图,基于CIBERSORT算法进行肿瘤微环境免疫浸润分析。收集化痰祛瘀法代表方金福安汤的入血成分,获取化合物成分及潜在作用靶点。结合蛋白质-蛋白质相互作用(protein-protein interaction,PPI)网络筛选出核心靶点,通过基因本体论(gene ontology,GO)和京都基因与基因组百科全书(Kyoto encyclopedia of genes and genomes,KEGG)富集分析探索其机制通路,最终构建“药物-化合物-靶点”网络图。结果构建了一个由ADIPOR2、HSPA4、PPIA、TMSB10基因组成的预后风险评估模型,ROC曲线计算曲线下面积(area under curve,AUC)分别为1年0.73,3年0.65,5年0.68,外部验证集的高风险组患者的总生存时间比低风险组更短(P<0.05)。动态列线图使用风险评分、肿瘤分期和年龄性别进行构建,校准曲线表明其具有良好的预测能力。免疫量化分析显示在高风险患者群体中,M2型巨噬细胞和浆细胞的比例较高,而CD8+T细胞的比例较低。得到23个金福安汤治疗肺癌的核心靶点。GO和KEGG功能富集分析表明金福安汤治疗高风险肿瘤患者主要涉及肿瘤生长、细胞周期调控、细胞凋亡以及免疫反应等过程和信号通路。结论基于免疫相关基因的预后模型能够为肺癌患者提供准确的预后评估。通过网络药理学初步揭示了金福安汤治疗非小细胞肺癌(non-small cell lung cancer,NSCLC)的作用机制,体现了其多成分、多靶点、多通路的作用特点。

Objective This study aimed to explore the potential anti-tumor mechanisms of Jinfu’an Decoction,a representative formula of the phlegm-resolving and stasis-removing approach,through the construction of lung cancer prognostic model based on immune-related genes and network pharmacology.Methods RNA expression data and clinical information of lung adenocarcinoma(LUAD)patients were downloaded from public databases.Weighted gene co-expression network analysis(WGCNA)was used to construct a gene association network,identifying the gene set most relevant to patient prognosis.The gene set was intersected with genes from the Immunology Database and Analysis Portal(Immport).Univariate and multivariate COX proportional hazard analyses were then performed to screen out risk genes.An immune-related prognostic model was constructed,and dynamic nomograms were drawn to visually assess the prognosis of patients.The prognostic model was validated using Kaplan-Meier survival curves,receiver operator characteristic curve(ROC),and external validation cohorts,with calibration curves used to evaluate prediction accuracy.The impact of risk genes on patient prognosis was further validated using GEPIA.A tumor-gene-protein interaction network was constructed,and tumor microenvironment immune infiltration analysis was conducted using the CIBERSORT algorithm.Additionally,the blood-active components of Jinfu’an Decoction,a representative formula of the phlegm-resolving and stasisremoving approach,were collected to obtain compound components and potential targets.Core targets were selected through protein-protein interaction(PPI)network analysis,and their mechanisms and pathways were explored via gene ontology(GO)and Kyoto encyclopedia of genes and genomes(KEGG)enrichment analyses.Ultimately,a“drug-compound-target”network diagram was constructed.Results A prognostic risk assessment model consisting of ADIPOR2,HSPA4,PPIA,and TMSB10 genes was constructed.The area under the curve(AUC)calculated by ROC was 0.73 for 1-year,0.65 for 3-year,and 0.68 for 5-year,respectively.External validation showed that patients in the high-risk group had a significantly shorter overall survival compared to those in the low-risk group(P<0.05).Dynamic nomograms incorporating risk scores,tumor stage,age,and sex demonstrated good predictive performance,as confirmed by calibration curves.Immune quantification analysis revealed a higher proportion of M2 macrophages and plasma cells in high-risk patients,while the proportion of CD8+T cells was lower.A total of 23 core targets associated with the Jinfu’an Decoction for lung cancer treatment were identified.GO and KEGG enrichment analyses indicated that the therapeutic effects of Jinfu’an Decoction on high-risk tumor patients mainly involved processes and signaling pathways related to tumor growth,cell cycle regulation,cell apoptosis,and immune responses.Conclusion The prognostic model based on immune-related genes provides accurate prognostic assessments for lung cancer patients.Preliminary results through network pharmacology reveal the mechanisms of action of Jinfu’an Decoction in treating non-small cell lung cancer(NSCLC),demonstrating its characteristics of multi-component,multi-target,and multi-pathway effects.