基于生物信息学和实验验证分析巨噬细胞极化在肺结节病中的作用

Analysis of the role of macrophage polarization in pulmonary sarcoidosis based on bioinformatics and experimental validation

目的通过生物信息学分析和实验验证探讨巨噬细胞极化在肺结节病中的作用。方法本研究为实验研究。从基因表达综合数据库下载肺结节病相关RNA-seq数据集GSE83456、GSE34608、GSE42834、GSE16538、GSE18781、GSE19314、GSE19976和GSE37912。将GSE83456数据集作为训练集,其余数据集作为验证集。采用非随机抽样的方法选取2021年1月至2023年1月于成都医学院第一附属医院住院的15例肺结节病患者(Ⅰ期或Ⅱ期)、15例肺结核患者和15例肺腺癌患者,以及同期接受体检的20名健康受试者为研究对象,收集肺结节病患者和健康受试者的外周血样本,收集肺结节病、肺结核和肺腺癌患者的纵隔淋巴结样本(肺腺癌患者淋巴结转移阴性)。通过"xCell"包分析GSE83456数据集中巨噬细胞、M1巨噬细胞和M2巨噬细胞在肺结节病患者中的丰度变化,并通过流式细胞术检测肺结节病患者和健康受试者外周血M1和M2巨噬细胞比例。通过基因集富集分析探索肺结节病组与对照组、高M1巨噬细胞组与低M1巨噬细胞组、高M2巨噬细胞组与低M2巨噬细胞组之间的潜在生物学差异。加权基因共表达网络分析筛选与巨噬细胞极化高度相关的模块基因,将GSE83456数据集中肺结节病组和对照组的差异基因与相关模块基因取交集,得到巨噬细胞极化相关差异基因,并对其进行基因本体论富集分析。通过Lasso回归和随机森林算法筛选出肺结节病患者巨噬细胞极化关键基因,并在验证集中验证关键基因的表达情况,对GSE83456数据集中关键基因与巨噬细胞丰度的关系进行Pearson相关分析。通过蛋白质印迹法验证关键基因在肺结节病、肺结核和肺腺癌患者纵隔淋巴结中的蛋白表达水平。绘制受试者工作特征曲线分析关键基因表达蛋白鉴别肺结节病和肺结核的价值。结果在GSE83456数据集中,肺结节病组巨噬细胞、M1巨噬细胞丰度评分均高于对照组[(0.27±0.09)分比(0.12±0.07)分,(0.18±0.07)分比(0.12±0.04)分,均P<0.001]。肺结节病患者外周血M1巨噬细胞比例和M1/M2巨噬细胞比值均高于健康受试者[(0.60±0.12)比(0.47±0.11),(1.23±0.07)比(0.84±0.06),均P<0.05]。炎症反应和干扰素反应等免疫相关通路富集于肺结节病组和高M1巨噬细胞组。加权基因共表达网络分析与差异基因分析的交集共获得32个巨噬细胞极化相关的差异基因,这些基因主要富集于免疫相关的生物过程和分子功能。Lasso回归和随机森林算法筛选出肺结节病患者巨噬细胞极化关键基因ANKRD22。在7个验证集中,肺结节病组ANKRD22 mRNA表达量均高于对照组,并且在GSE42834数据集中肺结核组ANKRD22 mRNA表达量高于肺结节病组(均P<0.001)。Pearson相关分析显示,在GSE83456数据集中,ANKRD22 mRNA表达量与巨噬细胞丰度(r=0.65)、M1巨噬细胞丰度(r=0.54)和单核细胞丰度(r=0.45)均呈正相关(均P<0.001)。蛋白质印迹法显示,肺结核患者和肺结节病患者的ANKRD22蛋白相对表达量均高于肺腺癌患者,且肺结核患者高于肺结节病患者[(0.98±0.02)比(0.38±0.03),(0.62±0.02)比(0.38±0.03),(0.98±0.02)比(0.62±0.02),均P<0.05]。ANKRD22蛋白鉴别肺结节病和肺结核的曲线下面积为0.802,最佳截断值为0.58,敏感度为73.3%,特异度为80.0%。结论M1巨噬细胞介导免疫炎症反应参与肺结节病的发生发展。肺结节病巨噬细胞极化关键基因ANKRD22可以作为肺结节病的潜在生物标志物,并具有鉴别肺结节病和肺结核的潜在价值。

Objective To explore the role of macrophage polarization in pulmonary sarcoidosis by bioinformatics analysis and experimental validation.Methods This was an experimental study.RNA-seq datasets GSE83456,GSE34608,GSE42834,GSE16538,GSE18781,GSE19314,GSE19976,and GSE37912 related to pulmonary sarcoidosis were downloaded from the Gene Expression Omnibus(GEO)database.The GSE83456 dataset was set as the training set and the remaining datasets were considered as the validation set.Fifteen patients with pulmonary sarcoidosis in stageⅠ-Ⅱ,15 patients with pulmonary tuberculosis,and 15 patients with lung adenocarcinoma who were hospitalized in the First Affiliated Hospital of Chengdu Medical College from January 2021 to January 2023 were included using non-random sampling.During the same period,20 healthy subjects who received physical examinations were included as the research subjects using non-random sampling.Peripheral blood samples were collected from patients with pulmonary sarcoidosis and healthy individuals,and mediastinal lymph node samples were collected from patients with pulmonary sarcoidosis,pulmonary tuberculosis,and lung adenocarcinoma(lung adenocarcinoma patients with negative lymph node metastasis).Macrophage,M1 and M2 macrophage abundances in patients with pulmonary sarcoidosis of GSE83456 dataset were assessed using the"xCell"package and the ratio of M1 and M2 macrophages in peripheral blood of pulmonary sarcoidosis patients and healthy subjects were detected by flow cytometry.Gene set enrichment analysis(GSEA)was performed using the GSEA software,thus exploring potential biological differences between the pulmonary sarcoidosis group and the control group,between the high M1 macrophage group and the low M1 macrophage group,and between the high M2 macrophage group and the low M2 macrophage group.Weighted gene co-expression network analysis(WGCNA)was performed using the"WGCNA"package to screen module genes that were highly correlated with macrophage polarization.Differentially expressed genes between pulmonary tuberculosis and control groups in the GSE83456 dataset were intersected with the key module genes to obtain the macrophage polarization-related differential genes,followed by the gene ontology enrichment analysis.Lasso regression and random forest algorithm were used to screen out key genes from macrophage polarization-related differential genes,and verified in validation sets.Pearson correlation analysis was performed on the relationship between key genes and macrophage abundance in the GSE83456 dataset.The expressions of key genes were verified by Western blotting in mediastinal lymph nodes of patients with pulmonary sarcoidosis,pulmonary tuberculosis,and lung adenocarcinoma.Receiver operator characteristic curve was drawn to analyze the value of protein encoded by key genes in distinguishing pulmonary sarcoidosis from pulmonary tuberculosis.Results In the GSE83456 dataset,patients with pulmonary sarcoidosis had significantly higher abundance scores of macrophages and M1 macrophages compared with those of control group([0.27±0.09]points vs[0.12±0.07]points,[0.18±0.07]points vs[0.12±0.04]points,both P<0.001).The proportion of M1 macrophages and M1/M2 macrophage ratio in peripheral blood of patients with pulmonary sarcoidosis were significantly higher than those of healthy subjects([0.60±0.12]vs[0.47±0.11],[1.23±0.07]vs[0.84±0.06],both P<0.05).Immune-related pathways such as inflammatory response and interferon response were significantly enriched in pulmonary sarcoidosis group and the high M1 macrophage abundance group.The intersection of WCGNA and differentially expressed genes obtained 32 macrophage polarization-related differential genes,which were mainly enriched in immune-related biological processes and molecular functions.Lasso regression and random forest analysis screened for the key gene ANKRD22 for macrophage polarization in pulmonary sarcoidosis.In all seven validation sets,the mRNA expression level of ANKRD22 was higher in the pulmonary sarcoidosis group than that in the control group.Besides,in the GSE42834 dataset,the mRNA expression level of ANKRD22 was significantly highly expressed in pulmonary tuberculosis than those with pulmonary sarcoidosis(all P<0.001).Pearson correlation analysis showed that in the GSE83456 dataset,the mRNA expression level of ANKRD22 was positive associated with macrophage abundance(r=0.65),M1 macrophage abundance(r=0.54)and monocyte abundance(r=0.45)(all P<0.001).Western blotting on clinical samples showed that the protein level of ANKRD22 was higher in patients with pulmonary tuberculosis and pulmonary sarcoidosis than in patients with lung adenocarcinoma,and the highest expression was found in tuberculosis samples([0.98±0.02]vs[0.38±0.03],[0.62±0.02]vs[0.38±0.03],[0.98±0.02]vs[0.62±0.02];all P<0.05).The area under the curve for ANKRD22 to discriminate pulmonary sarcoidosis from pulmonary tuberculosis was 0.802,with the optimal cut-off value,sensitivity and specificity of 0.58,73.3%and 80.0%respectively.Conclusions M1 macrophage-mediated immune-inflammatory responses are involved in the development of pulmonary sarcoidosis.ANKRD22,a key gene for macrophage polarization in pulmonary sarcoidosis,may serve as a potential biomarker for pulmonary sarcoidosis and has potential value in differentiating pulmonary sarcoidosis from pulmonary tuberculosis.