胰腺癌预后免疫基因对特征筛选及免疫评估

Screening of immune gene pairs signatures and assessment of immune status in pancreatic cancer

目的筛选胰腺癌(pancreatic cancer,PC)预后免疫相关基因对特征(immune-related gene pairs,IRGPs),构建风险评估模型及评估机制。方法ImmPort免疫基因集与癌症基因组图谱-胰腺癌(the cancer genome atlaspancreatic adenocarcinoma,TCGA-PAAD)及临床蛋白质组肿瘤分析协作组3期(clinical proteomic tumor analysis consortium 3,CPTAC-3)基因信息取交集获得IRGPs,Lasso-Cox比例风险回归法确定预后IRGPs并构建最佳风险评分(risksocre,RS)模型,随后对RS划分的高低风险组进行Kaplan-Meier生存分析、单因素和多因素Cox分析、cibersort和estimates免疫评估。通过相关性分析、基因本体论(gene ontology,GO)及京都基因和基因组百科全书(kyoto encyclopedia of genes and genomes,KEGG)富集分析评估IRGPs单基因涉及的分子机制。结果构建了基于19个IRGPs的PC RS模型。与低风险组相比,高风险组1年总生存率缩短,估计评分及免疫评分降低,肿瘤纯度则增加,以上差异均有统计学意义(P<0.001);高风险组的肿瘤组织幼稚B细胞、CD8^(+)-T细胞、滤泡辅助T细胞、调节性T细胞、单核细胞的浸润比例下降,以上差异均有统计学意义(P<0.05),而活化CD4^(+)-记忆T细胞、M1或M2巨噬细胞、静息树突状细胞、嗜酸性粒细胞、中性粒细胞比例则升高,以上差异均有统计学意义(P<0.05)。与肿瘤纯度呈显著负相关的基因包括:ITGAL、PIK3R5、ADA2、GZMB、TRBJ1.5、CXCL9、IGKV1.39、APLNR、IGLV7.46 IGHJ2、IL3RA、ACKR1、AKT3、SEMA3G、PLA2G2A、FGF2,而呈正相关的为:LTBP3、GRK2、BIRC5、IL22RA1、IL20RA,它们参与了免疫细胞活化及体液免疫应答过程,具有受体激活、抗原结合等功能。结论RS模型可反映PC高风险群体的免疫抑制状态,是评估PC预后的有效工具。

Objective:To screen the prognostic immune-related gene pair signatures(IRGPs)in pancreatic cancer(PC),build a risk score model,and explore the mechanism.Methods:ImmPort immune gene set,TCGA-PADD,and CPTAC-3 gene information were intersected to obtain IRGPs,Lasso-Cox proportional hazards regression method was used to determine the prognosis of IRGPs and construct the optimal risk score model(RS),followed by kaplan-meier survival analysis,univariate and multivariate cox analysis,cibersort-and estimates-based immune assessment for the high-and lowrisk groups divided by RS.correlation analysis,GO,and KEGG enrichment analysis evaluated the molecular mechanisms involved in single genes of IRGPs.Results:A PC RS model based on 19 IRGPs was constructed.Compared with the lowrisk group,the 1-year overall survival rate in the high-risk group was significantly shortened(P<0.001),the Estimate score and Immune score were significantly reduced,and Tumor purity was increased(P<0.05),naive B cells,CD8^(+)-T cells,follicular helper T cells,regulatory T cells,and monocytes were significantly lower expressed(P<0.05),while activated CD4^(+)-memory T cells,M1 or M2 macrophages,resting dendritic cells,eosinophils,and neutrophils were significantly highly expressed(P<0.05).Genes that were significantly negatively correlated with tumor purity included:ITGAL、PIK3R5、ADA2、GZMB、TRBJ1.5、CXCL9、IGKV1.39、APLNR、IGLV7.46 IGHJ2、IL3RA、ACKR1、AKT3、SEMA3G、PLA2G2A and FGF2,and the positively correlated genes were:LTBP3,GRK2,BIRC5,IL22RA1 and IL20RA.They were involved in immune cell activation and humoral immune response and had receptor activation and antigen binding functions.Conclusion:RS model can reflect the immunosuppressive status of high-risk PC patients,and is an effective tool for evaluating the prognosis of PC.