Liu Hanjie, You Maochun, Yang Hui, Wu Xiao, Zhang Siyu, Huang Sihan, Gao Huijuan, Xie Lushuang
School of Clinical Medicine, Tsinghua Medicine, Tsinghua University, Haidian District, 100084, Beijing, P.R. China.
Chengdu University of Traditional Chinese Medicine, Chengdu, 611137, Sichuan, P.R. China.
Metab Brain Dis. 2025 Jan 22;40(1):109. doi: 10.1007/s11011-025-01540-x.
The immune system has emerged as a major factor in the pathogenesis of Alzheimer's disease (AD). PANoptosis is a newly defined programmed cell death mechanism related to many inflammatory diseases. This study aimed to identify the differentially expressed (DE) PANoptosis-related genes with characteristics of immune dysregulation (PRGIDs) in AD using bioinformatics analysis of bulk RNA-seq and single-nuclei RNA sequencing (snRNA-seq) data. To improve the robustness of gene selection, we integrated 3 microarray and 6 snRNA-seq datasets from the Gene Expression Omnibus (GEO), which allowed us to not only examine overall gene expression patterns but also assess the cellular specificity of gene expression at the single-cell level. This approach helped to identify cell-type-specific gene alterations that may be masked in bulk RNA-seq analyses. Relevant PANoptosis, immune dysregulation, and AD-related genes were obtained from the Genecards database. The AlzData database was also used in this study. Expression validation, the least absolute shrinkage and selection operator (LASSO) regression model, and CytoHubba algorithms were applied for key DE-PRGIDs selection. LASSO, Logistic, and Cox regressions were used to construct prognostic models. The receiver operating characteristic (ROC) curve and correlation analyses were conducted on key DE-PRGIDs. The Seurat package in R software was employed for performing snRNA-seq data processing. Uniform manifold approximation and projection (UMAP) was utilized for cell type annotation and PRGID cell visualization. The violin plot was applied for displaying expression levels of PRGIDs. High-dimensional consensus weighted gene co-expression network analysis (hdWGCNA) was conducted on microglia to identify gene modules and hub genes. Venn diagram analysis identified 250 PRGIDs and 39 DE-PRGIDs. NFKBIA was identified as the key gene. Prognostic models based on the expression level of NFKBIA were obtained. ROC curve analysis revealed its area under the curve (AUC) value: 0.661 in training set and 0.836 in validation set. The heatmap displayed the result of correlation analysis. SnRNA-seq data analysis identified 7 cell types. The UMAP and violin plots revealed highly expressed PRGIDs in microglia with remarkable differences between healthy controls and AD. hdWGCNA identified PVT1 and APOE as hub genes associated with microglia. In conclusion, our findings provide evidence that PANoptosis may play a role in the immune dysregulation observed in AD. PVT1 has been implicated in AD pathogenesis, potentially exerting its effects through the miR-488-3p/FOXD3/SCN2A axis.
免疫系统已成为阿尔茨海默病（AD）发病机制中的一个主要因素。PANoptosis是一种新定义的与许多炎症性疾病相关的程序性细胞死亡机制。本研究旨在通过对批量RNA测序（RNA-seq）和单核RNA测序（snRNA-seq）数据进行生物信息学分析，鉴定AD中具有免疫失调特征的差异表达（DE）PANoptosis相关基因（PRGIDs）。为了提高基因选择的稳健性，我们整合了来自基因表达综合数据库（GEO）的3个微阵列和6个snRNA-seq数据集，这使我们不仅能够检查整体基因表达模式，还能在单细胞水平评估基因表达的细胞特异性。这种方法有助于识别在批量RNA-seq分析中可能被掩盖的细胞类型特异性基因改变。从Genecards数据库中获取了相关的PANoptosis、免疫失调和AD相关基因。本研究还使用了AlzData数据库。应用表达验证、最小绝对收缩和选择算子（LASSO）回归模型以及CytoHubba算法进行关键DE-PRGIDs的选择。使用LASSO、逻辑回归和Cox回归构建预后模型。对关键DE-PRGIDs进行了受试者工作特征（ROC）曲线和相关性分析。使用R软件中的Seurat包进行snRNA-seq数据处理。采用均匀流形近似和投影（UMAP）进行细胞类型注释和PRGID细胞可视化。使用小提琴图展示PRGIDs 的表达水平。对小胶质细胞进行了高维共识加权基因共表达网络分析（hdWGCNA）以识别基因模块和枢纽基因。维恩图分析确定了250个PRGIDs和39个DE-PRGIDs。NFKBIA被确定为关键基因。获得了基于NFKBIA表达水平的预后模型。ROC曲线分析显示其曲线下面积（AUC）值：训练集中为0.661，验证集中为0.836。热图展示了相关性分析的结果。snRNA-seq数据分析确定了7种细胞类型。UMAP和小提琴图显示小胶质细胞中PRGIDs高表达，在健康对照和AD之间存在显著差异。hdWGCNA确定PVT1和APOE为与小胶质细胞相关的枢纽基因。总之，我们的研究结果提供了证据表明PANoptosis可能在AD中观察到的免疫失调中起作用。PVT1已被证明与AD发病机制有关，可能通过miR-488-3p/FOXD3/SCN2A轴发挥作用。
