Department of Computational Biology and Medical Sciences, Graduate School of Frontier Sciences, The University of Tokyo, 7-3-1 Hongo, Bunkyo-ku, Tokyo, 113-0033, Japan.
Sci Rep. 2023 Apr 26;13(1):6817. doi: 10.1038/s41598-023-34021-1.
Alzheimer's disease (AD) is the most prevalent dementia disorder globally, and there are still no effective interventions for slowing or stopping the underlying pathogenic mechanisms. There is strong evidence implicating neural oxidative stress (OS) and ensuing neuroinflammation in the progressive neurodegeneration observed in the AD brain both during and prior to symptom emergence. Thus, OS-related biomarkers may be valuable for prognosis and provide clues to therapeutic targets during the early presymptomatic phase. In the current study, we gathered brain RNA-seq data of AD patients and matched controls from the Gene Expression Omnibus (GEO) to identify differentially expressed OS-related genes (OSRGs). These OSRGs were analyzed for cellular functions using the Gene Ontology (GO) database and used to construct a weighted gene co-expression network (WGCN) and protein-protein interaction (PPI) network. Receiver operating characteristic (ROC) curves were then constructed to identify network hub genes. A diagnostic model was established based on these hub genes using Least Absolute Shrinkage and Selection Operator (LASSO) and ROC analyses. Immune-related functions were examined by assessing correlations between hub gene expression and immune cell brain infiltration scores. Further, target drugs were predicted using the Drug-Gene Interaction database, while regulatory miRNAs and transcription factors were predicted using miRNet. In total, 156 candidate genes were identified among 11046 differentially expressed genes, 7098 genes in WGCN modules, and 446 OSRGs, and 5 hub genes (MAPK9, FOXO1, BCL2, ETS1, and SP1) were identified by ROC curve analyses. These hub genes were enriched in GO annotations "Alzheimer's disease pathway," "Parkinson's Disease," "Ribosome," and "Chronic myeloid leukemia." In addition, 78 drugs were predicted to target FOXO1, SP1, MAPK9, and BCL2, including fluorouracil, cyclophosphamide, and epirubicin. A hub gene-miRNA regulatory network with 43 miRNAs and hub gene-transcription factor (TF) network with 36 TFs were also generated. These hub genes may serve as biomarkers for AD diagnosis and provide clues to novel potential treatment targets.
阿尔茨海默病（AD）是全球最常见的痴呆症，目前仍没有有效的干预措施来减缓或阻止潜在的致病机制。有强有力的证据表明，在 AD 大脑中，神经氧化应激（OS）和随之而来的神经炎症与症状出现前后的进行性神经退行性变有关。因此，OS 相关生物标志物可能对预后有价值，并为早期无症状阶段的治疗靶点提供线索。在目前的研究中，我们从基因表达综合数据库（GEO）中收集了 AD 患者和匹配对照的大脑 RNA-seq 数据，以鉴定差异表达的 OS 相关基因（OSRG）。使用基因本体论（GO）数据库分析这些 OSRG 的细胞功能，并用于构建加权基因共表达网络（WGCN）和蛋白质-蛋白质相互作用（PPI）网络。然后构建接收者操作特征（ROC）曲线来识别网络枢纽基因。基于这些枢纽基因，使用最小绝对收缩和选择算子（LASSO）和 ROC 分析建立诊断模型。通过评估枢纽基因表达与免疫细胞脑浸润评分之间的相关性，检查免疫相关功能。此外，使用药物-基因相互作用数据库预测靶向药物，使用 miRNet 预测调节性 miRNA 和转录因子。总共从 11046 个差异表达基因、7098 个 WGCN 模块基因和 446 个 OSRG 中鉴定出 156 个候选基因，并通过 ROC 曲线分析鉴定出 5 个枢纽基因（MAPK9、FOXO1、BCL2、ETS1 和 SP1）。这些枢纽基因在 GO 注释中富集于“阿尔茨海默病途径”、“帕金森病”、“核糖体”和“慢性髓性白血病”。此外，预测到 78 种药物可以靶向 FOXO1、SP1、MAPK9 和 BCL2，包括氟尿嘧啶、环磷酰胺和表柔比星。还生成了一个包含 43 个 miRNA 的枢纽基因-miRNA 调控网络和一个包含 36 个 TF 的枢纽基因-转录因子（TF）网络。这些枢纽基因可能作为 AD 诊断的生物标志物，并为新的潜在治疗靶点提供线索。
