Department of Dermatology, The First Affiliated Hospital of Chongqing Medical University, Chongqing, China.
PLoS One. 2024 Nov 7;19(11):e0313344. doi: 10.1371/journal.pone.0313344. eCollection 2024.
Psoriatic arthritis (PsA) and rheumatoid arthritis (RA) are the most common types of inflammatory musculoskeletal disorders that share overlapping clinical features and complications. The aim of this study was to identify shared marker genes and mechanistic similarities between PsA and RA.
We utilized datasets from the Gene Expression Omnibus (GEO) database to identify differentially expressed genes (DEGs) and perform functional enrichment analyses. To identify the marker genes, we applied two machine learning algorithms: the least absolute shrinkage and selection operator (LASSO) and the support vector machine recursive feature elimination (SVM-RFE). Subsequently, we assessed the diagnostic capacity of the identified marker genes using the receiver operating characteristic (ROC) curve and decision curve analysis (DCA). A transcription factor (TF) network was constructed using data from JASPAR, HumanTFDB, and GTRD. We then employed CIBERSORT to analyze the abundance of immune infiltrates in PsA and RA, assessing the relationship between marker genes and immune cells. Additionally, cellular subpopulations were identified by analyzing single-cell sequencing data from RA, with T cells examined for trajectory and cellular communication using Monocle and CellChat, thereby exploring their linkage to marker genes.
A total of seven overlapping DEGs were identified between PsA and RA. Gene enrichment analysis revealed that these genes were associated with mitochondrial respiratory chain complex IV, Toll-like receptors, and NF-κB signaling pathways. Both machine learning algorithms identified Ribosomal Protein L22-like 1 (RPL22L1) and Lymphocyte Antigen 96 (LY96) as potential diagnostic markers for PsA and RA. These markers were validated using test sets and experimental approaches. Furthermore, GSEA analysis indicated that gap junctions may play a crucial role in the pathogenesis of both conditions. The TF network suggested a potential association between marker genes and core enrichment genes related to gap junctions. The application of CIBERSORT and single-cell RNA sequencing provided a comprehensive understanding of the role of marker genes in immune cell function. Our results indicated that RPL22L1 and LY96 are involved in T cell development and are associated with T cell communication with NK cells and monocytes. Notably, high expression of both RPL22L1 and LY96 was linked to enhanced VEGF signaling in T cells.
Our study identified RPL22L1 and LY96 as key biomarkers for PsA and RA. Further investigations demonstrated that these two marker genes are closely associated with gap junction function, T cell infiltration, differentiation, and VEGF signaling. Collectively, these findings provide new insights into the diagnosis and treatment of PsA and RA.
银屑病关节炎（PsA）和类风湿关节炎（RA）是最常见的两种炎症性肌肉骨骼疾病，它们具有重叠的临床特征和并发症。本研究旨在确定 PsA 和 RA 之间的共同标记基因和机制相似性。
我们利用基因表达综合数据库（GEO）中的数据集来识别差异表达基因（DEGs）并进行功能富集分析。为了识别标记基因，我们应用了两种机器学习算法：最小绝对值收缩和选择算子（LASSO）和支持向量机递归特征消除（SVM-RFE）。随后，我们使用接收者操作特征（ROC）曲线和决策曲线分析（DCA）评估所识别标记基因的诊断能力。使用 JASPAR、HumanTFDB 和 GTRD 中的数据构建转录因子（TF）网络。然后，我们使用 CIBERSORT 分析 PsA 和 RA 中的免疫浸润细胞丰度，评估标记基因与免疫细胞之间的关系。此外，通过分析 RA 的单细胞测序数据来鉴定细胞亚群，使用 Monocle 和 CellChat 检查 T 细胞的轨迹和细胞通讯，从而探索它们与标记基因的联系。
在 PsA 和 RA 之间共鉴定出 7 个重叠的 DEGs。基因富集分析表明，这些基因与线粒体呼吸链复合物 IV、Toll 样受体和 NF-κB 信号通路有关。两种机器学习算法均将核糖体蛋白 L22 样 1（RPL22L1）和淋巴细胞抗原 96（LY96）鉴定为 PsA 和 RA 的潜在诊断标记物。这些标记物通过测试集和实验方法进行了验证。此外，GSEA 分析表明，间隙连接可能在两种疾病的发病机制中起关键作用。TF 网络表明，标记基因与核心富集基因之间可能存在与间隙连接相关的潜在关联。CIBERSORT 和单细胞 RNA 测序的应用提供了对标记基因在免疫细胞功能中的作用的全面理解。我们的结果表明，RPL22L1 和 LY96 参与 T 细胞发育，并与 T 细胞与 NK 细胞和单核细胞的通讯有关。值得注意的是，RPL22L1 和 LY96 的高表达与 T 细胞中 VEGF 信号的增强有关。
我们的研究确定了 RPL22L1 和 LY96 作为 PsA 和 RA 的关键生物标志物。进一步的研究表明，这两个标记基因与间隙连接功能、T 细胞浸润、分化和 VEGF 信号密切相关。总之，这些发现为 PsA 和 RA 的诊断和治疗提供了新的见解。
