Ling Qiang, Liu Mingqi, Xu Wei, Liao Chunhua, Pang Guijian
Department of Urology, The Sixth Affiliated Hospital of Guangxi Medical University, The First People's Hospital of Yulin, Yulin, Guangxi Zhuang Autonomous Region, China.
PLoS One. 2025 Jun 3;20(6):e0324948. doi: 10.1371/journal.pone.0324948. eCollection 2025.
This study aimed to deepen understanding of the molecular mechanisms and key characteristic genes of non-obstructive azoospermia (NOA).
A systematic retrieval method was used to collect the mRNA expression data of NOA and obstructive azoospermia (OA) samples from the GEO database. Data preprocessing, differential gene expression screening, functional annotation, and signal pathway enrichment analysis were conducted using R software. The differences in immune microenvironment between NOA and OA samples were compared through CIBERSORT analysis. LASSO and SVM-RFE, two machine learning algorithms, were applied to select NOA-related characteristic genes. Subsequently, our investigation further identified genes differentially expressed in NOA that are associated with inflammatory responses. NOA samples were clustered based on these inflammation-related genes, while molecular features between different types were explored through pathway enrichment analysis of gene set variation analysis (GSVA). Finally, potential traditional Chinese medicine components targeting these inflammation-related genes were screened from the Chinese medicine database, followed by drug-protein docking simulations.
The study identified 772 DEGs mainly involved in the generation and maturation of sperm. Immune microenvironment analysis revealed significant differences in the infiltration levels of resting NK cells and activated dendritic cells between NOA and OA samples. Eight NOA-related characteristic genes were identified through LASSO and SVM-RFE algorithms. Further analysis revealed that three inflammation-related genes, namely LAMP3, PROK2, and CD14, exhibited significant differential expression in samples of NOA and OA. After clustering of these NOA samples based on the three inflammation-related DEGs, GSVA pathway enrichment analysis revealed molecular features between different NOA subtypes. Finally, potential traditional Chinese medicine components targeting these inflammation-related genes were selected.
This study revealed the key molecular mechanisms and characteristic genes of NOA, especially the role of inflammation-related genes, providing new therapeutic targets and directions for the treatment of NOA.
本研究旨在加深对非梗阻性无精子症(NOA)分子机制和关键特征基因的理解。
采用系统检索方法从基因表达综合数据库(GEO数据库)收集NOA和梗阻性无精子症(OA)样本的mRNA表达数据。使用R软件进行数据预处理、差异基因表达筛选、功能注释和信号通路富集分析。通过CIBERSORT分析比较NOA和OA样本免疫微环境的差异。应用最小绝对收缩和选择算子(LASSO)和支持向量机递归特征消除(SVM-RFE)这两种机器学习算法来选择与NOA相关的特征基因。随后,我们的研究进一步鉴定了在NOA中差异表达且与炎症反应相关的基因。基于这些与炎症相关的基因对NOA样本进行聚类,同时通过基因集变异分析(GSVA)的通路富集分析探索不同类型之间的分子特征。最后,从中药数据库中筛选靶向这些与炎症相关基因的潜在中药成分,随后进行药物-蛋白质对接模拟。
该研究鉴定出772个差异表达基因,主要参与精子的生成和成熟。免疫微环境分析显示,NOA和OA样本之间静息自然杀伤细胞和活化树突状细胞的浸润水平存在显著差异。通过LASSO和SVM-RFE算法鉴定出8个与NOA相关的特征基因。进一步分析表明,3个与炎症相关的基因,即溶酶体相关膜蛋白3(LAMP3)、前动力蛋白2(PROK2)和脂多糖结合蛋白(CD14),在NOA和OA样本中表现出显著差异表达。基于这3个与炎症相关的差异表达基因对这些NOA样本进行聚类后,GSVA通路富集分析揭示了不同NOA亚型之间的分子特征。最后,选择了靶向这些与炎症相关基因的潜在中药成分。
本研究揭示了NOA的关键分子机制和特征基因,尤其是与炎症相关基因的作用,为NOA的治疗提供了新的治疗靶点和方向。