不同阻滞阶段非梗阻性无精子症的生物信息学分析

目的：利用生物信息学筛选不同阻滞阶段非梗阻性无精子症（NOA）相关的调控因子及其调控途径,为后续的功能研究提供依据。方法：从GEO（Gene Expression Omnibus）数据库中获得GSE45885的基因表达谱矩阵,筛选其中的正常样本以及不同阻滞阶段NOA样本,包括4例正常生精样本和20例NOA患者样本,20例NOA患者中有2例患者处于减数分裂前期阻滞阶段（PRE）,7例处于减数分裂期阻滞阶段（MEI）,11例处于减数分裂后期阻滞阶段（POST）。将正常生精样本设置为对照组（CON）,使用GEO2R在线工具鉴定正常生精样本与不同阻滞阶段NOA样本之间的差异基因,对GEO芯片中不同阻滞阶段的NOA数据整合归一化并进行矫正后取交集。对差异表达的基因以及其靶基因进行GO（Gene Ontology）和KEGG（Kyoto Encyclopedia of Genes and Genomes）富集分析,获取关键通路;基于在线工具（STRING）构建蛋白质相互作用（PPI）网络图。然后利用Cytoscape中的CytoHubba插件对枢纽基因进行筛选。结果：在PRE中筛选出463个上调基因,12个下调基因;在MEI和POST中分别发现2个和3个下调基因,无上调基因。PRE中上调基因通过一些重要途径,如精子发生、精子细胞发育、精子的运动、纤毛运动和微管形成等功能来调控生精过程。在上述的3个阻滞阶段NOA中筛选出2个共同差异基因,均为微小RNA（miRNA）,2个miRNA有58个共同靶基因。利用在线生物信息学工具STRING成功构建PRE中上调基因的PPI网络图,并使用Cytoscape筛选出网络中前10个枢纽基因,包括DNAI1、DNAI2、PGK2、ROPN1L、ARMC4、DRC1、DNAAF3、CABYR、ZMYND10和CCDC65。结论：识别枢纽基因和共同差异基因及其通路为后续研究NOA的发生机制提供了有价值的参考,DRC1、ARMC4、MIR15A和MIR509-3可作为后续NOA发病机制研究的潜在靶点。

Bioanalysis of Non-Obstructive Azoospermia at Different Arrest Stages

Objective: To screen the regulatory factors and the related pathways of non-obstructive azoospermia (NOA), and to explore the underlying molecular mechanisms by bioinformatics. Methods: The gene expression matrix of GSE45885 was obtained from GEO database. The samples of the NOA at the different arrest stages and the controls were screened, including 20 NOA samples and 4 normal samples. Among the 20 NOA patients, 2 patients were in the premeiotic arrest stage (PRE), 7 patients in the meiotic arrest stage (MEI) and 11 were in postmeiotic arrest stage (POST). The GEO2R online tool was used to identify the differentially expressed genes (DEGs). The data of three arrest stages of NOA in GEO chip were integrated and normalized, and then the intersection was obtained after correction. Gene Ontology（GO）enrichment analysis and Kyoto Encyclopedia of Genes and Genomes（KEGG）pathway enrichment analysis were performed for the differentially expressed genes and their target genes to obtain key pathways. Protein-protein interaction (PPI) network graph was constructed based on online tool (STRING). Finally, CytoHubba in Cytoscape was used to screen Hub genes. Results: 463 up-regulated genes and 12 down-regulated genes were screened in PRE. Two and three down-regulated genes were found in MEI and POST, respectively. There were no up-regulated genes in MEI and POST. The up-regulated genes in PRE participated in the regulation of spermatogenesis through some important pathways such as spermatogenesis, spermatid development, sperm motility, motile cilium and microtubule formation. Two common differential expression genes were screened out from the above three arrest stages of NOAs, two of them were microRNAs (miRNA), and two miRNAs had 58 common target genes. An online bioinformatics tool STRING was used to successfully construct a PPI network map of up-regulated genes in PRE, and Cytoscape was used to screen out the top 10 Hub genes in the network including DNAI1, DNAI2, PGK2, ROPN1L, ARMC4, DRC1, DNAAF3, CABYR, ZMYND10, CCDC65. Conclusions: The identification of Hub genes, common differential genes and their pathways will provide valuable references for the subsequent study on the mechanism of NOA. DRC1, ARMC4, MIR15A and MIR509-3 may serve as potential targets for subsequent studies on the etiological mechanism of NOA.