研究他克莫司致移植后高血糖相关胰岛细胞差异表达基因、富集功能通路与共表达网络。

利用美国国立生物技术信息中心GEO公共数据库中GSE140230 RNA测序芯片分析他克莫司和PBS处理后人胰岛细胞差异表达基因，并对差异基因进行富集通路功能注释，通过加权基因共表达网络分析他克莫司相关共表达模块枢纽基因，并采用GSE156903 RNA测序芯片和他克莫司致小鼠高血糖动物模型对主要差异表达基因进一步行验证。实时荧光定量PCR(qRT-PCR)结果使用GraphPad Prism 6软件进行非配对t检验分析。P<0.05为差异有统计学意义。

GSE140230 RNA测序芯片分析共鉴定出他克莫司和PBS处理后人胰岛细胞差异表达基因268个，这些差异表达基因形成的蛋白-蛋白相互作用网络中有126个节点和176条边，其中INS、AGT、GAST和ADRA1D为枢纽基因。MCODE插件鉴定蛋白-蛋白相互作用网络中有5个紧密联系模块。基因集富集分析结果显示这些差异表达基因富集于细胞表面受体信号通路、磷代谢过程、含磷酸盐化合物代谢过程、多细胞生物过程正向调控及G蛋白偶联受体信号通路，上述富集通路中包含38个差异基因，其中上调差异基因均存在于G蛋白偶联受体信号通路，下调差异基因富集于其他4条功能通路。加权基因共表达网络分析(WGCNA)共聚类出12个共表达基因模块，其中黄色模块与他克莫司致胰岛细胞损伤正相关度最高(r＝0.64，P＝0.07)，主要富集于胰岛素抵抗和硫代谢等通路；黑色模块负相关度最高(r＝-0.60，P＝0.09)，主要富集于腺苷5′-单磷酸激活蛋白激酶信号通路、亨廷顿病、军团菌病及长寿调节途径-多物种等通路。WGCNA共表达网络分析结果提示黄色模块及黑色模块与他克莫司高度相关，其中黑色和黄色模块分别含有61、56个枢纽基因；黄色模块关键差异基因为LDLRAD3与JMY，黑色模块为IER3与KISS1R。qRT-PCR检测结果显示，DMSO处理组IER3相对表达量(0.20±0.02)高于他克莫司处理组(0.14±0.02)，差异有统计学意义(t＝3.288，P<0.05)，其余关键差异基因表达差异均无统计学意义(t＝0.486、0.742和1.731，P均>0.05)。GSE156903 RNA测序芯片分析显示，IER3为他克莫司处理组表达下调的基因。免疫荧光结果显示他克莫司处理组小鼠IER3蛋白表达量下降；荧光共定位分析显示DMSO处理组中约99%胰岛素阳性染色区域显示为IER3阳性染色，而他克莫司处理组仅约65%胰岛素阳性细胞共表达IER3。STRING数据库分析结果显示，IER3与蛋白磷酸酶2相关基因、丝裂原活化蛋白激酶相关基因、TNF和SLC37A2等存在蛋白-蛋白相互作用。

IER3可能为他克莫司致移植后新发糖尿病(PTDM)重要调节基因，可能可以作为PTDM防治潜在药物治疗靶点。

To investigated the functional pathways enriched and differential expressed genes in islet of tacrolimus induced post-transplantation hyperglycemia.

Differential expressed genes between tacrolimus and PBS treatment were filtered from the dataset GSE140230 from National Center for Biotechnology Information GEO database. Functional pathway annotations were conducted by gene set enrichment analysis and weighted gene co-expression network analysis. Master differential expressed genes screened from highly correlated modules of co-expression network analysis were then validated in dataset GSE156903 and tacrolimus induced diabetic mice islets. The results of quantiative real-time PCR (qRT-PCR) were compared with uncoupled t test in GraphPad Prism 6. A P<0.05 was considered statistically significant.

Altogether 268 differential expressed genes were filtered . The protein-protein interaction network constructed by these differential expressed genes had 126 nodes and 176 edges. INS, AGT, GAST and ADRA1D were the seed genes with highest degree of the protein-protein interaction network. MCODE plugin identified five close link modules in the protein-protein interaction network. Differential expressed genes mainly enriched in cell surface receptor signaling pathway, phosphorus metabolic process, phosphate-containing compound metabolic process, positive regulation of multicellular organismal process and G protein-coupled receptor signaling pathway. Altogether 38 differential expressed genes were contained in these pathways, and upregulated differential expressed genes were all enriched in G protein-coupled receptor signaling pathway, while downregulated differential expressed genes were enriched in other 4 pathways. Weighted gene correlation network analysis (WGCNA) screened 12 co-expression modules. Yellow module was the most positive correlated module (r=0.64, P=0.07), genes in yellow module were mainly enriched in insulin resistance and sulfur metabolism. Black module was the most negative correlated module (r=-0.60, P=0.09). Genes in black module were mainly enriched in adenosine 5′-monophosphate-activated protein kinase signaling pathway, Huntington disease, legionellosis and longevity regulatin pathway-multiple species. There were 61 and 56 seed genes in yellow module and black modules, respectively. The results of qRT-PCR showed that there was statistic difference for IER3 expression of islets between tacrolimus and DMSO treated mice (t=3.288, P<0.05), while KISS1R, LDLRAD3 and JMY were not (t=0.486, 0.742 and 1.731, all P>0.05). IER3 was also the differential expressed genes of tacrolimus treated islets in GSE156903. Immunofluorescence analysis showed that IER3 protein expression was decreased in tacrolimus-treated mice. The co-locoliazation analysis showed that almost 99% of insulin-positive staining area showed the postive-IER3 staining, while only 65% of the insulin-positive staining area was IER3 positive in tacrolimus-treated group. STRING database showed that IER3 was interacted with protein phosphatase 2 related genes, mitogen-activated protein kinase related genes, TNF and SLC37A2.

IER3 may play an important role in tacrolimus induced diabetes. Our findings may provide potential targets for prevention and treatment of tacrolimus induced post-transplantation diabetes mellitus.