目的：通过生物信息学手段筛选乳腺癌中差异表达的关键miRNA及其靶基因，干预其在乳腺癌细胞中的表达并观察对乳腺癌细胞功能的影响。方法：利用GEO数据库筛选在乳腺癌中差异表达的miRNA，ENCORI数据库验证差异miRNA的表达，以选定最显著的差异表达 miRNA 为研究对象；利用 Starbase、miRDB 和 miRWalk 数据库预测 miR-32-5p 的靶基因，利用DAVID数据库对靶基因进行GO分析和KEGG分析，利用String数据库联合Cytoscape3.6.2软件进行PPI网络分析及核心基因的筛选，从核心基因中选择相互联系紧密“度值”最显著的Dickkopf相关蛋白3（DDK3）基因进行后续实验。qPCR检测miR-32-5p在人正常乳腺细胞 MCF10A和人乳腺癌细胞MCF7、MDA-MB-231、MDA-MB-453细胞中的表达。向MDA-MB-231细胞中转染miR-32-5p mimic、miR-32-5p inhibitor及各自的对照（NC）序列，分别用CCK-8法、流式细胞术和Transwell实验检测过表达或抑制miR-32-5p对细胞增殖、凋亡和侵袭的影响。结果：从GEO数据库中获取的两个数据集共识别出两个差异miRNA，ENCORI数据库验证差异miRNA的表达发现miR-32-5p的表达水平与GEO数据库的结果一致，故选择其进行研究；预测得到198个miR-32-5p 潜在的靶基因并鉴定出 10 个核心基因（DKK3、WNT2B、SFRP5、SFRP2、SFRP1、LRP6、WNT6、KREMEN1、NEDD4L、TRIP12），其中DKK3的度值最大可能在乳腺癌中较为重要，于是选择miR-32-5p/DKK3轴进行后续研究。miR-32-5p在3种乳腺癌细胞中的表达水平显著高于正常乳腺细胞（均P<0.01），其中以MDA-MB-231细胞中表达最高。双荧光素酶基因报告实验验证了miR-32-5p与DKK3基因的靶向结合及其对后者表达的负向调控。转染miR-32-5p mimic、miR-32-5p inhibitor后成功提高或抑制了MDA-MB-231细胞中miR-32-5p的表达。与对照组相比，过表达miR-32-5p可抑制MDA-MB-231细胞的凋亡而促进细胞增殖和侵袭（P<0.05或P<0.01），敲低miR-32-5p则起相反的作用（均P<0.01）。结论：miR-32-5p/DKK3轴可能是影响乳腺癌发生发展的关键通路，过表达miR-32-5p能够抑制乳腺癌MDA-MB-231细胞的凋亡而促进细胞增殖和侵袭。

Objective: To screen the key differentially expressed miRNAs and their target genes in breast cancer using bioinformatics tools, and to observe the effect of interfering their expression on the function of breast cancer cells. Methods: The GEO database was used to screen the differentially expressed miRNAs in breast cancer, and the ENCOR1 database was further used to verify the expression of screened miRNAs to identify the most differentially expressed miRNA as the research target. Starbase, miRDB and miRWalk databases were used to predict the target genes of miR-32-5p. GO analysis and KEGG analysis of target genes were done by using DAVID database. PPI network analysis and screening of hub genes were performed using String database and Cytoscape3.6.2 software. The Dickkopf-related protein 3 (DKK3) gene with the most significant degree was selected from the hub genes for follow[1]up experiments. qPCR was used to detect the expression of miR-32-5p in human normal breast MCF10A cells and human breast cancer cells (MCF7, MDA-MB-231 and MDA-MB-453 cells). MiR-32-5p mimics, miR-32-5p inhibitors and their control (NC) sequences were transfected into MDA-MB-231 cells. The effects of over-expression or inhibition of miR-32-5p on cell proliferation,apoptosis and invasion were detected by CCK-8 method, flow cytometry, and Transwell experiment, respectively. Results: Two differentially expressed miRNAs were identified from the two datasets in GEO database. ENCORI database was adopted to verify the differentially expressed miRNAs and showed that the expression level of miR-32-5p was consistent with the result in the GEO database, so it was selected for subsequent research. 198 potential target genes of miR-32-5p were predicted and 10 hub genes (DKK3,WNT2B, SFRP5, SFRP2, SFRP1, LRP6, WNT6, KREMEN1, NEDD4L and TRIP12) were identified, among which DKK3 showed the most significant degree. So, miR-32-5p/DKK3 axis was selected for the subsequent research. The expression of miR-32-5p in three breast cancer was significantly higher than that in normal mammary gland cells (all P<0.01), with the highest expression in MDA[1]MB-231 cells. Dual-luciferase reporter gene assay verified cell lines that miR-32-5p tageted and down-regulated. After transfection with miR-32-5p mimics or miR-32-5p inhibitors, the expression of miR-32-5p in MDA-MB-231 cells was successfully increased or inhibited. Compared with the control group, overexpression of miR-32-5p could inhibit the apoptosis of MDA-MB-231 cells and promote cell proliferation and invasion (P<0.05 or P<0.01), while knocking down miR-32-5p played an opposite role (all P<0.01).Conclusion: miR-32-5p/DKK3 axis may be a key pathway affecting the development of breast cancer. Over-expression of miR-32-5p can inhibit the apoptosis but promote cell proliferation and invasion of breast cancer MDA-MB-231 cells.

海南省自然科学基金青年资助项目（No. 818QN314）