数据挖掘在NDRG2转录调控和信号转导研究中的应用

发布时间：2018-08-15 16:18

【摘要】：近年来，高通量技术在生物医学研究中的应用导致了大量生物数据的积累。这些数据涉及基因与蛋白质序列，DNA微阵列和生物医学图像等。为了利用这些数据发现可应用于生物医学研究的信息，数据挖掘技术得以开发和发展。在生物医学研究中，数据挖掘技术发挥了越来越重要的作用，并已成为生物医学发现过程的一部分。在我们的研究中，我们应用数据挖掘技术，首次从HepG2细胞的表达谱芯片中提炼出有效信息，创新性的预测了NDRG2的转录调控因子及NDRG2与Dusp6的相互作用分子。由于NDRG2在肿瘤细胞中的表达是降低的，为了全面理解NDRG2对肿瘤细胞的生物学效应，我们用表达谱芯片检测了HepG2细胞的基因表达，并对芯片数据进行了富集度分析。GO生物学过程分析表明参与G蛋白信号转导的基因表达增加了。我们选择其中5个基因进行qRT-PCR验证。而参与细胞M期的基因则降低了，这与细胞周期的分析结果一致。信号通路分析表明与血细胞分化和细胞粘附有关的分子表达明显增加，与蛋白的GPI修饰，蛋白降解和细胞分泌相关的基因表达降低。进一步通过模体分析和实验验证，我们发现NDRG2可以增加p38的磷酸化水平。通过富集度分析，我们成功从芯片数据中提取了有效信息，并为理解NDRG2在肿瘤细胞中的作用机制提供分子基础。为了理解NDRG2在不同条件下的表达模式，我们创新性的使用ARACNE算法和模体扫描预测了调节NDRG2表达的转录因子。通过模体扫描，我们发现在NDRG2启动子区域有129个转录因子的结合位点。为了提高预测的精确性，我们用ARACNE算法来估算NDRG2与这些转录因子的相关性。最终，我们得到了53个可能调节NDRG2基因表达的候选转录因子。在这些转录因子中，由于KLF4可以诱导结肠癌细胞分化，我们选择KLF4进行实验验证。对这些转录因子的功能分析表明，它们主要与细胞分化，器官发育和细胞内物质的运输和定位有关。这与之前的研究结果是一致。最后，为了发现NDRG2的相互作用分子，我们搜索了NDRG2在拟南芥中的同源物，并鉴定出NDL1,NDL2和NDL3为NDRG2的同源蛋白。这些NDL蛋白曾被报道可以与AGB1和RGS1相互作用，这提示NDRG2可以与这两个分子的人的同源物相互作用。AGB1为G蛋白三聚物的β亚单位。在人中，有5个AGB1的同源物，即GNB1—GNB5。RGS1为一GTP酶激活物，其可以与G蛋白三聚物的α亚单位相互作用,激活其活性使GTP发生水解，从而抑制G蛋白信号通路的转导。尽管只在拟南芥中只有一个RGS蛋白，但是，在人中却有20个左右的RGS蛋白。在这些蛋白中，RGS5与拟南芥中的RGS1的序列最相似，因而用于进一步的实验验证。通过免疫共沉淀实验和His-pulldown实验，我们发现RGS5可以与NDRG2相互作用。为了建立一种可以预测某一蛋白相互作用分子的方法，我们首次联合基因表达谱数据和蛋白质序列特征预测了Dusp6的相互作用分子。我们用MINDy算法来发现Dusp6的调节蛋白，用Pred_PPI来提高预测的准确性。通过这一方法，，我们成功预测了Mapk8为Dusp6的相互作用分子。总之，通过数据挖掘和实验验证的结合，我们成功建立了一种可以用来研究某一特定基因的转录调控和其参与的信号转导的策略。由于这些方法节省时间和资源，我们期望它们将来被用于其它基因的研究当中。
[Abstract]:In recent years, the application of high-throughput technology in biomedical research has led to the accumulation of a large number of biological data, including gene and protein sequences, DNA microarrays and biomedical images. In our research, we used data mining technology to extract effective information from HepG2 cell expression profiles chip for the first time, and innovatively predicted the transcriptional regulatory factors of NDRG2 and the phase of NDRG2 and Dusp6. Interaction molecules.
Because the expression of NDRG2 in tumor cells is decreased, in order to understand the biological effect of NDRG2 on tumor cells, we detected the gene expression of HepG2 cells with expression profiling chip, and analyzed the enrichment of the chip data. GO biological process analysis showed that the expression of genes involved in G protein signal transduction increased. Five of these genes were identified by qRT-PCR, while the genes involved in M phase were decreased, which was consistent with the analysis of cell cycle. Signal pathway analysis showed that the expression of genes related to blood cell differentiation and cell adhesion increased significantly, and the expression of genes related to protein GPI modification, protein degradation and cell secretion decreased. We found that NDRG2 could increase the phosphorylation level of p38 through the analysis of motifs and experiments. Through enrichment analysis, we successfully extracted effective information from the chip data and provided a molecular basis for understanding the mechanism of NDRG2 in tumor cells.
In order to understand the expression pattern of NDRG2 under different conditions, we used ARACNE algorithm and phantom scan to predict the transcription factors that regulate the expression of NDRG2. Through phantom scan, we found that there were 129 binding sites of transcription factors in the promoter region of NDRG2. Finally, we obtained 53 candidate transcription factors that might regulate the expression of NDRG2 gene. Among these transcription factors, KLF4 was chosen to induce colon cancer cells to differentiate. Functional analysis of these transcription factors showed that they were mainly related to cell differentiation and organogenesis. The transport and localization of substances in cells are consistent with previous studies.
Finally, in order to find the interaction molecules of NDRG2, we searched for the homologues of NDRG2 in Arabidopsis and identified NDL1, NDL2 and NDL3 as the homologous proteins of NDRG2. These NDL proteins have been reported to interact with AGB1 and RGS1, suggesting that NDRG2 can interact with the homologues of these two molecules in humans. AGB1 is a G protein trimer. In humans, five AGB1 homologues, GNB1-GNB5.RGS1, are GTP enzyme activators that interact with the alpha subunit of G-protein trimer, activate its activity to hydrolyze GTP and thus inhibit the G-protein signaling pathway. Although there is only one RGS protein in Arabidopsis, there are about 20 in humans. RGS protein. Among these proteins, RGS5 is the most similar to RGS1 in Arabidopsis and is therefore used for further experimental verification. Through immunoprecipitation and his-pulldown experiments, we found that RGS5 can interact with NDRG2. To establish a method for predicting a protein-protein interaction molecule, we first combined the bases. Interacting molecules of Dusp6 were predicted based on the expression profile data and protein sequence characteristics. We used MINDy algorithm to find the regulatory proteins of Dusp6, and used Pred_PPI to improve the accuracy of prediction.
In conclusion, by combining data mining with experimental verification, we have successfully developed a strategy that can be used to study the transcriptional regulation and signal transduction involved in a particular gene.
【学位授予单位】：第四军医大学
【学位级别】：博士
【学位授予年份】：2012
【分类号】：R363

【共引文献】