microRNA相关的生物信息学与进化分析

【作者】 杨桢；

【导师】 钟扬；

【作者基本信息】 复旦大学 ， 生物信息学， 2011， 博士

【摘要】 microRNA (miRNA)是一类长度大约为22个核苷酸左右的内源性非编码RNA分子,通过与mRNA互补结合抑制蛋白翻译或导致mRNA降解,干扰mRNA指导蛋白合成过程,从而负向调节蛋白编码基因的表达。作为一种重要的基因表达调控因子,miRNA在多种基因表达调控途径中发挥重要作用,调节胚胎发育、细胞分裂与分化、细胞凋亡、造血作用以及癌症发生在内的多个细胞过程。目前在人类基因组中已1000条miRNA基因。对于miRNA分子特点、转录机制、进化以及对已成为分子生物学与生物信息学研究的热点。在本文中,我们利用生物信息学及统计学方法,对包括癌症相关的miRNA筛选、miRNA对靶基因的表达调控模式、内含子编码的miRNA基因特征及miRNA基因家族的进化等问题进行了初步研究。miRNA与人类疾病关系是目前生物医学领域的一个研究热点,越来越多的证据表明,多种miRNA的突变或表达异常可能与癌症的发生密切相关。因此,miRNA可能成为一种用于癌症诊断与治疗的潜在分子标记。为了整合已有的癌症相关miRNA资源,我们从Gene Expression Omnibus数据库中搜索出miRNA在癌症中的表达谱数据,并利用芯片显著性分析方法挖掘出在miRNA在癌症组织中差异表达的信息,最终获得607个癌症相关的miRNA。我们同时设计了一个界面友好的数据库dbDEMC (a DataBase of Differentially Expressed miRNAs in human Cancers),以方便用户访问及获取相关信息。挖掘癌症相关miRNA为癌症相关研究提供了新的潜在分子靶标。miRNA对于靶基因的调控模式是另一个研究热点,miRNA对于靶基因可能存在两种不同的调控模式,即调节靶基因平均表达水平和降低靶基因表达波动性。我们从人类全基因组表达数据出发,首先从Gene Expression Omnibus数据库中选取多套基于HGU133 plus2.0芯片平台的蛋白编码基因表达谱,通过Meta-分析方法筛选出数百个稳定表达基因及波动表达基因。在此基础上利用miRNA靶基因预测软件分析了miRNA对两组基因的调控偏好性,分析表明miRNA更倾向于调控稳定表达基因,即miRNA可增强靶基因表达稳定性。此外,对于转录因子对靶基因的调节作用分析表明,转录因子和miRNA对于基因表达稳定性具有相反的效应。这一研究对揭示基因表达调控机制差异背后的内在机制提供了新的线索。已有研究表明,miRNA基因多数位于蛋白编码基因间区域,独立转录而行使功能。而某些miRNA基因可能位于蛋白编码基因的内含子内,其转录过程可能利用宿主基因的某些调控元件,因而表达状况可能与宿主基因存在一定关联。内含子编码的miRNA基因由于其表达调控的特殊性而越来越引起人们的兴趣。我们对编码miRNA的宿主基因特征进行了分析,发现miRNA宿主基因长度普遍高于非宿主基因,其原因可能为宿主基因内含子与外显子数量的显著增加。此外,对内含子编码的miRNA基因的系统发育分布分析后发现,其数量在脊椎动物,真兽类哺乳动物及灵长类动物形成过程中出现过三次爆发性增长。最后,我们以一个脊椎动物中特有的miRNA基因家族miR一181家族为例,分析了高等动物基因组中miRNA基因家族的起源与进化历程。该基因家族在低等尾索动物中只存在单基因拷贝,而在高等脊椎动物基因组中多存在六个拷贝。我们发现,该miRNA基因家族在脊椎动物基因组中由祖先基因经历两次整体复制及一次片段复制而成。对该基因家族调控的靶基因功能分析后发现,该miRNA调控的靶基因功能主要涉及基因表达调控、细胞间通讯、细胞生长、分裂、增殖等方面。更多还原

【Abstract】 microRNAs (miRNAs) are endogenously expressed small noncoding RNAs with transcripts ranging between 18-23 nt in length, they can regulate the expression of protein-coding genes through translational inhibition or processing mRNAs by base pairing to mRNAs 3’-untranslated regions (3’-UTRs). As an important kind of gene expression regulator, profound influences of miRNAs on diverse regulatory pathways, such as embryo development, cell division, differentiation, apoptosis, hematopoiesis and cancer development, are widely recognized. At this time, more than one thousand miRNA genes have been identified in the human genome. The properties, expression mechanism, evolution miRNAs have turned into the hotspots of molecular biology and bioinformatics. In this analysis, we mainly concentrate on the cancer related miRNAs, expression regulation mode of the target genes, intronic miRNAs and evolution of specific miRNA family, carried out some initial research by bioinformatic and statistical tools.Recently, more and more evidence suggests that aberrant expression of miRNAs may contribute to many types of human diseases, including cancer. miRNAs can be a novel kind of biomarker for the digonesis and treatment. For integrate the expression information about cancer related miRNAs, we first searched for the miRNA expression profiles in various kinds of human cancers in the Gene Expression Omnibus (GEO) database, and then used the method of Significance Analysis of Microarrays to screen the differentially expressed miRNAs in human cancers. Totally 607 cancer related miRNAs were finally identified. We also designed and constructed a database, called the dbDEMC, a database of differentially expressed miRNAs, to explore aberrantly expressed miRNAs among different cancers. This database is expected to be a valuable source for identification of cancer-related miRNAs, thereby helping with the improvement of classification, diagnosis and treatment of human cancers.The expression mode of miRNAs on target genes turned out to be another hotspot of bioinformatics. It is suggests that there are two different kind of gene expression mode of microRNAs:adjust the average expression level and/or reduce the fluctuation of target genes. We first screened out more than one hundred expression profiles based on the HGU133 plus2.0 microarray platform in the GEO database, and then filtered out different group of genes that have different expression mode, including the steady expressed genes and fluctuate genes, the preferential regulation of miRNA regulation on these genes was analyzed. We found that miRNAs preferentially regulate the steady expressed genes, which suggests they can enhance the stability of the expression of target genes. Further more, we found opposite effect between transcription factors and miRNAs. This analysis provide novel clue for understanding the mechanism of gene expression regulation.Most miRNAs genes locate in the intergenetic region, whereas some of them locate in the intronic region of protein coding genes. Their transcription may depend on the transcriptional elements of the host genes. Here we also analyzed the characteristics of the host gene of miRNAs. We found that the lengths of host genes are larger than the non-host genes, this is due to the increased number of their introns and exons. The phylogenetic distribution analysis of the intronic miRNAs among various species demonstrated that intronic miRNAs genes mainly emerged at three different time point, including the origin of vertebrate, eutheria and primate.At last, we use a vertebrate specific miRNA family, the miR-181 as an example, to analysis the origin and evolution of miRNA gene families in vertebrate genomes. There is only one copy of the miR-181 family was found in the urochordata genome, whereas there are six copies of this family in the vertebrate genome. We found that it is the whole genome duplication for twice and tandem duplication between them that form this family in the vertebrate genome. The functional analysis of the target genes of this family demonstrates that they mainly involved in the expression regulation, inter-cellular signaling, cell growth, division and proliferation.更多还原