【作者】 徐娟；

【导师】 李霞；

【作者基本信息】 哈尔滨医科大学 ， 生物物理学， 2011， 博士

【摘要】 microRNA（miRNA）是一类重要的非编码RNA分子，长约22个核苷酸，通过在转录后水平上降解靶基因mRNA或抑制靶基因的mRNA翻译来发挥作用。它们在许多生物和代谢过程中起着非常重要的作用，比如细胞增殖，信号传导，发育和凋亡。因此，miRNA功能的失调能影响细胞生物的很多特性，最终导致疾病的发生，甚至是癌症。人类基因组可以编码多于1100miRNAs，这些miRNA能调控大部分人类基因。有限的miRNA能调控大部分编码蛋白的基因被认为是通过协同作用完成的，也就是多个miRNA可以一起协同控制单个基因。因此，miRNA间的协同调控对于理解复杂的转录后调控是重要的。复杂疾病是由多个miRNA共同作用导致的，而不是单个miRNA的结果。我们面临着一些挑战：如何识别miRNA的协调作用，从而进一步从系统水平上判断miRNA的功能；从miRNA-miRNA协同网络的角度调查疾病miRNA特征。另外，分析协同调控miRNA在序列和结构上的关联及它们转录调控机制的一致性也是必要的。这里，我们借助miRNA对共调控的功能模块构建了miRNA-miRNA功能协同网络（MFSN）。功能模块有三个特征：被miRNA对共调控，富集在同一个GO功能类中，在蛋白质互作网络中距离近。如果一对miRNA至少协同调控一个功能模块，我们就认为它们有功能协同作用。预测的miRNA协同作用通过功能模块的高共表达和miRNA对与功能模块的显著负调控所证实。miRNA功能协同调控关系相互交织，形成一个复杂的miRNA功能协同调控网络。和其他的生物网络性质类似，MFSN呈现无尺度，小世界和模块化结构。这些网络特性允许多个miRNA同时发挥调控作用，从而对干扰快速地做出响应。我们用不同的miRNA靶点数据构建miRNA-miRNA功能协同调控网络，发现这些网络的结构特性对是一致的。我们发现疾病miRNA在MFSN中的拓扑测度是不同于非疾病miRNA的。疾病miRNA间有更多的协同作用，表明它们有更高的功能复杂性。他们还倾向定位在包含miRNA比较多的模块中，特别是这些模块的交叠处，表明疾病miRNA倾向是MFSN的全局中心，对不同或相似生物过程起到衔接作用。另外，和同一疾病相关的miRNA在网络中距离近，暗示着同一疾病的miRNA调控相同或者相似的功能。在不同miRNA靶点数据构建的miRNA协同调控网络中，疾病miRNA的拓扑性质也是一致的。最后，我们系统分析了协同调控miRNA的序列、结构和转录调控特征。和非协同调控miRNA对比较，协同调控miRNA之间的序列和结构相似性显著高；特别是相似性高的协同调控miRNA大部分都来自同一miRNA家族，随着相似性阈值的减小，不同家族miRNA间协同作用所占的比例也增加。进一步地，模块内miRNA间序列和结构相似性比MFSN中所有协同调控关系的序列和结构相似性还要显著高。另外，协同调控miRNA的序列相似性和其功能一致性存在显著的正相关，因此协同调控miRNA的序列相似性在一定的程度上能反映其功能的一致性。虽然结构相似性和功能一致性没有显著的相关性，但是有些miRNA间的协同调控是可以通过结构相似来解释的。我们还发现存在协同作用的miRNA显著共享TF，并且这些miRNA对的转录调控一致性也比非协同调控miRNA对的显著高。与序列或结构的特征类型，转录一致性高的miRNA对基本都来自同一miRNA家族。另外，这些共享转录因子的miRNA倾向是miRNA协同调控网络的中心节点。总之，miRNA-miRNA功能协同调控网络不仅可以用来深入理解miRNA的转录后调控模式，还为探索疾病miRNA的性质提供了一个新的视角。更多还原

【Abstract】 MicroRNAs are endogenous non-coding (~22nucleotide) RNAs that regulate theexpression of target mRNAs at the post-transcriptional level. They play critical rolesin various biological and metabolic processes such as cell proliferation, signaltransduction, development and apoptosis. Thus, it is intuitive that the abnormality ofmiRNA function can deeply impact multiple features of cell biology, ultimatelyresulting in complex pathological events including cancers. Some researchers claimedthat the human genome might encode more than1,100miRNAs and it was recentlydemonstrated that they potentially regulate the majority of all human genes, which cantherefore be used to improve our understanding of the mode of action of miRNAs andtheir functions. The limited miRNAs are thought to be able to control the larger set ofgenes through synergism, in which multiple miRNAs work synergistically to controlindividual genes. Therefore, synergistic regulations among multiple microRNAs areimportant to understand the mechanisms of complex post-transcriptional regulationsin humans. Complex diseases are affected by several miRNAs rather than a singlemiRNA. So, it is a challenge to identify miRNA synergism, thereby further determinemiRNA functions at a system-wide level and investigate disease miRNA features inthe miRNA-miRNA synergistic network from a new view. It is also nessesary toanalyze the sequence or structrue association between miRNAs with synergism.Here we constructed a miRNA-miRNA functional synergistic network (MFSN) viaco-regulating functional modules which have three features: common targets ofcorresponding miRNA pairs, enriched in the same Gene Ontology category and closeproximity in the protein interaction network. If a miRNA pair co-regulates at least onefunctional module, they are considered to be synergism. Predicted miRNA synergismis validated by significantly high co-expression of functional modules andsignificantly negative regulation to functional modules. The MFSN exhibits scale-free,small-world and modular architecture, so it has many advantages of structure as general biological networks such as protein interaction networks or friendshipnetworks in sociology. The structure of the MFSN is validated to be robust usingdifferent miRNA target datasets.Furthermore, the topological features of disease miRNAs in the MFSN aresignificantly distinct from non-disease miRNAs. They have more synergism whichindicates their higher functional complexity. Most disease miRNAs are located in theoverlap region of most tightly connected communities, as the global central cores ofthe MFSN. In addition, miRNAs for the same disease are close with each other in theMFSN to regulate targets with the same or similar functions. The topological featuresof disease miRNAs are validated to be robust using different miRNA target datasets.Finaly, we systemically analyzed the association between miRNAs with synergismat the view of sequence, tructure and transcriptional regulation. We found thatsynergistic miRNAs have significantly higher similarity of sequence or tructure thannon-synergistic miRNA pairs. Especially synergistic miRNAs pairs with the highestsimilarity are mostly from the same miRNA families, with decresing the thresholdvalue of similarity, the proportion of synergism between miRNAs from differentfamilies’ increases. The similarity of miRNA synergistic modules is significantlyhigher than the general level of the MFSN. In addition, the sequence similarity ofsynergistic miRNAs is significantly positive with their functional concordance, so thesequence similarity might reveal the functional association between miRNAs, to acertain extent. Althogh the structure similarity is not significant associated with thefunctional association, we found several miRNA synergism could be explained by it.We also found that synergistic miRNAs significantly share transcriptional factors(TFs), and their regulational concordance is higher than non-synergistic miRNA pairs.Most miRNA pairs with highest regulational concordance are from the same miRNAfamilies. At last, miRNAs sharing TFs tend to be the centre of the MFSN.Overall, miRNA-miRNA functional synergistic network not only increases our understanding of their mechanism in post-transcriptional regulation, but also providesa new insight into the global topological properties of disease miRNAs.更多还原