Systematic Analysis Of MiRNA Related Variants At DNA And RNA Level

MicroRNAs (miRNAs) are a family of approximately22nucleotides in length endogenous small noncoding RNAs. They are able to bind to the mRNA3’untranslated region (3’UTR) complementary thereby inhibitingmRNA translation or mediating mRNA decay. As a crucial class of post-translational regulators, miRNAs are involved in various developmental and physiological processes in animals and plants and also have been reported to play important roles in diseases. MiRNA nucleotide changes can happen at the DNA level (Single-nucleotide polymorphism, SNP) or at the RNA level including RNA editing and RNA addition. The nucleotide changes, variations of miRNA splice sites or expression differences are all likely to affect the miRNA function, thereby affecting a variety of biological processes. However, most of the studies on miRNA function influencing factors were based on a few of candidate miRNAs at single tissue or single sample level, systematically analyzing this at genome-wideand multi-organizations level will be able to deepen the understanding of miRNA function.In this study, by using bioinformatics tools, we integrated abundant of public data to study the miRNA functional variations.At the DNA level, we systematically analyzed miRNA related SNPs and their effects.At the RNA level, we analyzed the features and functions of several miRNA variations, such as RNA editing, RNA addition and splicing site shift. We have obtained the following important findings.Initially, using the latest data from public databases, we identified the SNPs which can influence miRNA expression or (and) miRNA/target binding.By comparing the location of miRNA genes with SNP location, we found a total of757SNPs in the human miRNA genes,50SNPs of which are in human miRNA seed regions and225,759SNPs in gene3’UTR regions. SNP density analysis showed that miRNA genes are more conservative than flanking regions, seed regions are especially conservative and conserved miRNA and miRNAs in clusters tend to have fewer SNPs. Combining two miRNA target prediction methods, we analyzed the affect of SNPs on miRNA/target gene binding. The results showed that SNPs in miRNA seed regions have enormous impact on the miRNA, which may not only affect the yield of the miRNA, but also greatly change the target profiles. Among those SNPs in the3’UTR regions, we identified98,008SNPs which can either disrupt the original miRNA target sites or creating novel miRNA/target binding sites, thus affecting the regulatory function of the miRNA.Furthermore, we experimentally confirmed seven loss-of-function SNPs and one gain-of-function SNP by luciferase assay. This is the first case to prove that a SNP in miRNA will create a novel miRNA target binding by experimental validation. All useful data about these miRNA and related SNPs were complied into miRNASNP, a user-friendly free online database (http://www.bioguo.org/miRNASNP/).Finally, we completed a meta-analysis of miRNA variant by processing410high-quality miRNA sequencing datasets, which includes more than20different organs or diseases. Weestablished a standard pipeline to process miRNA sequencing data.The results showed that a large number of miRNA isoforms exists in the sequencing data except canonical miRNAs. These isomiR may be due to the RNA editing, addition, splice sites change, and sequence degradation and so on. Further analyses showed that:miRNA addition mainly contains A and U addition; except for A->G editing, other types of editing also exist; the change of splice site may cause a5’ shift of the miRNA sequence, thus change the seed regions; miRNA degradation causes lots of shorter isomiR to exist. In addition, results show that the expressions of70%of the known human miRNAs are very low in each sample, whilethe expressions of21%(428) of miRNAsare in middle level and only9%(184) miRNAs express at high level. This suggests that researchersneedto pay more attention to miRNA expression to avoid introducing non-expressed miRNAs in a tissue or disease before they scan gene related miRNAs or perform network analysis.Our study provided a useful resource for studying miRNA function, related SNP and other miRNA influencing factors.