Transcriptional Regulation Of Non-coding RNA In Human Cancer

Cancer is one of the most serious diseases that threaten human life and health,the global morbidity and mortality rates of cancers are higher than other diseases.Currently more than 100 kinds of cancers have been found.Studies have shown that non-coding RNAs,such as miRNAs,lncRNAs play important regulatory roles in carcinogenesis and cancer development.Therefore,the study of transcriptional regulation of non-coding RNAs in cancer has important significance for understanding and treating cancer.However,the classic molecular biology experiments are not enough to uncovering or understanding the transcriptional regulation mechanisms of non-coding RNAs when considering the large-quantity of non-coding RNAs and the complexity of regulatory network.In recent years,the development of bioinformatics algorithms and high-throughput sequencing technology has provided new theories and methods for the study of transcriptional regulation of non-coding RNA in cancer and new chances to study and discuss scientific questions related to cancer transcriptional regulation in a more comprehensive way.In this dissertation,we mainly studied the transcriptional regulation of miRNA and IncRNA in cancer.The first part of dissertation is to study the combinational regulation of miRNA and transcription factor in various cancer and normal tissue cell lines.First,we examined the relationship between the status of genes that whether the genes were regulated by miRNA or not and transcription factor binding site number of genes in the 19 human cell lines.The results not only confirmed that miRNAs preferentially targeted the genes with high transcriptional regulation complexity,but also implied that cancer and normal tissue cell lines have similar regulatory mechanisms in this respect.The result demonstrates the universal significance of combinational regulation of miRNA and transcription factor.The study of the location characteristics of transcription factor binding sites implied that whether a gene was regulated by miRNAs mainly depended on the transcription factor binding site information in gene TSS region.On this basis,we constructed a mathematical model which used all transcription factor binding sites' information as features to predict whether the gene is regulated by miRNAs,and achieved good performance.By applying this model to predict target genes of single miRNA,we found that the combinational regulation of miRNAs and transcription factors can provide a new idea for optimizing algorithms of microRNA-target recognition.The second part of dissertation is a comparative study of IncRNA-mRNA type ceRNA regulatory networks in cancer and adjacent tissues.First,we constructed ceRNA regulatory networks for cancer and adjacent tissues in 13 types of cancer,respectively.We found that the adjacent tissues had more ceRNA regulatory effects,and only a small percentage of ceRNA regulation was preserved during the carcinogenesis by comparing ceRNA networks between the cancer and adjacent tissues in the same type of cancer.With the help of functional analysis in ceRNA regulation that emerging or disappeared during the carcinogenesis,we found that the emerging ceRNA regulation always led to cell cycle fluctuations,however,the disappeared ceRNA regulation affected the Wnt signaling pathway and the apoptotic pathway in many types of cancers.The third part of dissertation is to build a human cancer metastasis database(HCMDB).First of all,we manually extracted the metastasis-related function information of miRNAs,lncRNAs and protein coding genes from thousands of publications,and we found totally 2183 genes were associated with cancer metastasis,including 1901 protein coding genes,24 lncRNAs and 203 miRNAs.We then collected 152 gene expression datasets related to cancer metastasis from the GEO and TCGA databases,and performed metastasis-related gene identification and regulation network construction based on more rational designs of sample grouping.Our work constructed the first database of gene expression and function related to cancer metastasis based on data mining of publications and high-throughput data,which provides data support for exploring the function of non-coding RNAs and protein coding genes in cancer metastasis.In a word,we made a tentative study on several cancer-related mechanisms of transcriptional regulation by non-coding RNAs and counstructed a database of gene expression and function related to cancer metastasis based on a large number of common transcriptome and epigenetic data,which may be of great significance to the study of non-coding RNAs in cancers.Although our work only uncovered a small part of knowledge for comprehensive understanding of cancer-related mechanisms of transcriptional regulation,but I hope the dissertation could provide meaningful ideas and data mining platform for the cancer-related research of transcriptional regulation in the future.