The Dynamic Expression Of Circular RNA And Its Application In Disease Diagnosis

Circular RNA(circ RNA)is a type of non-coding RNA with a covalently linked loop structure.It is produced by a back-splicing event in the nucleus,which is then transported and enriched in the cytoplasm.It is ubiquitously expressed in various biological tissues.The biological function of most circ RNAs are not yet clear.However,the tissue and temporalspecific expression of circ RNAs suggests they may play an important role in certain biological processes.In this thesis,we performed an in-depth investigation on the dynamic expression of circ RNAs.We explored the correlations between the dynamic expression pattern of circ RNAs and tissue phenotypes using the Rat Body Map data.Furthermore,we investigated the potential of using circ RNA expressions in human peripheral blood as a diagnostic marker for human diseases.The main results of this thesis consist the following three parts.· A model-based tool to quantify the expression of circ RNAs from RNA-seq dataThe very first step in the study of circ RNA dynamic expressions is to estimate the expression level of circ RNAs in biological samples.Conventional bioinformatics tools took the number of reads that supported back-splicing junctions as the expression level of circ RNA,which had inferior precisions.Although model-based quantification tools have been used to accurately estimate the relative expression abundance of linear m RNAs from RNA-seq data,they are not applicable to circular transcripts.To fix this gap,we presented a method to estimate the relative expression level of circ RNAs using existing model-based framework.To include circular transcripts in the framework of expression quantification,we transformed a circular transcript to a pseudo-linear transcript and treated circular transcripts as linear transcripts in expression quantification phase.Reads from the circular RNA would be mapped to the corresponding pseudo-linear transcripts,which assists the quantification of the original circular transcripts.The application on simulated RNA-seq datasets confirmed that our proposed method can effectively obtain the accurate estimation of the expression level of both linear and circular transcripts.The expression level of linear RNAs was estimated more accurately as well,since circular transcripts were often ignored in previous tools.Using the empirical ribo-depleted RNA-seq data,we showed that,in comparison to several existing circ RNA quantification tools(CIRI,KNIFE,circ RNA_finder),the new method had the strongest correlations with the expression level of circ RNAs determined by q RT-PCR.On the other hand,our method had the smallest standard deviation of expression levels when estimating the circ RNA expression level from RNA-seq data of biological replicates.This suggests that the method we proposed here had a superior performance than existing tools.· The dynamic expression of rat circ RNAs and its relation to tissue phenotypesFirst,we systematically analyzed the expression levels of both circular and linear transcripts using the newly developed tool on the rat Body Map dataset,which consists of samples from 11 organs at 4 different developmental stages.We first constructed a complete repertoire of rat circ RNAs,and profiled the expressions of the whole rat transcriptome in rat Body Map dataset.We found that sequence in circ RNAs were more conserved.In comparison to linear RNAs,circ RNAs had much higher tissue specificity.When investigating the relationship between the dynamic expression pattern of circ RNAs and its corresponding biological function,we found a strong correlation between tissue-specific circ RNAs and tissue's biological function.The temporal-specific pattern of circ RNA expressions showed that circ RNA expression level in the testis changed dramatically with time.It rapidly accumulates with the progress of sexual maturation,which is followed by a sharp decay when aging.The expression pattern of circ RNAs,which fluctuates in time,is closely related to the spermatogenesis process.Furthermore,there was no correlation between the expression level of dynamic expressed circ RNAs and their cognate linear RNAs.This indicates that circ RNAs may perform important biological functions,which is independent of linear RNAs,in physiological processes like spermatogenesis.· The potential of using human PBMC circ RNA expression for tuberculosis diagnosisSince the dynamic expression of circ RNAs has important biological functions,we hypothesized that changes in the expression of circ RNAs can reflect the state of human diseases.Since the early diagnosis of active tuberculosis is clinically important,the search for diagnostic markers of pulmonary tuberculosis from peripheral blood has important values in clinical application.We explored the possibility of using the changes of circ RNA expression in human PBMCs as a diagnostic marker for tuberculosis.Analysis on RNA-seq data from different components in human peripheral blood revealed that circular RNA was abundantly expressed in the main components of peripheral blood.Also,there was a significant difference in circ RNA expression level in PBMC between tuberculosis patients and normal controls.In addition,we found several pathways that are related to the development and progression of tuberculosis.We selected seven circular RNAs from those KEGG pathways,and used them as a molecular signature for TB diagnosis.We validated this molecular signature in a validation cohort using q RT-PCR.We calculated the tuberculosis index(TB index)based on the expression level of those seven circ RNAs,and discriminated TB patients from healthy controls using TB index.The diagnosis power of this molecular signature was shown to be high(AUC = 0.946).This indicates that circ RNA expression in human PBMCs can be used as a biomarker for the diagnosis of active tuberculosis.In summary,we developed a new quantification method for circular RNA with improved accuracy and stability.Using rat Body Map dataset,we then observed a strong relationship between the dynamic expression pattern of circ RNAs and tissue's physiological function.Finally,we suggested the diagnosis power of using human PBMC circ RNA expression for human diseases.All these would lay a foundation for the subsequent analysis and application of circ RNAs.