The HPV-associated Whole Genomics Based On High-throughput Analyses And The Effects And Mechanisms Of Specific MiRNAs During Cervical Carcinogenesis

Cervical cancer (CC) is one of the most common cancer in women worldwide and exhibits a causal relationship to infection by high-risk human papillomaviruses (HR-HPVs). It displays the unique feature of well-defined clinical stages that are associated with different steps of tumourigenesis. Thus, cervical malignant abnormalities present a unique opportunity to profile neoplastic transformation, and it takes a good chance to prevent or target cervical intraepithelial neoplasia (CIN). However, there is still no druggable therapies except surgery treatments. The concept of translational study which is the essence of our study inspired us to identify crucial targets or pathways using clinical models of progressive disease, and then to guide clinical applications.The start point of this study is to profile miRNA expression in HPV-infected samples and relate this to histological and grade-specific alterations in the spectrum of cervical carcinogenesis, which facilitate the recognization of miRNAs as candidate biomarkers or targets. Since regulation of gene expression by miRNAs is a fundamental mechanism for controlling many biological processes, not only gene abundance but also gene structures such as splicing isoforms, it is necessary to integrate data of transcriptomes with miRNAomes. The advent of next-generation sequencing (NGS) technologies combining with microarray analyses provides new opportunities to deeper biological insights into the miRNA targeting mechanisms as evidenced by in vitro 3'-UTR assay validation and in vivo genome-wide miRNA target prediction. It would build a solid foundation for the whole RNAomics study of HPV infection and cervical carcinogenesis. In addition, the reproductive tract in female is a complicated ecosystem composed of various microbiome communities, it is unclear whether there are clinical associations between persistent HPV infection and dysbiosis, or there are any infective pathogens that might reinforce persistence or malignant transformation of HPV infection. Even for the virology study of HPV there are limited data about HPV genomes or transcription maps in vivo, since HPVs are non-cultivable viruses and could not be efficiently propagated in tissue cultures. As phi29 polymerase-mediated rolling cycle amplification (RCA) allows the efficient amplification of whole genomes independently of microbiome cultivation, the deep sequencing of the RCA product could be used for whole genome analysis. It would contribute to systematically study on HPV genomes, expression patterns, phylogenetic relationships and associations with diverse microbiomes in the reproductive system. The concept of metagenome would be introduced into the mechanism study of HPV infection and cervical carcinogenesis for the first time.Our study aimed to identify candidate biomarkers or druggable targets of HPV infection and cervical cancer. Whole genomic analyses of miRNAome, trancriptome and metagenome in cervical tissues were performed through microarray or next-generation sequencing (NGS) technologies. Then functional studies of specific miRNAs were followed based on miRNA expression profiling and bioinformatic analyses, and would provide a theorotic basis for the clinical application. From this study, we constructed a well-established system represented by miRNAs to pave the way for translational study of cervical cancer. ObjectiveTo identify candidate biomarkers or druggable targets of HPV infection and cervical cancer based on the high-throughput analyses of miRNAome, transcriptome and metagenome. It would set a solid foundation for the following functional studies and pave the way for translational study of cervical cancer (CC).MethodsFirst, we profiled miRNA expression signatures in various cervical tissues by microarray analysis, including 6 normal cervical tissues without HPV infection,7 normal ones with HPV 16 infection,6 CIN2-3 tissues with HPV 16 infection and 6 CC tissues with HPV 16 infection. Differentially expressed miRNAs were identified by significance analysis of microarray (SAM) and in-depth analysis. A network of "HPV related miRNAs/target gene" was established through gene ontology (GO) analysis and a series of prediction programs to identify miRNA targets. Then,6 miRNAs (miR-29a↓,miR-375↓,miR-195↓,miR-99a↓,miR-155↑,miR-92a↑)were validated in 133 cervical tissues to confirm the consistency between microarray and Real-time PCR analysis.Second, directional ligation sequencing (DeLi-seq) technology was applied to present NGS-based qualitative and quantitative evaluation of the HPV and cellualar transcriptome in 3 normal cervical tissues without HPV infection,4 CIN2-3 tissues and 3 CC tissues with HPV 16 infection.Finally, we used rolling cycle amplification (RCA) to enrich short circular DNA and amplificate long linear DNA fragments in 20 cervical tissues (10 CIN2-3 and 10 CC). Then we constructed genomic DNA libraries sequenced by Illumina Genome AnalyzerⅡ. All the raw data were analyzed by effective bioinformatic algorithms and tools. Through mapping all sequencing reads to 132 annotated HPV genomes, genomic information about HPV DNA in the clinical samples were revealed, including viral copies, HPV physical statuses (episomal or integration, simple infection or mulitiple infection), nucleotide mutations, integration sites, and et al. Also we could identify if there were novel HPV genotypes or subtypes, or other known or unknown pathogens in cervical environment from de novo assembly strategies.Results1. There were 52 miRNAs differentially expressed among four groups, including normal cervical tissues without HPV infection, normal ones with HPV 16 infection, CIN2-3 with HPV 16 infection and CC with HPV 16 infection, by ANOVA analysis (p<0.05).31 miRNAs were differentially expressed among three groups including normal cervical tissues without HPV infection, CIN2-3 and CC with HPV 16 infection (ANOVA analysis, p<0.05).2.6 miRNAs(miR-29a,miR-375,miR-195,miR-99a,miR-155,and miR-92a)were validated by Real-time PCR in 133 cases of cervical tissues. PCR results were quite consistent with microarray data, and provided detailed information about expression signatures of the 6 miRNAs in three different groups including normal cervical tissues without HPV infection,CIN2-3 and CC tissues with HPV 16 infection.3.All the 31 differentially expressed miRNAs were annotated by bioinformatic gene ontology (GO) analysis. Combined with miRNA target database and GENE database (NCBI), candidate HPV related miRNA targets were recognized and HPV-associated miRNA-mRNA regulatory networks were subsequently established. 4. Whole cervical transciptome and metagenome were analyzed by DeLi-seq and deep sequencing based on RCA. Dynamic expression abundance and transcript structures of both HPV and cellular genes were profiled, and whole genomic mutation and transcription maps of HPV 16 and HPV58 were present.Conclusions1. There are 52 miRNAs differentially expressed in various cervical tissues at different HPV infection stages. A total of 31 ones show significant and continuous expression along with progression from normal cervical tissue to cancer, and miR-375,29a,99a,195 are in a down-regulating trend, while miR-92a,155 are in an upregulating one. Our data suggested that the two platforms used in this study showed similar results, with qRT-PCR being more prominent.2. From our data, a putative HPV-associated miRNA-mRNA regulatory network is established, suggesting that miR-29 is the most highly enriched. Based on GO analysis of the 31 differentially expressed miRNAs, the high-enrichment GO terms are cell cycle, apoptosis, proliferation and et al.3. The transriptomic and metagenomic analyses present dynamic and genomic changes of expression abundance and transcipt structures of HPV and cellular trancripts. The genomic mutation and transcription maps of HPV 16 and HPV58 are constructed, which would contribute to the following functional studies. ObjectiveTo determine that miR-29 could target the HPV related genes YY1/CDK6 directly and plays important roles in the regulation of cellular biological behaviors by a series of correlation analyses in cervical tissues in vivo and functional analysis in vitro. Also to examine the pros and cons of relationships between HPV infection and specific miRNA abnormality, including miR-29 and miR-224.MethodsExpression of YY1/CDK6 was examined by Real-time PCR and immunohisto-chemistry, and correlations between miR-29 were determined by Spearman's correlation analysis. Then we assessed the restored expression of miR-29 on YY1/CDK6 gene expression by Real-time PCR and Western blot, and observed changes of cell cycle and apoptotic rates by flow cytometry in SiHa cells. Next, we examined the pros and cons of relationships between HPV infection and miR-29 or miR-224 abnormality. Correlations between miR-29 and E6/E7 were analyzed in CIN2-3 and CC tissues with HPV 16 infection based on Real-time PCR. Expression of miR-29 in normal cervical tissues with or without HPV infection was examined by Real-time PCR in order to get some clues. We assessed E6/E7 expression level in SiHa cells transfected with miR-29 mimics by Real-time PCR and Western blot; also miR-29 expression level in SiHa cells transfected with siRNAs targeting E6/E7 or in C33A cells with E6/E7 expression plasmids. As for miR-224, we determined its expression level in various cervical tissues and raft culture tissues infected with HPV by Real-time PCR and Northern Blot. Correlations between miR-224 and E6/E7 were analyzed. Through knockdown or overexpression of E6/E7 we observed expression changes of miR-29, and vice versa. Biological behaviors such as cell cycle and apoptotic rate were assessed by flow cytometry. Also we tried to identify target genes and pathways regulated by miR-224 in cervical carcinogenesis by a set of functional studies.Results1. It was shown that expression of YY1/CDK6 was negatively correlated with miR-29 in cervical tissues. There were significant differences of expression levels of YY1/CDK6 among different groups, and uniformly progressive upregulation of YY1 and CDK6 was determined both at mRNA and protein levels.2. Restored miR-29 reduces YY1 and CDK6 expression in SiHa cells at mRNA and protein levels. Transferred miR-29 triggered a significant accumulation of cells at the G1 phase and increased apoptotic rate in comparison with blank and negative control cells.3. We demonstrated a strongly inverse correlation of miR-29a expression with E6 and E7 expression in CIN2-3 and CC tissues with HPV16 infection. Neither ectopic expression of miR-29 could influence E6/E7 mRNA or protein levels, nor transfection with siRNAs that targeting HPV16 E6/E7 could change miR-29 expression level. There was no statistic difference of miR-29 expression between the osteosarcoma cell line U2OS expressing E6 gene (U2OSE6tet24) and control U2OS cells, nor between SiHa/CaSki cells and HPV-negative C33A cells. Moreover, there was no change of miR-29a in C33A cells transfected with E6 genes.4. There was a dynamic and bidirectional expression pattern of miR-224 during HPV infection in HPV infected cervical tissues in vivo and raft culture tissues in vitro. But there were no directional regulations between miR-224 and HPV E6/E7 in further functional studies in vitro.Conclusions1. Our data indicates that miR-29 restrains cell cycle progression and induces apoptosis via YY1 and CDK6, promoting malignant transformation induced by HPV and involved in the process of cervical carcinogenesis. we hypothesize that miR-29 could be potential tumor suppressors, while YY1 or CDK6 would serve as candidate HPV-induced oncogenes in cervical cancer.2. We find that there is a dynamic expression signature of miR-224 during cervical carcinogenesis, though the abnormality of miR-224 in HPV-infected cells might be regulated in an indirect way.