The Mechanism And Functional Study Of C-Myc-mediated Epigenetic Silencing Of MiR-101in Hepatocellular Carcinoma

【Background】MicroRNAs (miRNAs) constitute a new class of small non-coding RNAs, whichwidely exist in eukaryotic cells. Global down-regulation of miRNAs represents a generalfeature of human cancers. Through sequence-specific interactions with the3’-untranslatedregions (UTRs) of cognate mRNA targets, miRNAs suppress gene expressionpost-transcriptionally via both translational inhibition and mRNA degradation.Interestingly, most miRNAs are highly conserved throughout evolution, suggesting thatthey might play crucial roles in multiple key biological processes. The human genomeencodes more than one thousand miRNAs, which may target about1/3of mammaliangenes and participate in almost all cellular pathways. Therefore, the global down-regulation of tumor-suppressive miRNAs in HCC will contribute to the activation oflarge amounts of oncogenes which are important for hepatocarcinogenesis. It is worthnoticing that while great efforts have been made in uncovering differentially expressedmiRNAs between malignant and normal tissues, and then investigating their functionalroles, much less insight has been gained into the regulatory mechanisms underlying theiraltered expression. If we can elucidate the mechanisms for the widespread repression oftumor-suppressive miRNAs in HCC, it is possible to reverse this process, thus paving anew road for the treatment of liver cancer.Histone methyltransferase EZH2plays a critical role in the initiation and progressionof many human cancers. Besides its ability to silence diverse tumor suppressor genes,EZH2can also suppress multiple tumor-suppressive miRNAs. However, the mechanismsunderlying EZH2-mediated site-specific silencing are largely unknown. Does EZH2cooperate with any oncogenic transcription factor to repress tumor-suppressive miRNAsin HCC? Which miRNAs are involved in this process? What kind of roles these miRNAsplay in hepatocarcinogenesis? All these issues are still elusive and need to be clarified.【Aims】To identify the oncogenic transcription factor associated with EZH2in HCC and toscreen for the tumor-suppressive miRNAs repressed by both EZH2and this transcriptionfactor, with the aim of better elucidating the epigenetic mechanisms underlying globaldown-regulation of tumor-suppressive miRNAs and the functional roles of these miRNAsduring hepatocarcinogenesis, and providing powerful new tools for the early diagnosis andtreatment of HCC.【Methods】1. The EZH2-involved protein complex in HCC cells was assessed by tandem massspectrometry (LC-MS/MS) to identify the oncogenic transcription factor associated withEZH2;2. Co-immunoprecipitation assays were performed to validate the physicalinteraction between c-Myc and EZH2and to map the critical domain for this interaction;3.The miRNAs repressed by both c-Myc and EZH2were identified by miRNA profiling of HepG2cells in which c-Myc or EZH2had been knocked down previously, and the resultsobtained by miRNA profiling were confirmed using qRT-PCR assay;4. Followingtreatment of HCC cells with epigenetic drugs (5-Aza-CdR and/or TSA), qRT-PCR assaywas carried out to analyze the effect of epigenetic regulation on miR-101expression;5.The transcription start sites of miR-101were determined by5’-RACE assay, and bisulfitesequencing was used to detect the methylation status of miR-101promoter regions;6.Following treatment of HCC cells with DZNep (a pharmacological inhibitor of EZH2) orsiRNAs againt c-Myc/EZH2, qRT-PCR assay was carried out to analyze the effect ofc-Myc and EZH2(especially the enzymatic activity of EZH2) on miR-101expression;7.ChIP-qPCR was performed to assess the binding status of c-Myc and PRC2componentsin miR-101promoter regions;8. Luciferase reporter assays were carried out to analyze theeffect of c-Myc and EZH2on miR-101promoter activity;9. Soft agar colony formationassay, transwell invasion assay, endothelial cell tube formation assay and in vivo tumorformation assay were performed to assess the functional roles of miR-101in regulating themalignant behaviors of HCC cells;10. An unbiased computational screen was performedto identify the potential targets of miR-101by integrating the results of multiple predictionalgorithms (Targetscan, PicTar, and miRanda), and the predicted targets were functionallyclassified using DAVID Bioinformatics Resources;11. Luciferase reporter assays werecarried out to identify the target genes directly repressed by miR-101, and the resultsobtained by reporter assays were validated at the protein level by western blot analysis;12.Soft agar colony formation assay, transwell migration assay, VEGF ELISA and endothelialcell tube formation assay were performed to assess whether the newly identified targetgenes of miR-101are responsible for the aggressive phenotypes of HCC;13. Followingtreatment of HCC cells with mature miR-101mimics, qRT-PCR was carried out to analyzethe effect of miR-101overexpression on the transcription of endogenous miR-101;14.miRNA in situ hybridization and IHC staining were performed to assess the expressionstatus of miR-101, c-Myc, EZH2, EED, JUNB, CXCR7and STMN1in xenograft tissuesand HCC tissues, and the expression levels were further quantified in fresh samples from3normal liver and7HCCs to analyze the clinical relationship between miR-101and its regulators/targets;15. Tissue microarrays comprising54cases of HCCs were used toassess the effect of c-Myc and EZH2on miR-101expression, and the prognosis of thiscohort of HCC patients was analyzed according to the expression status of c-Myc andEZH2.【Results】1. The LC-MS/MS results revealed that, besides other components of PRC2, EZH2also associates with c-Myc, an important oncogenic transcription factor in HCC cells;2.Co-immunoprecipitation assays demonstrated that c-Myc physically interacts with EZH2,and both the bHLH-LZ domain and MBII domain of c-Myc are required for its interactionwith EZH2;3. The microarray results showed that10miRNAs including miR-101arerepressed by c-Myc and EZH2simultaneously, and the qRT-PCR data validated thatmiR-101is a true target of both proteins;4. The qRT-PCR results revealed that, miR-101expression levels were remarkably up-regulated in HCC cells upon combined treatment of5-Aza-CdR and TSA, while no obvious changes were observed upon treatment of onedrug alone;5. The results of5’-RACE assay showed that the TSS of miR-101-1transcriptis located9.3kb upstream of its mature sequence and the TSS of miR-101-2transcript islocated near the miR-101-2locus, subsequent bisulfate sequencing revealed that there areno significant DNA hypermethylation in miR-101promoter regions in both HCC cell linesand HCC tissues;6. The qRT-PCR results showed that the expression levels of miR-101precursors were significantly increased upon c-Myc or EZH2knockdown, and inhibitionof EZH2enzymatic activity also up-regulated the expression of both mature and precursormiR-101;7. The results of ChIP-qPCR assays showed that miR-101promoter regions arehighly enriched for c-Myc, EZH2, EED and H3K27me3in HCC cells, with binding peaksnear the TSSs;8. The results of luciferase reporter assays revealed that miR-101is directlyrepressed by the synergistic effect of c-Myc and EZH2in HCC cells;9. The results of invitro and in vivo functional study showed that the soft-agar colony-formation ability,invasion ability, tumor formation ability and angiogenesis-promoting ability of HCC cells were greatly impaired upon miR-101overexpression;10. The results of bioinformaticanalysis revealed that miR-101could affect the process of chromatin remodeling,angiogenesis, cell growth and cell migration by targeting EZH2, EED, JUNB, STMN1andCXCR7;11. The results of luciferase reporter assays and western blot analysis validatedthat EZH2, EED, JUNB, STMN1and CXCR7are direct targets of miR-101;12. Theresults of in vitro and in vivo functional study demonstrated that these newly identifiedtarget genes of miR-101are responsible for the aggressive phenotypes of HCC;13. TheqRT-PCR results showed that a single transient transfection of mature miR-101mimicscould efficiently initiate the endogenous expression of pre-miR-101, and this effect wasstable and could be sustained for a long period;14. The results of miRNA in situhybridization and IHC staining showed that miR-101was lost while c-Myc, EZH2, EED,JUNB, STMN1and CXCR7were overexpressed in xenograft tissues and HCC tissues,further analysis revealed that there is a significant (P <0.05) inverse correlation betweenexpression levels of miR-101and its regulators/targets;15. The results of tissuemicroarray analysis showed that the miR-101levels in c-Myc+/EZH2+HCCs were muchlower than others, and the prognosis of those patients bearing c-Myc+/EZH2+tumors wasthe poorest.【Conclusions】1. In HCC cells, c-Myc first binds to the promoter region of miR-101and furtherrecruits EZH2-involved PCR2complex, which contributes to the enrichment ofH3K27me3in miR-101promoter region and epigenetic silencing of miR-101eventually;2. Overexpression of miR-101could reverse multiple malignant behaviors of HCC cells,and the underlying mechanism is that miR-101could regulate target genes STMN1, JUNBand CXCR7, thereby modulating the proliferation, angiogenesis and migration processesof HCC cells;3.miR-101could also inhibit the expression of two components of PRC2complex (EZH2and EED), thus creating a double-negative feedback loop betweenmiR-101and PRC2, and the disturbance of this feedback loop contributes to the completesilencing of miR-101and abnormally activation of downstream pathways.