Integrative study of aberrant DNA methylation and chromatin regulators as important epigenetic variations in cancer development

Abstract

Cancer is the term referring to over 100 types of related diseases and the second leading death cause globally after cardiovascular diseases. Both genetics and epigenetics are involved in cancer development, and the importance of cancer epigenetics is increasingly being appreciated. Compared to genetic mutations, epigenetic alterations are naturally reversible, making it very suitable for therapeutic manipulation. The study in this thesis focused on aberrant DNA methylation and chromatin regulators (CRs), both of which are important epigenetic alterations in cancer development.

Firstly, we developed a series of computation and experimental strategies to distinguish driver DNA methylation variations involved in cancer development from the massive passengers. Direct Regulatory Elements (DRE) along with their gene targets were predicted, and many were validated by the CRISPR-dCas9 epigenetic editing system in this study. We found that changing DNA methylation status of driver DRE could significantly affect target gene expression and cancer cell behaviors, very similar to the effects of promoter DNA methylation. This strategy recovered lots of annotated/unannotated regulatory regions with many known/novel oncogenes or tumorsuppressors.

Secondly, we designed computational and experimental strategies to identify novel CR drivers and its downstream TF effectors in cancer development. We conducted an integrated analysis of the genomic and transcriptomic data of TCGA to identify key CR drivers affected by aberrant genomic alterations. Some novel CR drivers were demonstrated to play master regulator roles in cancer initiation and progression, and confirmed with experiments including dCas9 epigenetic editing and DOX inducible shRNA. It is possible that CRs mediate tumorigenesis by modulating the activity of some downstream key TFs. To understand how CR modulates the activity of downstream TFs, we designed an algorithm based on differential connectivity to identify CR-TF pairs with significant function interaction, but without expression correlation. The top-ranked master regulatory CRs plus one modulation pair of CR-transcription factors (TF) interaction predicted by the analyses were experimentally verified in liver cancer cell lines. We verified our prediction of the interaction between SIRT7 and NFE2L2 in liver cancer. SIRT7, a member of the sirtuin family and an important epigenetic regulator, was identified as a driver of tumorigenesis in our study. Our experiments validated that SIRT7 positively modulated the transcriptional program of NFE2L2, and they physically interacted with each other at protein level. Thus we uncovered the role of SIRT7 in tumorigenesis by regulating antioxidant signaling pathway in liver cancer.

Thirdly, we elucidated how cancer development was determined by the interplay between CR and DNA-methylation with one example. Here, gene Vacuolar Protein Sorting 72 Homolog (VPS72), a CR, was studied and proved as a master regulator (MR) in liver hepatocellular carcinoma (LIHC). VPS72 was found to control the expression of its target genes (e.g. HDAC11 and SMYD5) by decreasing the methylation level in their regulatory sites through depositing with H2A.Z, a variant of histone H2A, into the nucleosome.

Taken together, these findings reveal that aberrant DNA methylation and CRs are important epigenetic alterations in tumorigenesis, understanding of which can provide mechanical insight as well as design of new epigenetic therapies.