TARGETS AND REQUIREMENTS OF METHYLATION IN HUMAN CELLS

点击翻译按钮获取中文摘要

De novo methylation has been reported to occur in mammalian cells during embryogenesis, carcinogenesis, and foreign DNA integration. Although the methylation patterns of many de novo methylated genes have been studied extensively, the cis and trans determinants of de novo methylation are unknown and the rules that govern de novo methylation are entirely unclear. We propose to use two experimental systems, an oriP- based stable episomal system and a flp-recombination mediated chromosomally-integrated target system, to answer some of the essential questions regarding de novo methylation. In preliminary studies, we have found that the EBNA1 coding sequence exhibits reproducible de novo methylation both on the episome and in the chromosome when the de novo methyltransferase 3a or 3b is present. This is a useful starting point to define the requirements and targets of these two methyltransferases. In addition to the EBNA1 coding sequence, we have also examined the CpG island upstream of the endothelin receptor B gene. The chromosomal copies of this CpG island are methylated in a prostate cancer cell line. However, it is not methylated on the stable episome in this prostate cancer cell line or in cells expressing either of the de novo methyltransferases. A chromosomally-integrated target system utilizing flp-recombination will allow us to dissect the requirement for CpG island methylation in cancer cells at its natural location. Experiments described in this proposal are designed to answer the following questions: 1) How do the known de novo methyltransferases, Dnmt3a and Dnmt3b, target DNA for methylation? 2) What are the requirements for de novo methylation by these two methyltransferases? 3) Are there boundary elements or factors that can protect regions from de novo methylation? 4) What are the targets of Dnmt3a and Dnmt3b? 5) How is a CpG island targeted for methylation in cancer cells?