Epigenetics of tandem DNA repeats in FSH muscular dystrophy, the DNMT3B-deficient ICF syndrome, and ovarian epithelial tumors
Description
Long tandem repeated DNA sequences in the human genome have often been thought to not play very specific roles in cell physiology. Now the view is changing with new discoveries that transcription is permissible in centromeres [1]. In addition, the binding of transcription factors to long tandem repeated DNA sequences has been reported and they may act as biologically important storage sites for these factors in the nucleus [2, 3] Facioscapulohumeral muscular dystrophy (FSHD) is closely linked to the shortening of a tandem array that consists of 3.3-kb D4Z4 repeat units in one of the two 4q35.2 alleles [4, 5]. Patients with immunodeficiency, centromeric region instability, and facial anomalies (ICF) syndrome have naturally occurring DNA hypomethylation and abnormalities localized largely to tandemly repeated DNA (satellite 2) in the juxtacentromeric regions of chromosome 1 and chromosome 16 [6, 7]. In human cancers, the strong tendency for the satellite 2 DNA in the juxtacentromeric regions to be hypomethylated was first reported in breast adenocarcinomas by our lab [8] and further confirmed by our and other labs in a wide variety of human cancers including ovarian epithelial carcinomas [9-13] In this thesis, I describe my results from epigenetic studies of FSHD cells, ovarian tumors, and ICF cells. By using two marker sequences in a 4q35.2 subregion that were 0.3 Mb and 0.8 Mb proximal to D4Z4, I demonstrated that both subregions behaved like unexpressed gene regions with regard to the DNA replication timing in S phase in FSHD and control samples. This suggests that this whole region is not like constitutive heterochromatin. In the D4Z4 arrays and their surrounding regions in 4q35.2 region, I provided evidence of in vitro G-quadruplex formation from single-stranded oligonucleotides, suggesting a special chromatin structure in these arrays. In my studies of ovarian tumors, I showed overall hypomethylation and hypermethylation at one CpG site in a 0.2-kb subregion of Sat2 in ovarian carcinomas compared with somatic control tissues For my analysis of ICF cells, which has constitutive hypomethylation of Sat2, I examined methylation in the promoter regions or the 5' end of several genes, which had been shown by other lab members to be upregulated in RNA levels in ICF vs. control LCLs. I found no ICF-dependent methylation changes in these regions. These genes might be downstream of some other genes with ICF-specific promoter hypomethylation or the hypomethylation of Sat2 DNA might indirectly affect the expression of these genes. These studies helped elucidate the biology of DNA repeats in human diseases, although there were many questions still need to be answered