Research Interests My research projects on naturally programmed DNA methylation have spanned studies of microbial methylation of cytosine at the 5 and N4 positions to analysis of human DNA methylation leading to the first discovery of a protein that binds preferentially to methylated DNA and the first report of aberrant methylation in various human cancers. This interest led to my founding the international DNA Methylation Society in 1994, an e-mail-based society.

---

We showed that inhibitors of DNA methylation, but not other genotoxins, can target at extraordinarily high frequencies the formation of rearrangements of chromosome 1, and to a lesser extent, chromosome 16, in cultured human cells of normal origin. The spectrum of resulting chromosomes, visualized at metaphase, includes whole-arm deletions (Fig. 1, C), isochromosomes with two long arms and no short arms joined at the centromere (Fig. 1, D), and strange multibranched chromosomes with 3 - 7 arms joined in the centromeric region (Fig. 1, E-J and L). We are pursuing this line of research using cultured cells from patients with a rare human disease which is diagnosed by exactly the same spectrum of chromosomal abnormalities.

In this research, we hope to gain insight into the functions of DNA methylation in normal development and in carcinogenesis. This disease is called ICF (immunodeficiency, centromeric region instability, and facial anomalies) and is often fatal in childhood. It is the only known disease with aberrations in DNA methylation transmitted by Mendelian inheritance. Patients with this disease have abnormal hypomethylation of the chromosomal regions that are prone to rearrangements, namely, the heterochromatin around the centromeres of chromosomes 1 and 16. We have shown that these chromosomal regions are also undermethylated in breast cancer, Wilms tumors, and ovarian cancers, three types of cancer in which rearrangements in the vicinity of the centromeres of chromosomes 1 and 16 have been implicated in carcinogenesis. Our research in this area includes characterization of which genes are hypomethylated in cells from ICF patients and why ICF cells are prone to these chromosomal characteristic rearrangements. We are also examining two types of cancer, ovarian epithelial carcinomas and Wilms tumors, to analyze relationships between cancer-associated DNA hypomethylation and chromosome rearrangements (including ICF-like rearrangements) or cancer progression. Our studies of hypomethylation of DNA in cancer extend to the exciting field of gene regulation and intranuclear localization. Our recent findings from microarray analysis on ICF-specific alterations in gene expression suggest that methylation of constitutive heterochromatin in the vicinity of centromeres is involved in the repression of some genes throughout the genome in trans by altering interchromosomal heterochromatin-euchromatin interactions, and, therefore, changing the intranuclear localization of these genes. A precedent for this model is trans control of gene expression by interactions between early lymphogenesis genes, centromeric heterochromatin, and the Ikaros transcription factor. We are testing the hypothesis that one of the ways in which DNA hypomethylation in cancer favors carcinogenesis is by relieving trans-repression of gene expression by centromeric region heterochromatin.

In addition we are studying lymphoblastoid and myoblast cell cultures from a dominant genetic disease, facioscapulohumeral muscular dystropy (FSHD) to understand the origin of this unique syndrome. FSHD patients have 1-8 copies of a 3.3-kb subtelomeric repeat on one homologue of chromosome 4. In contrast, normal individuals have 10 to about 150 copies of this repeat on both of their chromosome 4 homologues. FSHD is thought to be due to heterochromatinization of genes adjacent to the repeat region only when the copy number of the repeat is high enough. It is hypothesized that inappropriate expression of such genes, especially in muscle cells, in patients with the deletion results in the disease phenotype. High levels of DNA methylation are often found in heterochromatin and are causally related to compaction of chromatin structure. We recently demonstrated for the first time that this FSHD-linked repeat is highly methylated in normal tissues and myoblasts. Also, we have provided evidence from histone chromatin immunoprecipitation (ChIP) assays, immunofluorescence in situ hybridization (immuno-FISH), and gene expression analysis by quantitative RT-PCR, that the chromatin adjacent to this repeat is not heterochromatic in normal or FSHD cells. Therefore, it is not prone to disease-triggering loss of this heterochromatinization depending on the number of copies of the FSHD-linked DNA repeat, as had previously been proposed. We are investigating the proteins that bind to the DNA repeat and their involvement in FSHD. Also, the relationship between the number of copies of this repeat and other types of chromatin structure changes, such as long- distance chromatin looping and changes in the chromatin structure of the repeat region itself, is under investigation in our lab.

[Home] [General Info] [Research Interests] [Publications] [Tulane Medical School] [Hayward Genetics Center] [E-Mail]