The laboratory studies biological mechanisms involved in human cancer. Our goal is to determine the molecular mechanisms that are involved in formation of genetic changes (gene mutations) and epigenetic changes (DNA methylation and histone modifications) in the human genome.

The main research topic in the laboratory is epigenetic gene regulation in development and disease. Many genes are silenced by DNA methylation and by other repressive epigenetic mechanisms in human tumors. Methylated genes hold great promise not only as functionally relevant silenced genes involved in tumorigenesis but also as biomarkers for cancer diagnosis. We developed a sensitive method for analysis of DNA methylation on a genome-wide scale, the methylated-CpG island recovery assay (MIRA). MIRA is used to identify new commonly methylated genes in human tumors. Homeobox genes, known to be involved in developmental processes and tissue specification, are very frequently methylated in lung cancers and many other human tumors. We found that the mechanisms of cancer-specific DNA hypermethylation often involves targeting of specific DNA sequences, in particular homeobox genes, by the Polycomb repression complex and this pathway is already prevalent in inflamed tissue predisposing epithelial cells to malignant transformation. Using bioinformatics approaches, we are trying to uncover mechanisms dependent on DNA sequence features and DNA binding proteins that determine the specificity of DNA methylation changes in human cancers. We are also investigating changes in the epigenome in response to exposure of cells to environmental carcinogens including ultraviolet and ionizing radiation. Another project will attempt to establish a link between tissue aging and malignant transformation based on common changes in epigenetic regulation that underlie these processes. Very recently, previously unrecognized DNA bases, including 5-hydroxymethylcytosine, 5-formyl-, and 5-carboxy-cytosine, have been detected in certain mammalian tissues and cell types. We developed methodology to determine the sequence location of 5-hydroxymethylcytosine in mammalian DNA and found that it is targeted to promoters and gene bodies. In collaboration with Piroska Szabó’s lab, we showed that 5-methylcytosine is oxidized to 5-hydroxymethylcytosine selectively in the paternal genome of fertilized oocytes as part of a transgenerational epigenetic reprogramming mechanism. We showed that 5-hydroxymethylcytosine is strongly depleted in human cancers and could therefore be used as a biomarker for malignancy. Future work will focus on the distribution patterns and on the enzymology of 5-hydroxymethylcytosine formation and removal and its biological function in normal and malignant tissues.