Full Professor,
  Ph.D., Hebrew Univ. (Jerusalem)

  Room: MCMED 1309/1310
  Phone: 398-7107
  Fax: 398-6690

  E-Mail: mszyf@pharma.mcgill.ca

General Interests of the laboratory:

1. Epigenome

2. DNA methylation

3. Cancer is an epigenetic disease

4. Therapeutic implications of the epigenome on anticaner therapy


What is the epigenome?

While genomic information is uniform in the different cells of complex organisms, the epigenome controls the differential expression of genes in specific cells. The programming of gene expression profiles is therefore dependent on the epigenome. The epigenome is composed of two modules, a component that is part of the covalent structure of DNA, methylated cytosines located in the dinucleotide sequence CG and a noncovalent module. Our understanding of the noncovalent module of the epigenome the chromatin and its associated chromatin modifying and remodeling activities is rapidly expanding in recent years (Strahl and Allis, 2000). It is now becoming clear that modifications of histones and their tails by acetylation, phosphorylation, and methylation plays an important role in determining the positioning of nucleosomes on DNA and the compactness of chromatin. Chromatin structure determines the state of activity of genes by gating the access of the transcription machinery to transcriptional regulatory regions. Chromatin structure plays a role in other genomic activities such as recombination and repai. Changes in chromatin structure play an important role in the silencing of certain genes in cancer and histone deacetylase inhibitors have demonstrated anticancer effect.
Back to top

What is DNA methylation?

In addition to this flexible and dynamic module of the epigenome, which is associated with the genome but is not part of its covalent structure, the genome is covalently modified by addition of a methyl group at the 5th position of the cytosine ring. Since the methyl group is connected to DNA by a strong chemical bond, it is considered a stable and fixed mark. The vast majority of methylated cytosines in vertebrate genomes are found in the dinucleotide sequence CG. Not all CG dinucleotide sequences are methylated, different CGs are methylated in different tissues creating a pattern of methylation that is gene and tissue specific. Thus, there is a good correlation between the state of activity of genes and lack of methylation of CGs in their regulatory regions (Razin and Riggs, 1980). Remarkably, there is also a tight correlation between the chromatin structure and the status of DNA methylation (Razin and Cedar, 1977). We propose that the DNA methylation pattern is steady state equilibrium of DNA methylation and demethylation reactions. DNA methylation is catalyzed by DNA methyltransferase enzyme and demethylation catalyzed by demethylase enzymes. The DNA methylation reaction is stimulated by inactive chromatin and demethylation by active chromatin. Thus the direction of the DNA methylation reaction is determined by chromatin structure.

Back to top

Cancer is an epigenetic disease

It is now clear that cancer is an epigenetic disease. In most tumors the DNA methylation pattern is defective. Vast regions of the genome lose their methylation while specific regions are heavily methylated.
Understanding the mechanisms that link chromatin structure and DNA methylation are extremely important for unraveling the possible mechanisms responsible for the DNA methylation pattern in normal tissues, possible physiological alterations of DNA methylation patterns throughout life and during the aging process, and its aberration in cancer. Before we are able to properly target DNA methylation in cancer therapy we ought to understand which of the changes in the DNA methylation machinery and DNA methylation pattern are causal for cancer and which are a consequence of the transformation process.

My laboratory is interested in understanding the links between chromatin, DNA methylation and cancer therapy. We want to understand what mechanisms define the DNA methylation pattern and why is the DNA methylation pattern tightly correlated with chromatin structure. Our working hypothesis is that both the DNA methylation and demethylation machineries are directed by chromatin structure and that both are defective in cancer. We want to understand how chromatin directs DNA methyltransferase and demethylase and how why they are involved in cancer and metastasis.
Back to top


Therapeutic implications of the epigenome on anticancer therapy

Inhibition of the main DNA methylation enzyme DNMT1 and the demethylase/ MBD2 results in a marked anticancer effect and has potential in anticancer therapy. We developed antisense and direct inhibitors of DNMT1 with demonstrated preclinical anticancer effect. A DNMT1 antisense molecule is currently in clinical trials (directed by MethylGene Inc.). Inhibition of MBD2/demethylase also showed anticancer activity in preclinical studies. We are interested in identifying under members of the epigenome that might serve as anticancer targets.

Discussion of our model on the anticancer potential of DNA methylation and demethylation inhibitors will be found in: {Szyf M., 1994, Trends Pharmacol Sci, 15(7), 233-8, Szyf M., 1996, Pharmacol Ther, 70(1), 1-37; Szyf M., 1998, Cancer Metastasis Rev, 17(2), 219-31, Szyf M. et al., 2000, Ann N Y Acad Sci, 910, 156-74; discussion 175-7, Szyf Moshe, 2000, Current Drug Targets, 1(1), 101-118, Szyf M., 2001, Trends Pharmacol Sci, 22(7), 350-354., Szyf M and Detich N, 2001, Prog Nucleic Acid Res Mol Biol, 69, 47-79}
Back to top