The long term goal of this project is to understand how m5-cytosine methylases (m5C-methylases) achieve their exquisite specificity in recognizing DNA sequences. To achieve this goal we are studying a set of enzymes that are related both structurally and functionally. Comparison of the sequences of a large number of methylases that form 5methylcytosine in DNA shows there is a great,deal of similarity. We propose to study three of these m5C-methylases in great detail These are the MspI methylase (forms meCCGG), the HpaII methylase (forms CmeCGG) and the HhaI methylase (forms GmeCGC). We have prepared an overexpressing clone of the MspI methylase and have purified this methylase to homogeneity. We are now attempting the crystallization of this protein either alone, complexed with DNA or complexed with known inhibitors of the enzyme such as S-adenosylhomocysteine or sinefungin. These structural studies will be complemented by biochemical and genetic approaches. We have also completed the sequence of the HpaII methylase gene and find that it shows striking similarities both to the Mspl methylase gene and the HhaI methylase gene. A key goal for the immediate future is to define in detail the DNA recognition domains for at least one of these methylase genes. This will be accomplished through a series of domain swap experiments between pairs of methylases. These domain swaps will involve both exact exchanges at selected sites, using a PCR-based strategy, as well as more random exchanges followed by genetic screening to detect recombinant methylases. We will also try to swap in other DNA-recognizing domains, such as homeo-box domains, so as to change methylase specificity. Truncated versions of the methylases will be prepared and tested to see if functional domains can be isolated that are able to bind DNA. We will use genetic techniques to analyse these methylases and their hybrids and derivatives, and will search for specific mutants that would extend the usefulness of the methylases for genome tagging and mapping studies.
Showing the most recent 10 out of 30 publications
