¹ANDREY V. MATVEYEV, ¹JULIE RUMBLE, and ^2*JEFF ELHAI

¹Dept. of Biology, University of Richmond, Richmond, VA 23173, USA; ²Dept. of Biology, Virginia Commonwealth University, Richmond, VA 23284, USA

     Many members of the gamma-proteobacteria (e.g. Escherichia coli) and those of the alpha-proteobacteria (e.g. Caulobacter crescentus) possess specific DNA methyltrans-ferases (MTases) that have no corresponding restriction enzymes in the host strains (1). These solitary MTases have amongst their functions the regulation of DNA synthesis (2,3). Cyanobacteria also possess a solitary MTase, which, like that of the gamma-proteobacteria, recognizes the sequence GATC. Anabaena PCC 7120 is unusual in that it has two functional GATC-specific MTases, DmtA and M.AvaV (1), neither with a corresponding restriction enzyme. A protein structurally similar to DmtA is in Synechocystis PCC 6803 and probably in most cyanobacteria. In contrast, M.AvaV is structurally unusual, similar only to a few noncyanobacterial MTases.

     Why should Anabaena possess two MTases with identical recognition sequences? Loss of M.AvaV resulted in a higher than normal frequency of heterocysts, including occasional double and triple heterocysts, which are never seen in the wild-type strain under standard growth conditions. In contrast, additional copies of the 5' region of avaMV (encoding M.AvaV) elicited a drastic decrease in heterocyst frequency to about 3% in N-free medium. The lower frequency could be overcome by introduction of a hetR-bearing plasmid.

     Efforts to produce insertionally inactivated dmtA in wild-type Anabaena by double recombination have all failed, starting with a variety of single-recombinants. The gene is evidently essential under standard growth conditions. However, it was possible to gain a double-recombinant, inactivated dmtA gene in an avaMV- background. Such mutants grew poorly and invariably contained unsegregated copies of the wild-type chromosome. The level of GATC-methylation varied with the growth condition. Reintroduction of wild-type avaMV into the strain has proven difficult and may be impossible.

     The functions of the two MTases, DmtA and M.AvaV, are thus not identical and in fact may antagonize each other. We have not yet found a way to distinguish their activities in vitro.

References
1. Matveyev, A.V., Young, K.T., Meng, A, and Elhai, J. (2001). DNA methyltrans-ferases of the cyanobacterium Anabaena PCC 7120. Nucl Acids Res 29:1491-1506.
2. Marinus, M.G. (1996). Methylation of DNA. In: Escherichia coli and Salmonella: Cellular and molecular biology, 2nd ed. (Eds: Neidhardt, F.C., Curtiss III,R., and Ingraham, J.L.) American Society for Microbiology, Washington, D.C, 782-791.
3. Zweiger, G., Marcynski, G., and Shapiro, L. (1994). A Caulobacter DNA methyltransferase that functions only in the predivisional cell. J. Mol. Biol. 235:472-485.

*Author for correspondence, e-mail: Cyano@Richmond.Edu