Andrey Matveyev,* Kathryn T. Young, Julie Rumble, Jennifer Lee, Elaine Bucheimer, Andrew Meng, Jeff Elhai
Dept. of Biology, University of Richmond, Richmond VA 23173, U.S.A.

The spaced heterocysts that appear in Anabaena PCC 7120 in response to nitrogen starvation have been explained by a model resting on two assumptions: (1) the first cells to begin differentiation arise at random along the filament, and (2) an inhibitor of differentiation diffuses laterally from differentiating cells. We have considered an alternative model, one that postulates nonrandom initiation of differentiation governed by a cell's position in the cell cycle. The model has been tested in two ways. First, the pattern of cell division within a filament has been studied, to see if it could underlie the ultimate pattern of differentiating cells. Cell division was found not to occur at random. Rather, contiguous cells are likely to share a common position in the cell cycle. Second, we have attempted to alter the normal regulation of DNA synthesis and cell division by manipulating genes encoding DNA methyltransferases, enzymes that have been shown to control both DNA replication and differentiation in other bacteria. Four such genes have been cloned. Two (dmnA and dmnD) encode enzymes that methylate the same sequence (GATC). Loss of dmnA appears to be lethal, while its presence in multiple copies leads to a striking drop in the frequency of heterocysts. Another gene (dmnB) encodes a GGCC-specific DNA methyltransferase. Different manipulations of the dmnB region has led to either the loss of the ability to differentiate or to heterocyst formation in the presence of nitrate. These results will be discussed in terms of opposing models to explain the spaced differentiation of heterocysts.