The level of DNA superhelicity in E. coli is set by the balance of the DNA-supercoiling activity of DNA gyrase and the DNA-relaxing activity of topoisomerase I. We have shown that the synthesis of both enzymes is in turn controlled by the level of supercoiling; if, for example, the cell's DNA is not supercoiled enough, more DNA gyrase and less topoisomerase I are made, so that supercoiling then increases toward a more normal level. We have recently studied the expression of the genes for the two gyrase subunits (gyrA and gyrB) so as to learn how this control works. A deletion analysis of both promoters shows that the region needed for DNA relaxation-induced expression is very small, no more than 21 base pairs. Both of these promoters are also atypical in that down-stream sequences are more important, and upstream sequences less important, than is normal for E. coli promoters. We have determined the DNA sequence of the E. coli gyrB gene and used the results to locate within the native protein the cleavage that produces a partly active Gyr B fragment in cell extracts. The fragment is the C-terminal half of the molecule. We have also compared this sequence with that of the gyrB gene of B. subtilis. The E. coli Gyr B protein (MW 90,000) can be shown to contain one extra loop of almost 200 amino acids. A study of the DNA-dependent ATPase activity of DNA gyrase shows that DNA binding sites in both half-molecules of the A2B2 enzyme must be filled before ATP hydrolysis can occur. Binding of short DNA chains and of ATP have both been shown to be highly cooperative. We have made a rough spatial model of the enzyme-DNA complex based on electro-dichroism experiments.
