tif-dependent induction of colicin E1, prophage lambda, and filamentation in Escherichia coli K-12.

To help understand how the tif-1 mutation of the recA gene of Escherichia coli confers adenine activability on the recA protein, we used the fact that cytidine plus guanosine inhibits induction of prophage lambda and cell filamentation in a tif-1 mutant, and that adenine reverses this inhibition. We varied the amount of adenine in agar plates containing a fixed amount of cytidine and scored for survivors of three different tif-dependent lethal induction processes. Much more adenine was required for cell killing when cytidine was present than when it was absent. Therefore adenine does not override cytidine inhibition, but instead appears to compete with it for a site of action which may be on the recA protein. The competition is not at the cell transport level. Our results lead to a model in which the tif form of the recA protein is an allosteric enzyme that binds both negative and positive modulators. By varying the adenine-cytidine ratio of the medium it is possible to control the degree of induction in a tif-1 cell. For the three different tif-dependent inductions studied here, least adenine was required for lambda induction and most for lethal filamentation, presumably reflecting requirements for different amounts of activated recA protein in each process. Varying the adenine-cytidine ratio revealed two stable intermediate stages in lambda induction, as well as a stage of colicin E1 induction in which the cells produced colicin without cell death. The rate of filament formation could be similarly controlled. Experiments with tif (ColE1, lambda) gave evidence of a competition between colicin repressor and lambda repressor for activated recA protein.