Mechanism of colicin E3 production in strains harboring wild-type or mutant plasmids.

As previously reported by others, more than 90% of the colicin E3 synthesized soon after induction of colicinogenic bacteria was found to be cell bound, about half of it being in a salt-extractable state at the cell surface. Evidence is presented that the colicin molecules remain preferentially bound to the cell which produced them, rather than being secreted and randomly distributed in the cell population. Secretion of colicin E3 may in fact never occur, all or most of the colicin found free in the medium perhaps being released during lysis of the producing cells long after induction. Among 19 mutant plasmids selected on the basis of their inability to produce an active colicin, only 3 were found to code for a protein which although it lacked any bactericidal activity, had the same molecular weight as wild-type colicin E3 and displayed a reaction of immunological identity with it. These three inactive colicins fail to be exported to the cell surface and seem to be blocked at some intermediate stage in the export process.