Mechanism of export of colicin E1 and colicin E3.

The mechanism of export of colicins E1 and E3 was examined. Neither colicin E1, colicin E3, Nor colicin E3 immunity protein appears to be synthesized as a precursor protein with an amino-terminal extension. Instead, the colicins, as well as the colicin E3 immunity protein, appear to leave the cells where they are made, long after their synthesis, by a nonspecific mechanism which results in increased permeability of the producing cells. Induction of ColE3-containing cells with mitomycin C leads to actual lysis of those cells, as some time after synthesis of the colicin E3 and its immunity protein has been completed. Induction of ColE1-containing cells results in increased permeability of the cells, but not in actual lysis, and most of the colicin E1 produced never leaves the producing cells. Intracellular proteins such as elongation factor G can be found outside of colicinogenic cells after mitomycin C induction, along with the colicin. Until substantial increases in permeability occur, most of the colicin remains cell associated, in the soluble cytosol, rather than in a membrane-associated form.     

Conformational changes of urea-denatured colicin E1 induced by phospholipid membranes.

The membrane insertion of urea-denatured colicin E1 was studied by using fluorescence spectroscopy, circular dichroism and monolayer techniques. The results showed that the denatured colicin E1 taking mainly the 'random coil' conformation may recover its orderliness to a certain extent under the induction of the phospholipid membrane and insert spontaneously into phospholipid membrane, indicating that unfolding of colicin E1 does not inhibit its membrane insertion. Among the four tryptophan residues of the membrane-bound colicin E1 molecules, at least two were accessible by the quenchers, i.e. not inserted into the membranes. Although urea-denatured colicin E1 interacted preferentially with negatively charged phospholipids, it seems less dependent on the negatively charged lipid than colicin A. The addition of urea increased the speed of the adsorption of colicin E1 to the membrane, but did not affect obviously its membrane insertion ability.     

Direct participation of lexA protein in repression of colicin E1 synthesis.

Spontaneous colicin E1 production by plasmid RSF2124 in a recA lexA(spr) strain of Escherichia coli was about 10-fold greater than that observed in a wild-type strain. The synthesis was repressed nearly to the level of a recA strain in the presence of the plasmid pMCR551, which carries the lexA gene.     

Effect of base substitutions in the colicin E1 gene on colicin E1 export and bacteriocin activity

Summary Base substitutions have been introduced into the segment of the colicin E1 gene corresponding to the polypeptide region between the 404th and the 502nd residues which was considered to participate in colicin E1 export and bacteriocin activity. The methods used were in vitro localized mutagenesis with sodium bisulphite and in vivo mutagenesis using either nitrosoguanidine or ethyl methane sulphonate. Cells carrying mutagenized plasmids were screened by their inability to form a clear zone on a lawn of colicin E1 sensitive cells. Mutation sites were determined from the nucleotide sequence analysis and the altered amino acid residues were reduced. The mutant proteins were analysed for their ability to be exported to the periplasmic space and for their bacteriocin activity. Out of eight mutants obtained, three had a single amino acid replacement. Mutant proteins that had Ser and Glu in place of Pro-462 and Gly-502, respectively, showed a decrease in both the export and the bacteriocin activity. A mutant protein having Arg in place of Gly-439 showed a decrease only in the bacteriocin activity. These results suggest that the target region of colicin E1 contributes to the export as well as the bacteriocin activity but the two functions are supported in part by different amino acid residues of the protein.     

Gating processes of channels induced by colicin A,its C-terminal fragment and colicin E1 in planar lipid bilayers

The dependence on pH and membrane potential of the pore formed by colicin A and its C-terminal 20 kDa fragment has been measured using planar lipid bilayers. The single channel conductance of the pore formed by both colicin A and the fragment increases with pH with an apparent pK of 6.0. At pH 5.0 the gating by membrane potential of the channels formed by either colicin A or its fragment is identical. At the same pH, quite similar pore properties were found when using the related bacteriocin, colicin E₁. In agreement with previous studies, these data indicate that the protein structure containing the lumen of the pore resides in the 20 kDa C-terminal part of the colicin A and favours the recently proposed model, based on protein sequence analysis, which proposes that colicin A, E₁ and I_B C-terminal domains are folded in the same three-dimensional structure. However, it is also shown that colicin A and not its C-terminal fragment undergoes a pH dependent transition between an acidic and a basic form of the pore with an apparent pK of 5.3. The two forms of the pore differ by their gating charge but not by the channel size. These results suggest that there is a pH dependent association between the C-terminal domain carrying the lumen of the pore and another domain of the molecule which affect the pore sensitivity to membrane potential.     

The molecular basis for the pH-activation mechanism in the channel-forming bacterial colicin E1

The in vitro activity of the channel-forming bacteriocins such as colicin E1 in model membranes requires the specific activation of the protein by an acidic environment in the presence of a membrane potential. Acid activation of the C-terminal domain results in the formation of an insertion-competent intermediate with an enhanced ability to penetrate and perforate cell membranes. We report novel findings of this activation process through the design and study of mutant proteins involving the replacement of conserved Asp residues Asp-408, Asp-410, and Asp-423 within helices 5a and 4 in the colicin E1 channel domain that resulted in enhanced membrane binding, bilayer insertion rates, and ion channel activities at near neutral pH values. This activation process involves the destabilization of a critical salt bridge (Asp-410 and Lys-406) and H-bonds (Asp-408 and Ser-405 main chain; Asp-423 and Lys-420 main chain). The helix-to-coil transition of this motif was identified previously by time-resolved Trp fluorescence measurements (Merrill, A. R., Steer, B. A., Prentice, G. A., Weller, M. J., and Szabo, A. G. (1997) Biochemistry 36, 6874-6884), and here we use this approach to demonstrate that disruption of the helical structure of helices 4 and 5a results in a shift in this equilibrium to favor the coil state. Finally, we show that the essential components of the pH trigger motif are conserved among the channel-forming colicins and that it likely exists within other bacterial proteins and may even have evolved into more sophisticated devices in a number of microbial species.     

Mapping of colicin E2 and colicin E3 plasmid deoxyribonucleic acid EcoR-1-sensitive sites.

Colicin plasmids E2 and E3 (Col E2 and Col E3) deoxyribonucleic acid (DNA) has been shown to contain, respectively, two and three EcoR1 restriction endonuclease-sensitive sites. This was determined by measuring the DNA fragments generated after EcoR1 endonuclease treatment by agarose gel electrophoresis and electron microscopy. The structure of heteroduplex Col E2-col E3 DNA molecules formed from EcoR1-generated fragments permitted a localization of the EcoR1-sensitive sites on the plasmid chromosomes.     

Temporal control of colicin E1 induction.

The expression of the gene encoding colicin E1, cea, was studied in Escherichia coli by using cea-lacZ gene fusions. Expression of the fusions showed the same characteristics as those of the wild-type cea gene: induction by treatments that damage DNA and regulation by the SOS response, sensitivity to catabolite repression, and a low basal level of expression, despite the presence of the fusion in a multicopy plasmid. Induction of expression by DNA-damaging treatments was found to differ from other genes involved in the SOS response (exemplified by recA), in that higher levels of DNA damage were required and expression occurred only after a pronounced delay. The delay in expression following an inducing treatment was more pronounced under conditions of catabolite repression, indicating that the cyclic AMP-cyclic AMP receptor protein complex may play a role in induction. These observations also suggest a biological rationale for the control of cea expression by the SOS response and the cyclic AMP-cyclic AMP receptor protein catabolite repression system.     

Isolation and characterization of a mutant ColE1 plasmid that allows constitutive colicin E1 synthesis.

It has been possible to isolate a ColE1 mutant which synthesizes colicin E1 constitutively. This result shows that there must be a gene(s) responsible for the regulation of colicin E1 synthesis.     

Ribosome degradation in Escherichia coli induced by colicin E2

The mode of action of colicin E₂ on ribosomes in $\text{Escherichia coli}$ cells was investigated by zonal centrifugation analysis. Ribosome particles, both 50S and 30S, were degraded to smaller contents with the lapse of time by the action of colicin E₂. Gradual reduction of S values of each particles could not be observed and degradative intermediates of possible RNA-protein complex were detected only at the position between 30S and 4S in the zonal centrifugation profile, which indicated the destruction of ribosome in burst-out attitude. 50S ribosome fraction influenced by colicin E₂ contained both 23S and half-sized RNA. From these data, the mode of action of colicin E₂ on ribosomes in $\text{E. coli}$ was discussed.     

Fine cleavage map of a small colicin E1 plasmid carrying genes responsible for replication and colicin E1 immunity.

A small plasmid (pAO2, 1 megadalton) carrying genes responsible for replication and colicin E1 immunity has been constructed from colicin E1 plasmid (A. Oka, K. Sugimoto, and M. Takanami, Proc. Mol. Biol. Jpn., p. 113-115, 1976). pAO2 DNA was cleaved into unique fragments with seven restriction endonucleases (R.HaeII,R.HaeIII,R.HapII,R.HhaI,R.AluI,R.HgaI, and R.HinfI). R.HaeII cleaved pAO2 DNA at two sites, R.HaeIII at four sites, R.HapII at nine sites, R.HhaI at eight sites, R-AluI at nine sites, R.HgaI at two sites, and R.HinfI at four sites, respectively. The order of HaeIII fragments of pAO2 was deduced from the physical map of colicin E1 plasmid previously reported (A. Oka and M. Takanami, Nature (London) 264:193-196, 1976). HapII, HhaI, and AluI fragments of pAO2 were assigned by analyzing overlapping sets of fragments arising upon digestion of individual HaeIII fragments with one of R.HapII, R.HhaI, or R.AluI, and upon their reciprocal digestion. The cleavage sites for R.HaeII, R.HgaI, and R.HinfI were localized on HapII, HhaI, and AluI fragments by combined digestion. On the basis of these data and estimates of the size of each fragment, a fine cleavage map of pAO2 was constructed.