Vol. 169, No. 6JOURNAL OF BACTERIOLOGY, June 1987, p. 2466-24700021-9193/87/062466-05$02.00/0Copyright © 1987, Americah Society for Microbiology

Four Plasmid Genes Are Required for Colicin V Synthesis,Export, and Immunity

LYNNE GILSON, HARE KHRISNA MAHANTY,t AND ROBERTO KOLTER*Department ofMicrobiology and Molecular Genetics, Harvard Medical School, Boston, Massachusetts 02115

Received 22 December 1986/Accepted 26 February 1987

The colicin V production and immunity genes were isolated from plasmid pColV-K30. A HindIH-to-SalIfragment of 9.4 kilobases was cloned into the compatible vectors pBR322 and pACYC184. Mutants defectivein colicin production were generated by TnS insertions and by constructing deletions in vitro. Physical analysisof these mutations identified a 4.4-kilobase region of this DNA which contains all the plasmid genes (cva) neededfor the production of colicin V. The colicin V immunity determinant (cvI) iS in a 700-base-pair fragment locatedwithin one end of this region. Complementation tests Identified three genes, called cvaA, cvaB, and cvaC,required for colicin production. Analysis of the proteins labeled in minicells harboring various TnS insertionsallowed us to identify protein products for the cvaA and cvaC genes. Mutations in cvaA and cvaB eliminatedcolicin activity in culture supernatants, but not within the cells. Mutations in cvaC, however, eliminated alldetectable activity. From these results we conclude that the cvaC gene codes for the structural gene for colicinV, while cvaA and cvaB are apparently needed for the normal export of the colicin.

Colicin V is a small, proteinaceous toxin whose activityalong with an immunity determinant is encoded on large,low-copy-number plasmids (14). ColV plasmids have beenfound naturally occuring in many strains of Escherichia coliand other members of the family Enterobacteriaceae. Thesebacteria also define the activity range for colicin V. Its targetfor growth inhibition is thought to be the cytoplasmic mem-brane, where it prevents the formation of membrane poten-tial (32). ColV plasmids are often associated with E. coliinvasiveness and pathogenicity (29, 30). These plasmids alsooften carry genes which may enhance the ability of cells toproliferate within the host. Examples of these are theaerobactin iron uptake genes (31) and a gene for increasedserum resistance (5). In addition, an enhanced adherence tointestinal epithelial cells has been noted in strains harboringColV plasmids (11). Colicin V production does not appear tobe a virulence determinant, but it has been hypothesized thatit may help to selectively maintain these genes (26, 31).The colicin V toxin is distinguished from other colicins by

the small size of the active protein (13) and by its consti-tutive, rather than SOS-inducible, synthesis (15). There isalso no evidence that, like many colicins, colicin V accumu-lates in the cell before its release or has a lysis gene productresponsible for its release (25).

Frick et al. (13) cloned a 900-base-pair (bp) region of thepColV-B188 plasmid which included the colicin V immunitygene (cvi) and an apparent colicin V structural gene (cva).However, cells harboring this cloned fragment did not pro-duce growth inhibition zones on a lawn of sensitive cells, andculture supematants did not contain assayable amounts ofcolicin. The killing activity coded by this 900-bp fragmentcould only be assayed after lysing the cells and appeared tobe fourfold-less potent than that coded by a much largerfragment (13). In this paper we show that a 4.4-kilobase (kb)region of DNA containing cvi and at least three other genesis necessary for the normal production and export of killingactivity.

* Corresponding author.t Present address: Department of Plant and Microbial Sciences,

University of Canterbury, Christchurch-1, New Zealand.

MATERIALS AND METHOIDS

Bacterial strains and plasmids. The bacterial strains andplasmids used in this study are listed in Table 1.Media and culture conditions. Media were prepared as

described by Miller (24). Antibiotics were used at the fol-lowing concehtrations (micrograms per milliliter): ampicillin,150; chloramphenicol, 20; and kanamycin, 50.DNA manipulations and colicin V assays. Plasmid DNA

preparation, digestions, ligations, transformations, and elec-trophoresis were performed as described previously (6, 12,21). Colicin V activity and immunity were determined asdescribed previously (13). Activity was quantitated by thecritical dilution method (22).TnS mutagenesis. Transposition of TnS into pHK11 and

pHK22 was accomplished by introducing these plasmids intostrain DB1358 (28). To ensure independent insertions intomany different sites, 20 colonies of DB1358 harboring eachplasmid were grown independently, and plasmid DNA wasprepared. This DNA was then used to transformn MC4100,selecting for resistance to kanamycin and ampicillin for thepHK11 derivatives and to kanamyciri and chloramphenicolfor the pHK22 derivatives. About 10% of these colonies nolonger produced colicin V. To map the TnS insertion sites,BglII, HindIII and BamHI digests were done, thereby usingthe precisely known restriction sites within the IS50 of TnS(16) and the unique sites in pHK11 or pHK22.Complementation testing. Each cva : :Tn5 mutant from

pHKll was used to cotransform ZK4 in combination witheach cva::TnS mutant from pHK22. Transformants wereselected as being simultaneously resistant to kanamycin,amnpicillin, and chloramphenicol. From each transformation,six colonies were tested for colicin V production. The abilityto form a zone of growth inhibition was interpreted aspositive complementation. No cases of partial complemen-tation were observed. In every case, all six colonies testedgave the same results. In addition, we confirmed the pres-ence of two separate plasmids on several of the doubletransformants.

Preparation of minicells, protein labeling, and protein anal-ysis. Strain P678-54T- was transformed with pHK22 and

2466

on March 30, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

COLICIN V PRODUCTION GENES 2467

TABLE 1. Bacterial strains and plasmids

Strain or Genotype or Referenceplasmid description

MC4100 F- araD139 AlacUl69 rpsL 8relA thiA

ZK4 MC4100 recA56 This workDB1358 F- ilv::Tn5 28P678-54T- thr leu thiA supE lacYffhuA 1

gal mal xyl ara mtl thy minJM83 ara A(lac-proAB) rpsL 4)80 33

lacZM1571-18 A(lac-proAB) thi supE F' lacIq 33

lacZM15

pColV-K30 ColV+, 90 kb 31pBR322 Apr Tcr, 4.3 kb 7pACYC184 Cm' Tcr 9pUC19 Apr LacZa' 33pMK16 Kmr Tcr 17M13tgl31 LacZa+ 18

appropriate pHK22:: TnS insertion mutants. Minicells wereisolated and labeled for 10 min as described previously (23).Labeled proteins were electrophoresed on either 21 or 12.5%sodium dodecyl sulfate (SDS)-polyacrylamide gels (19) at130 V. Prestained size standards were purchased fromBethesda Research Laboratories, Inc., Gaithersburg, Md.

RESULTS

Cloning of colicin V genes. HindIII fragments from acomplete digest of pColV-K30 DNA were ligated withHindIII-cut pBR322. This ligation mixture was used totransform strain MC4100. Ampicillin-resistant transformantswere selected and screened for colicin V production on thebasis of zones of growth inhibition surrounding the colonies.Apparent colicin-producing clones were further purified andtested for colicin V activity and immunity. Plasmid DNAwas then subjected to restriction analysis. For each of theclones tested, the pBR322 plasmid derivative contained a20-kb HindIII fragment insertion. The two possible orienta-

kb

pHK4

pHK I I

0

_ Fto

.' ax l,l

5

EE

10

mua ,_)

tions of the fragment inserted were designated pHK3 andpHK4.

After a restriction map of pHK4 was generated (Fig. 1),deletions were made in the plasmid by cutting withendonucleases and religating. One deletion, a SalI digestionfollowed by religation, still maintained both colicin V pro-duction and immunity. This plasmid was called pHKll (Fig.1). The growth inhibition zones generated by pHKll andother high-copy-number clones were larger than those ob-served with the wild-type pColV-K30.

Subcloning of pHK11. Three fragments of pHKll weresubcloned separately into pMK16, and the resultant con-structs were assayed for colicinogenic and immunity prop-erties (Fig. 2). The plasmids containing 2.2- or 1.5-kbBamHI-EcoRI fragments did not confer colicin productionor immunity properties on the cells. Cells which containedthe SalI-BamHI (5.2-kb) fragment did not produce colicin Vbut were immune to colicin V (Cva- Cvi+). Similarly, theKpnI-BglII (4.2-kb) fragment cloned into pUC19 resulted incells (strain JM83) with a Cva- Cvi+ phenotype. However,cells (strain 71-18) with M13tgl31 containing the HindIll-BglIl (7.2-kb) fragment were Cva+ Cvi+.TnS mutagenesis and complementation analysis. To better

localize the colicin production region and to perform com-plementation experiments, we constructed pHK22, a deriv-ative of pACYC184 which contains the same Sail-HindIll(9.4-kb) fragment as pHKll and is compatible with pHK11.Inhibition zones from strains harboring pHK22 were approx-imately the same size as those from strains containingpHK11. Both pHKll and pHK22 were mutagenized withTnS as described in Materials and Methods. Only twophenotypes with respect to colicin V production were ob-tained: colonies which still produced normal-sized growthinhibition zones and colonies which completely failed toproduce growth inhibition zones.The insertion sites for 11 cva::TnS mutants in pHKll and

for 13 such mutants in pHK22 were determined by restric-tion site analysis (Fig. 2). All Cva- insertions mapped withinthe BglII-KpnI (4.2-kb) fragment. The results of complemen-tation experiments with all possible combinations of mutantsindicated' that there were three distinct complementationgroups. Insertion mutations pHK11-5, -7, -10, and -8 and

15

SEa

20

a .. u

24

o,

-4FeSi

Cva Cvi

+ +

+ 4-

pHK 12

pHK 3

pHK 14

+

FIG. 1. Restriction map of pHK4 and its derivatives. Plasmids derived from pHK4 were made by single restriction enzyme digestsfollowed by religation. Vector pBR322 is represented by a heavy line.

VOL. 169, 1987

_toj o

iI

on March 30, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

2468 GILSON ET AL.

0

kb

la

2 4

_ i

E0

hi

pHKII ::Tn5

6

,0

4* + 15 108 11 47

FE

433 6

a 94l

aCO)A

2 91

pHK22::Tn5

cva

3 8 2 5 413 9 14

10

7 11 6

.. A .-... B -

£tiFIG. 2. Restriction map of the HindIII-SalI (9.4-kb) fragment cloned into pBR322 and pACYC184 to generate pHKll and pHK22,

respectively. Open arrows indicate sites of TnS insertions in each plasmid which gave a Cva- phenotype. The filled arrow represents the siteof the Cva+ insertion closest to the KpnI site. The approximate boundaries for the three cva complementation groups, cvaA, cvaB, and cvaC,are the results of complementation analysis and are represented by the dotted lines. The dashed line represents the 700-bp region which whencloned confers ColV immunity on host cells.

pHK22-8, -2, and -13 did not complement each other butwere complemented by all other TnS mutations. Likewise,pHK1-11, -4, -3, -6, and -2 and pHK22-1, -5, -9, -10, -4, -14,-7, and -11 formed another complementation group, and a

third group was formed by pHKll-9 and -1 and pHK22-6.We have designated these contiguous genes cvaA, cvaB, andcvaC (Fig. 2). The left-hand boundary of cvaA must liebetween pHK22-3, a Cva+ TnS insertion, and the Kpnl site,while the right-hand boundary lies between pHK11-8 andpHK11-11. The boundaries for cvaB are between pHK11-8and pHK11-11 on the left and between pHK11-2 andpHK11-9 on the right. The cvaC gene boundaries must liebetween pHK11-2 and pHK11-9 on the left and betweenpHK22-6 and the BglII site on the right (Fig. 2). Theapproximate gene sizes are estimated as 850 to 1,450 bp forcvaA, 1,900 to 2,550 bp for cvaB, and 150 to 1,050 bp forcvaC.

Localization of colicin V activity. Colicin V activity inculture supernatants and lysed cells was quantitated forpHK11 and pHK11 cva::TnS mutants to determine whetherany of the cva genes are involved in the release of colicin V(Table 2). While mutations in cvaA (pHK11-8) and cvaB(pHK11-4) resulted in an approximately twofold reduction incolicin V activity in lysed cells, they prevented any detect-able activity in the supematant. A mutation in cvaC(pHK11-1) eliminated activity in both the supernatant andthe cells. The colicin V structural gene therefore appears tobe cvaC, while cvaA and cvaB are responsible for its release.Mapping of the immunity region. All cva:: Tn insertion

mutants were tested for immunity to colicin V and were

Cvi+. A BamHI-BglII deletion in pHK11 (Fig. 2) gave a Cvi-phenotype. We then used the HindIll site from within the

IS50 of pHK11-1 to clone a 1.9-kb HindIII-BglII fragmentcontaining 1,195 bp from TnS and 700 bp of ColV DNA intoanother vector, pUC19. When JM83 cells were transformedwith this pUC19 derivative, the cells were Cvi+. Therefore,the cvi gene(s) lies within the 700-bp region between theBglII site and the rightmost TnS insertion in pHK11,pHK11-1 (Fig. 2).

Identification of cva gene products. The protein products ofpHK11, pHK22, and the TnS insertion derivatives wereanalyzed in minicells. Two or three mutants in each comple-mentation group were analyzed in this way with identicalresults. The results shown in Fig. 3 and 4 are from pHK22,pHKI22-6, pHK22-9, and pHK22-8. The 12.5% SDS-polyacrylamide gel shows two proteins, with masses of 27and 43 kilodaltons (kDa), which are affected by a disruptionof the cvaA gene (Fig. 3). In Fig. 4, the 21% gel shows aprotein of approximately 6 kDa as a gene product for cvaC.Lane 5 of Fig. 4, which contains plasmid proteins from cvaC

TABLE 2. Colicin V activity

Colicin V activity (U)a in:Plasmid Genotype

Lysed cells Supernatant

pHKll cva + 16 32pHK11-8 cvaA 10 <1pHK11-4 cvaB 10 <1pHK11-1 cvaC <1 <1pHK11A BamHI- cvaABC <1 <1

BglIla Units determined by the critical dilution method (22) as described in

Materials and Methods.

J. BACTERIOL.

on March 30, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

COLICIN V PRODUCTION GENES 2469

1 2 3 4

-68

.-a~~~ ~ ~ ~ ~~~.E e :-~~...... -25.7T~~~~~~~~~ ... ... ....... . . . j2.-14.3

~~>js= -8.

-43

FIG. 4. Identification of gene product from cvaC and cvi on a21% SDS-polyacrylamide gel. Lanes show labeled proteins fromminicells harboring pHK22 (lane 1); pHK22-6 (lane 2); pHK22-9(lane 3); pHK22-8 (lane 4), pHK11-1 with a SalI deletion using a Sallsite from within Tn5 (see Fig. 2) (lane 5); and a pUC19 derivativewith the HindIII-BglII fragment from pHK11-1 (as described in thetext) (lane 6). Two proteins of approximately 6,000 and 7,000 MWare marked by arrows; these are the putative cvaC and cvi products,respectively. Size standards are indicated in kilodaltons.

FIG. 3. Identification of gene products from cvaA by 12.5%SDS-polyacrylamide gel analysis of 35S-labeled proteins fromminicells harboring pHK22-derived plasmids prepared as describedin Materials and Methods. Lanes: 1, pHK22; 2, pHK22-6; 3,pHK22-9; 4, pHK22-8. The arrows indicate the positions of the 27-and 43-kD gene products of cvaA. Size standards are indicated inkilodaltons.

cvi mutant cells, lacks both the 6,000-molecular-weight(MW) protein and an approximately 7,000-MW protein. Thelatter protein is seen as a very faint band produced by thepUC19 plasmid with the 1.9-kb cvi-containing insert de-scribed above. Therefore, this protein is likely to be respon-sible for colicin V immunity.

DISCUSSIONWe have identified a 4.2-kb region of plasmid pColV-K30

which is responsible for the production of colicin V andimmunity to it. TnS mutagenesis and complementation anal-ysis defined three complementation groups involved in nor-mal colicin V production. The three genes defined by thecomplementation groups, cvaA, cvaB, and cvaC, are en-

coded contiguously (Fig. 2) and could be transcribed as partof an operon. The lack of strong polarity by the TnSinsertions suggests that each gene has its own transcriptionalpromoter. However, instances in which transposon inser-tions do not show polarity have been described (27). Theseinstances could be due to outward promoters from the endsof transposons (4, 10) or to transcription antitermination(20). The immunity gene, cvi, is within a 700-bp regionadjacent to, or possibly overlapping, the cvaC gene (Fig. 2).

Mutations in cvaA abolished the production in minicells oftwo proteins of approximately 27 and 43 kDa. It is possiblethat these proteins are the products of overlapping genes. Noobvious product change was apparent in the cvaB mutants.Mutations in cvaC abolished a protein of approximately6,000 MW. A deletion in the cvi region resulted in thedisappearance of a 7,000-MW protein which can be seen inpreparations from a plasmid with the cloned cvi gene.

In an earlier study of the genes necessary for colicin Vactivity and immunity, Frick et al. (13) obtained a plasmidessentially identical to pHKll from pColV-B188. In thatstudy, the structural gene for colicin V and the immunitygene were mapped to the EcoRI-BglII fragment (Fig. 2). Thefragment carrying these genes was then reduced to a 900-bpHaeIII-RsaI fragment. These 900 bp would cover the regionof cvaC and cvi defined in the present study. While the900-bp fragment coded for colicin V activity, this activitycould only be detected in extracts of lysed cells. In ourstudy, lysed-cell extracts from cvaA and cvaB mutantsshowed activity, while those from cvaC mutants did not.Supernatants from all three mutant cultures also showed nodetectable activity. Since strains harboring plasmids with thewild-type HindIII-SalI fragment yielded active colicin Vboth within the cells and in the supernatant, these resultssuggest that cvaC is the structural gene for colicin V and thatthe products of the other two genes, cvaA and cvaB, arenecessary for the proper export or release of the colicin.Comparison of the cloned region of pColV-B188 (13) withthat of pColV-K30 shows extensive homology. There aresome minor differences in the number of and distancebetween several restriction sites, which could reflectpolymorphisms between the two parent ColV plasmids.Most colicin plasmids contain a cluster of three genes

involved in the production and release of the colicin: thecolicin structural gene, the colicin immunity gene, and a lysisgene (25). The results described here indicate that the genesinvolved in colicin V production do not follow this pattern.Our work on the genetic arrangement of colicin V productionand immunity genes, along with the evidence that colicin Vis not induced by DNA damage (15), supports the suggestionthat colicin V can be considered a microcin (32). Microcinsconstitute a family of low-molecular-weight antibiotic sub-stances produced by many Enterobacteriaceae (2, 3). Unlikethe colicins, none of the microcins is inducible by DNA-damaging agents.

Further research into the production and function ofcolicin V and microcins in general should prove to be usefulin the understanding of peptide and small-protein transportin E. coli. It is likely that colicin V must cross the doublemembrane and cell wall of the producer cell and must then

.1.

VOL. 169, 1987

:1::m ::::::-::w-

'44..s; :s-

--

on March 30, 2020 by guest

http://jb.asm.org/

Dow

nloaded from

2470 GILSON ET AL.

span at least part of the envelope barrier of the susceptiblecell. The export and uptake of this protein could be differentfrom that of the larger colicins, which are thought to firstaccumulate in the producing cell and then to rely on a lysisgene product for export (25).

ACKNOWLEDGMENTS

We thank Geraldine Duncan for assistance in the initial charac-terization of colicin V clones.

This work was supported by an American Cancer Society grant toR.K. L.G. was the recipient of a Public Health Service predoctoraltraining grant from the National Institutes of Health, and R.K. wasa Charles King Trust Fellow.

LITERATURE CITED1. Adler, H. I., W. D. Fisher, A. Cohen, and A. A. Hardigree. 1967.

Miniature Escherichia coli cells deficient in DNA. Proc. Natl.Acad. Sci. USA 57:321-326.

2. Asensio, C., J. C. Perez-Diaz, M. C. Martinez, and F. Baquero.1976. A new family of low molecular weight antibiotics fromenterobacteria. Biochem. Biophys. Res. Commun. 69:7-14.

3. Baquero, F., and F. Moreno. 1984. The microcins. FEMSMicrobiol. Lett. 23:117-124.

4. Berg, D. E., A. Weiss, and L. Crossland. 1980. Polarity of TnSinsertion mutations in Escherichia coli. J. Bacteriol. 142:439-446.

5. Binns, M. M., D. C. Davies, and K. G. Hardy. 1979. Clonedfragments of the plasmid ColV, I-K94 specifying virulence andserum resistance. Nature (London) 279:778-781.

6. Birnboim, H. C., and J. Doly. 1979. A rapid alkaline extractionprocedure for screening recombinant plasmid DNA. NucleicAcids Res. 7:1513-1517.

7. Bolivar, F., R. L. Rodriguez, P. J. Greene, M. C. Betlach, H. L.Heynecker, and H. W. Boyer. 1977. Construction and character-ization of new cloning vehicles. II. A multipurpose cloningsystem. Gene 2:95-113.

8. Casadaban, M. J. 1976. Transposition and fusion of the lacgenes to selected promoters in Escherichia coli using bacterio-phage lambda and Mu. J. Mol. Biol. 104:541-555.

9. Chang, A. C. Y., and S. N. Cohen. 1978. Construction andcharacterization of amplifiable multicopy DNA cloning vehiclesderived from the P15A cryptic miniplasmid. J. Bacteriol.134:1141-1156.

10. Ciampi, M. S., M. B. Schmid, and J. R. Roth. 1982. TransposonTnlO provides a promoter for transcription of adjacent se-quences. Proc. Natl. Acad. Sci. USA 79:5016-5020.

11. Clancy, J., and D. C. Savage. 1981. Another colicin V pheno-type: in vitro adhesion of Escherichia coli to mouse intestinalepithelium. Infect. Immun. 32:343-352.

12. Currier, T. C., and E. W. Nester. 1976. Isolation of covalentlyclosed circular DNA of high molecular weight from bacteria.Anal. Biochem. 76:431-441.

13. Frick, K. K., R. L. Quakenbush, and J. Konisky. 1981. Cloningof immunity and structural genes for colicin V. J. Bacteriol.148:498-507.

14. Hardy, K. G. 1975. Colicinogeny and related phenomena.Bacteriol. Rev. 39:464-515.

15. Herschman, H. R., and D. R. Helinski. 1967. Comparative studyof the events associated with colicin induction. J. Bacteriol.94:691-699.

16. Jorgensen, R. A., S. J. Rothstein, and W. S. Reznikoff. 1979. Arestriction enzyme cleavage map of Tn5 and localization of aregion encoding neomycin resistance. Mol. Gen. Genet. 177:65-72.

17. Kahn, M., R. Kolter, C. Thomas, D. Figurski, R. Meyer, E.Remaut, and D. R. Helinski. 1979. Plasmid cloning vehiclesderived from plasmids ColEl, F, R6K and RK2. MethodsEnzymol. 68:268-280.

18. Kieny, M. P., R. Lathe, and J. P. Lecocq. 1983. New versatilecloning and sequencing vectors based on bacteriophage M13.Gene 26:91-99.

19. Laemmli, U. K. 1970. Cleavage of structural proteins during theassembly of the head of bacteriophage T4. Nature (London)227:680-685.

20. Li, S. C., C. L. Squires, and C. Squires. 1984. Antitermination ofE. coli rRNA transcription is caused by a control regionsegment containing lambda nut-like sequences. Cell 38:851-860.

21. Maniatis, T., E. F. Fritsch, and J. Sambrook. 1982. Molecularcloning: a laboratory manual. Cold Spring Harbor Laboratory,Cold Spring Harbor, N.Y.

22. Mayr-Harting, A., A. J. Hedges, and R. C. W. Berkeley. 1972.Methods for studying bacteriocins. Methods Microbiol.7A:315-422.

23. Meagher, R. B., R. C. Tait, M. Betlach, and H. W. Boyer. 1977.Protein expression in E. coli minicells by recombinant plasmids.Cell 10:521-536.

24. Miller, J. H. 1972. Experiments in molecular genetics. ColdSpring Harbor Laboratory, Cold Spring Harbor, N.Y.

25. Pugsley, A. P. 1984. The ins and outs of colicins. I. Production,and translocation across membranes. Microbiol. Sci. 1:168-175.

26. Quakenbush, R. L., and S. Falkow. 1979. Relationship betweencolicin V activity and virulence in Escherichia coli. Infect.Immun. 24:562-564.

27. San Millan, J. L., R. Kolter, and F. Moreno. 1985. Plasmid genesinvolved in microcin B17 production. J. Bacteriol. 163:1016-1020.

28. Silverman, M., and M. Sinon. 1980. Phase variation: geneticanalysis of switching mutants. Cell 19:845-854.

29. Smith, H. W. 1974. A search for transmissible pathogeniccharacters in invasive strains of Escherichia coli: the discoveryof a plasmid-controlled toxin and plasmid-controlled lethalcharacter closely associated or identical with colicin V. J. Gen.Microbiol. 83:95-111.

30. Smith, H. W., and M. B. Higgins. 1976. Further observations onthe association of the colicin V plasmid of Escherichia coli withpathogenicity and with survival in the alimentary tract. J. Gen.Microbiol. 92:335-350.

31. Williams, P. H. 1979. Novel iron uptake system specified byColV plasmids: an important component in the virulence ofinvasive strains of Escherichia coli. Infect. Immun. 26:925-932.

32. Yang, C. C., and J. Konisky. 1984. Colicin V-treated Esche-richia coli does not generate membrane potential. J. Bacteriol.158:757-759.

33. Yanisch-Perron, C., J. Vieira, and J. Messing. 1985. ImprovedM13 phage cloning vectors and host strains: nucleotide se-quence of the M13mpl8 and pUC19 vectors. Gene 33:103-119.

J. BACTERIOL.

on March 30, 2020 by guest

http://jb.asm.org/

Dow

nloaded from