Klebsiella pneumoniae - Journal of Bacteriology - American Society ...

2 downloads 0 Views 649KB Size Report
Received 3 December 1984/Accepted 13 June 1985. Plasmid pCUl was Kik+ (promotes killing of KlebsieUa pneumoniae). All TnS insertions within the tra ...
Vol. 163, No. 3

JOURNAL OF BACTERIOLOGY, Sept. 1985, p. 1296-1299

0021-9193/85/091296-04$02.00/0 Copyright © 1985, American Society for Microbiology

Regions

on

Plasmid pCU1 Required for the Killing of Klebsiella pneumoniae V. THATTE, S. GILL, AND V. N. IYER*

Department of Biology and Institute of Biochemistry, Carleton University, Ottawa, Ontario, Canada KJS 5B6 Received 3 December 1984/Accepted 13 June 1985

Plasmid pCUl was Kik+ (promotes killing of KlebsieUa pneumoniae). All TnS insertions within the tra region of pCUl were Kik-. Two other regions, kikA and kikB, were needed. They may be separated on different lilasmids, but both must be mobilized into KlebsieUa pneumoniae. Establishment of one kik region in K. pneumoniae followed by receipt of the second did not lead to killing. Kik was therefore intracellular and required concerted and transient action of both regions.

Incompatibility group N plasmids in Escherichia coli and closely related bacteria usually determine a particular conjugative transfer system, the N tra system. This transfer system operates efficiently when bacterial matings are conducted on the surfaces of solid media (5) or in liquid media providing increased surface tension (14). It has beenl shown that IncN plasmids also determine the production of short, pointed, and rigid pili on the cell surface of their bacterial hosts (3), and the presence of these pili is positively correlated with mating proficiency. These pili provide sites for the attachment of different groups of bacteriophages (3, 9, 12). Plasmids of this incompatibility group have also been shown to mediate the killing of Klebsiella pneumoniae when the latter is used as a recipient in conjugations with N+ strains as donors (13). Of 34 naturally occurring IncN group plasmnids we tested, 33 had this property. The exceptional plasmid was transfer deficient. To begin to understand the basis of this phenotype, we undertook a structural and functional analysis of pCU1, one of the better-studied members of this group (11), by transposon and deletion mutagenesis and molecular cloning. The study identified two separate regions on pCU1, both of which are required for killing K. pneumoniae and provide clues to the mechanism of killing. The E. coli strains and the materials and methods used for transposon and deletion mutagenesis and structural analysis of the plasmid were as described by Thatte et al. (16). Strains and methods used for assaying killing were described previously (13). A region of pCU1 between 27.5 to 11.2 kilobases (kb) (Fig. 1) was cloned to give rise to the hybrid pCU109, which contains the entire tra region of pCU1 but with the replicon of the nonconjugative plasmid pACYC184 substituted for that of pCU1. The derivation of this plasmid is shown in Fig. 2. E. coli carrying pCU109 killed K. pneumoniae efficiently. Over a hundred different TnS insertions mapping to different complementation groups of the tra region of either pCU1 or pCU109 (16) abolished killing. Therefore, either the products of the tra genes when present in E. coli or transfer of the plasmid to K. pneumoniae is necessary for killing. Some of the Tra- TnS insertion derivatives of either pCU1 or pCU109 produced N pili, and others did not (16). The collection of Tn5 insertions in pCU1 also yielded three different plasmid derivatives (pCU167, pCU184, and pCU66; Fig. 1). Strains carrying any one of these plasmids did not kill K. pneumoniae. The three insertions mapped at *

Corresponding author.

different positions closely clustered between 31.0 to 32 kb at one end of tra (Fig. 1). To determine the orientation of TnS in these derivatives, plasmid DNA from each was digested with BamHI and transformed into C600T (1). Kanamycinresistant (Ktnm colonies were screened for the loss of spectinomycin resistance, streptomycin resistance, and ampicillin resistance (Spr, Smr, and Apr, respectively) (note the BamHI site of TnS was asymmetrically placed and did not inactivate kanamycin resistance; reference 2). DNA isolated and analyzed from such derivatives (pCU401, pCU402, and pCU403) indicated that the region in the interval of approximately 21 to 31.5 kb was deleted in them (Fig. 1). This showed that in all three derivatives, the insertion of TnS was clockwise. The findings do not allow a definitive discrimination between insertions within this region and insertions which may be in its proximity and polar on its expression. However, the results do show that the conjugal transfer functions of the plasmid are not sufficient for this phenotype. When this phenotype was first described in 1981, we referred to it as the Kil phenotype (13). Recently, however, a phenotype that is operationally different has also been called Kil. In this phenotype that has been observed in pKM101, a plasmid that is closely related to pCU1, a particular region from the plasmid could not be cloned without other regions from the same plasmid being present in the cell. Both in pKM101 (17) and IncP group plasmids (6, 15), such nonclonable regions have been called kil, and their cognate-protecting region(s) have been called kor. To reduce confusion in the literature, we will hereafter refer to the killing of K. pneumoniae as the Kik phenotype and the regions determining this phenotype as kikA and kikB. The three deletion derivatives pCU401, pCU402, and pCU403 (Fig. 1) were Tra+ but Kik-. pCU171 is a Tra- Kik- TnS insertion derivative within 0.5 kb of the insertion in pCU166 which is Tra+ Kik-; kikA is thus next to the left end of tra (Fig. 1). The plasmid pCU101 is a derivative of pCU1 that contains all of the regions present in pCU109 (including kikA) except for a continuous 3-kb region at the end of tra opposite to kikA (Fig. 1). This plasmid does not promote killing of K. pneumoniae. This 3-kb region, therefore, defines a second region, kikB, that is also required for killing K. pneumoniae. The kikB region may actually be only a very small part of this 3-kb region, because 33 TnS insertions (16) in this region did not affect the Kik phenotype. The end of tra closest to kikB but separated from it included a region of pCU1 that must be present in cis for transfer (Fig. 1). By analogy with a similar region present on 1296

VOL. 163, 1985

NOTES

1297

FIG. 1. Plasmid derivatives that were constructed and used to ascertain the number and position of regions on pCU1 that were functionally necessary for the Kik phenotype. See Table 1 for the Kik phenotypes of these derivatives. The lines with circles at the end represent Tn5 insertions that were mapped to the positions indicated. Open circles indicate that the Tra phenotype was not affected, and solid circles indicate that it was. The respective pCU designations of these insertion derivatives are shown by the side of each circle. Only a small fraction of TnS insertions that were isolated is shown. The broken line indicates regions that were deleted to give rise to the particular deletion derivatives that are indicated by numbers within each triangle. The six outermost arcs of solid lines containing or not containing broken lines (deletions) indicate regions present in cloned derivatives (vector, pACYC184). Abbreviations: Tra, conjugal transfer; kikA and kikB, regions required for the killing of K. pneumoniae; bom, a region required in cis for plasmid transfer by conjugation; Ap, Sp, and Sm, regions determining resistance to ampicillin, spectinomycin, and streptomycin, respectively; Rep, a region required for plasmid maintenance; BI, BamHI; Bg,

BgIII; HIII, HindIII; and K, KpnI.

the plasmid ColEl (4), this region on pCU1 is called bom. E. coli carrying kikA and kikB on separate compatible plasmids will kill K. pneumoniae but only provided both have a bom region and at least one of the two plasmids provides the tra functions (Table 1). We have also done experiments in which the positions of kikA and kikB have been rearranged on a single plasmid. E. coli carrying this plasmid (pCU114; Table 2) killed K. pneumoniae efficiently. When pCU101 (carrying tra, bom, and kikA) was mated with K. pneumoniae carrying an established pCU404 (carrying bom and kikB), no killing was detected although it could be shown that the resultant K. pneumoniae not only carried both plasmids but, in a second mating experiment, could also efficiently kill a nalidixic acid-resistant K. pneumoniae strain. Similar results were obtained in a reciprocal experiment, that is, when a kikB region (in pCU167) was introduced into an established K. pneumoniae strain carrying kikA (pCU58 or pCU53 introduced previously by transformation). Thus, for killing to occur, not only must both kikA and kikB be transferred to K. pneumoniae, but such a transfer must occur simultaneously. Several conclusions can be derived concerning the mechanism of the Kik phenotype. (i) The function of N-type pili in mating with K. pneumoniae is not intrinsically lethal as has been suggested (14). This is inferred from the fact that all Tra- mutants do not kill K. pneumoniae, regardless of

whether or not they produce functional N pili, and the observation that nalidixic acid, which inhibits pCUl DNA transfer (7), also inhibits killing (13). (ii) Killing occurs within K. pneumoniae because it depends on the transfer of both kikA and kikB to K. pneumoniae. (iii) Killing is associated with the transient expression of kikA and kikB in K. pneumoniae. (iv) Methods for DNA-mediated transformation of K. pneumoniae are inefficient; this has prevented us from using them for the direct delivery of kikA and kikB into K. pneumoniae. Since the method of delivery has depended on the N tra region, which itself also enters K. pneumoniae as a helper, we cannot at present exclude a killing role for a gene or genes from the tra region in addition to kikA and kikB. (v) TnS insertions in the kikA region abolish the Kik phenotype and also PRD1 sensitivity. The relationship between these two phenotypes is unknown at present, although the insertion mutants do produce N pili and are transfer proficient. The PRD1 resistance of these insertion mutants can be complemented in trans (Table 1). This suggests that kikA function is mediated by a diffusible product. (vi) pCUl is structurally related to pMK101, although the plasmids have different origins (10). In pKM101, loci called kilA and kilB, each of which can apparently kill E. coli in the absence of overriding loci called korA and korB, have been reported (17). It is possible that similar regions exist on pCU1.

1298

J. BACTERIOL.

NOTES

pcu I Km

FIG. 2. Diagram illustrating the methods used for the construction of plasmid derivatives containing Tn5 insertions (triangles) at different sites on pCUl (pCU29 and pCU88), a deletion derivative of pCU29 (pCU51), or containing different regions of pCUl cloned in pACYC184 (pCU101, pCU109, pCU58, and pCU114). The position of the triangle indicates the site of TnS insertion. kikA and kikB, Regions required for the killing of K. pneumoniae; bom, region required in cis for plasmid transfer; Cm, chloramphenicol; Km, kanamycin; B, BamHI; K, KpnI; H, HindIlI. See Table 1 for the Kik phenotype of the derivatives.

pKM101 is Kil- but Kik+; (ii), in pKM101, no kil region has been found at a position comparable to that of kikB of pCU1; however, at such a position in pKM101, a region analogous to kikB of pCU1 has been found (S. C. Winans and G. C.

However, it is unlikely that kilA and kilB are analogous to kikB in these plasmids for the following reasons. (i) Although kilA of pKM101 maps at a position comparable to that of kikA of pCU1, a mutant with an insertion mapping at kilA of

TABLE 1. Physical and genetic regions present on pCU1 derivatives in relation to their ability to promote the killing of K. pneumoniae Relevant regions present on the plasmidb Phenotypec Position on pCU1 Plasmid(s) present in E. Tra Kik IKe kikB tra kikA (kb)a coli strain PRD1 bom

pCU1 pCU101 pCU51

0-39 27.5-8.2 8.2-21

pCU101 and pCU51

+ +

+

+ +

+

+ +

+ +

s s

s

-

-

+

-

-

r

r

_

+

+

-

+

+

s

s

+

+ +

-

+

s s r s

s s r s

+ +

+

+ + + +

+ +

-

+ + +

+

+

_

+

-

+

+

-

-

-

+

+

+

s s s

+ +

-

-

s s s

-

(-)

r r s

r r r

+

s

s

+

s

-

+

pCU114 pCU109 pCU404 pCU101 and pCU404 pCU53 pCU167 pCU53 and pCU167 pCU57 pCU58 pCU167 and pCU57

pCU167 and pCU58

Rearranged pCU29 27.5-11.2 7-11.5 27.3-3.3 pCU1::Tn5

0-8.2 27.7-34 and 5.1-8.2

+ +

+

-

-

(-) -

+ -

+ -

+

-

-

(-)

+

+

-

+ + + + +

(-)

+

+

+

+

-

-

+

See Fig. 1 for the relative position of these coordinates on pCU1. (-), The kikA region is present but functionally inactivated by the TnS insertion. c r, Resistance; s, sensitivity; +, proficient; -, nonproficient.

a

b

-

+

NOTES

VOL. 163, 1985

1299

TABLE 2. Bacterial plasmids Plasmid

pCU1 and its derivatives pCU1 pCU29 pCU1O1

pCU88 pCU109

Derivation and reference or source

Relevant phenotype conferred'

Apr Smr Spr Tra+; pCU1 replicon Apr Smr Spr Tra+ Kmr (Tn5); pCU1 replicon Cmr Tra+ plSA replicon Apr Smr Spr Tra+ Kmr (Tn5); pCU1 replicon Cmr Tra+ plSA replicon

(11) TnS insert at 8.2 kb of pCU1 (Fig. 1) 19.4-kb Hindlll fragment of pCU29 cloned into pACYC184 (Fig. 2) TnS insert at 11.2 kb of pCU1 (Fig. 1) 22.4-kb HindlIl fragment of pCU88 cloned into pACYC184 (Fig. 2) TnS insert at 32.5 kb of pCU1 (Fig. 1) TnS inserts at 32.0 to 33.0 kb of pCU1 (Fig. 1)

Apr Smr Spr Kmr (Tn5) Tra-; pCU1 replicon Apr Smr Spr Kmr (TnS) Tra+ IKes PRDlr; pCU1 replicon BamHI deletion derivative of pCU29 (Fig. 2) Kmr Tra- Bom- pCU1 replicon CUS1 Cmr Kmr Tra+ pCU1 replicon plus plSA replicon Fusion of pCU101 and pCU51 at BamHI site pCU114 KpnI deletion derivatives of pCU101 and Cmr Tra- Bom+; p1SA replicon pCU58 and pCU405 pCU109, respectively (Fig. 1) Tn5 insert at 30.5 kb of pCU1 Apr Smr Spr Tra+ KMr (Tn5); pCU1 replicon pCU97, pCU79, and pCU94 BamHI deletion derivatives of pCU167, pCU184, pCU401, pCU402, and pCU403 Kmr Tra+ IKes PRD1' pCU1 replicon and pCU66, respectively Fusion of 3.8-kb BglII fragment of pCU1 at Apr Kmr Tra- Bom+ ColEl replicon pCU404 BamHI site of pMK2004 (17) Plasmids used as cloning vectors

pCU171 pCU66, pCU167, and pCU184

(nonconjugative) pACYC184 pMK2004

(6) (8)

Cmr Tcr Apr Tcr Kmr

a Abbreviations: Ap, ampicillin; Cm, chloramphenicol; Km, kanamycin; Sm, streptomycin; Sp, spectinomycin; Tc, tetracycline; Tra, conjugal transfer; 13om, basis of mobilization; IKe and PRD1, N-pilus-specific bacteriophages; r or s, resistance or sensitivity; +, proficient; -, nonproficient.

Walker, personal communication); (iii) the kilB region of pKM101 maps at a position within its tra region and different from the position of kikB of pCU1. The physiological mechanisms of action of kil genes and kik genes in IncN group plasmids and their relationship, if any, to one another remains to be explored. This study was supported by an Operating Research Grant from the Medical Research Council of Canada and by an Equipment Grant from the Natural Sciences and Engineering Research Council of Canada. We are grateful. For gifts of antibiotics, we thank Bristol Myers of Canada. We also acknowledge with appreciation the enthusiastic participation of Minna Rotheirn and Bruce Love in discussions in the summer of 1984 that bear on some of these observations. LITERATURE CITED 1. Appleyard, R. K. 1954. Segregation of new lysogenic types during growth of a lysogenic strain derived from Escherichia coli K-12. Genetics 13:17-28. 2. Berg, D. E., and C. M. Berg. 1983. The prokaryotic transposable element Tn5. Biotechnology 1:417-435. 3. Bradley, D. E. 1979. Morphology of pili determined by the N incompatibility group plasmid N3 and interaction with bacteriophage PR4 and IKe. Plasmid 2:632-636. 4. Clark, A. J., and G. H. Warren. 1979. Conjugal transmission of plasmids. Annu. Rev. Genet. 13:99-125. 5. Dennison, S., and S. Baumberg. 1975. Conjugational behaviour of N plasmids in Escherichia coli K-12. Mol. Gen. Genet.

138:323-331. 6. Figurski, D. H., R. F. Pohlnan, D. H. Bechhofer, A. S. Prince, and C. A. Kelton. 1982. The broad host range plasmid RK2 encodes multiple kil genes potentially lethal to Escherichia coli

host cells. Proc. Natl. Acad. Sci. USA 79:1935-1939. 7. Gill, S., and V. N. Iyer. 1982. Nalidixic acid inhibits the conjugal transfer of conjugative N incompatibility group plasmids. Can. J. Microbiol. 28:256-258. 8. Kahn, M., R. Kolter, C. Thomas, D. Figurski, R. Meyer, E. Remaut, and D. R. Helinski. 1979. Plasmid cloning vehicles derived from plasmids ColEl, F, R6K, RK2. Methods Enzymol. 68:268-280. 9. Khatoon, H., R. V. Iyer, and V. N. Iyer. 1972. A new filamentous bacteriophage with sex factor specificity. Virology 48:145-155. 10. Konarska-Kozlowska, M., and V. N. Iyer. 1981. Physical and genetic organization of the IncN-group plasmid pCU1. Gene 14:195-204. 11. Konarska-Kozlowska, M., V. Thatte, and V. N. Iyer. 1983. Inverted repeats in the DNA of plasmid pCU1. J. Bacteriol. 153:1502-1512. 12. Olsen, R. H., J.-S. Siak, and R. H. Gray. 1974. Characteristics of PRD1, a plasmid-dependent broad host range DNA bacteriophage. J. Virol. 14:689-699. 13. Rodriguez, M., and V. N. Iyer. 1981. Killing of Klebsiella pneumoniae mediated by conjugation with bacteria carrying antibiotic resistance plasmids of the group N. Plasmid

6:141-147.

14. Singleton, P. 1983. N-mating in Escherichia coli is promoted by foaming. J. Gen. Microbiol. 129:3697-3699. 15. Smith, C. A., and C. M. Thomas. 1983. Deletion mapping of kil and kor in the trfA and trIB regions of broad host-range plasmid RK2. Mol. Gen. Genet. 190:245-254. 16. Thatte, V., D. E. Bradley, and V. N. Iyer. 1985. N conjugative transfer system of plasmid pCUt. J. Bacteriol. 163:1229-1236. 17. Winans, S. C., and G. C. Walker. 1985. Identification of pKM101-encoded loci specifying potentially lethal gene products. J. Bacteriol. 161:417-424.