Vol. 172, No. 2

Natural Transformation in Campylobacter SpeciesYING WANG' AND DIANE E. TAYLOR' 2*

Departments of Microbiology' and Medical Microbiology and Infectious Diseases,2*University of Alberta, Edmonton, Alberta, Canada T6G 2H7

Received 14 July 1989/Accepted 1 November 1989

Growing cells of Campylobacter coli and C. jejuni were naturally transformed by naked DNA without therequirement for any special treatment. Transformation frequencies for homologous chromosomal DNA wereapproximately lo-3 transformants per recipient cell in C. coli and 1o-4 in C. jejuni. Maximum competence wasfound in the early log phase of growth. Campylobacters preferentially took up their own DNA in comparisonwith Escherichia coli chromosomal DNA, which was taken up very poorly. Three new Campylobacter spp.-to-E.coli shuttle plasmids, which contained additional cloning sites and selectable markers, were constructed fromthe shuttle vector pILL550A. These plasmid DNAs were taken up by campylobacters much less efficiently thanwas homologous chromosomal DNA, and transformation into plasmid-free cells was very rare. However, withthe use of recipients containing a homologous plasmid, approximately 10-4 transformants per cell wereobtained. The tetM determinant, originally obtained from Streptococcus spp. and not heretofore reported inCampylobacter spp., was isolated from an E. coli plasmid and was introduced, selecting for tetracyclineresistance, by natural transformation into C. coli.

Bacteria have evolved different mechanisms for taking upDNA from the environment and incorporating it into theirgenomes (for reviews, see references 22 and 24). Suchnatural transformation may provide a potential advantage forthe bacterium, because it enables an individual cell toaccumulate advantageous mutations originating from sepa-rate individuals or even to acquire genetic material fromother species. It has been successfully used in geneticstudies and molecular cloning in bacteria, such as in Bacillus(12) and Streptococcus (16) spp.

Natural transformation systems appear to fall into twogroups, one typical of gram-positive organisms and the otherfound in some gram-negative bacteria (22). There is a

marked difference in the process of binding and transport ofDNA through the envelope of the recipient cell. In Bacillussubtilis, Streptococcus pneumoniae, and S. sanguis, a singlecompetent cell can bind and take up a large number ofDNAmolecules regardless of their sources. During the uptakeprocess, one strand of the DNA is degraded while thecomplementary strand is transported into the cell (22, 24). Ofthe gram-negative bacteria that possess natural transforma-tion systems, Haemophilus species bind and take up DNApossessing a specific 11-base-pair sequence (7). DNA uptakeby Neisseria gonorrhoeae also involves recognition of a

specific 10-base-pair sequence (11). In both cases, only a fewmolecules of homologous DNA can be taken up by a

competent cell, and heterologous DNA can be taken up onlyat much lower frequency (10). Natural transformation ofplasmid DNA is normally rare in both gram-positive andgram-negative bacteria, because the duplex DNA is nickedand partially degraded during both binding to and enteringinto the cell (24). The frequency of plasmid transformationcan be increased by using a recipient containing a homolo-gous plasmid. Thus, the incoming plasmid can be rescued bythe resident plasmid through homologous recombination (12,16).

In genetic engineering studies, transformation is ofteninitiated by artificial means, e.g., calcium chloride treatment

* Corresponding author.

for Escherichia coli and some other gram-negative bacterialspecies (6), polyethylene glycol-mediated transformation forprotoplasts of gram-positive bacteria (4) or whole cells ofgram-negative bacteria (5), or electroporation with high-voltage discharge to introduce DNA into a cell (20). Theseartificial methods are exploited probably because most bac-terial species are not naturally transformable and becauseplasmid and bacteriophage DNAs are believed to be intactafter entering into a cell by these artificial pathways.Campylobacter coli and C. jejuni are gram-negative mi-

croaerophilic bacteria, frequently responsible for gastroen-teritis in humans. Shuttle vectors which can transfer from E.coli to C. jejuni were constructed by Labigne-Roussel et al.(14), and electroporation of C. jejuni with plasmid DNA was

demonstrated by Miller et al. (20). In a preliminary report,several C. coli strains were found to take up DNA from theirenvironment (A. D. E. Fraser and E. M. Riche, Campylo-bacter IV, abstr. no. 185, p. 333-334, Goterna, Kungalv,Sweden, 1988). In this study, we demonstrate that most C.coli and some C. jejuni strains are naturally competentduring the logarithmic phase of growth and that they showstrong selectivity for taking up their own DNA. Plasmid andchromosomal DNA transformation was investigated, severalcloning vectors were constructed, and the streptococcaltetM gene was transfered by natural transformation into C.coli, where it was found to express tetracycline resistance.

MATERIALS AND METHODSStrains and culture conditions. The Campylobacter strains

used in this study are listed in Table 1. E. coli JM107 (30) wasalso used. The plasmids used were pUC13 (30), pUA466(25), pJI3 (kindly provided by V. Burdett [3]), and the shuttlevector pILL550A (kindly provided by A. Labigne-Roussel),which is a derivative of pILL550 (14) containing additionalPstI and SmaI sites (W. Yan and Y. Wang, unpublisheddata). Plasmids pUOA11, pUOA13, pUOA15, and pUOA17were constructed in this study. C. coli and C. jejuni strainswere cultured in Mueller-Hinton (MH) broth or on MH agarat 37°C under 7% C02, and E. coli strains were grown in LBbroth or on LB agar at 37°C. When necessary, the mediumwas supplemented with ampicillin (100 p.g/ml), tetracycline

949

JOURNAL OF BACTERIOLOGY, Feb. 1990, p. 949-9550021-9193/90/020949-07$02.00/0Copyright © 1990, American Society for Microbiology

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

31

Dec

embe

r 20

21 b

y 18

3.98

.44.

140.

950 WANG AND TAYLOR

TABLE 1. Campylobacter strains used in this study

Strain Chromosomal Plasmid content Plasmid Source or referenceresistance phenotypea (size [kb]) phenotypea

C. coliUA417 Nalr _b H. Lior (Canada)UA417R Nalr StrF This studyUA420 M. A. Karmali (Canada)UA585 Eryr C. D. Ribeiro (Wales)UA724 Eryr Gmr Kmr pUA724 (ca. 30) - R. Gomez-Lus (Spain)BM2509 Eryr pIP1433 (47.2) Kmr Tcr P. Courvalin (France [29])

pIP1445 (4.57)

C. jejuniUA67C Nalr 26UA466 pUA466 (45) Tcr 25UA466R Nalr Strr pUA466 (45) Tcr This studyUA649 pUA649 (40.8) - pUA466AtetO (25)UA650 Plasmid-cured UA466 (25)UA697 Eryr I. Phillips (United Kingdom)C31 J. Miller (United States [20])a Abbreviations: Eryr, erythromycin resistance; Gmr, gentamicin resistance; Kmr, kanamycin resistance; Nalr, nalidixic acid resistance; Ster, streptomycin

resistance; Tcr, tetracycline resistance.b-, Not detected.' This strain was formerly called C. jejuni SD2.

(12 tag/m1), kanamycin (16 to 32 ,ug/ml), nalidixic acid (24pug/ml), and streptomycin (10 ,ug/ml).Plasmids and chromosomal DNA isolation. Plasmid DNA

extractions were carried out as described by Birnboim andDoly (2), followed by cesium chloride-ethidium bromidedensity gradient centrifugation when necessary. Chromo-somal DNA was prepared by the method of Marmur (18).

Transformation. Campylobacter transformation was per-formed either on an agar surface or in a biphasic system. Fortransformation on MH agar, fresh recipient cells (24-hgrowth on MH agar) were spread on MH agar at about 5 x107 cells per plate and incubated for 6 h. Aliquots of DNA(ca. 0.2 ,ug in 5 ,ul ofMH broth, TE buffer, or ligation buffer[17]) were spotted directly onto the inoculated agar withoutadditional mixing or spreading, and incubation was contin-ued for 5 h. For transformation in a biphasic system, cellsuspensions (1 x 107 to 5 x 107 cells per ml of MH broth)were transferred (0.2 ml per tube) to test tubes (10 by 120mm) containing 1.5 ml MH agar and incubated for 2 to 6 h.DNA samples were added, and incubation was continued for3 to 5 h. DNase I and MgCl2 were added to final concentra-tions of 25 ,ug/ml and 5 mM, respectively, at various times ifneeded. E. coli was transformed by the CaCl2 procedure (6).Transformants were selected on MH or LB agar containingappropriate antibiotics.DNA abIng and uptake experiments. 32P-labeled trans-

forming DNA was prepared in vitro by either nick transla-tion, end labeling, or the random primer labeling method.Nick translation was performed by the method of Maniatis etal. (17), except that only 1/50 of the amount of DNase I wereused and DNA was preincubated with DNase I at 15°C for 30min before DNA polymerase I, [ot-32P]dATP, dCTP, dGTP,and dTTP were added. End-labeled DNA was prepared byfilling in with DNA polymerase I (Klenow fragment) in thepresence of deoxynucleoside triphosphates. Plasmid DNAwas linearized by the restriction enzyme XbaI for end filling.The reaction (typically 0.5 pg of DNA in 20 ,ul of mixture)was carried out at room temperature for 20 min. Randomprimer labeling was performed as described previously (9),except that following the labeling, 0.1 mM cold deoxynu-cleoside triphosphates and 3 U of DNA polymerase I (Kle-

now fragment) were added and the reaction was continued atroom temperature for 30 min. All the labeled DNA sampleswere collected by ethanol precipitation with 1 ,ug of tRNAand were suspended in MH broth.[32P]DNA uptake was performed by using the biphasic

system. Typically, the competent cells were incubated with[32P]DNA (0.1 K.g/ml). At various times, DNase I and MgCl2were added for 1 min. The cells were centrifuged, washedtwice with TE buffer, dissolved in lysing solution (TE, 1%sodium dodecyl sulfate), and then counted in a liquid scin-tillation counter. Competition for DNA uptake was carriedout by using a procedure similar to that described by Scoccaet al. (21).

Detection of DNA hydrolysis. The production of DNase bycampylobacters was detected as described previously (15).

Chemicals. Restriction endonucleases and enzymes in-volved in nucleic acid metabolism were purchased fromBoehringer Mannheim Canada Ltd., Dorval, Quebec, Can-ada. [_a-32P]dATP was purchased from Du Pont, NENResearch Products, Boston, Mass.; MH medium was pur-chased from Oxoid Ltd., Basingstoke, England; and antibi-otics and chemicals were purchased from Sigma ChemicalCo., St. Louis, Mo.

RESULTS

Isolation from mutants. A spontaneous Strr mutant of theNalr C. coli UA417 (designated UA417R) was isolated byplating fresh UA417 cells onto MH agar containing strepto-mycin (10 ,ug/ml). The mutation frequency for Strf in thisstrain was less than 10-'. We were unable to isolate aspontaneous Strr mutant from C. jejuni UA466. Therefore,some UA466 Strr mutants were first isolated by transforma-tion with UA417R DNA as the donor. The UA466R (Nal'Str') strain was then isolated from one of the Nal' Strrtransformants by spontaneous mutation. Mutation frequen-cies for Nal' in most C. coli and C. jejuni strains were about5 x 10-9 (28).

Natural competence of C. coli and C. jejuni and transfor-mation of chromosomal DNA. All five C. coli strains andthree of six C. jejuni strains tested were naturally competent

J. BACTERIOL.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

31

Dec

embe

r 20

21 b

y 18

3.98

.44.

140.

TRANSFORMATION IN CAMPYLOBACTER SPP. 951

TABLE 2. Transformation of different strains of C. coli andC. jejuni with homologous chromosomal DNA'

No. of transformants/Recipient spot (ca. 5 x 106 cells) DNAhydrolysis'

Nalr Strr

C. coliUA417 NAC 3,500 +UA420 0 3,000 +UA585 6,000 4,000 +UA724 1,000 1,000 +BM2509 0 4,000 +

C. jejuniUA67 NA 0 + +UA466 300 1,000 +UA649 200 800 +UA580 150 600 -UA697 0 0 -C31 0 0 +

a Transformation was performed on MH agar (see Materials and Methods).The donor DNA (0.2 SLg in 5 tLI of TE buffer) was from C. coli UA417R fortransformation of C. coli strains or from C. jejuni UA466R for transformationof C. jejuni strains. The cells within the DNA spot were scraped up and spreadon selective media with or without dilution. All experiments were repeated atleast twice.

b Symbols: + +, strong positive; +, weak positive; ±, very weak reactionafter 2 to 3 days of incubation; -, negative.

c NA, Not applicable.

for DNA uptake (Table 2). Transformation frequencies oftwo chromosomal markers (Nalr and Str) were about 10-3per cell in C. coli and 10-4 per cell in C. jejuni withsaturating DNA. C. coli UA420 and BM2509 could not betransformed to Nalr by UA417R DNA. C. coli UA585 washighly competent in this preliminary study and was chosenfor most of the transformation experiments. C. coli UA585cells could be transformed to Nalr by C. jejuni UA466RDNA at about 20% efficiency compared with homologousDNA transformation, but these interspecies Nalr transfor-mants grew more slowly than either parent. C. jejuni UA466could be transformed to Strr by C. coli UA417R DNA at 1%efficiency; however, these transformants exhibited normalgrowth rates.

Transformation frequency increased as donor DNA con-centration increased, and the saturation level of transform-ing DNA was about 1 ,ug/ml when cells were incubated withDNA for 30 min (Fig. 1). The transformation efficiencyobtained in the sample at a concentration of 0.01 jig ofDNAper ml (4 x 105 transformants per ,ug of DNA) was greaterthan that obtained at a concentration of 1 ,ug ofDNA per ml(8 x 104 transformants per jig of DNA). The transformationfrequency also increased as the incubation time with DNAincreased, and no transformants were obtained when DNaseI and MgCl2 were added to 0 min (Fig. 2).

Transformation of both Nalr and Strr markers was alsoperformed. The transformation frequencies of UA585 wereapproximately 1.2 x 10-3 for the Nalr marker, 4 x 10-4 forthe Strr marker, and about 2 x 10-7 for the Nalr Strrcotransformants. The results indicated that these two siteswere unlinked in this transformation system.DNA hydrolysis in these Campylobacter strains was ex-

amined. There was no clear correlation between their ownDNase activity and competence, and production of a smallamount of extracellular DNase activity did not appear tointerfere with chromosomal DNA transformation when thedonor DNA was suspended in TE buffer (Table 2). Additionof 10 mM MgCl2 to donor DNA did not affect the chromo-

S

C

x

h.S

C

aE

C

Ca

-

a

zC

Cz

104

1C3

1 o

DNR ( jig / 2H10 cells / ml )

FIG. 1. Dependence of transformation frequency on the concen-tration of donor DNA. The recipient was C. coli UA585, and thedonor was C. coli UA417 DNA. Transformation was performed in abiphasic culture system. DNA was added at the indicated concen-trations for 30 min before DNase I and MgCl2 were added. Nalrtransformants were selected after 3 h of incubation.

somal transformation frequency in UA585, which had littleDNase activity, but reduced the transformation frequenciesin UA417 and UA466 to about 40% (data not shown), both ofwhich showed some DNase activity.

Influence of growth phase on competence. To study thedevelopment of competence, we performed the biphasictransformation procedure with different growth phases of C.coli UA585 cultures as recipients and UA417R DNA as thedonor. UA585 cells were competent constitutively through-out their growth cycle (Fig. 3). A maximum number oftransformants was obtained after 6 h of incubation, but thetransformation frequency (2 x 10-4 per cell) was lower thanthat of 2-h samples (5 x 10-4 per cell). This indicated that

4

fi2x1&~e

E

L.

*) 4

104

zC

Z 00 15 30

Time of RdditIon of DNase ( Min )FIG. 2. Dependence of transformation frequency on the incuba-

tion time with donor DNA. The recipient was UA585 (5 x 108/ml),and the donor was UA417 DNA (1 ,g/ml). DNase I and MgCl2 wereadded at the indicated times after DNA was added. Nalr transfor-mants were selected after 3 h of incubation.

VOL. 172, 1990

I

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

31

Dec

embe

r 20

21 b

y 18

3.98

.44.

140.

952 WANG AND TAYLOR

2x 10

-

c

C.c

E

CC'a)

105

-

c

a

1-

0

Cz

Time ( Hours )

FIG. 3. Effect of growth phase on the competence of C. coliUA585. Transformation was performed in the biphasic system.UA417R DNA (in TE buffer) was added to 1 ,ug/ml at the indicatedtime intervals, and the cell number was counted by plating onto MHagar. Strr transformants were selected after 4 h of incubation.Symbols: 0, CFU per milliliter; A, number of transformants permilliliter.

the early-log-phase bacteria were slightly more competentthan the late-log-phase cells.

Construction of shuttle vectors. To study plasmid transfor-mation and to improve techniques for molecular cloning inCampylobacter spp., we constructed three shuttle vectorsby using pILL55OA as the parent (Fig. 4). Plasmid pUOA13consisted of the entire pUC13 plasmid (cut at the uniqueAatII site) and an EcoRI-SalI fragment of pILL550A whichcontains a replication origin of C. coli plasmid pIP1445, aKmr gene, and an origin of transfer (oriT) of the RK2 plasmid(14). This plasmid encodes Apr and Kmr and the ability tocomplement a defective ,-galactosidase in E. coli, whereasonly Kmr is expressed in Campylobacter spp. PlasmidpUOA15 was derived from pUOA13 by replacing the Kmrgene with the tetO gene from pUOA3 (25). The tetO deter-minant is expressed in both E. coli and Campylobacter spp.Plasmid pUOA17 was obtained by deleting the ClaI-ScaIfragment from pUOA13, thus removing part of the E. coliApr gene.[32P]DNA uptake and competition studies. Figure 5 shows

the kinetics of irreversible uptake of homologous (C. coliUA417), heterologous (E. coli), or plasmid (pUOA13) DNAby C. coli UA585 cells. Uptake of C. coli chromosomal DNAcontinued to increase up to 28 min of incubation, whereas E.coli and plasmid pUOA13 DNAs were taken up in barelydetectable amounts. Similar results were obtained whenDNA labeled by end filling or the random primer labelingmethod was used as the donor (data now shown).When unlabeled C. coli, C. jejuni, and E. coli DNAs were

used to compete with [32P]DNA for uptake, we found that C.coli and C. jejuni DNAs competed for uptake with approx-imately equal efficiency, whereas E. coli DNA did notinterfere with the uptake of C. coli DNA (Fig. 6).

Transformation of plasmid DNA. The frequency of trans-

pUOA1 7

pUOA13

_l a} X S $ $t \ /_~~~~~~~~~~~~~~~~~~~~~U_1~~1 '' - . '

5 6I b

lar' bla

i I

tetO

pUOAI 5FIG. 4. Restriction maps of the shuttle vectors pUOA13, pUOA15, and pUOA17. Symbols: -, fragment containing Km gene or tetO

gene; , DNA sequence from the Campylobacter plasmid pIP1445; , oriT DNA; , pUC13 DNA. Numbers represent kilobasepairs.

oI/ 1K 3 4

.,

J. BACTERIOL.

1104

I

I

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

31

Dec

embe

r 20

21 b

y 18

3.98

.44.

140.

TRANSFORMATION IN CAMPYLOBACTER SPP. 953

15x10

E

0)-

z41 0

0

0

W~~~~~~

L g

1 0 20 30

Time ( Min)FIG. 5. Kinetics of DNA uptake by C. coli UA585 cells. C. coli

UA585 cells were incubated with [32P]DNA from C. coli UA417 (0),E. coli (0), or plasmid pUOA13 (A). The uptake assay is describedin Materials and Methods.

formation of plasmid pUOA17 (8.3 kilobase pairs [kb]) andpILL55OA (8.6 kb) into plasmid-free UA585 cells (Table 3)was about 1,000 times lower than that of the chromosomalmarkers (Table 2), and we were unable to obtain plasmidtransformants (shuttle vectors) by using C. coli UA417 or C.jejuni UA649 as the recipient. When strain UA585 contain-ing a homologous plasmid was used as the recipient, trans-formation frequencies of the shuttle plasmids were increased100-fold, although they were still 10 times lower than that ofchromosomal markers (Table 3). We were unable to trans-form the 45-kb plasmid pUA466 into plasmid-free C. coli orC. jejuni strains. However, Tcr transformants could beobtained by transformation of pUA466 into C. jejuni UA649which contains a 4.2-kb deletion of plasmid pUA466 and inwhich most of the tetO determinant has been deleted (25). Inthis strain the transformation frequency of Tcr was close tothat of chromosomal markers. Also, in this natural transfor-mation system the transforming activity of pUOA17 when itwas isolated from E. coli was not significantly lower than

6xl1

_ 4x1l0)

-WC

Z 2x11

4

1 0

0 0.4 0.8 1.

Competing DNR ( jug / 5x108cells )FIG. 6. Competition by unlabeled DNA in the uptake of radio-

active C. coli DNA. C. coli UA585 cells were incubated with32P-labeled C. coli UA417 DNA (0.2 ,g/ml) for 15 min. Unlabeled C.coli (-), C. jejuni (0), or E. coli (A) DNA was present at theindicated concentrations.

TABLE 3. Transformation of plamid DNA into C. coliand C. jejuni strainsa

No. of trans-Donorb Recipient formants/spot

(marker selected)

pILL550A (C. jejuni) C. coli UA417 0 (Km)pILL550A (C. jejuni) C. coli UA585 8' (Km)pUOA17 (E. coli) C. coli UA585 3c (Km)pUOA17 (C. coli) C. coli UA585 6C (Km)pUOA17 (E. coli) C. coli UA585(pUOA15) 500 (Km)pUOA17 (C. coli) C. coli UA585(pUOA15) 640 (Km)pUOA15 (C. coli) C. coli UA585(pUOA13) 1200 (Tc)pILL550A (C. jejuni) C. jejuni UA649(pUA649) 0 (Km)pUA466 (C. jejuni) C. jejuni UA649(pUA649) 200 (Tc)pUA466 (C. jejuni) C. jejuni UA650 0 (Tc)

a Transformation was performed on MH agar. Approximately 5 x 106 cellswithin the DNA spot were scraped up and spread on MH agar containingantibiotics. The number of transformants is the average of three experiments.

b Plasmid DNA was isolated from the bacterial species indicated in paren-theses.

' These transformants were confirmed by isolation of plasmid DNA andelectrophoresis through a 0.7% agarose gel.

that obtained when it was isolated from Campylobacter spp.(Table 3).

Cloning and expression of streptococcal tetM gene in C. coli.The tetM gene was originally cloned into the E. coli vectorpACYC177 at the Hincll site (pJI3) (3). The gene has neverbeen demonstrated in Campylobacter spp. (23). The 5-kbHinclI fragment containing the tetM determinant is veryunstable when cloned in a high-copy-number E. coli plasmidsuch as the pUC series of vectors (13; Y. Wang, and D. E.Taylor, unpublished observations). A stable tetM-pUC13clone was obtained by cleaving pJI3 with HinclI, partiallydigesting it with BAL 31, and inserting it into the SmaI siteof pUC13 (termed pUOA11; 6.9 kb). The plasmid pUOA11(0.3 ,ug) was linearized with HinclI and ligated with SmaI-cut pILL55OA (0.3 ,ug), and the mixture was used to trans-form UA585(pUOA13) cells on MH agar. Two Tcr colonieswere obtained from some of the DNA-treated cells. PlasmidDNA was isolated from these two transformants and verifiedby ClaI cutting, which has one site in the tetM gene (19) andanother site in the vector pILL550A (Fig. 4). The MICs oftetracycline for both clones were 256 ,ug/ml.

DISCUSSION

In this study we have demonstrated that most C. coli andsome C. jejuni strains are naturally competent for DNAuptake. Transformation with both Nal and Str markersindicates that C. coli is able to take up only a very limitednumber ofDNA molecules at any given time. The frequencyof Nalr Strr cotransformants (2 x 10-7) is close to theproduct of the frequencies of Nalr transformants and Strrtransformants (1.2 x 10-3 multiplied by 4 x 10-4 = 4.8 x10-7); therefore, each competent cell probably takes up anaverage of two DNA molecules.The results obtained from both [32P]DNA uptake and

competition experiments demonstrate that the DNA uptakesystem of C. coli is specific for DNA derived from this andclosely related species. Such a mechanism is very similar tothat seen in Haemophilus (21) and Neisseria (8) species.Although direct evidence for a recognition sequence is stilllacking, our results favor the view that a specific recognitionsequence is present in both C. coli and C. jejuni. Thepossibility that DNA recognition involves interaction with

L

VOL. 172, 1990

3xl1~ [

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

31

Dec

embe

r 20

21 b

y 18

3.98

.44.

140.

954 WANG AND TAYLOR J. BACTERIOL.

modified residues (21) is unlikely, because the transforma-tion efficiencies of plasmid DNA isolated from E. coli and C.coli were similar (Table 3). The possibility that homologousrecombination is required during the DNA uptake process(24) also appears unlikely, because C. coli cells take up C.jejuni DNA as efficiently as they take up C. coli DNA (datanot shown), yet the two DNAs share only a limited amountof homology (approximately 32 to 48% [1]). Furthermore,the frequency of transformation of the shuttle plasmid intoC. coli cells containing a homologous plasmid was still muchlower than that of homologous chromosomal DNA.

It has been demonstrated that a restriction-modificationsystem exists in C. jejuni C31 (14, 20), which acts on theEcoRI recognition site. DNA isolated from Campylobacterspp. was resistant to EcoRI digestion and became suscepti-ble if the same DNA was transfered and isolated from E.coli. Moreover, similar results were obtained in electropo-ration experiments in which pILL550 DNA isolated from E.coli transformed C. jejuni poorly (20). In contrast, duringnatural transformation of Campylobacter spp., the trans-forming activity of the plasmid DNA isolated from E. coliwas about equivalent to that of DNA isolated from Cam-pylobacter spp. This result suggests that the incoming DNAwas protected from the action of restriction endonucleases,and it is consistent with the protection of transforming DNAfrom cellular nucleases observed in other natural transfor-mation systems (22).The frequencies of plasmid transformation were about

1,000 times lower than those of chromosomal markers withthe use of the highly competent UA585 strain and smallplasmids (e.g., pUOA17 [8.3 kb]). We were unable totransform UA417 or UA466 with plasmid DNA, possiblybecause these two strains have some extracellular DNaseactivity (Table 2). Such low transformation frequencies maybe due to very inefficient uptake of the shuttle plasmid byCampylobacter spp. (Fig. 5), combined with partial digestionof the donor DNA during uptake, which has been noted in allother well-characterized natural transformation systems(24). The damage to transforming plasmid DNA can becompensated for by the use of recipients containing ahomologous plasmid; thus, the incoming plasmid can berescued by the resident plasmid through homologous recom-bination (24). Our results showed that transformation fre-quencies of the shuttle plasmids increased to about 100-foldin this system.

Natural transformation of Campylobacter spp. should beuseful in chromosome mapping, as well as in the develop-ment of improved methods for gene replacement mutagene-sis. Workers in our laboratory are constructing additionalvectors as well as developing genetic mapping technique forCampylobacter spp., which could be applied to the study ofthe biology and pathogenesis of these microorganisms.

ACKNOWLEDGMENTSWe thank A. Labigne-Roussel for providing plasmid pILL550A

and K. Roy for helpful suggestions.This work was supported by an operating grant from the Natural

Sciences and Engineering Research Council of Canada. D.E.T. wasa recipient of a Heritage Medical Scholarship.

LITERATURE CITED1. Beiland, R. J., and T. J. Trust. 1982. Deoxyribonucleic acid

sequence relatedness between thermophilic members of thegenus Campylobacter. J. Gen. Microbiol. 128:2515-2522.

2. Birnboim, H. D., and J. Doly. 1979. A rapid alkaline extractionprocedure for screening recombinant plasmid DNA. Nucleic

Acids Res. 7:1513-1523.3. Burdett, V., J. Inamine, and S. Rajagopalan. 1982. Heterogene-

ity of tetracycline resistance determinants in Streptococcus spp.J. Bacteriol. 149:995-1004.

4. Chang, S., and S. N. Cohen. 1979. High frequency transforma-tion of Bacillus subtilis protoplasts by plasmid DNA. Mol. Gen.Genet. 168:111-115.

5. Chung, C. T., S. L. Niemela, and R. H. Miller. 1989. One-steppreparation of competent Escherichia coli: transformation andstorage of bacterial cells in the same solution. Proc. Natl. Acad.Sci. USA 86:2172-2175.

6. Cohen, S. N., A. C. Y. Chang, and L. Hsu. 1972. Nonchromo-somal antibiotic resistance in bacteria: genetic transformation ofEscherichia coli by R-factor DNA. Proc. Natl. Acad. Sci. USA69:2110-2114.

7. Danner, D. B., R. A. Deich, K. L. Sisco, and H. 0. Smith. 1980.An eleven-base-pair sequence determines the specificity ofDNA uptake in Haemophilus transformation. Gene 11:311-318.

8. Dougherty, J. J., A. Asmus, and A. Tomasz. 1979. Specificity ofDNA uptake in genetic transformation of gonococci. Biochem.Biophys.. Res. Commun. 86:97-104.

9. Feinberg, A. P., and B. Vogelstein. 1983. A technique forradiolabeling DNA restriction endonuclease fragments to highspecific activity. Anal. Biochem. 132:6-13.

10. Goodgal, S. H. 1982. DNA uptake in Haemophilus transforma-tion. Annu. Rev. Genet. 16:169-192.

11. Goodman, S. D., and J. J. Scocca. 1988. Identification andarrangement of the DNA sequence recognized in specific trans-formation of Neisseria gonorrhoeae. Proc. Natl. Acad. Sci.USA 85:6982-6986.

12. Gryczan, T., S. Contente, and D. Dubnau. Molecular cloning ofheterologous chromosomal DNA by recombination between aplasmid vector and a homologous resident plasmid in Bacillussubtilis. Mol. Gen. Genet. 177:459-467.

13. Hill, C., G. Venema, C. Daly, and G. F. Fitzgerald. 1988.Cloning and characterization of the tetracycline resistance de-terminant of and several promoters from within the conjugativetransposon Tn919. Appl. Environ. Microbiol. 54:1230-1236.

14. Labigne-Roussel, A., J. Harel, and L. Tompkins. 1987. Genetransfer from Escherichia coli to Campylobacter species: devel-opment of shuttle vectors for genetic analysis of Campylobacterjejuni. J. Bacteriol. 169:5320-5323.

15. Lior, H., and A. Patel. Improved toluidine blue-DNA agar fordetection of DNA hydrolysis by Campylobacters. J. Clin.Microbiol. 25:2030-2031.

16. Macrina, F. L., P. H. Wood, and K. R. Jones. 1980. Genetictransformation of Streptococcus sanguis (Challis) with crypticplasmids from Streptococcus ferus. Infect. Immun. 28:692-699.

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

18. Marmur, J. 1961. A procedure for the isolation of deoxyribo-nucleic acid from micro-organisms. J. Mol. Biol. 3:208-218.

19. Martin, P., P. Trieu-Cuot, and P. Courvalin. 1986. Nucleotidesequence of the tetM tetracycline resistance determinant of thestreptococcal conjugative shuttle transposon Tn1545. NucleicAcids Res. 14:7047-7058.

20. Miller, J. F., W. J. Dower, and L. S. Tompkins. 1988. High-voltage electroporation of bacteria: genetic transformation ofCampylobacter jejuni with plasmid DNA. Proc. Natl. Acad.Sci. USA 85:856-860.

21. Scocca, J. J., R. L. Poland, and K. C. Zoon. 1974. Specificity indeoxyribonucleic acid uptake by transformable Haemophilusinfluenzae. J. Bacteriol. 118:369-373.

22. Smith, H. O., D. B. Danner, and R. A. Deich. 1981. Genetictransformation. Annu. Rev. Biochem. 50:41-68.

23. Sougakoff, W., B. Papadopoulou, P. Nordmann, and P. Courva-lin. 1987. Nucleotide sequence and distribution of gene tetOencoding tetracycline resistance in Campylobacter coli. FEMSMicrobiol. Lett. 44:153-159.

24. Stewart, G. J., and C. A. Carlson. 1986. The biology of naturaltransformation. Annu. Rev. Microbiol. 40:211-235.

25. Taylor, D. E. 1986. Plasmid-mediated tetracycline resistance in

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

31

Dec

embe

r 20

21 b

y 18

3.98

.44.

140.

VOL. 172, 1990 TRANSFORMATION IN CAMPYLOBACTER SPP. 955

Campylobacterjejuni: expression in Escherichia coli and iden-tification of homology with streptococcal class M determinant.J. Bacteriol. 165:1037-1039.

26. Taylor, D. E., N. Chang, R. S. Garner, R. Sherburne, and L.Mueller. 1986. Incidence of antibiotic resistance and character-ization of plasmids in Campylobacter jejuni: strains isolatedfrom clinical sources in Alberta, Canada. Can. J. Microbiol.32:28-32.

27. Taylor, D. E., K. Hiratsuka, H. Ray, and E. Manavathu. 1987.Characterization and expression of a cloned tetracycline resis-tance determinant from Campylobacterjejuni plasmid pUA466.

J. Bacteriol. 169:2984-2989.28. Taylor, D. E., L. Ng, and H. Lior. 1985. Susceptibility of

Campylobacter species to nalidixic acid, enoxacin, and otherDNA gyrase inhibitors. Antimicrob. Agents Chemother. 28:708-710.

29. Trieu-Cuot, P., G. Gerbaud, T. Lambert, and P. Courvalin.1985. In vivo transfer of genetic information between Gram-positive and Gram-negative bacteria. EMBO J. 4:3583-3587.

30. Vieira, J., and J. Messing. 1982. The pUC plasmids, anM13mp7-derived system for insertion mutagenesis and sequenc-ing with synthetic universal primers. Gene 19:259-268.

Dow

nloa

ded

from

http

s://j

ourn

als.

asm

.org

/jour

nal/j

b on

31

Dec

embe

r 20

21 b

y 18

3.98

.44.

140.