Cloning and Physical Characterization of Chromosomal Conjugative ...

JOURNAL OF BACTERIOLOGY, June 1986, p. 972-977

Vol. 166, No. 3

0021-9193/86/060972-06$02.00/0 Copyright © 1986, American Society for Microbiology

Cloning and Physical Characterization of Chromosomal Conjugative Elements in Streptococci MOSES N. VIJAYAKUMAR, SCOTT D. PRIEBE,t GIANNI POZZI,t JUDITH M. HAGEMAN,§ AND WALTER R. GUILD* Department of Biochemistry, Duke University, Durham, North Carolina 27710 Received 2 January 1986/Accepted 22 March 1986

We used a directed insertion method to introduce a nonreplicating vector plasmid into the large conjugative cat-tet element foubd in the chromosome of Streptococcus pneumoniae BM6001 and derivatives. To direct insertion preferentially to the conjugative element, we transferred it by conjugation to Streptococcus faecalis and then used DNA from this strain as a source of restriction nuclease fragments for ligation to digests of the vector pVA891, which can replicate in Escherichia coli but not in streptococci. This ligation mix was used to transform pneumococcal cells carrying the cat-tet element, with selection for the erythromycin resistance carried by pVA891. Eight such isolates were found, and transformation and conjugation tests showed that in each case the vector had inserted into the conjugative element, as expected. DNA from these pneumococcal strains generated a variety of E. coli plasmids which provide tools for obtaining a detailed restriction map and for defining other structural features of the streptococcal conjugative element.

The recent appearance and horizontal spread of multiple antibiotic resistance among clinical strains of pneumococci (Streptococcus pneumoniae) and other gram-positive- organisms is of considerable medical importance and biological interest (2, 6). Among streptococci, most drug resistance determinants are parts of large nonhomologous insertions into the chromosome (8, 9, 18). Several of these elements exhibit a new form of conjugal transfer within and among species carrying no detectable plasmids (3, 8, 18, 18a). Two of these conjugative elements, Tn916 (4) and Ql(cat tet erm) of Streptococcus agalactiae B109 (19), are known to transpose to plasmids, and the term "conjugative transposon" has been applied to them. From sedimentation and transformation studies, it was found that fQ(cat-tet) of S. pneumoniae BM6001, fl(cat-erm-tet) of S. agalactiae B109, fl(cattet-erm-aphA) of S. pneumoniae BM4200, and fl(cat-tet) of S. pneurhoniae N77 were located in the chromosome (6). Using tet-3, a point mutation conferring drug sensitivity, it was shown that the tet determinants derived from various fQ elemenits were homologous to each other, whereas plasmidderived tet genes were not (20). The results thus far indicate that there is substantial homnology among the nl elements, which suggests that the various drug resistance determinants insert within or delete from a basic QI tet unit (6). If so, a detailed analysis of one such element should provide a framework to gain understanding of other related elements. Hence, we chose to study the well-characterized (6, 18, 20) fl(cat-tet) strain BM6001 element in detail. The initial scope of the project involved cloning fragments of the fl element in Escherichia coli and generating a restriction map. One could then focus on the location of the drug resistahce and transfer genes, the target sites for the insertion of the fl element in the pneumococcal chromo-

some, and other features which may be pertinent to the

mode of its transfer via transformation or conjugation. However, molecular cloning of DNA specifically from the conjugative element was difficult due to the scarcity of selectable markers that would express well in E. coli. One approach (9) involved cloning the cat gene from the Qi(caterm-tet) B109 element in E. coli as a starting point. This plasmid was used to probe a cosmid library to screen for overlapping DNA fragments. An alternative approach to be described here offers several advantages and enabled us to clone segments of DNA directly from different regions of the £t(cat-tet) BM6001 element at the same time. Also, this method enables us to disrupt, delete, or add any fragment of DNA within the element at will and to select for such mutants with ease. For selective recovery of fi DNA for cloning, we employed the method of directed insertion of a nonreplicating plasmid vector via genetic transformation in pheumococci (12, 13, 21). With this strategy, such insertion mutants would be obtained by selecting for an E. coli vector marker. Chromosomal DNA from the transformants would then be digested with a variety of restriction endonucleases, ligated, and used to transform E. coli to generate plasmids containing DNA flanking the vector in the pneumococcal insertion mutants. These plasmids could then be used as a source of DNA fragments for subsequent cycles of directed-insertion mutagenesis and cloning. This "chromosome walk" could be continued until the entire element had been cloned. These plasmids could be physically analyzed to provide a restriction map of the element. In this paper, we describe the method of generating pneumococcal mutants with the vector inserted specifically in the fQ DNA, the biological properties of these strains, and the generation of recombinant plasmids in E. coli by using the DNA from these strains.

* Corresponding author. t Present address: Biology Department, Brookhaven National

MATERIALS AND METHODS Bacterial strains, growth, transformation, and conjugation. Rxl is our standard wild-type S. pneumoniae laboratory strain (17). DP1322 is an Rxl derivative which received the

Laboratory, Upton, NY 11973. t Present address: Istituto di Microbiologia, Universita di Siena, 53100 Siena, Italy. § Present address: Department of Molecular, Cell and Developmental Biology, University of Colorado, Boulder, CO 80309. 972

VOL. 166, 1986

STREPTOCOCCAL CONJUGATIVE ELEMENTS 0.00 0.1 0

0.28 0.5 9-

BOm

HI

JSphI So/I -Nrul

I.034- AvOI

2.34-fEcoRI lam 0

3.72 3.82

Xbo 1 Cla I

973

rpm for 6 h. The top 20 ml was removed with care, using a bent-tip Pasteur pipette, before the bottom 10 ml of the gradient, which contained the chromosomal DNA and the membrane debris was collected. This material was extracted once with an equal volume of phenol-chloroform-isoamyl alcohol mixture (25:24:1) and twice with chloroform-isoamyl alcohol (24:1). The aqueous phase was dialyzed twice against 2 liters of TE (10 mM Tris hydrochloride, 1 mM disodium EDTA, pH 8) at 4°C. The dialysate was used for restriction nuclease digestions without further purification. Chemicals. Restriction nucleases and T4 DNA ligase were purchased from Bethesda Research Laboratories, and International Biotechnologies, Inc. Proteinase K was from Boehringer Mannheim GmbH. Enzyme reactions were carried out according to the supplier's recommendations.

3 5.54 HEind m 5.88 BomHI FIG. 1. Restriction enzyme map of pVA891. The sizes of single and multiple restriction fragments of pVA891 were measured after electrophoresis in 0.8 and 1.5% horizontal agarose gels by standard methods (22).

fQ(cat-tet) BM6001 element via transformation from strain DP1302 (17). DP1333 is DP1322 that had spontaneously lost chloramphenicol resistance (Cmr) and acquired the tet-3 point mutation conferring drug sensitivity (20). JF16 is an Rxl strain carrying the str-J chromosomal point mutation leading to resistance to streptomycin (Strr). DP1617 is an S. pneumoniae strain with multiple chromosomal markers, of which streptomycin resistance was used as reference for transformation procedures. DP1315 is an Rxl derivative containing the fl(cat-tet) BM6001 element and is our most proficient donor of this element in conjugation. Streptococcus faecalis JH2-2 was used as a conjugation recipient with DP1315 to construct JG14. E. coli DH1 cells (7) were transformed according to the method of Hanahan (7). Growth of pneumococcal strains, genetic transformation, conjugation, and scoring for transformants and transconjugants were done as described previously (5, 18a, 18b). S. faecalis cells were grown without aeration at 37°C in CAT medium, a rich broth containing casein hydrolysate and tryptone (5). DNA preparation. S. faecalis cells were lysed as described (3), and the DNA was purified by the method of Marmur (11). E. coli transformants were screened for plasmids by using rapid alkaline lysates (1) for agarose gel electrophoresis. Plasmid DNA from E. coli was purified according to standard methods involving lysozyme treatment and ultracentrifugation in CsCl-ethidium bromide density gradients. For restriction nuclease digestions and ligations, pneumococcal chromosomal DNA was isolated as follows. The cells were grown to 2 x 108 to 3 x 108 CFU/ml in 250 ml of CAT broth with 1 ,ug of erythromycin per ml. The cells were centrifuged, washed in 0.15 M NaCI-0.015 M trisodium citrate, resuspended in 2 ml of lysis buffer containing 0.1% sodium deoxycholate, 0.01% sodium dodecyl sulfate, and 30 mM disodium EDTA, and incubated at 37°C for 7 min or until the turbid culture cleared. The lysate was transferred to 60°C, and 0.5 ml of Proteinase K (5 mglml) was added. After 1 h, the lysate was layered onto a 25-ml 5 to 20% sucrose gradient containing 10 mM Tris hydrochloride (pH 8.0), 1.5 mM disodium EDTA, and 0.9 M NaCl on top of a 4-ml cushion of 70% sucrose and 3 mM disodium EDTA. The lysate was centrifuged in an SW25.1 rotor at 17°C at 22,000

RESULTS The insertion vector used in these studies, pVA891, was derived from pACYC184 by replacing the 0.2-kilobase (kb) HindIII-ClaI segment in the promoter region of the tet gene with a HindlIl-ClaI fragment containing a streptococcal erm determinant (10). Both erythromycin resistance (Emr) and Cmr markers are expressed in E. coli, whereas in pneumococci the plasmid is incapable of autonomous replication and confers resistance to erythromycin only when integrated into the chromosome. Because it was essential for this study to be able to identify vector and passenger DNA segments with precision and the restriction map for pVA891 had been indicated only sketchily (10), we redetermined this map (Fig. 1) using molecular size standards whose DNA sequences (15, 16) are currently available. We found the overall size to be 5.9 kb, 0.5 kb larger than previously reported. An independent confirmation of the scale factor was obtained by noting that the BamHI-to-AvaI region of pACYC184 is the same as that in pBR322, which matches our measurements closely. Directed insertion of the vector plasmid. A chimeric circular DNA constructed in vitro by ligating a fragment of "directing" chromosomal DNA to the foreign DNA can integrate during transformation at a region homologous to the passenger fragment. The resulting product contains a duplication of the directing DNA bracketing the heterologous DNA. This method has been useful in directing the insertion of heterologous DNA to specific regions of the genome (12, 13). TABLE 1. Phenotypic properties of pVA891 insertion strains of DP1322a Strain

Tra

Emr

Tcr

Cmr

GP41 GP42

+

+

+

GP43 GP44

+ + + +

+ +

+ + +

GP45 GP46 GP47 GP48

+ + +

_

+ + + + + + +

_

_

+ + +

+ + +

Loss of the inserted vector (%)b

0 0 13 7 0 5 7 16

a Emr, Tcr, and Cmr denote resistances to erythromycin, tetracycline, and chloramphenicol, respectively. Tra indicates transfer proficiency by conjuga-

tion. b Cells were grown for 50 generations at 37°C in the absence of selection and were plated for viable cells. The next day, 100 colonies were stabbed on blood agar plates containing 5 pLg of erythromycin per ml.

974

J. BACTERIOL.

VIJAYAKUMAR ET AL.

To direct the insertion of the vector plasmid pVA891 specifically within the conjugative element, we employed the following strategy. The Ql element was first introduced by filter mating into S. faecalis from pneumococcal strain DP1315 to create the transconjugant strain JG14. The chromosomal DNA of JG14 was digested with Sau3A and ligated to BamHI-cleaved pVA891. The ligation mixture was used to transform DP1322 competent cells. We assumed that the chromosome of S. faecalis did not have much homology to the pneumococcal genome and thus expected that the chromosomal insertion of the vector molecules would take place only when directed by the ligated passenger fragments derived from the fl element in this homology-dependent recombination reaction (13, 21). Such an event was expected to be rare due to the nature of the reaction (14) and the competing mass of nonhomologous DNA. Indeed the observed frequency of Emr transformants was about 5 x 10-6 of that of a chromosomal point marker for the amount of donor DNA used. Properties of Emr pneumococcal clones. Eight Emr transformants were isolated, grown, and tested for drug resistance and transfer properties (Table 1). Six contained resistances to all three drugs, chloramphenicol, tetracycline, and erythromycin. Except for GP45, all strains were able to transfer the fl element to wild-type cells by conjugation on filters (Table 1). Moreover, these transfer-proficient strains mobilized the Emr determinant to the recipients at the same frequency as they did Cmr or Tcr. However, the transfer did not include the donor chromosomal marker, str-J (data not shown). To further confirm the insertion of the vector within the fQ element, the transformation properties of the insertion mutants were studied. If the plasmid had inserted into the element in the mutants, then transformation of the plasmid marker, Emr, of a recipient strain bearing the element should be more efficient than transformation of the wild-type strain, Rxl. In this case, transformation of Rxl for Emr would require the insertion of all or a part of the large fl element, not just the plasmid, and this event, requiring the entry and integration of a longer stretch of DNA, has been shown to be relatively inefficient (18). On the other hand, if the plasmid had inserted into the host chromosomal DNA, then both TABLE 2. Transformation donor properties of pVA891 insertion derivatives of DP1322 Emr tranformants Donor"

per ml

DP1322

GP41 GP42 GP43 GP44 GP45 GP46 GP47 GP48 DP1617C

95,000 74,000 115,000 36,200 40,000 64,000 94,800 19,800 112,000

Rxl

600 93 147 208

1,400 62 600 5,200

Resistances of the Rxl Emr transformantsb Cmr Tcr Emr only Cmr Tcr

24 0 25 26 0 24 24 1

0 22 0 0 0 0 0 15

6 0 5 4 0 6 6 0

0 8 0 0 30 0 0 14

186,000

Donor cells were grown, processed, and lysed essentially as described (18), followed by addition of 0.1 ml-samples of lysate to 1 ml of competent DP1322 or Rxl cells, incubation at 37°C for expression, and scoring for Emr transformants. b Thirty Emr transformant colonies of Rxl were tested from each donor by toothpick transfer to blood agar plates containing chloramphenicol, tetracycline, or both. c Purified DP1617 DNA was added to control cells at 1 ,ug/ml. Transformants were scored for Strr conferred by the str-l point marker. a

TABLE 3. Transformation of pneumococcal recipients by E. coli plasmids derived from pneumococci with pVA891 insertionsa Parent strain

Plasmid

Size (kb)

GP42 GP44 GP45 GP48

pDP30 p44 pVJ31 pE7

22.5 8.0 16.0 17.0

Emr transformants per ml DP1322 Rxl

54,080 27,680