Characterization of a plasmid isolated from Branhamella catarrhalis and detection of plasmid sequences within the genome of a B. catarrhalis strain

PLASMID20,158-162 (1988)

Characterization of a Plasmid Isolated from Branhamella catarrhalis and Detection of Plasmid Sequences within the Genome of a B. catarrhalis Strain DANIELLE BEAULIEU,*,~ MARC OUELLETTE,*-~’ MICHEL G. BERGERON,-~# AND PAUL H. RoY*,~‘~ *DJpartement de Biochimie, Fact& des Scienceset g&tie, Universite’Lava/, Quebec, Canada GlK 7P4; ?Service d’lnfectiologie, Centre Hospitalier de I’llniversite’ Laval, Quebec, Canada Gl V 4G2; and $Dkpartement de Microbiologic, Fact&h de mgdecine, Universite’Lava/, Quebec, Canada GIK 7P4 Received April 25, 1988; revised July 11, 1988 We isolateda 12.2-kbplasmid from two clinical strainsof Branhamelia catarrhalis and evaluated its distribution among other B. catarrhalis strains by colony hybridization experiments using the whole plasmid as a probe. Homology was detected with the two plasmid-beating strains and also with a third B. catarrhalis strain named E7, which is plasmidless. Southern transfer analysis of total digestedE7 DNA using the purified plasmid as a probe revealeda singleband of hybridization, different from those observed in plasmid-bearing strains, for each restriction enzyme used. The region of the plasmid hybridizing with DNA of strain E7 was located within a 4.5-kb &I-EcoRV fragment. No homology was noted between our B. catarrhalis plasmid and about 20 other strains of different genera tested by colony hybridization. Q 1988 Academic PESS, hc.

Branhamella catarrhalis is an aerobic gramnegative diplococcus that was until recently considered a harmless nonpathogenic species of the upper respiratory tract of humans (Doern and Morse, 1980; Henriksen, 1976; Lennette et al., 1985). B. catarrhalis is now regarded as a significant pathogen causing lower respiratory tract infections, especially in winter (Rikitomi et al., 1986). Resistance to P-la&am drugs in B. catarrhalis is mediated by any one of severalnovel &lactamaseswhich differ from the TEM enzymes (Buu Hoi Dang Van et al., 1978; Eliasson and Kamme, 1985; Riley, 1988). Resistance was first noticed in 1977(Ninane et al., 1977;Percival et al., 1977) and was shown to be transferable by conjugation within the genus Branhamella and also from Moraxella nonliquefaciens to B. catarrhalis (Doem et al., 1980; Kamme et al., 1983, 1984). This led to the hypothesis that at least ’ Present address:Molecular Biology Department, The Netherlands Cancer Institute, Plesmanlaan 121,1066 CX Amsterdam, The Netherlands. * To whom correspondenceshould be addressed. 0147-619X/88 $3.00 Copyright 8 1988 by Academic Press, Inc. All rigbls of reproduction in any fom reserved.

one of these B-lactamaseswas plasmid mediated (Kamme et al., 1986; El&son and Kamme, 1985). Until now, no one has successfully isolated extrachromosomal DNA in B. catarrhalis; some attribute these failures to high endonuclease activity (Kamme et al., 1984, 1986; Buu Hoi Dang Van et al., 1978). We screened 17 clinical isolates of B. catarrhalis isolated during an investigation on acute otitis media in children (Bergeron et al., 1987). Of these 17 strains, 12 produced a plactamase (as determined by a nitrocefin hydrolysis test; O’Callaghan et al., 1972). The 17 strains were screened for the presence of plasmid DNA by minipreparations (Bimboim and Doly, 1979; Maniatis et al., 1982), and 2 P-lactamase-producingstrains, E22 and Po34, were found to harbor extrachromosomal elements of approximately 12 kb, which were named pLQ5 10 and pLQ5 11, respectively. The pLQ5 10 plasmid DNA was isolated by the Ish-Horowitz modification of the alkaline lysis method as described by Maniatis et al. ( 1982), to which we added the following steps: (1) after the isopropanol precipitation of DNA,

158

SHORT COMMUNICATIONS

the pellet was dissolved in TE (0.01 M Tris0.001 M EDTA, pH 8); (2) DNA was deproteinized twice with phenol equilibrated with Tris and once with chlorofornnisoamyl alcohol (24: l), and extracted twice with ether; and (3) the resulting solution was ethanol precip itated (2-3 vol) overnight at -20°C and the DNA was collected by centrifugation at 7000g for 30 min at 4°C. The DNA was purified by cesium chloride-ethidium bromide gradient ultracentrifugation (Maniatis et al., 1982). A physical map of plasmid pLQ5 10 (Fig. 1A) was then constructed by single and double restriction enzyme digestswhich were analyzed by agarosegel electrophoresis. The molecular massof the plasmid, as determined by restriction enzyme digests,was calculated as 12.2 kb. A preliminary attempt to transform this plasmid into Escherichia coli HB 101 by the CaC12 method (Maniatis et al., 1982), using ampicillin as a selective marker, was unsuccessful either because ampicillin resistance may be chromosomally determined or because the plasmid may not replicate in E. coli. We also found a plasmid of the same molecular mass in a type I B. cutarrhalis strain (N3034, kindly provided by ProfessorJ. Y. Riou) which produces the most commonly found B. caturrh-

159

alis @-lactamase,BRO-1 (Philippon et al., 1986; Riou and Guibourdenche, 1986). Since preparation of plasmid DNA from B. cutarrhalis is difficult, we decided to evaluate pLQ5 10 distribution by nucleic acid hybridization. Hybridization conditions (50% formamide, 42°C) and posthybridization washing conditions (0.1% SSC,0.1% SDS, 68°C) were asdescribedby Quellette et al. ( 1987)and were very stringent so that only strong homology would appear. We radiolabeled purified pLQ5 10 and used it as a molecular probe to screen the 17 B. catarrhalis strains. Three of these strains gave a positive signal: E22 and Po34, known plasmid carriers, and a third strain, E7, with which all attempts to isolate plasmid DNA remained negative. To evaluate whether pLQ5 10 plasmid DNA would show homology to microorganisms other than Brunhamella, we tested the probe against 10 ATCC Neisseriu species(N. flavescens(ATCC 13120), N. elonguta (ATCC 25295), N. elongata (ATCC 29315), N. cinereu (ATCC 14685), N. luctumica (ATCC 23970), N. subflaw (ATCC 14799), N. mucosa (ATCC 19696), N. mucosa (ATCC 25999), N. siccu (ATCC 9913), and N. siccu (ATCC 29259)), 2 clinical Neisseria species (N. gonorrhoeae

FIG. 1. Restriction map of pLQS 10 from E. cararrhnlis strain E22. The molecular massof the plasmid is 12.2 kb. (A) Unique cleavage sites on pLQ510 are BgfII, ClaI, EcoRI, and NruI; there are two sites for EcoRV, HueII, HincII, F’vuI, and SstI, and there are three for BcZl. There were no sites for the following enzymes:AvaI, BumHI, KpnI,NaeI, PstI, SmuI, SphI,SstII, and XhoI. (B) Physical map showing restriction fragments used as probes. The region of homology between pLQ5 10 and E7 chromosomal DNA is shown as a thick line. Probe A: 6.0-kb SrfI fragment; Probe B: 6.2-kb SstI fragment; Probe C: 5.2-kb EcoRV fragment; Probe D: 7.0-kb EcoRV fragment.

160

SHORT COMMUNICATIONS

(661452) and N. meningitidis (V261 l)), 3 animal sourceNeisseriu species(N. cuviue (N4 18 ATCC 14659), N. cuniculi (N420 ATCC 14688), and N. ovis (N453 ATCC 19575), kindly provided by Professor Riou), 3 Moraxella species(M nonliquefaciens(N534 I), M lucunata (N5297), provided by professorRiou, and M urethralis (BD468)), 2 E. coli (HB 101 and ATCC 34248), 2 Haemophilus injluenzae (H395 and ATCC 33533), and 3 other nonclinical B. catarrhalis strains (types I and II, provided by Professor Riou, and ATCC 25240). The results of the colony hybridizations show that the plasmid probe gave a positive signal with plasmid-carrying B. cutarrhalis strains (type I, E22, and Po34) and with E7 exclusively. None of the 25 other strains tested gave a positive signal. To better define the extent of the homology observedwith pLQ5 10 DNA, the plasmid was digested with SstI (Fig. 1B); the two resulting fragments were purified on low-melting-point agarose as described by Ouellette and Roy (1986) and used as separateprobes in colony hybridization experiments. The 6.0-kb SstI fragment probe gave a slight hybridization to B. catarrhalis E7, while the 6.2-kb SstI fragment retained the homology displayed by the whole plasmid. A similar experiment was performed using the two plasmid fragments generated by an EcoRV digestion (Fig. 1B). The results showed that the 7.0-kb EcoRV fragment hybridized to strain E7 while the 5.2-kb EcoRV fragment did not. The 6.2-kb SstI fragment and the 7.0-kb EcoRV fragment share a contiguous region of 4.5 kb and we conclude that the 4.5-kb SstI-EcoRV fragment, or parts thereof, contains the homology between E7 chromosomal DNA and pLQ5 10 DNA. We also noted a weak hybridization signal with the 7.0-kb EcoRV probe for B. catarrhalis strain Po51 (plasmidless and p-lactamaseproducer) and for E. coli (HB 101). The significance of this result is unknown, but it is probably an artifact since only one spot out of four (colonies were in duplicate and two identical filters were processed)gavethis signal for both strains mentioned. The slight hybridization to strain E7 noted for the 6.0-kb SstI

probe is much weaker than that observedwith the 6.2-kb SstI probe, meaning that the region of homology might extend beyond the St1 site, to a limited extent, on the 4.5-kb SstI-EcoRV fragment. It could also indicate that the probes were slightly contaminated by each other, resulting in this cross-hybridization. Thus we have defined the 4.5-kb SstI-EcoRV fragment of pLQ5 10 as the region encompassing the most significant homology. To determine the source of the hybridization signal observed in E7, we performed Southern transfer experiments using the method of Maniatis et al. (1982). Total DNA, isolated by the method of Chesney et al. (1979), from B. caturrhalis strain E7 digested with EcoRI and ClaI, purified pLQ5 10 DNA, and undigested DNA from B. catarrhalis strains E7 (plasmidless,,&lactamaseproducer), E8 (plasmidless, ,&lactamase producer), and E22 (plasmid pLQ5 10, &lactamase producer) and from N. elongata (ATCC 25295) were used. No signal was observed for the negative control strain N. elonguta (Fig. 2, lane 2) or for undigested DNA from B. catarrhalis strain E8 (Fig. 2, lane 4). For strain E22, labeled pLQ5 10 hybridized to a weak band of 12.2 kb (linearized pLQ5 10) and to a stronger band which corresponds to the open circular form of pLQ510 (Fig. 2, lane 3). Positive controls, pLQ5 10 purified plasmid DNA, linearized with EcoRI and undigested, hybridized strongly with the probe (Fig. 2, lanes 8 and 9, respectively). There is a signal from undigested E7 DNA showing the existence of some homology between pLQ5 10 and the chromosomal DNA of this strain (Fig. 2, lane 7). The &I-digested E7 DNA (Fig. 2, lane 5) showed one band of hybridization at 9 kb, while the EcoRI-digested E7 DNA (Fig. 2, lane 6) showed one band at approximately 15 kb. Neither of these bands corresponded to those from digested pLQ510 DNA (Fig. 2, lane 8) or to the open circular form or the covalently closed circular form of pLQ5 10 (compare Fig. 2, lanes 6 vs 3 and 9). A second Southern blot was done using restriction enzyme BclI (which has three cleavage sites in pLQ5 10) and the samecontrols as above; a hybridizing band of

SHORT COMMUNICATIONS

A 1

161

B 2

3

4

5

6

7

8

9

i0

123456789

FIG. 2. Southern transfer hybridization using plasmid pLQSl0 as a probe. (A) 0.7% agarose gel, (B) autoradiogram. The lanes contain the following DNAs: 1, I-kb ladder, 2, N. eiongutu (ATCC 25295) uncut; 3, E22 uncut; 4, E8 uncut; 5, E7IClaI; 6, E7fEcoRI; I, E7 uncut; 8, pLQSlO/EcoRI; 9, pLQ510 uncut; 10, supercoiled ladder. CCC, covalently closed circular form; OC, open circular form; L, linear fonn of pLQ5 10 DNA.

chromosomal DNA from strain E7 of approximately 7 kb was observed (results not shown). E22 total DNA digestedby either enzyme cited above and hybridized with a pLQ5 10 DNA probe showed a single band of hybridization at 12.2 kb (results not shown). This band corresponds to linearized plasmid DNA present in strain E22, and the absence of other hybridization indicates that this plasmid-carrying strain does not seem to possess the chromosomal homology to pLQ5 10. These results confirm that there is probably only one copy of a sequence homologous to the plasmid in B. cutu&zlis E7 chromosomal DNA. Moreover, the 4.5-kb SstI-EcoRV fragment described above contains a BclI site (seeFig. 1A). This, together with the fact that the BclI E7 digest yielded only one band of hybridization in the Southern transfer experiment, implied that the region of homology could be restricted to a smaller portion of pLQ510. Furthermore, other colony hybridization experiments (not shown) indicate that the 1-kb HinclI-EcoRV fragment included in the 4.5-kb %I-EcoRV fragment (see Fig. I)

does not hybridize with E7, so that the region of homology is now approximately within a 3.5-kb HincII-SstI fragment. The length of this region of homology permits us to hypothesize about its nature. We believe it could be an insertion sequence or parts of a transposable element which may have integrated into the genome of strain E7. Chromosomal integration of plasmids has been known to occur in other genera. For example, Hagblom et al. (1986) report the integration of the N. gonorrhoeue cryptic plasmid, in whole or in part, into the chromosomal DNA of plasmid-bearing and plasmid-free strains. Also, in H. injluenzae plasmid-free resistant strains, the 30-MD plasmid that usuahy codesfor /3-lactamaseproduction is integrated into the genome (Murphey-Corb et al., 1984). It is suggested,although not proved, that the stable integration of this 30-MD plasmid is mediated by insertion sequences, copies of which can be found on plasmids or on chromosomal DNA, either alone or associatedwith composite transposons (Murphey-Corb et al., 1984). We know that the homology of part of

162

SHORT COMMUNICATIONS

pLQ5 10 with the chromosomal DNA of strain HENRIKSEN,S. D. (1976). Moraxella, Neisseria, Branhamella and Acinetobacter. Annu. Rev. Microbial. 30, E7 does not come from the resistance gene 68-83. since 12 of the 17 B. catarrhalis isolates proKAMME, C., VANG, M., AND STAHL, S. (1983). Transfer duce a p-lactamase and that this homology is of &lactamase production in Branhamella catarrhalis. seen only in one particular strain. Further Stand. J. Infect. Dis. 15, 225-226. studies to investigate the chromosomal and KAMME, C., VANG, M., AND STAHL,S. (1984). Intragenic and intergenic transfer of Branhamella catarrhalis fiplasmid sequenceswill tell us more about the lactamase production. Stand. J. Infect. Dis. 16, 153nature of this homology. 155. ACKNOWLEDGMENTS We thank ProfessorJ. Y. Riou of PasteurInstitute, Paris, and the CHUL’s microbiology department for providing USwith bacterial strains; L. Bissonnette for advice with plasmid isolation; D. Gag& for typing this manuscript; R. Levesque for critical reading of the manuscript; and “Le Service d’Infectiologie” of the CHUL for encouragement and helpful comments. This work was supported by Grants A6774 and G 1541from the Natural Sciencesand Engineering Council (NSERC) and Grant EQ-1700 from the “Fonds de Formation des Chercheurset Aide B la Recherche” (FCAR) to P.H.R., D.B. and M.O. are NSERC predoctoral fellows.

REFERENCES BERGERON,M. G., AHRONHEIM,G., RICHARD, J. E., RIDING, K., CHON, C., BOYER,D., MACDONALD,N., BOUCHARD,M., YOUNG, J., AND DEMPSEY,E. E. ( 1987). Comparative efficacies of erythromycin-sulfisoxasole and cefaclor in acute otitis media: A double blind randomized trial. Pediatr. Infect. Dis. J. 6, 654660.

BIRNBOIM,H. C., AND DOLY, J. (1979). A rapid alkaline extraction procedure for screeningrecombinant plasmid DNA. Nucleic Acids Res. 7, 1513-l 523. Buu HoY DANG VAN, A., BRIVE-LE BOUGUENEC,C., BARTHELEMY,M., AND LABIA, R. (1978). Novel @tctamase from Branhamella catarrhalis. Ann. Microbial. Institut Pasteur B 129,397-406.

CHESNEY,R. H., SCOTI’,J. R., AND VAPNEK,D. (1979). Integration of the plasmid prophages Pl and P7 into the chromosome of Escherichia coli. J. Mol. Biol. 130, 161-173. D~ERN, G. V., AND MORSE,S. A. (1980). Branhamella (Neisseria) catarrhalis: Criteria for Laboratory identification. J. C/in. Microbial. 11, 193-195. D~ERN, G. V., SIEBERS,K. G., HALLICK, L. M., AND MORSE,S. A. (1980). Antibiotic susceptibility of &lactamase producing strains of Branhamefla (Neisseria) catarrhalis. Antimicrob. Agents Chemother. 17,24-29.

ELIASSON,I., AND KAMME, C. (1985). Characterization of the plasmid-mediated b-lactamase in Branhamella catarrhalis with special reference to substrate affinity. J. Antimicrob. Chemother. 15, 139-149. HAGBLOM,P., KORCH, C., JONSSON,A-B., AND NORMARK, S. (1986). Intragenic variation by site specific recombination in the CIyptic plasmid of Neisseria gonorrhoeae. J. Bacterial. 167, 23 l-237.

KAMME, C., ELIASSON,I., KNUTSON,B. K., AND VANG, M. (1986). Plasmid-mediated Blactamase in Branhame/la catarrhalis. Drugs 31 (Suppl 3), 55-6 1. LENNETTE,E. H., BALOWS,A., HAUSLER,W. J., AND SHADOMY,H. J. (1985). “Manual of Clinical Microbiology,” 4th ed., pp. 183-190. American Society for Microbiology, Washington, DC. MANIATIS,T., FRITSCH,E. F., AND SAMBROOK,J. (1982). “Molecular Cloning: A Laboratory Manual.” Cold Spring Harbor Laboratories, Cold Spring Harbor, NY. MURPHEY-CORB, M., NOLAN-WILLARD,M., AND DAUM, R. S. (1984). Integration of plasmid DNA coding for j!?lactamase production in the Haemophilus influenzae chromosome. J. Bacterial. 160, 8 15-8 17. NINANE, G., JOLY, J., PIOT, P., AND KRAYTMAN, M. (1977). Branhamella (Neisseria) catarrhalis as a pathogen. Lancet 2, 149. ~‘CALLAGHAN, C. H., MORRIS,A., KIRBY, S. M., AND SHINGLER,A. H. (1972). Novel method for detection of fl-lactamasesby using a chromogenic cephalosporin substrate.Antimicrob. Agents Chemother. 1,283-288. OUELLETTE,M., ROSSI,J. J., BAZIN, R., AND ROY, P. H. (1987). Oligonucleotide probes for the detection of TEM-I and TEM-2 Blactamase genes and their transposons.Canad. J. Microbial. 33, 205-211. OUELLETTE,M., ANDROY,P. H. (1986).Analysis by using DNA probes of the OXA- Blactamase gene and its transposon.Antimicrob. Agents Chemother. 30,46-5 1. PERCIVAL,A., CORKILL,J. E., ROLANDS,J., AND SYKES, R. B. (1977). Pathogenicity of /3-lactamaseproduction by Branhamella (Neisseria) catarrhalis. Lancet 2,1175. PHILIPPON,A., RIOU, J. Y., GUIBOURDENCHE, M., AND SOTOLONGO, F. (1986). Detection, distribution and inhibition of Branhamella cararrhalis &lactamases.Drugs 31 (Suppl 3), 64-69. RIKITOMI, N., MBAKI, N., MATSUMOTO,K., AND NAGATAKE,T. (1986). Seasonalchangesof in vitro adherence of Branhamella catarrhalis (B.C.) to human pharyngeal cells:Closecorrelation to the seasonaloccurrence of B.C. pulmonary infections. In “Abstracts of the 26th Interscience Conference on Antimicrobial Agents and Chemotherapy.” p. 259. RILEY, T. V. (1988). Antimicrobial resistance in Branhamella catarrhalis. J. Antimicrob. Chemother. 21, 147-

150. RIOU, J. Y., AND GUIBOURDENCHE,M. (1986). New methods of bacterial Diagnosis. Drugs 31 (Suppl3), I 6. Communicated by Gary Dunny