The csp operon of Streptococcus salivarius encodes two ... - CiteSeerX

Download PDF
3 downloads 0 Views 387KB Size Report
Comment
A Tn917 mutant library was generated to identify genes involved in the biogenesis of Streptococcus salivarius fimbriae. A fimbria-deficient mutant was isolated ...
Microbiology (2004), 150, 189–198

DOI 10.1099/mic.0.26592-0

The csp operon of Streptococcus salivarius encodes two predicted cell-surface proteins, one of which, CspB, is associated with the fimbriae Ce´line Le´vesque,3 Christian Vadeboncoeur and Michel Frenette Correspondence Michel Frenette [email protected]

Received 24 June 2003 Revised 18 September 2003 Accepted 22 September 2003

Groupe de Recherche en E´cologie Buccale (GREB), Faculte´ de Me´decine Dentaire and De´partement de Biochimie et de Microbiologie, Universite´ Laval, Que´bec, Canada G1K 7P4

A Tn917 mutant library was generated to identify genes involved in the biogenesis of Streptococcus salivarius fimbriae. A fimbria-deficient mutant was isolated by negative selection using an immunomagnetic separation technique with specific anti-fimbriae polyclonal antibodies (pAbs). The transposon was inserted in an ORF, called orf176, which encoded a protein of unknown function. The transposon prevented the transcription of orf176 as well as two genes located downstream, which are designated cspA and cspB and which form the csp operon. Sequence analyses of CspA and CspB revealed that both proteins possessed the classic cell-wall-anchoring motif (LPXTG) of Gram-positive bacterial surface proteins. Recombinant CspA (rCspA) and CspB (rCspB) proteins were generated in Escherichia coli and used to produce pAbs. Immunolocalization experiments showed that anti-rCspB, but not anti-rCspA antibodies specifically recognized S. salivarius fimbriae. Our results suggested that the csp operon encoded predicted cell-surface proteins, one of which, CspB, was associated with the fimbriae.

INTRODUCTION Bacteria produce a vast array of surface structures, including fimbriae. Fimbriae are hair-like appendages that are composed of protein subunits with diameters ranging from 2 to 8 nm and that usually extend 1 to 2 mm from the cell surface (Low et al., 1996). They enable bacteria to establish microbial communities by recognizing specific receptors in their natural environments (Ferna´ndez & Berenguer, 2000; Klemm & Schembri, 2000). In some cases, a single bacterial species can express a wide variety of fimbriae with different binding specificities (Low et al., 1996). The fimbriae of Gram-negative bacteria have been extensively studied, particularly those of Escherichia coli and Salmonella spp. (Mol & Oudega, 1996; Soto & Hultgren, 1999). In these organisms, fimbrial proteins are encoded by plasmid or chromosomal genes organized into operons. In most cases, the structural and regulatory genes involved in the biogenesis of fimbriae are contiguous. However, in some cases they are separated and constitute distinct operons (Thanassi & Hultgren, 2000). Unlike the fimbriae of Gram-negative bacteria, little information is available on the 3Present address: Oral Microbiology, Faculty of Dentistry, University of Toronto, 124 Edward St, Toronto, ON, Canada M5G 1G6. Abbreviations: IMS, immunomagnetic separation; pAbs, polyclonal antibodies; PTA, phosphotungstic acid. The GenBank accession number for the sequence reported in this paper is AF353638.

0002-6592 G 2004 SGM

Printed in Great Britain

structure, composition and genetics of the fimbriae of Gram-positive bacteria. The fimbriae of Actinomyces naeslundii (Yeung, 2000) and Streptococcus parasanguinis (Wu & Fives-Taylor, 2001), two members of the indigenous oral flora, have been the most extensively studied. Fimbriae have also been reported in Streptococcus salivarius, a pioneer species that colonizes the human oral mucosa (Handley et al., 1984). Recent work suggested that S. salivarius fimbriae are involved in the coaggregation of S. salivarius with the periodontopathogen Prevotella intermedia (Le´vesque et al., 2003). Although fimbriae have been observed on up to 50 % of S. salivarius cells in the human oral cavity, no information is available on their biochemical composition and no molecular genetic studies have been done to investigate the genes involved in their biosynthesis and assembly (Handley et al., 1999). We purified S. salivarius fimbriae and determined a partial internal amino acid sequence for the fimbrillin. The internal amino acid sequence was composed of a repeated motif of two amino acids alternating with two modified residues: A/X/T-E-Q-M/W. We postulated that these residues (X and/or W) might be glycosylated amino acids, since carbohydrates are present in pure preparations of fimbriae (Le´vesque et al., 2001). To date, no effective methods have been described for the complete dissociation of S. salivarius fimbriae into subunits, suggesting that the fimbrial subunits may be covalently linked. To identify the genes involved in the biogenesis of S. salivarius fimbriae, transposon mutagenesis was performed to generate 189

C. Le´vesque, C. Vadeboncoeur and M. Frenette

mutants with altered fimbrial expression. In this paper, we report the isolation of a fimbria-deficient mutant of S. salivarius and the identification of a fimbria-associated protein.

were incubated at 37 uC for 48 h. Clones with altered fimbrial expression were identified by colony immunoblot using pAb HL-72 and observed by transmission electron microscopy as described below. Colony immunoblot analysis. Bacterial colonies were transferred

METHODS Bacterial strains and culture conditions. All S. salivarius strains

were grown in a Hogg–Jago lactose (HJL) medium (Marciset & Mollet, 1994) and incubated at 37 uC, or at 30 uC when they harboured pTV1-OK. A37 is a hyperfimbriated mutant isolated from S. salivarius ATCC 25975 by positive selection for resistance to 0?5 mM 2-deoxyglucose (Brochu et al., 1993; Gauthier et al., 1990) and was grown in HJL broth supplemented with 0?5 mM 2-deoxyglucose. E. coli strains were grown aerobically at 37 uC in Luria–Bertani (LB) medium, or at 30 uC in LB-kanamycin broth when they harboured pTV1-OK. When needed, antibiotics were added as follows: 50 mg ampicillin ml21 or 30 mg kanamycin ml21 for E. coli, and 10 mg erythromycin ml21 or 500 mg kanamycin ml21 for S. salivarius.

to nitrocellulose membranes (Micron Separations) and allowed to dry at 37 uC for 10 min. The membranes were incubated in 20 mM Tris/HCl (pH 7?5)/0?5 M NaCl (TBS) containing 3 % (w/v) skim milk for 1 h. All procedures were carried out at room temperature with gentle agitation. The membranes were then transferred to TBS/ 3 % skim milk containing pAb HL-72 diluted 1 : 10 000 and incubated for 1 h. The membranes were washed twice (15 min each time) in TBS containing 0?05 % (v/v) Tween 20 (TTBS) and then incubated with alkaline phosphatase-conjugated goat anti-rabbit IgG (Jackson ImmunoResearch Laboratories) diluted 1 : 5000 in TBS/3 % skim milk for 1 h. Finally, the membranes were washed with TTBS twice (15 min each time) and with TBS once (15 min), and the antibody conjugate was detected with nitro blue tetrazolium/5-bromo4-chloro-3-indolyl phosphate as the substrate. The reaction was stopped by washing the membranes in distilled water. Negative controls were prepared by replacing pAb HL-72 with the same dilution of normal rabbit serum.

Transposon mutagenesis. Plasmid pTV1-OK is a thermosensitive

delivery vector for Tn917 mutagenesis (Gutierrez et al., 1996). It harbours the aphA-3 gene, which confers kanamycin resistance on both E. coli and Gram-positive organisms, and the Tn917 transposon, which confers erythromycin resistance on Gram-positive organisms. Plasmid pTV1-OK was obtained by transformation of E. coli HB101 and selection at 30 uC on LB-kanamycin agar. The transformation of S. salivarius ATCC 25975 with pTV1-OK was performed using the technique described by Buckley et al. (1999). The transformants were selected at 30 uC on HJL-kanamycin agar. Transformants were negatively stained and observed with a transmission electron microscope as described below to ensure that the streptococcal cells still produced fimbriae. Tn917 mutagenesis was performed as described by Gutierrez et al. (1996) with the following modifications: a single transformed colony was inoculated into 10 ml HJL-kanamycin broth and grown overnight at 30 uC. The culture was diluted 1 : 1000 in HJL supplemented with erythromycin at a sublethal concentration of 0?04 mg ml21 and incubated at 30 uC for 3 h. Independent pools of Tn917 insertion mutants were generated by a temperature shift to 42 uC with overnight incubation. Immunomagnetic separation (IMS) procedure. Immunomagnetic beads [2?8 mm diam., 6?56108 beads ml21 (Dynabeads M-280; Dynal)]

with covalently linked sheep anti-rabbit immunoglobulin G (IgG) were used. The beads (approx. 6?56107) were washed twice with sterile PBS (pH 7?4) supplemented with 0?1 % (w/v) bovine serum albumin (BSA), separated by magnetic force, and then resuspended to the original volume in PBS–BSA. They were then coated with capture antibodies by incubating them for 3 h at 4 uC with specific antifimbriae polyclonal antibodies (pAb HL-72) (Le´vesque et al., 2001) diluted 1 : 50 in PBS–BSA. The mixture was rotated at 50 r.p.m. to prevent settling of the beads. The pAb-coated magnetic beads were stored at 4 uC and washed before use as described by the manufacturer. The S. salivarius transposon mutant library was washed twice with sterile PBS (pH 7?4) and sonicated on ice (2620 s, 20 % duty cycle, power level 3; Branson Sonic Power Co.) in PBS (pH 7?4) to break up streptococcal chains. A 15 ml volume of pAb-coated beads was mixed with 1 ml of the S. salivarius transposon mutant library (approx. 26106 cells) and incubated at 4 uC overnight with rotation at 50 r.p.m. Beads with attached bacteria were discarded by magnetic force. The supernatant was recovered and incubated again at 4 uC overnight with a 15 ml volume of pAb-coated beads. The isolation of S. salivarius Tn917 insertion mutants with altered fimbrial expression required eight rounds of IMS. The final supernatant was spread on HJL agar supplemented with 10 mg erythromycin ml21 and the plates 190

Electron microscopy. Ten-microlitre samples of bacterial sus-

pensions (approx. 86107 cells ml21) were applied to carbon-coated Formvar copper grids (Canemco) and negatively stained with 1 % (w/v) phosphotungstic acid (PTA) (pH 7?0) for 10 s. The grids were air-dried prior to examination with a JEOL 2000 transmission electron microscope operating at 80 kV. For immunogold labelling experiments, 10 ml samples of bacterial suspensions (approx. 86 107 cells ml21) were applied to carbon-coated Formvar copper grids. After 30 min, the grids were blocked with PBS (pH 7?2) containing 1 % (w/v) IgG-free and protease-free BSA (Jackson ImmunoResearch Laboratories) for 30 min. The grids were then incubated with specific rabbit pAbs diluted in blocking buffer for 1 h and rinsed in PBS (pH 7?2) and distilled water. The grids were incubated for 1 h with 12 nm colloidal gold-labelled donkey anti-rabbit IgG (Jackson ImmunoResearch Laboratories) diluted 1 : 40 in blocking buffer and rinsed in PBS (pH 7?2) and distilled water. The samples were negatively stained and examined as described above. DNA manipulations. Routine DNA manipulations were performed

as described by Sambrook et al. (1989). Large-batch preparations of plasmid DNA were obtained using a Qiagen Plasmid Maxi kit. Genomic DNA was isolated from streptococci using a Puregene DNA Isolation kit (Gentra Systems). For the Southern blot experiments, the 3?6 kb KpnI–StuI and 1?0 kb EcoRV fragments from pTV1-OK used as Tn917 and vector probes, respectively, were labelled with [a-32P]dATP by the random priming DNA labelling technique (Sambrook et al., 1989). Based on the nucleotide sequence of Tn917 (Shaw & Clewell, 1985), the oligonucleotide primers Tn917-L (59-TGTCACCGTCAAGTTAAATG-39) and Tn917-R (59GAACTATTTTTTGGGCGACG-39) were designed for the extremities of the transposon. Inverse PCR using the Tn917-L and Tn917-R primers was performed to recover the Tn917 insertion site from mutant D37 chromosomal DNA. Inverse PCR was carried out in a 100 ml volume containing 10 ml 106 PCR buffer (Amersham Biosciences), 5 ml 106 deoxynucleoside triphosphate mix (2 mM each dATP, dCTP, dGTP and dTTP), 1 ml each primer stock solution (50 mM), 5 ml template DNA (200 ng EcoRI-restricted and selfligated mutant D37 chromosomal DNA) and 2?5 U Taq DNA polymerase (Amersham Biosciences). A single 785 bp PCR product was amplified using a three-step profile: 1 min denaturation at 94 uC, 1 min annealing at 60 uC and 2 min extension at 72 uC for a total of 35 cycles. The 785 bp product was sequenced with the Tn917-L and Tn917-R primers. Isolation of DNA regions contiguous to the insertion site was carried out by chromosome walking using appropriate synthesized oligonucleotides. DNA sequencing Microbiology 150

Streptococcus salivarius csp operon was performed by the DNA sequencing service of Universite´ Laval. Computer-assisted DNA and protein analyses were performed using the Genetics Computer Group Sequence Analysis software package, Version 9.1 (Devereux et al., 1984). Northern blots. Total RNA was extracted from streptococci using

the RNeasy kit (Qiagen), according to the manufacturer’s protocol, and separated on 1 % (w/v) agarose/formaldehyde gels (Sambrook et al., 1989). Northern hybridization procedures were performed as described by Lortie et al. (2000). Immobilized RNA was hybridized with DNA probes labelled with [a-32P]dATP by the random priming DNA labelling technique (Sambrook et al., 1989). The orf176 and csp probes were amplified by PCR from S. salivarius ATCC 25975 genomic DNA using the following primers: D37-8 (59-GGAGAAACATATGAATACAC-39) and D37-9 (59-TTTAAGATCTATTATGCAAGTC-39) for orf176, and D37-5 (59-CTACGACTTCTGAAAATGCC-39) and D3713 (59-AATAGAGCCATGTCCTTTCG-39) for csp. Cloning into the expression vector. The full-length coding regions

of cspA and cspB were first amplified by PCR using the following primers: D37-14 (59-AAAGGAAATTGCTTACATATGG-39; with a 2 bp substitution creating an NdeI site) and D37-15 (59-GTCCTTTCGTGTCATATGAG-39; with a 2 bp substitution creating an NdeI site) for cspA, and D37-24 (59-ATTCATATGTTTAAAGACTCAA-39; with a 2 bp substitution creating an NdeI site) and D37-25 (59AAAGGATCCTTTAAGAAGTG-39; with a 2 bp substitution creating a BamHI site) for cspB. PCR products were then cloned into pCR2.1-TOPO (Invitrogen) as described by the manufacturer. Recombinant plasmids designated pSS301 (1446 bp amplified fragment of cspA in pCR2.1-TOPO) and pSS307 (2576 bp amplified fragment of cspB in pCR2.1-TOPO) were used to transform E. coli DH10B. The 1413 bp NdeI fragment of pSS301 and the 2563 bp NdeI–BamHI fragment of pSS307 were subcloned in-frame downstream from the hexa-His sequence in the T7 expression vector pET28a(+) (Novagen) precut by the same enzymes and transformed into E. coli DH10B. The nucleotide sequences of the inserts were confirmed by sequence analysis. Plasmids designated pSS305 and pSS308 were then transformed into E. coli BL21(DE3) to express the hexa-His-tagged CspA and hexa-His-tagged CspB recombinant proteins, respectively. Expression of recombinant proteins. E. coli BL21(DE3)(pSS305)

and E. coli BL21(DE3)(pSS308) cells were incubated aerobically at 37 uC in 100 ml LB-kanamycin supplemented with 1 % (w/v) glucose. When the OD600 reached 0?3, the cells were transferred to room temperature and incubated until the culture reached an OD600 of 0?6. IPTG was then added at a final concentration of 1 mM to induce expression of recombinant proteins and the incubation was continued for a further 3 h at room temperature. The cells were collected by centrifugation (12 000 g, 10 min, 4 uC), resuspended in 2 ml 16 binding buffer (Novagen) and disrupted on ice by sonication (10620 s, 45 % duty cycle, power level 4?5). The soluble fractions of the disrupted cells were recovered by centrifugation (12 000 g, 5 min, 4 uC). The hexa-His-tagged recombinant proteins were then purified by affinity chromatography on Ni2+-nitrilotriacetic acid (Ni-NTA) resin (Novagen) as described by the manufacturer. SDS-PAGE. Approximately 30 mg purified recombinant CspA

(rCspA) and recombinant CspB (rCspB) proteins were submitted to SDS-PAGE using the buffer system of Laemmli (1970) at a constant voltage (200 V) with a 10 % polyacrylamide separating gel and a 4?5 % polyacrylamide stacking gel. The proteins were then transferred to a nitrocellulose membrane as described by Sambrook et al. (1989). After staining with 10 % (w/v) Ponceau Red in 1 % (v/v) acetic acid for 2 min, the membrane was briefly destained in distilled water and the spots corresponding to rCspA and rCspB proteins were excised and kept at 4 uC until used. http://mic.sgmjournals.org

Production of antibodies. pAbs were raised against rCspA (pAb

KL-46) and rCspB (pAb KL-49), respectively, by immunizing two female New Zealand White rabbits. Approximately 30 mg purified recombinant proteins that had been transferred to nitrocellulose membrane were emulsified with 200 ml PBS (pH 7?2) and 200 ml TiterMax adjuvant (Cederlane Laboratories) and injected intramuscularly into each hind limb of the animals. Two weeks later, the rabbits received a second injection (as described above) and the blood samples were taken two weeks after the final injection.

RESULTS Construction of a Tn917 mutant library of S. salivarius To identify genes involved in the biogenesis of S. salivarius fimbriae, we generated a transposon mutant library using the Tn917 delivery vector pTV1-OK (Gutierrez et al., 1996). Approximately 100 randomly selected erythromycin-resistant (Emr) mutants were replicated on medium supplemented with kanamycin to determine the percentage of Tn917 insertion mutants that had lost the plasmid backbone. We found that 90 % of the selected Emr clones were kanamycinsensitive (Kms), while 10 % were kanamycin-resistant (Kmr). This suggested that 90 % of the Emr clones had integrated the transposon into the chromosome (Emr) and had lost the plasmid backbone (Kms). Southern blot analyses of chromosomal DNA from 15 randomly selected Emr/Kms mutants digested with StuI (StuI cuts once in pTV1-OK) and probed with Tn917 DNA indicated that transposition was random and resulted in a single chromosomal insertion (data not shown). When analysed by Southern blotting, the StuI-digested chromosomal DNA of ten randomly selected Emr/Kmr clones reacted with the Tn917 and vector probes, indicating that replicon fusion had occurred (data not shown). Cointegrated structures have already been reported for transposons of the Tn3 family, to which Tn917 belongs (Brown & Evans, 1991; Heffron, 1983). The transposon mutant library of S. salivarius ATCC 25975 was used to identify fimbria-related gene(s) by isolating mutants with altered fimbrial expression. Screening the transposon library for fimbria-deficient mutants To enrich the library of S. salivarius mutants with altered fimbrial expression, we developed an IMS procedure using pAbs directed specifically against S. salivarius fimbriae. Colony immunoblot experiments conducted with over 4000 S. salivarius mutants obtained by IMS resulted in the isolation of three mutants with altered fimbrial expression: C34 (Emr/Kms), D11 (Emr/Kms) and D37 (Emr/Kmr). Transmission electron microscopy observations clearly showed fimbriae extending from the cell surfaces of mutants C34 and D11, but in very small amounts compared to the wild-type strain, whereas mutant D37 was completely devoid of fimbriae. As shown in Fig. 1, immunolabelling with pAb HL-72 stained fimbriae extending from the cell surface of wild-type cells (Fig. 1a), while no organized 191

C. Le´vesque, C. Vadeboncoeur and M. Frenette

(a)

(b)

Fig. 1. Immunolocalization of fimbriae on S. salivarius ATCC 25975 (a) and mutant D37 (b) cell surfaces by electron microscopy. Bacterial cells were successively incubated with pAb HL-72 and 12 nm colloidal gold-conjugated donkey anti-rabbit IgG, and then negatively stained with 1 % PTA (pH 7?0). Bars, 100 nm.

structures resembling fimbriae were recognized by these pAbs on mutant D37 cells (Fig. 1b). Mutant D37 was selected for further investigation. Recovery of the Tn917-interrupted gene The presence of Tn917 in the chromosome of mutant D37 was confirmed by Southern blot analysis of StuI-digested chromosomal DNA using the Tn917 DNA probe. Hybridization with the vector probe confirmed that the plasmid had integrated into the chromosome of this mutant (data not shown). Repeated subculturing in selective media confirmed that D37 was stable. To recover the interrupted gene, chromosomal DNA from D37 was digested with EcoRI and selfligated. Inverse PCR performed using the Tn917-L and Tn917-R primers, which are complementary to the left and right ends of the transposon, respectively, allowed the insertion site to be isolated. An ORF, named orf176, was found to be interrupted by the insertion of Tn917 and the plasmid backbone at the TATTA target site (Fig. 2). orf176 was predicted to encode a 176 aa protein with an estimated molecular mass of 20?4 kDa and an isoelectric point (pI) of 4?39. In silico structural prediction analyses suggested that the protein was cytoplasmic. BLAST searches 192

of current databases did not reveal any significant identity with bacterial fimbrial or non-fimbrial proteins. A potential ribosome-binding site was identified 6 bp upstream from the predicted start codon of orf176 (Fig. 2b). A promoter sequence (TTGCTC-N18-TACACT) was also identified 50 bp upstream from the ATG start codon by its similarity to the consensus 235 and 210 sequences (Lisser & Margalit, 1993). The TRTG motif (216 region) found in several Bacillus subtilis and other Gram-positive bacteria promoters (Voskuil & Chambliss, 1998) was also present. A terminator-like structure (factor-independent terminator) was located 32 bp downstream from the stop codon of orf176 (Fig. 2b). DNA sequence analysis of the chromosomal regions adjacent to orf176 To investigate the possibility that the insertion of a large DNA fragment could have affected the expression of contiguous genes, DNA regions flanking orf176 were isolated using direct and inverse PCR approaches. The organization of the ORFs located upstream and downstream from orf176 is shown in Fig. 2(a). Nucleotide sequence analysis of the 1?2 kb portion upstream from orf176 revealed the presence of an ORF oriented in the same transcriptional direction. Microbiology 150

Streptococcus salivarius csp operon (a)

(b)

Fig. 2. (a) Map of the S. salivarius ATCC 25975 csp locus. Arrows indicate the direction of transcription. The insertion of Tn917 and the plasmid backbone inserted at the TATTA target site in the chromosome of mutant D37 is indicated by a triangle. The 785 bp product obtained by inverse PCR using the Tn917-L (L) and Tn917-R (R) primers is also indicated. The orf176 and csp DNA probes used in the Northern blots are indicated by hatched boxes under the corresponding DNA fragments. (b) Partial nucleotide and amino acid sequences of the S. salivarius ATCC 25975 csp locus. Putative ”35, ”16 and ”10 promoter sites and putative ribosome-binding sites (RBS) are indicated. The insertion of Tn917 and the plasmid backbone at the TATTA target site (underlined) in the chromosome of mutant D37 is indicated by an arrow. The factorindependent terminator-like structure between orf176 and cspA is underlined with a broken line. The LPXTG motifs are in bold.

The deduced amino acid sequence of this ORF had significant levels of identity with the DNA topoisomerase I of Enterococcus faecalis (Bidnenko et al., 1998) (E value 1e231) and the topoisomerase III of B. subtilis (Kunst et al., 1997) (3e214) and E. coli (Perna et al., 2001) (3e211). This ORF was therefore designated topA. Nucleotide sequence analysis of the 4?2 kb portion downstream from orf176 revealed the presence of two ORFs transcribed in the same direction as orf176 (Fig. 2). No ORFs of significant size were detected on the reverse strand. The first ORF was 147 bp downstream from orf176. It was predicted to encode a 459 aa protein with an estimated molecular mass of 46?2 kDa and a pI of 4?20. The first 31 aa were predicted to compose a signal peptide. The putative leader sequence cleavage site of the precursor protein was predicted between the alanine and the aspartic acid at positions 31 and 32, respectively. A cell-wall-anchoring motif (LPXTG) of Gram-positive bacterial surface proteins http://mic.sgmjournals.org

(Fischetti et al., 1990) followed by a hydrophobic domain and a charged tail were also identified at the C terminus of the protein. This ORF was thus named cspA since the encoded polypeptide was predicted to be a cell-surface protein. BLAST searches revealed significant identity with SAG1996, a putative cell-surface anchor protein of Streptococcus agalactiae (Tettelin et al., 2002) (E value 2e212). The second ORF, named cspB, was 17 bp downstream from cspA and was predicted to encode an 846 aa protein with an estimated molecular mass of 90?6 kDa and a pI of 4?35. The first 42 aa were predicted to compose a signal peptide. Thus, the putative leader sequence cleavage site of the precursor protein encoded by cspB was predicted to be between the alanine and the aspartic acid at positions 42 and 43, respectively. An LPXTG motif followed by a hydrophobic domain and a charged tail were also identified at the C terminus of the protein. BLAST searches of current databases revealed significant identity with cell-surface proteins of Gram-positive bacteria that belong to two families of 193

C. Le´vesque, C. Vadeboncoeur and M. Frenette

adhesins (E values ranging from 9e232 to 8e24). The first family includes members of the antigen I/II family of oral streptococci (Jenkinson & Demuth, 1997): SspA and SspB of Streptococcus gordonii; PAc, SpaP and Sr of Streptococcus mutans; and SpaA of Streptococcus sobrinus. The second family includes the aggregation substances of enterococci (Weaver, 2000): Asa373, Asc10 and Asa1 of E. faecalis; and Ash701 of Enterococcus faecium. Potential ribosome-binding sites were located 9 and 5 bp upstream from the predicted start codons of cspA and cspB, respectively (Fig. 2b). A promoter sequence (TTGAGT-N16TATCAT) was located 37 bp upstream from the ATG start codon of cspA. There were no potential promoter sequences or apparent transcription-termination sequences between cspA and cspB, suggesting that they were cotranscribed. A terminator-like structure (factor-independent terminator) was found 68 bp downstream from cspB. Transcriptional studies Northern blot analyses of total RNA isolated from S. salivarius ATCC 25975 and fimbria-deficient mutant D37 using the orf176 coding region as a DNA probe allowed the detection of a single 0?6 kb mRNA transcript in wild-type cells (data not shown). The size of this transcript was consistent with the transcription of orf176 into a monocistronic mRNA. Since orf176 was interrupted by the insertion of Tn917 as well as the plasmid backbone in the chromosome of mutant D37, the 0?6 kb orf176 transcript was not detected in these cells (not shown). To assess the effect of the cointegrated structure on the transcription of the genes located downstream, Northern blots were also conducted using a csp DNA probe. The probe allowed the detection of a single 4 kb mRNA transcript in wild-type cells (Fig. 3, lane 1). The size of this transcript was consistent with the cotranscription of cspA and cspB. Interestingly, no mRNA transcript was detected with the csp probe in the fimbria-deficient mutant D37 cells (Fig. 3, lane 2).

1

2

kb

4

2.9

23S

1.5

16S

Fig. 3. Northern blot of total RNA isolated from S. salivarius ATCC 25975 (lane 1) and mutant D37 (lane 2) hybridized with the csp DNA probe. The arrow indicates the cspA-cspB 4 kb transcripts. The bands at 1?5 and 2?9 kb correspond to nonspecific hybridization to 16S and 23S rRNA, respectively.

Immunolocalization of CspA and CspB The fact that the csp operon (i) was located downstream from the Tn917-interrupted gene, (ii) contained genes encoding predicted cell-surface proteins and (iii) was not transcribed in the fimbria-deficient mutant D37 suggested that this operon could encode fimbria-related proteins. We therefore used pAbs raised against the rCspA and rCspB proteins to determine whether there was an association between CspA and CspB and S. salivarius fimbriae. As shown in Fig. 4, immunogold labelling with pAb KL-46 raised against rCspA did not stain the fimbriae (indicated by the arrow) extending from the surface of wild-type cells. Immunogold labelling with pAb KL-46 was also conducted using whole cells of A37, a hyperfimbriated mutant of S. salivarius ATCC 25975. As with the wild-type cells, no labelling was observed with A37 even though the mutant produced many more fimbriae than the parental strain (data not shown). Dot blot experiments conducted using fimbriae 194

Fig. 4. Immunolocalization of CspA on the S. salivarius ATCC 25975 cell surface by electron microscopy. Bacterial cells were successively incubated with pAb KL-46 and 12 nm colloidal goldconjugated donkey anti-rabbit IgG and then negatively stained with 1 % PTA (pH 7?0). Fimbriae that are not labelled by the antibodies are indicated by the arrow (see text). Bar, 100 nm. Microbiology 150

Streptococcus salivarius csp operon

purified from glass bead extracts (Le´vesque et al., 2001) did not react with the anti-rCspA pAbs (data not shown). These results suggested that CspA was not associated with the fimbrial structure extending outside the cell wall. Immunogold labelling with pAb KL-49 raised against rCspB stained the fimbriae extending from the surface of wild-type (Fig. 5a) and hyperfimbriated mutant A37 (Fig. 5b) cells. As can be seen in Fig. 5, the hyperfimbriated mutant cells showed more extensive labelling of fimbriae than the wild-type cells. No immunolabelling by pAb KL-49 of the fimbria-deficient mutant D37 was observed (Fig. 5c). When dot blot experiments were conducted using purified fimbriae, the results showed that anti-rCspB pAbs reacted with the fimbrial preparation, confirming that CspB was associated with S. salivarius fimbriae.

DISCUSSION Bacteria produce a variety of surface structures and appendages that are involved in host colonization. In the oral cavity, bacterial adherence to salivary components, the oral mucosa and other bacteria is often mediated by fimbriae (Hamada et al., 1998; Handley et al., 1999; Lamont & Jenkinson, 2000). However, little is known about the structure, composition and genetics of the fimbriae of Grampositive bacteria. The difficulty in studying these fimbriae is largely due to their resistance to extraction from the cell surface coupled with possible covalent links between the fimbrial subunits, making them impossible to dissociate into individual subunits for SDS-PAGE analysis. In this study, we sought to identify the genes involved in the biogenesis of S. salivarius fimbriae. We thus devised an original immunogenetically based strategy that enabled us to isolate the fimbria-deficient mutant D37. The interrupted gene of mutant D37 was recovered by inverse PCR and the gene was sequenced in the wild-type strain. Inactivation of orf176 in S. salivarius generated a fimbriadeficient phenotype. Since the structural and regulatory genes involved in the biogenesis of fimbriae are often clustered within a single chromosomal region (Low et al., 1996; Mol & Oudega, 1996; Thanassi & Hultgren, 2000), DNA regions flanking orf176 were isolated and sequenced. orf176 was located between topA, which encodes a putative topoisomerase, and the csp operon containing two genes (cspA and cspB), which encode predicted cell-surface proteins. Interestingly, the csp operon was not transcribed in the fimbriadeficient mutant D37. This may suggest that transcription of the csp operon was linked to orf176 expression and that the Orf176 protein acted as its transcriptional regulator. However, in silico structural prediction analyses of Orf176 failed to assign the classic helix–turn–helix motif to a portion of the polypeptide. The absence of this DNAinteracting motif and the fact that Orf176 did not share significant identity with any proteins in current databases, including unfinished microbial genomes, suggested that Orf176 might be a novel, fairly rare regulatory protein. http://mic.sgmjournals.org

Although transcription of orf176 generated a single monocistronic transcript, we cannot exclude the possibility that the insertion of a large fragment of DNA (at least two copies of Tn917 along with the plasmid backbone) might affect the architecture of this chromosomal region, thus altering the transcription of the csp operon. Gene knockout of orf176 combined with complementation experiments will be required to discriminate between these hypotheses. However, until now, all our attempts to inactivate orf176 in the wild-type strain by allelic exchange or to complement the mutant strain have proven unsuccessful due to the resistance of S. salivarius to genetic manipulation. The cellular location of CspA and CspB, two predicted cellsurface proteins, was examined by immunolabelling experiments using pAbs directed against purified CspA and CspB recombinant proteins. Although CspA was not associated with the fimbrial structure extending outside the cell wall, we could not rule out the possibility that it was associated with the cell-wall-embedded portion of the fimbriae, a portion that is not accessible for detection by antibodies. CspA might also be an accessory protein that is not integrated into the fimbrial structure, but is needed for the biogenesis of the fimbriae. In A. naeslundii T14V, six accessory fimbrial proteins, as well as the fimbrillin, are required for the correct assembly of type 1 fimbriae (Yeung, 2000; Yeung & Ragsdale, 1997). Immunolocalization experiments with anti-rCspB pAbs showed that CspB was associated with the fimbriae. Although CspB was not the S. salivarius fimbrillin as it did not possess the A/X/T-E-Q-M/W repeated motif previously reported (Le´vesque et al., 2001), this protein was also a fimbrial subunit since gold particles were observed attached to the fimbriae. In E. coli type 1 fimbriae, fimbrillin as well as other fimbrial subunits are interspersed at intervals along the fimbriae (Klemm & Krogfelt, 1994; Krogfelt & Klemm, 1988). In addition to its role in the structural organization of fimbriae, CspB may also have adhesin functions. Indeed, it was shown to share identity with several proteins of the streptococcal antigen I/II family. Members of the antigen I/II family are multifunctional adhesins that mediate interactions between oral streptococci and other oral bacteria, cell matrix proteins and human salivary glycoproteins (Jenkinson, 1994; Jenkinson & Demuth, 1997; Jenkinson & Lamont, 1997). Consistent with this hypothesis, we recently demonstrated that fimbriated S. salivarius cells coaggregated with P. intermedia while fimbria-deficient mutant D37 cells were unable to interact with this periodontopathogen (Le´vesque et al., 2003). The CspA and CspB proteins contained a cell-anchoring motif at the carboxyl terminus, suggesting that they are probably C-terminal-linked surface proteins. Indeed, all C-terminal-linked surface proteins from Gram-positive bacteria have a specific sorting signal composed of similar arrangements of amino acids (Fischetti, 2000; Navarre & Schneewind, 1999). CspA and CspB are the first putative C-terminal-linked surface proteins to be identified in 195

C. Le´vesque, C. Vadeboncoeur and M. Frenette

(a)

(b)

(c)

Fig. 5. Immunolocalization of CspB on S. salivarius ATCC 25975 (a), hyperfimbriated mutant A37 (b) and fimbria-deficient mutant D37 (c) cell surfaces by electron microscopy. The bacterial cells were successively incubated with pAb KL-49 and 12 nm colloidal gold-conjugated donkey anti-rabbit IgG and then negatively stained with 1 % PTA (pH 7?0). Bars, 100 nm.

S. salivarius. The only surface protein identified in S. salivarius is b-D-fructosyltransferase, which catalyses the polymerization of the fructose moiety of sucrose in fructans (Rathsam et al., 1993). However, this protein does not possess the LPXTG motif and its anchoring mechanism has not yet been determined (Rathsam & Jacques, 1998). Based on the general model for C-terminal-linked surface proteins of Gram-positive bacteria, these proteins are cleaved between the threonyl and glycyl residues of the LPXTG motif and the newly liberated carboxy terminus of threonine is anchored to cell-wall peptidoglycan (Navarre & Schneewind, 1994, 1999; Schneewind et al., 1995). Since CspB is associated with the fimbriae, it is more probable that the carboxy terminus of threonine, rather than being 196

anchored to the cell-wall peptidoglycan, is covalently linked to another fimbrial subunit, as proposed by Yeung et al. (1998) for A. naeslundii fimbriae and Ton-That & Schneewind (2003) for Corynebacterium diphtheriae fimbriae. This covalent linkage between fimbrial subunits would also account for the inability to dissociate S. salivarius, A. naeslundii and C. diphtheriae fimbriae into individual subunits. This work represents the first molecular genetic study undertaken to identify the genes involved in the biogenesis of S. salivarius fimbriae. Further studies will be needed to define the mechanisms used by Gram-positive bacteria to assemble fimbrial structures and regulate their synthesis. Microbiology 150

Streptococcus salivarius csp operon

ACKNOWLEDGEMENTS

Jenkinson, H. F. & Demuth, D. R. (1997). Structure, function and

We would like to thank J. A. Gutierrez for providing the pTV1-OK plasmid, L.-A. Lortie for assisting with the DNA sequence analyses and A. Zelada for helpful discussions. This research was supported by Canadian Institutes of Health Research (CIHR) operating grant MOP 36338 and the Fonds pour la Formation de Chercheurs et l’Aide a` la Recherche (FCAR) from the Province of Quebec (research team funding program).

immunogenicity of streptococcal antigen I/II polypeptides. Mol Microbiol 23, 183–190. Jenkinson, H. F. & Lamont, R. J. (1997). Streptococcal adhesion and colonization. Crit Rev Oral Biol Med 8, 175–200. Klemm, P. & Krogfelt, K. A. (1994). Type 1 fimbriae of Escherichia coli. In Fimbriae, Adhesion, Genetics, Biogenesis and Vaccines, pp. 9–26. Edited by P. Klemm. Boca Raton, FL: CRC Press. Klemm, P. & Schembri, M. A. (2000). Bacterial adhesins: function and structure. Int J Med Microbiol 290, 27–35.

REFERENCES

Krogfelt, K. A. & Klemm, P. (1988). Investigation of minor

Bidnenko, V., Ehrlich, S. D. & Jannie`re, L. (1998). In vivo relations

components of Escherichia coli type 1 fimbriae: protein chemical and immunological aspects. Microb Pathog 4, 231–238.

between pAMbeta1-encoded type I topoisomerase and plasmid replication. Mol Microbiol 28, 1005–1016. Brochu, D., Trahan, L., Jacques, M., Lavoie, M. C., Frenette, M. & Vadeboncoeur, C. (1993). Alterations in the cellular envelope of

spontaneous IIILman -defective mutants of Streptococcus salivarius. J Gen Microbiol 139, 1291–1300.

Brown, N. L. & Evans, L. R. (1991). Transposition in prokaryotes:

transposon Tn501. Res Microbiol 142, 689–700. Buckley, N. D., Vadeboncoeur, C., Leblanc, D. J., Lee, L. N. & Frenette, M. (1999). An effective strategy, applicable to Streptococcus

salivarius and related bacteria, to enhance or confer electroporation competence. Appl Environ Microbiol 65, 3800–3804. Devereux, J., Haeberli, P. & Smithies, O. (1984). A comprehensive

set of sequence analysis programs for the VAX. Nucleic Acids Res 12, 387–395. Ferna´ndez, L. A. & Berenguer, J. (2000). Secretion and assembly of

regular surface structures in Gram-negative bacteria. FEMS Microbiol Rev 24, 21–44. Fischetti, V. A. (2000). Surface proteins on Gram-positive bacteria.

In Gram-Positive Pathogens, pp. 11–24. Edited by V. A. Fischetti, R. P. Novick, J. J. Ferretti, D. A. Portnoy & J. I. Rood. Washington, DC: American Society for Microbiology. Fischetti, V. A., Pancholi, V. & Schneewind, O. (1990). Conservation

of a hexapeptide sequence in the anchor region of surface proteins from Gram-positive cocci. Mol Microbiol 4, 1603–1605. Gauthier, L., Bourassa, S., Brochu, D. & Vadeboncoeur, C. (1990).

Control of sugar utilization in oral streptococci. Properties of phenotypically distinct 2-deoxyglucose-resistant mutants of Streptococcus salivarius. Oral Microbiol Immunol 5, 352–359. Gutierrez, J. A., Crowley, P. J., Brown, D. P., Hillman, J. D., Youngman, P. & Bleiweis, A. S. (1996). Insertional mutagenesis and

recovery of interrupted genes of Streptococcus mutans by using transposon Tn917: preliminary characterization of mutants displaying acid sensitivity and nutritional requirements. J Bacteriol 178, 4166–4175. Hamada, S., Amano, A., Kimura, S., Nakagawa, I., Kawabata, S. & Morisaki, I. (1998). The importance of fimbriae in the virulence and

ecology of some oral bacteria. Oral Microbiol Immunol 13, 129–138. Handley, P. S., Carter, P. L. & Fielding, J. (1984). Streptococcus

salivarius strains carry either fibrils or fimbriae on the cell surface. J Bacteriol 157, 64–72. Handley, P. S., McNab, R. & Jenkinson, H. F. (1999). Adhesive

surface structures on oral bacteria. In Dental Plaque Revisited. Oral Biofilms in Health and Disease, pp. 145–170. Edited by H. N. Newman & M. Wilson. London: Eastman Dental Institute. Heffron, F. (1983). Tn3 and its relatives. In Mobile Genetic Elements, pp. 223–260. Edited by J. A. Shapiro. London: Academic Press Jenkinson, H. F. (1994). Cell surface protein receptors in oral streptococci. FEMS Microbiol Lett 121, 133–140.

http://mic.sgmjournals.org

Kunst, F., Ogasawara, N., Moszer, I. & 149 other authors (1997).

The complete genome sequence of the Gram-positive bacterium Bacillus subtilis. Nature 390, 249–256. Laemmli, K. (1970). Cleavage of structural proteins during the assembly of the head of bacteriophage T4. Nature 227, 680–685. Lamont, R. J. & Jenkinson, H. F. (2000). Adhesion as an ecological determinant in the oral cavity. In Oral Bacterial Ecology: the Molecular Basis, pp. 131–168. Edited by H. K. Kuramitsu & R. P. Ellen. Wymondham: Horizon Scientific Press. Le´vesque, C., Vadeboncoeur, C., Chandad, F. & Frenette, M. (2001).

Streptococcus salivarius fimbriae are composed of a glycoprotein containing a repeated motif assembled into a filamentous nondissociable structure. J Bacteriol 183, 2724–2732. Le´vesque, C., Lamothe, J. & Frenette, M. (2003). Coaggregation of Streptococcus salivarius with periodontopathogens: evidence for involvement of fimbriae in the interaction with Prevotella intermedia. Oral Microbiol Immunol 18, 333–337. Lisser, S. & Margalit, H. (1993). Compilation of Escherichia coli mRNA promoter sequences. Nucleic Acids Res 21, 1507–1516. Lortie, L.-A., Pelletier, M., Vadeboncoeur, C. & Frenette, M. (2000).

The gene encoding IIABLMan in Streptococcus salivarius is part of a tetracistronic operon encoding a phosphoenolpyruvate : mannose/ glucose phosphotransferase system. Microbiology 146, 677–685. Low, D., Braaten, B. & Van der Woude, M. (1996). Fimbriae. In Escherichia Coli and Salmonella. Cellular and Molecular Biology, pp. 146–157. Edited by F. C. Neidhardt. Washington, DC: American Society for Microbiology. Marciset, O. & Mollet, B. (1994). Multifactorial experimental designs for optimizing transformation: electroporation of Streptococcus thermophilus. Biotechnol Bioeng 43, 490–496. Mol, O. & Oudega, B. (1996). Molecular and structural aspects of fimbriae biosynthesis and assembly in Escherichia coli. FEMS Microbiol Rev 19, 25–52. Navarre, W. W. & Schneewind, O. (1994). Proteolytic cleavage and cell wall anchoring at the LPXTG motif of surface proteins in Grampositive bacteria. Mol Microbiol 14, 115–121. Navarre, W. W. & Schneewind, O. (1999). Surface proteins of Grampositive bacteria and mechanisms of their targeting to the cell wall envelope. Microbiol Mol Biol Rev 63, 174–229. Perna, N. T., Plunkett, G., III, Burland, V. & 25 other authors (2001).

Genome sequence of enterohaemorrhagic Escherichia coli O157 : H7. Nature 409, 529–533. Rathsam, C. & Jacques, N. A. (1998). Role of C-terminal domains in surface attachment of the fructosyltransferase of Streptococcus salivarius ATCC 25975. J Bacteriol 180, 6400–6403. Rathsam, C., Giffard, P. M. & Jacques, N. A. (1993). The cell-bound fructosyltransferase of Streptococcus salivarius: the carboxyl terminus specifies attachment in a Streptococcus gordonii model system. J Bacteriol 175, 4520–4527. 197

C. Le´vesque, C. Vadeboncoeur and M. Frenette Sambrook, J., Fritsch, E. F. & Maniatis, T. (1989). Molecular Cloning:

a Laboratory Manual, 2nd edn. Cold Spring Harbor, NY: Cold Spring Harbor Laboratory.

Voskuil, M. I. & Chambliss, G. H. (1998). The 216 region of Bacillus subtilis and other Gram-positive bacterial promoters. Nucleic Acids Res 26, 3584–3590.

Schneewind, O., Fowler, A. & Faull, K. F. (1995). Structure of the cell

Weaver, K. E. (2000). Enterococcal genetics. In Gram-Positive

wall anchor of surface proteins in Staphylococcus aureus. Science 268, 103–106.

Pathogens, pp. 259–271. Edited by V. A. Fischetti, R. P. Novick, J. J. Ferretti, D. A. Portnoy & J. I. Rood. Washington, DC: American Society for Microbiology.

Shaw, J. H. & Clewell, D. B. (1985). Complete nucleotide sequence of macrolide-lincosamide-streptogramin B-resistance transposon Tn917 in Streptococcus faecalis. J Bacteriol 164, 782–796.

Wu, H. & Fives-Taylor, P. M. (2001). Molecular strategies for fimbrial

Soto, G. E. & Hultgren, S. J. (1999). Bacterial adhesins: common

Yeung, M.

themes and variations in architecture and assembly. J Bacteriol 181, 1059–1071. Tettelin, H., Masignani, V., Cieslewicz, M. J. & 40 other authors (2002). Complete genome sequence and comparative genomic

analysis of an emerging human pathogen, serotype V Streptococcus agalactiae. Proc Natl Acad Sci U S A 99, 12391–12396. Thanassi, D. G. & Hultgren, S. J. (2000). Assembly of complex

organelles: pilus biogenesis in Gram-negative bacteria as a model system. Methods 20, 111–126.

expression and assembly. Crit Rev Oral Med 12, 101–115. K. (2000). Actinomyces: surface macromolecules and bacteria–host interactions. In Gram-Positive Pathogens, pp. 583– 593. Edited by V. A. Fischetti, R. P. Novick, J. J. Ferretti, D. A. Portnoy & J. I. Rood. Washington, DC: American Society for Microbiology.

Yeung, M. K. & Ragsdale, P. A. (1997). Synthesis and function

of Actinomyces naeslundii T14V type 1 fimbriae require the expression of additional fimbria-associated genes. Infect Immun 65, 2629–2639.

Ton-That, H. & Schneewind, O. (2003). Fimbrial assembly in

Yeung, M. K., Donkersloot, J. A., Cisar, J. O. & Ragsdale, P. A. (1998). Identification of a gene involved in assembly of

Corynebacterium diphtheriae. ASM Washington, DC, Abstract B-098.

Actinomyces naeslundii T14V type 2 fimbriae. Infect Immun 66, 1482–1491.

198

103rd

General

Meeting,

Microbiology 150