Characterization of Streptococcus suis Genes Encoding Proteins Homologous to Sortase of Gram-Positive Bacteria

Abstract

Many surface proteins which are covalently linked to the cell wall of gram-positive bacteria have a consensus C-terminal motif, Leu-Pro-X-Thr-Gly (LPXTG). This sequence is cleaved, and the processed protein is attached to an amino group of a cross-bridge in the peptideglycan by a specific enzyme called sortase. Using the type strain of Streptococcus suis, NCTC 10234, we found five genes encoding proteins that were homologous to sortases of other bacteria and determined the nucleotide sequences of the genetic regions. One gene, designated srtA, was linked to gyrA, as were the sortase and sortase-like genes of other streptococci. Three genes, designated srtB, srtC, and srtD, were tandemly clustered in a different location, where there were three segments of directly repeated sequences of approximately 110 bp in close vicinity. The remaining gene, designated srtE, was located separately on the chromosome with a pseudogene which may encode a transposase. The deduced amino acid sequences of the five Srt proteins showed 18 to 31% identity with the sortases of Streptococcus gordonii and Staphylococcus aureus, except that SrtA of S. suis had 65% identity with that of S. gordonii. Isogenic mutants deficient for srtA, srtBCD, or srtE were generated by allelic exchanges. The protein fraction which was released from partially purified cell walls by digestion with N-acetylmuramidase was profiled by two-dimensional gel electrophoresis. More than 15 of the protein spots were missing in the profile of the srtA mutant compared with that of the parent strain, and this phenotype was completely complemented by srtA cloned from S. suis. Four genes encoding proteins corresponding to such spots were identified and sequenced. The deduced translational products of the four genes possessed the LPXTG motif in their C-terminal regions. On the other hand, the protein spots that were missing in the srtA mutant appeared in the profiles of the srtBCD and srtE mutants. These results provide evidence that the cell wall sorting system involving srtA is also present in S. suis.

Many cell wall surface proteins of gram-positive bacteria are covalently anchored to the cell wall by a mechanism requiring a C-terminal anchoring motif, which consists of a conserved amino acid sequence, Leu-Pro-X-Thr-Gly (LPXTG, where X is any amino acid), followed by a hydrophobic domain and, in most cases, a tail of positively charged residues (11, 47). A series of elegant studies by Schneewind and his colleagues (32, 35, 45, 46) have determined the function of a membrane-localized cysteine protease, named sortase, using protein A of Staphylococcus aureus as the model. Sortase specifically cleaves the LPXTG sequence in protein A between the threonine and glycine residues and catalyzes the transfer of the processed protein to the free amino group of pentaglycine cross-bridges in the staphylococcal peptidoglycan. Thus, sortase appears to be a multifunctional protease-transpeptidase (58, 59).

Computational searches of complete and preliminary genome sequence data have revealed the presence of genes homologous to srtA, a sortase-encoding gene, in a variety of gram-positive bacteria (32, 41). There is usually more than one srtA homolog in each genome (41). Although the deduced amino acid sequences of sortase-like proteins show rather low sequence similarity (less than 25% identity) (32, 41), alignment of these proteins reveals strong conservation of the sequence surrounding a cysteine residue in the predicted catalytic active site (41). Recently, Ilangovan et al. (21) proposed dividing the sortase and sortase-like proteins into two classes. Class A enzymes, including the S. aureus sortase mentioned above, contain an N-terminal segment of hydrophobic amino acids that appears to function as both a signal sequence for secretion and a stop-transfer signal for membrane anchoring. Class B enzymes contain an N-terminal signal sequence as well as a C-terminal segment of hydrophobic amino acids that appears to serve as a membrane anchor. While the presence of multiple sortase-like proteins in various bacteria suggests that the proteins may have different substrate specificities, the roles of these individual proteins have yet to be studied extensively. Further studies will be needed to clarify the function of class B proteins as well as the biological function of the sortases in other gram-positive bacteria.

Surface proteins of gram-positive pathogens promote bacterial adhesion to specific tissues, resistance to phagocytic killing, or invasion of host cells (1, 36). It has been observed that a mutation in the srtA gene of S. aureus caused defects in anchoring of such surface proteins, including several surface-associated adherence factors (31), and consequently caused a significant decrease in either colonization to the host tissue or virulence in a mouse model (31).

Recently, the srtA gene of Streptococcus gordonii was identified by probing the homologous sequence using PCR amplification with primers which were complementary to the most highly conserved regions within the srtA homologs of several other bacteria, and a mutant deficient for the srtA gene was constructed by allelic exchange. The srtA mutant of S. gordonii exhibited reductions in the surface expression of recombinant M protein, which contains the LPXTG motif, and in colonization to the oral mucosa of mice (3). These results suggest that Streptococcus spp. also possess sortases whose structure and function resemble those of S. aureus sortase. These findings prompted us to examine the srtA homologs of other Streptococcus spp. of veterinary importance.

Streptococcus suis is a gram-positive coccus that has been implicated as the cause of meningitis, septicemia, arthritis, and sudden death in pigs (6). It can also cause human meningitis (2, 28). Thirty-five capsular serotypes have so far been described (14, 15, 19, 42), and some serotypes, especially serotype 2, are more frequently isolated from diseased pigs than others (6, 17, 18). However, not all strains of S. suis serotype 2 are virulent, and there is variation in the degree of virulence of virulent strains (61, 63). Comparisons between virulent and avirulent strains of S. suis have led to the proposal of several cellular and extracellular components as candidates for virulence markers (7, 13, 16, 23, 24, 61, 62). One of these candidates, a muramidase-released protein (MRP), is a cell wall-linked protein and contains the LPXTG motif in its C-terminal region (51). However, mutant strains of S. suis with impaired expression of MRP have been shown to induce disease in pigs, and thus MRP has not been proven to specify the virulence of this bacterium (52). Although a considerable number of virulence factors of pathogenic bacteria are either secreted or located on the cell surface, no cell wall-linked proteins of S. suis other than MRP have been described to date. Besides, until now, nothing has been known about the sortase of S. suis. Therefore, we considered that a study of the sortase in S. suis would provide additional knowledge of sortase-like proteins and lead to the discovery of novel cell wall-linked proteins in this bacterium.

We found that the S. suis type strain, NCTC 10234, possessed five genes encoding proteins similar to the sortase or sortase-like proteins of other bacteria. This report describes the genetic organization of the chromosomal regions encoding these proteins. Generation of mutants deficient for the five genes in conjunction with analysis of the protein in knockout mutants enabled us to identify a genetic determinant required for cell wall sorting of proteins containing LPXTG in S. suis.

MATERIALS AND METHODS

Bacterial strains, plasmids, media, and culture conditions.

Bacterial strains and plasmids used in this study are listed in Table 1. S. suis strains were grown in Todd-Hewitt (Difco Laboratories, Detroit, Mich.) broth or agar medium supplemented with 2% yeast extract (THY) at 37°C under 5% CO₂ for 18 h. Escherichia coli strains were cultured in Luria-Bertani broth or agar medium (Difco Laboratories) at 37°C for 18 h. When necessary, antibiotics were added to culture media at the following concentrations: for E. coli, ampicillin, 50 μg/ml; kanamycin, 25 μg/ml; chloramphenicol (Cm), 10 μg/ml; and spectinomycin (Spc), 50 μg/ml; for S. suis, Cm at 5 μg/ml and Spc at 100 μg/ml.

TABLE.

Bacterial strains and plasmids used in this study

Strain or plasmid	Descriptiona	Source or reference
E. coli
DH5α	Host for cloning vector	44
XL-1 Blue MRF′	Host for phage library	Stratagene
XLOLR	Host for plasmid subclone rescued from phage library	Stratagene
S. suis
NCTC 10234	Type strain, serotype 2	NCTCb
SRTΔA1	NCTC 10234 ΔsrtA1::cat	This study
SRTΔA2	NCTC 10234 ΔsrtA2::cat	This study
SRTΔBCD1	NCTC 10234 Δ(srtBCD)1::cat	This study
SRTΔBCD2	NCTC 10234 Δ(srtBCD)2::cat	This study
SRTΔE1	NCTC 10234 ΔsrtE1::cat	This study
SRTΔE2	NCTC 10234 ΔsrtE2::cat	This study
Plasmids
pUC19	E. coli multicopy vector, Apr	65
pCR2.1	E. coli vector for cloning of PCR fragment	Invitrogen
pSET1	S. suis-E. coli shuttle vector, source of cat	55
pSET3	S. suis-E. coli shuttle vector with MCS of pUC19, Cmr Spcr	55
pSET4s	Gene replacement vector with MCS of pUC19, Spcr	56
pR326	E. coli vector, source of cat	5
pGA4C	Subclone carrying gyrA region	This study
pMO17	Subclone carrying srtE region	This study
pSAD1	pSET4s carrying cat flanked by 5′ and 3′ ends of srt A	This study
pSBD1	pSET4s carrying cat flanked by srtB 5′ end and srtD 3′ end	This study
pSED1	pSET4s carrying cat flanked by 5′ and 3′ ends of srtE	This study
pSAComp1	pSET3 carrying intact srtA	This study

^acat, chloramphenicol acetyltransferase gene; Ap^r, ampicillin resistant; Cm^r, chloramphenicol resistant; Spc^r, spectinomycin resistant; MCS, multiple cloning site.

^bNCTC, National Collection of Type Cultures, Central Public Health Laboratory, London, England.

DNA manipulations.

Restriction enzymes and DNA-modifying enzymes were purchased from Takara Shuzo Co., Ltd. (Tokyo, Japan) and used according to the manufacturer's recommendations. Minipreparations of recombinant plasmids from E. coli and transformation of E. coli were performed by standard procedures (44). Isolation of plasmids from S. suis and transformation of S. suis were carried out by the methods described previously (54, 55). Genomic DNA of S. suis was isolated by the method described previously (40). A genomic library of S. suis NCTC 10234 using a plasmid vector was constructed as follows. The genomic DNA was partially digested with Sau3AI and size fractionated by standard procedures (44). DNA fragments approximately 1 kb in size were ligated with pUC19 vector and transformed into E. coli DH5α. The S. suis NCTC 10234 genomic library which was constructed with a phage vector was described previously (55). The library was screened by plaque hybridization. Phages from the hybridizing plaques were purified, and the phagemids were rescued and used to infect E. coli XLOLR to obtain plasmid subclones as described previously (48).

Genomic Southern hybridization and plaque hybridization were performed by the procedures described previously (48), except that hybridization was carried out at 65°C. For preparation of probes, DNA fragments were labeled with digoxigenin (DIG) using either a DIG DNA labeling and detection kit or DIG-PCR labeling mixture (Roche Diagnostics GmbH, Mannheim, Germany) according to the manufacturer's instructions.

Oligonucleotides were purchased from Hokkaido System Science Co., Ltd. (Sapporo, Japan) and are listed in Table 2. PCR was performed using a Perkin-Elmer thermal cycler model 2400 or 9600 (PE Biosystems Japan, Tokyo, Japan). Takara Ex Taq polymerase and a Takara LA PCR kit were used for conventional and inverse PCR, respectively, according to the manufacturer's instructions, except that the concentration of MgCl₂ was 3 mM, as described previously (48, 55). The rationale and the protocol for inverse PCR were essentially the same as described elsewhere (38).

TABLE.

Primers used for PCR

Primer and use	Sequence (5′ - 3′)a	Location or description
Conventional PCR
gA-F	GATRTTCGWGATGGSTTRAARCCKGTTCA	Degenerate primer for gyrA
gA-R	CAGTYGMWCCATTRACCAARAGGTTTGGAA	Degenerate primer for gyrA
SRT1	GTACARGTSACYAARGTHAC	Degenerate primer for srtB
SRT2	ARGAAAGARGAYCARATGGGAGG	Degenerate primer for srtB
SRT3	ATYTAYGGWGCWGGWACNATGAA	Degenerate primer for srtC
SRT4	AGTACAGGTTAGCAAGGT	Degenerate primer for srtC
SAKO5′-F	CTATCGAGGCAGTTAAGCTGGTTATGGACA	5′ flanking region of srtA with EcoRI site
SAKO5′-R	GCATGCATGGTGTTGAATATCAATGCCAAGGA	5′ region of srtA with EcoT22I site
SAKO3′-F	GCATGCATCGGACTATTATGCTACGCAACGTA	3′ region of srtA with EcoT22I site
SAKO-3′-R	CGGGATCCGGACTATTATGCTACGCAACGTA	3′ flanking region of srtA with BamHI site
SBKO5′-F	AACTGCAGTGTAGGAGCACATACATGAGGT	5′ flanking region of srtB with PstI site
SBKO5′-R	GGATATCGAACGTCTCAGAAGTTCCTGCATA	5′ region of srtB with EcoRV site
SDKO3′-F	GGATATCAGTCAACGCTTATCGCGAATCAGA	3′ region of srtD with EcoRV site
SDKO3′-R	GGGGTACCGGTCGCATCATCGACCTTGCTA	3′ flanking region of srtD with KpnI site
SEKO5′-F	CGGATCCAATAATGTCAGAGGAGCTGGATAA	5′ flanking region of srtE with BamHI site
SEKO5′-R	GACTAGTGGTATTGTCCCCTTTTCGAGCGTGT	5′ region of srtE with SpeI site
SEKO3′-F	GACTAGTCTTATGGTGTTAATAGCCACCGAT	3′ region of srtE with SpeI site
SEKO3′-R	GGAATTCATTTCCCTGAGTTCATCATAGGTA	3′ flanking region of srtE with EcoRI site
CT1	TAGTATGCATTAATTCGATGGGTTCCGAGG	3′ end of cat with EcoT221 site
CT2	TCACATGCATCACCGAACTAGAGCTTGATG	5′ end of cat with EcoT221 site
CS1	AATCGGAAACTAGTTTCCGA	3′ end of cat with SpeI site
CS2	TCACACTAGTCACCGAACTAGAGCTTGATG	5′ end of cat with SpeI site
SAcomp-F	CATGCCATGGCCTTTGCCTTGGTAGATGCTGCCGA	5′ flanking region of srtA with NcoI site
SAcomp-R	CATGCCATGGTTATTGTCCATAATCATACTGATTATAACT	3′ flanking region of srtA with NcoI site
DPS1-D	GTWGCWCCWGAYACWACWGC	Degenerate primer for protein spot 1
DPS2-D	GARGARATHYTNACWACWCC	Degenerate primer for protein spot 2
DPS3-D	GCWGTWCARATYATGGGHGT	Degenerate primer for protein spot 3
DPS5-D	ATHWTNGCNGCNGAYACNAA	Degenerate primer for protein spot 5
T3	AATTAACCCTCACTAAAGGG	Universal primer for cloning vector
T7	GTAATACGACTCACTATAGGGC	Universal primer for cloning vector
Inverse PCR
gAinv-U	CGTACAGAATAACCGAACTTAGTCCCCAT	3′ region of gyrA (upstream)
sAinv-D	CAACCTTCCTATTTTTAAAGGGGTCTTCA	5′ region of srtA (downstream)
sA3′inv-U	ACCAGTCCTAGCTTTTCGCAGGA	5′ region of orf201 (upstream)
sA3′inv-D	CATTTCCAGAAATACCAATAAC	3′ region of orf201 (downstream)
sBinv-U	CGTGCATACTCTACCCGTCCAGCTTCATGT	Central region of srtB (upstream)
sBinv-D	CTTAACGGCTCATACAGGTTTACCCAAAGC	Central region of srtB (downstream)
sCinv-U	AGATTGGTAAATAGCGAAGCATTAG	Central region of srtC (upstream)
sCinv-D	GGACCAGATTTTAGTTGTTGAACCG	Central region of srtC (downstream)
PS1inv-U	GCCGAATTCTGCAGATATCCATCAC	PCR fragment from primer DPS1-D
PS1inv-D	CTTGCAGCAGCTAAGGCAATTTTGG	PCR fragment from primer DPS1-D
PS3inv-U	CCTTCCAATCGCGCTGTCCCCGTCA	PCR fragment from primer DPS3-D
PS3inv-D	CGTGTTCAAGCTGGGGATATGGTTG	PCR fragment from primer DPS3-D
PS5inv-U	ACGCACGTTTGAGCGGTCTCCGTTT	PCR fragment from primer DPS5-D
PS5inv-D	GGATTCCTACACATTTCAGGAGCAAA	PCR fragment from primer DPS5-D
PS2inv-U	AGCTTCCTCTGTGGTGTCAGATTC	PCR fragment from primer DPS2-D
PS2inv-D	TAGCAAAAACAGCTGTTCTTATCG	PCR fragment from primer DPS2-D

^aUnique restriction cleavage sites introduced in the oligonucleotides are underlined. R, A or G; W, A or T; S, C or G; K, G or T; Y, C or T; M, A or C; H, A or C or T; N, A or G or T or C.

DNA sequencing and analysis.

Sequencing of cloned DNA fragments and various PCR products was carried out by dye terminator chemistry with specifically designed primers using either an Applied Biosystems model 373A automated DNA sequencer (PE Biosystems) or a model RISA-384 automatic sequencer (Shimadzu Co., Ltd., Kyoto, Japan). The sequences were assembled and analyzed with Sequencher Ver. 2 (Hitachi Software Engineering Co., Ltd., Yokohama, Japan) and Genetyx-Mac Ver. 10.1 (Software Developing Company, Tokyo, Japan). Sequences were searched against current DNA databases by using either BlastX, TBlastN, or TBlastX program network services available at the National Center for Biotechnology Information, Bethesda, Md. (http://www.ncbi.nlm.nih.gov). Further DNA comparisons were made with the preliminary or complete sequence data released by genome sequencing projects at various institutions (The Institute for Genomic Research, Rockville, Md. [http://www.tigr.org/cgi-bin/BlastSearch/blast.cgi?organism=s_pneumoniae and http://www.tigr.org/tdb/s_gordonii.shtml]; University of Oklahoma, Norman [http://www.genome.ou.edu/strep.html (9) and http://www.genome.ou.edu/smutans.html]; and Sanger Centre, Cambridge, United Kingdom [http://www.sanger.ac.uk/Projects/S_equi/]).

Procedures for identification of srtA homologs.

Three independent approaches were used to identify srtA or srtA homologs in S. suis: (i) comparison of the genetic organization flanking srtA and srtA homologs of other streptococci, followed by cloning of the neighboring gene region from S. suis, (ii) probing by PCR amplification with primers designed from srtA-homologous sequences, and (iii) shotgun sample sequencing of the S. suis genomic library.

(i) Cloning of a neighboring gene.

Degenerate primers gA-F and gA-R were designed from the nucleotide sequences of the gyrA genes that were found in the database. PCR amplification was conducted with genomic DNA of S. suis NCTC 10234 as a template. The amplified fragment of approximately 400 bp was cloned into pCR2.1 and sequenced. Among 10 clones sequenced, 1 was highly homologous to gyrA of Streptococcus pneumoniae. This cloned fragment was labeled with digoxigenin and used as a hybridization probe against an S. suis phage library (55). After plaque hybridization, the phages from the hybridizing plaques were purified, and the phagemids were rescued to create plasmid subclones. A plasmid subclone containing a 3.2-kb fragment, designated pGA4C, was obtained and sequenced. An inverse PCR was employed to amplify a DNA fragment containing the downstream DNA region using primers gAinv-U and sAinv-D. Genomic DNA of S. suis NCTC 10234 was digested with SpeI and self-ligated. The ligation mixture was used as a template for the inverse PCR, and the amplified product was sequenced directly. On the basis of the sequence, another set of primers, sA3′inv-U and sA3′inv-D, were designed for inverse PCR to amplify the region further downstream. Using ClaI-digested and self-ligated genomic DNA as a template, the DNA fragment was amplified and sequenced directly.

(ii) PCR probing.

Two sets of degenerate primers, SRT1 plus SRT2 and SRT3 plus SRT4, were designed from highly conserved regions among srtA homologs of Streptococcus pyogenes, S. mutans, S. equi, and S. pneumoniae found in the database. DNA fragments amplified by PCR with genomic DNA of S. suis NCTC 10234 were cloned into pCR2.1 and sequenced. On the basis of the sequences, two sets of primers, sBinv-U plus sBinv-D and sCinv-U plus sCinv-D, were designed for inverse PCRs. Using either EcoRI- or ClaI-digested and self-ligated genomic DNA of S. suis NCTC 10234 as a template, DNA fragments were amplified and sequenced directly.

(iii) Shotgun sample sequencing.

From the S. suis genomic library constructed with a plasmid vector, approximately 300 subclones were purified and sequenced in both strands (about 600 reads). The deduced translation products of the open reading frame (ORFs) found in a shotgun sample sequencing with an average length of approximately 400 bp were searched against the databases using the TBlastN program. One sequenced fragment which was found to be part of a gene encoding a sortase-like protein was labeled with digoxigenin and used as a probe for screening the S. suis phage library (55) via plaque hybridization. After hybridization, the plasmid subclone, designated pMO17, was rescued from the positive plaques and sequenced.

Gene replacement in S. suis.

(i) Construction of knockout vectors.

DNA fragments were amplified from genomic DNA of S. suis NCTC 10234 by PCR with the primer sets SAKO5′-F plus SAKO5′-R or SAKO3′-F plus SAKO3′-R. The PCR products were digested with EcoRIand EcoT22I or EcoT22I and BamHI, fused at the EcoT22I site, and cloned between the EcoRI and BamHI sites of pUC19. The unique EcoT22I site was then used to add a Cm acetyltransferase (cat) gene, which was amplified from pR326 with the primers CT1 and CT2, followed by digestion with EcoT22I. The resulting insert was recloned between the EcoRI and BamHI sites of pSET4s, generating knockout vector pSAD1 (Fig. 1A). Similarly, DNA fragments were amplified with the primer sets SBKO5′-F plus SBKO5′-R or SDKO3′-F plus SDKO3′-R. The PCR products were digested with PstI and EcoRV or EcoRV and KpnI, fused at the EcoRV site, and cloned between the PstI and KpnI sites of pUC19. The unique EcoRV site was used to add a cat gene, which was cut out from pR326 with EcoRI and HindIII, followed by filling-in with Klenow enzyme. The resulting insert was cut out with PstI and XhoI and recloned between the PstI and SalI sites of pSET4s, generating knockout vector pSBD1 (Fig. 1B). DNA fragments were amplified with primers SEKO5′-F plus SEKO5′-R or SEKO3′-F plus SEKO3′-R. The PCR products were digested with BamHI and SpeI or SpeI and EcoRI, fused at the SpeI site, and cloned between the BamHI and EcoRI sites of pUC19. The unique SpeI site was used to add a cat gene, which was amplified from pSET1 with the primers CS1 and CS2, followed by digestion with SpeI. The resulting insert was recloned between the BamHI and EcoRI sites of pSET4s, generating knockout vector pSED1 (Fig. 1C).

FIG. 1.

Physical and genetic map of srtA (A), srtBCD (B), and srtE (C) loci in NCTC 10234 and construction of the mutant strains. orf204 contains a contiguous region of sntA. The wild-type srtA, srtBCD, and srtE were replaced by the ΔsrtA::cat, ΔsrtBCD::cat, and ΔsrtE::cat alleles carried on the knockout vectors pSAD1, pSBD1, and pSED1, respectively, as described in Materials and Methods, to generate the mutants. Only restriction endonuclease sites used for inverse PCR, construction of the plasmid, or genomic Southern hybridization are indicated. Black arrows, srtA homologs; gray arrows, other genes in the srtA-homologous regions; open boxes, directly repeated segments; tnp, a pseudogene potentially encoding a transposase; gradated arrow, cat gene; B, BamHI; C, ClaI; E, EcoRI; H, HindIII; N, NcoI; RV, EcoRV; S, SpeI; Sal, SalI; T22, EcoT22I; X, XhoI.

(ii) Generation of mutants.

Procedures for selection of mutants whose genes were replaced by allelic exchange via double crossover were described previously (56). Briefly, S. suis NCTC 10234 was transformed with a knockout vector, and the cells were grown at 28°C in the presence of Cm and Spc. The cells at mid-logarithmic growth phase were diluted with THY broth containing Cm and grown at 28°C to early logarithmic phase. The cultures were then shifted to 37°C and incubated for 4 h. Subsequently, the cells were spread on THY agar containing Cm and incubated at 37°C. Temperature-resistant Cm-resistant (Cm^r) colonies obtained were screened for loss of vector-mediated Spc resistance (Spc^r) to detect putative mutants which had exchanged their wild-type allele for a genetic segment containing the cat gene as a consequence of homologous recombination via a double crossover. Finally, the genetic organization of the mutant alleles was analyzed by Southern hybridization.

(iii) Recombinant plasmid for complementation analysis.

A DNA fragment was amplified from genomic DNA of S. suis NCTC 10234 by PCR with the primers SAcomp-F plus SAcomp-R. The fragment was digested with NcoI and cloned into the NcoI site of pSET3, which was located in the cat gene of the vector, generating recombinant plasmid pSAcomp1.

Partial purification of cell wall proteins.

S. suis was grown in 200 ml of THY broth to an optical density at 600 nm of approximately 0.6. After centrifugation at 8,000 × g for 15 min at 4°C, the cell pellet was washed twice with 100 ml of 50 mM Tris-HCl (pH 7.5) (Tris buffer) and suspended in 12.5 ml of Tris buffer containing DNase (50 μg/ml) and RNase (50 μg/ml). Complete Protease Inhibitor Cocktail (Roche Diagnostics) was added to this suspension according to the manufacturer's instructions. The sample was then sonicated on ice with a Sonifier 250 (Branson Ultrasonics Corp., Danbury, Conn.) with a microtip. One minute of sonication with a 1-min interval was repeated 15 times with the instrument parameters set at 50% duty cycle and an output of 6. After sonication, unbroken cells were removed by centrifugation three times at 2,000 × g for 10 min at 4°C. Disrupted cell materials in the supernatant were pelleted by centrifugation at 50,000 × g for 30 min at 4°C and washed under the same conditions three times each with 10 ml of Tris buffer and Tris buffer containing 1 M NaCl, successively. The pellet was then suspended in 10 ml of Tris buffer containing 1% (vol/vol) Triton X-100 and incubated at 37°C for 1 h with gentle rotation. The Triton-insoluble materials were then pelleted by centrifugation at 100,000 × g for 30 min at 20°C. The pellet was suspended in 10 ml of 4 M guanidine hydrochloride and incubated at 37°C for 1 h with gentle rotation. The guanidine-insoluble materials were then pelleted by centrifugation at 100,000 × g for 30 min at 20°C, washed under the same conditions three times with 10 ml of distilled water, suspended in 3 ml of distilled water, and stored at −80°C until use.

Proteins were released from the guanidine-insoluble materials by incubation in a solution containing 1 mM Tris-HCl (pH 7.5), 50 μM MgCl₂, and N-acetylmuramidase SG (30 μg/ml; Seikagaku Kogyo, Tokyo, Japan) at 37°C for 1 h with gentle rotation. After centrifugation at 50,000 × g for 30 min at 4°C, the supernatants were collected and the protein concentrations were determined using a protein assay kit (BCA-200; Pierce Chemical Co., Rockford, Ill.). The samples were then lyophilized and stored at 4°C until use.

2D-PAGE and protein sequencing.

Equivalent amounts of the lyophilized samples (75 μg) were resolved directly in 30 μl of rehydration buffer (9 M urea, 2% [vol/vol] NP-40, 2% [vol/vol] 2-mercaptoethanol, 2% [vol/vol] Ampholine pH 3.5-10 [Amersham Pharmacia Biotech UK Ltd., Buckinghamshire, England]). Two-dimensional polyacrylamide gel electrophoresis (2D-PAGE) was performed according to the method of O'Farrell et al. (39) as follows. In the first dimension, nonequilibrium pH gradient electrophoresis was carried out in a apparatus (AE-6050A; Atto Co., Tokyo, Japan) with rectangular gels containing 3.6% acrylamide, 9.5 M urea, 2% (vol/vol) NP-40, and 2% Ampholine (pH 3.5-5 and pH 3.5-10 mixed at a ratio of 1:1; Amersham Pharmacia). The 30-μl sample was loaded at the acidic end of the gel and electrophoresed for 1,600 Vh. After completion of the run, gels were incubated in equilibration buffer (0.125 M Tris-HCl [pH 6.8], 5% [vol/vol] 2-mercaptoethanol, 2% [wt/vol] sodium dodecyl sulfate [SDS], 0.01% bromophenol blue) for 10 min or stored frozen in 10% (vol/vol) glycerol at −80°C.

In the second dimension, polyacrylamide gels with a ratio of 10% acrylamide monomer to 1% bisacrylamide (Bio-Rad Laboratories, Hercules, Calif.) containing 0.1% (wt/vol) SDS, 0.37 M Tris base, and 0.27 M Tris-HCl were used. The sample gels were placed on top of the second-dimension gels by submerging the gels in a warm gel overlay agarose solution (0.5% [wt/vol] agarose [SeaPlaque]; FMC, Rockland, Maine) dissolved in gel buffer. Electrophoresis in the second dimension was performed at 12.5 mA per gel for about 2 h, until the dye front reached the bottom of the gel.

After electrophoresis, protein spots were visualized by silver staining using a silver stain kit (Dai-ichi Kagaku Co., Tokyo, Japan). Based on the protein migration on the gel, protein spots were numbered. For protein sequencing, samples in a 2D-PAGE gel were electroblotted onto a polyvinylidene difluoride membrane (Problott; PE Biosystems) in 10 mM CAPS (N-cyclohexyl-3-aminopropanesulfonic acid) and 10% methanol (pH 11.0). The sheet was washed with 40% methanol-1% acetic acid, stained with 0.1% Coomassie brilliant blue R-250 in 40% methanol and 1% acetic acid, and destained with 50% methanol aqueous solution. The stained spots chosen for protein sequencing were cut and collected from several membranes and subjected to sequential degradation using an automated peptide sequencer (ABI model 477A; PE Biosystems).

Procedures for sequencing of genes encoding proteins corresponding to selected spots.

On the basis of the N-terminal amino acid sequences of protein spots, four degenerate primers, DPS1-D, DPS2-D, DPS3-D, and DPS5-D, were designed. DNA fragments amplified by PCRs with one of the degenerate primers plus either T7 or T3 primer using the S. suis phage library (55) as a template were cloned in pCR2.1 and sequenced to confirm that the fragments contained genetic segments corresponding to the protein spots. From the sequences of the fragments, four sets of primers, PS1inv-U plus PS1inv-D, PS2inv-U plus PS2inv-D, PS3inv-U plus PS3inv-D, and PS5inv-U plus PS5inv-D were designed for inverse PCRs. Using either EcoRI-, ClaI-, or SpeI-digested, self-ligated genomic DNA of S. suis as a template, DNA fragments were amplified and sequenced directly.

Nucleotide sequence accession numbers.

The nucleotide sequences determined in this study have been deposited in the DDBJ/EMBL/GenBank database under accession nos. AB066353, AB066354, AB066355, AB066356, and AB066357.

RESULTS

Identification of srtA.

We previously reported that several genes and their neighboring regions which had already been cloned and sequenced from S. suis showed the same or similar genetic organizations compared with the corresponding regions of other streptococci (40, 48). Comparison of the genetic organization of srtA and srtA homologs among S. gordonii, S. pyogenes, S. pneumoniae, S. mutans, and S. equi downloaded from the databases revealed that at least one of the srtA or srtA homologs was located adjacent to gyrA, a housekeeping gene encoding DNA gyrase A subunit. A plasmid subclone, pGA4C, containing the gyrA region of S. suis was therefore obtained and sequenced.

The region contained the entire length of the putative gyrA gene (84% identity to gyrA of S. pneumoniae [AF170993]) (8) and one incomplete ORF starting 7 bp downstream of gyrA. The deduced amino acid sequence of the ORF was found to be homologous to the sortase of S. gordonii (described in detail below), and the gene was tentatively designated srtA. A DNA fragment containing the downstream DNA region was amplified by two consecutive inverse PCRs and sequenced directly. The entire DNA sequence determined was 5,490 bp and contained six putative ORFs (Fig. 1A). srtA was preceded by a typical Shine-Dalgarno (SD) sequence, which overlapped the 3′ end of gyrA and was located 15 bp upstream of the putative ATG start codon.

The region downstream of srtA contained three ORFs, one on the same DNA strand as srtA and two on the complementary strand (Fig. 1A). The first ORF encoded a protein of 226 amino acids which showed 59% identity with a putative DNA repair protein of S. pyogenes (AE006355) (9), and the gene was designated radC. The deduced amino acid sequences of the second and third ORFs, designated ORF201 (230 amino acids) and ORF202 (139 amino acids) (Fig. 1A), showed 47% identity with S. pyogenes hypothetical protein Spy1119 (AE006554) and 74% identity with Spy1120 (AE006544), respectively. The genetic order radC-orf201-orf202 in S. suis was the same as in S. pyogenes, i.e., radC-spy1119-spy1120, although the genetic segment was located apart from the gyrA locus of S. pyogenes (9).

Identification of srtBCD.

Two DNA fragments obtained by PCR probing were found to overlap, and the entire DNA sequence determined was 5,606 bp. The sequence contained five putative ORFs (Fig. 1B). Three genes encoding sortase-like proteins were found in this region, and the genes were designated srtB, srtC, and srtD. The three srt genes overlapped each other in a head-to-tail manner, with a typical SD sequence located 6 to 12 bp upstream of a putative start codon of each gene. The remaining ORFs, which were located upstream and downstream of srtBCD, were designated ORF203 and ORF204, respectively. ORF203 was encoded on the same DNA strand as srtBCD, whereas ORF204 was on the complementary strand. Thus, the genes identified in this region were in the order orf203-srtB-srtC-srtD-orf204, as shown in Fig. 1B.

ORF203 encoded a protein of 284 amino acids which showed 29% identity with a hypothetical protein of Lactococcus lactis subsp. lactis (AE006252). This protein contained a putative N-terminal signal sequence and a putative C-terminal structure of membrane-spanning hydrophobic residues followed by a positively charged tail. The C-terminal end of the ORF203 protein was preceded by an IPYTG sequence, which is similar to the LPXTG motif, with substitution of Leu for Ile. ORF204 encoded a protein of 396 amino acids, of which the N-terminal end was truncated. The protein showed a typical cell wall-anchoring structure, LPATG, at the C-terminal region and was 64% identical with a putative 5′-nucleotidase of S. dysgalactiae subsp. equisimilis (AJ133440).

Furthermore, analysis of the DNA sequence revealed the presence of three segments of approximately 110-bp direct repeats with some mismatches (Fig. 1B). One, designated DR1, started 165 bp upstream of srtB and the remaining two, designated DR2 and DR3, started 79 and 375 bp downstream of srtD, respectively. The repeated DNA segments found in this region were similar to a repeated DNA element which had been found in intergenic regions of the genome sequence of S. pneumoniae (25, 27, 29, 57). The repeat unit found in S. pneumoniae consists of various combinations of three subunits, named boxA, boxB, and boxC, and is scattered throughout the genome of the organism. The repetitive moieties of S. suis consisted of combined segments that were organized in tandem repeats of two subunits; one was similar to boxA, and the other was similar to boxC.

Identification of srtE.

The sequence of pMO17 determined was 4,991 bp and contained six putative ORFs, as shown in Fig. 1C. One gene encoding a sortase-like protein was designated srtE. The srtE gene was preceded by a putative SD sequence 20 bp upstream of the putative start codon. Two ORFs located upstream of the srtE gene potentially encoded proteins of 202 and 128 amino acids. The former showed 40% identity with type I signal peptidase B of S. aureus (U65000), and the gene was designated spsB, whereas the latter showed no significant homology to proteins in the database and was designated ORF205 (Fig. 1C).

The ORF just downstream of srtE potentially encodes a protein of 138 amino acids, which showed 49% identity with a mercury resistance operon-regulatory protein of B. subtilis (X92868), and the gene was designated merR. The two remaining ORFs, designated ORF206 and ORF207, potentially encode proteins of 103 and 277 amino acids, respectively. These two proteins showed homology to S. pyogenes hypothetical proteins Spy0589 (39% identity [AE006514]) and Spy0590 (53% identity [AE006514]), respectively, which were also colocalized in S. pyogenes (9).

The six genes found in this region, except orf206, were located on the same DNA strand, and the genes identified in this region were in the order spsB-orf205-srtE-merR-orf206-orf207, as shown in Fig. 1C. There were relatively long intergenic spaces comprising 355 and 382 bp in the 5′- and 3′-flanking regions, respectively, of srtE, where no ORF could be found. However, homology searches of these DNA sequences compared to the databases using the TBlastN program revealed that the intergenic sequence in the 5′-flanking region of srtE had 75% homology with the 3′ end of a transposase gene in IS 1515 described for S. pneumoniae (34). This gene was located on the cDNA strand to srtE and potentially starts from the 5′ end of srtE with a 30-bp overlap (Fig. 1C). Comparison of the sequences showed that nucleotide substitutions and deletions were found in the sequence of S. suis, resulting in gene inactivation by a frameshift. Thus, this intergenic region constitutes a pseudogene. No notable feature was found in another intergenic region at the 3′ end of srtE.

Amino acid sequence analysis of srt gene products.

The predicted products of the five srt genes and the homology of their potential protein products with two representative sortases of other bacteria are summarized in Table 3. SrtA of S. suis possessed an N-terminal segment of hydrophobic amino acids that would function as a signal sequence for secretion, whereas the other four Srt proteins of S. suis were characterized by both the N-terminal signal sequence and a C-terminal segment of hydrophobic amino acids. On the basis of the criterion proposed by Ilangovan et al. (21) that the hydrophobic membrane anchoring domain appears in the C terminus of the sortase-like proteins, SrtA of S. suis was considered to be a class A enzyme and the other four Srt proteins to be class B enzymes.

TABLE.

Predicted gene products of five srt genes and their amino acid identity with sortases of other bacteria

Gene	Product size		Predicted pI	Signal sequencea	Hydrophobic residuesa	% Amino acid identityb with product from:
	Amino acids	kDa				S. gordonii	S. aureus
srtA	249	27.9	5.43	+	−	65/244	23/182
srtB	295	33.8	9.48	+	+	23/128	26/171
srtC	285	32.8	8.96	+	+	23/252	31/121
srtD	273	31.8	9.62	+	+	18/197	31/115
srtE	275	30.4	5.07	+	+	25/217	25/118

^aSignal sequence, N-terminal signal sequence for secretion; hydrophobic residues, C-terminal hydrophobic region; +, present; −, absent.

^bPercent identity/number of amino acids evaluated.

From the results of homology searches, all these Srt proteins generally exhibited a low level of homology with the sortases of S. gordonii and S. aureus, except that a high level of homology was noted between SrtA of S. suis and the sortase of S. gordonii (Table 3). Furthermore, a high level of sequence identity throughout almost the whole length of the proteins, ranging from 55 to 84%, was also noted when these two proteins and other streptococcal sortase-like proteins whose genes were linked to gyrA were compared. The amino acid sequences of the five Srt proteins were aligned with those of two representative sortases. As shown in Fig. 2, the predicted catalytic active sites, i.e., the His₁₄₁ residue and the region preceding the Cys₂₀₆ residue in SrtA of S. suis, were conserved in all five proteins, as has been reported for sortase-like proteins of other bacteria.

FIG. 2.

Sequence alignment of SrtA, SrtB, SrtC, SrtD, and SrtE of S. suis as well as SrtA of S. aureus and S. gordonii. The presumed C-terminal hydrophobic segments of class B enzymes are underlined. Black background, amino acids identical among all the sequences aligned; gray background, amino acids functionally conserved among the sequences aligned; dashes, gaps in the aligned sequences.

Generation of knockout mutants.

To generate isogenic mutants in which one of the srt gene loci was replaced by a cat cassette, we constructed three knockout plasmids, pSAD1, pSBD1, and pSED1, as described in Materials and Methods. pSAD1 had a ΔsrtA::cat allele composed of a fragment containing the first 102 nucleotides of srtA and part of gyrA located upstream of cat and a fragment containing the last 131 bp of srtA and radC as well as orf201, located downstream of cat, which was sufficient for homologous recombination to occur (Fig. 1A). pSBD1 had a ΔsrtBCD::cat allele composed of a fragment containing the first 236 nucleotides of srtB and orf203 located upstream of cat and a fragment containing the last 182 bp of srtD and part of orf204, located downstream of cat (Fig. 1B). pSED1 had a ΔsrtE::cat allele composed of a fragment containing the first 227 nucleotides of srtE and spsB as well as orf205 located upstream of cat and a fragment containing the last 191 nucleotides of srtE and merR, orf206, and part of orf207 located downstream of cat (Fig. 1C).

Using the three knockout vectors, we generated Cm^r Spc^s mutants. Two independently derived mutants from each knockout vector, designated SRTΔA1 and SRTΔA2 from pSAD1, SRTΔBCD1 and SRTΔBCD2 from pSBD1, and SRTΔE1 and SRTΔE2 from pSED1, were chosen for further study. Genomic DNA was isolated from these mutants, digested with NcoI, and examined by Southern hybridization using cat and either srtA, srtBCD, or srtE region-specific probes. The results showed that the srtA, srtBCD, and srtE probes hybridized with fragments of the expected sizes, and these fragments also hybridized with the cat probe, which did not hybridize with the DNA from the wild-type strain, NCTC 10234 (data not shown). On the basis of these results, we concluded that those mutants were chromosomal mutants in which one of the srt gene loci had been replaced by the cat gene cassette and were derived from allelic exchange via double crossover (Fig. 1A, 1B, and 1C).

Role of srt genes in sorting of cell wall proteins.

Cell wall materials of the knockout mutant strains as well as their parent strain, NCTC 10234, were prepared, and guanidine-insoluble muramidase-released proteins of the cell wall materials were analyzed by 2D-PAGE. Representative protein profiles are shown in Fig. 3. The protein profiles of SRTΔA1 and SRTΔA2 showed a dramatic change, especially; more than 15 protein spots disappeared compared with the gel of the parental strain (Fig. 3A and B). These results suggest that srtA may be involved in the processing of proteins to the cell wall in S. suis NCTC 10234.

FIG. 3.

2D-PAGE profiles of guanidine-insoluble muramidase-released proteins from cell wall materials of mutant and parental strains of S. suis. (A) Parental strain NCTC 10234, (B) SRTΔA1, (C) SRTΔBCD1, (D) SRTΔE1, (E) SRTΔA1(pSAComp1). Equivalent amounts of lyophilized samples resolved in rehydration buffer were loaded on each gel. After electrophoresis, gels were silver stained. Protein spots marked by the rectangles were subjected to N-terminal amino acid sequence determination. The following molecular weight standards were used in the second dimension: myosin (200,000), β-galactosidase (116,300), phosphorylase b (97,400), bovine serum albumin (66,300), glutamate dehydrogenase (55,400), lactate dehydrogenase (36,500), carbonic anhydrase (31,000), and trypsin inhibitor (21,500).

To confirm these observations, we performed complementation analysis. Since the SD sequence of srtA found in this study overlapped gyrA, as described above, it was possible that srtA was transcribed from the promoter of gyrA. Therefore, a fragment of srtA including its SD sequence was placed under the control of the cat gene promoter in the recombinant plasmid pSAcomp1. SRTΔA1 was transformed with pSAcomp1, and the protein profile of the transformant was analyzed in the same manner. As shown in Fig. 3E, the mutant phenotype of SRTΔA1 was completely complemented by the cloned srtA. On the other hand, the protein spots which disappeared in SRTΔA1 and SRTΔA2 appeared in the protein profiles of SRTΔBCD1, SRTΔBCD2, SRTΔE1, and SRTΔE2. (Fig. 3C and D). These results suggest that srtBCD and srtE did not mediate the processing of those proteins.

Identification of genes encoding the protein spots.

From the 2D-PAGE profile, eight protein spots, spots 1 through 8, which were absent from the profiles of SRTΔA1 and SRTΔA2 were selected and subjected to N-terminal amino acid sequence determination. On the basis of the N-terminal amino acid sequences, four putative genes could be identified, and the nucleotide sequences were determined. The putative genes found in the sequences were named according to their closest relative, probable function, or the original gene designation described elsewhere. The results are summarized in Table 4.

TABLE.

Predicted genes corresponding to selected protein spots and their deduced translational products

Gene	Predicted translational product			LPXTG motif	Description and function of closest relativeb	Spot no.	N-terminal amino acid sequencec
	Amino acids	kDaa	pIa
mrp	1,256	130.6	4.61	LPNTGE	Muramidase-released protein of S. suis [X64450] (100/1,256)	8	48-DETVASSE
						1	183-ALDTVAPDTTA
sntA	813	85.4	4.44	LPATGE	Cyclo-nucleotide phosphodiesterase of S. dysgalactiae subsp. equisimilis [AJ133440] (55/666)	2	28-EEILNTTPAS
sntB	674	69.7	4.48	LPNTGQ	5′-Nucleotidase of S. pyogenes [AE006537] (58/675)	3	29-DELAVQIMGVNDFHGAL
sntC	724	74.0	4.31	LPNTGD	5′-Nucleotidase of S. aureus N315 [AP003129] (39/724)	4	28-AETTTAATTTNQPAT
						5	347-SDKLLGEASLISAADTKNVTPNAKIAAL
						6	373-ALVDEIKAKYEAENA
						7	380-AKYEAENAQVVIENN

^aValues are calculated on the basis of the sequence excluding the N-terminal signal sequence or starting methionine residue.

^bThe accession numbers in the DDBJ/EMBL/GenBank database are shown in brackets; numbers in parentheses are percent identity/number of amino acids evaluated.

^cNumbers indicate the amino acid at which the sequence started.

The deduced protein sequences of the mrp, sntA, sntB, and sntC genes possessed an N-terminal signal sequence and a typical C-terminal cell wall-anchoring structure, LPXTG, followed by a hydrophobic domain consisting of about 20 amino acids and a tail of positively charged amino acids. mrp, which encoded MRP, is a known gene, as described in the introduction (51). Among the N-terminal sequences of the eight protein spots, spots 8, 2, 3, and 4 coincided with the N-terminal region of the deduced translational products, excluding the putative signal sequence, of mrp, sntA, sntB, and sntC, respectively (Table 4). The molecular sizes and isoelectric points estimated from the 2D-PAGE profiles of spots 2, 3, and 4 were in good agreement with the values calculated for the putative translational products of sntA, sntB, and sntC, respectively. The estimated sizes of spots 1 and 5 through 8 were smaller than those deduced from the gene sequences, and the N-terminal sequences of spots 1, 5, 6, and 7 coincided with the translated internal amino acid sequences of the corresponding genes (Table 4). These results suggest that proteolytic degradations had occurred due to residual proteases present in the cell wall materials in the course of partial purification, presumably during the process of N-acetylmuramidase digestion.

sntA encodes a 5′-nucleotidase and contains a contiguous region of orf204, which is located downstream of the srtBCD cluster, as described above. Interestingly, an Arg-Gly-Asp (RGD) sequence, a tripeptide motif that is commonly recognized by integrin receptors of mammalian cells (4, 20), was found at amino acid positions 505 to 507 of the SntA protein. Repetition of a short peptide segment was seen in the N-terminal region of SntC, but not in those of SntA or SntB.

DISCUSSION

Although Bolken et al. successfully amplified the srtA gene from S. gordonii by PCR (3), their PCR primers did not amplify a specific DNA fragment from S. suis genomic DNA in our preliminary experiments (unpublished observations). Many sortase-like proteins show a limited degree of sequence similarity, and many bacteria possess two or more srtA homologs (32, 41). Therefore, we adopted the three independent experimental approaches described above to find srtA homologs in S. suis. Our findings demonstrated the presence of five srtA homologs in S. suis NCTC 10234. The number of srtA homologs found in this strain was more than the number described for other streptococci, i.e., four in S. pyogenes and S. pneumoniae and one in S. equi and S. mutans (21, 41). However, there have been estimates of six or more srtA homologs in some bacteria (41), raising the possibility that the S. suis strain possesses an additional srtA homolog(s). This possibility is also supported by the fact that we found one srtA homolog, named srtE, from sample sequence data for only a limited number of shotgun clones (representing about 1% of the whole genome). On the other hand, the number of srtA homologs and their distribution may vary among S. suis strains of diverse origins. Therefore, further studies are needed to clarify the prevalence of the five srtA homologs in other S. suis strains.

In this study, the genetic structures of srtA, srtBCD, and srtE in S. suis and their neighboring regions indicated that the orders of the genes within a limited chromosomal segment were sometimes conserved among some Streptococcus species, although their overall chromosomal organizations might be quite different. These findings implied that it might be possible to identify novel genes by means of gene walking if the genes of interest are found to be linked to a specific gene, such as conserved housekeeping genes, in the chromosome of the counterpart whose genome sequencing has been completed or is being investigated. In addition to our previous observations (40, 48), we identified additional conserved genetic segments which were exemplified by a gyrA-srtA segment, a radC-orf201-orf202 segment in the srtA region, and an orf206-orf207 segment in the srtE region. Among them, the most notable one was the gyrA-srtA segment because the linkage between gyrA, an essential gene in prokaryotes, and srtA (or srtA homologs) was conserved among at least six Streptococcus species. As discussed below, it is evident that srtA in S. suis is responsible for processing many cell wall proteins; therefore, it should be a priority to examine the function of the gyrA-linked srtA homologs in other streptococci.

In contrast to srtA in S. suis, direct linkages of the neighboring genes relative to srtBCD or srtE were not observed in other Streptococcus species. However, structural features of interest were found close to srtBCD and srtE, i.e., directly repeated segments in the former and a transposase-like gene in the latter. Redundancy of the box segments may reflect genetic plasticity. Recent studies of the complete sequences of bacterial genomes have revealed that long target duplications are often found at the junction regions where an exogenous segment is apparently incorporated; for example, in the vicinity of restriction-modification genes (37, 64). Other extragenic repeated sequences found in diverse bacterial species are also known to play a role in recombination events (50). Taking all these observations together, we can predict that the genetic configuration of the srtBCD region represents a consequence of gene rearrangement. On the other hand, a genetic segment found in the intergenic region of srtE was apparently a piece of a genetic segment encoding a transposase that was originally described as a functional gene in S. pneumoniae (34). This finding may suggest that a genetic conversion driven by the transposase took place, although the pseudogene contained DNA only approximately half the size of the entire transposase gene and overlapped srtE. From the present results, however, we could not provide any evidence implying that srtBCD and srtE were acquired from other bacteria by lateral transfer. Regardless of whether the genetic configurations found in the srtBCD and srtE regions were derived from other bacteria or are the outcome of genetic conversion within the S. suis genome, it will be interesting to determine whether the genetic structures are preserved in other S. suis strains.

The amino acid sequences of the SrtA through SrtE proteins in S. suis show conserved amino acid residues in their core regions, as was reported for other sortase-like proteins (21, 32, 41). However, one structural feature in the N-terminal region distinguished the SrtA protein from the others, i.e., only the SrtA of S. suis is a class A enzyme, like the sortases of S. aureus and S. gordonii. While sortases generally show low sequence identity, SrtA of S. suis was highly homologous to the sortase of S. gordonii and other streptococcal sortase-like proteins whose genes are linked to gyrA. This suggests that these proteins have the same biological function as the sortase of S. gordonii.

Functional analyses of srtA have so far been carried out using the staphylococccal protein A system in S. aureus or a recombinant streptococcal M protein system in S. gordonii. However, the lack of such a well-established foreign gene expression system in S. suis led us to perform a partial purification of cell wall materials, followed by 2D-PAGE analysis of guanidine-insoluble muramidase-released proteins to assess the functions of the five srt genes. The 2D-PAGE profile was expected to provide a general picture of proteins which are covalently linked to the cell wall, because guanidine hydrochloride was used as a chaotrophic agent to remove most of the noncovalently associated proteins from the cell wall (26). However, miscellaneous contaminants, some of which would not be covalently linked, could also appear, since it is impossible to remove them completely through partial purification. Nevertheless, we showed that a genetically defined srtA mutant was deficient in displaying more than 15 major spots in the 2D-PAGE profile and that this phenotype was completely complemented by the srtA cloned from S. suis, indicating that the deficiency in protein sorting was caused by the disruption of srtA.

On the other hand, no significant difference was found in the 2D-PAGE profiles of srtBCD and srtE mutants. In particular, the spots that were missing in the profile of the srtA mutant were present in those of the srtBCD and srtE mutants. This was consistent with a recently reported observation that another srtA homolog in S. aureus, named srtB, did not function in sorting of protein A, FnbA, FnbB, or ClfA (33). However, it is important to note that not all cell wall-linked proteins appeared in the 2D-PAGE profiles. Proteins with isoelectric points outside the range used would not appear and/or there must be a number of proteins present in amounts not sufficient to be visible. Since it is also conceivable that srtBCD and srtE catalyze similar reactions using different surface protein substrates that are as yet unknown minor proteins, we cannot rule out the possibility that srtBCD and srtE are involved in the processing of cell wall proteins. Although the functions of srtBCD and srtE have yet to be identified, studies of the genetically defined srtBCD and srtE mutants may provide a better understanding of the activities of these enzymes.

N-terminal amino acid sequencing of eight protein spots that were missing in the srtA mutant enabled us to identify four genes that may encode these proteins. However, the N-terminal sequences determined were quite limited in length. Therefore, further analysis, e.g., determination of the internal amino acid sequences of those spots, may be required to confirm the identity of the genes. Whether the protein spots, especially the three novel proteins that we found, are in fact linked to the cell wall and exposed on the cell surface should also be clarified. Nevertheless, the N-terminal sequences of eight protein spots which were definitely associated with srtA coincided with the internal sequences of proteins encoded by the four genes identified. Moreover, the proteins encoded by the four genes contained the LPXTG motif in their C-terminal regions. Therefore, the simplest and most likely explanation for these results is that sorting of the cell wall proteins containing the LPXTG motif requires the function of srtA in S. suis; that is, that srtA encodes a sortase.

Among three novel cell wall proteins containing the LPXTG motif found in this study, SntA possessed another amino acid motif, a tripeptide RGD. The RGD motif has been shown to be biologically important, as it binds to integrins of mammalian cells. Integrins are heterodimeric membrane proteins located on the surface of mammalian cells that participate in cell-to-cell adhesion and cellular differentiation, migration, and attachment to the extracellular matrix (20). The three-amino-acid sequence RGD is critical for ligand recognition by many integrins (4). A variety of microbial pathogens bind integrins, usually through an RGD motif located on cell surface proteins. Several of these proteins are proven virulence factors (10, 12, 22, 30, 43, 53, 63). We have not examined the function or the molecular structure of the SntA protein. However, because the SntA protein was present in the cell wall fraction and SntA contained a cell wall-sorting signal represented by the LPXTG motif, it is possible that the SntA protein is located on the bacterial cell surface. This leads to the notion that the RGD motif found in the SntA protein may directly bind integrin and play an important role in host-pathogen interactions; that is, SntA may represent one of the virulence factors of S. suis.

Recently, computational searches of complete or preliminary genome sequences have made it possible to identify novel cell wall-linked proteins of some bacteria, and several attempts to characterize their functions have been carried out (9, 57). However, this strategy cannot be used for a bacterium, such as S. suis, about which we have limited genome information. In this regard, the experimental procedures described in this study will constitute a system for discovering novel cell wall-linked proteins, some of which may act as virulence factors. The method can be used not only for S. suis but also for other gram-positive bacteria, especially those whose genome sequences have not been completely determined.