| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 Channing Laboratory, Department of Medicine, Brigham and Women's Hospital, Harvard Medical School, Boston, MA 02115, USA
2 Departments of Laboratory Medicine and Pathology and Pediatrics, University of Minnesota Medical School, Minneapolis, MN 55455, USA
3 The Institute for Genomic Research, 9712 Medical Center Drive, Rockville, MD 20850, USA
Correspondence
Lawrence C. Paoletti
lpaoletti{at}channing.harvard.edu
Received 19 July 2005
Accepted 7 February 2006
Abbreviations: CPS, capsular polysaccharide; GBS, Group B Streptococcus; MLST, multilocus sequence typing; NT, nontypable; ST, sequence type.
The GenBank/EMBL/DDBJ accession numbers for the sequences reported in this paper are DQ359707–DQ359714.
 |
INTRODUCTION |
|---|
 TOP INTRODUCTION METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Capsular polysaccharides (CPSs) are major virulence factors and the main components of glycoconjugate vaccines. In addition, these surface carbohydrates provide useful epidemiologic markers, as GBS isolates are serotyped on the basis of their CPS antigens; nine distinct serotypes (Ia, Ib, and II–VIII) have thus far been described. The prevalence of GBS serotypes fluctuates over time and by geographic location (Le Thomas-Bories et al., 2001). Although serotype III strains are responsible for the majority of cases of neonatal meningitis worldwide, serotype V strains are isolated with increased frequency from all age groups within a population, and an estimated 24–31 % of the cases of invasive GBS disease in non-pregnant adults in the USA are caused by this serotype (Blumberg et al., 1996; Farley, 2001; Le Thomas-Bories et al., 2001).
Isolates without identifiable CPS by the standard Lancefield serotyping method are considered nontypable (NT). NT isolates may account for up to 4 % of all GBS isolates (Ferrieri et al., 1997), but account for up to 12 % in collections studied in some laboratories (Palacios et al., 1997). The extent of sequence variation in the cps genes of clinical isolates within a serotype, and among different serotypes, has not been well studied. It has been observed that closely and distantly related clones of GBS may share the genes encoding a specific CPS, suggesting that capsule switching may occur in GBS (Davies et al., 2004).
Most NT isolates express surface proteins: alpha and beta components of the C protein or one or more of the R protein antigens, Rib, and Alp1–4 proteins (Benson et al., 2002; Kong et al., 2002b). These isolates can be genotyped and distinguished by PFGE and by the expressed surface-protein antigens (Ferrieri et al., 1997; Ferrieri & Flores, 1997; Sellin et al., 2000). The C proteins were recognized by Lancefield as protective determinants found in GBS of serotypes Ia and Ib (reviewed by Lindahl et al., 2005). One of the C proteins, the beta C protein, binds to human IgA and is found mainly in serotype Ib GBS. The other C protein, the alpha C protein, has been recognized to be the prototype for a family of proteins (alpha-like proteins) that is found in most GBS and includes Rib and Alp1–4. The protein Rib is most frequently associated with type II and III GBS, while Alp1 is found frequently in serotype Ia GBS, and Alp3 most frequently in serotype V. R proteins (R1–4) have been defined immunologically on the basis of cross-reactivity among streptococcal strains, but are less well defined at the genetic level (Flores & Ferrieri, 1989), except that the gene for the R4 protein has been cloned and is identical to the Rib gene (Smith et al., 2004). The other R proteins appear to be closely related or identical to members of the Alp family. While certain surface proteins tend to be associated with certain capsular types, there is no genetic linkage, and numerous exceptions occur (Tettelin et al., 2005).
Herein we performed immunologic and genetic analysis of eight clinical isolates that were NT, but expressed R1, R4 proteins associated with serotype V. We hypothesized that these isolates might harbour genetic changes in the cps locus that lead to an NT phenotype. Sequencing of the 18 genes in the cps locus revealed unique mutations in each of the eight GBS isolates.
|
METHODS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Double-immunodiffusion reactions. Five millilitres of 7 h cultures were inoculated into 50 ml modified THB and incubated overnight at 37 °C. Cells were extracted with hot 0.2 M HCl, concentrated 5x in a Microcon YM-10 centrifugal filter device (Amicon) and serotyped by double immunodiffusion in agarose using type V monospecific antisera (Slotved et al., 2002). Isolates A1 and A2 were used to immunize rabbits designated R-41 and R-43, respectively. Antisera obtained from the two rabbits, along with hyperimmune rabbit antiserum to prototype serotype V, were used in double-immunodiffusion reactions in agarose (Ferrieri & Flores, 1997). These reactions were also used to identify surface-protein antigens (Flores & Ferrieri, 1989).
Inhibition ELISA. An inhibition ELISA was performed to assess the relative amounts of CPS produced by the eight isolates. All eight isolates were grown in modified THB to an OD650 of 0.3. The CPSs were extracted with mutanolysin (Ferrieri & Flores, 1997). Briefly, the harvested cell pellet was washed twice with 50 mM sodium phosphate buffer (pH 7.0) and then resuspended in 750 µl of 40 % sucrose in 50 mM sodium phosphate buffer (pH 7.0). To each sample, 250 µl mutanolysin (1 mg ml–1 in 50 mM sodium phosphate buffer, pH 7.0) was added, followed by incubation at 37 °C overnight with end-over-end mixing. Protoplasts and cell debris were removed by centrifugation, and the supernatants containing CPS were transferred to fresh tubes and stored at –20 °C. The ELISA plates were coated with type V CPS–human serum albumin conjugate, as described previously (Baker et al., 2004). Mutanolysin-extracted CPS was added as an inhibitor, and threefold serial dilutions were made with PBS. Rabbit serum (diluted 1 : 200 000) containing antibodies raised against type V–tetanus toxoid conjugated vaccine was added. After incubation, a goat anti-rabbit IgG–alkaline phosphatase conjugated secondary antibody (diluted 1 : 2000) was added to the plates, and the reaction plates were read in a Biotek microtitre plate reader (Biotek Instruments). The experiments were done in duplicate for all the isolates, and the results from the ELISA plates were averaged. Relative percentage inhibition of each preparation was determined by comparison to that obtained with purified type V CPS, which resulted in complete (100 %) inhibition. Percentage inhibition was calculated as [(Ano inhibitor–Atest)/Ano inhibitor]x100. The concentration of type V CPS in each preparation was determined by multiplying the amount of purified type V CPS that resulted in 60 % inhibition (linear portion of the standard curve) by the dilution of the extract that resulted in the same degree of inhibition.
PFGE. A rapid PFGE assay was performed on isolates according to a published protocol (Benson & Ferrieri, 2001; Fasola et al., 1993; Tenover et al., 1995). Briefly, bacteria were embedded in agarose plugs and treated with mutanolysin and proteinase K to lyse the cell wall and precipitate the proteins. The DNA was digested in situ with SmaI restriction enzyme, and the resulting fragments were resolved by PFGE. The PFGE profiles of the isolates were compared visually to the band profiles of the prototype isolate using the criteria of Tenover et al. (1995).
Multilocus sequence typing (MLST). Chromosomal DNA was extracted from cells using Qiagen genomic-tip kits. PCR reactions of the seven housekeeping genes adhP, pheS, atr, glnA, sdhA, glcK and tkt were performed using primers described previously (Jones et al., 2003). The PCR products were then purified using Montage PCR centrifugal filter devices (Millipore). The amplicons were sequenced on both strands of DNA with internal nested primers using ABI Prism BigDye Terminator reaction mix. The reaction products were separated and detected with an ABI Prism 3700 DNA analyser (Applied Biosystems). A consensus sequence was generated using the MacVector program (Accelrys) for the seven gene fragments. This sequence was then uploaded to the GBS database at http://pubmlst.org/sagalactiae/ to obtain an allele designation. The combination of alleles at the seven loci provided a sequence type (ST) for each isolate (Jones et al., 2003; Maiden et al., 1998).
Molecular serotyping. PCRs were performed with new sets of primers (L. C. Madoff, unpublished results) along with primer combinations shown elsewhere to amplify regions of CPS locus thought to be specific for the nine serotypes (Kong et al., 2002a). DNA extracts from the prototype strains of the nine serotypes were used as positive controls. In addition, the cfb (CAMP factor) gene was used as a control for GBS DNA.
Surface-protein genes. PCRs were performed with primer combinations shown elsewhere to amplify regions of the bca (C alpha protein), bac (C beta protein), rib (Rib protein), alp1 (C alpha-like protein) and alp3 (C alpha-like protein) genes (Kong et al., 2002b). In addition, the cfb gene was used as a control for GBS DNA. DNA from the prototype strains was used as a positive control for the presence of the surface-protein genes.
CPS locus sequencing. All eight NT isolates plus the four serotype V clinical isolates were sequenced for polymorphisms in the CPS locus. The 18 genes (SAG1158–SAG1175) found in the CPS locus were amplified by PCR (Takara LA PCR kit, Chemicon International) using standard conditions, and the resulting amplicons were treated with shrimp alkaline phosphatase (Amersham Biosciences) and exonuclease I (Amersham Biosciences). The purified PCR products were then sequenced on both strands using the BigDye terminator cycle sequencing kit (Applied Biosystems) according to standard protocols. A total of 80 sequencing primers were used to cover the entire CPS locus at 500 bp intervals. For strains B2 and B3, strain-specific primers were designed to generate sequence data in regions where initial primers did not anneal. The sequencing reactions were run on 3730XL automated DNA sequencers (Applied Biosystems). The data from the sequencing instrument was assembled using TIGR assembler software (Pop & Kosack, 2004; Sutton et al., 1995). The edited and assembled sequence was then compared with the sequence from strain 2603V/R deposited in GenBank (accession no. AE009948) using MacVector software.
|
RESULTS |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
). In addition, concentrated HCl extracts from isolates A1–A4, but not isolates B1–B4, reacted, albeit weakly, with type V CPS-specific antibody. These data suggest that isolates A1–A4 were atypical type V, whereas isolates B1–B4 were nontypable. All eight isolates expressed R1 and R4 surface-localized proteins, as determined by double-immunodiffusion reactions with specific antisera (Flores & Ferrieri, 1989). The results suggested that A1–A4 isolates were poorly encapsulated or had a modified type V CPS, whereas the NT isolates B1–B4 were acapsular, expressed an altered capsule, or had a small amount of CPS that was not detectable in a double-immunodiffusion reaction.
|
, panel IA. In contrast, extracts prepared from isolates B1–B4 each contained <0.05 µg ml–1 of type V CPS.
Clonality of isolates
All eight isolates expressed R1, R4 proteins, and their PFGE band profile patterns showed that they were all similar and belonged to DNA profile group 4 (Fig. 2). This banding pattern was similar to that of a prototype V/R1, R4 isolate when matched with the PFGE database. Three of these isolates (B2–B4) had a one-band difference and were assigned to profile group 4a.
|
). The other seven isolates (A2–A4, B1–B4) were all ST-1. These molecular typing results suggest that these eight isolates are genotypically similar.
|
Surface-protein genes
Previous studies (Ferrieri & Flores, 1997; Kong et al., 2002b) have shown strong associations between capsule serotypes and surface-protein genes. All eight isolates were found to have alp3, which is significantly associated with GBS type V (Kong et al., 2002a).
CPS locus sequencing
About 17.7 kb of chromosomal DNA from each of the eight isolates was sequenced for polymorphisms in the CPS genes. Four additional serotype V isolates were sequenced to identify the diversity of polymorphisms in the cps locus and also to serve as a control sequence for interpretation of the sequence data from the atypical V and NT/R1, R4 isolates. The data from the sequencing reactions were used to build a contig and were then compared with the published 2603V/R genome sequence. Sequencing results for the four atypical V isolates (A1–A4) and the nonserotypable isolates (B1–B3) indicated that all the genes (SAG1158–SAG1175) were present without major deletions or insertions.
Isolate B4 had an 861 bp insert at the third base of codon 373 in cpsE (Fig. 3). This insert was identified as an insertion sequence, IS1381. The insertion sequence contained a 20 bp terminal inverted repeat. The transcriptional direction of the two ORFs encoding putative transposases in the insertion sequence was reversed compared with that of cpsE.
|
); 12 of these mutations were synonymous changes, 13 were nonsynonymous changes, two were insertion mutations, and one was a single substitution mutation in the cpsD–cpsE intergenic region. In addition, each isolate (A1–A4, B1–B4) showed one to three unique mutations (Table 3). The 28 common polymorphisms found in the eight isolates were also found in the three control typable V isolates whose ST was ST-1, suggesting a strong linkage disequilibrium between the polymorphisms found in the cps locus and the ST of the isolates. Another typable V isolate with an ST-7 profile shared most of the polymorphisms with 2603V/R (ST-19) strain, except for two polymorphisms (Table 2). All 12 isolates, independent of ST, had a 9 bp insertion at codon 217 of cpsD.
|
|
|
DISCUSSION |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
PFGE showed that all eight isolates had a profile similar to that of a GBS type V containing R1 and R4 proteins (Ferrieri et al., 1997). Previous studies have shown an association of isolates with specific CPS serotypes with certain protein profiles, e.g. alpha C protein with type Ia, alpha and beta C proteins with type Ib, Rib protein with type III, and R1, R4 proteins with type V (Ferrieri & Flores, 1997; Flores & Ferrieri, 1989; Kong et al., 2002b). Serotype V isolates have been found to fall into a limited number of profile groups, and serotype V isolates producing R1, R4 proteins tend to belong to profile group 4. Also, NT isolates with the R1, R4 protein profile tend to belong to PFGE group 4, suggesting that there is restricted genetic diversity among isolates producing R1, R4 proteins (Ferrieri et al., 1997). The PFGE profile group 4 is identical to PFGE subtype O described elsewhere (Elliott et al., 1998). The recent increase in the recovery of isolates with the NT/R1, R4 profile from patients may indicate a periodic shift of this clonal complex with improved genetic fitness. This periodic shift may also account for the observed restriction in genetic variation among NT/R1, R4 isolates. Future surveillance data will assess the extent of clonal divergence that may arise from this group of isolates. All eight isolates cluster into a single clonal group and belong to the ST-1 complex, which includes the newly identified ST-173 (isolate A1), a single-locus variant of ST-1. GBS isolates of both human and bovine origin belong to the ST-1 complex (Bisharat et al., 2004). STs can be grouped into lineages or clonal complexes by the BURST program, and isolates of the ST-1 complex are grouped as BURST lineage 1 (Bisharat et al., 2004; Jones et al., 2003). The typing assays indicate that the four atypical V/R1, R4 isolates and the four NT/R1, R4 isolates are genetically similar, but that they differ in the amount of type V CPS produced and/or may have a modified type V CPS.
The cpsO gene in the cps locus specifically associated with serotype V was present in all eight isolates. Although the presence of cpsO does not indicate whether the gene product is expressed, the expected size of the amplicon obtained (642 bp, data not shown) rules out large insertions or deletions in the gene fragment that could potentially result in an NT phenotype. This result was corroborated by the CPS locus sequencing data. PCR analysis of the surface-protein genes identified the presence of alp3 in all eight isolates. Our results did not identify the rib gene in any of the eight isolates, although all of them produced R1, R4 proteins. This finding concurs with the recent observation that isolates co-expressing the R1 and R4 proteins show a negative reaction in both PCR and Southern hybridization assays with an r4 gene probe (Smith et al., 2004). The strong association between the presence of alp3 and the co-expression of R1, R4 proteins suggests that alp3 encodes a protein that reacts immunologically with both R1 and R4 antisera. Recent studies involving nucleotide and amino acid sequencing of the R4 protein indicate that the proteins R4 and Rib are identical and encoded by the r4/rib gene (Smith et al., 2004). The gene that encodes the R1 protein has not been identified, and further study of the R1 and Alp family of proteins at the genetic level is warranted.
All 18 genes in the CPS locus were sequenced to find the genetic changes that would explain the phenotypic differences between the four atypical and the four NT isolates. Except for isolate B4, no major insertions or deletions were found in the isolates. Isolate B4 had an IS1381 insertion in the C terminus of cpsE. Originally identified in a Streptococcus pneumoniae R6 strain (Sanchez-Beato et al., 1997), IS1381 has been reported in GBS isolates of both human and bovine origin (Dmitriev et al., 2004; Shakleina et al., 2004; Tamura et al., 2000), although this is the first observation of the disruption by IS1381 of a gene in the GBS cps locus. cpsE is thought to encode a glycosyltransferase in type Ia GBS and is proposed to have a role as the initiating enzyme in the biosynthesis of polysaccharide repeating units (Cieslewicz et al., 2001). Indeed, a cpsE knockout mutant in a serotype III strain has reduced CPS in both cell wall and protoplast fractions when compared with the wild-type parent strain (Chaffin et al., 2005). This finding is consistent with the reduced or disrupted expression of CPS due to apparent loss of the function of glycosyltransferase by isolate B4. Sequencing analysis also identified a 9 bp repetitive sequence in the C terminus of cpsD. All isolates used in this study, except for the 2603V/R isolate, had this repetitive sequence. This finding extends and corroborates the finding elsewhere (Kong et al., 2002a) that all serotype V isolates have this repetitive sequence.
Sequencing analysis also showed the distribution of 28 polymorphisms across 10 CPS genes found in common in all eight isolates. The three typable isolates with the ST-1 profile (BURST lineage 1) had the same 28 polymorphisms, suggesting a linkage disequilibrium between the ST and the polymorphisms, and indicative of the clonal structure of these isolates. Furthermore, another typable isolate with ST-7 (BURST lineage 3) and a 2603V/R isolate (single-locus variant of ST-19, BURST lineage 6) had different sets of polymorphisms (Jones et al., 2003). The presence of these common sets of polymorphisms in both typable and NT isolates suggests that these polymorphisms are not involved in the NT phenotype.
In addition to the 28 common polymorphisms seen, one to three unique mutations were observed in seven of the eight isolates. Other than the IS1381 insertion, no unique single-nucleotide polymorphisms were found in isolate B4. Isolate B1 had a termination mutation in cpsH, a gene known to encode a polymerase in both serotypes Ia and III (Cieslewicz et al., 2001). This isolate also had a nonsynonymous mutation in two other genes, cpsJ and cpsF, whose products, CpsJ and CpsF, are annotated to be a glycosyltransferase and a polysaccharide biosynthetic protein, respectively, in the sequenced genome of 2603V/R (Tettelin et al., 2002). A combination of these three mutations may account for the NT phenotype in this isolate. Four isolates (A1–A3, B3) had two mutations, one each in the cpsG and cpsA genes. cpsA is thought to encode a transcriptional regulator (Cieslewicz et al., 2001), and cpsG is thought to encode a ß-1,4-galactose transferase in both serotypes Ia and III (Cieslewicz et al., 2005). Three of the isolates (A2, A3 and B3) had a single-nucleotide deletion in codon 84 of cpsG, resulting in a frame-shift mutation that potentially results in loss of function of this glycosyltransferase. This truncation mutation, along with the other nonsynonymous changes found, may account for the NT phenotype in these isolates. Isolates A4 and B2 have a single nonsynonymous substitution in the cpsM and cpsE genes, respectively.
In conclusion, we have identified new genetic correlates for the NT phenotype in GBS. Knowledge of the composition and function of the various cps genes in different serotypes is warranted, as it provides insight into the dynamics of capsule biosynthesis, surface-protein expression, and the evolution of GBS serotypes.
|
ACKNOWLEDGEMENTS |
|---|
|
REFERENCES |
|---|
|
TOPINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  F. Kong, L. M. Lambertsen, H.-C. Slotved, D. Ko, H. Wang, and G. L. Gilbert Use of Phenotypic and Molecular Serotype Identification Methods To Characterize Previously Nonserotypeable Group B Streptococci J. Clin. Microbiol., August 1, 2008; 46(8): 2745 - 2750. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
H.-C. Slotved, F. Kong, L. Lambertsen, S. Sauer, and G. L. Gilbert Serotype IX, a Proposed New Streptococcus agalactiae Serotype J. Clin. Microbiol., September 1, 2007; 45(9): 2929 - 2936. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | J MED MICROBIOL | MICROBIOLOGY | J GEN VIROL | ALL SGM JOURNALS |
+ß); 12, PFGE control Ib/c(