Portal de Periódicos da CAPES

Peptide Mapping of a Novel Discontinuous Epitope of the Major Surface Adhesin from Streptococcus mutans

2004; Elsevier BV; Volume: 279; Issue: 21 Linguagem: Inglês

10.1074/jbc.m400820200

1083-351X

Craig J. van Dolleweerd, Charles Kelly, Daniel Chargelegue, Julian K‐C.,

Monoclonal and Polyclonal Antibodies Research

Guy's 13 is a mouse monoclonal antibody that specifically recognizes the major cell-surface adhesion protein SA I/II of Streptococcus mutans, one of the major causative agents of dental caries. Passive immunization with Guy's 13 prevents bacterial colonization in humans. To help elucidate the mechanism of prevention of colonization conferred by this antibody, the SA I/II epitope recognized by Guy's 13 was investigated. It was previously established that the epitope is conformational, being assembled from two non-contiguous regions of SA I/II. In the current study, using recombinant fragments of SA I/II and, ultimately, synthetic peptides, the discontinuous epitope was localized to residues 170–218 and 956–969. This work describes the mapping of a novel discontinuous epitope that requires an interaction between each determinant in order for epitope assembly and recognition by antibody to take place. Guy's 13 binds to the assembled epitope but not to these individual epitope fragments. The assembled epitope results from the interaction between the individual antigenic determinants and can be formed by mixing together determinants present on separate polypeptide chains. The data are consistent with one of the epitope fragments adopting a polyproline II-like helical conformation. Guy's 13 is a mouse monoclonal antibody that specifically recognizes the major cell-surface adhesion protein SA I/II of Streptococcus mutans, one of the major causative agents of dental caries. Passive immunization with Guy's 13 prevents bacterial colonization in humans. To help elucidate the mechanism of prevention of colonization conferred by this antibody, the SA I/II epitope recognized by Guy's 13 was investigated. It was previously established that the epitope is conformational, being assembled from two non-contiguous regions of SA I/II. In the current study, using recombinant fragments of SA I/II and, ultimately, synthetic peptides, the discontinuous epitope was localized to residues 170–218 and 956–969. This work describes the mapping of a novel discontinuous epitope that requires an interaction between each determinant in order for epitope assembly and recognition by antibody to take place. Guy's 13 binds to the assembled epitope but not to these individual epitope fragments. The assembled epitope results from the interaction between the individual antigenic determinants and can be formed by mixing together determinants present on separate polypeptide chains. The data are consistent with one of the epitope fragments adopting a polyproline II-like helical conformation. The Gram-positive bacterium Streptococcus mutans is one of the main causative agents of dental caries (1Loesche W.J. Microbiol. Rev. 1986; 50: 353-380Crossref PubMed Google Scholar). Adhesion of the bacterium to the tooth surface is mediated by a major bacterial cell surface protein, termed streptococcal antigen I/II (SA I/II) 1The abbreviations used are: SA I/II, streptococcal antigen I/II; aa, amino acid residues; ASM, alanine scanning mutagenesis; ELISA, enzyme-linked immunosorbent assay; PBS, phosphate-buffered saline; HRP, horseradish peroxidase; PP II, polyproline II. (2Russell M.W. Lehner T. Arch. Oral Biol. 1978; 23: 7-15Crossref PubMed Scopus (114) Google Scholar). This protein has been identified in a number of strains and serotypes of S. mutans and has also been called SR (3Ogier J.A. Wachsmann D. Schöller M. Lepoivre Y. Klein J.P. Arch. Oral Biol. 1990; 35 (suppl.): 25S-31SCrossref PubMed Scopus (5) Google Scholar) or PAc (4Okahashi N. Sasakawa C. Yoshikawa M. Hamada S. Koga T. Mol. Microbiol. 1989; 3: 221-228Crossref PubMed Scopus (94) Google Scholar). Sequencing of the spaP gene encoding SA I/II (5Kelly C. Evans P. Bergmeier L. Lee S.F. Progulske-Fox A. Harris A.C. Aitken A. Bleiweis A.S. Lehner T. FEBS Lett. 1989; 258: 127-132Crossref PubMed Scopus (92) Google Scholar) and alignment of the predicted amino acid sequence with the antigen I/II sequences from other streptococcal strains and species reveals many common features (6Jenkinson H.F. Demuth D.R. Mol. Microbiol. 1997; 23: 183-190Crossref PubMed Scopus (178) Google Scholar), and on this basis the proteins have been divided into a number of distinct regions (Fig. 1). These include an alanine-rich repeat (A) region (residues 121–447 of SA I/II from S. mutans NG5) consisting of four tandem repeats of an 82 amino acid (aa) sequence, a variable (V) region, where most of the sequence variation between antigen I/II homologues occurs and a proline-rich repeat (P) region (aa 840–983) consisting of three tandem repeats of a 39-aa sequence, together with a fourth, more degenerate repeat. The mouse monoclonal antibody Guy's 13, which recognizes SA I/II of S. mutans and S. sobrinus (7Smith R. Lehner T. Oral Microbiol. Immunol. 1989; 4: 153-158Crossref PubMed Scopus (21) Google Scholar), has been used successfully to prevent S. mutans colonization and the development of caries in non-human primates (8Lehner T. Caldwell J. Smith R. Infect. Immun. 1985; 50: 796-799Crossref PubMed Google Scholar) and prevents bacterial colonization in human clinical trials (9Ma J.K.C. Hunjan M. Smith R. Lehner T. Clin. Exp. Immunol. 1989; 77: 331-337PubMed Google Scholar, 10Ma J.K.C. Hunjan M. Smith R. Kelly C. Lehner T. Infect. Immun. 1990; 58: 3407-3414Crossref PubMed Google Scholar). Guy's 13 has been expressed, as a secretory immunoglobulin molecule in transgenic Nicotiana tabacum plants (11Ma J.K.C. Hiatt A. Hein M. Vine N.D. Wang F. Stabila P. van Dolleweerd C. Mostov K. Lehner T. Science. 1995; 268: 716-719Crossref PubMed Scopus (478) Google Scholar) and as a single chain Fv on the surface of the commensal bacterium Lactobacillus zeae (12Kruger C. Hu Y. Pan Q. Marcotte H. Hultberg A. Delwar D. van Dalen P.J. Pouwels P.H. Leer R.J. Kelly C.G. van Dollenweerd C. Ma J.K. Hammarstrom L. Nat. Biotechnol. 2002; 20: 702-706Crossref PubMed Scopus (151) Google Scholar), offering possible strategies for passive mucosal immunization against dental caries (13Ma J.K.C. Hikmat B.Y. Wycoff K. Vine N.D. Chargelegue D. Yu L. Hein M.B. Lehner T. Nat. Med. 1998; 4: 601-606Crossref PubMed Scopus (412) Google Scholar). Previous work by our group has established that the epitope recognized by Guy's 13 monoclonal antibody is conformational, being assembled from two regions of SA I/II that are noncontiguous in sequence (14van Dolleweerd C.J. Chargelegue D. Ma J.K.-C. Infect. Immun. 2003; 71: 754-765Crossref PubMed Scopus (26) Google Scholar). Using immunoblotting, it was shown that Guy's 13 recognizes two non-overlapping recombinant polypeptides that show no sequence homology. One part of the discontinuous epitope was shown to reside in a recombinant polypeptide (fragment 114, see Fig. 1) spanning aa 45–457 of SA I/II. The partial epitope residing in this fragment, which includes the A-region of SA I/II, is designated the A-region binding site or A site. The other part of the epitope was localized to a fragment (Pro0-IV) spanning aa 816–983 of SA I/II. The partial epitope residing in this fragment (which spans the P-region of SA I/II) is called the P-region binding site or P site. In ELISA-based experiments, binding of Guy's 13 to these individual fragments could not be demonstrated unless a recombinant polypeptide containing the A site was incubated simultaneously with a recombinant fragment containing the P site. Additionally, we showed that the two binding sites are able to associate with each other, leading to a structure that can subsequently be recognized by Guy's 13. This suggested that the A and P sites, which are widely separated in the linear sequence, are situated in close proximity in the native SA I/II. The aim of the current work was to map both binding sites in greater detail and, in the process, identify critical residues responsible for forming the discontinuous Guy's 13 epitope. This in turn may give a better understanding of the structure of SA I/II, the nature of the antibody-antigen interaction, and provide an insight into how the antibody prevents S. mutans colonization of the oral cavity. Bacterial Strains and Antigen I/II—Escherichia coli BL21(DE3)-pLysS was from Novagen. SA I/II was prepared from S. mutans Guy's strain (serotype c) (7Smith R. Lehner T. Oral Microbiol. Immunol. 1989; 4: 153-158Crossref PubMed Scopus (21) Google Scholar) as described previously (15Russell M.W. Bergmeier L.A. Zanders E.D. Lehner T. Infect. Immun. 1980; 28: 486-493PubMed Google Scholar). Antibodies—Guy's 13 monoclonal antibody and all other primary and secondary antibodies were used as described previously (14van Dolleweerd C.J. Chargelegue D. Ma J.K.-C. Infect. Immun. 2003; 71: 754-765Crossref PubMed Scopus (26) Google Scholar). Cloning and Expression of Recombinant spaP Gene Fragments— Genomic DNA of S. mutans was isolated according to the method of Bollet et al. (16Bollet C. Gevaudan M.J. de Lamballerie X. Zandotti C. de Micco P. Nucleic Acids Res. 1991; 19: 1955Crossref PubMed Scopus (110) Google Scholar). Amplification, cloning, and expression, using the pEXssΔ3 vector, of recombinant spaP gene fragments (see Fig. 1) was performed as described previously (14van Dolleweerd C.J. Chargelegue D. Ma J.K.-C. Infect. Immun. 2003; 71: 754-765Crossref PubMed Scopus (26) Google Scholar). The identity of the recombinant clones was confirmed by sequencing. Recombinant polypeptides are produced as a fusion with a C-terminal, 13-amino acid, E Tag peptide (17Schier R. Balint R.F. McCall A. Apell G. Larrick J.W. Marks J.D. Gene (Amst.). 1996; 169: 147-155Crossref PubMed Scopus (56) Google Scholar). An anti-E Tag antibody (Amersham Biosciences) was used to confirm expression of the recombinant polypeptides by immunoblotting (not shown). The residues of SA I/II encoded by the recombinant polypeptides are shown in Fig. 1. Peptide Synthesis—Peptides, corresponding to the SA I/II sequence from S. mutans NG5 strain, were synthesized as peptide amides by manual procedure on a solid phase scaffold using standard Fmoc (N-(9-fluorenyl)methoxycarbonyl) chemistry (18Wellings D.A. Atherton E. Methods Enzymol. 1997; 289: 44-67Crossref PubMed Scopus (240) Google Scholar) in a Biotech Instruments BT7400 peptide synthesis block (19Dörner B. Ostresh J.M. Houghten R.A. Frank R. Tiepold A. Fox J.E. Bray A.M. Ede N.J. James I.W. Wickham G. Chan W.C. White P.D. Fmoc Solid Phase Peptide Synthesis: A Practical Approach. University Press, Oxford2000Google Scholar), as described by Kelly et al. (20Kelly C.G. Younson J.S. Hikmat B.Y. Todryk S.M. Czisch M. Haris P.I. Flindall I.R. Newby C. Mallet A.I. Ma J.K.-C. Lehner T. Nat. Biotechnol. 1999; 17: 42-47Crossref PubMed Scopus (97) Google Scholar). Peptide compositions were determined by mass spectrometry. Immunoassays—For ELISAs, Maxisorp 96-well ELISA plates (Nunc, Roskilde, Denmark) were coated with SA I/II (2 μg/ml solution in PBS) or cell lysates from E. coli cultures expressing recombinant SA I/II fragments. The plates were incubated overnight at 4 °C and washed three times with H2O. Free protein binding sites were blocked with 2.5% (w/v) bovine serum albumin (Sigma, Fraction V) in PBS for 2 h at 37 °C. The plates were washed three times with H2O containing 0.1% (v/v) Tween 20, air-dried, and stored at -20 °C until required. Competition ELISA to measure the ability of lysates or peptides to inhibit the binding of Guy's 13 IgG to solid phase SA I/II has been described previously (14van Dolleweerd C.J. Chargelegue D. Ma J.K.-C. Infect. Immun. 2003; 71: 754-765Crossref PubMed Scopus (26) Google Scholar). Essentially, 65 μl of a 1 μg/ml solution of Guy's 13 IgG in 5× DB (12.5% (w/v) bovine serum albumin, 0.5% (v/v) Tween 20 in PBS) was added to 130 μl of inhibitor-1 (either neat lysate, 250 μm peptide in PBS, or PBS) and 130 μl of inhibitor-2 (neat lysate or PBS). Aliquots of 100 μl were added in triplicate to the wells of SA I/II-coated ELISA plates. Following overnight incubation at 4 °C, the wells were washed six times with H2O containing 0.1% (v/v) Tween 20. Detection of bound Guy's 13 IgG was with a horseradish peroxidase (HRP)-conjugated secondary antibody and incubation at 37 °C for 2 h. HRP activity was quantitated by the addition of 3,3′,5,5′-tetramethylbenzidine (Sigma). Enzyme activity was stopped with 1 m H2SO4 and the A450 measured. Direct ELISA to measure the ability of peptides or lysates to promote the binding of Guy's 13 IgG to immobilized recombinant SA I/II fragments was described previously (14van Dolleweerd C.J. Chargelegue D. Ma J.K.-C. Infect. Immun. 2003; 71: 754-765Crossref PubMed Scopus (26) Google Scholar). Peptides were diluted to 100 μm in DB. Aliquots (240 μl) of peptide (100 μm), E. coli lysate, or DB were mixed with 10 μl of Guy's 13 IgG (5 μg/ml solution in DB). An isotype-matched mouse IgG1κ monoclonal antibody (MOPC 31C) was used as a negative control. Aliquots of 100 μl were added in triplicate to the wells of an ELISA plate coated in advance with recombinant E. coli lysates. Following overnight incubation at 4 °C non-bound fluid phase components were removed and bound Guy's 13 IgG detected as described above. Mapping of the A-region Binding Site Using Competition ELISA—We have previously demonstrated the simultaneous requirement for both the A and P sites in fluid phase assays (14van Dolleweerd C.J. Chargelegue D. Ma J.K.-C. Infect. Immun. 2003; 71: 754-765Crossref PubMed Scopus (26) Google Scholar). In a competition ELISA, Guy's 13 IgG binding to solid phase SA I/II was inhibited using a lysate from E. coli expressing a recombinant fragment of the P-region mixed with lysate from E. coli expressing an A-region fragment. A-region fragments 114, 115, 116, and 121 (Fig. 1) were mixed with recombinant P-region fragment Pro0-IV (Fig. 2A). When only one fragment was incubated with Guy's 13 (the “PBS” columns), antibody binding to solid phase SA I/II was unaffected compared with the control wells which received only Guy's 13 (the “no inhibitor” column). Mixing of 114 lysate with Pro0-IV lysate resulted in over 90% inhibition of Guy's 13 IgG binding to solid phase SA I/II (OD450 < 0.1), consistent with our previous report (14van Dolleweerd C.J. Chargelegue D. Ma J.K.-C. Infect. Immun. 2003; 71: 754-765Crossref PubMed Scopus (26) Google Scholar). When A-region fragments 115 or 121 (which both span the first alanine-rich repeat) were mixed with Pro0-IV lysate, >90% inhibition was also observed. In contrast, when A-region fragment 116 (representing the C-terminal half of 114) was mixed with Pro0-IV, inhibition of only 30% was observed. Recombinant polypeptide 110 (spanning aa 984–1539 of SA I/II), which was shown previously to have no Guy's 13-binding activity in immunoblotting or in ELISAs (14van Dolleweerd C.J. Chargelegue D. Ma J.K.-C. Infect. Immun. 2003; 71: 754-765Crossref PubMed Scopus (26) Google Scholar), was used as a negative control. The lack of inhibition observed following mixing of the 110 lysate with either Pro0-IV lysate or with each of the A-region lysates (Fig. 2A) confirmed the requirement for the simultaneous presence of both A- and P-region binding sites. These results show that all of the recombinant fragments of the A-region were able to inhibit Guy's 13 IgG binding to solid phase SA I/II when mixed with Pro0-IV recombinant polypeptide, although the extent of inhibition varied depending on the recombinant polypeptide used. None of the recombinant fragments were able to inhibit Guy's 13 binding on their own. In view of the repeat nature of the A-region it is, perhaps, not surprising that all of the A-region recombinant polypeptides tested gave some levels of inhibition. The recombinant polypeptide 121 (aa 121–218 of SA I/II) inhibited Guy's 13 binding by >90% when mixed with Pro0-IV polypeptide. Two further non-overlapping clones of A-region fragment 121 were generated: fragment 128 (aa 121–169 of SA I/II) and fragment 129 (aa 170–218) (see Fig. 1). In combination with Pro0-IV fragment, A-region fragment 129 is an effective inhibitor of Guy's 13 binding to solid phase SA I/II (see Fig. 2B), giving levels of inhibition (80%) nearly equivalent to 121 (88.5%), while fragment 128 did not inhibit Guy's 13 IgG binding above controls. The A-region binding site therefore lies within aa 170–218 of SA I/II. An A-region binding site can therefore be localized to this fragment and the ELISA results indicate that homologous A-region binding sites may be present in at least one of the remaining A-region tandem repeats. Mapping of the Binding Sites Using Synthetic Peptides—The P-region binding site had previously been mapped to a recombinant polypeptide, representing aa 816–983 of SA I/II (14van Dolleweerd C.J. Chargelegue D. Ma J.K.-C. Infect. Immun. 2003; 71: 754-765Crossref PubMed Scopus (26) Google Scholar). To further characterize the binding of Guy's 13 to SA I/II, 19 overlapping 20-mer peptides spanning the P-region were synthesized (Table I). A further peptide, p1037, which lies C-terminal to the P-region, was included as a negative control. Each of the peptides was tested for their ability to inhibit Guy's 13 IgG binding to solid phase SA I/II in combination with A-region fragment 121. Three independent experiments were performed, and the results were plotted as the mean percentage inhibition (±S.D.) relative to a no inhibitor control, which represents Guy's 13 binding to solid phase SA I/II in the absence of any inhibitor lysate or peptide (Fig. 3). As a positive control, a mixture of 121 and Pro0-IV lysates produced high levels of inhibition (89%). No inhibition (<5%) was observed with any fragment or peptide in combination with control recombinant polypeptide 110 (gray bars). Most of the 121 lysate + peptide combinations also did not give any inhibition (<5%), comparable with the no inhibitor control and with the samples utilizing the negative control 110 lysate + peptide combinations. The exceptions were peptides p886, p926, p955, and p965, which produced varying levels of inhibition at 27%, 52%, 87%, and 20%, respectively.Table INomenclature and sequence of peptides spanning amino acids 803-1004 of the P-region of SA I/II from S. mutans NG5 Peptide p1037 was included as a negative control. Amino acid numbering is from SA I/II of S. mutans strain NG5.PeptideNumberingSequencep803803-822ETGKKPNIWYSLNGKIRAVNp814814-833LNGKIRAVNLPKVTKEKPTPp824824-843PKVTKEKPTPPVKPTAPTKPp834834-853PVKPTAPTKPTYETEKPLKPp847847-866TEKPLKPAPVAPNYEKEPTPp857857-876APNYEKEPTPPTRTPDQAEPp867867-886PTRTPDQAEPKKPTPPTYETp877877-896KKPTPPTYETEKPLEPAPVEp886886-905TEKPLEPAPVEPSYEAEPTPp896896-915EPSYEAEPTPPTRTPDQAEPp906906-925PTRTPDQAEPNKPTPPTYETp916916-935NKPTPPTYETEKPLEPAPVEp926926-945 (=887-906)EKPLEPAPVEPSYEAEPTPPp935935-954EPSYEAEPTPPTPTPDQPEPp945945-964PTPTPDQPEPNKPVEPTYEVp955955-974NKPVEPTYEVIPTPPTDPVYp965965-984IPTPPTDPVYQDLPTPPSIPp975975-994QDLPTPPSIPTVHFHYFKLAp985985-1004TVHFHYFKLAVQPQVNKEIRp10371037-1056ETTSFVLVDPLPSGYQFNPE Open table in a new tab Peptides p926 and p955 (mixed with 121 lysate) inhibited the binding of Guy's 13 to solid phase SA I/II by more than 50%. Sequence homology between these two peptides is the likely reason for this. Indeed, all four inhibitory peptides (p886, p926, p955, and p965) share a common sequence motif, discussed later. Dose-dependent inhibition of Guy's 13 binding to immobilized SA I/II has been demonstrated before for both the A and P sites (14van Dolleweerd C.J. Chargelegue D. Ma J.K.-C. Infect. Immun. 2003; 71: 754-765Crossref PubMed Scopus (26) Google Scholar). In the presence of saturating amounts of the 121 lysate, dose-dependent inhibition by peptide p955 was demonstrated. Maximal inhibition was achieved with 40 μm p955, with an IC50 (50% inhibitory concentration) of 3.3 μm (data not shown). Six 20-mer peptides spanning aa 161–230 of SA I/II (each peptide overlapping by 10 residues) were synthesized to map more finely the A-region binding site. However, none of these peptides significantly inhibited the binding of Guy's 13 to solid phase SA I/II when mixed with lysate from E. coli expressing recombinant fragment Pro0-IV (results not shown). Homologues of Peptide p955—The peptide p955 sequence is located within tandem repeat IV of the proline-rich region of SA I/II. The repeat nature of the P-region suggested that p955-homologous peptides from the other P-region tandem repeats might also inhibit Guy's 13 binding. Table II shows the sequence of p955 aligned with the three other peptide homologues from P-region tandem repeats I, II, and III, which were synthesized and tested for their ability to inhibit Guy's 13 IgG binding to solid phase SA I/II by preincubation with the 121 fragment (or with the 110 fragment as a negative control).Table IISynthetic peptides from the P-region of SA I/II from S. mutans NG5 homologous to peptide p955 The location of the peptides within each of the P-region repeats is shown. Amino acid numbering is from SA I/II of S. mutans strain NG5.PeptideNumberingRepeatSequencep853853-872IPAPVAPNYEKEPTPPTRTPDp892892-911IIPAPVEPSYEAEPTPPTRTPDp931931-950IIIPAPVEPSYEAEPTPPTPTPDp955955-974IVNKPVEPTYEVIPTPPTDPVY Open table in a new tab The results (Fig. 4A) showed that, in addition to p955, peptides p892 and p931 (from P-region repeats II and III, respectively) resulted in 88% inhibition of binding of Guy's 13 to solid phase SA I/II, whereas p853 (from P-region repeat I) was not an effective inhibitor. Alanine Scanning Mutagenesis of Peptide p955—ASM of peptide p955 was performed. Thirteen peptide analogues (A3 to A15) were synthesized, each containing a single alanine substitution at positions 3 to 15 of peptide p955. Each of the alanine-substituted peptides was tested with the 121 fragment (or with the 110 fragment as a negative control) for inhibition of Guy's 13 IgG binding to solid phase SA I/II (see Fig. 4B). As a positive control, Guy's 13 IgG preincubated with fragment 121 and either Pro0-IV or p955 resulted in high level (90%) inhibition. The results enabled the alanine substitutions to be classified into three groups: (i) those alanine substitutions that had no effect on the inhibition by p955 of Guy's 13 binding to SA I/II (positions 4, 7, 9, 10, 13, and 14), (ii) those substitutions that completely abolished the inhibitory effects of p955 (positions 5 and 8), and (iii) those alanine substitutions that resulted in partial abolition of the inhibitory effects of p955 (positions 3, 6, 11, 12, and 15). N- and C-terminal Deletions of Peptide p955—Ten peptides were synthesized that exhibited progressive deletions of amino acid residues from the N terminus of peptide p955. A further 10 peptides were synthesized with progressive deletions of amino acid residues from the C terminus of peptide p955. The results of competition ELISAs using the N-terminal deleted peptides showed that deletion of the N-terminal residue (Asn) of p955 had no effect on the inhibition of Guy's 13 binding to solid phase SA I/II. Deletion of the first 2 residues (Asn-Lys), however, reduced by 38% the inhibition exhibited by p955, while deletion of the first 3 residues completely abolished the inhibitory effect of p955 (data not shown). The results of the competition ELISAs using the C-terminal deleted peptides showed that deletion of the last 5 residues of p955 (Thr-Asp-Pro-Val-Tyr) had very little effect on the ability of the peptide to inhibit Guy's 13 binding to solid phase SA I/II. Deletion of the last 6–8 residues resulted in progressively lower levels of inhibition, culminating in total loss of inhibition when the last 9 residues were removed (data not shown). Binding of Guy's 13 IgG to Solid Phase A-region or P-region Binding Site in the Presence of Peptides or Lysates—To confirm the results of the competition ELISAs, the ability of peptides or lysates to promote binding of Guy's 13 to solid phase-associated binding sites was investigated. The results show that when A-region fragment 121 was adsorbed to the solid phase of an ELISA plate, Pro0-IV lysate and peptide p955 both promoted the binding of Guy's 13 (Fig. 5A). No effect was seen with a control monoclonal antibody (MOPC 31C) (gray bars). The results using P-region tandem repeat homologues of p955 (see Table II) show that p892 and p931 also promote binding of Guy's 13 to 121 while p853 does not. Lysates 121, 129, and 110 did not promote binding of Guy's 13 to immobilized 121. These findings are consistent with the results of the competition ELISAs (see Figs. 2 and 4A) and confirm the simultaneous requirement for both the A- and P-region binding sites. The results using the alanine-substituted peptide analogues of p955 in this direct binding ELISA (see Fig. 5B) also mirrored the results obtained using these peptides in the competition ELISAs. Thus, peptides (such as A7, A10, and A13), which are strong promoters of Guy's 13 binding to solid phase 121 fragment, are also strong inhibitors (in combination with fragment 121) of Guy's 13 binding to immobilized SA I/II (see Fig. 4B). As a control, Pro0-IV lysate (encoding the P-region binding site) coated onto the solid phase did not bind Guy's 13 in the presence of fluid phase p955 but did bind Guy's 13 in the presence of fluid phase 121 and 129 lysates, which both encode A-region binding sites (data not shown). Negative control 110 fragment coated onto the solid phase did not bind Guy's 13 IgG in the presence of p955 or any of the A- or P-region lysates (data not shown). In an attempt to delineate the Guy's 13 conformational epitope, we made use of earlier findings that the Guy's 13 epitope was formed from non-contiguous sequences that, when added together, were able to reconstitute the Guy's 13 epitope. The epitope is unusual because it can be reassembled in vitro irrespective of whether the two regions lie in cis or in trans. This allowed each binding site to be studied in turn and enabled each to be more finely mapped by cloning and expression of successively smaller recombinant fragments and, ultimately, with synthetic peptides. The A site was mapped to a recombinant polypeptide (fragment 129) spanning aa 170–218 of the A-region of SA I/II. This sequence is identical to aa 170–218 of PAc from S. mutans serotype c strain MT8148 (21Okahashi N. Sasakawa C. Yoshikawa M. Hamada S. Koga T. Mol. Microbiol. 1989; 3: 673-678Crossref PubMed Scopus (117) Google Scholar). The data indicated that at least one further binding site was present in the A-region, consistent with the repeat nature of this region. Attempts to map the binding site in the 129 fragment using overlapping 20-mer peptides were unsuccessful. Collectively, alignment of the 20-mer synthetic peptides p886, p926, p955, and p965, the results of ASM, the use of N- and C-terminally truncated peptides, and a comparison of the peptide homologues allowed a consensus P-region binding site to be defined by a 14-amino acid sequence: XPX(E/D)PXYXXX-PXPP. This sequence can be found at aa 893–906, 932–945, and 956–969 of SA I/II and implies the presence of at least three P-region binding sites for Guy's 13. ASM of p955 showed that residues Pro957, Pro960, Tyr962, Pro966, Thr967, Pro968, and the acidic residue Glu959 were most important for peptide binding. Although the P site is shown to reside on a peptide, the data from the ASM show that this partial epitope is not defined simply by a linear sequence of residues (22Geysen H.M. Barteling S.J. Meloen R.H. Proc. Natl. Acad. Sci. U. S. A. 1985; 82: 178-182Crossref PubMed Scopus (233) Google Scholar, 23Geysen H.M. Rodda S.J. Mason T.J. Tribbick G. Schoofs P.G. J. Immunol. Methods. 1987; 102: 259-274Crossref PubMed Scopus (723) Google Scholar). Thus many of the residues in this partial epitope are most likely not involved in making direct contact with the complementarity determining region residues of the antibody. Similarly, most of the residues of the A site are also unlikely to be involved in interaction with the paratope. This indicates that each binding site still retains some secondary structure, consisting of independent contact residues interspersed among non-contact residues. Non-contact residues may be required for either the formation of a stable secondary structure or for interactions between the A- and P-region binding sites, which then allows the assembled epitope to be recognized by Guy's 13. Other workers (24Korth C. Stierli B. Streit P. Moser M. Schaller O. Fischer R. Schulz-Schaeffer W. Kretzschmar H. Raeber A. Braun U. Ehrensperger F. Hornemann S. Glockshuber R. Riek R. Billeter M. Wuthrich K. Oesch B. Nature. 1997; 390: 74-77Crossref PubMed Scopus (538) Google Scholar, 25Reineke U. Sabat R. Misselwitz R. Welfle H. Volk H.D. Schneider-Mergener J. Nat. Biotechnol. 1999; 17: 271-275Crossref PubMed Scopus (55) Google Scholar, 26Prodinger W.M. Schwendinger M.G. Schoch J. Kochle M. Larcher C. Dierich M.P. J. Immunol. 1998; 161: 4604-4610PubMed Google Scholar) have described the mapping of discontinuous epitopes using synthetic peptides. Those reports have all described antibodies that are able to bind independently to at least two, non-overlapping peptides, indicating that the epitopes are discontinuous and that there is little interaction between the epitope fragments (binding sites) in the assembled epitope. In contrast, Spendlove et al. (27Spendlove L. Li L. Potter V. Christiansen D. Loveland B.E. Durrant L.G. Eur. J. Immunol. 2000; 30: 2944-2953Crossref PubMed Scopus (31) Google Scholar) described an antibody that, while able to bind independently to three peptides, showed increased binding when the peptides were mixed together. This synergistic binding is indicative of cooperativity between each of the binding sites within the assembled epitope. The discontinuous Guy's 13 epitope represents an extreme form of cooperativity. Under the denaturing and reducing conditions of immunoblotting Guy's 13 was shown to interact with SA I/II through two binding sites in an independent fashion (14van Dolleweerd C.J. Chargelegue D. Ma J.K.-C. Infect. Immun. 2003; 71: 754-765Crossref PubMed Scopus (26) Google Scholar). While this is not likely to be representative of the conditions under which Guy's 13 would be expected to encounter cell-surface SA I/II, it nevertheless proved invaluable in establishing the discontinuous nature of the epitope. In fluid phase immunoassays, cooperativity between the binding sites was essential for Guy's 13 binding to take place. Taken together the results indicate that the two binding sites interact with each other to form the assembled epitope and that Guy's 13 interacts with both binding sites within the assembled epitope. Structural analysis of proline-rich regions in other proteins (28Blundell T.L. Pitts J.E. Tickle I.J. Wood S.P. Wu C.-W. Proc. Natl. Acad. Sci. U. S. A. 1981; 78: 4175-4179Crossref PubMed Scopus (294) Google Scholar, 29Darbon H. Bernassau J.-M. Deleuze C. Chenu J. Roussel A. Cambillau C. Eur. J. Biochem. 1992; 209: 765-771Crossref PubMed Scopus (77) Google Scholar), where proline is present as every third residue, suggests that these regions adopt a conformation with 3 residues per turn, known as the polyproline II (PP II) helix, with a 3.1-Å rise/residue (30Adzhubei A.A. Sternberg M.J.E. J. Mol. Biol. 1993; 229: 472-493Crossref PubMed Scopus (434) Google Scholar). By analogy it is predicted that the P-region of SA I/II, which also contains frequently repeating (PXX)n motifs, may also adopt a PP II helix-like conformation (31Kelly C. Evans P. Ma J.K.C. Bergmeier L.A. Taylor W. Brady L.J. Lee S.F. Bleiweis A.S. Lehner T. Arch. Oral Biol. 1990; 35 (suppl.): 33S-38SCrossref PubMed Scopus (19) Google Scholar). Feng et al. (32Feng S. Chen J.K. Yu H. Simon J.A. Schreiber S.L. Science. 1994; 266: 1241-1247Crossref PubMed Scopus (744) Google Scholar) have proposed a model for the PP II helix of Src homology 3 ligands. This model, adapted for peptide p955 and depicted in Fig. 6 is consistent with the ASM results. The critical, as defined by ASM, acidic (Glu959) and tyrosine (Tyr962) residues lie in close proximity on one edge of the proposed PP II helix model. The residues that make a smaller entropic contribution lie in the lower plane, while residues that form the upper edge appear to make little contribution to peptide p955 function. The results of this study have demonstrated that A-region sequences of SA I/II are able to interact with P-region sequences. Thus, we hypothesize that the PP II-like helix of the proline-rich region interacts with the alanine-rich region, which itself is predicted to adopt an α-helical structure (5Kelly C. Evans P. Bergmeier L. Lee S.F. Progulske-Fox A. Harris A.C. Aitken A. Bleiweis A.S. Lehner T. FEBS Lett. 1989; 258: 127-132Crossref PubMed Scopus (92) Google Scholar, 33LaPolla R.J. Haron J.A. Kelly C.G. Taylor W.R. Bohart C. Hendricks M. Pyati J.P. Graff R.T. Ma J.K.C. Lehner T. Infect. Immun. 1991; 59: 2677-2685Crossref PubMed Google Scholar, 34Demuth D.R. Golub E.E. Malamud D. J. Biol. Chem. 1990; 265: 7120-7126Abstract Full Text PDF PubMed Google Scholar, 35Demuth D.R. Irvine D.C. Infect. Immun. 2002; 70: 6389-6398Crossref PubMed Scopus (23) Google Scholar). The interaction of PP II-like helices with α-helices has been described before for turkey pancreatic polypeptide, a 36-amino acid hormone (28Blundell T.L. Pitts J.E. Tickle I.J. Wood S.P. Wu C.-W. Proc. Natl. Acad. Sci. U. S. A. 1981; 78: 4175-4179Crossref PubMed Scopus (294) Google Scholar, 36Glover I. Haneef I. Pitts J. Wood S. Moss D. Tickle I. Blundell T. Biopolymers. 1983; 22: 293-304Crossref PubMed Scopus (192) Google Scholar), where residues 2 to 8 form a PP II-like helix that is closely packed against an α-helix formed from residues 14–32. Homologous polypeptides from the pancreas of a wide range of species have been isolated and sequenced, and despite sequence variation, secondary structure analyses (37Glover I.D. Barlow D.J. Pitts J.E. Wood S.P. Tickle I.J. Blundell T.L. Tatemoto K. Kimmel J.R. Wollmer A. Strassburger W. Zhang Y.-S. Eur. J. Biochem. 1984; 142: 379-385Crossref PubMed Scopus (138) Google Scholar) and immunochemical studies (38Taylor T.C. Thompson D.O. Ebner K.E. Kimmel J.R. Rawitch A.B. Mol. Immunol. 1988; 25: 961-973Crossref PubMed Scopus (3) Google Scholar) indicate that the tertiary structure involves the interaction of the PP II- and α-helices. This idea is now proposed here as the structural basis for the recognition of SA I/II by Guy's 13. Thus, hydrophobic interactions between the proline side chains of the PP II-helical P-region and side chains from the α-helical A-region permit the assembly of the Guy's 13 epitope. Thus, although the Guy's 13 epitope is formed from two discontinuous binding sites within SA I/II, each site has a conformational requirement determined by its secondary structure. Recently, the crystal structure of a recombinant protein spanning the variable (V) region of antigen I/II (SR) from S. mutans serotype f strain OMZ 175 has been determined (39Troffer-Charlier N. Ogier J. Moras D. Cavarelli J. J. Mol. Biol. 2002; 318: 179-188Crossref PubMed Scopus (53) Google Scholar). While this recombinant protein does not encompass the A- and P-region binding sites defined in our studies, it does include some flanking alanine-rich and proline-rich sequences. The structure shows that the alanine-rich sequences form an α-helix that, indeed, lies in close proximity to the proline-rich sequences. The structure also suggests that the A- and P-regions are orientated in an antiparallel configuration. Juxtaposition of the A- and P-regions is also suggested by our present report as well as our earlier work (14van Dolleweerd C.J. Chargelegue D. Ma J.K.-C. Infect. Immun. 2003; 71: 754-765Crossref PubMed Scopus (26) Google Scholar). In this regard it is interesting to note that one of the A-region 82 aa repeats, with a predicted α-helical structure, has approximately the same length (123 Å)as one of the P-region 39 aa repeats, with a proposed PP II-helical structure (121 Å), suggesting that an interaction between the two regions might require the longitudinal association of a PP II helix with an α-helix. If this is true, it may explain why an A-region binding site could not be localized to a 20-mer peptide. The minimal P-region binding site was defined by the 14-residue sequence XPX(E/D)PXYXXXPXPP, which if it adopts a PP II helix conformation, would have a length of 43.4 Å. This distance can only be spanned by an α-helical peptide of ≥29 residues. The binding of antibodies, such as Guy's 13, to S. mutans cell surface SA I/II might block adhesion epitopes by exerting their effect locally (by steric hindrance) or at a distance (through induced conformational changes in the SA I/II molecule). Such antibodies, when used as a passive mucosal vaccine, could affect the adhesin function of the protein, leading to prevention of colonization of the oral cavity by the bacterium. These studies, which begin to map the interaction of Guy's 13 with SA I/II at a molecular level, represent an initial step in understanding these mechanisms. We gratefully acknowledge Karen Homer for assistance with mass spectrometric analysis of the peptides.