Species-specific binding of human secretory component to SpsA protein of Streptococcus pneumoniae via a hexapeptide motif

Authors


Abstract

SpsA, a pneumococcal surface protein belonging to the family of choline-binding proteins, interacts specifically with secretory immunglobulin A (SIgA) via the secretory component (SC). SIgA and free SC from mouse, rat, rabbit and guinea-pig failed to interact with SpsA indicating species-specific binding to human SIgA and SC. SpsA is the only pneumococcal receptor molecule for SIgA and SC as confirmed by complete loss of SIgA and SC binding to a spsA mutant. Analysis of recombinant SpsA fusion proteins showed that the binding domain is located in the N-terminal region of SpsA. By the use of different truncated N-terminal SpsA fusion proteins, the minimum binding domain was shown to be composed of 112 amino acids (residues 172–283). The sequence of this 112-amino-acids domain was used to spot synthesize 34 overlapping peptides, consisting of 15 amino acids each, with an offset of three amino acids on a cellulose membrane. One of the peptides reacted specifically with both SIgA and SC. By using a second membrane with immobilized synthetic peptides of decreasing length containing parts of the identified 15-amino-acid motif a hexapeptide, YRNYPT was identified as the binding motif for SC and SIgA. SpsA proteins with a size smaller than the assay-positive domain of 112 amino acids were able to inhibit the interaction of SIgA and pneumococci provided they contained the binding motif. The results indicated that the hexapeptide YRNYPT located in SpsA of pneumococcal strain type 1 (ATCC 33400) between amino acids 198 and 203 is involved in SIgA and SC binding. Because synthetic peptides containing only parts of the hexapeptide also assayed positive, these results further suggest that at least the amino acids YPT of the identified hexapeptide are critical for binding to SC and SIgA. Amino acid substitutions in the identified putative binding motif abolished SC-/SIgA-binding activity of the mutated SpsA protein, confirming the functional activity of this hexapeptide and the critical role of the amino acids YPT in SC and SIgA binding. Identification of this motif, which is highly conserved in SpsA protein among different serotypes, might contribute towards a new peptide based vaccine strategy.

Introduction

Streptococcus pneumoniae is an important human pathogen that colonizes the upper respiratory tract thereby causing invasive diseases, such as meningitis and sepsis (Tuomanen et al., 1995). In spite of the availability of antibiotics, mortality due to pneumococcal infection remains high, indicating a need for an effective vaccine (Gillespie, 1989). The virulence factors of S. pneumoniae include capsular polysaccharides and choline-binding proteins (ChBPs). The ChBPs are displayed on the cell surface and belong to a family of proteins characterized by their non-covalent anchoring to the surface via a choline-mediated interaction of the carboxy-terminal located repeat units with the teichoic and lipoteichoic acids of the bacteria (Yother and White, 1994). The most well-known members of this pneumococcal protein family are the LytA, the major autolysin of pneumococci (Garcia et al., 1986), PspA, the lactoferrin-binding protective protein (Yother and Briles, 1992; Hammerschmidt et al., 1999) and SpsA, the secretory immunglobulin A (SIgA)-binding protein (Hammerschmidt et al., 1997). Because of their biological activities, all of these proteins represent important virulence factors (Paton, 1998). SIgA represents the major adaptive mucosal defence factor against infectious agents (Underdown and Schiff, 1986) and therefore, the interaction of SpsA with SIgA might have important biological consequences. The SIgA precursor, an IgA dimer joined by the J chain, is produced by local mucosal and glandular plasma cells in which the addition of the J chain occurs just before secretion (Parkhouse and Della Corte, 1973; McCune et al., 1981). To interact with pathogens on the luminal side of the mucosa, the complex of the J chain with dimeric IgA (Vaerman et al., 1998) has to be transported across the epithelium barrier by the polymeric Ig receptor (pIgR). Upon binding to the pIgR on the epithelial basolateral surface, a process of endocytosis, phosphorylation, transcytosis and proteolysis is initiated (Mostov, 1984). This results in release of the secretory component (SC) either in a free form or complexed to pIg as SIgA at the apical surface into the secretions.

One problem of using surface proteins of S. pneumoniae as vaccine candidates is the serotype specific sequence variation. This problem can be overcome by using functional domains instead of whole proteins. Therefore identification of functional domains has direct implications in vaccine development. Despite a variation of SpsA among different pneumococcal serotypes, expression of SpsA and binding to SC was observed in at least 73% of S. pneumoniae isolates indicating a highly conserved binding domain for SC (Hammerschmidt et al., 1997). Characterization of these conserved functional domains and their corresponding synthetic peptides might lead to development of vaccines effective against many serotypes. In the present study, the species-specific SC-mediated binding of human SIgA by SpsA was confirmed and a hexapeptide motif containing the amino acid sequence required for SC binding was identified using spot-synthesized synthetic peptides and competitive inhibition experiments with purified recombinant proteins. The SC-/SIgA-binding function of the identified motif was validated using site-directed mutagenesis in the hexapeptide motif of the recombinant SpsA protein SH2. The results indicated that the motif Y(H/R)RNYPT is required for SC and SIgA binding by SpsA of S. pneumoniae.

Results

Specificity of the SpsA protein for human SC and human SIgA

Binding of human SC and SIgA to SpsA encoded by pQSH1 was compared with the binding of SC and SIgA purified from mouse, rat, rabbit and guinea-pig respectively. Using the dot-blot technique, an interaction of purified SpsA fusion protein expressed by pQSH2 with human SC was observed with 5 ng of spotted SpsA and a concentration of 0.8 µg ml−1 SC. When using 0.16 µg ml−1 SC, the SpsA protein reacted only with spots from 10 ng upwards. In contrast, no interaction of SpsA with SIgA or SC from rat bile and rabbit, mouse and guinea-pig milk, respectively, was observed using Western blot analysis with SpsA run on SDS–PAGE and 10 µg ml−1 each SIgA or SC (Table 1). Performing reverse Western blots with 30 µg of applied human SC, rat SC and rat SIgA, respectively, and 1 µg ml−1 SpsA fusion protein, only human SC showed reactivity (Table 1). In reverse Western blots, no reactivity was observed with SIgA and SC from mouse, rabbit, and guinea-pig, indicating a species-specific reaction of SpsA with human SC (Table 1). When submitting mixtures of radioiodinated SpsA with a 12-fold molar excess of human SIgA to sucrose gradient ultracentrifugation, 87% of the radioactivity sedimented with two fast peaks at positions corresponding to those of human 11S dimeric SIgA and heavier SIgA, with only 13% at a slowly sedimenting peak corresponding to that of free SpsA, on the top of the gradient. Similar mixtures of labelled SpsA with rat, rabbit, mouse and guinea-pig SIgA yielded radioactivity elution profiles similar to that of labelled SpsA alone, as did mixtures of labelled SpsA with human and rat serum dimeric IgA, which lack SC (data not shown). These data further confirm the species specificity of the reaction of SpsA with human SIgA.

Table 1. Binding of SpsA to SIgA and SC from various species.
  Methods 

Species

Protein

Dot blota
Western
blotb
Reverse
Western blotc

SDGUd
 
  • a

    . SpsA (1 µg) applied on the membrane; after blocking, reaction with SIgA or SC, followed by appropiate biotinylated anti-IgA or anti-SC Ab (1/1000–1/5000), then HRP-or alkaline-phosphatase coupled Extravidin.

  • b

    . SpsA (10 µg) run on SDS–PAGE, electrotransfer to nitrocellulose, blocking, reaction with SIgA or SC, followed by appropiate biotinylated anti-IgA or anti-SC Ab (1/1000–1/5000), then HRP-or alkaline-phosphatase coupled Extravidin.

  • c . SIgA (25–50 µg) or SC (5–10 µg) run on 4–15% SDS–PAGE, electrotransfer to nitrocellulose, blocking, reaction with SpsA (20 µg ml −1), then rabbit anti-SpsA (1/500), then biotynilated goat anti -rabbit Ig Ab absorbed with appropiate (1/2000), then HRP-or alkaline-phosphatase coupled Extravidin.

  • d

    . SDGU, Sucrose density gradient ultracentrifugation of mixture of radiolabelled SpsA with SIgA (molar ratio 1:12.5). If bound to SigA, radiolabelled SpsA sediments run much faster than free labelled SpsA.

  • e

    . NA, not applicable, because free SC is not well separated from free labelled SpsA by SDGU.

  • f

    . ND, not done, because biotinylated goat anti-rabbit Ig Ab reacts with rabbit SIgA.

HumanSIgA++++ 
HumanSC+++NAe 
RabbitSIgANDf 
RabbitSCNAe 
RatSIgA 
RatSCNAe 
MouseSIgA 
MouseSCNAe 
Guinea-pigSIgA 
Guinea-pigSCNAe 

Binding of SIgA and SC to spsA mutants

To confirm that the interaction between pneumococcal cells and SIgA is only mediated through its binding to the SpsA protein, genetic mutants were constructed. Functional inactivation of SpsA of pneumococcal strains type 35A (NCTC 10319), type 2 (ATCC 11733) and R6x was achieved by gene disruption as described in detail in Experimental procedures. Analysis of SIgA binding performed by a bacterial binding assay with soluble radioiodinated SIgA and by Western blot analysis revealed complete loss of the binding activity in the spsA mutants (Fig. 1B and C). Furthermore, mutants also failed to react with anti-SpsA antiserum (data not shown). These results indicated that SpsA protein is the sole SC binding component of S. pneumoniae and confirms our earlier results using Scatchard analysis (Hammerschmidt et al., 1997).

Figure 1.

Molecular and functional analysis of pneumococcal spsA-deficient mutants.

A. Southern blot analysis of genomic DNA digested with HindIII and hybridized with a DIG-labelled PCR fragment (SH39–SH36) of spsA. The binding pattern of the mutants showed the integration of the vector pJDC9 into the chromosome and a HindIII DNA fragment at 0.94 kb as a result of the chromosomal duplicated spsA fragment, which was BamHI–HindIII subcloned in pJDC9.

B. Immunoblot analysis of SIgA binding to pneumococcal wild types and corresponding spsA mutants. Lanes: 1, S. pneumoniae type 35A (NCTC10319); 2, NCTC 10319 spsA mutant SPMU37 (pJDC9::spsA); 3, R6x; 4, R6spsA. The results obtained for S. pneumoniae type 2 (ATCC 11733) and its corresponding mutant SPMU51 (pJDC9::spsA) were identical.

C. Reactivity of spsA-deficient mutants with SIgA. The binding assay was performed with 125I-labelled SIgA to S. pneumoniae strains type 35A (NCTC10319), its corresponding spsA-mutant SPMU37 (pJDC9::spsA), R6x and R6spsA. The values represent the means and ranges from triplicate experiments. Binding of radiolabelled SIgA to SPMU51 (pJDC9::spsA) was also negative. Binding of 100% represents the total binding of radiolabelled SIgA to fetal calf serum.

Subcloning of the human SC-/SIgA-binding sequence of the SpsA protein

Binding of human SIgA to the pneumococcal surface protein SpsA is mediated by SC. In our previous study, the human SC-binding region was localized in the N-terminal part of SpsA. The residues between amino acids 160 and 324 of the SpsA protein of pneumococcal type 1 strain ATCC 33400 encoded by plasmid pQSM1 represented the smallest binding region identified so far (Hammerschmidt et al., 1997). In the present study, the sequence encoding the SC-binding domain was further subcloned. Derivatives of pQSM1 with N-terminal as well as carboxyl-terminal deletions were constructed by cloning PCR-amplified spsA fragments into expression vector pQE (Fig. 2A and B). Deleting the sequence coding for the proline-rich domain resulted in plasmid pQSM2, a subclone expressing a protein of 14.3 kDa. This protein and a 12.9 kDa protein encoded by pQSM5, a subclone that expresses amino acids 173–283 of SpsA, were both positive for SC and SIgA binding by immunoblot analysis (Fig. 2C). Truncation of the SpsA protein in the amino-terminal part up to amino acid 214 resulted in pQSM3 and pQSM4, which express amino acids 215–324 and amino acids 215–283 of SpsA respectively. However, the resulting proteins of 12.7 kDa and 7.9 kDa expressed by pQSM3 and pQSM4 interacted neither with SC nor with SIgA (Fig. 2C). These results indicate that the SC-binding motif must be located between amino acids 173 and 283 of SpsA. Therefore, further recombinant SpsA clones expressing amino acids 160–257 (pQSM6), amino acids 173–257 (pQSM7), amino acids 173–225 (pQSM8) and amino acids 192–257 (pQSM9) were constructed (Fig. 2A and B). The resulting purified hybrid proteins encoded by the cloned spsA sequences were also assayed by immunoblotting for their ability to bind SC and SIgA. However, none of the recombinant proteins showed binding to SIgA under denaturing (Fig. 2C) or native conditions (results not shown). Although the recombinant clones expressing SpsA fusion proteins were constructed in a manner such that binding to at least one of the hybrid proteins should be positive, analysis of the SIgA and SC overlay assay indicated a decreasing affinity of both SIgA and SC to SpsA proteins with decreasing molecular masses. Although SpsA protein SM6 differs by only 1.7 kDa in the molecular weight from SpsA protein SM5, SM6 showed only a very low reactivity with SIgA (not shown) but not SC (Fig. 2C). However, SpsA hybrid proteins with lower molecular masses than SM6 did not react at all with either SIgA or SC (Fig. 2C).

Figure 2.

Binding of human SC to recombinant clones analysed by Western blotting. Schematic representation of the SpsA fusion proteins and their binding activity to human SC as analysed by immunoblotting (A). Coomassie-stained purified SpsA fusion proteins (B). Immunoblot analysis with human SC (C). Lanes: 1, pQSH1 expressing amino acids 38–324; 2, pQSH2, expressing amino acids 38–283; 3, pQSH3, expressing amino acids 38–159; 4, pQSM1, expressing amino acids 160–324; 5, pQSM2, expressing amino acids 160–283; 6, pQSM3, expressing amino acids 215–324; 7, pQSM4, expressing amino acids 215–283; 8, pQSM5, expressing amino acids 173–283; 9, pQSM6, expressing amino acids 160–257; 10, pQSM7, expressing amino acids 173–257; 11, pQSM8, expressing amino acids 173–225; 12, pQSM9, expressing amino acids 192–257.

Analyses of SC/SIgA binding to spot synthesized peptides

Based on the results obtained from the recombinant SpsA proteins analysed by immunoblotting, spot-synthesized peptides were used to map the human SC-/SIgA-binding motif in SpsA protein. For this purpose, the 112 amino acids that are expressed by pQSM5 were divided into 34 overlapping peptides, consisting of 15 amino acids each, with an offset of three amino acids. The synthetic peptides were assayed for their capability to bind purified human SC and human SIgA. The results showed that only a single spot (Fig. 3A), the sequence of which is also shown in Fig. 3A, among the immobilized synthetic peptides interacted with both human SC and SIgA (Fig. 3A). Therefore, a second spot membrane with an offset of one amino acid and peptide length from six amino acids up to 15 amino acids residues per spot was prepared. The synthetic peptides contained at least parts of the 15 amino acid sequence that was assayed positive for SC/SIgA binding. The results showed strong binding of a six-amino-acid peptide with the sequence YRNYPT to both SC and SIgA (Fig. 3C). This sequence is located in SpsA of pneumococcal strain type 1 (ATCC 33400) between amino acids 198 and 203. Furthermore, synthetic peptides extended at the C-terminal end but starting with this hexapeptide were also scored positive (Fig. 3C and D). In addition, binding of SC and SIgA, respectively, was also positive to synthetic peptides starting with the arginine or asparagine of the hexapeptide and to one synthetic peptide with the sequence YPTITYKT (Fig. 3C and D). Synthetic peptides, containing amino acids preceding the hexapeptide sequence, did not exhibit binding, confirming the results of the first spotted membrane, with only one spot assaying positive. Densitometric analysis revealed that in this hexapeptide the sequence YPT in the binding motif is critical for binding to SC and SIgA respectively (Fig. 3C and D).

Figure 3.

Determination of the hexapeptide motif YRNYPT required for SIgA and SC binding in SpsA.

A. Spot membrane of the 112 amino acids encoded by pQSM5 divided into 34 overlapping peptides of 15 amino acids each, with an offset of three amino acids, analysed with SIgA. Arrow shows the only positive spot reacting with SIgA.

B. Amino acid sequence used for the synthesis of the spot membrane in C.

C. Spot membrane with 145 overlapping peptides, with peptide length from six amino acids up to 15 amino acids residues per spot analysed with SIgA.

D. Analysis of the spot membrane in C showing peptides recognized by SIgA. The quantitative analysis was carried out using densitometric measurement and resulted in relative intensity of the spots. The spot with the highest intensity was set to 100%.

The sequence locus encoding the SpsA protein was identified independently by other research groups and designated cbpA (Rosenow et al., 1997), pbcA (Hostetter et al., 1997) and pspC (Brooks-Walter et al., 1999) respectively. Comparison of amino acid sequences deposited in the database (Accession nos Y10818, AJ002054, AJ002055, AF019904, AF067128, U72655, AF068645, AF068646, AF068647, AF068648, AF068649 and AF068650) and SpsA sequences of clinical isolates among different serotypes (Hammerschmidt et al., 1998) indicated that the first tyrosine of the hexapeptide could be exchanged by an arginine or histidine and that the isoleucine following the hexapeptide could be also exchanged by a asparagine, as shown for strains that are capable of binding SC and SIgA (Hammerschmidt et al., 1998). Using computer-aided secondary structural analysis (Garnier et al., 1978) the hexapeptide identified as the binding motif for SC and SIgA, respectively, was located in a turn between α-helical structures in the SpsA protein (data not shown). These results strengthened the idea that the function of the hexapeptide is to act as the binding motif for SC, because this sequence should thus be accessible for the SpsA-SC/SIgA interaction.

Competitive inhibition of human SIgA binding to pneumococci by a synthetic peptide and recombinant SpsA proteins

As already described, purified N-terminal His-tagged SpsA protein (amino acids 38–324) that is encoded by pQSH1 and lacking the choline-binding repeats could completely inhibit the binding of human SIgA to pneumococci (Hammerschmidt et al., 1997). To verify the results obtained by epitope mapping with the spot membranes, a 15 amino acid synthetic peptide recognized by SC and SIgA was synthesized and used to competitively inhibit binding of SIgA to pneumococcal cells. However, this peptide in concentrations up to a 6250-fold molar excess over radiolabelled SIgA was not able to inhibit the binding of SIgA to pneumococci (data not shown). Because the conformation of the applied peptide may be important for binding, recombinant SpsA proteins purified by affinity chromatography were used to inhibit competitively binding of SIgA to pneumococci. The SpsA proteins selected for this assay were expressed by pQSH2, which encodes for a protein similar to pQSH1 except the proline rich domian of SpsA, pQSM2, pQSM4, pQSM5 and pQSM8. The SpsA protein encoded by pQSM8 was used as the representative of protein SM6–SM9. Out of these the protein encoded by pQSM4 did not contain the mapped binding motif. As expected, purified SpsA protein encoded by pQSM4 could not competitively inhibit binding of SIgA to pneumococci (Fig. 4A). In contrast, all SpsA fusion proteins that contain the SIgA-binding motif were able to competitively inhibit SIgA binding to pneumococci of different serotypes regardless of their molecular masses (Fig. 4A). However, the strength of inhibition depended on the length and molecular mass of the fusion proteins because the concentrations of unlabelled SpsA fusion proteins causing 50% displacement of 125I-labelled SIgA increased from 0.28 nmol for the protein expressed by pQSH2 to 85.8 nmol for the protein expressed by pQSM8 (Fig. 4B). The SM8 protein is a 53-amino-acid-long SpsA protein (aa 173–225) with a molecular mass of 6.2 kDa. In comparison, the concentration of unlabelled SIgA necessary to reach 50% blocking of binding of the radiolabelled SIgA to pneumococci was 17.3 nmol (Hammerschmidt et al., 1997). These results confirmed the identified motif from the spot membranes, but they also suggested that the affinity of SpsA fusion proteins for SIgA depends on the length of the proteins as well as on the surrounding amino acids and therefore on structure of the SpsA protein.

Figure 4.

Competitive inhibition of human SIgA binding to pneumococci by selected recombinant SpsA fusion proteins. Recombinant protein SM8 was used as representative of protein SM6 through SM9.

A. Inhibition was performed with 125I-labelled SIgA in the presence of increasing concentrations of unlabelled purified SpsA hybrid proteins. B. Concentration of the different SpsA proteins required for 50% inhibition of 125I-SIgA binding to S. pneumoniae type 1 (ATCC 33400) and type 35A (NCTC 10319) respectively.

Effect of site-directed mutagenesis of SpsA on SIgA-binding activity

To validate the effect of amino acid substitutions in the proposed SC-/SIgA-binding motif YRNYPT (amino acids 198–203) of SpsA, recombinant SpsA protein SH2 encoded by pQSH2 was selected for individual amino acid substitutions. Residues 200–202 were selected for individual aspartic acid (Asp) and glutamic acid (Glu), respectively, substitutions. The sequences of all mutated spsA DNAs were confirmed by DNA sequencing. The resultant recombinant and mutated SpsA proteins were expressed and the correct molecular weight of the induced proteins was visualized by Coomassie brilliant blue staining (Fig. 5A). The mutated SpsA proteins also reacted with anti-SpsA antiserum in immunoblot analysis (Fig. 5B). To determine the effect of single Asp or Glu substitutions on SIgA-binding activity, the binding of SIgA to mutated SpsA proteins was investigated by immunoblot analysis. The recombinant SpsA200 protein with a single Asp substitution at residue 200 showed no alteration in its ability to bind SIgA (Fig. 5C). In contrast, SpsA201 protein with an Asp substitution at amino acid 201 and SpsA202 protein with a Glu substitution at amino acid 202 completely abolished the binding to SIgA (Fig. 5C). These results verified the function of the sequence YRNYPT as the SC-/SIgA-binding motif in SpsA and furthermore, supported the critical role of the sequence YPT in the SIgA-binding activity.

Figure 5.

Effect of Asp and Glu substitutions in the proposed SC-/SIgA-binding motif YRNYPT (amino acids 198–203 in parental SpsA) on SIgA binding of recombinant SpsA(200−202) proteins.

A. Sequence of recombinant SpsA proteins with individual Asp substitution at position 200–201 and Glu substitution at position 202. B. Coomassie staining of recombinant SpsA protein SH2 and recombinant SpsA(200−202) proteins.

C. Human SIgA binding activity analysed by immunoblotting.

D. Immunoblot analysis with anti-SpsA antiserum SH2. Lanes: Purified His-tagged fusion protein SH2 (1), lysates of the recombinant clones expressing SpsA200−202 with Asp substitution at residue 200 (2), Asp substitution at residue 201(3), and with Glu substitution at residue 202 (4).

Discussion

S. pneumoniae is an important human pathogen capable of causing relatively harmless otitis media as well as life-threatening pneumonia and meningitis (Musher, 1992). How the pneumococci counteract the immune defence on mucosal surfaces and make the transition from commensal to an invasive pathogen is still not well known. One mechanism could be the existence of phase variants with a transparent and opaque phenotype (Weiser et al., 1994) that correlates with the capability to survive in different host niches due to an altered amount of C-polysaccharide (Kim and Weiser, 1998) and an altered distribution of choline-binding proteins (ChBPs) displayed on the surface of S. pneumoniae (Rosenow et al., 1997). Therefore, ChBPs might play an important role in the pneumococal pathogenesis.

Among the known ChBPs, the SpsA protein exhibits a unique interaction with human free SC and the SC part of human SIgA (Hammerschmidt et al., 1997). Because this binding occurs via the SC, SpsA does not interact with serum IgA. Many bacterial species interact with serum IgA (Jerlström et al., 1996), but SpsA is unique because of its interaction with SC. Recently, it was shown that Clostridium difficile toxin A binds also to the human milk secretory component and that deglycosylated SC binds much less than glycosylated SC (Dallas and Rolfe, 1998). SIgA displays important biological functions such as inhibition of microbial adherence (Williams and Gibbons, 1972), neutralization of toxins and enzymes, inhibition of antigen penetration through mucosal surfaces, opsonization for mucosal polymorphonuclear leucocytes, mediation of monocyte dependent bactericidal activity and antibody dependent cellular toxicity (Kilian et al., 1988). Therefore, the specific interaction of SIgA, a glycoprotein carrying a high load of N- and O-linked oligosaccharides (Baenzinger and Kornfeld, 1974; Hughes et al., 1999), with pneumococci, can have biological consequences. The species-specific binding of SpsA to human SIgA and human SC as shown in this study further underlines the importance of this interaction for S. pneumoniae, which is an exclusively human specific pathogen. Besides binding to SIgA, SpsA has also been shown to contribute to adherence of pneumococci to human epithelial as well as to endothelial cells (Rosenow et al., 1997), and more recently, to elicit cross-reactive antibodies to PspA and to provide immunity against pneumococcal bacteremia (Brooks-Walter et al., 1999). S. pneumoniae spsA-deficient mutant not only showed reduced adherence but also significant reduced binding to purified immobilized glycoconjugates 6′sialyllactose and lacto-N-neotetraose (LNnT), which constitute receptors for pneumococci on cytokine-activated cells (Cundell et al., 1995). In spite of the fact that spsA-deficient mutants exhibit reduced capability to bind to these glycoconjugates, no competitive inhibition of radiolabelled SIgA binding to different pneumococci with increasing concentrations of soluble sialic acid and LNnT was observed (data not shown). This is consistent with the observation of Falk et al. (1993) that 6′sialyllactose and LNnT are not able to competitively inhibit binding of SIgA to Helicobacter pylori. The discrepancy, however, between reduced binding to the glycoconjugates by the spsA-deficient mutants and the failure of inhibition of SIgA binding to pneumococci by soluble glycoconjugates remains unresolved, but could be due to structural changes of the immobilized glycoconjugates. As a result of specific binding of SpsA to SigA, which is proposed to influence the adherence of bacteria to the mucosal surface by specific binding to and by effecting aggregation (Williams and Gibbons, 1988) as well as the production of an IgA1-protease (Male, 1979), both these pneumococcal factors might be involved in pneumococcal protection against the immune defence on mucosal surfaces.

The ChBPs, PspA and SpsA, represent important protective antigens (Paton, 1998). SpsA molecules are divided into two clades designated A and B that differ in one N-terminal region, region 2, that is only present in clade A (Brooks-Walter et al., 1999). PspA has an analogous sequence sharing homology to region 2. However, both proteins show antigen variation among different pneumococcal strains (Crain et al., 1990; Hammerschmidt et al., 1997; Brooks-Walter et al., 1999). Because there are more than 90 different serotypes, their use as vaccine candidates seems to be limited. However, more than 73% of pneumococcal strains belonging to different serotypes interact with SIgA (Hammerschmidt et al., 1997) and, as shown in this study by the absence of SIgA binding activity in spsA-deficient mutants, SpsA is the sole pneumococcal receptor for SIgA. Taken together, these results suggest that the SC/SIgA-binding domain in SpsA proteins is strongly conserved in different serotypes of S. pneumoniae. In order to map the exact binding domain in SpsA, two different approaches were used, namely the construction of truncated derivatives of SpsA and the peptide spot synthesized membrane. By immunoblot analysis of the reaction of the truncated derivates of SpsA with SC and SIgA we were able to narrow down the binding domain to 112 amino acids. Further mapping was performed using spot-synthesized membranes. This method allows simultaneous screening of large numbers of different peptides (Frank, 1992). Because the ligands can be stripped off from the membrane, it can be reused for the same or a different ligand. Analysis of the membrane with 34 peptides, consisting of 15 amino acids each, with an offset of three amino acids resulted in only one positive spot for SIgA-binding activity. This result indicates that the positive spot represents the peptide which contains a SIgA specific binding motif. Spot synthesis of further peptides based on the positive spot with an offset of one amino acid led to the identification of a motif YRNYPT located between amino acid 198 and 203 in SpsA of S. pneumoniae type 1 (ATCC 33400). Densitometric analysis suggested that not all of the amino acids in the identified hexapeptide are necessary for binding of SC and SIgA but that the sequence YPT in the binding motif is critical for binding of SC and SIgA. Furthermore, the analysis of the second spot membrane strongly indicated that the flanked amino acids also play a pivotal role in the binding to SIgA. Spots containing amino acids preceding the hexapeptide sequence exhibited no binding to SC and SIgA, suggesting that the negatively charged amino acids abolished the binding. Sequence comparison and structural analysis (Garnier et al., 1978) confirmed the results obtained from the analysis of the spot membranes and competitive inhibition. In fact, sequence comparison showed that the tyrosine at position one of the hexapeptide is exchangeable for histidine or arginine (Hammerschmidt et al., 1997; Brooks-Walter et al., 1999). Furthermore, secondary structure analysis located the hexapeptide in a turn of a helix–turn–helix module of the SpsA molecule, leading to the assumption of an accessible sequence for the interaction with SC and SIgA respectively.

The validation of YRNYPT as the SC-/SIgA-binding motif of SpsA was obtained by site-directed mutagenesis within the spsA gene. The aspartic acid substitution of tyrosine (residue 201) and glutamic acid substitution of proline completely abolished binding to SIgA whereas the substitution of asparagine did not have any effect. These results clearly indicate that YRNYPT is indeed the binding motif and, as shown by spot analysis, the sequence YPT is critical for binding. However, in competitive inhibition experiments, comparison of SpsA mutant proteins, each with single amino acid substitutions or with simultaneously mutated amino acids in the binding motif, will show the quantitative effect on SIgA binding to pneumococci.

The binding data and the effects of amino acid substitutions in the motif were supported by inhibition data with recombinant fusion proteins of SpsA. All SpsA proteins containing the identified binding motif YRNYPT competitively inhibited the binding of SIgA to pneumococci whereas the peptide SM4 without this motif had no effect. However, a 15 amino acid synthetic peptide, also containing the binding motif, failed to inhibit binding, indicating that the affinity of SpsA fusion proteins to SIgA depends on the molecular mass as well as on the charge of the flanking amino acids. This is supported by the fact, that with decreasing length of recombinant SpsA proteins, higher concentrations are required to achieve 50% blocking of SIgA binding to pneumococci and that spot synthesized peptides with negatively charged amino acids preceding the hexapeptide YRNYPT showed no reactivity with SIgA. These findings are supported by others, which identified a motif MLKKIE located in the β-antigen of group B streptococci and recognized by serum IgA (Jerlström et al., 1996). Because the serum IgA binding motif MLKKIE is not present in SpsA this might be an explanation for the absence of SpsA reactivity with serum IgA. However, the identified binding motif is also absent in toxin A of Clostridium difficile, a recently identified protein capable of binding human milk SC (Dallas and Rolfe, 1998), suggesting that this interaction differs from the SpsA–SC interaction. Identification of a specific motif in SpsA for binding of human SIgA and human SC might also contribute towards vaccine formulation. Because this motif is present in SpsA protein from many pneumococcal strains of different serotypes, it could be important for a peptide-based conjugate vaccine effective against the different pneumococcal serotypes. Additionally, identification of YRNYPT could also improve analytical applications. The motif can be used to determine the occurrence of human SIgA in body fluids and also serves as a ligand for affinity purification of human SIgA and human SC.

Experimental procedures

Bacterial strains, media and growth conditions

Streptococcus pneumoniae were cultured in Todd–Hewitt broth (Oxoid) supplemented with 0.5% yeast extract (THY) to mid-log phase or grown on blood agar (BectonDickinson). Escherichia coli M15[pREP4] (Qiagen) was used as the host for recombinant pQE expression plasmids and cultured at 37°C on Luria–Bertani (LB) agar or grown on LB-agar containing 100 µg ml−1 ampicillin. Expression of the His-tagged fusion proteins was induced with 1 mM IPTG after the culture reached an OD600 of 0.8 and continued to grow at 30°C for 4 h.

Proteins and antisera

SIgA and free SC were purified as described earlier from milk or bile from humans (Kobayashi et al., 1973), rats (Vaerman et al., 1975a; Acosta Altamirano et al., 1980), mice (Lemaitre-Coelho et al., 1978; Pierre et al., 1995), rabbits (Delacroix et al., 1982) and guinea-pigs (Vaerman and Heremans, 1972; Vaerman et al., 1975b). Antiserum against SpsA was generated by Eurogentec. Rabbits were immunized subcutaneously with 50 µg of purified SpsA encoded by pQSH2 in 1 ml of 1:1 emulsion of buffer and complete Freund's adjuvant. The rabbits were boosted with 50 µg of the SpsA protein and incomplete Freund's adjuvant at days 14, 28 and 56. Preimmune serum was collected prior to immunization.

Competitive inhibition assay

Binding of radioiodinated ligands to pneumococci was performed as described previously (Hammerschmidt et al., 1997). In competitive inhibition experiments, the binding of 16 ng of 125I-labelled SIgA to 5 × 109 pneumococcal cells (type 1 strain ATCC 33400, type 35 A strain NCTC 10319) was determined in the presence of increasing concentrations of unlabelled purified SpsA protein. SpsA fusion proteins were used in concentrations from 0.01 ng up to 100 µg.

Electrophoresis and immunoblot analysis

Fusion proteins were subjected to sodium dodecyl sulphate polyacrylamide gel electrophoresis with 16% gels by the method described by Laemmli (1970) and either stained with Coomassie brilliant blue or subsequently transferred to a nylon membrane (Immobilon-P; Millipore), using a semidry blotting system (Bio-Rad). Binding of human SIgA was investigated using immunoblot analysis as described previously (Hammerschmidt et al., 1997). Binding of purified human free secretory component was detected with rabbit anti-SC antibody (diluted 1:1000 in PBS) followed by biotinylated goat anti-rabbit antibody (Dianova; 1:2000) and extravidin-peroxidase (Sigma; diluted 1:1000 in PBS) and a substrate solution containing 1 mg ml−1 4-chloro-1-naphthol and 0.1% H2O2 in PBS. Binding of purified SC and SIgA from mouse, rat, rabbit and guinea-pig was examined using dot-blot assay, Western blot and reverse Western blot analysis in which SC or SIgA was applied instead of SpsA fusion protein. Binding of SC and SIgA was detected using either polyclonal anti-SC or anti-IgA antibodies followed by biotinylated goat anti-rabbit and alkaline-phosphatase-conjugated extravidin. The binding activity was detected by a substrate solution containing 0.05 mg of 5-bromo-4-chloro-3-indolylphosphate and 1 mg ml−1 nitroblue tetrazolium in 100 mM Tris-HCl, pH 9.6 and 4 mM MgCl2.

Binding of radioiodinated SpsA to SIgA assayed by sucrose density gradient ultracentrifugation

SpsA (50 µg of SH1) was labelled with 0.5 mCi of carrier free 125iodine using the chloramin T method. Twenty picomoles of labelled SpsA (50 µl, about 4 500 000 c.p.m., > 99% precipitable in 10% trichloracetic acid) were mixed with 100 µg of human, rat, rabbit, mouse or guinea-pig SIgA (250 pmol), or with 100 µg of human or rat dimeric serum IgA, devoid of SC, or with PBS–BSA (0.25%) in a final volume of 150 µl. The mixtures were carefully layered on the top of isokinetic sucrose gradients (10 ml of 5–21% sucrose in PBS–BSA) and run for 18 h at 130000 g. After the run, the gradients were eluted from the bottom in 28 successive 0.35 ml fractions, which were counted in a gamma counter, as described elsewhere (Vaerman et al., 1998).

Recombinant DNA techniques and PCR

Transformation of E. coli with recombinant plasmids was achieved by electroporation (Calvin and Hanawalt, 1988). Plasmid DNA was isolated using the Qiagen Plasmid Kit (Qiagen) and PCR products were purified using the PCR Purification Kit (Qiagen). T4 DNA ligase and restriction enzymes were purchased from New England Biolabs and used according to the instructions of the manufacturer. The integrity of insert DNA was determined using ABI PRISM dye terminator cycle sequencing (Perkin-Elmer). PCR was performed on a thermocycler (Biozym) with 0.2 µg of chromosomal DNA as a template using the Ampli-Taq Gold™ polymerase under buffer conditions recommended by the manufacturer. Each reaction consisted of 35 cycles including 30 s of denaturation at 94°C, 30 s of annealing and an extension at 72°C.

Expression cloning procedure and sequencing

Primers incorporating a BamHI restriction site at the 5′ end and a HindIII site at the 3′ end were designed from the spsA sequence (EMBL database accession nos Y10818 and AJ002055) in order to amplify different fragments of spsA. Insert DNA was ligated into similarly digested expression vectors pQE30 and pQE31 to allow in frame expression of SpsA His-tagged fusion proteins. These proteins were purified under native conditions using affinity chromatography on Ni-nitrilotriacetic acid resins, according to the protocols of the manufacturer (Qiagen). Oligonucleotides were synthesized by Life Technologies. PCR oligonucleotides pairs used for construction of the recombinant spsA clones were as follows; SH39 (5′-GGATCCACAGAGAACGAGGGAAGTACCC-3′) and SH36 (5′-TTTTTCTTAAGCTTTATCTTCTTCTGC-3′) resulting in pQSH2; SH35 (5′-AACCGGATCCTTCAGATACAGCG-3′) and SH36 resulting in pQSM2; SH37 (5′-TTGAAATTGCTGGATCCGATGTGG-3′) and SH25 (5′-CTCAGCTATTAAGCTTTTTTGGAGTAGATGGTTGTGCTGG -3′) resulting in pQSM3; SH37 and SH36 resulting in pQSM4; SH43 (5′-CGACGGATCCAGAAAAAAAGGTAGC-3′) and SH36 resulting in pQSM5; SH35 and SH44 (5′-TTCTAAGCTTGTAGCCTCAGTTTCTTC-3′) resulting in pQSM6; SH43 and SH44 resulting in pQSM7; SH43 and SH46 (5′-GTTCAAGCTTCGCTTTTTTAACTTCC-3′) resulting in pQSM8; and SH45 (5′-CCAAGGATCCAAAAGAAGAAGATTACC-3′) and SH44 resulting in pQSM9. The recombinant plasmids pQSH1 and pQSM1 have been described in detail previously (Hammerschmidt et al., 1997). Subclone pQSH3, a derivative of pQSH1, was constructed through deletion of a 495 bp HindIII fragment.

Construction of spsA-deficient S. pneumoniae mutants

Three spsA-deficient mutants were constructed by insertion duplication as described previously (Morrison et al., 1984). Using the TOPO™ TA Cloning® Kit (Invitrogen), a 250 bp and a 514 bp fragment of spsA, corresponding to nucleotides 112–352 of spsA of S. pneumoniae strain type 2 (ATCC11733) or to nucleotides 112–626 of spsA of S. pneumoniae strain type 35A (NCTC10319) and R6x (Tiraby and Fox, 1973) respectively, were cloned individually into the pCR®–TOPO vector. The spsA fragments were generated by PCR with the oligonucleotides SH39 (5′-GGATCCACAGAGAACGAGGGAAGTACCC-3′) and SW5 (5′-AAGCTTCTGACGGCAACTCATCTTTCG-3′); and SH39 and SW9 (5′-AAGCTTGCTTGCTTAATTGTGCCCTCG-3′) respectively. In S. pneumoniae strain type 2 (ATCC11733), the 250 bp spsA fragment encodes amino acids 38–117. The second spsA fragment of 514 bp encodes the region for amino acids 38–208 in S. pneumoniae type 35A (NCTC 10319) and R6x. Insert DNA was digested with BamHI–HindIII and subcloned into the similarly digested integration vector pJDC9 and transformed by electroporation into E. coli DH5α. Transformants were selected on LB agar containing 250 µg ml−1 erythromycin, and plasmids were examined for pneumococcal insert DNA. S. pneumoniae mutants of the wild-type strains ATCC 11733, NCTC 10319 and R6x were obtained by transformation with plasmid DNA from a selected recombinant E. coli, using the competence stimulating peptide to enhance the transformation efficiency (Håvarstein et al., 1995). Transformants that did not express SpsA were detected by their lack of reactivity with SIgA and anti-SpsA antiserum in Western blot analysis. Integration of the vector disrupting the spsA gene was confirmed using PCR and Southern blot analysis (Fig. 1A). The mutants were designated SPMU51 (pJDC9::spsA) in the ATCC11733 background, SPMU37 (pJDC9::spsA) in the NCTC 10319 background and R6spsA in the R6x background respectively.

Site-directed mutagenesis of the spsA gene

Site-directed mutagenesis of the spsA gene was performed using the QuikChange™ site-directed mutagenesis kit (Stratagene) according to the manufacturer's instructions. Briefly, each mutant spsA DNA was generated using one pair of polyacrylamide gel electrophoresis-purified complementary oligonucleotide primers, containing GAC (code for aspartic acid) or GAA (code for glutamic acid) substitutions at the desired sites. Plasmid DNA pQSH2 (40 ng), which encodes the amino acids 38–283 of SpsA, was used as a DNA template. The resultant mutant spsA DNAs, obtained by temperature cycling using PfuTurbo DNA polymerase, were transformed into Epicurian Coli®XL1-blue supercompetent cells provided by the manufacturer. Expression of the mutated SpsA proteins was visualized using Coomassie brilliant blue staining and anti-SpsA antiserum in immunoblot analysis. Binding of human SIgA to mutated SpsA proteins was performed by immunoblot analysis as described previously (Hammerschmidt et al., 1997). The site-specific mutations were verified by sequence analysis using the ABI PRISM dye terminator cycle sequencing (Perkin-Elmer).

Preparation of spot synthesized peptides on membranes and immunoblot analysis of SC and SIgA binding

The SC and SIgA binding domain in SpsA encoded by the recombinant plasmids pQSM5 was divided into 34 overlapping synthetic peptides, consisting of 15 amino acids each, with an offset of three amino acids. The peptides were synthesized on cellulose paper (1Chr; 3 MM, Whatman) as described previously (Frank, 1992). On a second spot membrane, a 33-amino-acid region of the SC and SIgA binding domain (corresponding to amino acid 189–221 of SpsA of pneumococcal strain ATCC 33400) was divided into 145 overlapping synthetic peptides with an offset of one amino acid. The membranes consisted of 10 series with lengths from six amino acids up to 15 amino acids per spot. Each series spanned parts of the 15-amino-acid region, which was recognized to be positive for SC and SIgA binding in the spot membrane described above. For the SC and SIgA binding assay, the membrane was washed three times with TBS (50 mM Tris base buffered saline adjusted to pH 7.0) for 10 min and incubated overnight in blocking buffer (2 ml of Genosys-blocking buffer, 5% saccharose, 8 ml of T-TBS (TBS plus 0.05% Tween 20, pH 7.0). After washing with T-TBS, the membrane was incubated in blocking buffer containing 10 µg ml−1 SIgA or SC for 5 h followed by three washes with T-TBS. Detection of binding was carried out as already described in immunoblot analysis. Densitometric analysis of the spots resulting in relative intensity was carried out using the easy Win32 analysis software (Herolab).

Acknowledgements

We thank A. Langendries (ICP, Experimental Medicine Unit) for excellent technical assistance and D. A. Morrison, University of Illinois, Chicago, USA, for providing competence stimulating peptide (CSP). The authors are also grateful to R. Frank (GBF) for spot membrane synthesis and D. Heinz for critical reading of the manuscript. Part of this work was carried out as partial fulfilment of a diploma thesis of S.W.

Ancillary