Identification of the Bacteria-binding Peptide Domain on Salivary Agglutinin (gp-340/DMBT1), a Member of the Scavenger Receptor Cysteine-rich Superfamily*

Salivary agglutinin is encoded byDMBT1 and identical to gp-340, a member of the scavenger receptor cysteine-rich (SRCR) superfamily. Salivary agglutinin/DMBT1 is known for its Streptococcus mutans agglutinating properties. This 300–400 kDa glycoprotein is composed of conserved peptide motifs: 14 SRCR domains that are separated by SRCR-interspersed domains (SIDs), 2 CUB (C1r/C1s Uegf Bmp1) domains, and a zona pellucida domain. We have searched for the peptide domains of agglutinin/DMBT1 responsible for bacteria binding. Digestion with endoproteinase Lys-C resulted in a protein fragment containing exclusively SRCR and SID domains that binds to S. mutans.To define more closely the S. mutans-binding domain, consensus-based peptides of the SRCR domains and SIDs were designed and synthesized. Only one of the SRCR peptides, designated SRCRP2, and none of the SID peptides bound to S. mutans. Strikingly, this peptide was also able to induce agglutination of S. mutansand a number of other bacteria. The repeated presence of this peptide in the native molecule endows agglutinin/DMBT1 with a general bacterial binding feature with a multivalent character. Moreover, our studies demonstrate for the first time that the polymorphic SRCR domains of salivary agglutinin/DMBT1 mediate ligand interactions.

Until now, studies dealing with agglutinin-bacteria interactions have been focused mainly on the characterization of bacterial receptors (18,19) and identification of their cognate carbohydrate ligands on agglutinin (1, 20 -22). These studies revealed that carbohydrate residues play only a partial role in binding and aggregation of bacteria by agglutinin (1,19,20). For example, chemical modification of the carbohydrate residues of agglutinin only slightly impaired its agglutinating properties (20). On the other hand, treatments affecting the polypeptide moiety abolished binding to S. mutans completely, suggesting a dominant role for peptide domains (1,20,22). The present study was directed on identification of the peptide domains on agglutinin that are responsible for its bacteriaagglutinating properties.
From the DNA sequence of the gene encoding salivary agglutinin, DMBT1 (9,10), the architecture of the polypeptide chain of agglutinin was deduced. Salivary agglutinin (see Fig.  1A) is composed of 13 highly homologous SRCR domains (13,23) separated by SIDs (SRCR-interspersed domains), 2 CUB (C1r/C1s Uegf Bmp1) domains (24,25) separated by a 14th SRCR domain, and a ZP (zona pellucida) domain (26). The biological function of the highly conserved SRCR domains has not yet been established, but their presence in proteins with broad spectrum binding properties suggests a role in ligand binding or adhesion. We recently noted genetic polymorphism within DMBT1. These polymorphisms lead to DMBT1 alleles, giving rise to polypeptides with interindividually different numbers of SRCR domains and SIDs. Based on analogies to mucins, we have proposed that these polymorphisms may lead to a differential efficacy in mucosal protection (10,27). However, the major drawback of this hypothesis is that it remains to be shown that the SRCR domains and/or SIDs are involved in ligand binding.
The search for the peptide domains on salivary agglutinin/ DMBT1 involved in bacteria binding was therefore initially directed to this part of the molecule. On basis of the deduced amino acid sequence of agglutinin, we designed a digestion strategy to obtain a fragment consisting exclusively of SRCR and SID domains. This fragment preserved the S. mutans binding properties. To characterize the binding domain in greater detail, peptides were synthesized covering the complete SRCR and SID consensus sequence, and their binding to S. mutans was analyzed. Only one 16-mer peptide (QGRVEV-LYRGSWGTVC) of the SRCR domain was found to bind to S. mutans and to mediate agglutination.

EXPERIMENTAL PROCEDURES
Purification of Agglutinin-Human parotid saliva was collected with a Lashley cup. Twenty-five ml of parotid saliva was kept on ice water for 30 min to promote the formation of a precipitate. This precipitate was collected by centrifugation at 5,000 ϫ g for 20 min at 4°C. The resulting pellet was dissolved in 2.5 ml of PBS. The pellet was ϳ10-fold enriched in agglutinin. This crude agglutinin fraction was used as starting material for digestion studies and in the various binding studies. For further purification, the pellet was dissolved in buffer A (10 mM Tris, 10 mM EDTA, pH 6.9, 0.05% CHAPS) and applied on an UNO Q-6 column (Bio-Rad) equilibrated in buffer A and linked to fast protein liquid chromatography equipment (Amersham Biosciences). Proteins were eluted with a linear gradient from 0 to 0.5 M NaCl in buffer A. The eluate was monitored at 280 nm and analyzed by SDS-PAGE and Western blotting. The isolated agglutinin preparation contained no detectable protein impurities (less than 5%). Protein concentrations were determined with the BCA Protein Assay Kit (Pierce) according to the manufacturer's instructions.
Protein Digestion-0.5 g of endoproteinase Lys-C, sequencing grade from Lysobacter enzymogenes (Roche Diagnostics GmbH), was added to 50 l of crude agglutinin (50 g/ml), dissolved in 25 mM Tris, 1 mM EDTA (pH 8.5). After 18 h of incubation at 37°C, digestion was stopped by adding a mixture of proteinase inhibitors (Complete Mini tablets, Roche Diagnostics). For reduction, agglutinin was incubated with 10 mM dithiothreitol for 4 h at 4°C, and subsequently carboxymethylated with 20 mM 2-iodoacetamide for 4 h at 4°C.
Liquid-phase Binding Assay-Two hundred l of an S. mutans suspension (10 9 bacteria/ml in PBS) was centrifuged at 3,000 ϫ g for 10 min. The pellet was mixed with 100 l of digested agglutinin (50 g/ml) supplemented with 5 mM calcium chloride and incubated for 1 h at 37°C. Next, bacteria were collected by centrifugation at 3,000 ϫ g for 2 min, washed twice in PBS, and transferred to a new vial. Bacteriabound components were extracted by incubation with 10 mM EDTA in PBS for 1 h at 37°C. The extracts were examined by SDS-PAGE and Western analysis. Control extracts were obtained from bacteria that had not been incubated with (digested) agglutinin.
Overlay Adherence Assay-Adhesion of S. mutans to the Lys-C proteinase-digested crude agglutinin immobilized on nitrocellulose was studied using an overlay adhesion assay. Briefly, 40-l samples of digested agglutinin (50 g/ml) were separated by SDS-PAGE and blotted onto nitrocellulose membranes. The membranes were blocked with PBS supplemented with 0.1% Tween 20 and 2% bovine serum albumin (PBS-T-BSA) and incubated with S. mutans (10 9 bacteria in PBS-T-BSA) for 16 h at 4°C. The nitrocellulose membranes were washed twice with PBS-T-BSA. Bound bacteria were visualized with monoclonal antibody OMVU37, directed against S. mutans (31) with alkaline phosphatase-conjugated rabbit anti-mouse immunoglobulins (DAKO, Glostrup, Denmark) as the chromogenic substrate.
Peptide Design and Synthesis-Based on the published amino acid sequence of gp-340 (7), consensus sequences of the 13 SRCR domains and 11 SIDs were determined using alignment software (Vector NTI, InforMax Inc., Oxford, UK). Nine peptides, together spanning the complete sequence, were synthesized by solid-phase peptide synthesis using the T-bag method adapted for Fmoc (N-(9-fluorenyl)methoxycarbonyl) chemistry (32).
Peptide Purification-Peptides were purified by reversed-phase HPLC on a JASCO HPLC System (Tokyo, Japan). Peptides were dissolved in 0.1% trifluoroacetic acid and applied on a VYDAC C 18 -column (218TP, 1.0 ϫ 25 cm, 10-m particles, Vydac, Hesperia, CA) equilibrated in 0.1% trifluoroacetic acid. Elution was performed with a linear gradient from 30 to 45% acetonitrile containing 0.1% trifluoroacetic acid in 20 min at a flow rate of 4 ml/min. The absorbance of the column effluent was monitored at 214 nm, and peak fractions were pooled, lyophilized, and reanalyzed by reversed-phase-HPLC and by capillary electrophoresis on a Biofocus 2000 apparatus (Bio-Rad). The authenticity of the (monomeric) peptides was confirmed by quadrupole-time of flight mass spectrometry (Q-TOF MS) on a tandem mass spectrometer (Micromass Inc., Manchester, UK) as described previously (33). Despite the presence of cysteine residues, exclusively monomeric peptides were detected, indicating that peptide multimerization by disulfide bond formation had not occurred. The purity of the peptides was at least 90%.
Adhesion Assays-Bacterial adhesion was examined using a microtiter plate method based on labeling of microorganisms with cell-permeable DNA binding probes. 2 Microtiterplates Fluotrac 600 (Greiner, Recklinghausen, Germany) were coated with various amounts of synthetic peptides, crude agglutinin, or purified agglutinin. For reduction, the peptides and agglutinin were incubated with 10 mM dithiothreitol for 4 h at 4°C and subsequently carboxymethylated with 20 mM 2-iodoacetamide for 4 h at 4°C. The peptides were dissolved in coating buffer (100 mM sodium carbonate, pH 9.6) to a starting concentration of 40 g/ml and diluted serially. After incubation at 4°C for 16 h, plates were washed twice with Tris-buffered saline containing 0.1% Tween 20 and 1 mM Ca 2ϩ (TTC). Subsequently, 100 l of an S. mutans suspension (5 ϫ 10 8 bacteria/ml TTC) were added to each well and incubated for 2 h at 37°C. Plates were washed three times with 0.1% Tween 20 using a plate washer (Mikrotek EL 403, Winooski, VT). Bound bacteria were detected using 100 l/well of 1 mM SYTO-9 solution (Molecular Probes, Leiden, The Netherlands), a cell-permeable fluorescent DNA binding probe. Plates were incubated in the dark for 15 min at ambient temperature and washed three times with 0.1% Tween 20. Fluorescence was measured in a fluorescence microtiter plate reader (Fluostar Galaxy, BMG Laboratories, Offenburg, Germany) at 488-nm excitation and 509-nm emission wavelength.
Agglutination Assays-150 l of an S. mutans suspension (5 ϫ 10 8 bacteria/ml TTC) were mixed with 150 l of peptide solution at final peptide concentrations of 0 -200 g/ml in a 96-well microtiter plate (low affinity, PS Microlon-F, Greiner) and incubated for 2 h at 37°C. After agglutination and subsequent sedimentation of the bacteria, 10 l of the sediment was transferred on a microscopic slide. After heat fixation, bacteria were stained with a 20% crystal violet solution (Merck) and examined by light microscopy. Turbidometric analysis of the agglutination process was carried out using a spectrophotometer (UVICON 930, Kontron Instruments, Watford, UK). Bacterial suspensions (5 ϫ 10 8 bacteria/ml TTC) supplemented with either calcium chloride or EDTA were preincubated for 2 h at 37°C before peptides were added. The optical density of the bacterial suspensions was monitored at 700 nm for 100 min at 37°C. These experiments were repeated at least three times.

RESULTS
Lys-C Cleavage of Agglutinin/DMBT1-To investigate the role of the repeating SRCR and SID domains of agglutinin/ DMBT1 with respect to S. mutans binding, a selective proteo-lytic cleavage procedure was conducted using endoproteinase Lys-C (34 -36). Agglutinin/DMBT1 contains 10 lysine residues, which are all located in the C-terminal region (Fig. 1A). One lysine residue is located in the 13th SRCR domain, 3 are located in the first CUB domain, 1 is located in the second CUB domain, 4 are located in the ZP domain, and 1 is located in the unique sequence, located C-terminally of the ZP domain. Hydrolysis after the 1st and 2nd lysine residues (Lys-1722, Lys-1791) will not yield distinct cleavage products because the resulting fragments are held together by disulfide linkages between Cys-1665 and Cys-1729 and between Cys-1766 and Cys-1792, respectively (23,37). Hydrolysis at Lys-1812 will generate a fragment that contains 13 of the 14 SRCR domains as well as the interspersing SID domains (Fig. 1A). The re-FIG. 2. Lys-C digestion of agglutinin and S. mutans binding to the digested fragment. As shown in A, crude agglutinin was treated with endoproteinase Lys-C, and proteins were separated on 4 -15% SDS-PAGE, transferred to nitrocellulose, and immunoblotted with monoclonal antibody 213-6 (A and C). Lane 1, crude agglutinin; lane 2, Lys-C-digested agglutinin. After digestion, a new band appeared with a lower apparent molecular weight than the parent protein, representing a fragment that most probably contains predominantly SRCR domains and SIDs (amino acids 1-1812). As shown in B, digested agglutinin was blotted onto a nitrocellulose membrane and subjected to an overlay binding assay with S. mutans. After incubation with S. mutans, bound bacteria were visualized with an antibody directed against S. mutans. Bacteria were found at both bands, suggesting that S. mutans bound to both agglutinin and the Lys-C digested fragment. C, the digested agglutinin was subjected to a liquid-phase binding assay with S. mutans. Lane 1, starting material; lane 2, bound fraction; lane 3, unbound fraction. Both agglutinin and the digested fragment were found in the bound fraction but not in the unbound fraction, demonstrating that S. mutans binds both agglutinin and the Lys-C-digested fragment. As a molecular marker, we used the high range molecular mass standards (Bio-Rad): myosin (208 kDa), ␣-galactosidase (116 kDa), bovine serum albumin (84 kDa), and ovalbumin (47 kDa). maining C-proximal ZP and CUB domains contain additional Lys-C cleavage sites. Thus, Lys-C digestion would liberate a large SRCR/SID stretch in addition to a number of smaller fragments. SDS-PAGE analysis of the Lys-C-treated agglutinin revealed a new protein band with slightly lower molecular mass than the native agglutinin/DMBT1 ( Fig. 2A). This band, the putative SRCR/SID stretch of agglutinin/DMBT1, was recognized by monoclonal antibody 213-6 directed against a peptide epitope of salivary agglutinin/DMBT1 (29) (Fig. 2A, lane  2). Immunoreactive protein bands of lower molecular mass were not detected. Reduction and subsequent carboxymethylation virtually did not show different protein patterns on Western blot (not shown). Next to the putative SRCR/SID stretch, the native agglutinin/DMBT1 was still present, even after prolonged incubation with excess Lys-C or addition of 0.01% SDS. The triplet encoding for Lys-1812 spans the fusion site of exons 45 and 46 after splicing. Sequencing did not reveal heterozygosity of the saliva donor in the exon or intron sequences. Thus, partial digestion of salivary agglutinin/DMBT1 is not based on genetic polymorphism. Sequencing of 25 cDNA clones further ruled out that alternative splicing leads to omission of Lys-1812 (not shown).
Effect of Lys-C Digestion on Binding of Agglutinin to S. mutans-Whether binding sites for S. mutans were still present after digestion was examined using an overlay adherence assay and a soluble-phase binding assay (Fig. 2). Nitrocellulose membranes containing native agglutinin and Lys-Cdigested agglutinin overlaid with bacteria showed that S. mutans adhered to both agglutinin bands (Fig. 2B). This demonstrates that Lys-C digestion had not destroyed the binding domains for this bacterium. Controls, in which either bacteria incubation or antibody incubation was omitted, were negative. This was confirmed by the results of liquid-phase binding assays in which bacteria were incubated with a solution containing the digested agglutinin (Fig. 2C). Western analysis of the bacterial extracts and the corresponding supernatants demonstrated that both native agglutinin and the digested fragment were found in the bacterial pellets (Fig. 2C, lane 2). In contrast, the supernatants were completely devoid of agglutinin (Fig. 2C,  lane 3). In control experiments in which the bacteria were omitted, agglutinin and the digested fragment were only present in the supernatants.
Determination of Consensus Sequence of SIDs and SRCR Domains-To more specifically characterize the bacteria-binding region, we synthesized a series of peptides spanning the SIDs and SRCR domains of agglutinin. Using alignment software, the consensus sequences of the SIDs and SRCR domains were determined. This resulted in two consensus sequences for the SIDs, one 20-residue long sequence and one 22-residue long sequence (Fig. 1, B and C). Peptides representing these sequences were synthesized (designated SID20 and SID22, Table  I). The consensus sequence of the SRCR domain had a length of 109 amino acids (Fig. 1D). To cover this sequence, seven synthetic peptides were synthesized, representing loops in the native protein that run in between disulfide bridges (23, 37) (SRCR peptide (SRCRP) 1-7, Table I).
Bacteria Binding Properties of Synthetic SID and SRCR Peptides-Binding of S. mutans to these peptides was determined in a solid-phase adherence assay in which adhered bacteria were quantified using a fluorescent DNA stain. Crude agglutinin and purified agglutinin were included as controls. Besides binding to agglutinin, S. mutans adhered to only one of the nine peptides tested, SRCRP2 (Fig. 3). None of the other peptides tested, viz. SRCRP1, SRCRP3-7, SID20, and SID22, mediated adhesion of S. mutans. To examine the specificity of  the adherence, inhibition studies were conducted by preincubating S. mutans suspensions with all peptides at various concentrations. Also in these experiments, only SRCRP2 inhibited adhesion of S. mutans to agglutinin (Fig. 4), suggesting that the binding sites on S. mutans for SRCRP2 and agglutinin are identical or are located in close vicinity. Since the SRCRP2 fragment contains a cysteine residue, this peptide potentially forms disulfide-bonded species. Several results, however, indicate that monomeric SRCRP2 was the active species in our experiments. Mass analysis revealed that the SRCRP2 preparation used in the binding studies was composed of a single species with a molecular mass of the monomer. Furthermore, neither reduction/carboxymethylation of SRCRP2 nor substitution of its cysteine to alanine to prevent formation of disulfide-bonded species had any effect on its bacteria binding properties (not shown).
Other bacteria, including S. gordonii, S. sanguis, S. oralis, S. sobrinus, S. mitis I, S. mitis II, A. actinomycetemcomitans, P. intermedia, E. coli, B. fragilis, M. catarrhalis, P. micros, S. aureus, L. casei, and H. pylori were also bound by salivary agglutinin as well as by SRCRP2 but not by the other peptides tested (results not shown). These findings suggest that the SRCRP2 domain endows agglutinin with broad spectrum binding properties.
Agglutination Assays-Varying peptide concentrations were mixed with S. mutans suspensions and incubated for 2 h to allow precipitation of aggregates formed. Precipitates were examined by light microscopy, revealing that only SRCRP2 induced agglutination of S. mutans (Fig. 5). At increased calcium concentrations, lower peptide concentrations were required to induce agglutination, and the aggregates formed were clearly larger in size (Fig. 5C). In contrast, incubation in EDTA-containing buffer required higher concentrations of SRCRP2 and resulted in smaller aggregates (Fig. 5A). A similar dependence of calcium was found for the agglutinin-mediated bacteria agglutination. These findings were confirmed by the turbidometric assay (Fig. 6). DISCUSSION In the present study, we have identified a peptide domain on salivary agglutinin/DMBT1 that is involved in bacteria binding and agglutination. To confine the binding domain, we proteolytically cleaved the protein into smaller parts. The low prevalence of lysine residues in the repeating SID/SRCR domains, a single lysine at the C terminus of the 13th SRCR domain, allowed the use of Lys-C to separate the SID/SRCR part from FIG. 4. Inhibition of S. mutans binding to agglutinin by peptide SRCRP2. Purified agglutinin-coated microplates (20 g/ml) were incubated with an S. mutans suspension that was preincubated with various concentrations of peptide SRCRP2 (white bar, 50 g/ml; gray bar, 100 g/ml; black bar, 200 g/ml). Increased concentrations of peptide SRCRP2 decreased the amount of bacteria bound to agglutinin, demonstrating a competitive inhibition of peptide SRCRP2. All other peptides tested did not inhibit S. mutans binding to agglutinin at a concentration of 200 g/ ml; typical examples are peptides SRCRP1, SRCRP3, SID20, and SID22. the rest of the molecule. A complicating factor is the presence of disulfide bridges spanning potential Lys-C cleavage sites, which prevents liberation of the hydrolysis products. Residue Lys-1812 is the first Lys-C cleavage site resulting in distinct cleavage fragments. The SID/SRCR stretch of agglutinin/ DMBT1 obtained after Lys-C digestion most probably contains a part of the first CUB domain. In theory, by reduction of the disulfide bridges, this small CUB fragment could be removed. However, the bacteria binding properties of agglutinin/DMBT1 are also lost (22), hampering further identification of the binding domain using this approach. Next to the band for the putative SID/SRCR stretch, the native agglutinin/DMBT1 was still present ( Fig. 2A), even after prolonged incubation with excess Lys-C with the addition of 0.01% SDS or after reduction and carboxymethylation of agglutinin/DMBT1 prior to digestion. This was not due to genetic heterogeneity of agglutinin/ DMBT1 since it was obtained from a donor who was homozygous within the respective exons. Alternative splicing using donor/acceptor sites varying by one or few nucleotides cannot explicitly be ruled out but is unlikely to occur because none of 25 cDNA clones sequenced over the site showed a corresponding phenomenon. We speculate that the propensity of agglutinin/DMBT1 to form complexes may render a part of the lysine residues inaccessible to Lys-C. This is reasonable because if the SRCR domains are involved in ligand binding, it is likely that the constant C terminus, i.e. the CUB and ZP domains, is responsible for oligomerization. Alternatively, resistance to Lys-C digestion may be caused by differential N-and O-glycosylation of potential glycosylation sites close to the presumed cleavage site. In this context, it should be noted that CUB1 contains 21 potential O-glycosylation sites and 3 potential Nglycosylation sites (7).
Only one of the SRCR consensus peptides, SRCRP2, bound to a wide variety of Gram-positive and Gram-negative bacteria. Absence of binding by the other peptides does not necessarily mean that the corresponding domains are not involved in bacteria adherence to agglutinin/DMBT1 per se. It is possible that these peptides are derived from conformational domains of agglutinin/DMBT1 and that their lack of conformational restraints leads to poor binding of the bacteria tested.
A puzzling finding is that, whereas the bacteria binding properties of agglutinin/DMBT1 are destroyed upon reduction, the putative binding domain (the SRCRP2 peptide) is part of a disulfide bridged loop. Secondary structure analysis using protein analysis software (Vector NTI) predicts that the SRCRP2 peptide is composed of two short ␤-sheets separated by a ␤-turn (not shown). Strikingly, based on crystallographic analysis, the same secondary structure has been proposed for the corresponding sequence in the SRCR domain of the highly homologous Mac-2-binding protein (23) (Fig. 7). This indicates that formation of the disulfide bond does not introduce significant constraints on the SRCRP2 structure. Thus, the effect of reduction on the bacteria binding properties of the SRCRP2 stretch must be communicated by conformational changes elsewhere in the agglutinin/DMBT1 molecule. For example, it can be envisaged that upon breaking of disulfide bonds, the SRCR domains become shielded, thereby rendering them inaccessible to bacteria and antibodies (29).
The role of the carbohydrate moiety, which comprises 25% of the whole molecule on a weight basis (22), remains to be elucidated. Glycosylated regions will be present predominantly in FIG. 7. Schematic representation of the SRCR domain of the Mac-2-binding protein, highlighting the putative bacteria-binding domain of agglutinin/DMBT1. A model of the SRCR domain of the Mac-2-binding protein (Protein Data Bank accession code 1BY2) was constructed using RasWin Molecular Graphics. The stretch that corresponds to SRCRP2 is composed of two short ␤-sheets joined by a ␤-turn (highlighted in yellow) and highly homologous (Ͼ80%) to the Mac-2-binding protein amino acid sequence. The QXR motif (blue, Q18; red, R20) is located at the surface of the SRCR domain. the SIDS, which when compared with SRCR domains, contain a high density of potential O-glycosylation sites. The high density of glycans forces these regions in an extended conformation, thus creating a molecule with alternating stretched SIDS and globular SRCR domains, promoting multivalent interaction. In addition, the oligosaccharides may provide (low affinity) binding sites for microbial receptors, which in the mature agglutinin/DMBT1 may act in concert with the SRCR peptide domains.
A study on the bacteria binding of the human macrophage MARCO receptor (38), another member of the SRCR superfamily, has demonstrated that a peptide encompassing residues 431-441 (RGRAEVYYSGT) was responsible for binding of MARCO to S. aureus and E. coli (39). This region corresponds to residues 1-11 of SRCRP2 with 55% homology. Detailed analysis of MARCO variants demonstrated a crucial role for an arginine-rich segment for this function (40). More precisely, the motif RXR was identified as an essential element for highaffinity bacterial binding. Strikingly, the consensus sequence SRCRP2 does not contain such an RXR motif but rather a QXR motif. The present finding that only this peptide mediates binding to a wide variety of bacteria suggests that both motifs, RXR and QXR, could play a role in bacteria binding.
In summary, we propose that the SRCR domains of salivary agglutinin/DMBT1 are involved in bacteria binding. Furthermore, we find that only one SRCR domain consensus-based synthetic peptide of 16 amino acids, which is 100% identical to the actual amino acid sequence of 8 out of 13 SRCR domains, is responsible for the broad ligand binding behavior. The repeated presence of this peptide in the native molecule endows agglutinin/DMBT1 with a general bacterial binding feature with a multivalent character. This lends support for the hypothesis that numeric variations of the SRCR domains of salivary agglutinin/DMBT1 may interfere with efficient protection. The crucial next steps will be to determine which pathogen binding activities are qualitatively or quantitatively affected by these polymorphisms.