Salivary Statherin Peptide-Binding Epitopes of Commensal and Potentially Infectious Actinomyces spp. Delineated by a Hybrid Peptide Construct

Abstract

Adhesion of microorganisms to host receptor molecules such as salivary statherin molecules is a common event in oral microbial colonization. Here we used a hybrid peptide construct (with both a hydroxyapatite-binding portion and a test peptide portion) to map the interaction of Actinomyces species (and Candida albicans) with statherin. Adhesion to hybrid peptides and truncated statherin variants revealed three binding types, types I to III. (i) Type I strains of rat, hamster, and human infection origins bound C-terminal-derived QQYTF and PYQPQY peptides. The QQYTF peptide inhibited statherin binding for some strains but not for others. (ii) Type II strains of human and monkey tooth origins bound middle-region-derived YQPVPE and QPLYPQ peptides. Neither strain was inhibited by soluble peptides. (iii) Type III strains of human infection origins (and C. albicans) did not bind to either statherin-derived peptides or truncated statherin. Moreover, the type I strains inhibited by QQYTF were also inhibited by TF and QAATF peptides and were detached from statherin by the same peptides. In conclusion, it is suggested that commensal and potentially infectious microorganisms bind middle or C-terminal statherin differently and that other microbes might require discontinuous epitopes.

Adhesion of microorganisms to host receptor molecules (e.g., poly- and glycopeptides in secretions) is a crucial step in biofilm development. Adhesion sites on host molecules have been delineated with respect to carbohydrate and (though less elaborately) peptide recognition motifs (9, 11, 16). The recognition of RGD motifs in fibronectin and iC3b by Candida tropicalis and C. albicans, respectively (3), and the recognition of PQ in salivary acidic proline-rich proteins (APRPs) by Streptococcus gordonii and Actinomyces naeslundii genospecies (gsp) 2 (9, 16) are examples of proposed peptide recognition motifs.

Statherin, which is present in saliva (26) and nasal secretions (18), is a 43-residue phosphopeptide (26). It mediates adhesion of Actinomyces spp. (17), Porphyromonas gingivalis (1), and Fusobacterium nucleatum (28; S. Sekine, K. Kataoka, and S. Shizukuishi, Abstr. 80th Gen. Session IADR/AADR/CADR, abstr. 0735, 2002) (http://iadr.confex.com/iadr/2002SanDiego/techprogram/abstract_9594.htm), all of which are implicated in periodontitis, and C. albicans (14). Adhesion of F. nucleatum involves the middle Tyr₂₁Gln₂₂Pro₂₃Val₂₄Pro₂₅Glu₂₆ stretch of statherin (Sekine et al., Abstr. 80th Gen. Session IADR/AADR/CADR 2002), adhesion of A. viscosus 19246 involves the N-terminal Thr₄₂Phe₄₃ dipeptide of statherin (16), and adhesion of P. gingivalis fimbrillin involves the Leu₂₉Tyr₃₀ and Tyr₄₁Thr₄₂Phe₄₃ residues of statherin (1).

As well as binding to hydroxyapatite, the phosphorylated N-terminal 1- to 15-amino-acid (aa) segment of statherin also inhibits calcium phosphate precipitation (26). The C-terminal section of statherin, especially the QQYTF terminus and longer C-terminal fragments, displays bacterial growth inhibition activity (15). Statherin is encoded by the STATH gene. The STATH gene has high identity with (and is located near) the histatin genes (HIS1 and HIS2), suggesting that HIS1, HIS2, and STATH belong to a single gene family exhibiting accelerated evolution between the HIS and STATH coding sequences (25). The antifungal properties of histatins have been extensively described previously (21, 23, 24).

A. naeslundii gsp 2 is an early commensal colonizer of teeth and mucosal surfaces of humans and monkey, whereas A. naeslundii gsp 1 occurs at later stages of oral biofilm development (13). A. viscosus is an oral commensal bacterium in rats and hamsters (8). Actinomyces spp. have also been implicated in dental caries development (2, 4), root canal infections, and chronic, suppurative infections in the cervicofacial, thoracic, and abdominopelvic regions and in the central nervous system (5, 27).

A. naeslundii and A. viscosus recognize APRPs and statherin through type 1 fimbriae. The type 1 fimbriae specificity for APRPs versus statherin parallels structural variations in the fimP major subunit genes (12, 16) and coincides with Actinomyces tropism (17). Thus, A. naeslundii gsp 2 in humans binds preferentially to APRPs while A. viscosus colonizing rats and hamsters (as well as strains isolated from actinomycosis infections or blood) binds preferentially to statherin (17).

Recently a statherin-based hybrid fusion construct employing the high level of affinity of statherin for hydroxyapatite was found useful for leukocyte binding to peptides bearing the RGD epitope (10). The 15 N-terminal amino acids of statherin, including the two phosphoserines in positions 2 and 3, were fused to a test domain via a proline residue, which (through its conformational restraints) confers optimal presentation of the test epitope away from the hydroxyapatite surface.

The aim of the present paper was to use a hybrid fusion construct to delineate binding epitopes in statherin for commensal versus infectious Actinomyces spp. and C. albicans.

MATERIALS AND METHODS

Bacterial strains, culture growth, and radiolabeling.

The following microorganisms were used: A. viscosus strains ATCC 19246 (from the Culture Collection of the University of Göteborg [CCUG], Göteborg, Sweden) and R28 and T6-1600 (from the late M. Yeung, University of Texas Health Science Center, San Antonio, Tex.), A. naeslundii gsp 1 strains ATCC 12104, 35334, 29952, and 30267 (CCUG) and gsp 2 strains T14V (from P. Kolenbrander, National Institutes of Health, Bethesda, Md.) and M4356 (the late M. Yeung), A. radicidentis strain 42377 (CCUG), and C. albicans strain GDH 18 (from an oral isolate provided by L. P. Samaranayake, Hong Kong).

Actinomyces strains were metabolically labeled by adding 200 μCi of [³⁵S]methionine (Trans ³⁵S-Label; ICN Biomedicals, Irvine, Calif.) to bacteria suspended in 80 μl of 10 mM phosphate-buffered saline (pH 7.2). The suspension was spread on Columbia II agar base plates, supplemented with human erythrocytes, and grown overnight (19 h) at 37°C in an atmosphere with 5% CO₂. Labeled bacteria were harvested and transferred to adhesion buffer (ADH; 50 mM KCl, 1 mM CaCl₂•2H₂O, 0.1 mM MgCl₂•6H₂O, 0.62 mM K₂HPO₄, 1.4 mM KH₂PO₄, pH 6.5).

Candida cells were metabolically labeled by adding 10 μCi of [³⁵S]methionine to 10 ml of yeast nitrogen base (Difco Laboratories, Detroit, Mich.) with 20 mM glucose before overnight (19 h) growth under conditions of agitation at room temperature. Candida cells (blastospores) were harvested and transferred to ADH buffer.

After harvesting, all microorganisms were washed three times in adhesion buffer by centrifugation at 12,000 × g for 10 min and resuspended in adhesion buffer with 0.5% bovine serum albumin (BSA; Sigma Chemical Co, St. Louis, Mo.) to a density of 5 × 10⁸ cells/ml for Actinomyces and to 5.4 × 10⁷ cells/ml for C. albicans.

Saliva collection.

Lashley cups were used to collect parotid saliva in ice-chilled tubes. Secretion was stimulated by a mildly acidic lozenge (SST; Salix Pharma, Tystberga, Sweden).

Statherin purification.

Statherin was purified from freshly collected parotid saliva as described previously (14). The purity and identity of the purified statherin were confirmed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis and bacterium-binding investigations.

Synthetic peptides.

A set of synthetic hybrid peptides (Fig. 1) with a common hydroxyapatite-binding domain corresponding to residues 1 to 15 of statherin (Asp-pSer-pSerGluGluLysPheLeuArgArgIleGlyArgPheGly-; DpSpSEEKFLRRIGRFG) linked via a proline to various test fragments of 5 to 10 aa residues (10) and a set of custom hexa-, penta-, and dipeptides were synthesized.

FIG. 1.

Illustration of amino acid stretches covered by hybrid and custom peptides. “P” (statherin illustration) denotes phosphate on serine, and “pS” denotes phosphoserines in positions 2 and 3. “P” (hybrid peptides illustration) denotes proline linking the hydroxyapatite-binding and test segments.

The hybrid peptides contained the following test sequences, which, in total, covered residues 16 through 43 of statherin: hybrid peptide 1 contained -Tyr₁₆Gly₁₇Tyr₁₈Gly₁₉Pro₂₀ (YGYGP); hybrid peptide 2 contained -Tyr₂₁Gln₂₂Pro₂₃Val₂₄Pro₂₅Glu₂₆ (YQPVPE); hybrid peptide 3 contained -Gln₂₇Pro₂₈Leu₂₉Tyr₃₀Pro₃₁Gln₃₂ (QPLYPQ); hybrid peptide 4 contained -Pro₃₃Tyr₃₄Gln₃₅Pro₃₆Gln₃₇Tyr₃₈ (PYQPQY); hybrid peptide 5 contained -Gln₃₉Gln₄₀Tyr₄₁Thr₄₂Phe₄₃ (QQYTF); and hybrid peptide 6 contained -Pro₃₃Tyr₃₄Gln₃₅Pro₃₆Gln₃₇Tyr₃₈Gln₃₉Gln₄₀Tyr₄₁Thr₄₂Phe₄₃ (PYQPQYQQYTF). Hybrid peptides 1 to 4 and 6 were synthesized by Nordic BioSite, Täby, Sweden, and peptide 5 was synthesized by Biopeptide Co., San Diego, Calif.

In addition, the following custom peptides with unmodified endings were synthesized: TyrGlyTyrGlyPro (YGYGP; Biopeptide Co.), TyrGlnProValProGlu (YQPVPE; Biopeptide Co.), GlnProLeuTyrProGln (QPLYPQ; Biopeptide Co.), ProTyrGlnProGlnTyr (PYQPQY; Biopeptide Co.), GlnGlnTyrThrPhe (QQYTF; Biopeptide Co.), GlnGlnTyrAlaAla (QQYAA; Thermo Hybaid, Ulm, Germany), GlnAlaAlaThrPhe (QAATF; Thermo Hybaid), ThrPhe (TF; Thermo Hybaid), GlnTyr (QY; Thermo Hybaid), and Asp₁pSer₂pSer₃Glu₄Glu₅Lys₆Phe₇Leu₈Arg₉Arg₁₀Ile₁₁Gly₁₂Arg₁₃Phe₁₄Gly₁₅Tyr₁₆Gly₁₇Tyr₁₈Gly₁₉Pro₂₀Tyr₂₁-Gln₂₂Pro₂₃Val₂₄Pro₂₅-Glu₂₆Gln₂₇Pro₂₈Leu₂₉Tyr₃₀Pro₃₁Gln₃₂ (DpSpSEEKFLRRIGRFGYGYGPYQPV-PEQPLYPQ; Nordic Biosite). All peptides were high-pressure liquid chromatography purified to at least 95% purity and characterized by mass spectrometry by the respective companies.

Hydroxyapatite binding.

Adhesion of radiolabeled microorganisms to statherin (5 and 25 μg/ml [corresponding to 0.93 and 4.65 μM, respectively]) or to hybrid peptide (dilutions from 1.56 to 50 μg/ml [corresponding to 0.55 to 19.3 μM]) coated onto hydroxyapatite beads was measured as described previously (14). Briefly, 5 mg of hydroxyapatite (Macrosorb C; Microporous Materials Ltd., Deeside, United Kingdom), which had been prewashed three times with 0.2 M NaOH and three times with water and dried, was weighed into each well in a 96-well plate and equilibrated in ADH over night. The beads were washed with ADH and incubated with 125 μl of protein or hybrid peptide suspension for 1 h under conditions of constant rotation and then washed three times and blocked with ADH with 1% BSA. Labeled bacteria or Candida organisms (125 μl) were incubated with protein-coated beads for 1 h; after repeated washings, the percentage of bound microorganisms was determined by scintillation counting (using a 1214 Rackbeta liquid scintillation counter [LKB Wallac, Turku, Finland] and Ready Protein plus scintillation fluid [Beckman Coulter, Fullerton, Calif.]).

Binding inhibition.

Adhesion inhibition of microbial binding to statherin (4.5 μM), parotid saliva (diluted 1:1 with ADH buffer), or hybrid peptide (4.5 μM) coated onto hydroxyapatite was tested by preincubating the microorganisms with statherin (0.5 mg/ml, 92.9 μM) or synthetic peptide (3 mM) and serial dilutions of peptides ranging from 0.125 to 4 mg/ml (corresponding to 0.16 to 5.83 mM) for 30 min at room temperature.

Latex bead assay.

Calcium-phosphate-coated latex beads (Bangs Laboratories, Inc., Fishers, Ind.) (80 mg) were coated with statherin (25 μg/ml) and washed as described previously (17). Latex beads coated with BSA (2 mg/ml) were used as a control. The beads were suspended in 1.6 ml of phosphate-buffered saline-BSA (2 mg/ml). Binding inhibition was tested by preincubating microorganisms (15 μl of a 1.5 × 10⁹ cells/ml suspension) with either statherin (15 μl of a 0.5 mg/ml suspension) or various synthetic peptides (15 μl of a 3 mM suspension) at room temperature for 15 min. After the addition of statherin-coated latex beads (15 μl), aggregation was scored visually as follows: 0, no visible aggregates; 1, small uniform aggregates but no change in suspension turbidity; 2, more aggregates of slightly larger size than those that scored 1 but no change in suspension turbidity; 3, more and slightly larger aggregates than those that scored 2 and a slight decrease in turbidity; 4, larger aggregates than those that scored 3 but no change in number and a solution becoming clear; 5, larger, but fewer, aggregates in a clear solution (13).

Aggregation assay.

Aggregation was tested by incubating equal amounts of Actinomyces (5 × 10⁹ cells/ml) bacteria or Candida (1.1 × 10⁸ cells/ml) yeasts with statherin (1 mg/ml) or parotid saliva for 5 min at room temperature on a glass slide. Aggregation was scored visually as described above.

Desorbtion.

Equal volumes of statherin-coated latex beads (15 μl) and microbial cells (Actinomyces [5 × 10⁹ cells/ml] and Candida [1.1 × 10⁸ cells/ml]) were incubated on a glass slide under conditions of moderate tilting for 15 min at room temperature, and aggregation was scored. Synthetic peptides (15 μl [3 mM]) were added; after 3 min, aggregation was rescored visually using the criteria described above.

Statistics.

Data are presented as mean values of triple measurement and standard errors (SE) of the means for continuous, normally distributed variables. Assay reproducibility values are illustrated as SE and coefficient of variation. Differences between means for the various test groups were tested using the corresponding control group. Student's t test was used when comparing two groups; for comparisons among more than two groups, analysis of variance (followed by a multiple-mean, post hoc test [Tukey's test]) was employed. SPSS software (version 10.0) was used. A P value below 0.05 was considered indicative of a significant difference.

RESULTS

Microorganisms of various origins bind different statherin segments.

Adhesion of Actinomyces (and C. albicans) strains to statherin and truncated statherin revealed three binding types, types I to III (Table 1). (i) Type I strains of rat, hamster, and human infection origins bound to statherin (1 to 43 aa) avidly but only weakly to a statherin polypeptide lacking the C-terminal portion (1 to 32 aa). (ii) Type II strains of human and monkey tooth origin bound avidly to statherin and to the truncated statherin polypeptide. (iii) Type III strains of human infection origins (and C. albicans) bound to statherin but not to the statherin truncate. No microorganism or strain bound to a synthetic peptide of the N-terminal aa 1 to 15 portion of statherin (Table 1).

TABLE 1.

Binding to statherin reflecting hybrid peptides, purified statherin, and a statherin truncate

Species and strain	Binding typea	Results										Origin
		Statherin test domain			Hybrid peptide test domainb						Inhibition by epitope peptide
		aa 1-43	aa 1-15	aa 1-32	aa 33-43	aa 16-20 (YGYGP)	aa 21-26 (YQPVPE)	aa 27-32 (QPLYPQ)	aa 33-38 (PYQPQY)	aa 39-43 (QQYTF)
A. viscosus
R28	Ia	++++	−	+	+++	−	−	+	+++	+++	Yes	Rat mouth
T6-1600	Ia	++++	−	+	+++	−	−	+	+++	+++	Yes	Hamster mouth
19246	Ia	++++	−	−	++	−	−	+	++	++	Yes	Human actinomycosis
A. radicidentis
42379	Ib	++++	−	++	++	−	−	+	+++	++	No	Human osteitis
A. naeslundii gsp 1
30267	Ib	++++	−	++	++	−	−	−	++	+	No	Human blood
A. naeslundii gsp 2
T14V	II	+++	−	+++	+	−	+++	++++	++	++	No	Human dental plaque
M4356	II	++++	−	++++	+	−	++++	++++	−	−	No	Monkey dental plaque
A. naeslundii gsp 1
12104	III	+	−	−	−	−	−	−	−	−	No	Human sinus
35334	III	+	−	−	−	−	−	−	−	−	No	Human endocarditis
29952	III	++	−	−	−	−	+	−	−	−	No	Human blood
C. albicans
GDH18	III	+++	−	−	−	−	−	−	−	−	No	Human mouth

^aClassified on the basis of binding and binding inhibition patterns.

^baa, amino acid position in the 43-aa statherin polypeptide. Concentrations of 4.4 to 4.8 μM of the various hybrid peptides (12.5 μg of the penta- or hexapeptides/ml and 15.7 μg of the 11-aa peptide/ml) and 4.5 μM of statherin truncate (aa 1 to 32; 17.7 μg/ml) or statherin (25 μg/ml) were coated onto hydroxyaptite beads. ++++, binding of ≥60% of input cells; +++, ≥40 to ≤60%; ++, ≥25 to <40%; +, ≥10 to <25%; −, <10%. Scores were determined using triplicate measurements in each experiment. The experiments were repeated three times. Background binding to BSA was <6%, and the median coefficient of variation was 5.2% among all experiments.

Middle and C-terminal peptides of statherin mediate differential adhesion.

To further localize microbial binding to separate statherin segments, C-terminal and middle-region peptides were tested using the hybrid-construct assay to investigate levels of binding (Fig. 1). Type I strains bound avidly to hybrid constructs bearing the C-terminal-derived QQYTF and PYQPQY peptides and PYQPQYQQYTF peptides (Table 1 and Fig. 2a). Type II strains bound avidly to the middle region-derived YQPVPE and QPLYPQ peptides (Table 1 and Fig. 2b). Type III strains did not bind to any of the peptides presented in the hybrid construct (Table 1 and Fig. 2c).

FIG. 2.

Binding of A. viscosus, strain R28 (type I) (a), A. naeslundii gsp 2, strain T14V (type II) (b), and C. albicans, strain GDH18 (type III) (c), to increasing amounts of hybrid peptides. Data are expressed as percentages of added bacteria binding to hydroxyapatite (means and SE of triplicate measurements). S_16-20 denotes YGYGP (corresponding to aa 16 to 20 of statherin), S_21-26 denotes YQPVPE (aa 21 to 26), S_27-32 denotes QPLYPQ (aa 27 to 32), S_33-38 denotes PYQPQY (aa 33 to 38), and S_39-43 denotes QQYTF (aa 39 to 43).

C-terminal QQYTF inhibits adhesion of Actinomyces strains of infectious origin.

Adhesion of certain type I strains (type Ia), though not others (type Ib), was inhibited by QQYTF in solution regardless of whether the adhesion was to hybrid peptides, statherin, or saliva coated onto hydroxyapatite surfaces (Table 1 and Fig. 3). Inhibition occurred in a dose-dependent fashion, and 50% or more inhibition was achieved at 6 mM of peptide (data not shown). The same results were obtained by investigations (performed using latex beads coated with purified statherin) of aggregation inhibition or lack of inhibition in the presence of the custom peptides (data not shown). No other peptide displayed inhibition activity in solution regardless of the interaction mode or ligand (hybrid peptide, statherin, or saliva; data not shown).

FIG. 3.

Inhibition of binding of A. viscosus, strain R28 (type I) (a), and A naeslundii gsp 2, strain T14V (type II) (b), to hybrid peptide (4.5 μM), statherin (4.5 μM), and parotid saliva (diluted 1:1) coated onto hydroxyapatite beads with custom peptides in solution. “Preincubation” indicates that bacteria were treated with peptide in solution before added to ligand-coated hydroxyapatite. The custom peptides correspond to the statherin-binding epitopes with and without alanine replacements in the QQYTF peptide. Data are expressed as percentages of added bacteria binding (mean and SE of triplicate measurements). “Control binding” refers to binding without peptide preincubation. “ns” indicates a nonsignificant difference compared to the results seen with control binding. **, P < 0.01; ***, P < 0.001 (compared to the results seen with the respective controls).

Delineation of QAATF as an optimal inhibitor.

To delineate the epitope(s) in QQYTF, the inhibitory activity of alanine-substituted peptides was established. Adhesion of type 1 rat and hamster and human infection strains to hybrid peptide-coated or statherin-coated hydroxyapatite was inhibited by QQYTF, QAATF, and TF but not by QQYAA or PYQPQY peptides in solution (Fig. 3a), suggesting that binding activity resides in small and conformationally dependent epitopes in which TF and Q contribute to inhibition activity of QQYTF.

Desorbtion of statherin-bound bacteria by QQYTF and QAATF.

TF-containing peptides (QQYTF, QAATF, and TF) desorbed A. viscosus bound to statherin on latex beads, whereas peptides with only a QY motif (QQYAA, PYQPQY, and QY) had no effect. The QQATF peptide totally (aggregation score of 3 reduced to 0) and the QQYTF and TF peptides partially (aggregation score of 3 reduced to 2) desorbed A. viscosus from statherin.

Cryptic behavior of statherin.

Purified statherin could not inhibit bacteria binding to statherin or hybrid peptides coated onto hydroxyapatite or to purified statherin coated onto latex beads (data not shown). Further, purified statherin (1 mg/ml) did not cause aggregation of any of the bacteria (data not shown).

DISCUSSION

The present report extends current knowledge of bacterial recognition motifs by using a new hybrid peptide fusion construct. While the N-terminal stretch (1 to 20 aa) of statherin confers no bacterial binding, binding resides either in middle or C-terminal small epitopes or in more complex epitopes not yet identified. The hybrid peptide construct (proven useful for leukocyte binding to RGD) (10) confirmed the previous report by Li et al. that the C-terminal TF dipeptide is part of the statherin-binding site for A. viscosus strain 19246 (16). Advantages of the hybrid construct binding assay are that (i) no restraints are introduced to statherin-binding cooperativity and (ii) the linking proline residue orients the test epitope away from the hydroxyapatite and leaves it highly dynamic (10). However, the hybrid peptide does not necessarily mimic the conformation and configuration present in self-associating soluble or membrane-bound proteins. The present results, indicating small configuration-dependent epitopes, illustrate this aspect.

NMR and CD studies have revealed three distinct structural motifs of statherin (19, 20): (i) an alpha-helical structure at the N-terminal domain spanning Asp₁-Tyr₁₆ (found here not to expose any bacterial-binding motif); (ii) a polyproline type II segment at the middle proline-rich domain covering Gly₁₉-Gln₃₅, exposing binding sites for A. naeslundii gsp 2 (present study) and F. nucleatum (Sekine et al., Abstr. 80th Gen. Session IADR/AADR/CADR 2002); and (iii) a 3₁₀-helical TyrGln-rich structure at the C-terminal Pro₃₆-Phe₄₃ stretch exposing binding sites for A. viscosus (present study) and P. gingivalis (1). In addition to these linear epitopes, however, the presence of discontinuous or scattered binding epitopes is indicated by the lack of binding for some Actinomyces gsp 1 strains and C. albicans.

Adhesion of bacteria to the middle and C-terminal peptides did not reveal any consensus sequences. Whereas Q was the common denominator among binding motifs, TF was crucial for binding inhibition and desorbtion. Thus, QQYTF and TF, but not QQYAA, inhibited binding and desorbed bound bacteria. The lack of TF in the PYQPQY binding peptide may indicate that both TF and Q epitopes are active, though TF is a primary and high-affinity binding site. Notably, the enhanced inhibitory and desorbing effect seen when the TF dipeptides are preceded by QAA (i.e., QAATF) supports this hypothesis or argues for a role of QAA in TF configuration. The influence of single preceding amino acid residues on ligand efficacy has been shown for other peptides, i.e., RGD epitope induction of fungus cell signaling (6) and binding inhibition to the PQ epitope in APRPs (16). Nevertheless, both peptide and carbohydrate recognition may involve strikingly narrow epitopes for binding.

Low-affinity binding may allow bacteria to escape binding of inhibitors in solution, whereas high-affinity binding would not. Interestingly, commensal A. naeslundii gsp 2 strains displayed noninhibitable binding whereas periodontitis-associated F. nucleatum binds the same segment in an inhibitable way (Sekine et al., Abstr. 80th Gen. Session IADR/AADR/CADR 2002). Furthermore, binding of A. viscosus strain 19246 isolated from an infection site was found to be inhibitable by C-terminal peptides in solution and periodontitis-associated P. gingivalis recombinant fimbrillin behaves in the same way (1). Accordingly, potentially human infectious strains, including rat and hamster strains, may have difficulty becoming established in the human mouth because of statherin-derived peptides, even though statherin per se has cryptic receptors in human saliva (28; Sekine et al., Abstr. 80th Gen. Session IADR/AADR/CADR 2002, and data not shown). Similarly, statherin promotes growth of many bacteria (7) and statherin peptides inhibit growth of oral anaerobic bacteria (15). Other biofilm regulatory effects besides adhesion may thus be released upon proteolysis of statherin. Notably, statherin-derived peptides (due to endogenous or bacterial proteolysis) are present in saliva (22). The different statherin interaction modes may (to various degrees) reflect statherin-mimicking epitopes in other tissue components, as previously suggested for epithelial adhesion sites for C. albicans colonization (14).