INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

The caseins are the most abundant bovine milk proteins, and there are four major types: _s1-, _s2-, -, and -casein (23). All four caseins are phosphorylated on specific seryl residues, and in addition, -casein is glycosylated (4). -Casein is hydrolyzed by the enzyme chymosin between Phe¹⁰⁵ and Met¹⁰⁶, generating two polypeptides: a hydrophobic N-terminal para--casein polypeptide -casein (residues 1 to 105) [-casein(1-105)] and a hydrophilic phosphorylated and glycosylated C-terminal polypeptide -casein(106-169), known as the caseinomacropeptide (CMP). CMP is heterogeneous and contains all the posttranslational modification sites (glycosylation and phosphorylation) of -casein. Six potential glycosylation sites have been identified on CMP (18), and up to five different carbohydrate moieties may be attached at each site (20). Three genetic variants of CMP have also been identified, originating from the precursors -casein A, B, and E, with variants A and B being the most common in bovine milk (15).

CMP and CMP-derived peptides have been reported to have a variety of biological activities, such as suppression of gastric secretions (28), depression of platelet aggregation (2), inhibition of influenza virus hemagglutination (8), inhibition of cholera toxin binding (9), and immunomodulating activities (14). CMP has also been shown to incorporate into salivary pellicle and inhibit the adherence of Streptococcus mutans, the oral pathogen implicated in the development of dental caries (12, 17, 21, 25, 26). The incorporation of CMP into salivary pellicle is proposed to be the mechanism by which certain milk protein fractions, when added to the cariogenic diet of rats, substantially reduce caries activity and the recovery of mutans group streptococci from the experimental animals (7). Examination of the amino acid sequence of CMP revealed that the peptide contains residues that could form an amphipathic helical structure, and as such, the peptide may possess antibacterial properties.

Antibacterial peptides have been isolated and characterized from mucosal surfaces of the gastrointestinal tract and secretions of a wide variety of organisms (2, 13). In general, these peptides have been reported to contain a high percentage of basic amino acyl residues in an amphipathic structure. These characteristics have been proposed to facilitate interaction between the positively charged peptide and the negatively charged bacterial membrane (13). Antibacterial peptides have been characterized according to their size, conformation, and amino acid composition into -helical amphipathic peptides, cysteine-rich -sheet amphipathic peptides, disulfide ring peptides, linear peptides with proline-rich sequence motifs (13), and, more recently, phosphorylated and glycosylated anionic antibacterial peptides (22).

We present here evidence that CMP has activity against the oral opportunistic pathogens Streptococcus mutans and Porphyromonas gingivalis and against Escherichia coli. Fractionation of CMP, using reversed-phase high-performance liquid chromatography (RP-HPLC), revealed that the nonglycosylated, phosphorylated form of -casein(106-169), which we have designated kappacin, was the active form of the peptide against S. mutans. Hydrolysis of CMP with endoproteinase Glu-C generated a range of peptides, of which only Ser(P)¹⁴⁹-casein(138-158) exhibited growth-inhibitory activity. Peptides corresponding to Ser(P)¹⁴⁹-casein-A(138-158) and its nonphosphorylated counterpart, -casein-A(138-158), were chemically synthesized and tested for antibacterial activity. The phosphorylated peptide Ser(P)¹⁴⁹-casein-A(138-158) exhibited growth-inhibitory activity against S. mutans (MIC, 59 µg/ml [26 µM]); however, the nonphosphorylated peptide did not inhibit growth of the bacterium, indicating that phosphorylation is essential for antibacterial activity.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Preparation of CMP. Casein-HCl (Bonlac Foods, Melbourne, Australia) was dissolved in deionized water at 2.5% (wt/vol) and the pH was adjusted to 6.2 by the addition of 1 M NaOH. Chymosin (EC 3.4.23.4; Sigma, St. Louis, Mo.) was added at 0.001% (wt/vol), and the mixture was incubated for 1 h at 37°C. The hydrolysis was terminated by the addition of trichloroacetic acid to a final concentration of 4% (vol/vol). The solution was centrifuged (10,000 × g, 5 min) and the supernatant (0.5 ml) was applied to a TSK gel filtration column (30 cm by 7.8 mm; Supelco, Bellefonte, Pa.) connected to an Applied Biosystems Incorporated (ABI) HPLC system using 30% acetonitrile-0.1% (vol/vol) trifluoroacetic acid (TFA) in deionized water at a flow rate of 1.0 ml/min. The eluant was monitored using an ABI 1000S diode array detector at a primary wavelength of 215 nm. Fractions were collected and analyzed by mass spectrometry (MS) and N-terminal sequence analysis (see below). The fraction containing CMP was lyophilized and stored at 70°C.

Fractionation of CMP. CMP was dissolved at 6.0 mg/ml in 0.1% (vol/vol) TFA in deionized water (solvent A) and applied to a Brownlee (Alltech) C₁₈ preparative RP column (250 by 10 mm) installed in a Waters 440 HPLC system. The sample was eluted using a gradient of 15% solvent B for 5 min and a gradient of 15 to 20% solvent B in 1 min followed by a gradient of 20 to 50% solvent B in 155 min at a flow rate of 4.0 ml/min. Solvent B contained 90% acetonitrile-0.1% (vol/vol) TFA in deionized water. The eluant was monitored using a primary wavelength of 215 nm. All fractions were analyzed by MS and N-terminal sequence analysis, lyophilized, and tested for antibacterial activity (see below).

Endoproteinase Glu-C digestion of CMP and fractionation of generated peptides. CMP was dissolved in 50 mM ammonium acetate, pH 4.0, buffer at 1.0 mg/ml. Endoproteinase Glu-C (Sigma) was added to give a final concentration of 5.0 µg/ml, and the solution was then incubated at 37°C for 24 h. The hydrolysis was stopped by lowering the pH to 3.0 by the addition of glacial acetic acid. Generated peptides were separated by application of 100 µl to a Brownlee analytical C₁₈ column (see above) using a gradient from 0 to 60% solvent B in 24 min. Solvent A contained 0.1% (vol/vol) TFA in deionized water, and solvent B contained 90% acetonitrile-0.1% (vol/vol) TFA in deionized water. The fractions collected were characterized by MS and N-terminal sequence analysis and then tested for antibacterial activity (see below).

MS. Mass spectrometric analysis of peptides was performed using a Voyager linear matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectrometer equipped with delayed extraction (PerSeptive Biosystems). Samples were mixed (1:1, vol/vol) on the sample analysis plate with a saturated solution of 2,5-dihydroxybenzoic acid in 30% aqueous acetonitrile, containing 0.1% (vol/vol) TFA, and left to dry. All spectra were obtained in linear positive and negative ion modes with an accelerating voltage of 20 kV and pulse delay time of 125 ns.

N-terminal sequence analysis. Peptides were applied to a preconditioned Hewlett-Packard sequencing column in 1 ml of 2% (vol/vol) TFA in deionized water. The N-terminal amino acid sequence was determined using a Hewlett-Packard G1005A protein sequencing system.

Solid-phase peptide synthesis and purification. Peptides corresponding to Ser(P)¹⁴⁹-casein-A(138-158) and -casein-A(138-158) were synthesized using standard solid-phase peptide synthesis protocols for 9-fluoroenylmethoxy carbonyl (Fmoc) chemistry on an ABI 431 peptide synthesizer. The peptides were assembled as the carboxyl form using Pac-Peg-PS resin (PerSeptive Biosystems). Subsequent additions of the remaining Fmoc amino acids, including Fmoc-Ser[PO(OBzl)OH]-OH (Calbiochem-Novabiochem Pty Ltd, Sydney, Australia), were accomplished with O-benzotriazole-N,N,N,N-tetramethyluronium hexafluorophosphate-1 hydroxybenzotriazole (Auspep Pty Ltd, Melbourne, Australia).

Antibacterial growth assay The gram-positive bacterium S. mutans Ingbritt and the gram-negative bacteria P. gingivalis W50 (ATCC 53978) and E. coli NCTC 10418 were used for the antibacterial assay and were stored in 30% glycerol broth at 20°C or as lyophilized cultures. The antibacterial assay was conducted in sterile 96-well microtiter plates (Becton Dickinson). S. mutans was cultured in Todd-Hewitt broth (36.4 g/liter)-yeast extract (5 g/liter) broth containing 100 mM potassium phosphate, pH 6.28 (TYPB). E. coli was cultured in nutrient broth (36 g/liter), pH 6.28, and P. gingivalis was cultured in brain heart infusion broth (37 g/liter) supplemented with hemin (5 mg/liter) and cysteine (10 mM). To each well was added 250 µl of medium containing the RP-HPLC fractions or synthetic peptides in various concentrations and 50 µl of bacterial inoculum. The bacterial inocula were prepared by diluting exponentially growing cells in growth medium. For the S. mutans assays 4.5 × 10³ or 4.5 × 10⁵ viable cells/ml were used in each well. For the P. gingivalis and E. coli growth assays 8.3 × 10⁷ viable cells/ml were used in each well. Initial studies with P. gingivalis showed that this bacterium did not grow reproducibly in this assay when inocula with fewer viable cells were used. The P. gingivalis inoculum was prepared under strictly anaerobic conditions in an anaerobic workstation with an atmosphere of 85% N₂, 10% CO₂, and 5% H₂ (Mk3 Don Whitley Scientific Ltd, Adelaide, Australia). The P. gingivalis inoculum was added to the microtiter plate containing growth medium, which was then sealed. Control assays contained all components except the test fraction. The negative control wells each contained 245 µl of medium, 50 µl of inoculum, and 5 µl of gramicidin D (41 µM). Growth was monitored hourly over a 40-h period by measuring the optical density of the culture at 620 nm using an iEMS microplate reader (Labsystems OY Research Technologies Division) that incubated the cultures at 37°C. The MIC was determined as the lowest concentration of peptide required to completely inhibit the growth of the bacterium. For the MIC assay peptide concentration was varied using twofold serial dilutions.

Amino acid analysis. RP-HPLC fractions were subjected to gas phase hydrolysis using 6 M HCl with 1% (vol/vol) phenol for 24 h at 110°C. Hydrolysates were dried and resuspended in 10 µl of 0.25 M borate, pH 8.5, containing 10 µl of hydroxyproline. Amino acid analysis was performed by precolumn Fmoc derivatization using a GBC amino acid analysis system. The derivatized amino acids were separated on an octyldecyl silane RP column (150 by 4.6 mm; Hypersil) and detected by an LC1250 fluorescence detector. Each sample was analyzed in duplicate and at two different concentrations.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Antibacterial activity of CMP and components. The CMP preparation produced by chymosin digestion of casein followed by gel filtration-HPLC was heterogeneous, containing variably phosphorylated and glycosylated forms of the A and B genetic variants, as determined by MS and N-terminal amino acid sequence analysis. The CMP preparation had growth-inhibitory activity against both gram-positive and gram-negative bacterial species. The growth of S. mutans with an inoculum of 4.5 × 10³ viable cells/ml was inhibited by the CMP preparation in a concentration-dependent manner (MIC, 1.7 mg/ml) (Table 1). Increasing the inoculum to 4.5 × 10⁵ viable cells/ml resulted in a doubling of the MIC. The growth-inhibitory activity of CMP was also determined for two gram-negative species with inocula containing 8.3 × 10⁷ viable cells/ml. CMP inhibited the growth of both P. gingivalis and E. coli under these conditions (MICs, 3.8 and 4.3 mg/ml, respectively).

View this table: [in this window] [in a new window]   TABLE 1.   MICs of various forms of -casein(106-169) and synthetic -casein-A(138-158) against S. mutans

View larger version (19K): [in this window] [in a new window]   FIG. 1.   RP-HPLC of CMP. CMP was dissolved in solvent A (0.1% TFA in deionized water), applied to a preparative RP-HPLC column (C18), and then eluted using a linear gradient of 15% solvent B in 5 min, 15 to 20% solvent B (90% acetonitrile-0.1% [vol/vol] TFA in deionized water) in 1 min, and 20 to 50% solvent B in 155 min at a flow rate of 4.0 ml/min. Peaks were detected at 215 nm.

Enzymatic digestion of CMP and analysis of peptides. In order to characterize the active region of -casein(106-169), the peptide was subjected to hydrolysis with endoproteinase Glu-C. The resulting peptides were separated using RP-HPLC (Fig. 2) and 12 fractions were collected and tested for antibacterial activity. Fractions 1 to 11 contained mixtures of glycosylated -casein(106-169) fragments as determined by MS analysis. These fractions exhibited little growth-inhibitory activity against S. mutans, whereas fraction 12 at 0.14 mg/ml inhibited the growth of the bacterium by 84% (Table 3). Fraction 12 contained the nonglycosylated peptides Ser(P)¹⁴⁹-casein-A(138-158), Ser(P)¹⁴⁹-casein-A(148-169) and Ser(P)¹⁴⁹-casein-B(148-169) as determined by MS and N-terminal sequence analysis (Table 3). This fraction was subjected to gel filtration-HPLC, and the peptides Ser(P)¹⁴⁹-casein-A(148-169) and Ser(P)¹⁴⁹ -casein-B(148-169) were collected and shown to exhibit no growth-inhibitory activity against S. mutans. The peptide Ser(P)¹⁴⁹ -casein-A(138-158), however, did inhibit the growth of S. mutans, accounting for all the antibacterial activity of fraction 12. 

View larger version (20K): [in this window] [in a new window]   FIG. 2.   RP-HPLC of peptides generated by endoproteinase Glu-C hydrolysis of CMP. Generated peptides were applied to an analytical RP-HPLC column (C18), and peptides were eluted using a gradient of 0 to 60% solvent B in 24 min.

Antibacterial activity of synthetic peptides corresponding to -casein-A(138-158). To confirm that antibacterial activity was associated with Ser(P)¹⁴⁹-casein-A(138-158) and to determine the importance of phosphorylation of the Ser¹⁴⁹ residue, two peptides were chemically synthesized, corresponding to -casein-A(138-158), AVESTVATLEDSPEVIESPPE, and Ser(P)¹⁴⁹-casein-A(138-158), AVESTVATLEDPEVIESPPE, where  is phosphoserine. The synthetic peptides were purified by RP-HPLC and then analyzed by MS (Fig. 3). The m/z values obtained for the synthetic peptides were 2,200 and 2,280, corresponding to -casein-A(138-158) (calculated mass, 2,199 Da) and Ser(P)¹⁴⁹-casein-A(138-158) (calculated mass, 2,279 Da), respectively. The synthetic nonphosphorylated -casein-A(138-158) at concentrations up to 2.2 mg/ml (1 mM) did not inhibit the growth of S. mutans, whereas the synthetic, phosphorylated form of the peptide was inhibitory (MIC, 59 µg/ml [26 µM]) (Table 1).

View larger version (27K): [in this window] [in a new window]   FIG. 3.   RP-HPLC of synthetic -casein-A(138-158) (A) and Ser(P)149-casein-A(138-158) (C). The synthetic peptide was applied to the preparative column and eluted using a gradient of 0 to 100% solvent B in 30 min. Solvent A contained 0.1% TFA in deionized water, and solvent B contained 90% acetonitrile-0.1% TFA in deionized water. MS analysis of purified -casein-A(138-158) (B) and Ser(P)149-casein-A(138-158) (D). A sample of peptide was dissolved in 2,5-dihydroxybenzoic acid in 30% acetonitrile, containing 0.1% TFA, and applied to a MALDI-TOF mass spectrometer.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

Bovine milk contains a variety of biologically active peptides that may be released by enzymatic proteolysis during digestion and/or food processing. These peptides have been shown to exhibit a variety of bioactivities, such as antithrombotic, antihypertensive, immunostimulating, and opioidal properties (14). Peptides derived from the milk proteins lactoferrin (1) and _s2-casein (29) have been shown to exhibit activity against a range of gram-positive and gram-negative bacteria. It has been suggested that the release of these antibacterial peptides upon milk ingestion may assist the developing immune system of the neonate to control the intestinal microflora (1). However, whether these peptides are actually released in situ upon milk ingestion is unknown.

The peptide released by natural digestion of casein in the stomach, with chymosin, CMP (11) has been shown to substantially inhibit the adherence of Streptococcus mutans to salivary pellicle (17, 21, 25, 26), and this has been proposed to be the mechanism of anticariogenicity of milk protein fractions in animal caries models (7). In this study we have shown that CMP has growth-inhibitory activity against S. mutans and P. gingivalis as well as E. coli. CMP is the relatively hydrophilic C-terminal fragment of -casein(106-169) and contains all of the posttranslational modification sites. As a result, CMP is heterogeneous with different glycosylated and phosphorylated forms of the genetic variants.

CMP was fractionated by RP-HPLC, and the various fractions were tested for activity against S. mutans. The early eluting material upon RP-HPLC of CMP (RP1, RP2, and RP3) (Fig. 1) corresponded to glycosylated CMP, with m/z values of >7,400, as determined by MS analysis. The early elution of the glycosylated forms of CMP upon RP-HPLC is consistent with the previous studies of Molle and Leonil (16) and Minkiewicz et al. (15). No growth-inhibitory activity was observed with the glycosylated forms of CMP when tested against S. mutans. However, the later eluting fractions RP4 and RP5 did display growth-inhibitory activity against S. mutans (MICs, 0.68 and 1.04 mg/ml, respectively) (Table 2). Both RP4 and RP5 were shown by MS and sequence analysis to correspond to the nonglycosylated, phosphorylated -casein(106-169), with RP4 corresponding to variant A and RP5 corresponding to variant B. The later elution of variant B relative to variant A upon RP-HPLC is consistent with the increased hydrophobicity associated with the Thr¹³⁶Ile¹³⁶ and Asp¹⁴⁸Ala¹⁴⁸ substitutions in variant B. These substitutions also resulted in a slightly lower specific antibacterial activity of variant B (Table 2).

Analysis of the products of endoproteinase Glu-C hydrolysis of CMP revealed that the only fragment generated by the enzyme that exhibited antibacterial activity was Ser(P)¹⁴⁹-casein-A(138-158). Peptides corresponding to -casein-A(138-158) and Ser(P)¹⁴⁹-casein-A(138-158) were synthesized and tested for activity against S. mutans. This confirmed the growth-inhibitory activity of Ser(P) ¹⁴⁹-casein-A(138-158) and that phosphorylation of Ser¹⁴⁹ was critical for activity. An MIC of 59 µg/ml (26 µM) was obtained for the synthetic phosphopeptide Ser(P)¹⁴⁹-casein-A(138-158).

We have designated the antibacterial peptide from bovine milk -casein, Ser(P)¹⁴⁹-casein(106-169), kappacin, and this peptide shares very few characteristics with the cationic amphipathic antibacterial peptides so far characterized from vertebrates (13). These positively charged peptides are believed to initially make contact with the negatively charged phospholipid bilayer and insert into the microbial membrane. Once in the lipid bilayer the peptides, through an amphipathic helical structure, are thought to aggregate and form an ion channel (pore) spanning the membrane, allowing ions to equilibrate across the membrane, resulting ultimately in cell death (13). Molecular modeling and secondary-structure predictions of kappacin revealed that the peptide contained residues that could form an amphipathic helical structure. Structural studies have not yet been performed on kappacin, but recently Plowman et al. (19) showed using ¹H nuclear magnetic resonance spectroscopy that the nonphosphorylated -casein-B(130-153) formed an amphipathic helix spanning residues 136 to 149 at pH 5.0 in the presence of 40% (vol/vol) trifluoroethanol (19). These results are consistent with kappacin's containing residues that may form an amphipathic helical structure in a bacterial membrane environment. Kappacin molecules in this environment may aggregate to form an anionic pore, increasing the permeability of the membrane to cations. In an acidic environment, kappacin may allow the influx of hydrogen ions to lower intracellular pH. This mechanism is consistent with our finding that kappacin's activity against S. mutans increases at lower pH values (E. C. Reynolds, S. G. Dashper, M. Malkoski, N. M. O'Brien-Simpson, G. H. Talbo, and K. J. Cross, November 1989, Australia patent application PCT/AU98/0097; unpublished data). Earlier studies have shown that S. mutans will grow at pH 7.0 but not at lower pH values in the presence of the ionophore gramicidin, which collapses transmembrane cation gradients through formation of an anionic pore (3). CMP has been detected in the stomach, duodenum, and jejunum of humans after milk ingestion (11). The release of kappacin in the stomach therefore may be a mechanism to limit gastrointestinal tract infection in the developing neonate, by increasing the sensitivity of bacteria to gastric acid by collapsing essential transmembrane cation gradients. The reason why the glycosylated forms of kappacin lacked antibacterial activity is not clear, but it is interesting that molecular modeling suggests that the sugar moieties would block pore formation.

The exact antibacterial mechanism of kappacin in inhibiting the growth of bacteria remains to be determined. Recently Wu et al. (27) tested a range of cationic antibacterial peptides and showed that only a few actually caused an increase in membrane permeability, suggesting that the antibacterial activity might involve mechanisms other than the primary disruption of the microbial membrane. Kappacin does not exhibit sequence similarity with the cationic antibacterial peptides and apart from a propensity to form an amphipathic helical structure does not posses any of the other characteristics of these peptides. The sequence of kappacin is, in fact, unique among the known antibacterial peptides; however, the peptide does share some characteristics with a new class of anionic antibacterial peptides that includes chromacin and enkelytin. These peptides, identified in bovine adrenal medullary chromaffin granules, are anionic, contain a number of glutamyl residues, and are phosphorylated, similar to kappacin (6, 22). Further, like kappacin, the phosphorylation of enkelytin is essential for antibacterial activity (5). The antibacterial mechanism of the chromaffin peptides has not been elucidated; however, recent ¹H nuclear magnetic resonance spectroscopic analysis of synthetic nonphosphorylated enkelytin demonstrated that the peptide adopts a proline-kinked amphipathic helical structure in 50% (vol/vol) trifluoroethanol (10). The structure of the phosphorylated form of enkelytin has not been determined; however, phosphorylation is likely to change the peptide's conformation through electrostatic repulsion or by divalent metal ion binding (5, 10). The binding of divalent cations may act to stabilize the peptide structure but also may help to localize the peptide at the bacterial cell surface. It remains unclear how the negatively charged antibacterial peptides, including kappacin, interact with the bacterial cell surface. Phosphorylation of enkelytin and kappacin not only may be essential for conformation but also may allow divalent metal ion cross-linking with bacterial membrane phospholipids. It is interesting that the antibacterial fragment of kappacin, Ser(P)¹⁴⁹-casein-A(138-158), contains an internal prolyl residue, Pro¹⁵⁰, similar to enkelytin, such that the conformation of kappacin may also be a proline-kinked amphipathic helix, and therefore the mechanism of action of the two peptides may be similar.

In conclusion, we have demonstrated that CMP inhibits the growth of the oral pathogens S. mutans and P. gingivalis as well as E. coli, and we have identified the active component as the nonglycosylated Ser(P)¹⁴⁹-casein-A(106-169), designated kappacin. Kappacin is novel with respect to its amino acid sequence, but the peptide does share some characteristics with a new class of anionic antibacterial peptides. Kappacin may have utility in foods and oral care products to help lower the risk of dental caries and chronic periodontitis associated with S. mutans and P. gingivalis, respectively.