Basis of beneficial immunomodulation by ... - Wiley Online Library

RESEARCH ARTICLE

Basis of bene¢cial immunomodulation by monoclonal antibodies against Streptococcus mutans adhesin P1 Ryutaro Isoda, Rebekah A. Robinette, Trina L. Pinder, William P. McArthur & L. Jeannine Brady Department of Oral Biology, University of Florida College of Dentistry, Gainesville, FL, USA

Correspondence: L. Jeannine Brady, Department of Oral Biology, College of Dentistry, University of Florida, Gainesville, FL 32610, USA. Tel.: 1352 846 0785; fax: 1351 846 0786; e-mail: [email protected] Received 20 March 2007; revised 4 May 2007; accepted 7 May 2007. First published online 5 July 2007. DOI:10.1111/j.1574-695X.2007.00279.x Editor: Willem van Eden Keywords Streptococcus mutans ; immunomodulation; monoclonal antibodies; bacterial adhesin; discontinuous epitopes; Antigen I/II family.

Abstract We previously identified five monoclonal antibodies (MAbs) against Streptococcus mutans adhesin P1 that modulate the humoral response when bound to whole bacteria and immune complexes (ICs) are administered to BALB/c mice. The two MAbs that redirected the response towards increased efficacy recognize discontinuous epitopes involving pre-alanine-rich domain sequence; therefore, to evaluate whether epitope specificity contributes to a desirable outcome a further MAb with this characteristic was tested. A beneficial immune response was promoted. None of the three MAbs that promoted a beneficial response was opsonic, suggesting that increased uptake of ICs by phagocytes does not mediate the improvement of the IC-elicited antibodies to inhibit bacterial adherence. Finally, two of the six antiP1 MAbs activated complement but did not partition according to desirable vs. nondesirable effects.

Introduction The immune response against numerous antigens is different when the antigens are administered as part of immune complexes (ICs) containing specific polyclonal or monoclonal antibodies (MAbs) from when immunization is with antigen alone (reviewed in Brady, 2005; Getahun & Heyman, 2006). We have studied the practicality of utilizing MAbs to redirect the immune response towards one of enhanced benefit against the surface adhesin P1 of Streptococcus mutans (Brady et al., 2000; Oli et al., 2004; Rhodin et al., 2004a, b), the predominant etiologic agent of dental caries (Hamada & Slade, 1980). In our previous work, the induction of antibodies more inhibitory of bacterial adherence was increased significantly when ICs containing S. mutans and certain anti-P1 MAbs were administered mucosally (Brady et al., 2000) or parenterally (Oli et al., 2004). Differences in the specificity and isotype composition of IC-elicited antibodies were apparent compared with immunization with bacteria alone, and similar changes in the specificity of anti-P1 secretory IgA (sIgA) and serum IgG antibodies were observed (Rhodin et al., 2004a). P1, also called PAc (Okahashi et al., 1989) and Antigen I/II (Russell & Lehner, 1978), belongs to the Antigen I/II

2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

family of multi-domain adhesins expressed by most oral streptococci and involved in interactions with host macromolecules and other bacteria (reviewed in Jenkinson & Demuth, 1997). It mediates bacterial adherence to a highmolecular-weight glycoprotein, salivary agglutinin (SAG) (Demuth et al., 1988), now known to represent the human salivary scavenger protein gp340 (Prakobphol et al., 2000) contained within the salivary pellicle on the tooth surface. Antigen I/II family proteins all share a common architecture and contain an N-terminal signal sequence, a series of alanine-rich tandem repeats within the amino-terminal third of the molecule, a c. 150 amino acid residue variable region where most sequence variation is clustered, a series of proline-rich tandem repeats in the central portion of the molecule, and carboxy-terminal sequences characteristic of wall and membrane spanning domains of streptococcal surface proteins, including the LPXTG motif conserved among sortase substrates (Fischetti et al., 1990). A schematic representation of P1 is shown in Fig. 1. P1 has been widely studied as a vaccine candidate against dental caries, and both serum IgG and sIgA contribute to protection (reviewed in Michalek et al., 2001). Passive immunization of human subjects with the anti-Antigen I/II MAb Guy’s 13 has been reported to confer long-term

51 (2007) 102–111

103

Immunomodulation by streptococcal monoclonal antibodies

Signal A-Region Sequence (a.a. 1-38) (a.a. 186-464)

P-Region (a.a. 840-963)

Wall and Membrane Spanning Region (a.a. 1486-1556)

1-6F 4-9D 4-10A

+

5-5D, 6-11A

+

Cytoplasmic Tail (a.a. 1557-1561) 3-10E

Fig. 1. Schematic representation of P1 including relevant domains and the minimal sequences known to reconstitute epitopes recognized by anti-P1 MAbs. The A- and P-regions correspond to the alanine-rich and proline-rich tandem repeat domains, respectively. The intervening sequence between the A- and P-regions represents the variable region where the majority of amino acid differences between Antigen I/II family polypeptides are clustered. Continuous lines indicate that the epitope is achieved within that single polypeptide fragment. Discontinuous lines indicate that both segments are necessary and can be added together as separate fragments to achieve the epitopes (McArthur et al., 2007).

protection against recolonization with S. mutans (Ma et al., 1987). Because antibodies can act as immunomodulatory agents and influence the adaptive immune response against the antigens to which they bind, a more complete understanding of this phenomenon has important implications for both active and passive immunization approaches. Our earlier studies identified five IgG1 anti-P1 MAbs that modulate the immune response against P1 when bound to S. mutans and the ICs are administered intraperitoneally to immunize BALB/c mice (Oli et al., 2004). When included as part of an IC, MAbs 6-11A and 3-10E increased the formation of serum antibodies capable of inhibiting bacterial adherence to SAG, as measured using a whole bacterial cell Biacores surface plasmon resonance assay (Oli et al., 2006), but these MAbs themselves do not inhibit adherence to the immobilized ligand (Brady et al., 1992; Oli et al., 2004). MAbs 6-11A and 3-10E share the common characteristic of recognizing discontinuous epitopes involving a c. 100-amino acid segment immediately upstream of the alanine-rich repeat domain (A-region) (Rhodin et al., 2004b; McArthur et al., 2007) (see Fig. 1). In contrast, the pre-A-region does not contribute to epitopes recognized by MAbs 1-6F, 4-9D or 410A, all of which themselves inhibit adherence, but either negatively influence or have no observable effect on the efficacy of the elicited response when administered as part of an IC (Oli et al., 2004). Based on these findings we wished to determine whether the functional activity and specificity of an anti-P1 MAb could be used to predict a beneficial effect on the host response. We therefore evaluated a sixth IgG1 anti-P1 MAb, namely 5-5D. Like 6-11A and 3-10E, its epitope is contributed to by a sequence contained within the pre-Aregion segment (McArthur et al., 2007). Potential mechanisms of antibody-mediated immunomodulation include masking or exposure of dominant vs. 51 (2007) 102–111

cryptic epitopes upon antibody binding, alterations in proteolysis, antigen processing and presentation, changes in cytokine expression by antigen-presenting cells (APCs) and/ or T cells, increased uptake of antigen via Fc receptors (FcR) on APCs, differential engagement of stimulatory vs. inhibitory FcR, enhanced germinal centre (GC) formation, development of memory B cells and maturation of antibody affinity, changes in usage of germline-encoded VH genes, and induction of somatic hypermutation (for reviews see Brady, 2005; Getahun & Heyman, 2006). The formation of GC in which memory B cells are generated is facilitated by the trapping of IC and activation of complement in the network of follicular dendritic cells (DCs) and/or B cells in the lymphatic follicles. In addition, within APCs, major histocompatibility complex (MHC) class II antigen presentation is affected by the endocytic compartment used for the processing of the internalized antigen. Opsonization of antigen by C3b and uptake via complement receptors has been shown to alter intracellular trafficking of internalized antigen compared with uptake via the B-cell receptor in B lymphocytes (Perrin-Cocon et al., 2004), and the resultant change in epitopes favoured for display to CD4 helper T cells could have downstream effects on the nature and specificity of an elicited antibody response. In addition, the uptake of antibody-bound antigen via activating Fc receptors on APCs can influence the nature of the immune response both in terms of lowering the concentration of antigen necessary for presentation by APCs (Manca et al., 1991) and by inducing APC maturation via receptor cross-linking by IC (Regnault et al., 1999). To determine whether the immunomodulatory mechanism(s) of anti-P1 MAbs involves classical pathway complement activation and/or increased uptake of MAbcoated S. mutans by phagocytic cells, each of the six MAbs was also tested for its ability to promote deposition of C3 on the bacterial surface and for opsonophagocytic activity using the J774A.1 mouse monocyte cell line.

Materials and methods Bacteria and growth conditions Serotype c S. mutans NG8 (kindly provided by K. W. Knox, Institute for Dental Research, Sydney, Australia) was grown aerobically to stationary phase for 16 h in Todd Hewitt broth supplemented with 0.3% yeast extract (BBL, Cockeysville, MD).

Anti-P1 monoclonal antibodies A panel of anti-P1 MAbs was generated and characterized previously (Ayakawa et al., 1987; Brady et al., 1992, 1998; Seifert et al., 2004; Rhodin et al., 2004b; McArthur et al., 2007). The anti-P1 IgG1 MAbs evaluated in the present study, namely 6-11A, 3-10E, 5-5D, 4-10A, 4-9D, I-6F and 6-8C, were used as ascites fluids at the indicated dilutions. 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

104

The minimal linear sequence or combinations of sequences required to form the epitopes recognized by these MAbs are summarized in Fig. 1.

Murine immunization and sample collection Groups of six 8-week-old female BALB/c mice (Charles River Laboratories, Wilmington, MA) were immunized intraperitoneally as described on days 0 and 14 with c. 1.5 109 CFU in 150 mL of PBS of S. mutans alone or of S. mutans coated with either a saturating ([high]) or 0.1 subsaturating ([low]) concentration of MAb 5-5D (Oli et al., 2004). The dilution of MAb necessary to saturate the bacterial cells was predetermined by serial titration and dot blot analysis as previously described (Brady et al., 2000). Mice were exsanguinated on day 45, and serum samples were stored at 20 1C. Control groups received PBS or MAb only, and did not develop antibodies against S. mutans. Mice in the three treatment groups developed comparable levels of serum antibodies against whole S. mutans cells and purified full-length P1 as determined by enzyme-linked immunosorbent assay (ELISA) (data not shown).

Inhibition of S. mutans adherence Salivary agglutinin (SAG) was prepared by a modification of the method of Rundegren & Arnold (1987) (Brady et al., 1992). The relative ability of anti-P1 MAbs, or of pooled sera from each of the three treatment groups immunized with S. mutans or with S. mutans coated with anti-P1 MAb 5-5D, to inhibit P1-mediated adherence of S. mutans NG8 whole cells to immobilized SAG was evaluated in vitro as previously described using a Biacores surface plasmon resonance (SPR) assay (Oli et al., 2006). A spaP-negative isogenic mutant devoid of P1 (Crowley et al., 1999) that does not adhere above background levels to the control surface was used in this assay to validate that the detected change in resonance signal was P1-mediated and not the result of an interaction of non-P1 surface components with low levels of potential contaminants, such as sIgA, in the agglutinin preparation. The percentage inhibition of adherence of S. mutans to SAG in the presence of antibody was calculated as: [(change in resonance units (DRU) without test sera or MAb DRU with test sera or MAb)/DRU without test sera or MAb] 100. Each experiment was repeated four times.

Preparation of monocytic cells The mouse macrophage-like cell line J774A.1 (ATCC TIB67, purchased from ATCC) was maintained following the instructions provided by ATCC, and cultured in Dulbecco’s modified Eagle medium (D-MEM) (Invitrogen Corporation, Carlsbad, CA) supplemented with 100 Units mL 1 penicillin, 100 mg mL 1 streptomycin, 4 mM L-glutamine 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

R. Isoda et al.

and 10% heat-inactivated fetal bovine serum (Invitrogen Corporation). For each opsonophagocytosis experiment, J774A.1 cells were grown to near confluency in 75-cm2 cellculture flasks and cell-culture medium, and floating cells were removed by aspiration. Adherent cells were washed once with ice-cold Hank’s balanced salt solution (HBSS, Invitrogen Corporation). Fresh ice-cold HBSS was added to the flask and adherent cells were physically dislodged with a cell scraper, washed once by centrifugation at 250 g for 2 min, counted under a microscope following staining with trypan blue, and resuspended at 2.5 107 viable cells mL 1 with ice-cold HBSS. Only mononuclear cell recoveries with viability 4 85% were used in the opsonophagocytosis assay.

Opsonophagocytosis assay The opsonic activity of each anti-P1 MAb was examined using the method of Drevets & Campbell (1991) with the following modifications. In brief, 2.5 106 J774A.1 cells, 2.5 107 S. mutans and murine ascites fluid (1 : 1000) were mixed in 1 mL of HBSS and rotated end-over-end for 30 min at room temperature in polypropylene snap-cap tubes (12 by 75 mm; Falcon; Becton-Dickinson Labware, Lincoln Park, NJ). Bacteria not associated with the mononuclear cells were removed by centrifugation at 300 g for 8 min at 4 1C and two washes with 2.0 mL of ice-cold HBSS. Bacteria associated with the J774A.1 cells were enumerated by lysing the J774A.1 cells with 1.0 mL of sterile distilled water, making 10-fold dilutions of the lysed cells and plating onto Todd Hewitt containing 0.3% yeast extract (THYE) agar plates. Uptake of S. mutans, or lack thereof, by the monocytic cells was confirmed visually by light microscopy (magnification 1000), whereby the cells were resuspended in 1.0 mL PBS instead of water and 0.1 mL was removed to make cytocentrifuge slides, which were fixed, stained with Diff Quick (AHS delCaribe, Aquado, Puerto Rico), and examined under oil immersion (not shown).

Statistical analysis One-factor ANOVA was used to detect differences between the treatment groups. If significant, individual differences were evaluated further by the Scheffe´ test.

Complement activation by anti-P1 MAbs Streptococcus mutans NG8 cells were fixed with 2% paraformaldehyde at 4 1C for 2 h using the method of RuizOlvera et al. (2003). The fixed cells were washed with PBS, adjusted to 1 108 cells mL 1 in 1% bovine serum albumin (BSA)-PBS supplemented with 0.9 mM CaCl2 and 0.33 mM MgCl2, and 200 mL of the cell suspension was incubated for 1 h at room temperature with murine ascites fluids containing each anti-P1 MAb (1 : 100 dilution) or buffer only. 51 (2007) 102–111

105


(a) 600

Response difference (RU)

500

no MAb 3-10E 5- 5D 4- 9D

400 300 200 100 0

−50

0

50

−100

100

105

200

Time (sec.)

(b) 100% * 80% % Inhibition

Classical pathway vs. alternative pathway activation was distinguished as described (Wang et al., 2005) with some modification. Classical pathway activation was determined following a 3-min incubation with ImmunoOTM guinea pig complement (MP Biomedicals, Aurora, OH) (1 : 200) in the presence of test or negative-control antibodies. Alternative pathway activation was identified following a 30-min incubation with guinea pig complement in the absence of antibody. The classical pathway reaction was stopped by the addition of 1.2 mL of ice-cold stopping solution (1% BSA, 10 mM EGTA in PBS) and by immediately collecting the cells by centrifugation. Cells were washed twice with PBS and incubated for 1 h at room temperature in 100 mL of 1% BSA-PBS containing fluorescein-conjugated goat IgG fraction to guinea pig complement C3 (MP Biomedicals) (1 : 400) and tetramethyl rhodamine iso-thiocyanate (TRITC)-conjugated goat anti-mouse IgG (Southern Biotech, Birmingham, AL) (1 : 200). Cells were washed twice, resuspended in 50 mL of PBS, mixed with an equal volume of VectaSheild HardSetTM (Vector Laboratories, Burlingame, CA), and 10 mL of each suspension was mounted onto a glass slide and sealed with a coverslip. Stained cells were viewed with a Zeiss Axioplan 2 florescence microscope at 1000 magnification under oil immersion, and pictures were taken with an attached digital camera using SPOT advanced version 4.04 software (Diagnostic Instrument Inc., Sterling Heights, MI). Positive and negative controls for these experiments included a polyclonal murine antiserum raised against S. mutans NG8 whole cells in parallel with a sham-immunized vehicle-only serum, as well as the anti-P1 MAb 6-8C, which maps to the wall-anchored C-terminal portion of the protein (Brady et al., 1991) and whose epitope is thereby not exposed on S. mutans whole cells. As an additional control, the complement source was also heat-inactivated at 56 1C for 60 min.

60% 40% 20% 0% 3-10E

4-9D

5-5D

Fig. 2. Inhibition of Streptococcus mutans adherence to immobilized SAG by anti-P1 MAbs measured by Biacores surface plasmon resonance. (a) Sensorgram showing the interaction of S. mutans NG8 with SAG in the absence (red line) and presence of anti-P1 MAbs 3-10E (green line), 4-9D (black line), or 5-5D (blue line). Bacterial cells were injected for 60 s and the change in resonance signal (DRU) is indicated on the y-axis. (b) Percentage inhibition of adherence of S. mutans to SAG by anti-P1 MAbs was calculated as ([mean DRU without MAb mean DRU with MAb]/mean DRU without MAB) 100. Data represent the average of three independent experiments. Error bars indicate SD. (P o 0.01 compared with inhibition by an isotype-matched irrelevant negativecontrol Mab.)

Results Inhibition of S. mutans adherence to immobilized SAG by MAb 5-5D Because of the previously observed inverse relationship between the ability of anti-P1 MAbs to promote a desirable immune response and their ability to inhibit bacterial adherence directly (Oli et al., 2004), we first used a wholecell Biacores assay to test whether anti-P1 MAb 5-5D blocks the interaction of S. mutans with immobilized SAG (Fig. 2). A representative sensorgram of NG8 binding to immobilized SAG in the presence and absence of anti-P1 MAbs is shown in Fig. 2(a). The percentage inhibition of NG8 adherence to SAG in the presence of each MAb is shown in part (b). The MAbs 4-9D and 3-10E were used as positive and negative controls in this experiment. MAb 4-9D is known to be a strong inhibitor of bacterial adherence (Brady 51 (2007) 102–111

et al., 1991; Oli et al., 2004), whereas MAb 3-10E is no more inhibitory of adherence than an IgG1 isotype-matched irrelevant anti-Actinobacillus actinomycetemcomitans control (Oli et al., 2006). MAb 5-5D was no more inhibitory of S. mutans adherence to immobilized SAG than the negative control MAb 3-10E.

Inhibition of S. mutans adherence by sera from IC-immunized mice To continue to explore whether a beneficial immunomodulatory effect can be predicted based on the epitope specificity and functional activity of an anti-P1 MAb, mice were immunized with S. mutans alone or with S. mutans coated with two different concentrations of MAb 5-5D, and the antisera from the three treatment groups were compared 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

106

R. Isoda et al.

16

50%

*

*

14

Ave. CFU (x106)

% Inhibition

40% 30% 20%

*

12 10

*

8

*

6 4 2

10%

0

0% S. mutans

[High]

[Low]

Fig. 3. Inhibition of adherence of Streptococcus mutans to human SAG by sera from mice immunized with bacteria alone compared with bacteria coated with MAb 5-5D. The percentage inhibition of adherence of S. mutans to immobilized SAG was measured using the surface plasmon resonance assay described in the text. High and Low indicate groups of mice immunized with S. mutans coated with saturating and subsaturating concentrations of the MAb, respectively. (P o 0.05 compared with mice immunized with S. mutans alone.)

using the Biacores assay for their relative ability to inhibit bacterial adherence (Fig. 3). Depending on the MAb, immunomodulatory effects can vary with the ratio of antibody to antigen, and more antibody does not necessarily confer an optimal effect; therefore, immunizations were performed with IC containing saturating (high) or 0.1 subsaturating (low) concentrations of 5-5D as described previously (Brady et al., 2000; Oli et al., 2004). As expected, sera from mice immunized with S. mutans coated with 5-5D were significantly more inhibitory of bacterial adherence compared with sera from mice immunized with bacteria alone (P o 0.05), indicating that in this system a known functional characteristic, i.e. lack of adherence inhibition, and epitope specificity are useful predictors of a beneficial immunomodulatory effect. In the case of 5-5D, an increase in the formation of adherence-inhibiting antibodies was observed when mice were immunized with S. mutans coated with either high or low concentrations of MAb. Sera from each of the three treatment groups demonstrated similar anti-S. mutans and anti-P1 titers by ELISA (not shown). This is consistent with our previous findings that the change in adherence inhibition reflects qualitative changes, i.e. specificity and isotype composition, of elicited antibodies rather than simple quantitative differences in the overall intensity of the response (Brady et al., 2000; Oli et al., 2004).

Opsonic activity of anti-P1 MAbs Increased uptake of ICs via Fc and/or C receptors on APCs are potential mechanisms of antibody-mediated immunomodulation, and the J774A.1 macrophage-like cell line 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

6-8C 1-6F (Negative Control)

4-9D

4-10A

3-10E

5-5D

6-11A

MAbs

Fig. 4. Opsonic activity of anti-P1 MAbs. The uptake of Streptococcus mutans by J774A.1 macrophage-like cells was evaluated in the presence of the six anti-P1 test MAbs or the negative-control anti-P1 MAb 6-8C. J774 A.1 cells were washed, lysed with water, and S. mutans were enumerated by plating onto THYE agar plate. (P o 0.01 compared with uptake in the absence of antibody.)

represents an established APC model (Guy & Belosevic, 1993; Bohm et al., 1995). It expresses both FcgRI and complement receptor type 3 (CR3) (Sears et al., 1990; Guy & Belosevic, 1993; Taborda & Casadevall, 2002). To determine whether the beneficial immunomodulatory activity of anti-P1 MAbs is related to their ability to promote opsonophagocytosis, uptake of S. mutans by J774A.1 cells in the presence of each of the six anti-P1 MAbs was evaluated (Fig. 4). Another anti-P1 MAb, 6-8C, which maps to the C-terminus of P1, was included as a negative control because its epitope is masked when P1 is present on the cell surface and it does not bind to S. mutans whole cells (Brady et al., 1991). All six of the test MAbs are known to react with P1 on the cell surface (Brady et al., 1991). None of the three MAbs that redirect the host response towards one of increased benefit, namely 6-11A, 3-10E and 5-5D, was opsonic, suggesting that their mechanism of immunomodulation is not simply to facilitate Fc-mediated uptake of antigen by APCs. On the other hand, those three MAbs that significantly decreased the efficacy of the elicited antibody response (1-6F and 4-9D), or resulted in antibody changes that did not affect bacterial adherence (4-10A) (Oli et al., 2004), demonstrated significant opsonic activity (P o 0.01).

Complement activation by anti-P1 MAbs Complement activation and deposition of C3 on S. mutans whole cells in the presence and absence of anti-P1 antibodies was evaluated using fluorescein isothiocyanate (FITC)labelled goat anti-guinea pig C3 and immunofluorescent microscopy (Fig. 5a). A polyclonal murine anti-S. mutans NG8 whole-cell antiserum was used as a positive control, and sham-immunized serum was used as a negative control 51 (2007) 102–111

107


Fig. 5. Complement deposition on Streptococcus mutans in the presence and absence of MAbs. (a) Deposition of C3 on the surface of S. mutans NG8 was detected by immunofluorescent microscopy using FITC-conjugated anti-guinea pig C3 antibody. Classical pathway activation was evaluated after a 3-min incubation with a polyclonal anti-NG8 antiserum (P) or sham-immunized vehicle-only nonimmune control serum (N), with anti-P1 test MAbs as indicated, with a negative-control MAb that is nonreactive with the bacterial surface (6-8C), or in the absence of antibody AB( ). Alternative pathway activation (Alt) was evaluated after a 30-min incubation in the absence of antibody. HI indicates heat inactivation of the complement source. (b) As a control to ensure appropriate binding of anti-P1 antibodies, or lack thereof, cells were also stained with a TRITC-conjugated secondary antibody and the same fields were evaluated for red fluorescence.

51 (2007) 102–111


c

108

R. Isoda et al.

Table 1. Summary of characteristics of anti-P1 MAbs Isotype Adherence inhibition by MAb Change in adherence inhibition by MAb-modulated serum Contribution of pre-A-region sequence to epitope Opsonization Complement fixation Epitope exposed on bacterial surface

5-5D IgG1

6-11A IgG1

3-10E IgG1

1 1

1 1

1 1

1 1

1

1

4-10A IgG1

1-6F IgG1

4-9D IgG1

1 0

1

1

1

1 1 1

1

6-8C IgG1 NT

1 1

NT, not tested in immunization experiments. 0, no effect or no measurable change. 1, increase; , decrease.

in this experiment. Antibody-independent alternative pathway activation has been described for S. mutans (Inal et al., 1976) and was evident after a 30 min, but not after a 3 min, incubation of the bacteria with complement in the absence of antibody, indicating that antibody-mediated classical pathway activation and alternative pathway activation were differentiated in this assay. This activity was destroyed by heat inactivation of the complement at 56 1C for 60 min. Two of the six test MAbs, 5-5D and 1-6F, activated complement as evidenced by the detection of C3 on the bacterial surface. No C3 was detected on the bacteria in the absence of antibody or in the presence of any of the other anti-P1 Mabs, including the negative control MAb 6-8C, which is nonreactive with S. mutans whole cells. To ensure that differences in fluorescent detection of C3 were indeed reflective of complement activation, or lack thereof, and not of differences in the degree of antibody binding to the S. mutans cells, TRITC-labelled goat anti-mouse IgG was included in the secondary detection mixture and the same fields were also visualized for red fluorescence (Fig. 5b). Binding of the positive-control polyclonal antiserum and all anti-P1 MAbs, except the 6-8C negative control, was readily apparent. No reactivity of the secondary reagents was observed in the absence of primary antibody at either the 3- or 30-min timepoints, confirming the specificity of the detection.

Discussion To further our understanding of how MAbs against the S. mutans surface adhesin P1 influence the immune response against this important target of protective immunity when they are bound to the bacteria and antibody–antigen complexes are used as the immunogen, we examined the contribution of epitope specificity and functional activity of the MAbs themselves, as well as their ability to promote opsonophagocytosis and to activate complement. The results are summarized in Table 1. It is immediately apparent that a common denominator among the MAbs that promote the formation of antibodies that inhibit adherence 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

of S. mutans to immobilized salivary agglutinin is shared features of their cognate epitopes as well as their own inability to inhibit adherence directly. Anti-P1 (Antigen I/II) MAbs include several (4-10A, 6-11A, 5-5D, 3-10E, Guy’s 13) that recognize conformational determinants involving an interaction between the discontinuous A- and P-regions and/or segments flanking these domains (van Dolleweerd et al., 2003, 2004; Seifert et al., 2004; Rhodin et al., 2004b; McArthur et al., 2007) (see Fig. 1). A crystal structure of the intervening variable region suggests that the A- and P-regions would be in close proximity in the native molecule (Troffer-Charlier et al., 2002). In the context of our previous findings with the immunomodulatory MAbs 6-11A and 3-10E, it is telling that 5-5D demonstrates common characteristics. As shown in this report, it does not directly inhibit the adherence of S. mutans to salivary agglutinin, and it promotes the formation of adherence-inhibiting antibodies when included as part of an IC. Like 6-11A and 3-10E, it recognizes a complex epitope requiring the P-region and involving sequence immediately upstream of the A-region (McArthur et al., 2007). The three MAbs within this group are not identical, however, and can be distinguished from one another. For example, 5-5D is cross-reactive with SspB of Streptococcus gordonii and SpaA of Streptococcus sobrinus (Brady et al., 1991), and unlike 5-5D and 6-11A, 3-10E is completely nonreactive with a P1 deletion construct lacking amino acids 84–190 (McArthur et al., 2007). Therefore, the common immunomodulatory properties of these MAbs stem not from recognition of a single identical epitope, but from recognition of a family of related conformational epitopes requiring the interaction of similar discontinuous segments of P1. This implies that the mechanism responsible for their redirection of the immune response involves the nature of the antigen antibody interaction and requires binding that is dependent on structural elements of the adhesin itself. That binding to or interference with preA-region sequence may unveil epitopes recognized by adherence-inhibiting antibodies is illustrated by the increased

51 (2007) 102–111

109


reactivity of the adherence-inhibiting MAb 1-6F with anframe deletion construct lacking this segment compared to full-length P1 (McArthur et al., 2007). While antibodies can mask certain antigenic epitopes during an immune response, cryptic epitopes can be revealed as a consequence of antibody binding (Watts et al., 1998), presumably owing to alterations in protein conformation. An alteration in protein conformation would not only be expected to change the spectrum of B-cell clones stimulated during a humoral response, but also to impact antigenic processing and presentation and hence the stimulation of specific CD4 helper T-cell clones (Carmicle et al., 2007). The involvement of T cells in immunomodulation by anti-P1 MAbs has been suggested by changes not only in the specificity of antibodies in the serum of immunized animals but also in the relative concentrations of elicited IgG1, IgG2a and IgG2b isotypes (Brady et al., 2000; Oli et al., 2004; Rhodin et al., 2004a). The clinical association of recognition of particular Antigen I/II T-cell epitopes was highlighted in a study of naturally sensitized caries-resistant compared with caries-susceptible human subjects (Kelly et al., 1995). In contrast to 6-11A, 3-10E and 5-5D, MAbs 4-10A, 4-9D and 1-6F are strongly inhibitory of S. mutans adherence (Brady et al., 1992; Oli et al., 2004). MAb 4-10A had no observed effect on the formation of adherence-inhibiting antibodies when included as part of an immune complex (Oli et al., 2004) and recognized a discontinuous epitope fully reconstituted by the interaction of isolated A- and P-region polypeptides without a contribution of preA-region sequence (McArthur et al., 2007). The epitopes recognized by MAbs 4-9D and 1-6F do not depend on an A-region P-region interaction and also do not involve preA-region sequence. Both have a statistically significant negative influence on the generation of adherence-inhibiting antibodies when included as part of an IC (Oli et al., 2004), and both map to the variable region (McArthur et al., 2007), with 4-9D requiring additional contiguous sequence spanning into the A-region (Tucker et al., 2006). Adherence inhibition directly by 1-6F and 4-9D themselves suggests that epitopes in the proximity of the variable region represent a functionally important target of effective antibodies, and that IC containing these particular MAbs negatively influences the immunogenicity of this segment of P1. In terms of immunomodulation and functional capability, another clear-cut dichotomy among the six test MAbs was their opsonophagocytic activity compared with their beneficial vs. indifferent or deleterious effects (Table 1). ICs comprising antigen and antibody have been reported to interact with three broad types of receptors: the B-cell receptor, activating as well as inhibitory Fc receptors, and complement receptors, or combinations thereof (reviewed in Getahun & Heyman, 2006). Antibody-mediated immunomodulation can be either Fc-dependent or Fc-indepen51 (2007) 102–111

dent. Mice lacking FcRg, and hence functional FcgRI, FcgRIII, and FcgRIV activating receptors, show diminished IgG1-, IgG2a- and IgG2b-mediated enhancement of humoral responses (Wernersson et al., 1999). Activating FcgRs are expressed on APCs including macrophages, dendritic cells and follicular dendritic cells. Because an increase in antigen uptake via FcR on APCs represents a potential antibodymediated mechanism of immunomodulation (Lanzavecchia, 1990; Manca et al., 1991), we evaluated the ability of the anti-P1 MAbs to serve as opsonins and to promote phagocytosis by the macrophage-like J774.A1 cell line. Although 6-11A, 3-10E, and 5-5D all mediate an increase in the level of adherence-inhibiting antibodies, none of these MAbs was opsonic, indicating that a simple increase in Fcmediated uptake is not responsible for the beneficial change in the immune response observed when these antibodies are used to coat bacteria prior to immunization. Surprisingly, however, the other anti-P1 MAbs tested, 4-10A, 4-9D, and 1-6F, all demonstrated significant opsonophagocytic activity. As stated above, these three MAbs also influence the immunogenicity of P1, with the latter two significantly skewing antibody formation towards a less effective adherence-inhibiting response (Oli et al., 2004). Therefore, Fc-dependent uptake of S. mutans by an APC in the presence of certain antibodies apparently can shape the nature of a subsequent immune response; however, in this case the associated changes were not desirable. All six test MAbs are of the IgG1 subclass; therefore, it is not just the Fc-region of the antibody itself that dictates its opsonic activity, suggesting that in this system epitope specificity may govern the way each MAb interacts with P1 on the bacterial surface and can influence its ability to promote phagocytosis, perhaps owing to steric factors enabling some, but not all, bacterial-bound antibodies to interact functionally with Fc receptors on the APC. Finally, to eliminate the possibility that complement activation represents a primary mechanism of beneficial or detrimental MAb-mediated immunomodulation, we examined complement activation as evidenced by C3 deposition on the bacterial surface. Among the four subclasses of mouse IgG, the IgG1 subclass is known as a weak- or noncomplement-fixing subclass (Ey et al., 1979; Burton, 1987); however, classical pathway activation by a murine IgG1 MAb is not unknown (Wang et al., 2005). Two of the six anti-P1 IgG1 MAbs, 5-5D and 1-6F, were clearly able to activate complement as demonstrated by immunofluorescent microscopy. Because complement activation does not correspond to the known immunomodulatory properties of the anti-P1 MAbs it does not appear that this represents the major mechanism by which anti-P1 MAbs redirect the immune response towards one of improved vs. lessened benefit. However, it does not preclude that complement activation and engagement of complement receptors does 2007 Federation of European Microbiological Societies Published by Blackwell Publishing Ltd. All rights reserved

c

110

not in some way contribute to the composite effects of these two complement-fixing MAbs. In summary, our results indicate that the specificity and adherence-inhibiting function of anti-P1 MAbs are key factors in their effects on the adaptive humoral response. With the identification of 5-5D as a third MAb with desirable immunomodulatory properties, the reciprocal relationship between the ability to inhibit adherence and the promotion of an adherence-inhibiting response becomes more strongly established and makes sense in the context of revealed vs. masked epitopes. Lack of adherence-inhibiting activity in combination with recognition of an epitope involving pre-A-region sequence therefore predicts a beneficial immunomodulatory property of anti-P1 MAbs. In addition, the partitioning of the MAbs into two groups with regard to opsonophagocytosis and promotion of beneficial compared with nonbeneficial responses suggests that Fcmediated uptake of ICs by APCs does not contribute to desirable immunomodulatory properties of nonadherenceinhibiting anti-P1 MAbs, but paradoxically may be involved in making less effective the immune response against ICs containing adherence-inhibiting MAbs.

Acknowledgements This work was supported by grant NIH DE013882 from the National Institute of Dental and Craniofacial Research (NIDCR).

References Ayakawa GY, Boushell LW, Crowley PJ, Erdos GW, McArthur WP & Bleiweis AS (1987) Isolation and characterization of monoclonal antibodies specific for antigen P1, a major surface protein of mutans streptococci. Infect Immun 55: 2759–2767. Bohm W, Schirmbeck R, Elbe A, Melber K, Diminky D, Kraal G, van Rooijen N, Barenholz Y & Reimann J (1995) Exogenous hepatitis B surface antigen particles processed by dendritic cells or macrophages prime murine MHC class I-restricted cytotoxic T lymphocytes in vivo. J Immunol 155: 3313–3321. Brady LJ (2005) Antibody-mediated immunomodulation: a strategy to improve host responses against microbial antigens. Infect Immun 73: 671–678. Brady LJ, Piacentini DA, Crowley PJ & Bleiweis AS (1991) Identification of monoclonal antibody-binding domains within antigen P1 of Streptococcus mutans and cross-reactivity with related surface antigens of oral streptococci. Infect Immun 59: 4425–4435. Brady LJ, Piacentini DA, Crowley PJ, Oyston PC & Bleiweis AS (1992) Differentiation of salivary agglutinin-mediated adherence and aggregation of mutans streptococci by use of monoclonal antibodies against the major surface adhesin P1. Infect Immun 60: 1008–1017.


c

R. Isoda et al.

Brady LJ, Cvitkovitch DG, Geric CM, Addison MN, Joyce JC, Crowley PJ & Bleiweis AS (1998) Deletion of the central proline-rich repeat domain results in altered antigenicity and lack of surface expression of the Streptococcus mutans P1 adhesin molecule. Infect Immun 66: 4274–4282. Brady LJ, van Tilburg ML, Alford CE & McArthur WP (2000) Monoclonal antibody-mediated modulation of the humoral immune response against mucosally applied Streptococcus mutans. Infect Immun 68: 1796–1805. Burton DR (1987) Molecular Genetics of Immunoglobulin. Elsevier Science B.V, Amsterdam. Carmicle S, Steede NK & Landry SJ (2007) Antigen threedimensional structure guides the processing and presentation of helper T-cell epitopes. Mol Immunol 44: 1159–1168. Crowley PJ, Brady LJ, Michalek SM & Bleiweis AS (1999) Virulence of a spaP mutant of Streptococcus mutans in a gnotobiotic rat model. Infect Immun 67: 1201–1206. Demuth DR, Davis CA, Corner AM, Lamont RJ, Leboy PS & Malamud D (1988) Cloning and expression of a Streptococcus sanguis surface antigen that interacts with a human salivary agglutinin. Infect Immun 56: 2484–2490. Drevets DA & Campbell PA (1991) Roles of complement and complement receptor type 3 in phagocytosis of Listeria monocytogenes by inflammatory mouse peritoneal macrophages. Infect Immun 59: 2645–2652. Ey PL, Prowse SJ & Jenkin CR (1979) Complement-fixing IgG1 constitutes a new subclass of mouse IgG. Nature 281: 492–493. Fischetti VA, Pancholi V & Schneewind O (1990) Conservation of a hexapeptide sequence in the anchor region of surface proteins from gram-positive cocci. Mol Microbiol 4: 1603–1605. Getahun A & Heyman B (2006) How antibodies act as natural adjuvants. Immunol Lett 104: 38–45. Guy RA & Belosevic M (1993) Comparison of receptors required for entry of Leishmania major amastigotes into macrophages. Infect Immun 61: 1553–1558. Hamada S & Slade HD (1980) Biology, immunology, and cariogenicity of Streptococcus mutans. Microbiol Rev 44: 331–384. Inal S, Nagaki K, Ebisu S, Kato K & Kotani S (1976) Activation of the alternative complement pathway by water-insoluble glucans of Streptococcus mutans: the relation between their chemical structures and activating potencies. J Immunol 117: 1256–1260. Jenkinson HF & Demuth DR (1997) Structure, function and immunogenicity of streptococcal antigen I/II polypeptides. Mol Microbiol 23: 183–190. Kelly CG, Todryk S, Kendal HL, Munro GH & Lehner T (1995) T-cell, adhesion, and B-cell epitopes of the cell surface Streptococcus mutans protein antigen I/II. Infect Immun 63: 3649–3658. Lanzavecchia A (1990) Receptor-mediated antigen uptake and its effect on antigen presentation to class II-restricted T lymphocytes. Annu Rev Immunol 8: 773–793.

51 (2007) 102–111

111


Ma JK, Smith R & Lehner T (1987) Use of monoclonal antibodies in local passive immunization to prevent colonization of human teeth by Streptococcus mutans. Infect Immun 55: 1274–1278. Manca F, Fenoglio D, Li Pira G, Kunkl A & Celada F (1991) Effect of antigen/antibody ratio on macrophage uptake, processing, and presentation to T cells of antigen complexed with polyclonal antibodies. J Exp Med 173: 37–48. McArthur WP, Rhodin NR, Seifert TB, Oli MW, Robinette RA, Demuth DR & Brady LJ (2007) Characterization of epitopes recognized by anti-Streptococcus mutans P1 monoclonal antibodies. FEMS Immunol Med Microbiol. DOI: 10.1111/ j.1574-695X.2007.00260.x. Michalek SM, Katz J & Childers NK (2001) A vaccine against dental caries: an overview. BioDrugs 15: 501–508. Okahashi N, Sasakawa C, Yoshikawa M, Hamada S & Koga T (1989) Molecular characterization of a surface protein antigen gene from serotype c Streptococcus mutans, implicated in dental caries. Mol Microbiol 3: 673–678. Oli MW, Rhodin N, McArthur WP & Brady LJ (2004) Redirecting the humoral immune response against Streptococcus mutans antigen P1 with monoclonal antibodies. Infect Immun 72: 6951–6960. Oli MW, McArthur WP & Brady LJ (2006) A whole cell BIAcore assay to evaluate P1-mediated adherence of Streptococcus mutans to human salivary agglutinin and inhibition by specific antibodies. J Microbiol Meth 65: 503–511. Perrin-Cocon LA, Villiers CL, Salamero J, Gabert F & Marche PN (2004) B cell receptors and complement receptors target the antigen to distinct intracellular compartments. J Immunol 172: 3564–3572. Prakobphol A, Xu F, Hoang VM et al. (2000) Salivary agglutinin, which binds Streptococcus mutans and Helicobacter pylori, is the lung scavenger receptor cysteine-rich protein gp-340. J Biol Chem 275: 39860–39866. Regnault A, Lankar D, Lacabanne V et al. (1999) Fcgamma receptor-mediated induction of dendritic cell maturation and major histocompatibility complex class I-restricted antigen presentation after immune complex internalization. J Exp Med 189: 371–380. Rhodin NR, Van Tilburg ML, Oli MW, McArthur WP & Brady LJ (2004a) Further characterization of immunomodulation by a monoclonal antibody against Streptococcus mutans antigen P1. Infect Immun 72: 13–21. Rhodin NR, Cutalo JM, Tomer KB, McArthur WP & Brady LJ (2004b) Characterization of the Streptococcus mutans P1 epitope recognized by immunomodulatory monoclonal antibody 6-11A. Infect Immun 72: 4680–4688. Ruiz-Olvera P, Ruiz-Perez F, Sepulveda NV, Santiago-Machuca A, Maldonado-Rodriguez R, Garcia-Elorriaga G & GonzalezBonilla C (2003) Display and release of the Plasmodium falciparum circumsporozoite protein using the

51 (2007) 102–111

autotransporter MisL of Salmonella enterica. Plasmid 50: 12–27. Rundegren J & Arnold RR (1987) Differentiation and interaction of secretory immunoglobulin A and a calcium-dependent parotid agglutinin for several bacterial strains. Infect Immun 55: 288–292. Russell MW & Lehner T (1978) Characterisation of antigens extracted from cells and culture fluids of Streptococcus mutans serotype c. Arch Oral Biol 23: 7–15. Sears DW, Osman N, Tate B, McKenzie IF & Hogarth PM (1990) Molecular cloning and expression of the mouse high affinity Fc receptor for IgG. J Immunol 144: 371–378. Seifert TB, Bleiweis AS & Brady LJ (2004) Contribution of the alanine-rich region of Streptococcus mutans P1 to antigenicity, surface expression, and interaction with the proline-rich repeat domain. Infect Immun 72: 4699–4706. Taborda CP & Casadevall A (2002) CR3 (CD11b/CD18) and CR4 (CD11c/CD18) are involved in complement-independent antibody-mediated phagocytosis of Cryptococcus neoformans. Immunity 16: 791–802. Troffer-Charlier N, Ogier J, Moras D & Cavarelli J (2002) Crystal structure of the V-region of Streptococcus mutans antigen I/II at 2.4 A resolution suggests a sugar preformed binding site. J Mol Biol 318: 179–188. Tucker L, Robinette R, Pinder T, McArthur WP & Brady LJ (2006) Characterization of monoclonal antibodies that inhibit adherence of Streptococcus mutans. 35th Annual Meeting of the American Association for Dental Research, Orlando, FL. Abstract no. 1066. van Dolleweerd CJ, Chargelegue D & Ma JK (2003) Characterization of the conformational epitope of Guy’s 13, a monoclonal antibody that prevents Streptococcus mutans colonization in humans. Infect Immun 71: 754–765. van Dolleweerd CJ, Kelly CG, Chargelegue D & Ma JK (2004) Peptide mapping of a novel discontinuous epitope of the major surface adhesin from Streptococcus mutans. J Biol Chem 279: 22198–22203. Wang F, Nakouzi A, Alvarez M, Zaragoza O, Angeletti RH & Casadevall A (2005) Structural and functional characterization of glycosylation in an immunoglobulin G1 to Cryptococcus neoformans glucuronoxylomannan. Mol Immunol 43: 987–998. Watts C, Antoniou A, Manoury B, Hewitt EW, McKay LM, Grayson L, Fairweather NF, Emsley P, Isaacs N & Simitsek PD (1998) Modulation by epitope-specific antibodies of class II MHC-restricted presentation of the tetanus toxin antigen. Immunol Rev 164: 11–16. Wernersson S, Karlsson MC, Dahlstrom J, Mattsson R, Verbeek JS & Heyman B (1999) IgG-mediated enhancement of antibody responses is low in Fc receptor gamma chain-deficient mice and increased in Fc gamma RII-deficient mice. J Immunol 163: 618–622.


c