Peroxidase-labeled, human IgG, IgM, IgA and the unlabeled F(ab’)₂ and Fc fragments of human IgG were purchased from Jackson ImmunoResearch Labs, Inc. (West Grove, PA). Unlabeled animal IgG and sera were obtained from Sigma Aldrich (St. Louis, MO) and Innovative Research (Novi, MI). Human IgG₁, IgG₂, IgG₃ and IgG₄ were obtained from Athens Research & Technology, Inc. (Athens, GA). Human, plasminogen-free fibrinogen was purchased from Calbiochem (La Jolla, CA) and biotinylated by the method of Bayer et al. [19]. Neutravidin-peroxidase was obtained from Pierce (Rockford, IL). The preparation of rabbit antiserum against a synthetic peptide copying amino acid residues 3-17 of Mrp4 (anti-sMrp4(3-17)) and residues 1-30 of Emm4 (anti-sEmm4(1-30)) was previously described [5].

The construction of mutants utilizing the parental strain M type 4 S. pyogenes (SP4) was previously described [5]. These consisted of MP4, an Mrp-negative mutant; AR4, an Emm-negative mutant; EP4, an Enn-negative mutant; SF4, a SOF-negative mutant; and DS4, an Sof-negative and Sfbx-negative mutant. The mutant SP4ΔA, which expresses Mrp in which the A-repeats were deleted in-frame, was constructed by cutting the desired sequences from the pTrcHis vector that contained an insert of rMrpΔA DNA (see below, cloning of rMrp for details) and ligating the insert into pG+Host9, a temperature-sensitive shuttle vector generously provided by E. Maguin [20]. The vector was then introduced into SP4 via allelic exchange and a mutant expressing Mrp with an in-frame deletion of the A-repeats was selected by previously described methods [5]. The strains were grown overnight at 37°C in Todd-Hewitt broth supplemented with 1% yeast extract (THY) unless indicated otherwise.

DNA encoding the desired sequences of Mrp4 were amplified by PCR, ligated into pTrcHis, introduced into Escherichia coli Top10, expressed as histidine fusion products, and purified by metal affinity chromatography as previously described [5]. The recombinant proteins consisted of rMrp(1-328), rMrp(150-255), rMrp(150-185), rMrp(256-328), rMrp(1-184), rMrp(97-197). The numbers in each case indicate the amino acid residues that are spanned in the mature form of Mrp4. rMrpΔA, in which the DNA encoding the A-repeats was deleted in-frame, was constructed as follows. The region of mrp upstream of the A-repeats and the region of mrp immediately downstream of the A-repeats were amplified by PCR, sequentially inserted and ligated in-frame into the pTrcHis vector, then expressed and purified as above.

The wild type strain SP4 and the indicated mutants were grown overnight in THY, washed in phosphate buffered saline (PBS), and adjusted to an OD₅₃₀ of 0.4 in PBS. Microtiter wells were coated with 100 µl of the streptococcal suspension for 30 min at 37 °C, then washed and blocked with 1% BSA in PBS. The wells were then reacted with various concentrations of peroxidase-labeled, human IgG or biotinylated, human fibrinogen for 30 min at 37 °C. The wells treated with peroxidase-labeled, human IgG were then washed and 100 µl of the substrate tetramethylbenzidine (TMB) was added. For wells treated with biotinylated fibrinogen, the wells were incubated with 1 µg/ml of Neutravidin-peroxidase for 30 min at 37°C, then washed and the TMB substrate added. The absorbance at 650 nm was recorded after color development. Blanks consisted of wells coated with BSA instead of streptococci and then treated as described above. Assays were done in quadruplicate.

The parental strain SP4 and its mutant SP4ΔA were grown overnight in THY, washed in PBS and used to coat microtiter wells as described above. The coated wells were reacted for 30 min at 37°C with rabbit anti-sMrp4(3-17) and anti-sEmm4(1-30) diluted 1:500 in Tris-saline-BSA (0.05 M Tris-HCl, 0.15 M NaCl, 1 mg/ml bovine serum albumin, pH 7.4) containing 10% pig serum to block non-immune binding of immunoglobulins. The wells were then washed and reacted with a 1:2000 dilution of peroxidase-labeled, goat anti-rabbit IgG diluted in Tris-saline-BSA. After 30 min at 37°C the wells were washed, 100 µl of TMB substrate added and the absorbance at 650 nm was recorded after color development. Control wells were reacted under the same conditions with normal rabbit serum to determine the degree of non-specific binding. Assays were done in quadruplicate.

Microtiter wells were coated with 100 µl of 2.5 µg/ml of rMrp(1-328) in sodium bicarbonate (pH 9.5) for 1 hour at 37°C. The wells were washed and blocked with 1% BSA in PBS. Afterwards, 100 µl of 1 µg/ml of peroxidase-conjugated human IgG in a 10% solution of the indicated animal serum or in various concentrations of animal IgG diluted in Tris-saline-BSA were added to the appropriate wells. The positive control consisted of 1 µg/ml of peroxidase-conjugate human IgG in Tris-saline-BSA. Wells coated with BSA and treated as above served as negative controls. The wells were incubated at 37°C for 30 min, washed with Tris-saline, and 100 µl of the TMB substrate added. The absorbance at 650 nm was recorded after color development. Percent inhibition was calculated using the formula: % inhibition = [1- (A₆₅₀ of animal serum (or animal IgG)/A₆₅₀ of positive control)] x 100. Assays were done in quadruplicate.

Microtiter wells were coated with 2.5 µg/ml of rMrp(1-328) and then blocked with 1 mg/ml BSA in PBS. A 100 µl aliquot of a solution of 1 µg/ml peroxidase-conjugated human IgG and various concentrations of human IgG subclasses 1-4 or the F(ab’)₂ or the Fc fragment of human IgG in Tris-saline-BSA + 0.05% Tween-20 was added to the wells and incubated at 37°C for 30 min. The wells were then washed, TMB substrate added and the absorbance at 650 nm recorded after color development. Negative controls consisted of wells coated with BSA and treated as described above. Positive controls consisted of Mrp-coated wells treated with 1 µg/ml peroxidase-conjugated human IgG in Tris-saline-BSA + 0.05% Tween-20. Percentage of inhibition was calculated using the formula: % inhibition = [1- (A₆₅₀ of binding of IgG in solutions containing the F(ab’)₂ (or Fc, or IgG_1-4)/A₆₅₀ of positive control)] x 100. Assays were done in triplicate.

To map the IgG-binding domains of Mrp4, microtiter wells were coated with 10 µg/ml of the indicated recombinant peptides of Mrp for 30 min at 37 °C and then blocked with 1% BSA in PBS for 30 min at 37 °C. Various concentrations of peroxidase-labeled human IgG in Tris-saline-BSA were then added to the wells and incubated for 30 min at 37 °C. After washing, the TMB substrate was added, and the absorbance at 650 nm was recorded after color development. Negative controls consisted of wells coated with BSA and treated as described above. Control assays using antiserum against the polyhistidine tag indicated that all of the histidine-tagged recombinant proteins of Mrp coated the microtiter wells to a similar degree. Assays were done in triplicate.

Microtiter wells were coated with rMrp(1-328), blocked with BSA, and then reacted with peroxidase-labeled, human IgG in various concentrations of rMrp(150-185) (a single A-repeat) or rMrp(150-255) (three A-repeats) for 30 min at 37°C. The wells were then washed, the TMB substrate added and the absorbance at 650 nm recorded after color development. Assays were done in triplicate.

For growth assays in human blood, SP4 and its mutant SP4ΔA, were grown at 37°C in Todd-Hewitt broth supplemented with 1% yeast extract (THY) to an OD₅₃₀ of 0.08 and diluted 1:10,000 in THY. A 50 µl aliquot of this dilution was added to tubes containing 450 µl of heparinized, human blood. The mixtures were rotated for 3 hours at 37 °C and the number of CFU were determined by plating dilutions on blood agar plates. The experiments were designed to provide an inoculum in the range of 100 to 300 CFU and the number of CFU in the inoculum was determined by plating on blood agar. The blood of donors was screened to ensure that their blood did not contain antibodies that could opsonize S. pyogenes and alter their growth in blood.

All blood donors signed a written informed consent and the study and the consent form were approved by the Institutional Review Board of the University of Tennessee Health Science Center.

To evaluate the role of Mrp in the binding of IgG by S. pyogenes, the binding of human IgG to the parental, wild type strain SP4 was compared to that of the Mrp-negative mutant MP4 and various other mutants of SP4 defective in expressing other, major proteins on the surface of S. pyogenes. The wild type strain SP4 bound human IgG in a dose-related fashion (Figure 2). Inactivation of Mrp reduced IgG binding by ~70%, whereas inactivation of Emm, Enn, Sof, or Sfbx had only minor effects on binding. That inactivation of Mrp did not completely reduce IgG binding was not surprising as other streptococcal surface components such as Emm also bind IgG [7]. However, these findings clearly indicate that Mrp is the major IgG-binding protein on SP4.

To determine if Mrp4 selectively binds to human IgG, IgA, or IgM, microtiter wells were coated with rMrp and reacted with peroxidase-labeled human IgG, IgA, or IgM. Mrp4 strongly bound to human IgG but not to human IgM, or IgA (Figure 3A). These findings are consistent with those of others who also found that Mrp selectively bound to human IgG [7,16,18,21].

To examine the potential for host specificity in binding, various animal sera were tested for their ability to block the binding of human IgG to Mrp (Figure 3B). Human, pig and monkey sera were the best inhibitors of binding. Horse, rabbit and donkey sera partially inhibited binding. Goose, guinea pig, rat, chicken, cow, mouse and goat sera had little to no effect on binding of human IgG to Mrp.

It is possible that animal sera may contain other components that could bind to Mrp and sterically hinder interactions with IgG. Therefore, various concentrations of purified IgG from different animals were also tested for their ability to block the binding of human IgG to Mrp (Figure 4). The results essentially reflected those obtained with animal sera. For a better comparison of binding, the concentration required for 50% inhibition was calculated for each purified IgG (Table 1). Human IgG was the most effective inhibitor requiring 6 µg/ml to achieve 50% inhibition followed by pig (20 µg/ml), horse (29 µg/ml) rabbit (40 µg/ml) and monkey IgG (50 µg/ml). Thus, human IgG was > 3-fold better inhibitor than pig IgG, the one closest to human IgG in blocking binding. Mouse, rat, goat, sheep, cow, guinea pig, donkey and chicken IgG were ineffective inhibitors.

Most immunoglobulin-binding proteins of bacteria bind immunoglobulins via the Fc domain. Mrp is no exception. The Fc domain of human IgG blocked the binding of human IgG to rMrp(1-328), whereas the F(ab’)₂ fragment of human IgG had little effect on binding (Figure 5). These findings agree with those of Heath et al. [16] who found that Mrp from an M type 76 strain of S. pyogenes bound human IgG via its Fc domain.

To further examine the binding of Mrp to human IgG, each of the subclasses was tested for its ability to competitively inhibit the binding of peroxidase-labeled human IgG to Mrp (Figure 6). The concentration required to inhibit 50% binding of peroxidase-labeled IgG was 1.2 µg/ml for IgG₁, 15 µg/ml for IgG₂, 30 µg/m for IgG₃, and 7.4 µg/ml for IgG₄. The finding that IgG₃ was the least effective inhibitor was not unexpected because of previous reports that it bound poorly or not at all to Mrp [7,12,22].

To localize the IgG-binding domain of Mrp, various recombinant peptides of Mrp were constructed, purified and tested for binding IgG and fibrinogen by ELISA. A summary of the results is shown in Figure 7 (top panel). As expected, full-length rMrp(1-328) bound both human IgG and human fibrinogen. Both rMrp(150-185), which contains a single A-repeat, and rMrp(150-255) which contains three A-repeats, bound IgG but did not bind fibrinogen. Both rMrp(150-185) and rMrp(150-255) blocked the binding of human IgG to Mrp, but rMrp(150-255) was ~40-fold more effective inhibitor than rMrp(150-185) (Figure 8).

The above results indicate that the A-repeat region of Mrp4 contains the IgG-binding domain and is consistent with the results of Heath et al. [16] who found that human IgG binds to the A-repeats of Mrp from an M type 76 strain of S. pyogenes. To confirm that the A-repeats are the only IgG-binding domain within Mrp, a recombinant protein was constructed in which the DNA encoding the A-repeat region was deleted in-frame and tested for binding human IgG. This construct failed to bind human IgG but it still retained the ability to bind fibrinogen (Figure 7, top panel). Previous work [11] indicated that there are two fibrinogen-binding domains in the N-terminus of Mrp4 as illustrated in the middle panel of Figure 7. Our data indicate that the binding domains for IgG and fibrinogen are separate and distinct. Supporting this concept is the finding that purified, human fibrinogen did not block the binding of IgG to Mrp (data not shown).

The finding that inactivation of Mrp dramatically reduced IgG binding provided clear evidence that Mrp is a major IgG-binding protein in S. pyogenes that express Mrp. However, inactivation of Mrp not only reduces IgG binding by S. pyogenes, but it also reduces the binding of fibrinogen to S. pyogenes, which is involved in resistance to phagocytosis [5]. Therefore, evaluation of the role of IgG binding to Mrp in the growth of S. pyogenes in human blood must be done without altering the binding of fibrinogen to Mrp on the surface of the streptococci. To accomplish this a mutant, SP4ΔA, was engineered that expresses Mrp with an in-frame deletion of the DNA encoding the A-repeats of Mrp. We compared the levels of expression of Emm4 and Mrp4 of this mutant to that of wild type to ensure that the expression of these key proteins was not decreased by the introduction of this mutation. The expression of Mrp in SP4ΔA was virtually identical to that of its wild type parent, SP4 (Figure 9). There was a slight increase in reactivity of anti-sEmm4(1-30) serum with SP4ΔA as compared to that of SP4. The reason for this slight increase is not clear but it has been suggested that alterations of one surface protein may provide enhanced access of antibodies to another surface protein [5]. However, it is clear that there was no significant reduction in the expression of Mrp4 or Emm4 on the surface of the streptococci due to this mutation.

Next, it was critical to determine if mutant SP4ΔA that did not bind IgG would still bind fibrinogen. The binding of fibrinogen by SP4 and SP4ΔA was virtually identical, but there was a dramatic reduction in the binding of human IgG by SP4ΔA when compared to SP4 (Figure 10). These data clearly indicate that the A-repeat region of Mrp is the major IgG-binding receptor on the surface of SP4. Furthermore, the finding that fibrinogen binding of SP4ΔA was similar to that of SP4 provided additional evidence that the expression levels of Mrp was not altered by the in-frame deletion of the A-repeats, because Mrp is the major fibrinogen-binding protein in SP4 [5].

To evaluate the effect of this mutation on resistance to phagocytosis, the growth of SP4 and SP4ΔA in human blood was compared (Table 2). The multiplication factor of SP4 was 133.7 ± 16, whereas the multiplication factor of SP4ΔA was 27.1 ± -.7. This indicates that the loss of the A-repeats within Mrp reduced growth of S. pyogenes in human blood by ~80%.