A monoclonal antibody that recognizes the E domain of staphylococcal protein A

Abstract

Staphylococcal protein A (SpA) binds Fcγ and V_H3 clan Fab domains of human and animal immunoglobulin (Ig) with each of its five Ig binding domains (IgBDs), thereby supporting Staphylococcus aureus escape from opsonophagocytic killing and suppressing adaptive B cell responses. The variant SpA_KKAA cannot bind Ig yet retains antigenic properties that elicit SpA-neutralizing antibodies and disease protection in mice, whereas S. aureus infection or SpA-immunization cannot elicit neutralizing antibodies. As a test for this model, we analyzed here mAb 358A76, which was isolated from SpA-immunized mice. Unlike SpA_KKAA-derived mAbs, mAb 358A76 binds only the first IgBD (E) but not any of the other four IgBDs (D-A-B-C) and its binding can neutralize only the E domain of SpA, which does not provide disease protection in mice. These results are in agreement with a model whereby wild-type SpA-immunization generates a limited antibody response without neutralizing and/or disease protective attributes.

1. Introduction

Almost all clinical Staphylococcus aureus isolates express protein A (SpA) [1, 2], SpA is synthesized as a precursor that is cleaved and anchored by sortase A in the bacterial cell wall envelope [3, 4]. During staphylococcal growth, SpA molecules are also released into the extracellular medium [5]. SpA is comprised of four to five repeats of the 56-61 residue immunoglobulin binding domains (IgBDs) [6, 7]. Each IgBD folds into a triple-helical bundle with discrete binding sites for the Fcγ and V_H3 clan Fab domains of human and animal immunoglobulin (Ig) [8, 9]. Glutamine (Q) residues 9 and 10 in α-helix 1 of each IgBD capture Fcγ domains [9], thereby inhibiting antibody-mediated opsonization [10, 11]. Aspartic acid (D) residues 36 and 37 in the linker region between α-helix 2 and 3 of each IgBD capture the Fab portion of V_H3-type IgG and IgM [8]. SpA association with V_H3-type B cell receptors triggers B cell superantigen activity [11,12]. These interactions explain why S. aureus spa mutants, in comparison with wild-type, are more susceptible to opsonophagocytic killing and unable to block adaptive immune responses in a mouse model of infection [13].

SpA_KKAA, a variant with 20 amino acid substitutions replacing Q⁹ and Q¹⁰ with lysine (K) as well as D³⁶ and D³⁷ with alanine (A) in each of the five IgBDs, cannot bind Fcγ or V_H3 Fab domains [14]. When used as vaccine antigen in mice, SpA_KKAA, but not SpA, elicited protein A-specific immune responses and protection from disease [14]. SpA_KKAA was also used to study protein A-specific immune responses in humans and mice. Healthy volunteers with a history of staphylococcal infection did not harbor serum SpA_KKAA-specific antibodies [14]. Mice that had been injected with recombinant wild-type SpA or infected with wild-type S. aureus strains (expressing SpA) did not mount significant SpA_KKAA-specific antibody responses [15]. These data were incorporated into a model, whereby S. aureus infection does not elicit protective immune responses because of the B cell superantigen activity of protein A [13].

Mouse monoclonal antibodies (mAb) were isolated after immunization of animals with SpA_KKAA [16]. Protection against S. aureus disease by candidate SpA_KKAA-mAbs correlated with the ability to bind multiple IgBDs and inhibit protein A binding to Fcγ and Fab V_H3 domains of Ig [16]. United States patents US2008/0118937 A1 and US2010/0047252 A1 describe a murine hybridoma cell line derived from mice that had been immunized with wild-type SpA. The corresponding antibody, mAb 358A76, was reported to bind SpA, however the nature of this binding was heretofore not known. Here, we analyze the molecular attributes of SpA-mAb 358A76 in comparison with SpA_KKAA-mAb 3F6 [16].

2. Materials and methods

2.1. Bacterial strains and growth conditions

S. aureus strain USA300 LAC [17] was grown in tryptic soy broth (TSB) at 37°C. Escherichia coli strains DH5α and BL21 (DE3) were grown in Luria-Bertani (LB) broth with 100 μg·ml⁻¹ ampicillin at 37°C.

2.2. Monoclonal antibodies

The generation and characterization of the mouse monoclonal antibody 3F6 have been described previously [16]. Hybridoma cell line 358A76 was purchased from American Type Culture Collection (ATCC accession number PTA-7938) and expanded at the Fitch Monoclonal Antibody Facility, University of Chicago. Antibodies were affinity purified from the spent culture supernatant of cell lines as described previously [16].

2.3. Purification of recombinant proteins

Recombinant SpA variants tagged with N-terminal 6 histidine residues (SpA, SpA-E, SpA_KKAA, SpA-E_KKAA, SpA-D_KKAA, SpA-A_KKAA, SpA-B_KKAA, and SpA-C_KKAA) and synthetic peptide fragments (H1, H2, H3, H1+2, and H2+3), were produced from plasmids or purchased, respectively, as described previously [16]. Briefly, purification of recombinant SpA variants was performed by diluting overnight cultures of E. coli BL21 carrying specific plasmids, 1:100 into fresh media and grown at 37°C. When cultures reached an absorbance at 600 nm (A₆₀₀) of 0.5, isopropyl β-D-l-thiogalatopyranoside (IPTG) was added at a concentration of 1 mM and cultures were incubated at 37°C for an additional three hours. Bacterial cells were sedimented, suspended in column buffer [50 mM Tris-HCI (pH 7.5), 150 mM NaCI] and disrupted with a French pressure cell at 14,000 psi. Lysates were subjected to ultracentrifugation at 100,000 ×g for 30 min and proteins in the supernatant were subjected to nickel-nitrilotriacetic acid (Ni-NTA) affinity chromatography. Proteins were eluted in column buffer containing increasing concentrations of imidazole (up to 500 mM). Protein concentrations were determined by bicinchoninic acid (BCA) assay (Thermo Scientific).

2.4. Enzyme linked immunosorbent assay

To determine binding affinities of candidate mAbs, ELISA plates were coated with either full length SpA_KKAA (20 nM), individual IgBDs (100 nM), or synthetic peptides (H1, H2, H3, H1+3 and H2+3,100 nM) derived from IgBD domain E of SpA_KKAA (SpA-E_KKAA) in 0.1 M carbonate buffer (pH 9.5 at 4°C) overnight. Plates were blocked and incubated with variable concentrations of mAbs 358A76 and 3F6. For competition assay, plates were coated with 20 nM of SpA_KKAA, or wild-type SpA, or 100 nM SpA-E in 0.1 M carbonate buffer (pH 9.5 at 4°C) overnight. The following day, plates were blocked and incubated with dilutions of competing monoclonal antibodies at a starting concentration of 30 or 6 μg·ml⁻¹ prior to the incubation with HRP-conjugated SpA-specific mAbs (Innova Biosciences) or human IgG at a final concentration of 200 ng·ml⁻¹.

2.5. Mouse renal abscess model

Affinity purified antibodies in PBS were injected at 5 mg·kg⁻¹ into the peritoneal cavity of BALB/c mice (6 week old, female, Charles River Laboratories) 24 hours prior to challenge with S. aureus. Overnight cultures of S. aureus were diluted 1:100 into fresh TSB and grown for 2 hours at 37°C. Staphylococci were sedimented, washed and suspended in PBS to the desired bacterial concentration. Inocula were quantified by plating dilution aliquots on TSA and enumeration of colonies after incubation of plates at 37°C overnight. Animals were anesthetized via intraperitoneal injection with 65 mg·ml⁻¹ ketamine and 6 mg·ml⁻¹ xylazine per kilogram of body weight and infected by injection with 5×l0⁶ CFU S. aureus USA300 (LAC) into the periorbital venous sinus of the right eye. On day 4 following challenge, mice were killed by CO₂ inhalation. Kidneys were removed, and staphylococcal loads were analyzed by homogenizing renal tissue in PBS containing 0.1% Triton X-100. Serial dilutions of homogenates were plated on TSA and incubated for colony formation and enumeration.

2.6. Staphylococcal survival in blood

Whole blood was collected from BALB/c mice by cardiac puncture and coagulation was inhibited by adding 10 μg·ml⁻¹ lepirudin. Fifty μl of a suspension containing 5×10⁵ CFU S. aureus USA300 were mixed with 950 μl of mouse blood in the presence of 10 μg·ml⁻¹ of mAbs. Samples were incubated 30 min at 37°C with slow rotation, placed on ice with 1% saponin/PBS to lyse blood cells and serially diluted prior to plating on TSA and enumeration of colonies.

2.7. Ethics Statement

All mouse experiments were performed at least twice and conducted in accordance with institutional guidelines following experimental protocol review and approval by the Institutional Biosafety Committee (IBC) and the Institutional Animal Care and Use Committee (IACUC) at the University of Chicago.

2.8. Statistical analysis

Colony counts from infected animals were analyzed with the two-tailed Mann-Whitney test. Unpaired two-tailed Student’s t-tests were used to assess the statistical significance of ELISA data and ex vivo blood survival data. All data were analyzed with Prism™ (GraphPad Software, Inc.) and P values less than 0.05 were deemed significant.

3. Results

3.1. Monoclonal antibody mAb 358A76 recognizes the E domain ofSpA

To determine the affinity constant (K_a = [mAb·Ag]/[mAb]×[Ag]) and protein A binding site for SpA-mAb 358A76, microtiter plates were coated with SpA_KKAA, a protein A variant that cannot capture IgG by binding Fcγ or V_H3 Fab [13]. As control, the SpA_KKAA-derived mouse monoclonal antibody mAb 3F6 [16] bound with high affinity to SpA_KKAA [K_a 22.81(± 2.84)×10⁹ M⁻¹] (Fig. 1AB). We determined an affinity constant K_a 1.95(± 0.55)×10⁹ M⁻¹ for SpA-mAb 358A76, approximately 10 fold less than SpA_KKAA-mAb 3F6 (P=0.0004; Fig. 1AB). To identify the binding site of SpA-mAb 358A76, the five individual IgBDs - E_KKAA, D_KKAA, A_KKAA, B_KKAA and C_KKAA - were purified [16] and used to coat ELISA plates. Further, we used synthetic peptides representing one (H1, H2, and H3) or two adjacent α-helices (H1+2 and H2+3) of the IgBD-E_KKAA triple helical bundle as antigens in the ELISA assay. SpA_KKAA-mAb 3F6 bound each of the five IgBDs with similar affinity (K_a 12.41-27.46×10⁹ M⁻¹). In contrast, SpA-mAb 358A76 bound only to E_KKAA (K_a 0.21×10⁹ M⁻¹) but not to any of the other four IgBDs (D_KKAA, A_KKAA, B_KKAA and C_KKAA) (Table 1). Further, SpA-mAb 358A76 did not recognize synthetic peptides encompassing one or two ex-helices of IgBD-E_KKAA (H1, H2, H3, H1+2, and H2+3). By comparison, SpA_KKAA-mAb 3F6 bound to the helix 1+2 peptide, but not to single α-helix peptides (H1, H2, H3) or to the H2+3 peptide (Table 1). As reported previously [7], alignment of the amino acid sequences of the five IgBDs of protein A shows that the E domain is most dissimilar to the remaining four IgBDs with both conservative and non-conservative amino acid substitutions in α-helix 1 and 3 of the triplehelical fold [6, 9] (Fig. 1CD). The unique amino acid sequences of α-helix 1 and 3 of the E domain may explain why SpA-mAb 358A76 selectively binds only the first IgBD. Thus, immunization of mice with wild-type SpA led to the generation of an lgG_2a antibody, which recognizes the E-domain of protein A with K_a of 1×10⁹ M⁻¹.

Fig. 1.

SpA-mAb 358A76 specifically recognizes the E domain of protein A. ELISA examining the association of (A) mAbs 358A76 and (B) 3F6 with the immobilized non-toxigenic SpA variant (SpA_KKAA), each immunoglobulin-binding domain (lgBD-E_KKAA, -D_KKAA, -A_KKAA, -B_KKAA, and -C_KKAA), and the synthetic linear peptides derived from the three helices (H1, H2, H3, H1+2, H2+3) of IgBD-E_KKAA. (C) Alignment of amino acid sequences of the five IgBDs of SpA. Amino acid residues identical to that of IgBD-E (top sequence) are depicted with a dot whereas those unique to the E domain are highlighted in yellow boxes. The four amino acid residues substituted in each IgBD of the non-toxigenic SpA_KKAA variant are highlighted in the magenta boxes. (D) Amino acid sequence homology level was compared using ClustalW and the numbers represent the percent of amino acid homology between IgBDs.

TABLE 1.

Association constants for the binding of mAbs 358A76 and 3F6 to SpA_KKAA and its fragments

amAb		Association constant (×109 M−1) for antigen or antigen fragment
		IgG binding domains of protein A					Segments of the EKKAA triple-helical bundle
	SpAKKAA	EKKAA	DKKAA	AKKAA	BKKAA	CKKAA	H1	H2	H3	H1+2	H2+3
IgG2a
358A76	1.00	0.21	<	<	<	<	<	<	<	<	<
c3F6	22.97	17.69	12.41	20.15	27.46	26.46	<	0.01	<	0.41	0.01

^aAffinity purified antibodies (100 μg·ml⁻¹) were serially diluted across ELISA plates coated with antigens (20nM for SpA_KKAA and 100 nM for IgG binding domains) to calculate the association constant using Prism^® (GraphPad Software, Inc.). To study the binding of antibodies to protein A antigen, we used the SpA_KKAA variant (residues 1-291 of mature SpA harboring six N-terminal histidyl residues) with four amino acid substitutions in each of the five immunoglobulin binding domains (IgBD) of protein A [E (residues 1-56), D (residues 57-117), A (residues 118-175), B (residues 176-233) and C (residues 234-291)]. In each IgBD, the glutamines at position 9 and 10 (amino acid residues from IgBD-E) were replaced with lysine (Q⁹K, Q¹⁰K) and aspartic acids 36 and 37 were substituted with alanine (D³⁶A, D³⁷A). The same substitutions were introduced into proteins spanning individual IgBDs: E_KKAA, D_KKAA, A_KKAA, B_KKAA and C_KKAA (all expressed and purified with an N-terminal six histidyl tag). SpA_KKAA and individual IgBDs were purified by affinity chromatography from E. coli extracts. Peptides H1, H2, H3, H1+2, and H2+3 were synthesized on a peptide synthesizer and purified via HPLC. The peptides encompass helices 1 (H1: NH₂-AQHDEAKKNAFYQVLNMPNLNA-COOH), 2 (H2: NH₂NMPNLNADQRNGFIQSLKAAPSQ-COOH), 3 (H3: NH₃ AAPSQSANVLGEAQKLNDSQAPK-COOH), 1+2 (H1+2: NH₂-AQHDEAKKNAFYQVLNMPNLNADQRNGFIQSLKAAPSQ-COOH) or 2+3 (H2+3: NH₂-NMPNLNADQRNGFIQSLKAAPSQSANVLGEAQKLNDSQAPK-COOH) of the triple helical bundle of the E_KKAA IgBD (residues 1-56 of SpA_KKAA: NH₂-AQHDEAKKNAFYQVLNMPNLNADQRNGFIQSLKAAPSQSANVLGEAQKLNDSQAPK-COOH).

^bThe symbol < signifies measurements that were too low to permit the determination of the association constant.

^cAssociation constant was previously measured [16].

3.2. SpA_KKAA-mAB 3F6, but not SpA-mAb 358A76, prevents IgG binding to protein A

Do the monoclonal antibodies studied here bind protein A with sufficiently high affinity to block its binding to human immunoglobulin, which associates with SpA via its Fcγ and V_H3 Fab domains with affinities ranging from 0.1-1 ×10⁸ M⁻¹ [18, 19]? To address this question, we developed a competitive ELISA assay with either SpA_KKAA (control) (Fig. 2A) or SpA (Fig. 2B) or the SpA-E domain alone (Fig. 2C) and measured binding of HRP-conjugated antibodies. As a control, IgG_2a antibodies did not interfere with the binding of HRP-conjugated SpA-mAb 358A76 or SpA_KKAA-mAb 3F6 to SpA_KKAA (FIG. 2A). Increasing amounts of SpA-mAb 358A76 prevented the binding of HRP-mAb 358A76 to SpA_KKAA (Fig. 2A). In contrast, SpA-mAb 358A76 did not interfere with HRP-mAb 3F6 binding to SpA_KKAA (Fig. 2A). Further, SpA_KKAA-mAb 3F6 reduced the association between HRP-mAb 358A76 and SpA_KKAA more efficiently than SpA-mAb 358A76 (95% compared to 88.0% inhibition, P=0.0007) (Fig. 2A). These data suggest that SpA-mAb 358A76 binds a conformational epitope, presumably involving non-conserved amino acids in ex-helices 1 and 3 of the E domain (FIG. 1CD). This binding can be blocked by the higher affinity interaction of SpA_KKAA-mAb 3F6 with E_KKAA. The competitive ELISA assay was also used to address whether SpA-mAb 358A76 and SpA_KKAA-mAb 3F6 perturb the interaction between HRP-labeled human IgG and purified full-length protein A or its isolated E domain (Fig. 2BC). When added at 6 μg·ml⁻¹ and 30 μg·ml⁻¹, SpA_KKAA-mAb 3F6 reduced the binding of HRP-labeled human IgG to full length protein A as well as the E domain, whereas the IgG_2a control mAb did not (Fig. 2BC). At a concentration of 6 μg·ml⁻¹ or 30 μg·ml⁻¹, SpA-mAb 358A76 blocked the association of human IgG with the E domain of SpA, but failed to interfere with the binding of human IgG to full-length SpA (Fig. 2BC).

Fig. 2.

SpA-mAb 358A76 cannot block IgG binding to protein A. (A) ELISA assay to measure competitive inhibition of antibody binding. Horse radish peroxidase (HRP)-conjugated mAbs (358A76-HRP and 3F6-HRP) were added to ELISA plates coated with SpA_KKAA and where first incubated with either isotype control antibody (IgG_2a) or mAbs (358A76 and 3F6). Values at OD_405nm were recorded and normalized for the interaction of SpA_KKAA and HRP-conjugated SpA specific mAbs. Isotype control antibodies, SpA-mAb 358A76 or SpA_KKAA-mAb 3F6 were used to perturb the binding of human IgG to wild-type full-length SpA (B) or the E domain of SpA (C) immobilized on ELISA plates. The values were normalized to SpA interaction with human IgG in the absence of monoclonal antibodies. The asterisk denotes a P value less than 0.05.

3.3. SpA_KKAA-mAb 3F6, but not SpA-mAb 358A76, protects mice against S. aureus disease

Cohorts of BALB/c mice were injected into the intraperitoneal cavity with 5 mg·kg⁻¹ SpA-mAb 358A76, SpA_KKAA-mAb 3F6 or isotype control. Passively immunized animals were challenged by intravenous injection with S. aureus USA300 (LAC), a highly virulent MRSA strain and causative agent of the community-associated epidemic of S. aureus infections in the United States [17,20]. Compared to IgG_2a control, animals that received SpA_KKAA-mAb 3F6 harbored reduced bacterial loads in renal tissues (1.26 log₁₀CFU·g⁻¹ reduction; P=0.0021, Fig. 3A). Animals that received SpA-mAb 358A76 displayed a small reduction in bacterial load, as compared to the control group. However, this difference was not statistically significant (0.42 log₁₀CFU·g⁻¹ reduction; P=0.0948, Fig. 3A). When compared to SpA-mAb 358A76, passive transfer of SpA_KKAA-mAb 3F6 provided mice also with increased protection against CA-MRSA strain USA300, as evidenced by the reduced bacterial load (0.84 log₁₀CFU·g⁻¹ reduction; P=0.0011, Fig. 3A).

Fig. 3.

SpA-mAb 358A76 does not protect mice against S. aureus disease. (A) Cohorts of animals (n=10) were immunized by intraperitoneal injection with either mock (IgG_2a isotype control mAb), mAb 358A76 or mAb 3F6 at 5 mg· kg⁻¹. Twenty-four hours later, animals were challenged by intravenous inoculation with 5×10⁶ colony forming units (CFU) of S. aureus USA300. Four days post-challenge, animals were euthanized to enumerate staphylococcal loads in kidneys (CFU/ml of ground tissue kidney in suspension). (B) Lepirudin-treated mouse blood was incubated with 5×10⁵ CFU S. aureus USA300 (LAC) in the presence of PBS, IgG_2a isotype control mAb, mAb 358A76 or mAb 3F6 (10 μg·ml⁻¹) for 30 minutes; staphylococcal survival was measured by plating sample aliquots and enumeration of CFUs. The values were normalized to the CFU number obtained from at least three independent experiments.

During infection, association of IgG with surface exposed SpA blocks complement activation, engagement of C1q and opsonophagocytosis of staphylococci by phagocytes [10, 11,21]. Opsonophagocytic killing of invasive microbes is a key defense strategy that has been used as a correlate of protective immunity for many bacterial vaccines [22, 23]. We employed Rebecca Lancefield’s assay of bacterial killing in blood [24] to compare the activity of SpA-mAb 358A76 and SpA_KKAA-mAb 3F6. S. aureus USA300 cells were incubated in lepirudin-treated blood of naïve mice, in the presence or absence of 10 μg·ml⁻¹ mAb 358A76, mAb 3F6 or IgG_2a mAb control. Sample aliquots were lysed, plated on agar medium and staphylococcal load enumerated. In contrast to SpA_KKAA-mAb 3F6, which reduced the staphylococcal load in blood by 49%, SpA-mAb 358A76 and IgG_2a-control failed to promote opsonophagocytic killing of staphylococci (mAb 3F6vs. PBS, P<0.0001; mAb 3F6vs. 358A76, P=0.0007; Fig. 3B).

4. Discussion

Protein A has evolved to interfere with B cell adaptive immune responses that promote opsonophagocytic clearance of S. aureus in host tissues and/or the development of antibodies providing protective immunity against recurrent infection [13]. When inoculated into the bloodstream of naïve mice, S. aureus spa variants that cannot bind Ig are cleared more rapidly than wild-type staphylococci [13]. SpA-mediated resistance against opsonophagocytic killing requires Ig, as μMT mice lacking mature B cells and antibodies [25] do not display this phenotype [13]. Protein A is comprised of either 4 or 5 IgBDs with the sequence arrangement E-D-A-B (4 domains) or E-D-A-B-C (5 domains)[6, 7]. Comparing SpA IgBDs, Sjöquist and colleagues noted that the D-A-B-C domains are highly homologous, whereas the E domain differs in sequence [7]. When tested in isolation, each of the five IgBDs can bind Fcγ and V_H3-clan Fab [7], although more recent work suggests that the E domain may bind Ig with less affinity than the other four IgBDs [26]. Using protein A to assess antibody association in serum, Hjelm and co-workers observed complexes comprised of 2 mol IgG:1 mol SpA, i.e. SpA endowed with all five IgBDs [27]. To the best of our knowledge, the stoichiometry of SpA association with serum IgG or IgM has not yet been investigated in depth. It is also not known which of the 4 or 5 IgBDs are engaged in binding serum IgG or IgM and what may be the contribution of the E domain for these mechanisms. Although it is known that IgG binding to protein A on the staphylococcal surface interferes with neutrophil opsonophagocytosis of bacteria [10,11], the molecular basis of this interference is not appreciated. Protein A binding at Fcγ [9, 28], i.e. the C_H2-C_H3 region of IgG, interferes with antibody binding to complement component C1q of the classical pathway [21]. Nevertheless, protein A binding to IgG Fcγ certainly does not universally block association with Fcγ receptors, e.g. FcγRI and FcγRII, on the surface of immune cells [29]. Thus, the molecular mechanism of SpA-mediated resistance against opsonophagocytic clearance is not yet known.

Protein A also triggers proliferation of discrete lymphocyte populations [30], specifically V_H3-clan B cells, by crosslinking their IgM receptors [31, 32]. Clonal expansion of B cell populations is followed by their apoptotic collapse, and this B cell superantigen activity of protein A is associated with preventing antibody responses against S. aureus secreted antigens including protein A [12, 14]. In agreement with this model, animals or humans with a history of s aureus infection do not elaborate antibodies that recognize protein A as an antigen [14]. SpA_KKAA, the non-toxigenic (non-lg binding) variant of protein A, elicits antibodies that neutralize SpAand provide protection against S. aureus disease [14]. By comparison, immunization with SpA elicits only low-titer antibodies that are not associated with significant disease protection [14].

To test the model that immunization with wild-type SpA triggers B cell superantigen activity, we examined monoclonal antibody SpA-27, which had been raised by immunizing mice with whole cell preparation S. aureus Cowan I (Sigma, St. Louis, MO). SpA-mAb Spa27 was typed as mouse IgG₁, an antibody class whose Fcγ domain does not bind protein A [33]. Nevertheless, mouse IgG₁ antibodies can bind wild-type SpA via their V_H3-type Fab domains [32, 34]. Using the variants SpA_KK, which associates only with V_H3-type Fab domains, and SpA_AA, which binds only the Fcγ domain, we observed that SpA-mAb Spa27 binds wild-type SpA and SpA_KK, but not SpA_AA or SpA_KKAA [16]. Thus, SpA-mAb Spa27 represents a V_H3-clan IgG₁ antibody that does not specifically recognize protein A antigen and cannot protect mice against S. aureus disease [16]. As a second test for the model of SpA B cell superantigen activity, we analyzed SpA-mAb 358A76. This antibody recognizes the E domain, but not the other four IgBDs of SpA. Based on amino acid similarities between the five IgBDs and the epitope mapping data in Table 1, SpA-mAb 358A76 may specifically recognize a conformational epitope involving helix 1 and 3 of the E domain. Interestingly, SpA-mAb 358A76 can neutralize human IgG binding to the E domain, but fails to block association of the other four IgBD (D-A-B-C) with human IgG. SpA-mAb 358A76 did not provide immune protection against S. aureus disease, indicating that neutralization of the single IgBD E domain is not sufficient to provide protective immunity. These data suggest that only non-toxigenic variants of protein A, for example SpA_KKAA, are able to raise broad spectrum SpA-neutralizing antibodies that provide protection against S. aureus disease and enable antibody responses against other staphylococcal antigens [14].

Highlights:

• Protein A is a B cell superantigen of Staphylococcus aureus

• Staphylococcus aureus infection does not elicit protein A-neutralizing antibodies

• S. aureus infection generates an antibody that binds the E domain of protein A

• Antibody binding to the E domain does not neutralize protein A

• Antibody binding to the E domain does not provide disease protection