Native Human Antibody to Shr Promotes Mice Survival After Intraperitoneal Challenge With Invasive Group A Streptococcus

Abstract

Background

A vaccine against group A Streptococcus (GAS) has been actively pursued for decades. The surface receptor Shr is vital in GAS heme uptake and provides an effective target for active and passive immunization. Here, we isolated human monoclonal antibodies (mAbs) against Shr and evaluated their efficacy and mechanism.

Methods

We used a single B-lymphocyte screen to discover the mAbs TRL186 and TRL96. Interactions of the mAbs with whole cells, proteins, and peptides were investigated. Growth assays and cultured phagocytes were used to study the mAbs’ impact on heme uptake and bacterial killing. Efficacy was tested in prophylactic and therapeutic vaccination using intraperitoneal mAb administration and GAS challenge

Results

Both TRL186 and TRL96 interact with whole GAS cells, recognizing the NTR and NEAT1 domains of Shr, respectively. Both mAbs promoted killing by phagocytes in vitro, but prophylactic administration of only TRL186 increased mice survival. TRL186 improved survival also in a therapeutic mode. TRL186 but not TRL96 also impeded Shr binding to hemoglobin and GAS growth on hemoglobin iron.

Conclusions

Interference with iron acquisition is central for TRL186 efficacy against GAS. This study supports the concept of antibody-based immunotherapy targeting the heme uptake proteins to combat streptococcal infections.

We isolated native human monoclonal antibody (TRL186) against Shr and evaluated its efficacy and mechanism. TRL186 bestowed protection in both prophylactic and therapeutic modalities against aggressive group A Streptococcus (GAS) infection in mice and impeded hemoglobin binding and GAS growth on hemoglobin iron.

Group A Streptococcus (GAS) is the ninth leading infectious source of human morbidity and mortality, with a global burden estimated to exceed 500 000 deaths annually [1, 2]. GAS commonly colonizes the mucosal surfaces and skin, frequently causing pharyngitis and impetigo. These infections can lead to severe immune sequelae, such as acute rheumatic fever and glomerulonephritis [1]. Timely treatment with antibiotics can mitigate GAS infections and their complications, but resistance to penicillin alternatives are on the rise [1, 3, 4]. The frequency of GAS diseases has increased in the past 2 decades, reaching 7–10 cases per 100 000 in the United States and Canada [5, 6]. A large number of circulating serotypes pose a significant challenge for vaccine development, with none approved to date [7, 8]. Without a vaccine, the burden of GAS sequelae and invasive diseases is extreme, and the need for improved means to prevent and manage infections is high.

GAS is an iron-requiring bacterium that mostly relies on heme iron to satisfy its need for the metal [9]. Proteins involved in heme capture and import are critical for GAS survival in the host. The sia operon encodes the key heme acquisition pathway, including 2 surface receptors and an ABC transporter, which capture heme from the host (shr) shuttle it across the cell wall (shp) and through the cytoplasmic membrane (siaABC) [10–14]. Shr, the first receptor in the sia heme relay, binds to hemoglobin and other host hemoproteins [10]. The 145 kDa surface protein has a unique N-terminal region (NTR) followed by 2 near–iron transport (NEAT) domains [12]. Shr binds to hemoglobin using a novel mechanism through a domain (DUF1533) that appears twice in its NTR; this new hemoglobin binding module was named HID for hemoglobin-interacting domain [12, 15]. Following binding to hemoglobin, NEAT1 captures and transfers the heme to either NEAT2 or Shp [13, 16]. Inactivation of shr impairs GAS growth on hemoglobin as an iron source [12] or in human blood [17]. Shr also binds fibronectin and laminin in vitro [18], and deletion mutants show reduced binding to fibronectin or laminin in a strain-dependent manner [17–19]. Shr knockout mutants are attenuated in both zebrafish [18], and mouse models for invasive GAS infections [17].

The essential role Shr plays in GAS pathophysiology raised the possibility of targeting this protein for the development of antibacterial strategies. Shr is highly immunogenic, and immunizing mice intraperitoneally with the purified protein or intranasally with Shr-expressing Lactococcus lactis protects from an invasive GAS infection [20]. Moreover, rabbit Shr-antiserum administrated prophylactically also defends against GAS in a mouse model for passive immunity [20]. In this study, we used a B-lymphocyte screen to identify 2 native human monoclonal antibodies (TRL96 and TRL186) to Shr. We show that TRL186, but not TRL96, aids mice survival after intraperitoneal challenge with an invasive GAS strain in both prophylactic and therapeutic mouse models.

MATERIALS AND METHODS

Strains and Growth Conditions

The strains and plasmids are listed in Supplementary Table 1. Bacteria were grown at 37°C aerobically in Luria–Bertani broth (Escherichia coli) or statically in Todd–Hewitt broth with 0.2% w/v yeast extract (GAS). When necessary, ampicillin (100 µg/mL) or kanamycin (70 µg/mL) was used.

Single B-Lymphocyte Monoclonal Antibody Discovery Technology

Blood samples were collected from anonymized donors under informed consent approved by the Institutional Review Board of Stanford University (Stanford Blood Center, Stanford, California), and peripheral mononuclear cells were prepared as described. Using the CellSpot platform (Trellis Bioscience, Redwood City, California) [21], memory B cells were stimulated to proliferate and differentiate into plasma cells. The secreted immunoglobulin G (IgG) footprint (100 fg/cell over a 5-hour period) of individual cells was probed with fluorescent nanoparticles of distinguishable types conjugated with full-length recombinant Shr protein, the NEAT1 and NEAT2 domains, and bovine serum albumin (BSA; counter-screening bead). More than 30 monoclonal antibodies (mAbs) have been cloned based on their binding profiles. The V regions from messenger RNA of single B cells of interest were cloned onto an IgG1 constant region, which was then expressed by transient transfection in HEK293 cells.

Enzyme-Linked Immunoabsorbent Assays With Shr Proteins

Full-length Shr and Shr fragments (NTR, NEAT1, or NEAT2) from GAS were expressed and purified as described previously [12]. Antibody binding was evaluated by enzyme-immunosorbent assay (ELISA) using 96-well Enzyme Immuno Assay/Radioimmunoassay microplates. Microplates were incubated overnight at 4°C with 25 µg/mL bait protein in phosphate-buffered saline (PBS) (10 mM PBS, 100 nM sodium chloride, pH 7.4). Wells were coated with BSA, and uncoated wells served as controls. Plates were washed with PBS and 0.05% Tween (PBST), blocked with 5% soy milk (in PBST) for 1 hour at 37°C and washed again. TRL186 (2 µg/mL in PBST) was allowed to react with the bait for 1 hour at 37°C. Antibody binding was detected using alkaline phosphatase (AP)–conjugated anti-human IgG (Sigma). For affinity testing, TRL186 (in PBST) was added to a plate coated with Shr. Binding was detected using horseradish peroxidase–conjugated anti-human IgG.

ELISA With Synthetic Peptides

A library of overlapping peptides (15-mers with an offset of 3 amino acids) encompassing Shr NTR (Mimotopes, Melbourne, Australia) were screened for interactions with TRL186 by ELISA similar to that done with Shr proteins. Biotinylated peptides dissolved in 80% dimethyl sulfoxide were diluted 1/200 in PBST and immobilized onto Pierce streptavidin-coated high capacity plates. TRL186 (2 µg/mL in PBST) was allowed to interact with the coated wells for 30 minutes at room temperature, and binding was detected using AP-conjugated anti-human IgG.

ELISA With Immobilized Bacteria

Overnight grown cultures (5 mL of Optical density of 600 nm [OD₆₀₀] 0.9) of GAS were harvested, washed with PBS (5 mL), and used to coat microplates (50 µL) overnight at 4°C. After blocking, TRL186 and TRL96 were allowed to interact with the coated wells for 1 hour at 37°C and binding was detected as above. We used a human monoclonal antibody (TRL308) against respiratory syncytial virus as a control [22].

ELISA With Shr and Hemoglobin

Human hemoglobin was biotinylated using the EZ-Link Sulfo-NHS-SS-Biotinylation Kit. Microplates were coated overnight at 4°C with purified Shr-NTR or BSA (control). Plates were washed and blocked as above, and biotinylated hemoglobin (in saline) was allowed to react with the coated wells during overnight incubation at 4°C. Hemoglobin binding was detected with streptavidin-conjugated AP. To study the impact of TRL186 on NTR’s binding to hemoglobin, mAb in increasing concentrations was allowed to interact with the coated wells for 1 hour at 37°C before the addition of 50 nM biotinylated hemoglobin. After subtracting from the background (BSA), the percentage of inhibition was calculated with the following equation: 100 – [(NTR-Hb binding with TRL186) / (NTR-Hb binding) × 100].

GAS Growth in Hemoglobin and mAb

Todd Hewitt broth with 0.5% yeast extract (THYB) containing 3 mM 2,2′ dipyridyl (THYB-DP) with 5 µM hemoglobin and 20 µg/mL antibody (TRL186 or TRL96) were inoculated with overnight GAS cultures (starting OD₆₀₀ = 0.01). The culture optical density was determined following 20-hour incubation at 37°C.

Growth and Differentiation of HL60 Cells

HL60 promyelocytic leukemia cells (American Type Culture Collection, CCL240) were maintained, passaged, and differentiated into granulocytes (with 100 mM dimethylformamide) as described [23]. The viability of the differentiated cells was assessed using trypan blue exclusion and was considered acceptable if >90% of the cells excluded the azo dye. HL60 cells were used for the bacterial killing assay at day 5 postdifferentiation as long as the expression of CD35 (complement receptor 1) was up-regulated by ≥55% of the cell population and that of CD71 (transferrin receptor) was down-regulated by ≤15% of the cell population, as determined by flow cytometry (BD LSRFortessa) [24, 25].

Phagocyte-Dependent Killing

GAS killing was evaluated as described [23, 26]. In brief, mAbs diluted in Hanks balanced salt solution containing 0.1% gelatin (HBSS-G) was added to a 96-well microplate (10 µL/well). Twenty microliters of GAS culture (50 colony-forming units [CFUs]/µL) and 10 µL of 50% baby rabbit serum were added to wells, and the plates were incubated at 37°C for 30 minutes on an orbital shaker at 200 rpm. Forty microliters of differentiated HL60 cells (1 × 10⁴/µL) freshly prepared in HBSS-G was added, and the plates were incubated at 37°C for an additional 45 minutes with shaking. Surviving bacteria were enumerated by plating on THYA plates (100 µL/plate). GAS killing was calculated as following equation: (CFU [control] – CFU [mAb]) / CFU [control]) × 100.

Passive Vaccination and GAS Infection Model

MGAS5005 cells were harvested at the mid-logarithmic phase (OD₆₀₀ = 0.7), washed, and resuspended in 0.9% saline. Bacterial concentration was determined by microscopic counts and verified by plating. CD-1 mice (20–22 g, Charles River Laboratories) were infected by intraperitoneal injection of 0.1 mL cell suspension. Mice were weighed and administrated intraperitoneally with a single dose of TRL186 (15 mg/kg) or with PBS 1 hour before challenge with 5 × 10⁷ CFUs (prophylactic model) or 4 hours after infection with 1 × 10⁸ CFU (therapeutic model). Fifteen mice were used in experimental groups with at least 1 repeat. Mice were observed 4 times per day after the challenge. Morbid animals were euthanized according to protocols approved by the Georgia State University Institutional Animal Care and Use Committee.

Structural Prediction

The first DUF 1533 domain was predicted with Modeller [27]. The structure of the second DUF 1533 (PBD: 6DK) was used as a template [15]. RAMPAGE was used to generate a Ramachandran plot to verify the bond angles and torsional strain of the predicted structure [28]. The model was refined until 98% of the residues resided in the favored region, 2% in the allowed, and no in the outlier region.

Statistical Analysis

Data presented are averaged from experiments repeated at least twice. The Student t test was used for testing significance when comparing 2 groups (P ˂ .05). Kaplan–Meier plots of survival and log-rank tests were used for comparison of protection by immunization.

RESULTS

We selected 2 lead candidates, TRL96 and TRL186, based on high specificity and affinity to Shr, and investigated their efficacy in protection against GAS infection.

TRL96 and TRL186 Recognize Shr on GAS Surface and Enhance Killing by Cultured Phagocytes

To examine if TRL96 and TRL186 interact with GAS cells, we performed a whole-cell ELISA, where GAS cells were used to coat the microtiter plates and allowed to react with the mAbs (Figure 1A). We used an NZ131 mtsR mutant, in which the shr gene is deregulated since shr is repressed by MtsR in standard laboratory medium due to iron concentration [29]. Both mAbs generated significant binding to the bacteria with a stronger signal for TRL186 (OD₄₀₅ = 0.7), providing nearly half the signal as rabbit Shr-antiserum [20] or a commercially available GAS antibody reactions containing TRL308, a human mAb to respiratory syncytial virus, resulted only in a low background similar to normal rabbit serum or without primary antibody. We tested TRL186 and TRL96 binding to a series of NZ131 mutants with complete or various in-frame deletions in Shr [12, 18] (Supplementary Figure 1). A Δshr mutant did not react with TRL186 or TRL96, establishing Shr specificity in vivo. TRL186 did not bind to GAS expressing Shr mutants lacking the NTR, and TRL96 did not interact with strains expressing Shr variants without NEAT1 domain. We also examined the mAb impact on GAS killing by cultured phagocytes (Figure 1B). Preincubation with TRL186 or TRL96 resulted in 35% and 44% reduction in GAS recovery following incubation with differentiated HL-60 cells compared to the reactions containing normal rabbit serum.

Figure 1.

TRL96 and TRL186 bind to group A Streptococcus (GAS) and enhance opsonization. A, Microtiter plate wells coated with ZE491 GAS cells were allowed to react with monoclonal antibodies (mAbs), rabbit anti-GAS serum (type-specific carbohydrate, Abcam), normal rabbit serum (NRS), or no primary antibody control, and binding was detected by alkaline phosphatase–conjugated secondary antibodies. Data are from 2 experiments done in triplicate. B, GAS cells (1000 colony-forming units) preincubated with mAb or NRS in 50% baby rabbit serum were added to HL60 cells. GAS survival after 30 minutes of incubation was determined. Data were derived from 3 experiments. The error bars represent the standard deviation. Abbreviations: Ab, antibody; GAS, group A Streptococcus; NRS, normal rabbit serum; OD₄₀₅, Optical density at 405 nm; PBS, phosphate-buffered saline.

TRL186 Protects Mice From Intraperitoneal Challenge With an Invasive GAS Strain

Since NZ131 (and its isogenic derivatives) exhibits low virulence in murine models, we used the hypervirulent MGAS5005 [30] to evaluate the mAbs’ defense against an aggressive GAS infection. An ELISA with immobilized GAS cells confirmed binding of TRL186 and TRL96 to intact MGAS5005 cells, but no interactions were observed with the negative control, TRL308 (Supplementary Figure 2). We also tested the mAb in a prophylactic model (Figure 2A). A single dose of each mAb (15 mg/kg) was administered intraperitoneally 1 hour prior to infection with MGAS5005 (10⁷ CFUs). TRL186 prophylactic administration increased mice survival from 68% (mock) to 100%. In contrast, mice receiving TRL96 were as susceptible as the control group. Because TRL96 did not offer protection when administrated prophylactically, we only tested TRL186 in the therapeutic effectiveness. The mAb was injected 4 hours postchallenge with 10⁸ CFUs of MGAS5005 to model a more aggressive disease (Figure 2B). We observed a rapid progression of infection in the control mice, which resulted in only 27% survival after 5 days. In contrast, mice that received TRL186 had a 73% survival rate. The defense conferred by TRL186 in both infection models suggests the potential utility of mAbs targeting Shr in the treatment of GAS infection.

Figure 2.

TRL186 protects mice from a systemic group A Streptococcus (GAS) infection. A, Prophylaxis of TRL96 and TRL186. B, Therapeutic efficacy of TRL186. Kaplan–Meier survival curves of immunized mice injected intraperitoneally with monoclonal antibody (mAb; 15 mg/kg) 1 hour before infection (A, 5 × 10⁷ colony-forming units [CFUs]; phosphate-buffered saline, n = 25; mAb group, n =15; P = .017) or 4 hours after infection (B, 1 × 10⁸ CFUs; both groups, n =15; P = .009) with MGAS5005. The data shown are pooled data from 2 independent experiments. Statistical significance was determined by the log-rank test. Abbreviation: PBS, phosphate-buffered saline.

TRL186 Interacts With a Short Segment in the NTR of Shr

An ELISA with immobilized Shr showed that TRL186 binds to the full-length Shr with high potency (K_D is in the 100 pM range, Figure 3A). We examined TRL186 binding site using ELISA with the recombinant Shr proteins encompassing the full-length protein, NTR, NEAT1, or NEAT2 (Figure 3B). TRL186 generated a strong binding signal with the NTR fragment similar to intact Shr, whereas only background signals were recorded from the wells containing NEAT1 or NEAT2. The binding studies with purified proteins or with whole cells establish that TRL186 recognizes Shr NTR (Supplementary Figure 1). To further define TRL186 target, we screened a library of overlapping peptides (15-mers with 12 overlapping amino acids) covering Shr-NTR. Among 109 peptides covering the Shr-NTR region, only 2 peptides (numbers 33 and 34) exhibited significant reactivity with TRL186 in an ELISA (Figure 3C). This analysis identified a short Shr segment (IKKGDKVTFISA) located at the end of the first DUF1533/HID [12, 15] region in the interaction with TRL186, which we mapped onto a predicted ribbon structure (Figure 3D).

Figure 3.

TRL186 affinity and binding site. A, TRL186 binds to Shr with high affinity. Microtiter plate wells coated with Shr were allowed to react with TRL186 in serial dilution starting at 5 μg/mL (30 nM). Antibody binding was plotted as a function of TRL186 concentration. B, TRL186 binding is localized to Shr-NTR. TRL186 was allowed to interact with wells coated with the full-length Shr, NTR, NEAT1 (N1), or NEAT2 (N2). C, TRL186 binds to peptides from Shr-NTR. TRL186 was allowed to react with immobilized peptides derived from NTR. The reaction in wells without antibody (no antibody), uncoated wells (phosphate-buffered saline), or bovine serum albumin–coated wells is shown. The data are pooled from 3 independent experiments. Error bars represent standard deviation. The top panel in (C) shows a schematic representation of Shr. The black arrow indicates the TRL186 binding site. D, Ribbon diagram of predicted DUF 1533 structure. The first DUF1533 region is shown in blue, and the TRL186 binding location is in red. Abbreviations: Ab, antibody; BSA, bovine serum albumin; OD₄₀₅, Optical density at 405 nm.

TRL186 Interferes With NTR-Hemoglobin Binding and Hemoglobin-Dependent Growth of GAS

Since the NTR mediates Shr binding to hemoglobin, we examined the Shr-hemoglobin interactions in the presence of TRL186 in an ELISA. After optimizing the hemoglobin binding by purified NTR protein (Figure 4A), we evaluated the NTR-hemoglobin binding in the presence of increasing concentrations of TRL186. Hemoglobin binding by NTR was inhibited in a dose-dependent manner by TRL186 (Figure 4B). The maximum inhibition (43%) was achieved at 2 µg/mL concentration.

Figure 4.

TRL186 interferes with hemoglobin (Hb) binding by Shr and group A Streptococcus (GAS) growth on Hb iron. NTR was allowed to react with biotin-labeled Hb in the absence (A) or presence (B) of TRL186. Alkaline phosphatase–conjugated streptavidin was used to detect the Shr/Hb interaction at 405 nm. Inhibition of Hb binding (%) is plotted as a function of TRL186 concentration. C, NZ131 GAS cells were inoculated into Todd Hewitt broth with 0.5% yeast extract (THYB), iron-chelated THYB (DP), or THYB-DP supplemented with Hb with and without TRL186 or TRL96. Growth was monitored at 600 nm. Data are from 2 independent experiments done in duplicate. Error bars represent standard deviation. *statistical significance. Abbreviations: DP, 2,2′ dipyridyl; Hb, hemoglobin; ns, not significant; OD₄₀₅, optical density at 405 nm; OD₆₀₀, optical density at 600 nm; THYB, Todd Hewitt broth with 0.5% yeast extract.

GAS requires iron and uses Shr to capture and import heme from the host hemoglobin [10, 12]. We therefore tested if the anti-Shr mAbs can obstruct the use of hemoglobin iron by GAS (Figure 4C). GAS cells were cultivated in iron-deplete THYB (ie, THYB-DP) with or without hemoglobin and the mAbs. As expected, GAS did not grow in the iron-deplete medium, but the addition 5 µM hemoglobin reinstated growth. Notably, the inclusion of TRL186 but not of TRL96 in the medium prevented growth restoration by hemoglobin. The addition of TRL186 to GAS cultures grown in either THYB or THYB-DP had no impact (Supplementary Figure 3A). The negative control mAb, TRL308, did not influence GAS growth on hemoglobin iron (Supplementary Figure 3B). Furthermore, a Δshr mutant, which exhibited very limited growth in THYB-DP with 5 μM hemoglobin, was not impacted by the addition of TRL186 (Supplementary Figure 3C). Together, these data show that TRL186 specifically impedes GAS growth on hemoglobin iron by interfering with Shr function.

Discussion

Herein, we report TRL186, a high-affinity mAb against Shr that defends from aggressive GAS disease in mouse models. In this study, we used a single B-lymphocyte technology to isolate a pool of mAbs against GAS-Shr. The lead candidates, TRL186 and TRL96, are the first human mAbs described against a protein required for GAS heme acquisition.

Shr is exposed on the streptococcal surface [18], and whole-cell ELISA revealed that the epitopes on Shr are recognized by the mAbs (Figure 1A). The interaction between GAS and the mAbs was lost when the shr gene was deleted, establishing specificity (Supplementary Figure 1). We previously showed that Shr antiserum evokes opsonizing antibodies that facilitate phagocytosis [20]. Here we observed that both TRL96 and TRL186 improved phagocyte-mediated killing of GAS in vitro (Figure 1B). Additional work is needed to determine unequivocally that GAS killing by the cultured phagocytes resulted from phagocytosis. Nevertheless, studies in an animal model with prophylactic administration of the antibody revealed that only TRL186 was protective (Figure 2A). These observations suggest that additional attributes differentiating between TRL186 and TRL96 play a role in the defense from GAS in vivo.

TRL186 also provided partial defense when administrated postinfection with a higher GAS dose; mice that were given TRL186 were almost 50% more likely to survive than mock-treated mice (Figure 2B). These observations are encouraging since defense by antibody administrated after infection can be more difficult to achieve compared to a prechallenge vaccination, especially in a model for aggressive disease. Prophylaxis with a human mAb to the hemoprotein receptor IsdB of Staphylococcus aureus, for example, confers significant protection in a lethal sepsis model, but not when given after the challenge, although it is protective in a staphylococcal catheter colonization model when administrated either pre- or postinfection [31]. Together, the prophylactic and therapeutic qualities of treatment with TRL186 underscore the potential of an antibody-based treatment modality that targets Shr.

Shr is a large surface protein that belongs to the GAS core genome [10, 32]. To better understand the mechanism of TRL186 in vivo protection, we investigated the antibody binding region (Figure 3B and Supplementary Figure 1). We narrowed the binding region by a screen of overlapping peptide library to a short sequence (IKKGDKVTFISA) at the end of the first DUF1533 region (Figure 3C). Using the protein prediction software Modeller, we generated a ribbon structure of this region (Figure 3D). The antibody binding location appears to be at the end of the antiparallel β-sheet directly, a region that interacts with hemoglobin [15]. The entire Shr protein is highly conserved (≥99% amino acid identity) across GAS strains, with no variations in the Shr-binding segment. Conservation also exists among Shr homologs in other streptococcal species, such as Streptococcus dysgalactiae (86% identity) and Streptococcus canis (72% identity). This conservation suggests that an Shr-based therapy will have broad coverage among GAS isolates, and it might be possible to isolate antibodies that protect against multiple streptococcal pathogens.

The interaction between bacterial receptors and host hemoproteins is the first step in the process of heme uptake; in the case of Shr, this step is mediated by its NTR. A recent study extends the hemoglobin-capture mechanism by showing that 2 domains of unknown function (DUF1533) in the NTR, now named HID, bind hemoglobin independently [15]. We tested the ability of hemoglobin to bind to purified Shr in the presence of the mAb, and found that TRL186 inhibited NTR-hemoglobin binding in a concentration-dependent manner (Figure 4A and 4B), raising the possibility that TRL186 interferes with heme capture from hemoglobin. Growth experiment in iron-deplete medium supplemented with hemoglobin with or without various mAbs demonstrated TRL186 had no impact on GAS growth in the iron-rich THYB, in iron-deplete THYB without hemoglobin, or on the Δshr mutant (Supplementary Figure 3A and 3C). Still, TRL186 impeded GAS growth in THYB-DP supplemented with hemoglobin (Figure 4C). Together, these data establish that TRL186 specifically hinders GAS growth on hemoglobin iron by interfering with Shr function. TRL96 interacts with the heme-binding module, NEAT1, but did not influence GAS growth on hemoglobin iron. Therefore, either TRL96 does not interfere with NEAT1 function or NEAT1 role is redundant due to the presence of NEAT2. Of note, the deletion of both but not individual NEAT domains in Shr is required for growth phenotype [12]. Since only TRL186 but not TRL96 defends from GAS infection, we suggest that the ability to interfere with hemoglobin use is a key component in protection. Without the benefit of using hemoglobin (the largest iron pool in the body), GAS is likely starving for the growth-essential iron during infection.

The central role of iron in infection makes the iron uptake machinery an attractive vaccine target. Cattle vaccination against a siderophore receptor, for example, reduced the load of E. coli O157 in feces and the rectoanal mucosa [33, 34]. Heme acquisition satisfies the iron requirement for GAS, and each of the heme uptake components characterized in GAS contributes to virulence [17, 18, 35–38]. Therefore, an antibody-mediated treatment impacting heme intake may offer a promising treatment strategy. Blocking heme uptake is a significant part of the protection mechanism against S. aureus provided by human mAbs against the heme receptor IsdA [39]. Redundancy often limits protection by an antibody that targets a single antigen [40], and using an antibody cocktail improved efficacy and strain coverage in several models for staphylococcal infections [41]. Hence, a combination therapy that consists of several mAbs that may include the heme receptors Shp [11] and HupY [35] and the binding protein SiaA (HtsA) [10] might increase the protection observed with TRL186.

Recently strides were made in antibody discovery and engineering for pathogens including S. aureus, Bacillus anthracis, and Clostridium difficile [40, 42–44], but only a few protective mAbs were described in GAS. These include an mAb to the platelet-activating factor acetylhydrolase, SsE. Prophylaxis with SsE mAb defends mice from MGAS5005 subcutaneous challenge [45]. To the best of our knowledge, TRL186 is the first mAb that increases mouse survival rate in a therapeutic model. This TRL186 study serves as a proof of concept for using mAbs that target heme uptake in GAS for immunotherapy.

Supplementary Data

Supplementary materials are available at The Journal of Infectious Diseases online. Consisting of data provided by the authors to benefit the reader, the posted materials are not copyedited and are the sole responsibility of the authors, so questions or comments should be addressed to the corresponding author.

Notes

Financial support. This work was supported by the National Institute of Allergy and Infectious Diseases Small Business Innovation Research to Trellis Bioscience (grant number R43AI096625-02). K. V. L. is supported by a Georgia State University Molecular Basis of Disease Fellowship for predoctoral studies.

Potential conflicts of interest. L. M. K. is an employee and shareholder of Trellis Bioscience, which provided funding to Z. E., who has no financial interest in the company. All other authors report no potential conflicts of interest.

All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.