Identification of immunostimulatory antigens in Group A Streptococcus–derived vesicles

Significance

Group A streptococci (GAS) are responsible for a wide array of infections, ranging from mild throat and skin infections to less common but more severe invasive infections and autoimmune complications. Despite being a significant burden for the global healthcare system, there is no licensed vaccine against this pathogen. Extracellular vesicles (EVs) have emerged as a novel vaccine platform against bacterial infections. Here, we demonstrate that GAS-derived EVs are potent immunostimulants with self-adjuvating properties. Additionally, we identified seven highly conserved immunogenic antigens in the EVs, providing a foundation for rational antigen selection in future vaccine development. Our results highlight the potential of GAS EVs as a promising vaccine platform against GAS infections.

Abstract

An effective vaccine against Streptococcus pyogenes (Group A Streptococcus, GAS) is urgently needed. Bacteria-derived extracellular vesicles (EVs) are emerging as promising vaccine platforms. In this study, we evaluated the immunostimulatory properties of GAS-derived EVs when administered intranasally in mice. Immunization induced a strong humoral response, including pathogen-specific immunoglobulin G (IgG) and immunoglobulin A (IgA) antibodies. Additionally, we observed local and systemic T cell responses, specifically a robust Th17 response along with a modest but statistically significant Th1 response. Through immunoprecipitation coupled with mass spectrometry and western blotting, we identified seven immunogenic proteins, including the lipoproteins MtsA, BMP, PrsA1, PrsA2, MetQ, MalX, and the glutamine transporter GlnP. These antigens were recognized by human sera from both healthy individuals and patients with necrotizing soft tissue infection. Finally, we demonstrate that these antigens stimulate production and secretion of IL-17A by lung cells and splenocytes and that they are highly conserved across diverse GAS M-types, highlighting their potential as broadly protective vaccine candidates.

Streptococcus pyogenes, also known as Group A streptococci (GAS), is associated with mild and life-threatening infections, spanning from pharyngitis and impetigo to severe invasive infections such as septicemia, streptococcal toxic shock-like syndrome and necrotizing fasciitis. Prolonged or repeated exposure to GAS infections can lead to poststreptococcal autoimmune sequelae like acute rheumatic fever and rheumatic heart disease. The global burden of severe diseases due to GAS infections is estimated to be at least 500,000 deaths a year (1). Additionally, pharyngitis is a significant driver of antibiotic prescription (2), hence, the development of a vaccine that prevents GAS diseases and associated antibiotic use is an urgent need. The development of GAS vaccines has faced many challenges including M-type diversity, with more than 200 emm types reported, poorly defined correlates of protection, and safety concerns due to the development of cross-reactive antibodies (3–5). Despite these obstacles, many vaccines are currently under development (4).

Bacteria-derived extracellular vesicles (EVs) have emerged as promising vaccine candidates (6). The most notable example is the licensed meningococcal vaccine Bexsero, which includes Neisseria meningitidis outer membrane vesicles in the formulation (7, 8). The protective efficacy of EV-immunization of mice was also shown for Gram-positive pathogens like Streptococcus pneumoniae and Staphylococcus aureus (9–11). EV-based vaccines offer advantages over conventional vaccines as these nanostructures are loaded with a wide array of bacterial antigens, which are presented to immune cells in their native conformation. EVs also possess inherent adjuvant properties due to the abundance of pathogen-associated molecular patterns (PAMPs) and can elicit cellular and humoral immune responses (12, 13).

Like many other Gram-positive pathogens, GAS releases EVs enriched in lipoproteins, virulence-associated and immunomodulatory factors (14–16). In this study, we explore the possibility of using GAS-derived EVs as an antigen presenting platform in vaccine development. We show that intranasal immunization with GAS-derived EVs elicits production of pathogen specific antibodies together with a CD4⁺ T cell response in mice. We identify seven immunostimulatory antigens: the lipoproteins MtsA, BMP, PrsA1, PrsA2, MetQ, MalX, and the glutamine transporter GlnP. We propose that these antigens could be used as vaccine candidates since they are also immunogenic in humans, can stimulate T cell responses, and are highly conserved across different GAS strains.

Results

Intranasal Immunization with GAS EVs Generates GAS-specific Antibodies with Opsonophagocytic Activity.

EVs have emerged as promising vaccine candidates against bacterial infections (8–10). S. pyogenes spontaneously releases EVs (GAS EVs) that vary in size from 40 to 400 nm in diameter (median size: 145 nm) and appear as single membrane particles when visualized by cryoelectron microscopy (Cryo-EM, SI Appendix, Fig. S1). GAS EVs are loaded with ~700 different proteins, as determined by mass-spectrometry analysis (Dataset S1). The cargo of these vesicles includes surface antigens, virulence factors, and other proteins previously proposed as vaccine candidates.

To test the immunostimulatory potential of GAS EVs, male C57BL/6 mice were immunized with EVs formulated with aluminum hydroxide (alum). As illustrated in Fig. 1A, mice received two intranasal immunizations (days 0 and 14) with 10 µg of alum-adjuvated EVs, while the control group was immunized with alum in phosphate-buffered saline (PBS). 21 d after the primary immunization, sera, nasal wash, lungs, and spleen were collected to evaluate humoral and T cell responses upon GAS EV-vaccination. GAS-specific IgG and IgA levels in sera and nasal wash, respectively, were measured by whole-cell enzyme-linked immunosorbent assay (ELISA). The levels of GAS-specific IgG in sera and IgA in nasal wash were significantly increased in the GAS EVs-immunized group as compared to the alum control group (Fig. 1B). These results show that intranasal delivery of alum-adjuvated EVs elicits production of IgG and IgA that recognize antigens on the bacterial surface. Additionally, the functionality of the antibodies in sera was evaluated using an in vitro opsonophagocytic killing (OPK) assay. Sera from EV+ alum-immunized mice enhanced the HL-60 OPK capacity of S. pyogenes as compared to sera from alum-immunized mice (Fig. 1C).

Fig. 1.

Intranasal immunization with alum-adjuvated GAS EVs generates GAS-specific antibodies. (A) Immunization scheme. 5-wk-old male C57BL/6 mice were immunized intranasally with 50 µL of GAS EVs (10 µg) formulated with aluminum hydroxide (Alum, 10 mg/mL). The control group was immunized with alum in PBS. (B) GAS-specific IgG and IgA levels in serum and nasal wash, respectively, were analyzed by whole-bacteria ELISA (plates were coated with heat-inactivated GAS 5448 AP). Each symbol represents data from an individual mouse. (C) Survival of GAS 5448 following opsonophagocytic assay with HL-60 cells. Pooled sera from five independent experiments were tested. Bars represent mean ± SEM of three independent experiments. Significance was determined by the two-tailed unpaired t test (IgG and bacterial survival %) or Mann–Whitney test (IgA). **P ≤ 0.01; ***P ≤ 0.001.

Intranasal Immunization with GAS EVs Induces Th17 and Th1 Responses.

Previous reports have shown that Th17 and Th1 cell responses are important for protection against GAS infections (17–21). Here, flow cytometry analysis revealed that immunization with alum-adjuvated EVs enhanced CD4⁺ T cell recruitment to the lungs as seen by an increase in the percentage (%) and total number of cells compared to alum-immunized mice (SI Appendix, Fig. S2A). Moreover, a significantly greater % and total number of CD4⁺ T cells expressed CD44 (SI Appendix, Fig. S2B), suggesting these are activated effector and/or memory CD4⁺ T cells. Importantly, we observed a significant increase in the percentage of activated CD4⁺ T cells expressing RORγT (Fig. 2A), the transcription factor responsible for Th17 differentiation, suggesting EV immunization triggers the induction and recruitment of Th17 effector cells to the lungs. In line with this, following nonspecific restimulation with phorbol 12-myristate 13-acetate (specified in methods) (PMA) and ionomycin, activated CD4⁺ T cells produced significantly more IL-17A in EV immunized mice compared to alum-immunized mice (Fig. 2B). To assess antigen-specific cytokine production, lung cells were stimulated ex vivo with EVs and secreted IL-17A was measured in cell culture supernatants by ELISA. Stimulation of lung cells from EV-immunized mice induced a significant increase in the secretion of IL-17A as compared to alum-immunized mice (Fig. 2C). Overall, these data demonstrate that intranasal immunization with alum-adjuvated EVs induces a robust antigen-specific Th17 response in the lungs.

Fig. 2.

Intranasal immunization with alum-adjuvated GAS EVs elicits Th17 and Th1 responses in the lungs. Mice were immunized as described in Fig. 1A. Lungs were collected 7 d after the last immunization and stained for flow cytometry analysis or cultured ex vivo to measure cytokines response to antigen stimulation. Representative (A) histogram and percentage (%) of RORγT⁺ CD44⁺ CD4⁺ T cells. Representative (B) flow cytometry plots, percentage (%) and total number (#) of IL-17⁺ CD44⁺ CD4⁺ T cells. Lung cells were cultured ex vivo and stimulated with GAS EVs or media for 72 h; secretion of (C) Th17-specific IL-17A was measured by ELISA. Representative (D) histogram and percentage (%) of T-bet⁺ CD44⁺ CD4⁺ T cells. Representative (E) flow cytometry plots, percentage (%) and total number (#) of IFNγ⁺ CD44⁺ CD4⁺ T cells. (F) Quantification of Th1-specific IFNγ secreted into the culture supernatant upon stimulation with GAS EVs or media for 72h. Cytokine data are normalized to media stimulation controls. Bars represent mean ± SEM of two (flow cytometry) or three (cytokines) independent experiments. Each symbol represents data from an individual mouse. Significance was determined by the two- tailed unpaired t test (% and # of IL-17⁺ T cells, % of IFNγ⁺ and RORγT⁺ T cells) or Mann–Whitney test (# of IFNγ⁺ T cells, % of T-bet⁺ T cells and IL-17A and IFNγ secretion). **P ≤ 0.01; ***P ≤ 0.001; ****P ≤ 0.0001; ns, not significant.

A Th1 response is associated with IFNγ production and controlled by the transcription factor T-bet. Interestingly, we observed a significant increase in IFNγ production but not T-bet expression from CD4⁺ T cells in the lungs following GAS EVs immunization (Fig. 2 D and E). Additionally, ex vivo stimulation of lung cells from EV-immunized mice induced a significant increase in the secretion of IFNγ as compared to alum-immunized mice (Fig. 2F), indicating that a small but significant Th1 response is also induced in the lungs upon immunization with alum-adjuvated EVs.

In addition to eliciting a local response in the lungs, intranasal immunization with alum-adjuvated EVs also induced a systemic Th17 response as evidenced by the significant increase in the % of IL-17⁺ activated T cells in the spleen (Fig. 3A), and increased secretion of IL-17A by splenocytes stimulated ex vivo with GAS EVs (Fig. 3B). Although there was no increase in IFNγ production from CD4⁺ T cells (Fig. 3C), a significant increase in IFNγ secretion by splenocytes stimulated with GAS EVs was observed (Fig. 3D). These results suggest that immunization with alum-adjuvated EVs can elicit systemic activation of CD4⁺ T cells and induce a Th17 response.

Fig. 3.

Intranasal immunization with alum-adjuvated GAS EVs induces a T cell response in the spleen. Mice were immunized as described in Fig. 1A. Spleen samples were collected 7 d after the last immunization. Representative (A) flow cytometry plots, percentage (%) and total number (#) of IL-17⁺ CD44⁺ CD4⁺ T cells. (B) Splenocytes were stimulated for 72h with GAS EVs or media, before secreted IL-17A was quantified in the supernatant. Representative (C) flow cytometry plots, percentage (%) and total number (#) of IFNγ⁺ CD44⁺ CD4⁺ T cells. (D) Quantification of IFNγ secreted into the culture supernatant upon splenocyte stimulation with GAS EVs or media for 72h. Cytokine data are normalized to media stimulation controls. Bars represent mean ± SEM of two (flow cytometry) or three (cytokines) independent experiments. Each symbol represents data from an individual mouse. Significance was determined by the two- tailed unpaired t test (# of IL-17⁺ T cells, % and # of IFNγ⁺ T cells and IFNγ secretion) or Mann–Whitney test (% of IL-17⁺ T cells and IL- 17A secretion). *P ≤ 0.05; ***P ≤ 0.001; ****P ≤ 0.0001; ns, not significant.

Our data demonstrate that intranasal immunization with GAS EVs formulated with aluminum hydroxide elicits a strong local and systemic Th17 response, alongside increased IFNγ production. However, EVs exhibit self-adjuvating properties due to their high content of PAMPs (6), therefore, we tested whether intranasal immunization with GAS EVs in the absence of adjuvant could induce similar T cell responses as those observed in the presence of aluminum hydroxide. As expected, intranasal immunization with GAS EVs alone triggered the induction and recruitment of Th17 effector cells to the lungs and spleen as evidenced by the significant increase in the % and total number of activated CD4⁺ T cells expressing IL-17A (Fig. 4 A and B). Additionally, a significant increase in IFNγ production from CD4⁺ T cells was also observed in the lungs and spleen of EV-vaccinated mice (Fig. 4 C and D). The increase in IL-17A⁺ activated CD4⁺ T cells in the lungs and spleen observed upon immunization with GAS EVs alone is comparable to that observed in the presence of aluminum hydroxide (Lungs: Figs. 2B vs. 4A, and Spleen: Figs. 3A vs. 4B). A robust IFNγ production was observed in the absence of alum (Lungs: Figs. 2E vs. 4C and Spleen: Figs. 3C vs. 4D), confirming that GAS EVs possess inherent adjuvant properties and can elicit cellular immune responses even in the absence of adjuvants.

Fig. 4.

Intranasal immunization with GAS EVs alone elicits Th17 and Th1 responses in the lungs and spleen. Mice were immunized twice with 10 µg of EVs following the immunization scheme described in Fig. 1A. Lung and spleen cells were collected 7 d after the last immunization and stained for flow cytometry analysis following nonspecific restimulation with PMA and ionomycin. (A and B) Percentage (%) and total number (#) of IL-17⁺ CD44⁺ CD4⁺ in lungs and spleen samples, and (C and D) percentage (%) and total number (#) of IFNγ⁺ CD44⁺ CD4⁺ T cells in lungs and spleen samples. Bars represent mean ± SEM of one experiment. Each symbol represents data from an individual mouse. Significance was determined by the two-tailed unpaired t test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P ≤ 0.0001; ns, not significant.

Identification of Immunostimulatory Antigens on GAS EVs.

Thus, we demonstrated that immunization with GAS EVs elicits a humoral and cellular response in mice (Figs. 1–4). Next, we sought to identify the antigens responsible for these immunostimulatory properties of GAS EVs. Many proteins, ranging in size from ~30 to 80 kDa, were detected in GAS EVs when probed with sera from EV+ alum-immunized mice (Fig. 5A). These immunoreactive antigens were identified by coimmunoprecipitation coupled to mass spectrometry (Table 1 and Dataset S2). Five different lipoproteins (PrsA1, PrsA2, MtsA, SPy_1228, and SPy_0319) and a transmembrane protein (SPy_1315) were identified. PrsA1 and PrsA2 are extracellular parvulin-type peptidyl-prolyl isomerases that assist in the folding of exported proteins (22). MtsA, SPy_1228, and SPy_0319 are substrate binding lipoproteins of different ABC transporters. MtsA is involved in iron/manganese transport (23–26). SPy_1228 and SPy_0319 (from now on referred to as BMP and MetQ, respectively) have not been characterized in GAS but share sequence similarity with lipoproteins involved in the transport of ribonucleosides and methionine, respectively (27, 28). SPy_1315 is annotated as an ABC transporter substrate-binding protein/permease and shares sequence similarity with the glutamine permease GlnP (29).

Fig. 5.

Identification of immunostimulatory antigens in GAS EVs. (A) Two and five µg of GAS 5448 AP EVs were loaded on an SDS-PAGE and stained with Coomassie (Left). Western blot analysis of GAS EVs (2 µg) using sera from EV+ alum immunized mice (Right). (B) Antigen-specific IgG levels in sera vs. recombinantly expressed MtsA, BMP, PrsA1, PrsA2, MetQ, GlnP, and MalX were measured by ELISA. Bars represent mean ± SEM. Statistical comparisons were made using the Mann–Whitney test. *P < 0.05; **P < 0.01; ***P < 0.001; ****P ≤ 0.0001; ns, not significant.

Table 1.

Immunostimulatory antigens identified in GAS EVs

Antigen	MW	Annotation	Protein ID
SPy_1315 (GlnP)	78.4 kDa	ABC transporter substrate-binding protein/permease	WP_011285590.1
PrsA1	38.5 kDa	Peptidylprolyl isomerase PrsA	WP_002984003.1
SPy_1228 (BMP)	36.4 kDa	BMP (Basic membrane protein) family lipoprotein	WP_002989701.1
MtsA	34.3 kDa	Metal ABC transporter substrate-binding protein	WP_004218965.1
PrsA2	34.3 kDa	Peptidylprolyl isomerase PrsA	WP_010922719.1
SPy_0319 (MetQ)	31.1 kDa	MetQ/NlpA family ABC transporter substrate-binding protein	WP_010921899.1

To validate MtsA, BMP, PrsA1, PrsA2, MetQ, and GlnP as immunoreactive antigens, these were expressed and purified as recombinant proteins lacking signal peptide or membrane domains. Antigen-specific IgG levels were measured by ELISA and western blot; this dual approach allowed the detection of both conformational and linear epitopes, respectively. MalX, a lipoprotein found in pneumococcal vesicles and important for protection against invasive pneumococcal disease (9), was included as a control. ELISA results demonstrated that immunization with GAS EVs formulated with alum, but not alum alone, induced a significant increase in the IgG levels specific for GlnP. Additionally, a smaller but significant increase in IgG levels against MtsA, PrsA1, and MetQ were also observed (Fig. 5B). Western blot analysis revealed that GlnP, MtsA, and PrsA2 were strongly recognized by pooled sera from EV+ alum-immunized mice, followed by BMP and PrsA1, whereas MetQ and MalX were not detected (SI Appendix, Fig. S3). Altogether, these results confirm that immunization with GAS EVs elicits the production of antibodies specific to six different GAS surface proteins: MtsA, BMP, PrsA1, PrsA2, MetQ, and GlnP, highlighting their potential as vaccine candidates. These data also show that GlnP is the most immunogenic of all the six antigens (Fig. 5B and SI Appendix, Fig. S3).

To explore the relevance of these antigens in humans, recombinantly expressed proteins were probed with commercial human AB serum (Merck, H4522). BMP and GlnP were strongly recognized by human sera, while PrsA1, PrsA2, and MalX were also recognized but to a minor extent (Fig. 6A). Additionally, antigen-specific IgG levels in plasma from patients with GAS-associated necrotizing soft tissue infection (NSTI) were measured by ELISA. As shown in Fig. 6B, antibodies against all the antigens (MtsA, BMP, PrsA1, PrsA2, MetQ, GlnP, and MalX) were detected. Notably, the levels of anti-BMP, PrsA2, GlnP, and MetQ antibodies were higher than of anti-PrsA1, MalX, and MtsA antibodies. Overall, these results suggest that the immunoreactive antigens identified in GAS EVs (MtsA, BMP, PrsA1, PrsA2, MetQ, GlnP, and MalX) are also relevant for humans, highlighting their potential as vaccine candidates.

Fig. 6.

Immunostimulatory antigens identified in GAS EVs react with human plasma. (A) Western blot against recombinantly expressed MtsA, BMP, PrsA1, PrsA2, MetQ, GlnP, and MalX (2 µg of each protein, see SI Appendix, Fig. S3 for Coomassie) using commercial human AB serum (Merck, H4522). (B) Antigen-specific IgG levels in plasma from patients with NSTI vs. recombinantly expressed MtsA, BMP, PrsA1, PrsA2, MetQ, GlnP, and MalX were measured by ELISA. Samples incubated without human plasma were used as negative controls (Ctrl −). Bars represent mean ± SEM. Each symbol represents data from an individual patient. Statistical comparisons were made using the Mann–Whitney test (PrsA1 and MetQ) or two-tailed unpaired t test (MtsA, BMP, PrsA2, GlnP, and MalX). *P < 0.05; **P < 0.01; ns, not significant.

GAS EVs-Immunostimulatory Antigens Drive a Th17 Response.

Ex vivo stimulation of lungs and spleen cells from EV-immunized mice with GAS EVs induced a significant increase in the secretion of IL-17A and IFNγ (Figs. 2 C and F and 3 B and D). To determine whether the immunostimulatory antigens identified in GAS EVs are responsible for the CD4⁺ T cell activation, splenocytes and lung cells from alum- and GAS EVs-immunized mice were stimulated ex vivo with recombinantly expressed MtsA, BMP, PrsA1, PrsA2, MetQ, GlnP, or MalX for 72 h. Secreted IL-17A and IFNγ in cell culture supernatants were measured by ELISA. As shown in Fig. 7, all antigens can induce secretion of IL-17A by lung (Fig. 7A) and spleen (Fig. 7B) cells of mice immunized with GAS EVs, but not alum alone. Of all the tested antigens, BMP induced the highest levels of IL-17A secretion followed by MetQ. In the case of IFNγ, none of the recombinant proteins induced a significant increase in secretion by lung cells, while only BMP induced a significant increase in IFNγ secretion by splenocytes (SI Appendix, Fig. S4). These results show that MtsA, PrsA1, PrsA2, MetQ, GlnP, and MalX are Th17 antigens, while BMP can induce production and secretion of both IL-17A and IFNγ.

Fig. 7.

IL-17A response to antigen stimulation. (A) Lung cells and (B) splenocytes from mice immunized with alum or GAS EVs formulated with alum were collected 1 wk after the second immunization. Cells were cultured ex vivo and stimulated with 1 µg of recombinantly expressed proteins (MtsA, BMP, PrsA1, PrsA2, MetQ, GlnP, or MalX) or media for 72 h. The concentration of IL-17A in culture supernatants was measured by ELISA. Data are normalized to media stimulation controls and represented as mean ± SEM of two independent experiments. Statistical comparisons were made using the Mann–Whitney test. **P < 0.01; ***P < 0.001; ****P ≤ 0.0001.

Vaccine Candidates Identified in GAS EVs Are Conserved Across Different M-Types.

Ideal vaccine candidates should be conserved in different M-types and provide broad coverage across GAS strains. Bioinformatic analysis revealed that MtsA, BMP, PrsA1, PrsA2, MetQ, and GlnP are highly conserved across different M-types (>95% amino acid identity) and present in 100% of the 1340 genomes analyzed. MalX, which was tested due to its relevance as an immunogenic antigen in S. pneumoniae (9), also shares > 95% amino acid identity but is present only in a limited number of GAS strains (Dataset S3). Additionally, we demonstrate that the same protein profile observed in GAS EVs (purified from GAS 5448 AP) can be observed in the cell lysates of different invasive GAS isolates (M-types: 1, 3, 4, 6, 12, 28, 75, and 89) (Fig. 8A). Furthermore, serum IgG from mice vaccinated with EVs can bind to the surface of different invasive GAS isolates as demonstrated by whole-cell ELISA (Fig. 8B). These results suggest that the immunostimulatory antigens identified in GAS EVs are conserved across different M-types and that the antibodies generated upon immunization with GAS 5448 AP-EVs of M-type 1, cross-react with many other M-types.

Fig. 8.

Vaccine candidates identified in GAS EVs are conserved across GAS strains. (A) Western blot analysis of GAS EVs and bacterial lysates of different invasive GAS isolates (M-type is indicated in brackets). Samples were probed with sera from GAS EVs-immunized mice (Top panel) or anti-GAPDH as a loading control (Bottom panel). (B) IgG binding to different GAS strains was analyzed by whole-cell ELISA. Plates were coated with different invasive GAS strains, covering a range of M-types, and probed with sera from mice vaccinated with alum (light gray circles) or GAS EVs formulated with alum (blue circles). Bars represent mean ± SEM of four independent experiments. Each symbol represents data from an individual mouse. Significance was determined by the two-tailed unpaired t test (S073, S221, S338, and S530) or Mann–Whitney test (AP, S011, S137, and S150). ****P ≤ 0.0001.

Discussion

Group A streptococcal diseases represent a significant burden for public health, however, there is currently no licensed vaccine against this pathogen. In this work, we examined the immunogenicity of GAS-derived EVs. Our findings demonstrate that GAS EVs are potent immunostimulants that elicit local and systemic humoral and T cell responses upon intranasal delivery in mice. GAS EVs served as an antigen presenting platform that allowed identification of the immunostimulatory antigens MtsA, BMP, PrsA1, PrsA2, MetQ, MalX, and GlnP. These antigens elicit IgA, IgG, and IL-17A production in mice, are relevant in humans, and are highly conserved across different GAS M-types, highlighting their potential as broadly protective vaccine candidates.

Protective immunity to GAS infections is not well understood, however, it is likely to require the presence of pathogen specific antibodies that can promote opsonophagocytosis, neutralization of toxins/virulence factors and blocking of bacterial adhesion together with a CD4⁺ T cell response (3). Accumulating evidence suggest that induction of mucosal Th17 and antibody responses can protect from S. pyogenes infections (18, 19, 21, 30–34), probably by IL-17A mediated recruitment and expansion of neutrophils and macrophages, key cells in GAS clearance (34–37). In this work, we demonstrate that intranasal immunization with alum-adjuvated GAS EVs induces a strong mucosal response, characterized by production of IgA and IL-17A producing CD4⁺ T cells. Additionally, it elicits the production of opsonizing IgG antibodies in sera and a robust systemic Th17 response. A similar, local and systemic, Th17 response was also observed after immunization with GAS EVs in the absence of adjuvant, confirming their inherent adjuvant properties. Together, our data highlight the potency of GAS EVs as immunostimulants and their potential role in eliciting protective immunity against mucosal and systemic GAS infections.

Th1 responses are also important for protection against GAS infections, and these may play a key role in fighting invasive diseases, while Th17 responses may be more relevant against mucosal infections. In fact, it was previously shown that intranasal inoculation with heat killed GAS induced IL-17A-producing T cells, while intravenous and subcutaneous challenge induced IFNγ-producing cells (33). Moreover, two independent studies showed that GAS vaccines can protect from invasive disease when formulated with Th1-promoting adjuvants (17, 38). Although the T cell response we observed upon intranasal immunization with alum-adjuvated GAS EVs was dominantly Th17, we also found a small, but significant increase in the production of IFNγ. Importantly, IFNγ production was higher in lung and spleen cells from mice immunized with EVs in the absence of alum, which preferentially induces a Th2-like response (39). It is therefore possible that intranasal immunization with EVs alone can induce an immune response protective against invasive disease.

EVs have emerged as a novel vaccine platform against bacterial infections, not only due to their potent immunostimulatory and self-adjuvating properties but also due to the display of multiple antigens in their native conformation (6). In this work, GAS EVs served as an antigen presenting platform that allowed identification of the highly conserved immunostimulatory antigens MtsA, BMP, PrsA1, PrsA2, MetQ, and GlnP. Although MalX was not found in GAS EVs, it was included due to its role in EV-mediated protection against invasive pneumococcal disease (9). Of the seven antigens, six are lipoproteins involved in central metabolic processes. PrsA1 and PrsA2 are parvulin-type peptidyl-prolyl isomerases required to maintain the GAS proteome homeostasis by assisting in the folding and maturation of extracellular proteins (22). MtsA, BMP, MetQ, and MalX are substrate-binding lipoproteins of ABC transporters involved in the uptake of iron, zinc, and manganese [MtsA, (23–26)], ribonucleosides [BMP, (27)], methionine [MetQ, (40)] and maltose [MalX, (41, 42)]. GlnP is the product of Spy_1315, which encodes for an ABC transporter permease subunit involved in L-glutamine transport. GlnP shares high similarity with parts of the ABC transporter involved in lysine, arginine, and ornithine transport in Streptococcus infantarius, suggesting GlnP may also be involved in L-lysine transport, which is essential for bacterial growth (43). Previous work demonstrated that protein lipidation enhances immunogenicity and protective efficacy (9, 11, 44–46). Here, we show that EV-associated lipoproteins are recognized better than recombinant proteins lacking signal peptide and lipidation when probed with sera from GAS-EV immunized mice. This suggests that antigen delivery in EVs plays a key role in immunogenicity.

In this study, MetQ, MalX, and GlnP are reported as vaccine candidates in GAS. These three proteins were here shown to induce IL-17A production by lung cells and splenocytes from EV-vaccinated mice and are therefore good Th17 antigens. Additionally, antibodies against MetQ, MalX, and especially GlnP were found in sera from healthy individuals and NSTI patients. Of the seven antigens tested here, GlnP was the most immunogenic in mice and humans, which together with its key role in bacterial growth highlights its potential as a vaccine candidate.

MtsA, BMP, PrsA1, and PrsA2 have previously been identified as potential vaccine candidates (38, 47–50). Immunization with BMP, either alone or in combination with other antigens (Spy7 vaccine) protected against skin infection or bacterial dissemination, respectively (48, 50). Similarly, immunization with PrsA1 and PrsA2 conferred protection against invasive challenge (38). Our data support the role of MtsA, BMP, PrsA1, and PrsA2 as vaccine candidates. BMP was strongly recognized by sera from healthy individuals and NSTI patients, induced production of high levels of IL-17A in lung cells and splenocytes from EV-vaccinated mice and was the only antigen capable of stimulating IFNγ-production by splenocytes.

In conclusion, our work highlights the potential of using EVs including MtsA, BMP, PrsA1, PrsA2, MetQ, MalX, and GlnP as vaccine candidates. These membrane associated antigens play key roles in bacterial growth, nutrient uptake, biofilm formation, oxidative stress resistance, survival in murine macrophages and virulence (22, 26, 27, 29, 40, 42, 43, 51, 52). Additionally, these antigens are highly conserved across different M-types, are relevant in humans, and capable of eliciting immune responses associated with protection.

Materials and Methods

Human Subjects.

The study included analyses of plasma samples collected from patients enrolled in the INFECT patient cohort (ClinicalTrials.gov, NCT01790698). Samples from three S. pyogenes infected NSTI patients enrolled at Sahlgrenska University hospital were included. The study was approved by the Regional Ethics Committee in Gothenburg (ref. no. 930-12, amendment ref no. 2024-05651-02). The study was conducted in accordance with the Declaration of Helsinki, and all patients provided informed consent prior to enrollment and sample collection.

Bacterial Strains and Growth Conditions.

All bacterial strains used in this study are listed in SI Appendix, Table S1. S. pyogenes was grown on tryptic soy agar (TSA) plates supplemented with 5% sheep blood at 37 °C in a 5% CO₂ atmosphere. Liquid cultures were grown statically at 37 °C in either Todd-Hewitt broth alone (THB) or supplemented with 0.2% yeast extract (THY). Escherichia coli was grown aerobically in lysogeny broth (LB) at 37 °C with constant shaking. When required, E. coli cultures were supplemented with 100 μg/mL spectinomycin or 50 μg/mL kanamycin.

Isolation of EVs.

GAS 5448 AP was grown in THB to an OD₆₀₀ of 0.6. Bacteria were pellet by centrifugation at 8,000 rpm for 20 min, the supernatant was recovered and filtered through a 0.22uM filter (Filtropur S 0.2). Samples were then concentrated 10× with a tangential flow filtration capsule (MWCO: 100 kDa, PALL Corporation). EVs were pelleted by ultracentrifugation (30,000 rpm for 4 h at 4 °C) and further purified using OptiPrep (iodixanol) Density Gradient Medium (Sigma-Aldrich). Crude EVs were resuspended in 50% (v/v) Optiprep in solution A (0.25 M sucrose, 6 mM ethylenediaminetetraacetic acid (EDTA), 60 mM HEPES-NaOH, pH 7.4) and overlaid with 30% and 5% (v/v) Optiprep in solution B (0.25 M sucrose, 1 mM EDTA, 10 mM HEPES-NaOH, pH 7.4). Separation was achieved by ultracentrifugation at 35,000 rpm for 6 h. The interphase between the 5% and 30% Optiprep solutions was recovered and washed three times with PBS (for the first wash, EVs were centrifuged at 35,000 rpm for at least 6 h; 2nd and 3rd washes were centrifuged at 30,000 rpm for 4 h).

The EV protein concentration was determined using the Pierce bicinchoninic acid Protein Assay kit (Thermo Fisher Scientific). EVs were tested for sterility by spotting on blood agar plates and then stored at −80 °C until further use.

The GAS EVs particle size and count distribution was measured by dynamic light scattering using a Zetasizer Ultra instrument (Malvern Panalytical). Vesicles samples were diluted 1:250 in water and equilibrated for 2 min at 25 °C before each measurement.

Mice Immunization.

All animal experiments were approved by the local ethical committee (Stockholms Norra djurförsöksetiska nämnd). Groups (n = 5 or n = 4 for control groups) of 5- to 6-wk-old male C57BL/6 mice were immunized with 10 μg of GAS EVs either alone or formulated with 10 mg/mL aluminum hydroxide in PBS. Control groups were immunized with PBS or alum (10 mg/mL). Mice were anesthetized by isoflurane (Abbott) inhalation and then immunized intranasally with 50 µL/mouse of the formulations on days 0 and 14. No signs of discomfort or distress were observed after immunization with EVs, alum hydroxide, PBS, or any of the formulations used in this study. 7 d after the second immunization, mice were killed, and nasal wash (in PBS), blood (obtained via heart puncture), spleen, and lungs were recovered under aseptic conditions. Male mice were used due to their higher susceptibility to GAS infections than female mice (53–55).

Whole-Bacterial-Cell ELISA.

GAS 5448 AP or different invasive GAS isolates (SI Appendix, Table S1) were grown at 37 °C in THY to an OD₆₀₀ of 0.6. Bacteria were washed two times with PBS supplemented with EDTA-free protease inhibitor cocktail (cOmplete, Roche) and heat inactivated for 2.5 h at 60 °C. Nunc MaxiSorp plates (Thermo Scientific) were coated with 100 µL of the bacterial suspension (adjusted to an OD₆₀₀ of 0.6 in PBS + cOmplete, Roche) and incubated overnight at 4 °C. Next, plates were blocked for 2 h at room temperature with 200 μL of 2.5% skim milk in PBS. After blocking, plates were incubated for 2 h at room temperature, first with 100 µL of mice sera (1:500 in PBS) and then with 100 µL of goat anti-mouse IgG-HRP (Sigma) diluted 1:5000 in 1% BSA in PBS. Plates were washed three times with PBS after each incubation step. Finally, samples were developed with the 3,3‘,5,5’-tetramethylbenzidine substrate reagent set (BD OptEIA) for 20 min. The reaction was stopped with 50 µL of sulfuric acid 0.5 M and the absorbance was measured at 450 and 570 nm. For nasal washes, samples were not diluted and incubated overnight. Goat anti-mouse IgA-HRP (Life technologies) was used at a 1:1000 dilution in 1% BSA in PBS.

OPK Assay.

HL-60 cells were differentiated by culturing in RPMI 1640 with 0.6% N,N-dimethylformamide (Thermo Scientific Chemicals) for 5 d prior to OPK assay. 10 microliters of 1 × 10⁸ S. pyogenes 5448 were incubated with 20 µL pooled sera from alum- or EV+alum-immunized mice for 1 h at 37 °C. After incubation, bacterial samples were divided into duplicate wells. Fifty microliters of 1 × 10⁷ HL-60 cells/mL with 1:10 final volume of baby rabbit complement (Bio-Rad, Cat. No. C12CA) were added to one well. The other well was incubated with media and served as the bacteria only control. Samples were incubated for 1 h at 37 °C, afterward, samples were serially diluted in PBS and plated onto blood agar plates. Pooled sera from five independent experiments were run. Bacterial survival in presence of neutrophils was calculated by normalizing to the number of colony forming units in corresponding bacteria only control.

Western Blotting.

Two micrograms of protein samples (GAS EVs, recombinantly expressed proteins or cleared cell lysates) were mixed with 1X NuPAGE LDS sample buffer (Invitrogen) and boiled for 10 min. Samples were resolved by SDS-PAGE using 4 to 12% NuPAGE Bis-Tris protein gels (Invitrogen) and either stained with Imperial Protein Stain (Thermo Scientific) or transferred to PVDF (polyvinylidene difluoride) membranes. Membranes were blocked overnight with 5% skim milk in tPBS (PBS containing 0.1% Tween-20). After three washes with tPBS, membranes were incubated with primary antibodies: pooled sera from immunized mice (1:3,000 in PBS for 3 h), rabbit anti-GAPDH (1:20,000 in PBS for 1 h), or human serum from human male AB plasma (1:500 in 1% BSA, overnight. Merck). Membranes were washed three times with tPBS and then incubated with the HRP-conjugated secondary antibodies: goat anti-mouse IgG (1:3,000 in PBS for 2 h. BioRad), goat anti-rabbit IgG (1:10,000 in PBS for 1 h. Invitrogen), or goat anti-human IgG (1:1,000 in PBS for 3 h. Chemicon International). Membranes were washed with tPBS and further developed with Amersham enhanced chemiluminescence Plus western blotting detection system (GE Healthcare Life Sciences) using ChemiDoc XRS+ (Bio-Rad Laboratories).

Coimmunoprecipitation.

Protein G HP SpinTrap/Ab columns (GE Healthcare) were pre-equilibrated with PBS. Protein G Sepharose was removed from the columns and incubated with pooled sera (10 µL sera/400 µL PBS) from immunized mice (GAS EVs/alum or PBS/alum). After 1 h incubation at 37 °C, samples were centrifuged (100 × g for 1 min); after removing the supernatant, the pelleted Protein G Sepharose was washed four times with PBS. In parallel, 30 µg of GAS EVs were solubilized by incubation with Pierce IP lysis buffer (Thermo Scientific) at 4 °C for 1 h. Coupled Protein G Sepharose was then mixed with solubilized GAS EVs and incubated for 2 h at 4 °C. Finally, Protein G Sepharose was washed six times with PBS as described above. Two strategies were employed to elute coimmunoprecipitated proteins: i) coupled Protein G Sepharose was resuspended in 1X NuPAGE LDS sample buffer (Novex) and subjected to SDS-PAGE, after 10 min, running was stopped, and 1 cm gel slices were cut and used for mass spectrometry. ii) proteins were eluted from Protein G Sepharose with 400 µL of 0.1 M glycine-HCl, pH 2.7 followed by neutralization with 40 µL of 1 M Tris pH 8.0. Samples were concentrated using SpeedVac (Thermo Fisher) and sent for mass spectrometry.

Mass Spectrometry.

Coimmunoprecipitation samples.

Samples were prepared using a standard (in gel or in solution) alkylation and tryptic digestion protocol. Eluted peptides were cleaned up using C18 StageTip and dried down. Resolubilized peptides were analyzed on a Q Exactive HF instrument (Thermo Scientific) using a 60 min linear gradient elution on a 50 cm long C18 column. Acquisition method: DDA (Data-Dependent Acquisition). Tandem mass spectrometry (MS/MS) was performed applying HCD (higher-energy collisional dissociation). A database search was carried out using Mascot server (v2.5.2) for peptide matches and protein searches against the S. pyogenes ATCC 700294 (SF370, Serotype M1) genomic database UniProtKB (Proteome ID: UP000000750, downloaded in 2022).

GAS EV samples.

S. pyogenes EVs were prepared as described in previous sections. Three independent samples, consisting of 30 µg of total EVs proteins were sent to mass spectrometry and processed as described above. A database search was carried out using MSFragger for peptide matches and protein searches against the S. pyogenes ATCC 700294 (SF370, Serotype M1) genomic database UniProtKB (Proteome ID: UP000000750, downloaded in 2024).

Expression and Purification of Recombinant Proteins.

For protein purification, the soluble region of GlnP (codons 30-520, WP_011285590.1) or the mature form (lacking the signal peptide) of MtsA (codons 22-310, WP_004218965.1), Spy_1228 (BMP, codons 22-350, WP_002989701.1), PrsA1 (codons 24-351, WP_002984003.1), PrsA2 (codons 24-309, WP_010922719.1), Spy_0319 (MetQ, codons 23-281, WP_010921899.1), and MalX (codons 26-419, WP_010922390.1) were cloned into pET28a, which allows the expression of C-terminally 6xHis tagged proteins. Briefly, glnP, mtsA, BMP, prsA1, prsA2, metQ, and malX were amplified from gDNA using primer pairs glnP.fw/glnP.rv, mtsA.fw/mtsA.rv, BMP.fw/BMP.rv, prsA1.fw/prsA1.rv, prsA2.fw/prsA2.rv, metQ.fw/metQ.rv, and malX.fw/malX.rv (primer sequences are shown in SI Appendix, Table S2). The resulting PCR products were cloned into NcoI/XhoI digested pET28a using the NEBuilder HiFi assembly kit (New England biolabs) according to the manufacturer’s instructions. BL21 (DE3) (Invitrogen) cells containing the pET28a constructs were cultured in LB containing kanamycin at 37 °C with constant shaking until an OD₆₀₀ of 0.8 was reached. Then, protein production was induced by the addition of 0.5 mM IPTG (isopropyl β-D-1 thiogalactopyranoside). Bacterial growth continued for 4 h at 30 °C, then cells were harvested by centrifugation. Cells were resuspended in lysis buffer (5 mM imidazole, 150 mM NaCl, 20 mM Tris pH 7.5) and lysed by sonication followed by centrifugation to remove cell debris. Recombinant proteins in the cleared cell lysates were purified by affinity chromatography using Ni-sepharose (GE-healthcare). The eluted proteins were further purified by Superdex 200 size exclusion chromatography using an NGC Chromatography System (Biorad). PrsA2 was eluted in buffer containing 300 mM NaCl, 5 mM EDTA, 5% glycerol, and 20 mM Tris pH 7.5, while the other proteins were eluted in PBS. Protein concentration was determined using the Pierce BCA Protein Assay kit (Thermo Fisher Scientific).

Antigen-Specific Antibody ELISA.

Nunc MaxiSorp plates (Thermo Scientific) were coated overnight at 4 °C with 100 µL of recombinantly expressed antigens (10 µg/mL in PBS). Plates were blocked for 2 h at room temperature with 200 µL of 2.5% skim milk in PBS and then washed three times with PBS before the addition of mice sera (diluted 1:250 in PBS). Samples were incubated at room temperature for 2 h and then washed three times with tPBS. Next, plates were incubated for 1.5 h at room temperature with goat anti-mouse IgG-HRP (BioRad) diluted 1:4000 in 1% BSA in PBS. Plates were washed three times with PBS before developing with the TMB substrate reagent set (BD OptEIA) as specified before.

Human plasma samples from patients with NSTI, collected at the time of hospital admission (56), were diluted 1:5 in 1% BSA in PBS and added to plates coated with recombinantly expressed proteins (10 µg//mL). After overnight incubation at 4 °C, plates were washed three times with tPBS and incubated 2 h with goat anti-human IgG-HRP (Chemicon international) diluted 1:1000 in 1%BSA in PBS. After three final washes, samples were developed with the TMB substrate reagent set (BD OptEIA).

Tissue Processing.

Spleen and lungs were collected from mice immunized with GAS EVs/alum or PBS/alum. Lungs were diced into small pieces and digested with DNase (80 U/mL, Sigma) and liberase (0.8 U/mL, Sigma) in Hanks’ Balanced Salt Solution (Sigma) for 45 min at 37 °C. Digestions were stopped by the addition of lungs buffer (2 mM EDTA, 2% FBS in PBS). Single lungs and spleen-cell suspensions were prepared by homogenizing tissue through a 70 µm cell strainer. Red blood cells were lysed with RBS lysis buffer (BD Pharm). After washing with lungs buffer, cells were resuspended in restimulation media (RPMI supplemented with 10% FBS, 1% nonessential amino acids, 1% sodium pyruvate, 1% L-glutamine, 0.1% 2-mercaptoethanol, and 1% penicillin/streptomycin). Cell counts and viability were assessed by trypan blue staining. Single cell suspensions were used for flow staining or ex vivo stimulation (see below).

Flow Cytometry.

For intracellular cytokine analysis lung cells were stimulated with PMA (Phorbol 12-myristate 13-acetate, 30 ng/mL), ionomycin (0.5 µg/mL), and Brefeldin A (1X) for 3 h at 37 °C in a 5% CO₂ atmosphere. Samples were then stained with antibodies against surface markers including CD45 (30-F11), CD4 (RM4-5), TCRb (H57-597), CD44 (IM7) in the presence of an Fc Block (Biolegend and ThermoFisher). For intracellular staining, samples were permeabilized with FOXP3/Transcription factor staining buffer (Invitrogen) for 30 min before adding antibodies against IL-17A (TC11-18H10), IFNγ (XMG1.2), RORγT (B2D) and T-bet (4B10). Samples were acquired on BD Fortessa or Sony ID700 and analyzed using FlowJo software (Tree Star). Zombie ultraviolet light fixable dye (Biolegend) was used to exclude dead cells and FMOs were used as gating controls.

Ex Vivo Stimulation and Cytokine Quantification.

A total of 2 × 10⁶ lung cells or splenocytes were stimulated with 0.5 µg of GAS EVs, 1 µg of recombinantly expressed proteins or with media alone. Cells were incubated at 37 °C with 5% CO₂ for 72 h in restimulation media (see above). After centrifugation, culture supernatants were collected and stored at −20 °C before cytokine analysis.

IL-17A and IFNγ secreted into the culture supernatant was quantified using the mouse IL-17A (homodimer) or IFNγ uncoated ELISA kits (Invitrogen) following the manufacturer’s instructions. Data are normalized to the media stimulation controls and represented as mean ± SEM.

Bioinformatic Analysis.

To assess amino acid conservation of MtsA, BMP, PrsA1, PrsA2, MetQ, GlnP, and MalX across different GAS strains (representing different emm types), we performed TBLASTN v2.4.0 (57) using the full-length amino acid sequence of each protein as query against 1,340 genome assemblies from GenBank, as compiled by Davies et al. (58). Thresholds were E-value ≤1e−6, % identity ≥40%, and coverage ≥0.70 (alignment length/query length), for each assembly and gene the top bitscore hit was retained. For every gene, the mean pIdent (%) was calculated per M-Family, and a heatmap was generated using Matplotlib v3.8.0 (59).

Statistical Analysis.

Statistical analyses were performed using GraphPad Prism 5. Results are represented as the mean ± SEM. For pairwise comparisons of samples that follow normal distribution (Shapiro–Wilk test), a two-tailed unpaired t test was performed. Mann–Whitney test was performed when samples did not follow a normal distribution. P ≤ 0.05, was considered significant (*P ≤ 0.05, **P ≤ 0.01, ***P ≤ 0.001, ****P ≤ 0.0001; ns, not significant).

Supplementary Material

Appendix 01 (PDF)

Dataset S01 (XLSX)

Dataset S02 (XLSX)

Dataset S03 (XLSX)

Data, Materials, and Software Availability

All study data are included in the article and/or supporting information.