We initially sought to demonstrate immunity in the respiratory tract following an intranasal (IN) infection. Cohorts of BALB/c mice received either one or two non-lethal IN infections with mouse-passaged S. pyogenes 2031 (emm1), each 3 weeks apart ( Figure 1A ). To assess whether the number of prior intranasal exposures would determine the level of site-specific protection, mice received a final IN challenge with homologous S. pyogenes isolate three weeks following a single (1x) or two (2x) sequential infections infections. Mice were closely monitored for the appearance of clinical symptoms and scored based on an approved clinical scoring system (25). All mice, regardless of the number of prior IN exposures (1× or 2×), showed a significant reduction (p <0.01-0.001) in clinical scores when compared to the control group, which was not infected prior to challenge ( Figure 1B ). On day 2 post-challenge, mice were euthanized to assess bacterial burden (colony forming units [CFU]) in the nasal-associated lymphoid tissue (NALT) and in the lungs. Following 1x or 2x prior exposures, all mice had significant reductions in bacterial burden in the NALT (66.6%–92.6%; p <0.05) and lungs (90%–100%; p <0.01) when compared to the control group (0×) ( Figures 1C, D ). IN challenge of these mice resulted in a very low to undetectable bacterial burden in lymphoid organs (including spleen and cervical lymph nodes (CLN)), which was comparable across all groups, with or without prior exposure ( Figure 1E ).

To assess the longevity of protection in the respiratory tract, a separate cohort of mice (also previously infected 1x or 2x with S. pyogenes) was rested for 9 weeks before receiving a homologous IN challenge ( Figure 1F ). We observed that 1x or 2x infections ( Figures 1A–D ) induced enduring immunity, which did not wane after a 9-week rest period. All mice previously infected with S. pyogenes demonstrated a significant reduction in clinical scores compared to control mice (p <0.001) ( Figure 1G ). This was associated with significantly reduced bacterial burden in the NALT (96.3%–99.4%, p <0.05–0.01) and lungs (93.8%–99%, p <0.05–0.01) ( Figures 1H, I , respectively). Interestingly, IN challenge in mice not previously infected led to bacterial dissemination to lymphoid organs only in the older naive mice ( Figure 1J ), which presented a higher burden in the CLN and spleen, than 9-weeks younger naive mice (Figure 1E) (>99.9%, p <0.05), suggesting that age was associated with reduced innate protection. All mice with prior S. pyogenes infection demonstrated complete protection against dissemination to the CLN and spleen, which was significantly reduced (>99.9%, p<0.05) compared to the control group (0x) ( Figure 1J ).

Overall, we found that a single IN exposure to S. pyogenes resulted in immunity in the respiratory mucosa. Furthermore, our data show age-related susceptibility to bacterial dissemination following S. pyogenes IN challenge; however, this was completely prevented by previous exposure to the organism.

To investigate the role of humoral immunity in mucosal protection, we measured the levels of M-protein type-specific IgG in the sera of mice following each sequential infection. S. pyogenes M1 type-specific IgG antibodies were not evident following a single IN infection ( Supplementary Figure 1A ). Nevertheless, M1-specific IgG titers developed after two infections and remained consistent in cohorts that received subsequent IN infections ( Supplementary Figure 1A ).

The fact that M1-specific circulating antibodies were not evident in mice after a single S. pyogenes infection prompted us to investigate other immune correlates of protection in the respiratory mucosa. Salivary antibodies play important roles in protecting against S. pyogenes by preventing bacterial attachment to the mucosal epithelia, opsonizing the bacteria, and directly lysing them (26, 27). To investigate URT mucosal immunity, we measured the levels of M1-specific IgG and IgA in the saliva of the mice following each infection. M1-specific IgG and IgA titers were significantly increased in the saliva after a single infection ( Figures 2Ai, ii , respectively). A progressive increase in M1-specific IgG and IgA responses in saliva of 2x infected mice was also noted. To investigate the long-term humoral immunity mediated by antibody-secreting cells in the bone marrow (BM) (known as long-lived plasma cells (LLPCs)) (28, 29), we counted the number of M1-specific IgG-secreting LLPC in naive and infected mice using an ELISpot assay. We found that the number of M1-specific IgG- and IgA-secreting BM cells was significantly higher in mice that received 2x IN infections than in naïve mice ( Figures 2Bi, ii , respectively).

Next, to explore the role of effector cell-mediated responses in immunity to S. pyogenes, we assessed specific cell populations in the lungs and spleens of sequentially infected mice using flow cytometry. Mice that received either 1x or 2x IN infections had a significant increase in Ly6G⁺ neutrophils in both the lungs ( Figures 2C, Di ) and spleen ( Figure 2Dii ) when compared to naïve mice. Similarly, an increase in effector/memory CD4⁺ T cells was observed in mice that received 1x or 2x infections prior to an IN challenge ( Figures 2E, F ).

To assess whether infection site-specific IL-17 secretion was associated with immunity in the URT, we measured IL-17 secreted by lung cells and splenocytes from mice that received 1x, 2x or 3x IN infections or remained as naive infections. Immune cells isolated from the lungs and spleens were stimulated ex vivo with heat-killed (HK) homologous S. pyogenes 2031 prior to the detection of IL-17 in the culture supernatant by ELISA. Mice that received 1x, 2x or 3x IN infections produced significantly higher (p <0.001) levels of IL-17 in the lungs than to naïve mice ( Figure 2Gi ); however, at least 2x IN infections were required for induction of IL-17 in the spleen ( Figure 2Gii ).

Taken together, these data show that mucosal immunity following a single S. pyogenes IN infection is associated with an increase in M1-specific IgG and IgA antibodies in the saliva, alongside the expansion of neutrophils, effector/memory CD4⁺ T cells, and IL-17⁺ cells in the lung.

To assess the importance of IL-17 in mucosal immunity against S. pyogenes IN infection, we compared IL-17 knockout (IL-17^−/−) mice with wild-type (WT) BALB/c mice receiving no (0x) or 2x IN infections prior to the final IN challenge ( Figure 3A ). IL-17^−/− mice sequentially infected prior to challenge demonstrated significant increases in bacterial burden in throat swabs (p <0.01), NALT (p <0.01), lungs (p <0.05), and lymphoid organs (p <0.01) compared to 2x infected WT mice ( Figures 3B–E ). Sequentially infected IL-17^−/− mice showed an increased bacterial burden in the NALT (1.71-fold increase, non-significant), throat swab (3.04-fold increase, non-significant), and lymphoid organs (4.53-fold increase, non-significant) compared to control mice (IL17-/- 0x) ( Figures 3B–E ). In contrast, sequentially infected WT mice showed a reduction in bacterial burden in NALT (p <0.05), lungs (non-significant), and throat swabs (p <0.001) ( Figures 3B–D ). As shown previously (Figure 1E), the young and immunocompetent mice did not develop systemic infection post-challenge ( Figure 3E ).

The higher susceptibility to S. pyogenes mucosal infection in IL-17-/- mice was due to impaired humoral and cellular responses. This is evidenced by the significantly lower (p <0.001) M1-specific IgG antibodies in the saliva of 2x infected IL-17^−/− mice compared to WT mice ( Figure 4Ai ). No difference in M1-specific salivary IgA levels was observed between the sequentially infected IL-17^−/− and WT mice ( Figure 4Aii ). The number of M1-specific IgG⁺ and IgA⁺ LLPC in the BM of IL-17^−/− mice that received 2x infections, was significantly lower than that in WT mice (p <0.05–0.01) ( Figures 4Bi, ii , respectively). In contrast to the reduction observed in WT mice, M1-specific IgG titers did not change in the sera of IL-17^−/− mice after the 2x infections ( Supplementary Figure 2A ). No changes in IgA responses were observed in the sera of either WT or IL-17^−/− mice ( Supplementary Figure 2B ). The role of IL-17 in cell-mediated immunity was assessed in the lungs and spleen of sequentially infected mice. IL-17^−/− mice that received 2x infections showed a significant decrease in Ly6G⁺ neutrophils in the lungs ( Figures 4C, Di ), paralleled by an increase in the spleen ( Figure 4Dii ), when compared to the respective WT control. This suggests that mice with intact IL-17 are able to recruit neutrophils from the spleen into the lungs to combat mucosal infection.

Taken together, these data show that the absence of IL-17 prevents the development of humoral and cellular immunity against S. pyogenes in the respiratory tract.

Next, we modeled a scenario in streptococcal endemic settings to assess whether sequential IN infections could provide cross-compartmental protection to the skin and prevent systemic dissemination to lymphoid organs. We have previously demonstrated that at least two sequential homologous skin infections are required to generate skin immunity against a homologous skin challenge (5). In the current study, WT mice were given either 2x or 3x homologous IN infections 3 weeks apart or left naïve (cohort 1). Three weeks after the last IN infection, all mice received a skin challenge with the homologous isolate 2031 ( Figure 5A ). Six days after the skin challenge, the mice in each cohort were euthanized to assess the bacterial burden in various tissues. We found that none of the mice that received 2x or 3x sequential IN infections were protected from the skin challenge ( Figure 5B ). This was evident from the bacterial burden observed in the skin lesions of sequentially infected mice, which was comparable to that observed in control mice ( Figure 5B ). To further investigate whether a higher number of prior IN exposures could induce cross-compartmental protection in the skin, a second cohort of mice (cohort 2) received 4x infections, each 3 weeks apart, following which, 3 weeks later, they received a skin challenge ( Figure 5C ). The mice showed a significant reduction (p <0.01) in the bacterial burden in the skin (95%) and lymphoid organs (95%) compared to control mice ( Figures 5D, E ).

Therefore, cross-compartmental protection against S. pyogenes infection is hard to achieve and requires at least four previous homologous exposures.

To understand whether skin and systemic immunity observed after repeated IN infections are associated with the induction of immune responses in the spleen, we analyzed splenic CD4⁺ T cells and neutrophils using immunohistochemistry (IHC) staining. No difference was observed between mice that received 1x or 3x IN infections ( Figures 6A, B ). However, when mice received 4x infections, a significant increase (p <0.05) in the number of neutrophils and neutrophil elastase H-score (a surrogate marker for neutrophil extracellular trap formation (30)) was observed in comparison to mice that received 1x infection ( Figures 6A–C ). While neutrophils accumulated in the perifollicular spaces, increased numbers of CD4⁺ T cells were observed in the splenic follicles of mice that received 4x infections ( Figures 6A, D ). Interestingly, a significant increase in IL-17 secretion by splenocytes stimulated ex vivo with HK S. pyogenes 2031 was observed after 3x IN infections (Figure 6E); however, it did not correlate with protection from skin challenge ( Figures 5B ). CD4⁺ T cell accumulation in splenic follicles led to the hypothesis that germinal center responses may play a role in cross-compartmental protection. To confirm this, we investigated plasma CXCL13 levels. CXCL13 is a B cell chemoattractant used as a plasma marker of germinal center activity (31) and has also been shown to play a role in mucosal immunity (32). We assessed the level of CXCL13 in mouse sera collected three days and three weeks after each homologous sequential IN infection ( Figure 6F ). A significant transient increase in serum CXCL13 levels was noted at 3 days post 2^nd, 3^rd, and 4^th IN infections (4.4- and 5.5-fold increase, respectively); however, at 3 weeks post 2^nd and 3^rd infection, CXCL13 levels had dropped to levels comparable to those in the naive sera. However, following the 4^th IN infection, significantly increased CXCL13 levels were maintained for at least until 3-weeks post infection ( Figure 6F ). Notably, in these mice, a significant increase in CXCL13 levels following 2nd infection also coincided with an increase in serum IgG ( Supplementary Figure 1 ). Although CXCL13 demonstrated a transient, albeit significant, increase following each IN infection, serum IgG titers remained consistent.

Taken together, repeated URT mucosal infections generated cross-compartmental immunity in the skin and spleen, which was not associated with changes in IL-17 levels but rather correlated with neutrophil infiltration and germinal center activity, as evidenced by increased levels of serum CXCL13 and expansion of CD4⁺ T cells in the splenic follicles.

Mice were housed at the Animal Facility of Griffith University (Gold Coast, Australia). All experiments and animal procedures were approved by the Griffith University Animal Ethics Committee (Animal Ethics Approval GLY/04/18), in compliance with the Australian National Health and Medical Research Council Guidelines. Experimental protocols involving IL-17 knockout (IL-17^−/−) mice were reviewed and approved by the Office of the Gene Technology Regulator (OGTR). BALB/c mice (female, 4–6 weeks old) were sourced from the Animal Resource Centre, Western Australia. IL-17^−/− mice were obtained from Yoichiro Iwakura (Tokyo University of Science, Japan) under a Material Transfer Agreement (MTA) (67). Knockout mice were bred in-house at the Griffith University Animal Facility. The general health of the mice was monitored daily. Following the challenge, mice were monitored for signs of illness as per a score sheet approved by the Griffith University Animal Ethics Committee (68). The observer was blinded to the experimental groups.

S. pyogenes 2031 (emm1) was obtained from the Menzies School of Health Research (Darwin, NT, AUS). The isolate was serially passaged in mice to ensure virulence and was resistant to 200 µg/ml of streptomycin. The isolate was grown overnight in liquid Todd–Hewitt broth medium (THB; Oxoid, AUS), supplemented with 1% yeast and 1% neopeptone (THBYN; Difco, AUS). The isolate was 10-fold serially diluted and plated in duplicate on Columbia Blood Agar (CBA; Oxoid, UK) supplemented with 5% defibrinated horse blood (Equicell, AUS) and 200 µg/ml streptomycin (Sigma, China) to determine the number of colony-forming units (CFUs).

M-protein was extracted using pepsin digestion as described previously (69, 70). Overnight THBYN cultures were pelleted and resuspended in four times the weighted volume of PBS pH 5.8 twice. Resuspended pellets were pre-warmed to 37°C for enzymatic digestion with pepsin A (Merck, AUS) (1 mg pepsin per 10 g bacterial suspension) for 45 min with intermittent mixing. The digested suspension was pelleted, and the pepM extract buffer was exchanged with PBS pH 7 using a 10 kDa Amicon centrifugal filter unit (Merck, Ireland). PepM extracts were confirmed by SDS-PAGE and stored in solution at −20°C.

The mice received sequential intranasal infections 3 weeks apart. Mice were anesthetized via an IP injection of ketamine 100 mg/kg and xylazine 20 mg/kg (71). Using a pipette, 5 µL of bacterial inoculum was administered to each nare (1 × 10⁷ CFU in 10 µL/mouse) while the mouse remained on its back to ensure inhalation. Mice were monitored daily as described above.

Mice were challenged using a skin scarification model as previously described (72). The mice were anesthetized with an IP injection of ketamine 100 mg/kg and xylazine 20 mg/kg. The fur from the neck nape was removed using clippers and then superficially scarified. A 10 µL inoculum was topically applied; once the inoculum had absorbed into the skin, a temporary cover (Band-Aid™) was applied to the wound, and mice were housed individually. The mice were monitored daily for signs of illness, as described above.

Indirect-ELISA was used to quantify antigen-specific IgG and IgA antibody titers as described elsewhere (73). Goat anti-mouse IgG (Bio-Rad, AUS) or IgA (Invitrogen) horseradish peroxidase (HRP) linked antibodies were used to detect antigen-specific antibodies. Optical density (OD) at 450 nm was measured using a Tecan Infinite m200 Pro plate reader. Titers were defined as the highest dilution of serum for which the OD was >3 standard deviations (SD) above the mean OD of the control samples (naïve sera).

Serum samples were collected by puncturing the submandibular vein. Whole blood was allowed to clot and be removed prior to centrifugation for serum separation.

Throat swabs were performed using floq swabs (Interpath, USA) moistened in PBS prior to swabbing both sides of the throat. Swabs were squeezed into tubes containing PBS, 10-fold serially diluted, and plated onto CBA plates supplemented with 5% horse blood and 200 µg/ml of streptomycin. Following overnight incubation at 37°C, the CFU were counted to determine the bacterial load.

At designated time points, the mice were sacrificed via CO₂ inhalation, and whole blood was collected via cardiac puncture into tubes containing ethylenediaminetetraacetic acid (EDTA; Thermo, AUS). Tissues were collected and mechanically homogenized using a Bullet Blender™ (Next Advance, USA) following the manufacturer’s instructions. Samples were 10-fold serially diluted and plated in replicates on CBA plates supplemented with 5% horse blood and 200 μg of streptomycin. Following overnight incubation at 37°C, CFU were counted to determine bacterial load.

ELISpot was used to quantify the number and location of antibody-secreting cells (ASC) in the splenocytes, lungs, and long-lived plasma cells (LLPCs) in the bone marrow. Multiscreen HA filter plates (Merck, Ireland) were coated with 5 μg/mL M1 extract in carbonate coating buffer overnight at 4°C. Isolated cells were adjusted to 5 × 10⁶/mL and directly tested for IgG/IgA-secreting cells using previously published methods (74, 75). Spots were developed using an AEC substrate kit (BD, AUS) according to the manufacturer’s instructions and manually counted to determine ASCs.

The CXCL13 assay was performed according to the manufacturer’s instructions (R&D Systems, USA). Test sera were diluted 1:4 and assay standards were prepared in a 1:1 ratio of assay diluent in duplicate and incubated at room temperature for 2 h on a shaker. The plate was washed five times before conjugate incubation at room temperature for 2 h on a shaker. After washing, the substrate solution was added for 30 min at room temperature in the dark prior to the addition of the stop solution. The OD was measured at 450 nm with a correction of 540 nm using a Tecan Infinite m200 Pro plate reader.

IL-17A ELISA was performed according to the manufacturer’s instructions (Mouse IL-17A ELISA MAX; Biolegend, USA). Nunc MaxiSorp plates (Thermo, AUS) were coated overnight with an IL-17A capture antibody diluted in carbonate coating buffer. The plates were blocked with 1% BSA/PBS for 1 h prior to washing. Splenocyte supernatants (diluted 1:4 in duplicate) and standards were incubated at room temperature for 2 h. The detection antibody was incubated for 1 h prior to avidin–HRP incubation for 30 min. TMB substrate solution was incubated for 20 min, followed by the addition of an acid stop solution. The absorbance (OD) was measured at 450 nm with a correction of 570 nm. IL-17A concentration was determined by plotting the unknown samples on a standard curve to determine IL-17A pg/mL.

Lymphocyte populations were purified from the splenocytes, lungs, and bone marrow. Lungs were digested in Worthington collagenase III (Scimar, AUS) supplemented with DNase I (Merck, AUS), and all tissues were transferred through a 0.70 μM cell strainer (Corning, USA) to obtain single-cell suspensions. RBCs were lysed with ACK lysis buffer and washed with RPMI (Thermo Fisher Scientific, AUS).

Single cell suspensions of splenocytes were prepared as described above, counted using a hemocytometer in trypan blue (Sigma, AUS) and adjusted to 4 × 10⁶ cells/mL. Splenocytes were stimulated with heat-killed 2031 or medium as a negative control. After 72-hour stimulation at 37°C with 5% CO₂, the cells were centrifuged at 300g for 10 min, and the supernatant was collected. The supernatants were stored at −80°C. Cytokines in the supernatant were subsequently measured using a mouse Th1/Th2/Th17 CBA kit (BD, AUS) according to the manufacturer’s instructions. Samples were acquired using a BD LSR Fortessa cytometer, and data were analyzed using FCAP Array v3.1 (BD).

The cell populations were determined by flow cytometry. Single cell suspensions were prepared as described above and pre-incubated with the Fc block (CD16/32) for 15 min on ice. The cells were surface-stained with a master mix containing dead cell exclusion dye (NIR), CD4-FITC, CD62L-PE-Cy7, CD44-APC, CD45-BUV395, CD11b-BUV737, CD3-PE-CF594, and Ly6G-BV510. Cells were stained on ice in the dark for 40 min. Following incubation, the cells were washed in FACS buffer (2.5% fetal calf serum, 5mM EDTA in PBS) and fixed in 2% paraformaldehyde for 15 min. Samples were washed in PBS and acquired on a BD LSR Fortessa flow cytometer, and data were analyzed using FlowJo V.10.7 (BD).

Formalin-fixed paraffin-embedded (FFPE) spleens were sectioned at 3 μm thickness on SuperFrost+ slides. IHC staining of spleen sections for CD4 (1:200, D7D2Z; Cell Signaling Technology) and neutrophil elastase (1:200, E8U3X; Cell Signaling Technology) was performed using a Leica BOND™ RX auto-stainer (Leica, Nussloch, Germany) with the BOND Polymer Refine Detection (Leica) kit and developed with 3,3’-diaminobenzidine as the chromogen. The stained slides were mounted in Dako Mounting Medium (Dako) and coverslipped using a Dako coverslipper.

Images were acquired using an Olympus VS200 digital slide scanner (EVIDENT Life Science, USA) under bright-field emission. Each tissue section was acquired using a ×20 objective (UPLXAPO ×20; NA 0.8). Individual images were obtained from regions of interest (ROI) drawn using the VS200 software based on automatic outline thresholding. Cell counts were performed using QuPath (v 4.0.1), and the neutrophil elastase H-score was assigned in the range of 0–300. Cell counts are given as the average of cells per µm2Data was exported to Microsoft Excel for tabulation and plotted using GraphPad Prism.

The data were analyzed using Graph Pad PRISM version 10.7 for Windows. All data, except where noted, are presented as the geometric mean ± standard error of the mean (SEM). Statistical differences between the two groups were determined using the non-parametric Mann–Whitney t-test. When comparing more than two groups, analyses were performed using one-way ANOVA with p <0.05 considered to be statistically significant.