The S. pyogenes HA capsule promotes nasopharyngeal infection and enhances a cytokine response that supports the recruitment of neutrophils

In order to evaluate the influence of the HA capsule during experimental infections, we used the pG⁺host5 integration plasmid [32] (Table 1) to generate a markerless 1,212-bp in-frame deletion of hasA in the M18 serotype rheumatic fever isolate S. pyogenes MGAS8232 [33] (Table 1). M18 serotypes are well known for being highly encapsulated, a phenotype that has been traced to the mutations within the RocA regulatory protein [34]. The isogenic nature of the hasA mutant was determined by PCR and DNA sequencing with primers that flanked the deleted hasA region (S1 Table) and whole genome sequencing analysis. The correct ΔhasA deletion was confirmed, but compared to the wildtype MGAS8232 strain, two non-synonymous single nucleotide polymorphisms (SNPs) within pstI gene encoding the cytosolic protein enzyme I in the phosphoenolpyruvate phosphotransferase system (PTS) were also identified and resulted in two amino acid substitutions (Leu₁₉₄Phe and Ser₃₀₆Phe) (S2 Table). Thus, we generated a complementation strain using the pDCerm plasmid [35] (S1 Table) by expressing the hasA gene and its native hasA promoter in the ΔhasA mutant background (MGAS8232 ΔhasA + hasA). As predicted, MGAS8232 ΔhasA lost the large mucoid colony phenotype on sheep blood agar compared to wildtype MGAS8232, and capsule production was restored in the hasA-complemented strain (Fig 1A).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 1. Hyaluronic acid expression by Streptococcus pyogenes promotes nasal infection in B6_HLA mice.

(A) S. pyogenes constructs streaked onto TSA + 5% sheep blood agar plates. The plate figure are representative images of the wild-type MGAS8232, ΔhasA mutant and the hasA complemented strain. (B) B6_HLA mice were administered ~1×10⁸ CFUs of S. pyogenes MGAS8232 wildtype, ΔhasA, or ΔhasA +hasA strains intranasally and sacrificed 48 h later. Data points represent CFUs from cNTs of individual B6_HLA mice. Horizontal bars represent the geometric mean. Significance was determined by Kruskal Wallis one-way ANOVA with Dunn’s multiple comparisons test (****, P < 0.0001; **, P < 0.01). The horizontal dotted line indicates the theoretical limit of detection. (C) Heat-map of cytokine responses in cNTs of B6_HLA mice during S. pyogenes infection. Data shown represent normalized median cytokine responses from cNTs (n ≥ 3 per group). (D) Immunohistochemistry of infected cNTs at 24 h post-infection with wildtype S. pyogenes MGAS8232 and the ΔhasA mutant. Sections were stained with α-S. pyogenes (red), αB220 (blue), αCD3 (green), and αLy6G (white) antibodies. Panels are a close-up view from the boxed section. Arrows indicate regions with internalized S. pyogenes.

https://doi.org/10.1371/journal.ppat.1011013.g001

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Bacterial strains and plasmids used in this study.

https://doi.org/10.1371/journal.ppat.1011013.t001

The human-specific tropism of S. pyogenes represents a major challenge when conducting experimental infection models. We previously demonstrated that the use of transgenic mice that express human MHC class II molecules (herein referred to as B6_HLA mice) greatly enhances S. pyogenes nasopharyngeal infection due to the selective specificity of superantigens for human MHC class II molecules [36]. Using this infection model, we examined the influence of HA capsule expression on bacterial burden during acute nasopharyngeal infection through nasal inoculation (~1×10⁸ CFUs) of B6_HLA mice using the three MGAS8232 strains. The ΔhasA mutant resulted in a >1000-fold reduction in bacterial CFUs from the nasal mucosa at 48 hours compared to infection by wildtype S. pyogenes MGAS8232 (Fig 1B). We confirmed that capsule expression was specifically required for the nasopharyngeal infection phenotype as the complemented strain with restored capsule expression phenocopied the wildtype infection (Fig 1B). Consistent with the non-invasive nature of the model [36], mean bacterial dissemination of wildtype S. pyogenes remained below the limit of detection in the lungs, liver, spleen, heart, and kidneys (S1 Fig). We conclude that HA capsule expression improves experimental S. pyogenes nasopharyngeal infection in B6_HLA mice, and removal of the capsule does not increase bacterial dissemination of S. pyogenes MGAS8232.

To further assess the nasopharyngeal environment during wildtype S. pyogenes and ΔhasA infections, we conducted a multiplex cytokine and chemokine array using infected nasal turbinate homogenates [37]. Quantitative data is shown in supplemental figures (S2 Fig) and summarized as a heatmap showing normalized cytokine responses for any cytokine with an average concentration above 20 pg ml^-1 within a treatment group (Fig 1C). Uninfected mice demonstrated no apparent inflammatory signature from the complete nasal turbinate (cNT) homogenates whereas robust cytokine responses were evident in wildtype-infected mice and included: Th1-type cytokines (IL-1α and IL-1β); Th17-type cytokines (IL-6 and IL-17); chemokines (KC, IP-10, MCP-1, MIP-1α, MIP-1β, MIG, MIP-2, LIF and LIX); and growth factors (G-CSF) (Figs 1C and S2). In contrast, ΔhasA-infected cNTs presented a reduced inflammatory signature and revealed similar cytokine expression profiles as uninfected mice. Particularly, reductions were detected with pro-inflammatory cytokines, such as IL-1β, IL-6, and IL-17, and those involved in monocyte and neutrophil recruitment, including KC, IP-10, MCP-1, MIP-1β, MIG, and G-CSF (Figs 1C and S2). Restoring capsule expression in the ΔhasA mutant background induced a moderately inflamed environment, trending for greater concentrations of KC, MIP-1α, MIG, MIP-2, and G-CSF compared to ΔhasA mutant infections. Interestingly, IL-2, IL-12 (p40), IL-15, and IL-9 concentrations were higher in cNTs challenged with ΔhasA mutant compared to wildtype or ΔhasA + hasA strains; however, concentrations for these cytokines did not drastically differ from uninfected murine cNTs (S2 Fig). There were also numerous cytokines that did not have average concentrations above 20 pg ml^-1 in any treatment groups, were not different between treatment groups, or did not conform to any obvious trends (S2 Fig). These cytokine trends suggest that at 48 hours post-nasal inoculation with S. pyogenes, HA capsule expression is associated with higher concentrations of cytokines and chemokines that support inflammation and monocyte and neutrophil function.

To gain further insight into the interaction between S. pyogenes and host immune cells during nasopharyngeal infection, wildtype MGAS8232 and ΔhasA-infected cNTs were harvested and cryopreserved for immunohistochemistry. Due to the possibility that ΔhasA mutant could be completely cleared by 48 hours post-infection (Fig 1B), infected cNTs were collected at 24 hours post-infection and sections were stained with α-S. pyogenes (red), α-B220 (blue), α-CD3 (green), and α-Ly6G (white) fluorescent antibodies where indicated. Sections revealed that S. pyogenes was present with robust α-Ly6G neutrophil signals in both wildtype and ΔhasA-infected cNTs (Fig 1D). By 24 hours, neutrophils infiltrated to the central areas where S. pyogenes resided, whether the HA capsule was expressed or not (Fig 1D). Although few differences in immune cell percentages have been detected within the cNTs of wildtype-infected B6_HLA mice by 48 hours [36], infected nasal passages have demonstrated increased trends for neutrophil populations (GR1⁺) during wildtype S. pyogenes MGAS8232 infection [36], entirely consistent with these immunofluorescence experiments. Notably, S. pyogenes ΔhasA, but not wildtype, was detected within the epithelial cell layer in some areas (Fig 1D, denoted by arrows in the right panel). Together, these observations are consistent with a model whereby neutrophils are recruited to murine cNTs by 24 hours post-nasal infection with either wildtype MGAS8232 or ΔhasA strains; however, by 48 hours ΔhasA mutants are rapidly cleared and cannot remodel the nasopharynx to express favourable inflammatory responses.

The hyaluronic acid capsule blocks S. pyogenes invasion of pharyngeal cells and protects against killing by neutrophils in vitro

If S. pyogenes successfully escapes mucocilliary clearance, it proceeds to target and adhere to the underlying epithelial surface. Multiple studies have described the HA capsule as an important adhesin [19,20], and we next sought to evaluate and compare adherence capabilities of wildtype S. pyogenes MGAS8232 with its unencapsulated mutant. Collagen type IV and fibronectin make up a significant portion of the nasopharyngeal ECM, and therefore, bacterial binding to these structures may contribute to streptococcal infection [38–41]. Collagen type IV is the primary component of the ECM basement membrane that underlays epithelial cells, and fibronectin, while only a minor component, is frequently secreted to mediate adhesion and migration of host cells [42,43]. Bacterial binding was assessed by inoculating S. pyogenes onto wells pre-coated with either collage type IV or fibronectin. A decline in both fibronectin binding and collagen type IV binding was observed for the ΔhasA mutant (Fig 2A and 2B) demonstrating that under in vitro growth conditions, S. pyogenes HA capsule can likely adhere to the ECM through interactions with both collagen type IV and fibronectin.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 2. The S. pyogenes HA capsule inhibits host cell invasion but promotes survival from neutrophil-mediated killing.

Binding of S. pyogenes to wells pre-coated with 1 μg of human ECM components (A) fibronectin and (B) collagen type IV. (C) Adhesion of S. pyogenes to D562 pharyngeal epithelial cells. Confluent cell monolayers were cultured with S. pyogenes (MOI of 100) for 2 h at 37°C + 5% CO₂. Cells were washed with PBS and lysed with Triton X-100 for enumerating remaining adherent bacteria. (D) Internalization of S. pyogenes into D562 cells. Confluent D562 cells were cultured with S. pyogenes (MOI of 100) for 2 h at 37°C + 5% CO₂ followed by 1 h in media supplemented with 100 μg mL^-1 of gentamycin. Bars represent mean CFUs ± SEM and each dot represents a biological replicate. Statistical differences were evaluated by unpaired t-test (A–C) (**, P < 0.01; ****, P < 0.0001) or (D) one-way ANOVA (*, P < 0.05; ****, P < 0.0001). (E) Whole human blood survival assay. Heparinized blood from human donors were inoculated with ~10³ CFUs of S. pyogenes MGAS8232 at 37°C with rotation for 3 h. Data points represent geometric mean CFUs ± SD at each timepoint (n ≥ 3). Statistical significance was determined using one-way ANOVA with Friedman test (***, P < 0.001). (F) Neutrophil survival assay. Neutrophils were isolated from human blood by density centrifugation and inoculated with opsonized S. pyogenes at a MOI of 10. Surviving bacteria were enumerated after 60 mins at 37°C with rotation and calculated as the difference between survival in the no neutrophil control and in the presence of neutrophils. Each data point represents S. pyogenes CFUs from an individual donor. Data shown are the means of percent survival ± SD. Statistical analyses were performed using one-way ANOVA with Kruskal-Wallis test (*, P < 0.05).

https://doi.org/10.1371/journal.ppat.1011013.g002

Bacterial adherence was also evaluated using the pharyngeal cell line Detroit-562 (D562) due to its similarity of surface molecules with non-transformed pharyngeal cells [43], and its ability to induce streptococcal superantigens and DNAses that are otherwise weakly expressed by S. pyogenes [44,45]. There was a slight but statistically significant increase in the amount of S. pyogenes that adhered to D562 cells when HA capsule expression was absent (Fig 2C). Despite the ability to bind collagen type IV and fibronectin, these results conflict with reports indicating the HA capsule contributes substantially to bacterial adhesion properties, and suggests that the HA capsule may function in part to mask adhesins on the bacterial cell wall and obstruct adherence, at least with S. pyogenes MGAS8232 [15,20]. Next, we aimed to characterize the capacity of these constructs for epithelial cell internalization. We found a dramatic ~1000-fold increase in ΔhasA mutants recovered from lysed D562 cells following gentamycin treatment (Fig 2D). Given this dramatic phenotype, we also evaluated the capsule complemented strain (ΔhasA + hasA), which completely lost the invasion phenotype (Fig 2D). These data demonstrate that HA capsule expression by S. pyogenes MGAS8232 inhibits adhesion and represses internalization into pharyngeal epithelial cells.

Following attachment to epithelial cell surfaces, a critical mechanism during early colonization stages is to evade host immune responses. S. pyogenes MGAS8232 resistance to bacteriolysis was investigated to determine if the HA capsule improves immune evasion. Compared with wildtype S. pyogenes MGAS8232, growth and survival in whole human blood was markedly attenuated (~3 logs) in the absence of capsule expression (Fig 2E). Upon earlier histological assessment of infected nasal turbinates, neutrophils had accumulated in regions surrounding both wildtype S. pyogenes and ΔhasA at 24 hours post-infection, yet the ΔhasA strain had significantly less bacterial burden by 48 hours (Fig 1B and 1C). Furthermore, a significant decline in cytokines and chemokines involved in recruiting, modulating, and activating neutrophils were detected by 48 hours with ΔhasA infection (Figs 1C and S2). Therefore, we next sought to examine if the HA capsule resists neutrophil activity specifically. Unencapsulated bacteria were more susceptible to neutrophil-mediated killing demonstrated by a significant decline in ΔhasA mutants that survived in the presence of freshly isolated human neutrophils (Fig 2F). Indeed, the reduced bacterial survival in each condition was rescued with complementation of capsule expression in the MGAS8232 ΔhasA + hasA strain (Fig 2E and 2F). These results confirm an important role for the capsule in promoting resistance to killing by neutrophils, and thus, may provide a protective role against innate immune responses during early stages of acute infection.

Depletion of neutrophils restores S. pyogenes ΔhasA bacterial load during nasopharyngeal infection

Preventing opsonophagocytic bacterial clearance is one of the main proposed mechanisms for the HA capsule and has been repeatedly investigated using various in vitro bacterial survival assays [16,25,26]. Since neutrophil influx is a major feature of our experimental nasopharyngeal model and during natural infections [46], we aimed to explore the importance of neutrophils during nasopharyngeal infection and determine whether preventing phagocyte-mediated killing is a key molecular process by which HA capsule functions in this model. For this purpose, mice were depleted of neutrophils by administering the αLy6G monoclonal antibody (Fig 3A), which effectively depletes neutrophils from the peripheral blood of mice [47], with rat IgG2a used as an isotype control. To confirm the neutrophil depletions, we assessed innate immune cell populations from both the nasal passages and blood at 48 hours in wildtype MGAS8232 infected mice. This protocol successfully eliminated virtually all Ly6G⁺ cells from both sites (Fig 3B) and did not alter macrophage, monocyte or dendritic cell populations in the nasal passage (Fig 3C), however, a small but statistical increase in monocyte population was observed in the blood (Fig 3D). We next examined the effect of depleting neutrophils on nasal challenges with wildtype MGAS8232 or ΔhasA strains at both 24- and 48-hours post-infection. Although there were trends for reductions in CFUs, there were no statistical differences in the amount of wildtype S. pyogenes recovered from neutrophil depleted mice compared to control mice at either 24- or 48-hours post-infection (24 h, p = 0.5802; 48 h, p = 0.2803), indicating that wildtype S. pyogenes MGAS8232 infection is not impacted by neutrophil depletion in this model (Fig 3E). As expected, isotype treated mice showed a reduction for the ΔhasA mutant at both 24- (p < 0.05) and 48-hours (p < 0.0001) compared to wildtype-infected control mice (Fig 3E). While there was no difference in ΔhasA mutant recovery between control and neutrophil depleted mice at 24 hours post-infection (p = 0.9808), ΔhasA mutant burden substantially increased (~3 logs) at 48 hours post-infection in neutrophil depleted mice compared to control mice (p < 0.05). By 48 hours post-infection, no statistical differences were observed between the amount of ΔhasA mutants recovered from neutrophil depleted mice and control mice receiving wildtype S. pyogenes (p = 0.0989). Overall, these findings suggest neutrophil-mediated clearance mechanisms contribute substantially to the lower burden of unencapsulated S. pyogenes in the nasopharynx.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 3. Early clearance of the HA capsule-deficient mutant from murine nasal turbinates is due to enhanced susceptibility to neutrophil-mediated killing.

(A) Schematic outline for in vivo depletion of neutrophils with injections of 250 μg (500 μg total) of αLy6G or isotype control rat IgG2a 24 h prior to and 24 h post-intranasal challenge with 10⁸ CFUs of S. pyogenes wildtype or ΔhasA mutant strains. (B) Representative flow cytometric analyses of nasal and blood innate immune cells from the neutrophil depletion experiments at 48 h. Flow plots show live cells that were negative for CD4, CD45R and CD19, and gates were set on Ly6G⁺ and F4/80^- cells for neutrophils, and Ly6G^- and F4/80⁺ for macrophage populations. Percentage of innate immune cell populations from either nasal cell extracts (C) or blood (D) for the indicated treatment groups as percentage of live cells. Data points represent individual mice and the bars represent the mean. Significance was determined by Mann-Whitney test (*, P < 0.05) (E) Neutrophil effects on S. pyogenes survival in the nasopharynx. Data points represent CFUs from cNTs of individual mice 24 and 48 h post-infection. Horizontal bars represent the geometric mean. The horizontal dotted line indicates limit of detection. Significance was determined by two-way ANOVA with Tukey’s multiple comparisons (*, P < 0.05; **, P < 0.01; ****, P < 0.0001). Fig 3A was created using Biorender.com.

https://doi.org/10.1371/journal.ppat.1011013.g003

Bacterial burden and lesion pathology during skin infection are enhanced by S. pyogenes hyaluronic acid capsule expression

Pharyngeal colonization by S. pyogenes is believed to be the major reservoir for this pathogen in developed countries, yet skin infections (impetigo) tend to be more prevalent in resource-poor settings [48]. Since nasopharyngeal infection by S. pyogenes MGAS8232 was notably compromised by the loss of the HA capsule, we next performed a novel skin infection model to further assess whether capsule expression could promote experimental skin infection. To address this, B6_HLA mice were intradermally injected in each hind flank with 2.5×10⁷ CFUs of wildtype, ΔhasA, or ΔhasA + hasA strains. There was a ~10% decline in the weights of mice infected with wildtype S. pyogenes, a striking contrast to mice infected with the ΔhasA strain that gained weight over the 72 hour infection period (Fig 4A). The ΔhasA mutant strain revealed a clear reduction in virulence through considerably smaller lesions and less inflamed tissue over the infection period compared to wildtype-infected mice (Fig 4B and 4D). Significantly less bacterial CFUs were also recovered from each ΔhasA-infected lesion compared to wildtype-infected lesions (Fig 4C). Though weight loss and lesion sizes were only partially restored in ΔhasA + hasA-infected mice, bacterial CFUs recovered were fully complemented and did not differ from wildtype-infected mice (Fig 4B–4D). The moderate restoration of weight loss and lesion sizes in ΔhasA + hasA-infected mice may be due to the incomplete complementation of the hasA gene expressed in trans using the pDCerm plasmid, as erythromycin was not used topically to maintain plasmid expression and replication over the 72 hour infection period. This is supported by the observation of some capsule deficient reversion colonies on plated skin homogenates. Overall, these data demonstrate that the expression of the HA capsule in S. pyogenes supports virulence during acute skin infections in B6_HLA mice.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 4. Hyaluronic acid capsule expression by Streptococcus pyogenes promotes skin infection in B6_HLA mice.

B6_HLA mice were administered ~5 × 10⁷ CFUs of wildtype S. pyogenes MGAS8232, or the hasA mutant, or the ΔhasA + hasA complemented strain by intradermal injections in each hind flank. (A) Weights of B6_HLA mice at 24, 48, and 72 h following skin challenge. Data is represented as a percentage of day 0 weight. Data points represent the weight means ± SEM (n ≥ 5). (B) Skin lesion areas of mice at 24, 48, and 72 h after skin challenge. Data points represent individual lesion areas (2 per mouse), and the bars represent the mean. Significance was determined by two-way ANOVA with Geisser’s Greenhouse correction and Dunnett’s multiple comparisons test (*, P < 0.05, **, P < 0.01, ***, P < 0.001;) for panels A and B. (C) Data points represent CFUs from individual infected skin lesions from mice at 72 h. Horizontal bars represent the geometric mean. Significance was determined by one-way ANOVA with Kruskal-Wallis test (****, P < 0.0001; **, P < 0.01). The horizontal dotted line indicates the theoretical limit of detection. (D) Representative skin lesion images from B6_HLA mice 72 h following skin challenge.

https://doi.org/10.1371/journal.ppat.1011013.g004

Depletion of neutrophils recovers S. pyogenes ΔhasA bacterial load during skin challenge

To examine whether a lack of neutrophils would similarly enhance infection by unencapsulated S. pyogenes in the skin, B6_HLA mice were depleted of neutrophils as described above and challenged with subdermal infections with 2.5×10⁷ CFUs wildtype or ΔhasA S. pyogenes MGAS8232. Irrespective of depletion status, mice infected with the ΔhasA mutant revealed significantly less weight loss and considerably smaller lesions compared to wildtype-infected mice over the infection period (Fig 5A, 5B and 5D). Neutrophil depletion did not impact weight loss, lesion sizes, or the amount of wildtype S. pyogenes CFUs retrieved from each infected lesion (Fig 5A–5D). As expected, control mice receiving the isotype antibody showed a significant reduction of ΔhasA mutants recovered within lesions at 72 hours (p < 0.0001) compared to wildtype-infected control mice (Fig 5C). In contrast, neutrophil depleted mice displayed a sharp increase in ΔhasA CFUs recovered from each lesion compared to control mice that received infections with the ΔhasA mutant (p < 0.001), despite presenting similar lesion sizes (Fig 5C and 5D). Overall, these data support that S. pyogenes HA capsule expression is important for resisting bacterial killing by neutrophils during experimental skin infection.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Fig 5. The hyaluronic acid capsule is important for resisting neutrophil-mediated killing in the skin.

B6_HLA mice were administered ~5 × 10⁷ CFUs of wildtype S. pyogenes MGAS8232 or the ΔhasA or ΔhasA + hasA complemented strain intradermally in each hind flank. Mice received αLy6G or rat IgG2a isotype control antibodies intraperitoneally 24 h preceding and 24 h after skin infections. (A) Weights of B6_HLA mice at 24, 48, and 72 h following S. pyogenes skin challenge. Data is represented as a percentage of day 0 weight. Data points represent the weight means ± SEM (n ≥ 3). (B) Skin lesion areas of mice following at 24, 48, and 72 h after skin challenge. Data points represent individual lesion areas (2 per mouse), and the bars represent the mean. (C) Data points represent CFUs from individual infected skin lesions from mice at 72 h. Horizontal bars represent the geometric mean. The horizontal dotted line indicates limit of detection. Significance was determined by two-way ANOVA with Geisser’s Greenhouse correction and Dunnett’s multiple comparisons test (*, P < 0.05, **, P < 0.01, ***, P < 0.001;) for panels A and B, or one-way ANOVA with Kruskal Wallis (***, P < 0.001; ****, P < 0.0001) for panel C. (D) Representative skin lesion images from B6_HLA mice 72 h following skin challenge.

https://doi.org/10.1371/journal.ppat.1011013.g005

Ethics statement

Human venous blood was taken from healthy volunteer donors in accordance with human subject protocol 110859. The full study protocol was approved by the London Health Sciences Centre Research Ethics Board (University of Western Ontario, London, ON, Canada). Volunteers were recruited by a passive advertising campaign within the Department of Microbiology and Immunology at the University of Western Ontario, and following an outline of the risks, written consent was obtained from each volunteer before samples were taken. Following blood collection, samples were fully anonymized and no information regarding the identity of the donor, including sex and age, were retained.

All mouse experiments were conducted in accordance with the Canadian Council on Animal Care Guide to the Care and Use of Experimental Animals. The Animal Use Protocol (AUP) number 2020–041 was approved by the Animal Use Subcommittee at the University of Western Ontario (London, ON, Canada).

Bacterial strains, media, and growth conditions

Bacterial strains used in this study are listed in Table 1. The main bacterial model strain for our work is S. pyogenes MGAS8232, an M18 serotype and pharyngeal isolate from a patient with acute rheumatic fever [33]. S. pyogenes strains were grown statically in Todd Hewitt broth (BD Biosciences; Franklin Lakes, NJ, USA) supplemented with 1% (w/v) yeast extract (BD Biosciences) (THY) and 1 μg mL^-1 erythromycin when appropriate. For solid media preparation, 1.5% agar and/or 1 μg mL^-1 erythromycin were added to the media when applicable. Molecular cloning experiments utilized the E. coli XL1-Blue strain cultured in Luria-Bertani (LB) broth (Thermo Fisher Scientific, Waltham, MA, USA) aerobically at 37°C, or Brain Heart Infusion (BHI; BD Biosciences, Franklin Lakes, NJ, USA) media containing 1.5% (w/v) agar (Thermo Fisher Scientific). Media was supplemented with 150 μg mL^-1 erythromycin (Sigma-Aldrich Canada, Oakville, ON, Canada) as necessary. A complete list of plasmids used in this study can be found in Table 1.

Construction of recombinant S. pyogenes strains

References to genomic loci are based on the genome of S. pyogenes MGAS8232 [33]. In-frame genetic deletion in the hasA gene was generated using the Gram-positive-E. coli shuttle vector, pG⁺host5 [32] (Table 1). Using appropriate PCR amplification primers listed in S1 Table, the upstream and downstream regions flanking the hasA gene were amplified from the MGAS8232 genome. Allelic replacement of the wildtype hasA gene with the deletion mutant ΔhasA via homologous recombination was conducted as described previously [36]. For complementation of the hasA genetic deletion, DNA fragments containing the hasA open reading frame and its native promoter were amplified from the S. pyogenes MGAS8232 genome using complementation primers listed in S1 Table, and cloned into the XhoI and SpeI restriction sites of the plasmid pDCerm [35]. This construct (pDCerm::hasA) was electroporated into the MGAS8232 ΔhasA mutant strain to produce the hasA complementation strain (ΔhasA + hasA).

Genomic sequencing analysis

Genomic DNA preparations from S. pyogenes MGAS8232 wildtype and ΔhasA strains were sent for paired end Illumina sequencing at the John P. Robarts Research Institute sequencing facility (University of Western Ontario, London, Ontario). Illumina short-read sequence data were used to generate de novo assemblies using SPAdes v3.15 [69], which were annotated using Prokka v1.12 [70]. The assemblies for the isogenic MGAS8232 and MGAS8232 ΔhasA have been deposited at NCBI (Biosamples SAMN27522080 and SAMN27522081 [Bioproject: PRJNA825546]). Any sequence differences between the strains were determined using Snippy v4.1 (https://github.com/tseemann/snippy). The publicly available S. pyogenes MGAS8232 sequence was used as a reference (Bioproject: PRJNA286). SNPs unique to the ΔhasA mutant were reported (S2 Table).

Extracellular matrix binding assay

Corning Costar 9018 high-binding 96-well plates (Corning; Kennebuck, ME, USA) were coated with 1 μg of collagen type IV (Sigma-Aldrich) or fibronectin (Calbiochem, EMD Millipore Corporation; Temecula, CA, USA) dissolved in carbonate-bicarbonate buffer (0.2M sodium carbonate anhydrous, 0.2M sodium bicarbonate, pH = 9.6) and left overnight at 4°C. The following day, the plates were washed three times using PBS with 0.05% (v/v) tween-20 and blocked for two hours with 5% (w/v) skim milk at room temperature and then washed as described. Afterwards, 100 μl of bacteria containing 10⁷ CFUs were added in triplicate to pre-coated wells and left for 2.5 hours at 37°C. Plates were washed and fixed with 10% neutral buffered formalin (VWR International; Randor PA, USA) for 40 minutes and then washed again. Wells were incubated with 50 μl of 0.5% (w/v) crystal violet (Sigma-Aldrich) in 80% (v/v) sterile MilliQ water and 20% (v/v) methanol for 5 minutes at room temperature before being washed five times. Stain was solubilized in 5% (v/v) acetic acid with mild agitation for 10 minutes. Colorimetric analysis was measured at OD₅₉₀ using Synergy HTX Multi-Mode Microplate Reader (Biotek).

Human cell culturing

The Detroit-562 (ATCC CCL-138) human pharyngeal cell line was grown and maintained at 37°C in 5% CO₂ in minimal essential medium (MEM) and passaged every 2–3 days. All tissue culture basal media was supplemented with 10% (v/v) heat-inactivated fetal bovine serum (FBS; Sigma-Aldrich, St. Louis, MO, USA), 100 IU penicillin (Gibco, Life Technologies Inc, Carlsbad, CA, USA), and 100 μg mL^-1 streptomycin (Gibco-BRL, Life Technologies, Grand Island, N.Y.), all filtered through a 0.2 μm PES filter (Nalgene, Thermo Scientific, Waltham, MA, USA).

Epithelial cell adhesion and invasion

Detroit-562 cells were grown to confluence on 12 or 24-well TC-treated plates (Falcon, Corning). On the day of infection, cells were washed with PBS and maintained in their respective serum-free media for at least 1 h prior to infection. The average number of cells per well was calculated and used to determine the number of bacteria required for a multiplicity of infection (MOI) of 100. During this time, overnight streptococcal cultures were subcultured into pre-warmed media and grown to early exponential phase. Tissue culture plate wells were washed three times with PBS and inoculated with 500 μl of the resuspended bacterial dose per well and left for 2–3 hours at 37°C in a 5% CO₂ incubator to allow bacteria to adhere to the cells. Background adherence levels were measured by inoculating bacteria onto uncoated wells. Following incubation, wells were washed three times with PBS to remove non-adherent bacteria, and cell monolayers were lysed with 500 μl of 0.01% Triton X-100 for 5 minutes, followed by disruption of the wells by scraping with a 1 mL pipette tip. Solubilized wells were serially diluted 10-fold and plated onto TSA 5% sheep blood agar plates (BD Biosciences) to enumerate bacteria present.

For invasion experiments, bacterial were inoculated onto cell monolayers for 2–3 hours at 37°C with 5% CO₂ and washed as above, followed by the addition of media containing 100 μg mL^-1 of gentamicin for 1 hour at 37°C and 5% CO₂ to kill extracellular bacteria. Wells were then extensively washed to remove the gentamicin media. Cells were lysed, serially diluted, and plated as described above to estimate the number of intracellular bacteria.

Whole blood survival assay

Lancefield bactericidal assays were performed as previously described [71] with some modifications. Briefly, a volume of 10 μl containing 1000 bacterial CFUs was added to 990 μl heparinized whole human blood in 1.5 mL Eppendorf tubes and allowed to incubate at 37°C with rotation. After 30, 60, 90, 120, and 180 minutes, samples were serially diluted 10-fold and drop plated in triplicate onto 5% TSA blood agar plates to enumerate the surviving bacteria in whole blood at indicated time points.

Isolation of human polymorphonuclear neutrophils

Polymorphonuclear neutrophils (PMNs) were isolated using a Ficoll (GE-Healthcare) Histopaque (Sigma-Aldrich) density gradient centrifugation method. Briefly, heparinized blood from human donors was diluted with an equal volume of PBS (Wisent Bioproducts Inc.) and layered carefully onto a dual Ficoll-Histopaque gradient and centrifuged at 396 × g for 20 minutes without braking. The PMN layer was collected and washed with cold RPMI containing 0.05% human serum albumin (RPMI-HSA) and with addition of 1 mL ice-cold water to lyse residual erythrocytes. PMNs were then collected in RPMI-HSA following centrifugation and adjusted to 4 × 10⁶ cells mL^-1.

PMN survival assay

Overnight streptococcal cultures were grown to early exponential phase and diluted to 10⁴ CFU mL^-1 in RPMI containing 10% (v/v) normal serum for 30 minutes to assist with bacterial opsonization. A volume of 0.225 mL opsonized bacteria were co-cultured with 0.025 mL of isolated human PMNs at 4 × 10⁶ cells mL^-1 (1:10 bacterial CFUs to neutrophils) at 37°C with vigorous shaking. Control samples had no PMNs added to control for bacterial growth. Viable bacteria in each reaction mixture were measured after 60 mins by lysing cells with 750 μl of 0.025% Triton X100, followed by serial dilution and plating onto 5% TSA blood agar plates overnight at 37°C. Percent bacterial survival was calculated as average bacterial CFUs in the presence of neutrophils divided by bacterial CFUs in no PMN control samples.

Mice

Mice were bred in a barrier facility at the University of Western Ontario and genotyped routinely for appropriate transgene expression. Human MHC class II transgenic (B6_HLA) mice [72] were bred from McCormick laboratory colonies specifically for this study and were on a C57BL/6 (B6) background. During all breeding and experiments, mice were provided food and water ad libitum and appropriate enrichment was provided in all cages.

Nasopharyngeal infection model

Preparation of S. pyogenes MGAS8232 for nasal inoculation has been previously described [36,37,49]. Briefly, bacteria were grown to early exponential phase (OD₆₀₀ of 0.2–0.3), cells were centrifuged, washed, and resuspended in Hank’s buffered saline solution (HBSS) at ∼1 × 10⁸ CFUs per 15 μL. Mice were given 2 mg ml^-1 neomycin sulphate ad libitum in their drinking water two days prior to infection to reduce the nasal microbiota. Mice were anesthetized using FORANE (isoflurane, USP; Baxter Corporation; Mississauga, ON, Canada) and 7.5 μL of bacterial inoculum was administered into each nostril. Mice were sacrificed 24 or 48 hours post-infection and their complete nasal turbinates (cNTs), including the nasal-associated lymphoid tissue, nasal turbines, and maxillary sinuses, were extracted and homogenized using a glass homogenizer [37]. Murine lungs, liver, spleen, kidneys, and heart were also removed for bacterial enumeration where indicated. Organs were serially diluted in HBSS and plated on TSA supplemented with 5% sheep’s blood for bacterial enumeration. Counts less than 30 CFUs per 100 μL of tissue supernatant were considered below the theoretical limit of detection.

Skin infection model

The fur on the lower backs of B6_HLA mice between 8–12 weeks old was removed by shaving and hair removal cream the day prior to infection. S. pyogenes was grown and prepared as stated above and resuspended to 5 × 10⁸ CFU per mL in HBSS. Mice were anesthetized and a 50 μL dose containing ~2.5 × 10⁷ CFU was injected intradermally into each lower flank. On each day following infection, mice were weighed and lesions at the injection sites were measured using calipers. At 72 hours post-infection, mice were sacrificed and the skin around each injection site was harvested, homogenized, and plated on TSA with 5% sheep’s blood overnight at 37°C for bacterial enumeration. Bacterial burden was presented as CFUs from individual lesions.

Immunofluorescent histology

At the previously identified endpoint, cNTs were collected as described [37] and tissues were fixed in periodate-lysine-paraformaldehyde (PLP) and prepared for sectioning as previously [73]. Following fixation, cNT were passed through sucrose gradients and frozen in OCT (TissueTek) media. Serial sections (7 μm) were cut using a cryostat. Prior to staining, all slide-mounted tissue sections were blocked with PBS containing 1% BSA, 0.1% Tween-20, and 10% rat serum. Sections were stained with the following antibodies: anti-Streptococcus pyogenes Group A Carbohydrate (Abcam, ab9191), anti-B220 (Biolegend, RA3-6B2), anti-CD3 (Biolegend, 17A2), and anti-Ly6G (Biolegend, 1A8). After staining, sections were mounted with ProLong Gold Antifade Reagent (Invitrogen). Tiled images of whole cNT sections were collected using a DM5500B fluorescence microscope (Leica) at 10× and 20×.

Neutrophil depletion in vivo

The function of neutrophils during acute infections by S. pyogenes were examined in B6_HLA mice between 8–12 weeks old. Neutrophils were depleted in vivo by intraperitoneally injecting mice with 250 μg mAb αLy6G clone 1A8 (BioXcell, NH, USA) 24 hours before and 24 hours following nasopharyngeal and skin infections. Control mice received rat IgG2a clone 2A3 (BioXCell, NH, USA). Depletion of circulating neutrophils has been confirmed in previous studies using flow cytometry of Ly6G⁺ expressing populations in blood [47,74,75]. Bacterial burden in nasal turbinates was examined at both 24 and 48 hours following infection.

Flow cytometry analyses

To assess the neutrophil depletion experiments, we phenotyped innate immune cells from both the nasal tubinates and blood. Murine nasal turbinates were isolated as previously described [37] and collected in R10 media. Nasal turbinates were treated with 0.3 mg ml^-1 collagenase D (Sigma-Aldrich) in R10 media at 37°C for 30 minutes and then pushed through a 0.7-μm cell strainer. The single-cell suspension was then treated with ACK lysis buffer (Gibco) and washed with PBS containing 2% FBS. Blood was collected from mice via cardiac puncture. Following isolation, blood was treated with ACK lysis buffer (Gibco) and washed with PBS containing 2% FBS. Following isolation, both blood and nasal cells were stained and analysed as follows. Cell viability was first determined using Fixable Viability Dye eFluor506 (Thermo Fisher) and then subsequently stained anti-CD4-PE-Cy5 (clone RM4-5, Thermo Fisher) anti-CD45r(B220)-V450 (clone RA3-6B2, BD), anti-CD19-BV711 (clone 1D3, Thermo Fisher), anti-F4/80-A647 (clone BM8, Biolegend), anti-Ly6G-A700 (clone RB6-8C5, Biolegend), anti-Ly6C-PE (clone HK1.4, Biolegend), anti-CD11b-A488 (clone M1/70, Biolegend), and anti-CD11c-APC-Cy7 (clone HL3, BD). Cells were fixed overnight with 1% paraformaldehyde prior to analysis. Events were acquired using a LSR II (BD Biosciences), and data were analyzed using FlowJo v10.7.1 (TreeStar).

Cells were gated first on whether they were alive or dead, and the live population was subsequently gated for the singlet population. Single cells were checked for the expression of CD4, CD45r, and CD19 and the triple negative population was gated for further analysis. Cells were then assessed for expression of F4/80 and Ly6G. F4/80 positive and Ly6G negative cells were classed as macrophages. Cells that were subsequently high for Ly6G but negative for F4/80 were assessed for CD11b and Ly6C expression and classed as neutrophils. F4/80 and Ly6G double negative cells were gated and assessed for CD11b and Ly6C expression, with double positive cells classed as monocytes. Ly6G and F4/80 double negative cells were also gated and assessed for CD11c and Ly6C expression, with CD11c positive and Ly6C negative cells classed as dendritic cells. Data is presented as the percentage of live cells.

Detection of cytokines and chemokines in vivo

Cytokine and chemokine concentrations were determined from cNT homogenates of B6_HLA mice infected with S. pyogenes MGAS8232. Uninfected murine cNT homogenates were measured as a background control. Mutiplex cytokine array (Mouse Cytokine/Chemokine Array 32-Plex) was performed by Eve Technologies (Calgary, AB, Canada) and data on heat maps is presented as normalized median cytokine responses (X_normalized = [(x—x_min)/(x_max−x_min)]) from cNT homogenates.

Statistical analysis

All statistical analysis was completed using GraphPad Prism 9.3.1. Significance was calculated using the Student’s t test or one-way ANOVA with Dunnett’s or Tukey’s multiple comparisons post hoc test where indicated. A P value less than 0.05 was determined to be statistically significant.