Host-to-Host Group A Streptococcus Transmission Causes Infection of the Lamina Propria but Not Epithelium of the Upper Respiratory Tract in MyD88-Deficient Mice

ABSTRACT

To understand protective immune responses against the onset of group A Streptococcus respiratory infection, we investigated whether MyD88 KO mice were susceptible to acute infection through transmission. After commingling with mice that had intranasal group A Streptococcus (GAS) inoculation, MyD88^−/− recipient mice had increased GAS loads in the nasal cavity and throat that reached a mean throat colonization of 6.3 × 10⁶ CFU/swab and mean GAS load of 5.2 × 10⁸ CFU in the nasal cavity on day 7. Beyond day 7, MyD88^−/− recipient mice became moribund, with mean 1.6 × 10⁷ CFU/swab and 2.5 × 10⁹ CFU GAS in the throat and nasal cavity, respectively. Systemic GAS infection occurred a couple of days after the upper respiratory infection. GAS infects the lip, the gingival sulcus of the incisor teeth, and the lamina propria of the turbinate but not the nasal cavity and nasopharyngeal tract epithelia, and C57BL/6J recipient mice had no or low levels of GAS in the nasal cavity and throat. Direct nasal GAS inoculation of MyD88^−/− mice caused GAS infection, mainly in the lamina propria of the turbinate. In contrast, C57BL/6J mice with GAS inoculation had GAS bacteria in the nasal cavity but not in the lamina propria of the turbinates. Thus, MyD88^−/− mice are highly susceptible to acute and lethal GAS infection through transmission, and MyD88 signaling is critical for protection of the respiratory tract lamina propria but not nasal and nasopharyngeal epithelia against GAS infection.

INTRODUCTION

Streptococcus pyogenes or group A Streptococcus (GAS) is a major human pathogen that causes pharyngitis and skin infections (1). GAS can occasionally cause severe invasive infections (2). Among more than 200 M protein gene (emm) genotypes, the M1T1 clone of emm1 GAS that emerged in the 1980s has been globally disseminated (3, 4) and is responsible for about 20% of contemporary GAS pharyngitis and severe infections (2, 5). No licensed GAS vaccine is available despite substantial morbidity and global economic loss caused by GAS infections and the considerable efforts made to develop GAS vaccine over the last several decades. GAS pathogenesis is also not fully understood, hindering the development of a vaccine and therapies for treating severe invasive infections.

A variety of mouse and nonhuman primate models of GAS infections have been used to investigate GAS pathogenesis, gene expression or fitness, immune responses, and GAS vaccine candidates. Infections are induced in these models by inoculation of >10⁷ CFU GAS through intranasal, intraperitoneal, air sac, intratracheal, subcutaneous, and intramuscular routes (6–16). Characterization of pharyngeal GAS infection using murine models is limited in the literature. An invasive emm1 GAS isolate is primarily confined to the nasal-associated lymphoid tissue (NALT) after intranasal inoculation (7). A transmission model using FBV/NJ mice and an emm75 strain apparently leads to low levels of GAS shedding, indicating low levels of GAS colonization in recipient mice (15). No GAS transmission model is available for acute infection in recipient mice. Consequently, very little is known about immune protection and pathogenesis during the onset of acute pharyngeal infection.

Myeloid differentiation factor 88 (MyD88) is a central player in innate immune responses to GAS infection in mouse models of subcutaneous GAS infection (17, 18). We tested whether MyD88 knockout (MyD88^−/−) mice were susceptible for acute infection through transmission from mice with nasal GAS inoculation. We found that MyD88^−/− mice were highly susceptible to acute GAS infection in the nasal cavity and nasopharyngeal tract, GAS colonization in the throat, and lethal systemic infection through transmission. More interestingly, GAS infected the lamina propria but not the nasal cavity and nasopharyngeal tract epithelia of MyD88^−/− recipient mice after GAS transmission. The results indicate that MyD88 signaling is critical for protection of the respiratory tract lamina propria against GAS infection.

RESULTS

Recipient MyD88^−/− mice are susceptible to upper respiratory infection through transmission.

Transmission experiments involved cages of six mice/cage. Two starter mice in each cage were inoculated in the nares with GAS suspension and commingled with four recipient mice on day 0. Starter and recipient mice were euthanized on days 4 and 7, respectively. The throats of euthanized mice were swabbed to measure throat colonization, and part of the head containing the nasal cavity was homogenized to quantify GAS loads on blood agar plates. Starter mice were 1-month-old male and female C57BL/6J mice that were inoculated in the nares with 2.8 × 10⁸ CFU of MGAS2221 in one experiment and 3.4 × 10⁸ CFU of MGAS2221 in another experiment. Euthanized starter mice had median 8.8 × 10⁶ CFU/swab throat colonization and 2.5 × 10⁹ CFU GAS at the nasal cavity (Fig. 1A). C57BL/6 starter mice were noticeably sick on day 2, and half of them became moribund on day 4. Nineteen C57BL/6J mice were tested as GAS recipient control mice, and the nasal samples of eight mice were used for histological analyses. Twelve of the nineteen C57BL/6J mice were negative for throat GAS colonization, and seven of eleven C57BL/6J mice had no detectable GAS in the nasal cavity (Fig. 1B and C). The GAS-positive C57BL/6J mice had mean 53 CFU GAS/swab throat colonization and 674 CFU GAS in the nasal cavity. These results indicate that wild-type C57BL/6J recipient mice did not establish significant nasal cavity and throat GAS infection and colonization through transmission. In contrast, combined male and female MyD88^−/− recipient mice had median 2.0 × 10⁶ CFU GAS/swab throat colonization (MyD88^−/− versus C57BL/6J: P < 0.0001) and 6.3 × 10⁷ CFU GAS in the nasal cavity (MyD88^−/− versus C57BL/6J: P < 0.0001) (Fig. 1B and C). There was no significant difference in nasal infection (P = 0.4861) and throat colonization (P = 0.9809) between female and male MyD88^−/− recipient mice. Thus, MGAS2221 transmission can cause nasal infection and throat colonization in MyD88^−/− mice but not in C57BL/6J mice.

FIG 1.

Host-to-host group A Streptococcus (GAS) transmission resulted in GAS infection in the nasal cavity and throat colonization in MyD88^−/− but not C57BL/6 recipient mice. Two C57BL/6J starter mice were inoculated at the nares with 2.8 × 10⁸ CFU of MGAS2221 in experiment 1 and 3.4 × 10⁸ CFU of MGAS2221 in experiment 2. Starter mice were commingled with four C57BL/6 or MyD88^−/− recipient mice in each cage. The starter and recipient mice were euthanized on days 4 and 7, respectively. The throat was swabbed, and the nasal cavity was homogenized to quantify GAS loads. (A) GAS numbers in the homogenized nasal cavity and swab samples of C57BL/6J starter mice. (B and C) GAS numbers in the homogenized nasal cavity (B) and swab (C) samples of recipient mice. The data presented were combined from experiments 1 and 2. Statistical analysis: ****, P < 0.0001. ns, not significant.

Time course of GAS infection in MyD88^−/− recipient mice.

On day 2 after the start of commingling with starter mice, MyD88^−/− recipient mice had mean 1.2 × 10³ CFU/swab throat colonization and 1.4 × 10⁵ CFU GAS in the nasal cavity. These mean values increased to 3.1 × 10⁵ CFU/swab and 3.4 × 10⁷ CFU on day 5 and to 1.6 × 10⁷ CFU/swab and 2.5 × 10⁹ CFU in the throat and nasal cavity, respectively, beyond day 7 when mice became moribund (Fig. 2). Few GAS bacteria were detected in the liver and spleen on day 2, and the mean GAS loads were 2.5 × 10⁴ CFU/spleen and 669 CFU/g liver on day 5. Moribund recipient mice had mean 1.8 × 10⁹ CFU/spleen and 1.2 × 10⁸ CFU/g liver (Fig. 2). Thus, MyD88^−/− recipient mice rapidly establish acute GAS infection in the nasal cavity and GAS colonization in the throat, and lethal systemic GAS infection occurs a couple of days after the upper respiratory infection.

FIG 2.

Time course of GAS infection in MyD88^−/− recipient mice. The experiments were done as in Fig. 1 except that MyD88^−/− recipient mice were euthanized at the indicated days after the start of the commingling. The >7 days were days beyond day 7 when mice became moribund. This figure presents the GAS numbers of swabs and homogenized nasal cavity, spleen, and liver samples.

Lamina propria as the primary GAS infection site in the nasal cavity in recipient MyD88^−/− mice.

To determine where GAS infects the upper respiratory tract of MyD88^−/− recipient mice, sequential sagittal sections of the head were alternatively stained with hematoxylin-eosin (H&E) and Gram stain. Fig. 3A shows a representative Gram staining image of a sagittal section of the nasal cavity of five MyD88^−/− recipient mice on day 7 after the start of commingling with starter mice. The nasal cavity is divided into three turbinate regions, nasoturbinates, maxiloturbinates, and ethmoturbinates (19). There was extensive Gram staining in the upper lip and at the gingival sulcus of the incisor teeth. The lamina propria of the nasoturbinates, maxiloturbinates, and ethmoturbinates had patches of extensive Gram staining that were adjacent to the turbinate bone (Fig. 3B and D). The majority of the mucosal epithelial cells were intact (Fig. 3C). Some epithelial cells next to the GAS-positive lamina propria lost nuclear hematoxylin staining and were necrotic (Fig. 3E). No extensive Gram-positive staining was detected in the turbinate epithelial cells with or without nuclear staining with hematoxylin (Fig. 3B and D). To determine whether the bacteria in the nasal cavity were GAS, immunostaining of the nasal section of the MyD88^−/− recipient mice was performed using rabbit anti-GAS antibodies. The immunostaining pattern of the nasal cavity of MyD88^−/− mice were similar to the Gram-staining pattern shown in Fig. 3A (Fig. 4A). Strong anti-GAS staining was present in the lamina propria but not in the epithelium (Fig. 4B). A control mouse without GAS exposure had no significant anti-GAS immunostaining (Fig. 4C), demonstrating the specificity of the antibody against GAS. The immunostaining results are consistent with the finding that >95% of bacterial colonies from homogenized nasal cavities of MyD88^−/− recipient mice on day 7 were beta-hemolytic on blood agar plates. MyD88^−/− recipient mice at day 9 after the start of the commingling had more severe nasal infection, having necrotic epithelial cells and more bacteria in the turbinate lamina propria than at day 7 (Fig. 5A). However, no extensive Gram staining was present in the necrotic epithelium (Fig. 5A). Thus, GAS preferentially infected the lamina propria over epithelium of the nasal turbinates in MyD88^−/− recipient mice.

FIG 3.

GAS infection in the upper lip, the gingival sulcus of the incisor teeth, and the lamina propria but not the epithelium in the nasal cavity in MyD88^−/− recipient mice. (A) Representative Gram staining image of a sagittal section of the nasal cavity of MyD88^−/− recipient mice at day 7 after commingling with starter mice. The image was generated by compiling multiple partially overlapped image shots along particular tissue marks. (B) Gram staining image of the circled area in panel A showing patches of Gram staining in the lamina propria of turbinate and lack of Gram staining at the epithelium. (C) Hematoxylin-eosin (H&E) staining image of an area corresponding to the circle in panel A showing the intact epithelium. (D and E) Gram and H&E staining images of the boxed area in panel A showing the necrotic epithelial cells near extensive GAS Gram staining in the lamina propria.

FIG 4.

Anti-GAS immunostaining of the nasal cavity of MyD88^−/− recipient mice. A sagittal section of the nasal cavity that was adjacent to the one of the mouse analyzed in Fig. 3 was subject to immunostaining with anti-GAS antibodies. (A) Overall anti-GAS immunostaining image of the nasal cavity showing the positive anti-GAS immunostaining pattern that was similar to the Gram staining pattern in Fig. 3A. The image was generated by compiling multiple partially overlapped image shots along particular tissue marks. (B) Anti-GAS immunostaining image of the boxed area in panel A showing GAS in the lamina propria but not at the epithelium. (C) Lack of anti-GAS Vulcan Fast Red immunostaining in the nasal cavity of a MyD88^−/− control mouse that was not exposed to GAS.

FIG 5.

Lack of neutrophil infiltration in the GAS-containing lamina propria of the turbinate of MyD88^−/− recipient mice in GAS transmission. The MyD88^−/− recipient mouse analyzed was moribund and euthanized on day 9 after the start of the commingling with starter mice. (A) Gram staining image of a sagittal section of the nasal cavity showing severe GAS infection in the lamina propria of an ethmoturbinate and necrotic epithelium without GAS bacteria. (B) Lack of anti-MPO immunostaining image of the ethmoturbinate of a sagittal section that was adjacent to the one in panel A. (C) Anti-MPO immunostaining image of ethmoturbinates in a sagittal section of the nasal cavity of a C57BL/6J mouse on day 4 after intranasal GAS inoculation. The anti-MPO immunostaining was present in the necrotic epithelial and immune cell-filled space between the turbinates. PMN, polymorphonuclear leukocyte.

Lack of neutrophil infiltration in the lamina propria of the nasal turbinates of MyD88^−/− recipient mice.

To check whether there was neutrophil infiltration at infection sites in the lamina propria, immunostaining of nasal section of MyD88^−/− recipient mice was performed using anti-myeloperoxidase (MPO) antibodies. The lamina propria had intense Gram staining (Fig. 5A); however, the lamina propria did not show anti-MPO immunostaining (Fig. 5B). As a positive control for anti-MPO immunostaining, the nasal cavity of a C57BL/6J with nasal MGAS2221 inoculation had intense anti-MPO immunostaining between turbinates (Fig. 5C), indicating a significant neutrophil presence. These results demonstrate the absence of neutrophil infiltration to GAS infection sites in the lamina propria of MyD88^−/− recipient mice, confirming the previous finding that MyD88 is required for neutrophil recruitment in response to GAS infection (18).

Association of GAS with fibers in the lamina propria in MyD88^−/− recipient mice.

To determine whether GAS bacteria are associated with lamina or host cells in the lamina propria, scanning electron microscopy analysis was performed. Fig. 6A shows a section of the turbinate across the epithelium, lamina propria, and turbinate bone. A higher magnification image of the boxed area of panel A shows that numerous bacteria were associated with laminin fibers and filled the space surrounded by laminin next to the turbinate bone (Fig. 6B). Consistent with the negative anti-MPO immunostaining in the GAS-containing lamina propria, there are no signs of the presence of infiltrated intact or necrotic immune cells at the bacterial sites in the lamina propria. There were bacteria on the cilia surface (Fig. 6C) and at the lamina propria side of the basal densa (Fig. 6D). However, there were no bacteria in the epithelial cell layer. These results indicate that GAS bacteria in the lamina propria were associated with fibrous lamina but not with host cells. The results also suggest that bacteria in the lamina propria could not cross the basal densa to enter the epithelial cell layer and that bacteria at the cilia surface could not establish infection at the epithelial cell layer of the mucosa.

FIG 6.

SEM analysis of GAS infection of the lamina propria in MyD88^−/− recipient mice in GAS transmission. (A) SEM image showing a section across the epithelium, lamina propria, and bone of a turbinate. (B) SEM image showing the boxed area in panel A that contained a batch of numerous GAS bacteria associated with lamia fibers next to the turbinate bone. (C) SEM image showing association of a few bacteria with the cilia and bacteria associated with detached mucus. (D) SEM image showing the containment of GAS in the lamina propria by the basal densa.

GAS infection in the nasopharyngeal tract.

As with the nasal cavity, GAS was present in the lamina propria of the nasopharyngeal tract of MyD88^−/− recipient mice on day 7 after the start of commingling with starter mice, but the epithelial cell layer of the nasopharyngeal mucosa was intact and lacked Gram-stained bacteria (Fig. 7). Thus, GAS infected the lamina propria but not the epithelium of the nasopharyngeal tract in MyD88^−/− recipient mice.

FIG 7.

GAS infection in the nasopharyngeal tract of MyD88^−/− recipient mice. (A) Gram stain image showing the ethmoturbinates and part of the nasopharyngeal tract in a MyD88^−/− recipient mouse on day 9 after the start of commingling with starter mice. Gram staining was in the lamina propria, and epithelial cells of the ethmoturbinates were necrotic and contained a few GAS bacteria. (B) Gram staining image of the boxed area of the nasopharyngeal tract in A. (C) H&E stain image showing the intact epithelium of the nasopharyngeal tract in the area that corresponding to the boxed area in A.

Histology of MyD88^−/− and C57BL/6J mice with nasal GAS inoculation.

MyD88^−/− mice with direct nasal inoculation of MGAS2221 became moribund by day 2 after inoculation. Gram staining was present at the ethmoturbinate region, and the bacteria were primarily located in the lamina propria (Fig. 8A and B). The epithelial cells in the ethmoturbinate mucosa were necrotic and were associated with sparsely scattered GAS (Fig. 8B). GAS bacteria were absent at the nasoturbinates and maxiloturbinates (Fig. 8C), and the epithelium at the nasoturbinates and maxiloturbinates was largely intact and had normal nuclear H&E staining (Fig. 8D). GAS was apparently cleared at the nasoturbinates and maxiloturbinates but remained at the ethmoturbinates. These results suggest that GAS crossed the epithelium to enter the lamina propria of the ethmoturbinates. The results also indicate that the environment in the lamina propria was more suitable for GAS growth than that in the necrotic epithelium.

FIG 8.

GAS infection was primarily in the nasal lamina propria in MyD88^−/− mice with nasal GAS inoculation. MyD88^−/− mice were inoculated at the nares with 1.2 × 10⁸ CFU of MGAS2221 and moribund and euthanized on day 2 after inoculation. (A) Gram staining image of a sagittal section of the nasal cavity. The image was generated by compiling multiple partially overlapped image shots along particular tissue marks. (B) Gram staining image of the boxed area in A showing intensive Gram stain in the lamina propria of the ethmoturbinates and necrotic ethmoturbinate epithelium with fewer bacteria than in the lamina propria. (C) Gram staining image of the circled area in A showing the absence of Gram staining in the nasoturbinates and maxiloturbinates. (D) H&E staining image of the area corresponding to the circled area in A showing that the epithelial cells of the nasoturbinates and maxilloturbinates were largely normal.

C57BL/6J mice with direct nasal inoculation of MGAS2221 were obviously sick on day 2 after inoculation, and half of them became moribund on day 4. The nasal space at the nasoturbinates and maxiloturbinates was filled with necrotic immune cells (Fig. 9A). The mucosal epithelial cells of the nasoturbinates and maxiloturbinates were necrotic. Numerous GAS bacteria were associated with necrotic cells in the nasal cavity, and there were few GAS bacteria in the lamina propria of the nasal turbinates (Fig. 9B). Anti-MPO immunostaining was located at the necrotic epithelium and nasal cavity but not in the lamina propria (Fig. 5C). Thus, the competent immune responses in C57BL/6J mice apparently prevented GAS infection in the lamina propria and ethmoturbinates and retained GAS at the nasoturbinates and maxiloturbinates.

FIG 9.

GAS did not infect the lamina propria of the turbinates of C57BL/6J mice with nasal GAS inoculation. C57BL/6J mice were inoculated at the nares with 3.2 × 10⁸ CFU of MGAS2221, and half of them were moribund and euthanized on day 4 after inoculation for histological analyses. (A) Gram staining image of a sagittal section of the nasal cavity showing that the space at the nasoturbinates and maxiloturbinates was filled with necrotic immune cells and showing the lack of Gram staining in the lamina propria of the turbinates. The image was generated by compiling multiple partially overlapped image shots along particular tissue marks. (B) Gram staining image of the boxed area in A showing the lack of Gram stain in the lamina propria, necrotic epithelium, necrotic neutrophils, and bacteria that were associated with necrotic cells.

DISCUSSION

This study establishes a GAS transmission model for acute infection and colonization of the upper respiratory tract using MyD88^−/− mice. After commingling for 4 days with starter mice inoculated with nasal M1T1 GAS, recipient MyD88^−/− mice, but not recipient C57BL/6J control mice, developed acute nasal and systemic GAS infection and throat colonization by day 7. GAS infected the lamina propria of the nasal cavity turbinate and the nasopharyngeal tract in MyD88^−/− recipient mice. Although GAS did not infect the epithelium of the nasal turbinates and the nasopharyngeal tract, it could cause necrosis of nearby epithelial cells. Direct nasal GAS inoculation of MyD88^−/− mice mainly caused GAS infection in the lamina propria of the ethmoturbinates and led to association of scattered GAS bacteria with necrotic epithelial cells of the ethmoturbinates. In contrast, direct nasal GAS inoculation of C57BL/6J mice did not result in bacteria in the lamina propria but rather resulted in the nasal cavity being filled with necrotic immune cells and epithelial cells and numerous GAS bacteria associated with the necrotic cells. Thus, MyD88^−/− mice are highly susceptible to GAS infection through transmission, and GAS has an unexpected GAS tissue infection pattern of the upper respiratory tract in MyD88^−/− recipient mice. The findings indicate that MyD88 mediates protection against GAS infection in the lamina propria and that the epithelium at the respiratory tract has MyD88-independent protective mechanisms against GAS infection.

GAS is a human-specific pathogen, although murine and nonhuman primate models of GAS infection have been used to investigate GAS pathogenesis, gene expression, fitness gene, immune responses, and GAS vaccine candidates. These infection models usually involve inoculation of >10⁷ CFU GAS through intranasal, intraperitoneal, air sac, intratracheal, subcutaneous, and intramuscular routes (6–16). Murine GAS infection models through intranasal inoculation have provided valuable information on host and GAS factors for throat colonization, histopathology of the upper respiratory tract, shedding, asymptomatic carriage, and immune response (20). GAS bacterial numbers at the nasal cavity and NALT decrease within hours postinoculation and rebound by 24 h (7, 21). In our transmission model, most recipient C57BL/6J mice did not have GAS at the nasal cavity and throat, indicating that the murine immune system can efficiently kill small numbers of GAS bacteria acquired through transmission. In the direct nasal inoculation model, high GAS doses apparently overwhelm the immune response, leading to incomplete killing and accumulated innate immune evasion factors that compromise GAS killing to allow rebound of GAS numbers.

A couple of murine transmission models for respiratory bacterial pathogens have been reported. Streptococcus pneumoniae can be transmitted from infant mice with nasal inoculation to infant recipient mice during commingling; however, high percentages S. pneumoniae colonization of infant recipient mice requires influenza A infection (22, 23). FBV/NJ mice inoculated with an emm75 GAS strain shed more GAS bacteria than C57BL/6, CD-1, and BALB/c mice, and recipient FBV/NJ mice commingled with infected mice shed GAS (15). In our model, C57BL/6J recipient mice did not develop an acute and lethal infection in the GAS transmission study. In contrast, MyD88^−/− recipient mice acquired acute and lethal nasal and systemic GAS infection, indicating that host-to-host transmission models for GAS colonization and acute infection can be achieved by using immunocompromised mice that are deficient in immune responses critical for protection against GAS infections. GAS transmission using various immunocompromised mice are expected to be valuable in understanding host immune protection mechanisms against the onset of respiratory GAS infection and identifying GAS factors critical for infection. Transmission models using immunocompromised mice should also be valuable tools for studying host interactions with other bacterial pathogens as well.

MyD88 is a critical adapter for Toll-like receptors and interleukin-1 receptors. MyD88 deficiency leads to frequent and severe infections by pyogenic bacteria, including GAS (24, 25). MyD88 signaling is important for the inflammatory response against GAS infection in mouse models. MyD88-deficient macrophages and dendritic cells dramatically decrease the production of tumor necrosis factor-α (TNF-α), interleukin-6 (IL-6), and IL-12, and interferon-β (IFN-β) (17, 26), and MyD88 plays a critical role in the production of cytokines TNF-1, IL-6, IL-1b, and IL-12; in the production of chemokines MCP-1 and KC; and in the recruitment of neutrophils, macrophages, and dendritic cells to subcutaneous infection sites in mice (18). It would be significant to determine what cytokines, chemokines, and phagocytes play critical roles in host protection against the onset of upper respiratory tract GAS infection through transmission. We are in the process of examining GAS transmission in various gene knockout mice to dissect the role of MyD88-mediated inflammatory responses in the onset of respiratory colonization and infection.

One striking finding in this study is that GAS infects the lamina propria in the nasal cavity and nasopharyngeal tract of MyD88^−/− recipient mice. After direct nasal GAS inoculation of MyD88^−/−, GAS may damage the epithelium or cross the epithelium to enter the lamina propria. After GAS transmission in MyD88^−/− recipient mice, it is most likely that GAS enters the lamina propria from the infection at the upper lip and the gingival sulcus of the incisor teeth and then grows and spreads in the lamina propria. Further study is needed to understand how GAS enters the lamina propria, which may advance the understanding of tissue tropism of GAS in the absence of neutrophil responses. GAS can infect the lamina propria in MyD88^−/− mice but not in C57BL/6J mice, indicating that MyD88-mediated protection is critical to prevent GAS infection in the lamina propria. It is possible that GAS infection in the lamina propria is linked to the systemic dissemination of GAS in MyD88^−/− mice. Further study on how MyD88-mediated protection against GAS infection in the lamina propria and GAS virulence factors critical for lamina propria infection will be important for understanding invasive GAS infections.

Another striking finding is that GAS does not infect the nasal cavity and nasopharyngeal tract epithelia of MyD88^−/− recipient mice. These results suggest that GAS cannot efficiently infect the epithelium, even in the absence of MyD88-mediated innate immune responses. The inability of GAS to infect the respiratory epithelium of MyD88^−/− recipient mice could be due to a combination of several factors. First, GAS may not adhere well to the epithelium of the upper respiratory tract. Second, extracellular GAS on the surface of the nasal mucosa could be cleared in the absence of inflammatory cells by ciliary action. For example, inoculated GAS bacteria in the nasal cavity of MyD88^−/− mice were missing at the nasoturbinate and maxilloturbinate, supporting the clearance of inoculated GAS in the absence of inflammatory cells. Third, it is possible that GAS cannot efficiently invade the respiratory epithelium. Finally, another possibility is that GAS bacteria that enter epithelial cells are cleared inside the cell. Indeed, epithelial cells can clear invaded GAS by autophagy (27). Regardless of actual mechanisms for the finding, the inability of GAS to infect the ciliated columnar epithelium at the upper respiratory tract of MyD88^−/− recipient mice is consistent with the fact that human strep throat patients usually do not have nasal congestion. We plan a follow-up study to determine whether GAS can directly infect the epithelium in the nasal cavity and the nasopharyngeal tract in the absence of phagocytes.

MATERIALS AND METHODS

Declaration of ethical approval.

All animal experimental procedures were carried out in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health (28). The protocols for mouse experiments were approved by the Institutional Animal Care and Use Committee at Montana State University (MSU).

Bacterial strains and growth.

Serotype M1 strain MGAS2221 has been described (29). The strain was grown in Todd-Hewitt broth supplemented with yeast extract (THY) or on agar plates. For growth on THY agar plates, bacteria were streaked fully on plates and incubated for 8 h at 37°C in 5% CO₂. The bacteria were collected from plates using an inoculation loop, washed in 1.0 mL Dulbecco’s phosphate-buffered saline (DPBS), and resuspended in DPBS for inoculation

Mouse infections.

C57BL/6J and MyD88 KO (Myd88^tm1.1Defr/J) mice were bred at the Animal Resources Center of MSU using breeding mice from The Jackson Laboratory. Breeding mice used did not have throat colonization of beta-hemolytic bacterium. Murine GAS transmission experiments involved cages of six mice/cage: two starter mice with nasal GAS inoculation and four naive or recipient mice. Starter mice were 1-month-old C57BL/6J mice that were inoculated in the nostrils with 13 μl of MGAS2221 suspension in DPBS with an optical density at 600 nm of 30, and inocula were determined by plating. The starter mice were commingled with four 1-month-old MyD88^−/− or C57BL/6J (control) recipient mice and were euthanized on day 4 after the start of commingling or prior to day 4 when starter mice became moribund. Recipient mice were euthanized on day 7 or other specified days. Euthanized mice were swabbed using mini-tip polyester swabs, and the liver, the spleen, and the part of the head containing the nasal cavity were collected for GAS quantification. For some mice, the head without the tongue and lower jaw was collected for histological analyses.

Quantification of throat colonization and GAS numbers in the nasal cavity.

For determining throat colonization, euthanized mice were swabbed using mini-tip polyester swabs, and the swabs were put in 0.4 mL THY with 15% glycerol and vortexed for 2 min. The swab samples were plated at appropriate dilutions on trypticase soy agar plates with 5% sheep blood (blood agar plates), and beta-hemolytic colonies were counted as the number of GAS. The remaining swab samples were frozen at −80°C for plating at different dilutions if necessary.

For quantification of GAS numbers in the nasal cavity, the skin of the head, upper lip, lower jaw, and tongue were removed, and the nasal cavity was obtained by cutting the head at the dorsal side of the eyes. The nasal cavity samples were homogenized in 5 mL THY/15% glycerol by grinding the tissue on a wire mesh screen with a syringe pestle. Homogenized samples were vortexed for 2 min and plated at appropriate dilutions on blood agar plates. Remaining samples were stored at −80°C for plating at different dilutions if necessary.

Histological analyses.

Wholes heads were collected, and the nasal cavity was perfused with 10% formalin through the trachea. The heads were then incubated in 10% formalin for 48 h. Fixed heads were decalcified with 10% ETDA, pH 7.5, for 2 weeks. Decalcified heads were dehydrated with ethanol, cleared with xylene, and infiltrated with paraffin using a tissue-embedding console system (Sakura Finetek, Inc.). Four-μm sagittal sections of heads were obtained using Leica RM2155 Microtome, deparaffinized, and stained with hematoxylin and eosin (H&E) or with a Gram stain kit from Becton, Dickinson and Company. Microscopic images were taken with a Nikon ELCIPSE E800 fluorescence microscope.

Immunohistochemistry.

Four-μm sagittal sections of formalin-fixed heads were deparaffinized and then boiled in citrate solution, pH 6.0, for 25 min for antigen retrieval, washed with DPBS and treated twice with TBS for 2 min. The samples were blocked with BioCare Rodent Block M (catalog number RBM961H for 30 min, washed with 0.05% Tween 20 in TBS (TBS-T), and then incubated with rabbit anti-Streptococcus group A polyclonal antibodies (US Biological catalog number S7974-28) diluted 1:500 in 0.5% Tween 20 in TBS for 1.5 h. The samples were washed with TBS-T and incubated with the secondary antibody, BioCare rabbit AP polymer (catalog number RMR625H), for 30 min. The samples were then washed with TBS-T and developed for 7 min with the BioCare Vulcan Fast Red stain kit (catalog number FR805H). The samples were washed with water and counter stained with hematoxylin from Richard-Allan Scientific (catalog number 7211).

Scanning electron microscopy.

Four-μm sections of formalin-fixed nasal tissue were mounted on P-type silicon wafers and deparaffinized. The samples without coating were examined in a Zeiss SUPRA 55VP scanning electron microscope.

Statistical analyses.

The statistical analyses were performed using the Mann-Whitney test of the GraphPad Prism software (version 9.1.2).