Toll-like receptor 2 has a prominent but nonessential role in innate immunity to Staphylococcus aureus pneumonia.

Staphylococcus aureus is an important cause of acute bacterial pneumonia. Toll-like receptor 2 (TLR2) recognizes multiple components of the bacterial cell wall and activates innate immune responses to gram-positive bacteria. We hypothesized that TLR2 would have an important role in pulmonary host defense against S. aureus TLR null (TLR2-/-) mice and wild type (WT) C57BL/6 controls were challenged with aerosolized S. aureus at a range of inocula for kinetic studies of cytokine and antimicrobial peptide expression, lung inflammation, bacterial killing by alveolar macrophages, and bacterial clearance. Survival was measured after intranasal infection. Pulmonary induction of most pro-inflammatory cytokines was significantly blunted in TLR2-/- mice 4 and 24 h after infection in comparison with WT controls. Bronchoalveolar concentrations of cathelicidin-related antimicrobial peptide also were reduced in TLR2-/- mice. Lung inflammation, measured by enumeration of bronchoalveolar neutrophils and scoring of histological sections, was significantly blunted in TLR2-/- mice. Phagocytosis of S. aureus by alveolar macrophages in vivo after low-dose infection was unimpaired, but viability of ingested bacteria was significantly greater in TLR2-/- mice. Bacterial clearance from the lungs was slightly impaired in TLR2-/- mice after low-dose infection only; bacterial elimination from the lungs was slightly accelerated in the TLR2-/- mice after high-dose infection. Survival after high-dose intranasal challenge was 50-60% in both groups. TLR2 has a significant role in early innate immune responses to S. aureus in the lungs but is not required for bacterial clearance and survival from S. aureus pneumonia.

Introduction

Staphylococcus aureus is a leading cause of health care associated pneumonia and is growing in importance as a respiratory pathogen in the community (Kollef et al. 2005; Chambers and Deleo 2009; Jones 2010). The worldwide emergence of highly virulent, antibiotic resistant‐strains of S. aureus lends special urgency to understanding mechanisms of host resistance to this vigorous pathogen (Zetola et al. 2005; DeLeo and Chambers 2009). Successful defense against S. aureus infections of the lungs and other tissues is largely dependent on an innate immune response that recruits neutrophils to the site of infection (Foster 2005; von Kockritz‐Blickwede et al. 2008; Robertson et al. 2008; Kohler et al. 2011).

The activation of innate immune responses to bacteria in the lungs involves the detection of bacterial ligands by pattern recognition sensors such as Toll‐like receptors (TLRs) (Fournier and Philpott 2005; Kawai and Akira 2011). Of the TLRs that have been identified, TLR2 has emerged as playing the most important role in the activation of innate immune responses to S. aureus (Fournier and Philpott 2005; Fournier 2012). TLR2 recognizes diverse microbial structures and detects multiple components of the staphylococcal cell wall, including lipoteichoic acid and lipopeptides (Lien et al. 1999; Schwandner et al. 1999; Takeuchi et al. 1999; Underhill et al. 1999; Yoshimura et al. 1999; Fournier and Philpott 2005; Hashimoto et al. 2006b; Kurokawa et al. 2009; Muller‐Anstett et al. 2010; Fournier 2012). TLR2 mediates the secretion of pro‐inflammatory cytokines and chemokines by macrophages stimulated with S. aureus (Underhill et al. 1999; Ozinsky et al. 2000; Hoebe et al. 2005; Schmaler et al. 2009; Ip et al. 2010; Wolf et al. 2011; Yimin et al. 2013), and facilitates the recruitment of neutrophils to diverse sites of S. aureus infection (Miller et al. 2006; Mullaly and Kubes 2006; Sun et al. 2006; Nichols et al. 2009). TLR2 also stimulates the killing of S. aureus by neutrophils (Jann et al. 2011; Yipp et al. 2012). Overall, the importance of TLR2 in host resistance to S. aureus appears to vary with the locus of infection. TLR2 has been shown to exert a protective effect against intravenous or subcutaneous challenge with S. aureus (Takeuchi et al. 2000; Hoebe et al. 2005; Miller et al. 2006; Gillrie et al. 2010; Yimin et al. 2013), but to be dispensable or counter‐protective in peritoneal infection or brain abscess (Kielian et al. 2005; Mullaly and Kubes 2006; Nichols et al. 2009; Blanchet et al. 2014). The role of TLR2 in host defense against S. aureus pneumonia has not been fully investigated.

Functional TLR2 is expressed by diverse cell populations within the lungs (Opitz et al. 2010), including alveolar macrophages (Hoogerwerf et al. 2010; Juarez et al. 2010; Kapetanovic et al. 2011), airway epithelial cells (Hertz et al. 2003; Armstrong et al. 2004; Muir et al. 2004; Sha et al. 2004; Soong et al. 2004; Mayer et al. 2007), and pulmonary endothelial cells (Pai et al. 2012). TLR2 has been shown to mediate pro‐inflammatory cytokine responses of alveolar macrophages and airway epithelial cells to S. aureus in vitro (Muir et al. 2004; Soong et al. 2004; Kapetanovic et al. 2011), as well as pulmonary inflammatory responses to staphylococcal lipoteichoic acid in vivo (Knapp et al. 2008). Thus, we hypothesized that TLR2 would play an important role in the activation of innate immune responses to S. aureus infection of the lungs. We used a murine model of acute S. aureus pneumonia to test this hypothesis (Skerrett et al. 2004; Ventura et al. 2008a,b).

Materials and Methods

Bacteria

A human blood isolate of S. aureus designated JP1 was obtained from the microbiology laboratory of the Veterans Affairs Puget Sound Health Care System (Skerrett et al. 2004; Ventura et al. 2008a,b). This is a serotype 8, methicillin‐sensitive strain that produces alpha toxin (Chaffin et al. 2012). The genetic features of this organism in comparison with other strains of S. aureus, its transcriptomic response to pulmonary infection, and other host: pathogen interactions during experimental pneumonia have been described (Skerrett et al. 2004; Braff et al. 2007; Ventura et al. 2008a,b; Chaffin et al. 2012). The organism was grown overnight in Luria Bertani (LB) broth, diluted to 30% glycerol, aliquoted, flash frozen in dry ice and ethanol, and stored at −80°C. For each experiment, bacteria were seeded in LB broth and incubated for 6 h at 37°C on a shaking platform, then diluted 100‐fold into fresh LB broth. After 16–18 h incubation at 37°C with agitation, stationary phase bacteria were pelleted by centrifugation, washed twice in PBS, and suspended in PBS to a concentration estimated by optical density at 540 nm and confirmed by quantitative culture on LB agar.

Mice

Mice with targeted deletions of TLR2 were obtained from S. Akira (Osaka, Japan) and backcrossed to C57BL/6 mice for at least six generations, as described (Takeuchi et al. 2000; Skerrett et al. 2007). Male and female C57BL/6 mice 8–10 weeks of age and free of specific pathogens were purchased from Jackson Laboratories (Bar Harbor, ME) and served as wild type (WT) controls after being housed locally for at least 1 week. Mice were housed in laminar flow cages and permitted ad lib access to sterile food and water. All experiments were approved by the Institutional Animal Care and Use Committee of the University of Washington.

Experimental pneumonia and tissue harvest

Two methods of infection were used. For kinetic studies of bacterial clearance and host responses, mice were exposed to aerosolized bacteria in a whole animal exposure chamber, as described (Skerrett et al. 2004). Briefly, groups of WT and TLR2−/− mice were placed in individual wire mesh compartments within a 55L chamber and exposed for 30 min to bacterial aerosols generated by twin jet nebulizers (Salter, Arvin, CA). Three distinct inocula were tested (low, intermediate, and high), whereby the volume of broth in which the bacteria were cultured determined the concentration of bacteria in the final 20 mL volume of slurry used in the nebulizers. Actual bacterial deposition in each experiment was determined by quantitative culture of lung tissue harvested from sentinel mice immediately after infection (Table 1). At serial time points after infection (4 h, 24 h, 48 h, and 96 h), 3–5 mice in each group were euthanized with pentobarbital and exsanguinated by cardiac puncture. The left lung was harvested for quantitative culture and cytokine measurements, the spleen was taken for quantitative culture, and the right lung was lavaged by sequential instillation and retrieval of four 0.5 mL volumes of 0.9% sodium chloride with 0.6 mmol/L EDTA via a 22 g catheter in the trachea, as described (Morris et al. 2009). For survival studies mice were infected by the intranasal route, which permitted deposition of higher concentrations of bacteria in the lungs. Mice were anesthetized with isoflurane, held in the vertical position, and a 50 μL suspension containing 3 × 108 CFU bacteria in PBS was distributed between both nares (Ventura et al. 2008a,b). Mice were held upright for 1 min after inoculation and then were recovered in the prone position. Preliminary studies established that this dose of bacteria resulted in approximately 50% mortality in WT mice.

Table 1

Preparation and lung deposition of aerosolized Staphylococcus aureus

Inoculum	Volume of broth	Experiment	Nebulizer slurry (CFU/mL)	Lung depositiona (CFU/lung)
Low	100 mL	1	8.8 × 1010	3.5 ± 0.1 × 105
		2	7.5 × 1010	5.3 ± 1.4 × 105
Intermediate	1 L	1	1.4 × 1012	6.6 ± 0.7 × 106
		2	1.1 × 1012	2.9 ± 0.1 × 106
High	2 L	1	3.2 × 1012	1.0 ± 0.03 × 107
		2	3.2 × 1012	1.7 ± 0.2 × 107

Data are mean ± SEM, n = 3–4 mice, from which the left lung was homogenized and quantitatively cultured immediately after infection.

Bacterial clearance

Left lungs and spleens each were homogenized in 1 mL PBS, and 0.1 mL volumes of serial dilutions in PBS were cultured on duplicate LB agar plates. Colonies were counted after 48 h incubation at 37°C.

Bronchoalveolar lavage cell counts

At designated time points after infection mice underwent bronchoalveolar lavage (BAL) (Morris et al. 2009). The lavage fluid was centrifuged at 300  and supernatants were stored at −80°C. The cell pellets were resuspended in RPMI 1640 (Mediatech, Manassas, VA) containing 10% heat‐inactivated fetal bovine serum (HyClone Laboratories, Logan, UT). Cell counts were measured by hemocytometer and differential counts were determined by examination of cytocentrifuge specimens stained with a modified Wright‐Giemsa technique (Diff‐Quick, Dade Behring, Dudingen, Switzerland).

Measurement of cytokines, cathelicidin, and total protein

Lung homogenates in PBS were diluted 1:1 in lysis buffer containing 2× protease inhibitor cocktail (Roche Diagnostics, Mannheim, Germany), incubated on ice for 30 min, and then centrifuged at 1500 . The supernatants were harvested and stored at −80°C until assayed. Tumor necrosis factor‐α (TNF‐α), interleukin (IL)‐1β, IL‐6, IL‐10, IL‐12 p70, macrophage inflammatory protein 2 (MIP‐2; CXCL2), keratinocyte chemokine (KC; CXCL1), monocyte chemoattractant protein 1 (MCP‐1; CCL2), granulocyte‐macrophage colony stimulating factor (GM‐CSF), and interferon‐γ (IFN‐γ) were measured in lung homogenates by multiplex microbead array using a Luminex 100 analyzer (Austin, TX) and reagents purchased from R&D Systems (Minneapolis, MN). IL‐1α and IL‐17A were measured by ELISA using DuoSet kits from R&D Systems. Lung specimens from uninfected mice were used as controls. Mouse cathelicidin‐related antimicrobial peptide (CRAMP) was measured in BALF by ELISA, using reagents generously provided by R. Gallo (University of California, San Diego) (Braff et al. 2007). Total protein in BALF was measured using the bicinchoninic acid assay (Pierce, Rockford, IL).

Histopathology

A separate experiment was performed to obtain specimens for histological analysis. Four and 24 h after exposure to aerosolized bacteria (intermediate dose), four WT and four TLR2−/− mice were euthanized and exsanguinated. The trachea was cannulated, the chest opened, and the lungs were inflated to 20 cm pressure with 4% paraformaldehyde. The lungs were removed en bloc and fixed in 4% paraformaldehyde at 4°C for 24 h then transferred to 70% ethanol. The tissue was embedded in paraffin and stained with hematoxylin and eosin. Four widely spaced sections of lung from each animal were scored for the intensity of peribronchial, perivascular, and alveolar inflammation and necrosis, each on a scale of 0–4, by a veterinary pathologist who was blinded to the genotype and time after infection of each specimen (Morris et al. 2009; Gibson‐Corley et al. 2013).

Phagocytosis and killing of S. aureus by alveolar macrophages in vivo

Mice were exposed to aerosolized S. aureus at the low inoculum, a concentration that resulted in pulmonary deposition of bacteria below the threshold required to elicit a rapid neutrophil response (Toews et al. 1979). Both lungs were lavaged 4 h after infection as described (Skerrett et al. 1999). The BALF from each mouse was centrifuged at 300  at 4°C, and the supernatant stored at −80°C. The cell pellet was washed twice in cold Hank's Balanced Salt Solution and resuspended in 0.5 mL RPMI 1640. A 25 μL volume of this suspension was removed for counting in a hemocytometer and 75 μL were used for preparation of a cytocentrifuge slide stained with Diff‐Quik. The remaining cell suspension was lysed with 0.5% Triton X and quantitatively cultured. Preliminary studies established that this concentration of Triton X did not affect the viability of S. aureus. Complete lysis of the cell suspension was documented by microscopic analysis. The cytocentrifuge slides were scored for differential cell counts, the proportion of cells with cell‐associated bacteria, and the number of bacteria per cell (Skerrett et al. 2004).

Statistical analysis

Statistical analyses were performed with GraphPad Prism 7 software (La Jolla, CA). Interval data are expressed as mean ± SEM, and statistical comparisons between two experimental groups were made using unpaired two‐tailed T tests. Cytokine analyses were adjusted for multiple comparisons using the Holm‐Sidak method (GraphPad). Ordinal data (histological grading) are shown as median with interquartile range, and statistical comparisons were made using Mann–Whitney U tests. A P value of <0.05 was considered significant.

Results

Cytokine responses after inhalation of S. aureus were blunted in TLR2−/− mice

Levels of immunoreactive cytokines and chemokines were measured in lung homogenates harvested 4 h, 24 h, and 48 h after inhalation of S. aureus at two different inocula (Fig. 1). Dose‐ and time‐related induction of most proteins was evident in both WT and TLR2−/− mice, but responses were significantly blunted in TLR2‐deficient animals. TNF‐α, MIP‐2, and IL‐1α concentrations at the 4 h and 24 h time points were reduced by >50% in TLR2−/− mice at both levels of infectious challenge, suggesting that induction of these cytokines was largely dependent on TLR2. Lung levels of IL‐17A did not differ significantly between TLR2−/− and WT mice, and concentrations of IL‐10, IL‐12p70, and IFN‐γ were not significantly above those in uninfected controls in either group of mice (not shown). These findings indicate that TLR2‐dependent recognition plays an important role in the activation of innate immune responses to S. aureus in the lungs.

Figure 1

Lung cytokine responses to inhaled Staphylococcus aureus are blunted in TLR2−/− mice. Cytokines were measured by microbead array or ELISA in lung homogenates harvested 4 h, 24 h, and 48 h after deposition of 4.5 × 106 CFU/lung (intermediate dose; A–D, I–L) or 1.4 × 107 CFU/lung (high dose; E–H, M–P). Data are mean ± SEM; n = 8 except for 48 h after intermediate dose and 4 h after high dose where n = 4; and represent the combined results of two separate experiments at each bacterial inoculum. *P < 0.05 versus WT; **P < 0.005 versus WT. Protein levels in uninfected lung homogenates were as follows (mean ± SEM, n = 4, all pg/mL): TNF‐α, 41 ± 9; IL‐1β, 311 ± 55; IL‐1α, 229 ± 27; MIP‐2 (CXCL2), 67 ± 43; KC (CXCL1), 264 ± 45; MCP‐1 (CCL2), 236 ± 97; IL‐6, 48 ± 7; GM‐CSF, 45 ± 13.

Lung inflammation in response to S. aureus infection was mildly reduced in TLR2−/− mice

Inhalation of S. aureus resulted in a dose‐dependent influx of neutrophils into the airspaces of the lungs that was evident by 4 h after intermediate or high dose infection and persisted through 96 h of observation (Fig. 2; P < 0.005 intermediate versus high dose for both WT and TLR2−/−, at both 4 h and 24 h after infection). The number of bronchoalveolar neutrophils was reduced by more than 60% in TLR2−/− mice in comparison with WT controls 4 h after intermediate dose infection and remained significantly blunted in this group through the first 48 h after infection (Fig. 2A). Greater numbers of mononuclear cells were present in BAL samples of TLR2−/− mice 96 h after infection, suggesting delayed resolution of inflammation in these mice (Fig. 2A and B). After high dose challenge, in contrast, both WT and TLR2−/− mice mounted vigorous neutrophilic responses; significantly fewer neutrophils were found in BAL samples from TLR2−/− mice only at the 48 h time point, and there were no significant differences in mononuclear cells (Fig. 2C and D).

Figure 2

Reduced neutrophil response to inhaled Staphylococcus aureus in TLR2−/− mice. Bronchoalveolar lavage (BAL) neutrophils (PMN) and mononuclear cells (MN) were determined at serial time points after inhalation of S. aureus at two different inocula: 4.5 × 106 CFU/lung (intermediate dose; A, B), or 1.4 × 107 CFU/lung (high dose; C, D). Data are mean ± SEM, n = 8, except for 48 h after intermediate dose and 4 h after high dose where n = 4; and represent the combined results of two separate experiments at each inoculum. *P < 0.05 versus WT; **P < 0.005 versus WT.

Reduced lung inflammation in TLR2−/− mice also was evident from histological analysis. As shown in Figure 3, bronchiolar and alveolar inflammation was readily apparent 24 h after intermediate dose infection in WT mice, but these lesions were less severe in TLR2−/− animals, in agreement with the BAL findings. Using semi‐quantitative morphometry, only bronchiolar inflammation 24 h after inhalation of S. aureus was significantly reduced in TLR2−/− mice in comparison with WT mice (Fig. 4).

Figure 3

Reduced lung inflammation in TLR2−/− mice with Staphylococcus aureus pneumonia. Lung tissue was harvested 4 h and 24 h after inhalation of S. aureus (6.6 × 106 CFU/lung), sectioned, and stained with hematoxylin and eosin. Images of lesions representing the upper range of severity scores demonstrate less intense inflammation in TLR2−/− mice.

Figure 4

Reduced lung inflammation in TLR2−/− mice with Staphylococcus aureus pneumonia. Lung tissue was harvested 4 h and 24 h after inhalation of S. aureus (6.6 × 106 CFU/lung). Hematoxylin and eosin stained sections were scored for A. Bronchial, B. Alveolar, and C. Vascular inflammation and necrosis on a 4 point scale. Each data point represents the composite score from examination of 4 widely spaced sections of lung from an individual mouse. *P < 0.05 versus WT.

Cathelicidin release into airways after inhalation of S. aureus was decreased in TLR2−/− mice

We measured airway levels of CRAMP because this antimicrobial peptide has microbicidal activity against S. aureus (Bals et al. 1999; Jann et al. 2009; Alalwani et al. 2010), is inducible by TLR2 ligands (Liu et al. 2006), and is strongly upregulated in the lungs of mice with S. aureus pneumonia (Braff et al. 2007). Concentrations of CRAMP were undetectable in BALF from uninfected mice and in BALF harvested 4 h after infection with an intermediate dose of S. aureus (6.6 × 106 CFU/lung). By 24 h after infection CRAMP was readily detectable in BALF from both WT and TLR2−/− mice, but levels were reduced by more than 50% in TLR2−/− mice in comparison with WT mice (Fig. 5). Thus, the release of CRAMP into the airways after inhalation of S. aureus is significantly impaired in the absence of TLR2.

Figure 5

Decreased release of cathelicidin‐related antimicrobial peptide (CRAMP) in TLR2−/− mice. CRAMP was measured by ELISA in bronchoalveolar lavage fluid 24 h after inhalation of Staphylococcus aureus (deposition 6.6 × 106 CFU/lung). Each data point represents an individual mouse. **P = 0.001

The viability of S. aureus ingested by alveolar macrophages in vivo was increased in TLR2−/− mice

To assess the uptake and killing of S. aureus by alveolar macrophages in vivo, we exposed mice to low‐dose aerosols that resulted in bacterial depositions that were below the threshold required to elicit a rapid neutrophil response (Toews et al. 1979). Mice were lavaged 4 h after infection, yielding bronchoalveolar cell populations that were ≥99% alveolar macrophages (by Wright‐Giemsa morphology) in all animals. The number of BAL macrophages did not differ between the two groups of mice (Fig. 6A). Cell‐associated bacteria were visible in approximately 70% of alveolar macrophages from both groups of animals; there were no differences between WT and TLR2−/− mice in the proportion of cells harboring bacteria or in the number of bacteria per cell (Fig. 6B). However, the viability of S. aureus in cells harvested from TLR2−/− mice was significantly greater than in cells from WT mice (Fig. 6C). Culture of lung homogenates 24 h after low dose infection demonstrated that the bacterial burden in TLR2−/− mice was approximately twofold higher than that of WT controls (Fig. 6D). These data suggest that TLR2 is required for optimal bacterial killing by alveolar macrophages after inhalation of S. aureus in vivo, contributing to a mild impairment or delay in bacterial clearance after low‐dose infection.

Figure 6

Increased viability of Staphylococcus aureus ingested by TLR2‐deficient alveolar macrophages (AM) in vivo. Bronchoalveolar cells (≥99% AM) were harvested 4 h after low‐dose airborne infection with S. aureus (deposition 4.3 × 105 CFU/lung) (A). Cytocentrifuge samples were scored for the number of cell‐associated bacteria (B, mean ± SEM, n = 10). The cell pellets were lysed and quantitatively cultured (C). Lung homogenates were cultured 24 h after low‐dose infection (D). Data represent the combined results of two independent experiments. *P < 0.05; **P < 0.005.

Bacterial clearance was not impaired in TLR2−/− mice after intermediate or high‐dose infection

Despite the blunted inflammatory responses, reduced levels of CRAMP, and evidence of defective bacterial killing by alveolar macrophages, bacterial clearance from the lungs was not impaired in TLR2−/− mice after inhalation of S. aureus at intermediate or high inocula (Fig. 7A and C). Indeed, bacterial clearance by TLR2−/− mice was slightly more rapid after intermediate and high‐dose infection than in WT controls. By 96 h after infection, the number of bacteria persisting in the lungs was indistinguishable between the two groups of mice after both intermediate and high‐dose infection. Bacterial burdens in spleen, indicative of systemic spread of infection, did not differ significantly between the two groups of mice (Fig. 7B and D). Thus, TLR2 is not required for elimination of S. aureus from the lungs and spleen.

Figure 7

Bacterial clearance after inhalation of Staphylococcus aureus is unimpaired in TLR2−/− mice after intermediate or high‐dose infection. Bacterial burdens in lung and spleen were measured at serial time points after deposition of 4.5 × 106 CFU/lung (intermediate dose; A, B) or 1.4 × 107 CFU/lung (high dose; C, D). Data are mean ± SEM, n = 8, except for 48 h after intermediate dose and 4 h after high dose where n = 4; and represent the combined results of two separate experiments at each inoculum. **P < 0.005 versus WT.

Survival from staphylococcal pneumonia was not reduced in mice lacking TLR2

To determine if TLR2‐dependent signaling influenced the outcome of potentially lethal bolus infection, mice were infected by the intranasal route using an inoculum determined in pilot studies to result in approximately 50% survival in WT C57BL/6 mice. As shown in Figure 8, survival after infection did not differ significantly between TLR2−/− and WT mice. Lung and spleen tissues harvested 7 days after challenge revealed persistent infection in all surviving mice, with no significant differences in bacterial burden between the two experimental groups.

Figure 8

Survival from Staphylococcus aureus pneumonia is unimpaired in TLR2−/− mice. Anesthetized mice (n = 10 per group) were inoculated intranasally with S. aureus (5 × 108 CFU). After 7 days surviving mice were euthanized. Left lungs and spleens were homogenized for quantitative culture.

Discussion

Using a murine model of acute staphylococcal pneumonia, we found that the intrapulmonary production of proinflammatory cytokines and chemokines, the recruitment of neutrophils to the lungs, the release of cathelicidin into the airspaces, and bacterial containment by alveolar macrophages all were reduced in TLR2−/− mice in comparison with WT controls. However, bacterial clearance and survival were not adversely affected by the absence of TLR2. These findings indicate that TLR2 has a prominent but non‐essential role in pulmonary host defense against S. aureus infection.

Our results demonstrate that TLR2 is involved in triggering innate immune responses to S. aureus in the lungs. TLR2 is known to recognize multiple ligands present in the cell walls of S. aureus (Lien et al. 1999; Schwandner et al. 1999; Yoshimura et al. 1999), including lipoproteins (Hashimoto et al. 2006a,b; Kurokawa et al. 2009), lipoteichoic acid (Schwandner et al. 1999; Travassos et al. 2004; Hashimoto et al. 2006a), and, possibly, peptidoglycan (Schwandner et al. 1999; Dziarski and Gupta 2005; Muller‐Anstett et al. 2010; Fournier 2012), as well as secreted toxins such as the Panton‐Valentine leukocidin (Zivkovic et al. 2011). Diverse resident cell populations in human and murine lungs express functional TLR2, including alveolar macrophages (Hoogerwerf et al. 2010; Juarez et al. 2010; Kapetanovic et al. 2011), airway epithelial cells (Hertz et al. 2003; Armstrong et al. 2004; Muir et al. 2004; Sha et al. 2004; Soong et al. 2004; Mayer et al. 2007), and pulmonary endothelial cells (Pai et al. 2012). In vitro studies have demonstrated that TLR2‐mediated recognition of intact S. aureus or its purified surface components stimulates NFκB translocation and/or pro‐inflammatory cytokine production in alveolar macrophages and airway epithelial cells (Muir et al. 2004; Soong et al. 2004; Kapetanovic et al. 2011). In vivo experiments with mice have shown that intranasal instillation of staphylococcal lipoteichoic acid or peptidoglycan induces bronchoalveolar release of pro‐inflammatory cytokines and chemokines, associated with a rapid influx of neutrophils to the lungs (Leemans et al. 2002; Knapp et al. 2008; Poole et al. 2011). The response to LTA in these studies was demonstrated to be entirely dependent on TLR2 (Knapp et al. 2008). Similarly, endobronchial instillation of staphylococcal LTA in human volunteers stimulated local chemoattractant release and neutrophil recruitment (Hoogerwerf et al. 2008). Our data show that intrapulmonary production of pro‐inflammatory cytokines and chemokines in response to S. aureus infection was largely, albeit not completely, dependent on TLR2‐mediated signaling after both intermediate and high‐dose infection. These findings indicate that TLR2 plays an important but nonexclusive role in early recognition of S. aureus in the lungs.

The recruitment of neutrophils to the site of infection is an important early consequence of pathogen recognition, and neutrophils are required for effective resistance to S. aureus pneumonia (Robertson et al. 2008; Kohler et al. 2011; Robinson et al. 2014). TLR2 mediated signaling has been found to contribute to neutrophil recruitment in models of peritoneal, subcutaneous, corneal, and intracerebral infection with S. aureus (Miller et al. 2006; Mullaly and Kubes 2006; Sun et al. 2006; Nichols et al. 2009). Furthermore, intrapulmonary instillation of TLR2 ligands derived from S. aureus was found to induce an influx of neutrophils in both mice and humans (Leemans et al. 2002; Hoogerwerf et al. 2008; Knapp et al. 2008; Poole et al. 2011). TLR2‐mediated signaling also plays a role in the microbicidal actions of neutrophils. TLR2 contributes to the oxidative killing of S. aureus in vitro (Jann et al. 2011), and is required for neutrophil extracellular trap formation in response to S. aureus skin infection in vivo (Yipp et al. 2012). We found that neutrophilic inflammation in response to S. aureus lung infection was less dependent on TLR2 than the measured cytokine responses: early neutrophil recruitment was significantly blunted in TLR2−/− mice after intermediate but not high‐dose infection. These results support a redundant and inoculum‐dependent function for TLR2 in the pulmonary inflammatory response to S. aureus.

We found that bronchoalveolar levels of cathelicidin‐related antimicrobial peptide (CRAMP) were significantly reduced in TLR2−/− mice after inhalation of S. aureus. CRAMP is the murine homolog of the human cathelicidin, LL‐37 (Kovach et al. 2012; Vandamme et al. 2012; Beaumont et al. 2014). Cathelicidins are stored as inactive precursors in the secondary granules of neutrophils (Jann et al. 2009; Vandamme et al. 2012), and are inducible in other cell populations, including pulmonary macrophages and airway epithelial cells (Kovach et al. 2012; Vandamme et al. 2012). TLR2 ligands, in conjunction with vitamin D3, can stimulate cathelicidin expression in human monocytes and macrophages (Liu et al. 2006), and intraocular expression of CRAMP is blunted in TLR2−/− mice with staphylococcal endophthalmitis (Talreja et al. 2015). Cathelicidins have broad microbicidal activities but also exert immunomodulatory effects that influence the recruitment and activation of neutrophils at sites of infection (Alalwani et al. 2010; Vandamme et al. 2012; Beaumont et al. 2014). These peptides exhibit direct antimicrobial activity against S. aureus, and contribute to killing of S. aureus by neutrophils (intracellularly) and airway epithelial cells (extracellularly) (Bals et al. 1999; Jann et al. 2009; Alalwani et al. 2010). Cathelicidins are expressed in the lungs after bacterial infection (Schaller‐Bals et al. 2002; Braff et al. 2007; Kovach et al. 2012), and CRAMP‐deficient mice exhibit impaired resistance to pulmonary infection with Klebsiella pneumoniae and Pseudomonas aeruginosa (Kovach et al. 2012; Beaumont et al. 2014). Chimera studies demonstrated that resistance to K. pneumoniae was dependent on expression of CRAMP by marrow derived rather than structural (parenchymal) cells (Kovach et al. 2012). We have previously reported that CRAMP accumulates in bronchoalveolar lavage fluid after infection with S. aureus (Braff et al. 2007), and in the present study demonstrate that this response is partially dependent on TLR2. However, our observations do not establish the source of cathelicidin in this model, nor define a specific role for cathelicidin in resistance to S. aureus pneumonia.

Alveolar macrophages can ingest and kill S. aureus and are largely responsible for the elimination of staphylococci from the lungs after low‐dose infection that falls below the threshold for triggering a neutrophil response (Green and Kass 1964; Goldstein et al. 1974; Toews et al. 1979; Onofrio et al. 1983; Jubrail et al. 2016). Our studies demonstrated that the avid phagocytosis of S. aureus by alveolar macrophages in vivo was unimpaired in the absence of TLR2. However, the viability of ingested bacteria 4 h after infection was significantly greater in alveolar macrophages from TLR2−/− mice in comparison with WT controls, suggesting that the early killing of S. aureus by alveolar macrophages was partly dependent on TLR2. Prior in vitro studies with other macrophage populations have suggested a complex role for TLR2 in macrophage anti‐bacterial functions that may be cell‐specific. TLR2 was not required for the phagocytosis of S. aureus by bone marrow‐derived macrophages or peritoneal exudate macrophages, but differentially influenced the fate of intracellular bacteria in these cells (Blander and Medzhitov 2004; Watanabe et al. 2007). In bone marrow‐derived macrophages the fusion of S. aureus‐containing phagosomes with lysosomes was impaired in the absence of TLR2, indicating a TLR2‐dependent mechanism for phagosome maturation and bacterial disposal (Blander and Medzhitov 2004). On the other hand, superoxide release and bacterial killing after ingestion of S. aureus were enhanced in peritoneal exudate macrophages that lacked TLR2, supporting an inhibitory effect of TLR2 on the microbicidal activity of these cells (Watanabe et al. 2007).

Despite its role in triggering innate immune responses in the lungs, TLR2 was not required for resistance to S. aureus pneumonia. Survival from staphylococcal pneumonia was not impaired in mice lacking TLR2 and bacterial clearance from the lungs was minimally impaired in the absence of TLR2 only after low‐dose infection; indeed, elimination of S. aureus from the lungs was slightly accelerated in TLR2−/− mice challenged with higher inocula of bacteria. Our results contrast with the only prior study of S. aureus pneumonia in TLR2−/− mice, in which survival after intranasal infection was reduced in the absence of TLR2 (Blanchet et al. 2014). These data were presented in a supplemental figure restricted to survival analysis, which limits the context in which to draw comparisons with our observations (Blanchet et al. 2014). However, the investigators used a different strain of S. aureus and an inoculum more than threefold greater than the highest tested in our studies, suggesting that the contribution of TLR2 to the outcome of staphylococcal pneumonia may depend on the strain and infecting dose of bacteria. Defining the role of TLR2 in pulmonary defense against other isolates of S. aureus will require further investigation.

Research with different experimental models suggests that the role of TLR2 in resistance to S. aureus depends on the route of infection. After intravenous challenge, TLR2−/− mice have consistently exhibited reduced survival (Takeuchi et al. 2000; Hoebe et al. 2005; Yimin et al. 2013), in association with increased burdens of bacteria in blood and kidneys, but not liver, spleen, or lungs (Takeuchi et al. 2000; Hoebe et al. 2005; Schmaler et al. 2009; Gillrie et al. 2010; Yimin et al. 2013). Furthermore, intravenous infection of mice with S. aureus expressing mutant lipoproteins that are unrecognized by TLR2 resulted in increased lethality and higher bacterial burdens in kidney and liver than infection with the wild‐type strain (Bubeck Wardenburg et al. 2006). Subcutaneous infection of TLR2−/− mice with S. aureus resulted in increased local bacterial replication in comparison with wild‐type controls (Hoebe et al. 2005; Miller et al. 2006), without defective cytokine responses or neutrophil recruitment (Miller et al. 2006). Intra‐ocular infection with S. aureus led to blunted early cytokine responses, reduced neutrophil influx, and increased bacterial replication in TLR2−/− mice compared with wild‐type controls (Talreja et al. 2015). In contrast, TLR2−/− mice responded to intra‐peritoneal injection of S. aureus with a delayed inflammatory response but unimpaired bacterial clearance (Mullaly and Kubes 2006), and were protected from lethal challenge (Blanchet et al. 2014). In a murine model of staphylococcal brain abscess, the absence of TLR2 influenced cytokine responses and lymphocyte accumulation, but did not affect bacterial clearance, morbidity, or mortality (Kielian et al. 2005; Nichols et al. 2009). Collectively, these studies suggest that the role of TLR2 in mediating host resistance to S. aureus infection is tissue‐specific and inoculum‐dependent.

The contribution of TLR2 to pulmonary host defense against bacterial infection also appears to be pathogen‐specific. TLR2 partially mediates lung inflammatory responses to Streptococcus pneumoniae, but does not influence bacterial clearance or survival from experimental pneumococcal pneumonia (Knapp et al. 2004), similar to our findings with S. aureus. TLR2 does play an important role in host resistance to intracellular pathogens such as Legionella pneumophila and Francisella tularensis (Archer and Roy 2006; Hawn et al. 2006; Malik et al. 2006; Fuse et al. 2007; Abplanalp et al. 2009). In contrast, TLR2 has a modest suppressive effect on pulmonary defenses against extracellular gram‐negative pathogens such as Pseudomonas aeruginosa and Acinetobacter baumanii (Knapp et al. 2006; Skerrett et al. 2007).

TLR2 recognizes multiple surface components of S. aureus and plays a prominent role in the expression of innate immune responses to acute staphylococcal pneumonia. However, TLR2‐mediated signaling was not required for successful resistance to S. aureus in the lungs under the conditions tested. The contribution of TLR2 to host defense against S. aureus appears to depend on the site of infection.

Conflict of Interest

None declared.