Interleukin 1 Receptor-Like 1 (IL1RL1) Promotes Airway Bacterial and Viral Infection and Inflammation

Interleukin 1 receptor-like 1 (IL1RL1), also known as suppression of tumorigenicity 2 (ST2), is the receptor for interleukin 33 (IL-33) and has been increasingly studied in type 2 inflammation. An increase in airway IL-33/ST2 signaling in asthma has been associated with eosinophilic inflammation, but little is known about the role of ST2 in neutrophilic inflammation.

ABSTRACT

Interleukin 1 receptor-like 1 (IL1RL1), also known as suppression of tumorigenicity 2 (ST2), is the receptor for interleukin 33 (IL-33) and has been increasingly studied in type 2 inflammation. An increase in airway IL-33/ST2 signaling in asthma has been associated with eosinophilic inflammation, but little is known about the role of ST2 in neutrophilic inflammation. Airway Mycoplasma pneumoniae and human rhinovirus (HRV) infections are linked to neutrophilic inflammation during acute exacerbations of asthma. However, whether ST2 contributes to M. pneumoniae- and HRV-mediated airway inflammation is poorly understood. The current study sought to determine the functions of ST2 during airway M. pneumoniae or HRV infection. In cultured normal human primary airway epithelial cells, ST2 overexpression (OE) increased the production of neutrophilic chemoattractant IL-8 in the absence or presence of M. pneumoniae or HRV1B infection. ST2 OE also enhanced HRV1B-induced IP-10, a chemokine involved in asthma exacerbations. In the M. pneumoniae-infected mouse model, ST2 deficiency, in contrast to sufficiency, significantly reduced the levels of neutrophils following acute (≤24 h) infection, while in the HRV1B-infected mouse model, ST2 deficiency significantly reduced the levels of proinflammatory cytokines KC, IP-10, and IL-33 in bronchoalveolar lavage (BAL) fluid. Overall, ST2 overexpression in human epithelial cells and ST2 sufficiency in mice increased the M. pneumoniae and HRV loads in cell supernatants and BAL fluid. After pathogen infection, ST2-deficient mice showed a higher level of the host defense protein lactotransferrin in BAL fluid. Our data suggest that ST2 promotes proinflammatory responses (e.g., neutrophils) to airway bacterial and viral infection and that blocking ST2 signaling may broadly attenuate airway infection and inflammation.

INTRODUCTION

Asthma, particularly severe or refractory asthma (RA), remains a significant challenge in health care (1–3). RA is defined as a poorly controlled disease state despite appropriate therapy. It represents about 10% of the asthma population but is responsible for a disproportionately large amount of health care costs (4). Although airway eosinophilic inflammation has been extensively studied and although corresponding therapies (e.g., anti-interleukin 13 [IL-13] and anti-IL-5 antibodies) have become commercially obtainable, there are few options available to reduce excessive neutrophilic inflammation in RA (5, 6), especially during acute exacerbations that are often associated with bacterial and viral infections. Various bacteria (e.g., Mycoplasma pneumoniae) and viruses (e.g., rhinoviruses [RV]) have been detected in asthmatic airways not only during the acute exacerbations but also under stable conditions (7–9). What drives asthmatic airway neutrophilic inflammation associated with microbial infection remains poorly understood.

IL-33, a member of the IL-1 family, has been implicated in the pathogenesis of asthma (10, 11) and has been increasingly investigated in type 2 inflammation (12, 13). However, limited studies have touched on its role in neutrophilic inflammation. The effects of IL-33 are mediated by its binding to a heterodimer receptor complex, the transmembrane receptor interleukin 1 receptor-like 1 (IL1RL1, or ST2 [suppression of tumorigenicity 2]) and IL-1 receptor accessory protein (IL-1RAP). ST2, a member of the interleukin 1 receptor family, is specific to IL-33 signaling (14, 15). In humans, the ST2 gene is located on chromosome 2q12, an area of the genome linked to asthma (16). In mice, ST2 is located on chromosome 1. There are four splice variants (isoforms) of ST2 including ST2L (or ST2), soluble ST2 (sST2), ST2V, and ST2LV. These isoforms derive from a single transcript as a result of the proximal instead of distal promoter being used during gene transcriptional regulation (17). ST2 and sST2 are the two major isoforms. Many types of cells, including T cells, macrophages, epithelial cells, neutrophils, natural killer cells, mast cells, and basophils, express ST2 (10, 11). ST2 is involved in various disease processes (18–20), and both IL-33 and ST2 have been reported to be increased in asthmatic airways (e.g., airway epithelium) (21).

The involvement of the IL-33/ST2 pathway has not been well defined during pathogen infection associated with neutrophilic inflammation as the limited publications have suggested inconsistent results (22, 23). Both beneficial and detrimental effects of ST2 signaling on infection and inflammation have been reported (24–26). Thus, it is necessary to further determine the function of ST2 in the host defense against respiratory bacterial and viral infection.

The current study is aimed at utilizing a human in vitro model and a mouse in vivo model to test our hypothesis that ST2 enhances proneutrophilic inflammatory responses to both viral and bacterial infections. Specifically, we performed RV and M. pneumoniae infections in ST2-overexpressing human primary airway epithelial cells and in ST2-deficient mice and found that ST2 enhances the production of proinflammatory cytokines involved in neutrophil recruitment into the pathogen-infected lung.

RESULTS

ST2 OE in human airway epithelial cells increases proinflammatory cytokine production.

To determine if ST2 is involved in regulating the proinflammatory responses to M. pneumoniae and human rhinovirus (HRV) infection, ST2 overexpression (OE) experiments were performed by transducing human tracheobronchial epithelial (HTBE) cells from a healthy subject with lentiviruses encoding ST2 cDNA or a scrambled control (SC) cDNA. As shown in Fig. 1, both ST2 mRNA and protein were increased in cells transduced with ST2 lentiviruses compared to levels in the control cells.

FIG 1.

Lentivirus-mediated ST2 OE in human tracheobronchial epithelial (HTBE) cells from a healthy donor. (A) Increased ST2 mRNA expression in cells transduced for 5 days with ST2 cDNA-containing lentiviruses compared to that in virus expressing the scrambled control (SC) DNA sequence. (B) Western blot showing increased ST2 protein expression in cells transduced with ST2 cDNA-containing lentiviruses. replicates. Data are expressed as medians with interquartile ranges (n = 4 replicates).

In the absence of M. pneumoniae infection, ST2 OE increased IL-8 protein levels at 24 and 72 h of cell culture (Fig. 2A). At 24 h after M. pneumoniae infection there was a significant increase in IL-8 levels in the supernatant of the SC cells (Fig. 2A). ST2 OE cell supernatants still had significantly higher IL-8 levels than the SC cell supernatant after M. pneumoniae infection. However, ST2 OE cells did not show a further increase of IL-8 after M. pneumoniae infection compared to the level in noninfected ST2 OE cells (Fig. 2A). IL-33 protein levels were significantly higher in the ST2 OE cell supernatants under baseline conditions. Cell supernatants did not show a significant increase in IL-33 after M. pneumoniae infection, but the trend of ST2 OE cell supernatants having higher levels was maintained and was even significant at 72 h post-M. pneumoniae infection (Fig. 2B).

FIG 2.

Effect of ST2 OE on proinflammatory cytokine production in human tracheobronchial epithelial (HTBE) cells from a healthy donor. After HTBE cells were transduced with lentiviruses containing ST2 cDNA or the scrambled control (SC) DNA sequence for 5 days, cells were infected with M. pneumoniae or HRV1B for 24 and 72 h, and the indicated cytokines were measured in the supernatants. Data are expressed as medians with interquartile ranges (n = 4 replicates).

After human rhinovirus 1B (HRV1B) infection, cell supernatants did not show a significant increase in IL-8 levels compared to the level in noninfected cell supernatants although ST2 OE cell supernatants still showed an increase in IL-8 compared to the level in SC cell supernatants (Fig. 2C). The level of the proinflammatory cytokine IP-10 (CXCL-10) was significantly higher in the ST2 OE cell supernatants under baseline conditions. At 24 h after viral infection, ST2 OE cells continued to have significantly higher IP-10 levels (Fig. 2D). At 72 h, IP-10 levels were increased in ST2 OE cell supernatants with or without HRV1B infection. IL-33 levels were higher in ST2 OE cell supernatants at 24 h with or without HRV infection. At 72 h, IL-33 levels were similar to those at 24 h, but ST2 OE cell supernatants from HRV-infected cells had significantly less IL-33 (Fig. 2E).

ST2 overexpression in human airway epithelial cells increases the levels of M. pneumoniae but not of HRV1B.

M. pneumoniae levels trended to be higher (P = 0.11) in ST2 OE cell supernatants than those of the controls at 24 h postinfection (Fig. 3A). HRV1B loads were similar among the ST2 OE and control cells at both 24 and 72 h postinfection (Fig. 3B).

FIG 3.

Effect of ST2 OE on pathogen load in human tracheobronchial epithelial (HTBE) cells from a normal donor. After HTBE cells were transduced with lentiviruses containing ST2 cDNA or the scrambled control (SC) DNA sequence for 5 days, cells were infected with M. pneumoniae or HRV1B for 24 and 72 h. M. pneumoniae in the supernatants was measured by culture (A), and HRV1B in the cells was quantified by RT-PCR (B). Data are expressed as medians with interquartile ranges (n = 4 replicates). Mp, M. pneumoniae.

Effects of ST2 on mouse lung neutrophilic inflammation during pathogen infection.

(i) M. pneumoniae infection model. We performed 2 or 3 independent mouse model experiments to determine the role of ST2 in bacterial and viral infection. In wild-type (WT) mice, neutrophil levels (number/milliliter and total neutrophils) in bronchoalveolar lavage (BAL) fluid were significantly higher at 24 and 72 h post-M. pneumoniae infection than in the phosphate-buffered saline (PBS)-treated control groups. ST2 knockout (KO) mice demonstrated significantly fewer neutrophils in BAL fluid than the WT mice after 24 h. There were fewer neutrophils in the KO mice at 72 h postinfection, but this difference was no longer statistically significant (Fig. 4A). The percentage of neutrophils in BAL fluid followed a trend similar to that of the total neutrophil count (Fig. 4B). Neutrophil keratinocyte-derived chemokine (KC) levels in BAL fluid were similar among WT and ST2 KO mice with or without M. pneumoniae infection (Fig. 4C). IL-33 was detectable in BAL fluid of all groups of mice, but M. pneumoniae infection did not significantly alter IL-33 levels in either WT or ST2 KO mice (Fig. 4D).

FIG 4.

ST2 modulates lung inflammation in mice infected with M. pneumoniae. WT and ST2 KO mice were intranasally inoculated with M. pneumoniae or PBS (control) for 24 and 72 h. BAL fluid neutrophils were counted (A and B), and proinflammatory cytokines KC (C) and IL-33 (D) were measured in BAL fluid. Each color represents an independent experiment, and each symbol represents an individual animal (n = 10 to 16 mice/group from 2 or 3 independent experiments). Data are expressed as medians with interquartile ranges.

(ii) HRV1B infection model. After 16 h of viral infection, both strains of mice had significant increases in BAL fluid neutrophil levels, in both total numbers and percentages (Fig. 5A and B). Additionally, HRV1B infection increased BAL fluid neutrophils in WT mice after 72 h. However, unlike the M. pneumoniae infection model, ST2 KO and WT mice did not differ significantly in BAL fluid neutrophil levels. Both strains of mice infected with HRV1B for 16 h showed significantly higher levels of KC than their PBS-treated controls (Fig. 5C). Importantly, the ST2 KO mice had significantly lower KC levels than the WT mice at 16 h postinfection. This trend was flipped in the 72-h infection group, with ST2 KO mice showing significantly higher KC levels than WT mice.

FIG 5.

ST2 modulates lung inflammation in mice infected with HRV1B. WT and ST2 KO mice were intranasally inoculated with HRV1B or PBS (control) for 16 and 72 h. BAL fluid neutrophils were counted (A and B), and proinflammatory cytokines KC (C), IP-10 (D), and IL-33 (E) were measured in BAL fluid. Each color represents an independent experiment, and each symbol represents an individual animal (n = 10 to 16 mice/group from 2 or 3 independent experiments). Data are expressed as medians with interquartile ranges.

IP-10 is a sensitive marker of active rhinovirus infection (27) and has also been shown to contribute to recruitment of various types of leukocytes, such as neutrophils (28). After 16 and 72 h of HRV1B infection, WT and KO mice had increased IP-10 levels in BAL fluid (Fig. 5D). The ST2 KO mice infected with HRV1B had significantly less IP-10 in BAL fluid than WT mice, a result that is similar to that for KC. At 72 h post-viral infection, the IP-10 levels in the KO mice were significantly higher than those in the WT mice, a finding similar to BAL fluid neutrophil and KC levels. After 16 h of viral infection, IL-33 was significantly increased in WT mice compared to the level in the PBS-treated controls. Notably, IL-33 levels were significantly lower in ST2 KO mice than in WT mice after HRV1B infection (Fig. 5E). After 72 h of infection, this difference in IL-33 release into the BAL fluid no longer existed.

Effects of ST2 on mouse lung pathogen levels.

(i) M. pneumoniae infection model. ST2 KO mice had significantly lower M. pneumoniae levels in BAL fluid at 24 h postinfection than WT mice (Fig. 6A). The M. pneumoniae load in the homogenized lung tissue of ST2 KO mice also showed a significant reduction compared to that in WT mice at 24 h postinfection (Fig. 6B). After 72 h of infection, BAL fluid and homogenized lung bacterial loads in the WT and ST2 KO mice were similar.

FIG 6.

Effect of ST2 on bacterial or viral load in mice infected with M. pneumoniae or HRV1B. WT and ST2 KO mice were intranasally inoculated with M. pneumoniae, HRV1B, or PBS (control) for 16, 24, and 72 h. BAL fluid or lung tissue was processed to quantify M. pneumoniae (A and B) and HRV1B (C and D) loads by culture or quantitative RT-PCR. Each color represents an independent experiment, and each symbol represents an individual animal (n = 10 to 16 mice/group from 2 or 3 independent experiments). Data are expressed as medians with interquartile ranges.

(ii) HRV1B infection model. Similar to the M. pneumoniae model data, during the early infection (16 h), ST2 KO mice had a significantly lower HRV1B load in the BAL fluid than WT mice and a trend of reduced viral load in the lung tissue (Fig. 6C and D). At 72 h postinfection, the two strains of mice appeared to have similar levels of HRV1B in BAL fluid (Fig. 6C). However, after 72 h of infection, ST2 KO mice showed a significantly higher level of HRV1B in the lung tissue (Fig. 6D). Of note, both strains of mice showed a significant decrease in viral loads at 72 h postinfection versus the load at 16 h postinfection.

Effects of ST2 on host defense protein lactotransferrin expression during M. pneumoniae and HRV infection in mice.

ST2 KO mice had a higher baseline level of lactotransferrin than WT mice. M. pneumoniae infection in the WT and ST2 KO mice increased lactotransferrin levels at 24 h compared to levels in their PBS-treated controls. ST2 KO mice had slightly higher levels of lactotransferrin than the WT mice after 24 and 72 h of M. pneumoniae infection, but the difference was not statistically significant (Fig. 7A and E). After 24 and 72 h of M. pneumoniae infection, there was a weak, but insignificant, negative correlation between lactotransferrin level and M. pneumoniae load in the BAL fluid of ST2 KO mice (Fig. 7B) but not in the WT mice (r = −0.04, P = 0.83).

FIG 7.

Effect of ST2 on lactotransferrin production in BAL fluid of mice infected with M. pneumoniae or HRV1B. WT and ST2 KO were intranasally inoculated with M. pneumoniae, HRV1B, or PBS (control) for 16, 24, and 72 h. (A) BAL fluid was processed to quantify lactotransferrin by Western blotting in M. pneumoniae-infected mice. (B) Correlations between lactotransferrin protein levels and M. pneumoniae load in BAL fluid of ST2 KO mice after 24 and 72 h of infection. (C) Lactotransferrin levels in BAL fluid of HRV1B-infected mice. (D) Correlations between lactotransferrin protein levels and HRV load in BAL fluid of ST2 KO mice after 16 and 72 h of infection. (E) Representative Western blot image for lactotransferrin protein in BAL fluid collected at various time points. Each color represents an independent experiment, and each symbol represents an individual animal (n = 10 to 16 mice/group from 2 or 3 independent experiments). Data are expressed as medians with interquartile ranges.

After 16 h of HRV1B infection, ST2 KO mice had a significantly increased lactotransferrin protein level in the BAL fluid compared to that in WT mice (Fig. 7C and E). At 72 h postinfection, the trend of ST2 KO mice having more lactotransferrin was continued, but the difference was no longer statistically significant. A weak negative correlation between lactotransferrin level and HRV1B load was observed in the BAL fluid of ST2 KO mice (Fig. 7D) but not in that of the WT mice (r = 0.21, P = 0.32).

DISCUSSION

ST2 upregulation has been described in asthmatic airway cells, including epithelium (29), but its function remains controversial. Using the ST2 overexpression (OE) and knockout approaches in vitro and in vivo, our current study has revealed a proinflammatory function of ST2 in both bacterial and viral infections of the airways. Specifically, ST2 OE in human primary airway epithelial cells leads to excessive IL-8 production even in the absence of microbial infection. With rhinovirus infection, ST2 OE further enhances IP-10 production. The proinflammatory function of ST2 in both bacterial and viral infections was further confirmed in the mouse models. Together, given the controversies about the role of ST2 in lung infections, our research findings have helped clarify the role of ST2 during infections with two of the most common pathogens involved in asthma exacerbations and neutrophilic inflammation.

As the IL-33/ST2 signaling axis promotes type 2 inflammation (30), recent research has started to define the role of ST2 in airway infections of various pathogens. Infection of bacteria such as Haemophilus influenzae and Moraxella catarrhalis has been associated with high levels of IL-33 in nasopharyngeal fluids (31). Studies using the respiratory syncytial virus (RSV) and influenza virus mouse models suggest that while blocking the IL-33/ST2 axis reduced RSV-induced lung inflammation and damage (32), it impaired lung epithelial integrity and function during influenza virus infection (33). Thus, the role of ST2 during infection may depend on the species of the pathogens.

Airway infection with M. pneumoniae, an extracellular pathogen, has been shown to contribute to the pathobiology of both stable asthma and acute asthma exacerbations (34–36). However, whether ST2 is involved in M. pneumoniae-induced lung inflammation has never been investigated. By using an ST2-deficient mouse model of M. pneumoniae infection, we demonstrated that ST2 is essential to M. pneumoniae-induced lung neutrophilic inflammation for up to 3 days of infection. The reduction of lung neutrophils accompanied by a decrease in bacterial load at 24 h postinfection (the early phase of infection) in ST2-deficient mice suggests the potential detrimental role of ST2 during an acute M. pneumoniae infection. It remains to be determined if ST2 contributes to chronic or repeated exposure to M. pneumoniae infection.

Since the role of ST2 in infection may depend on the types or strains of pathogens, we sought to further determine if ST2 affects rhinovirus infection that is linked to acute exacerbations of various lung diseases, including asthma. A previous study (37) has shown that intraperitoneal injection of an anti-ST2 antibody in HRV1B-infected BALB/c mice was able to attenuate HRV-induced IP-10 and lung inflammation at 48 h postinfection. However, viral load data were not reported in this publication. Our data showing decreased lung KC and IP-10 levels in ST2-deficient mice at 16 h postinfection are consistent with the previous study. However, even by extending the infection to 3 days, we were not able to see the enhancing effect of ST2 on HRV-induced lung neutrophilic inflammation. Intriguingly, although neutrophil levels were low at 72 h post-viral infection, KC levels were higher in ST2-deficient mice than in the wild-type mice. Although the underlying mechanisms of this observation are unclear, it suggests that the effect of ST2 may also depend on the stage of the infection. Additionally, we found that HRV1B infection in wild-type mice increased the release of IL-33 into the airway lumen, which was in part inhibited by ST2 deficiency at 16 h but not at 72 h post-HRV infection. This novel finding suggests a potential role of HRV infection in promoting type 2 inflammation, which could be attenuated by blocking ST2 signaling. Another new observation in our study was that the effect of ST2 on HRV load may depend on the stage of infection. In the early phase of infection, ST2 promoted viral infection. However, ST2 may inhibit HRV infection at the late stage of infection. Thus, the timing effect should be considered for intervening in ST2 signaling during HRV infection. Together, our in vivo data indicated that during airway rhinovirus infection, ST2 enhanced viral infection and proinflammatory cytokine production but did not significantly affect neutrophil recruitment into the airway lumen. While the underlying mechanism for this discrepancy is unclear, it is possible that ST2 may increase the retention of neutrophils in the lung tissue, such as the airway wall, resulting in less neutrophil emigration into the airway lumen.

As airway epithelium is the major site of M. pneumoniae and HRV infection and as ST2 expression is upregulated in airways of asthmatics, particularly those with type 2-high inflammation (29), we determined if ST2 OE in human airway epithelial cells exerts a similar proinflammatory function. It has been found that IL-33 protein in cultured human airway epithelial cells can be released at low levels constitutively (38). Our data show that the control cells released IL-33 at a low level constitutively, and cells overexpressing ST2 significantly increased in IL-33 production. Higher levels of both ST2 and IL-33 are expected to increase the production of proinflammatory cytokines such as IL-8. Although the focus of our airway epithelial cell culture experiments was on the role of ST2 OE in airway inflammation, we also performed a pilot study to demonstrate the effect on inflammation of inhibiting the IL-33/ST2L pathway. Normal human tracheobronchial epithelial cells were stimulated with IL-33 in the absence or presence of sST2, an IL-33 decoy receptor. We found that sST2 significantly decreased IL-33-mediated IL-8 production. In future studies, we plan to test if sST2 is able to reduce the proinflammatory responses seen in ST2 OE cells exposed to viruses and bacteria. We still do not know why ST2 OE induces IL-33 release from human airway epithelial cells. One possibility is that the existence of baseline IL-33 and ST2 OE may result in a proinflammatory phenotype of the cells, as indicated by increased IL-8 production, which may subsequently secrete more IL-33.

How ST2 modulates pathogen load has not been well understood. In the current study, we explored the role of ST2 in regulating lactotransferrin, a protein shown to inhibit M. pneumoniae infection and marginally decrease HRV load (39, 40). We found a trend of increased lactotransferrin in M. pneumoniae-infected ST2-deficient mice compared to the level in the wild-type mice at 24 h, and this was accompanied by a lower bacterial load. Interestingly, ST2 KO mice had significantly more lactotransferrin protein at 16 h post-HRV1B infection than wild-type mice and also showed a significantly lower viral load in the BAL fluid. This provides more evidence that lactotransferrin not only acts as an antimicrobial but also could have antiviral properties. Collectively, the impact of ST2 on pathogen levels varies depending on the type and stage of an infection. We realize that an inverse relationship of lactotransferrin level with rhinovirus load does not necessarily indicate the direct antiviral activity of lactotransferrin against HRV as there are no publications on this topic. However, due to the potential of lactotransferrin to interact with viruses extracellularly and its nuclear localization in epithelial cells (41), it is worth performing future experiments to determine if lactotransferrin has a direct effect on viral entry into epithelial cells or on viral replication. Alternatively, in an in vivo setting, lactotransferrin may promote viral clearance indirectly through its immunomodulatory effect (42). Future studies are warranted to further determine how the IL-33/ST2 signaling pathway regulates lactotransferrin expression. Our cell culture data of increased M. pneumoniae load in ST2 OE cells is consistent with that in the mouse model of ST2 KO mice that showed a lower M. pneumoniae load than the wild-type mice. While the underlying mechanisms are not clear, this finding further suggests that ST2 inhibits the host defense mechanisms. Future studies using an RNA sequencing approach may help us find potential antimicrobial targets that are inhibited by ST2 overexpression.

As the current study was focused on the role of ST2 in bacterial and viral infection in nonallergic or non-type 2 inflammatory airways, the research findings from this report do not necessarily apply to the model of allergic or type 2 inflammatory airways. Thus, future experiments are needed to determine if ST2 promotes neutrophilic or even eosinophilic inflammation in M. pneumoniae- or HRV-infected human airway epithelium or mouse lungs exposed to allergens or type 2 cytokines, including IL-33 and IL-13. One limitation to our study is that there are multiple isoforms of ST2, including soluble ST2, and we focused on membrane-bound ST2 function. It is likely that various ST2 isoforms may be regulated during pathogen infection and ultimately determine the outcomes of airway inflammation and infection. It will be interesting to determine how different ST2 isoforms play a role in the host defense against pathogens, especially in asthma.

In summary, our research findings have advanced scientific knowledge about the function of ST2 in human airway and mouse lung defense against two of the relevant pathogens (M. pneumoniae and HRV) in asthma pathogenesis. Revealing the functions of ST2 in airway infections will provide insights into their contribution to asthma and other lung diseases and further support the concept of targeting excessive IL-33/ST2 signaling to reduce infection-mediated airway obstruction and disease exacerbations.

MATERIALS AND METHODS

HRV and M. pneumoniae preparation.

Human rhinovirus 1B (HRV1B) was purified and titrated in H1-HeLa cells as described previously (43). M. pneumoniae (strain FH; American Type Culture Collection 15531) was cultured in SP-4 broth for 6 days at 35°C. The adherent M. pneumoniae was harvested and prepared as previously reported (44, 45).

Lentivirus-mediated overexpression of ST2 in human primary airway epithelial cells.

To isolate HTBE cells, we obtained a lung from a deidentified human organ donor whose lung was not suitable for transplantation and was donated for medical research through the National Disease Research Interchange (Philadelphia, PA). There was no history of clinical lung diseases. The Institutional Research Board (IRB) at National Jewish Health approved this research.

Lentivirus-mediated ST2 overexpression (OE) was performed in HTBE cells using reagents from Genecopoeia (Bethesda, MD). ST2 cDNA or a scrambled control (SC) cDNA sequence was cloned into a lentiviral vector (pReceiver-Lv205) that contains a green fluorescent protein (GFP) reporter gene (IRES2-eGFP) and a puromycin resistance gene. ST2 or SC plasmid was transfected into 293FT packaging cells using a Lenti-Pac HIV Expression Packaging kit (Genecopoeia). After 48 h of transfection, the supernatants containing the packaged virus were harvested and used to transduce HTBE cells in six-well plates, as we described previously (46). Seventy-two hours after transduction, cells were passed into 12-well culture plates at 1 × 10⁵ cells/well for submerged culture. After 24 h, cells were treated with HRV1B (1 × 10⁵ PFU/well), M. pneumoniae (1 × 10⁵ CFU/well), or medium. Supernatants and cells were harvested at 24 h and 72 h postinfection.

HRV1B and M. pneumoniae infection in mice.

Wild-type (WT) BALB/c mice were obtained from Jackson Laboratories and then bred at the NJH Biological Resources Center. ST2 knockout (KO) mice were initially generated by the McKenzie group (47) using the replacement vector targeting ST2 gene exons 4 and 5. Although the initial genetic background of ST2 KO mice was 129×C57BL/6, in our previous publication the mice were then backcrossed to BALB/c mice for more than 8 generations (48). Thus, BALB/c mice were used as the wild-type control mice in the current study. All of the experimental protocols were approved by the Institutional Animal Care and Use Committee (IACUC) at National Jewish Health. WT and ST2 KO (8 to 12 weeks of age) mice were anesthetized by intraperitoneal injection of ketamine (70 mg/kg) and xylazine (10 mg/kg) and were intranasally inoculated with 50 μl of phosphate-buffered saline (PBS) as a control, with HRV1B at 1 × 10⁷ PFU per mouse, or with M. pneumoniae at 1 × 10⁷ CFU per mouse (49, 50). Mice were sacrificed at 16 h (peak of viral replication) and 72 h (resolution of viral infection) post-HRV1B infection, while those infected with M. pneumoniae were sacrificed at 24 h (peak of inflammation) and 72 h (resolution of inflammation) postinfection.

Mouse BAL fluid and lung tissue processing.

Mice were euthanized by intraperitoneal injection of pentobarbital sodium (Fatal-Plus, Vortech Pharmaceuticals, Dearborn, MI) in sodium chloride. Lungs were lavaged with 1 ml of saline solution. Cell-free BAL fluid was stored at −80°C for cytokine analysis. BAL fluid cell cytospin slides were stained with a Diff-Quick stain kit (IMEB, San Marcos, CA) for cell differential counts. Leukocyte differentials were determined as a percentage of 500 counted leukocytes. Left lungs and cell-free BAL fluid were used for RNA extraction and quantitative real-time PCR (RT-PCR) of genes of interest and Western blotting of lactotransferrin.

M. pneumoniae quantification in mouse BAL fluid and lung tissue and in human airway epithelial cell supernatants.

To quantify live M. pneumoniae bacteria existing in mouse lung tissues, half of the left lung lobe was homogenized in 500 μl of PBS. Ten microliters of lung homogenate or BAL fluid was placed onto PPLO (pleuropneumonia-like organism) culture plates (Remel, Lenexa, KS, USA) and incubated at 37°C in 5% CO₂ for 7 days to count CFU (51). M. pneumoniae culture from supernatants (10 μl) of cultured human airway epithelial cells was similarly performed.

Quantitative real-time PCR.

For cell culture experiments, the supernatant was removed, and 350 μl of RLT lysis buffer (Qiagen) was added to cells. Using a pipette tip, the cells were scraped to ensure complete lysis and then collected in a 1.5-ml microcentrifuge tube. Ethanol (70%) was added at 1:1 (vol/vol), and then the mixture was added to Econospin spin columns for RNA (Epoch Life Science) extraction using the protocol provided in the kit. Briefly, the RLT lysis buffer and 70% ethanol mixture in the spin column were centrifuged at 8,000 × g for 30 s. Different wash buffers were added and spun through at the same speed. Then 30 μl of water was used to elute the RNA from the column, some of which was used to quantify the RNA concentration and 500 ng of which was used for cDNA synthesis. The remaining left lung tissue of infected mice was homogenized and utilized for RNA extraction using the TRIzol reagent method. Fifty microliters of BAL fluid from infected mice was added to 350 μl of RLT lysis buffer for RNA extraction using Econospin spin columns for RNA (Epoch Life Science) (52, 53). A total of 500 ng of RNA underwent reverse transcription to produce cDNA using a Bio-Rad T100 thermocycler.

Custom-made gene expression assays for HRV, 16S rRNA, and ST2 obtained from Integrated DNA Technologies were used for quantitative real-time PCR. The specific primers and probe sequences for HRV have been previously reported (43). The specific primers and probes for ST2 are 5′-GGAGAGATATGCTACCTGGAGA-3′ (forward), 5′-GCTCGTAGGCAAACTCCTTAT-3′ (reverse), and 5′-ACCAACATACGAAAGAGCAGGCGG-3′ (probe). The housekeeping gene glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was evaluated as an internal positive control. The comparative cycle threshold (C_T) method (2^−ΔΔCT) was used to demonstrate the relative level of the target genes. To measure HRV load in mouse lung samples, 18S rRNA was used as the housekeeping gene.

ELISA.

Murine KC, IP-10, and IL-33 DuoSet enzyme-Linked immunosorbent assay (ELISA) kits (R&D Systems, Minneapolis, MN) were used to determine the cytokine levels in mouse BAL fluid per the manufacturer’s instructions.

Human IL-8, IP-10, and IL-33 DuoSet ELISA kits (R&D Systems, Minneapolis, MN) were used to measure cytokine levels in cultured human airway epithelial cell supernatants. Due to low protein concentrations of IL-33, supernatants were concentrated 10-fold using Corning Spin-X columns (Sigma, St. Louis, MO).

Western blot analysis.

Western blotting of ST2 was performed to confirm its overexpression in HTBE cells infected with ST2-expressing lentivirus. Cells were lysed in radioimmunoprecipitation assay (RIPA) lysis buffer (Fisher Scientific, Hampton, NH) with a protease inhibitor cocktail (ThermoFisher Scientific, Waltham, MA). Equal amounts of total protein were separated on 10% SDS-polyacrylamide gels, transferred onto nitrocellulose membrane, blocked with Western blocking buffer, and incubated with a goat anti-human ST2 antibody (R&D Systems, Minneapolis, MN) or an anti-β-actin antibody (Santa Cruz Biotechnology, Dallas, TX) overnight at 4°C. After washes in PBS with 0.1% Tween 20, the membranes were incubated with the appropriate horseradish peroxidase (HRP)-linked secondary antibodies and developed using a Fotodyne imaging system (Fotodyne, Inc., Hartland, WI).

Western blotting of lactotransferrin in BAL fluid was performed to quantify its levels as we previously reported (51). In brief, an equal volume (20 μl) of BAL fluid from each sample was electrophoresed by 12% SDS-polyacrylamide gel electrophoresis, transferred onto a nitrocellulose membrane, blocked with Western blocking buffer, and incubated with a rabbit anti-mouse lactotransferrin antibody (Millipore Sigma, Burlington, MA) overnight at 4°C. After washes in PBS with 0.1% Tween 20, the membranes were incubated with an anti-rabbit IgG conjugated to horseradish peroxidase. Densitometry was performed using NIH ImageJ software. The densitometric values (raw density) in equal volumes of BAL fluid sample from each mouse were used to indicate lactotransferrin levels (51). Membranes were also stained with Ponceau S solution to indicate that similar levels of protein (as indicated by the 68-kDa albumin) were loaded.

Statistical analysis.

For normally distributed data, analysis of variance (ANOVA) was used for multiple comparisons, and a Tukey’s post hoc test was applied where appropriate. Student's t test was used when only two groups were compared. For nonparametric data, a Kruskal-Wallis test was used. For all correlations, a nonparametric Spearman correlation test was used. A P value of <0.05 was considered significant.