Urban PM2.5 exacerbates allergic inflammation in the murine lung via a TLR2/TLR4/MyD88-signaling pathway

Nevertheless its mechanism has not been well explained yet, PM2.5 is recognized to exacerbate asthma. In the present study, the roles of toll-like receptor (TLR) 2, TLR4 and MyD88, in exacerbation of allergen-induced lung eosinophilia caused by urban PM2.5 was investigated. TLR2-, TLR4-, MyD88-deficient and WT BALB/c mice were intratracheally challenged with PM2.5 +/− ovalbumin (OVA) four times at 2-week intervals. PM2.5 increased neutrophil numbers and KC in bronchoalveolar lavage fluid and caused slight peribronchiolar inflammation in WT mice. However, these changes were attenuated, but not completely suppressed in gene-deficient mice, especially in MyD88−/− mice. In WT mice, PM2.5 + OVA exacerbated OVA-related lung eosinophilia. This exacerbation includes increase of IL-5, IL-13, eotaxin and MCP-3; infiltration of eosinophils into the airway submucosa; proliferation of goblet cells in the airway epithelium; and the production of antigen-specific IgE and IgG1 in serum. All these effects were stronger in TLR2−/− mice than in TLR4−/− mice. In MyD88−/− mice, this pro-inflammatory mediator-inducing ability was considerably weak and lung pathology was negligible. These results suggest that urban PM2.5 may exacerbate allergic inflammation in the murine lung via a TLR2/TLR4/MyD88-signaling pathway. PM2.5-bound trace microbial elements, such as lipopolysaccharide may be a strong candidate for exacerbation of murine lung eosinophilia.

In our previous work using a murine asthma model, we have shown that urban dust collected from the air in Beijing, China exacerbated ovalbumin (OVA)-associated murine lung eosinophilia 17 , as did PM2.5 collected from the air in Shenyang, China 15,16 and PM2.5-rich dust collected from the air in Fukuoka, Japan 18 . However, the precise exacerbating factors contained within urban PM2.5 and the mechanisms of action are not fully understood. Therefore, studies to clarify the mechanisms involved in exacerbation of murine lung eosinophilia are necessary.
Toll-like receptors (TLRs) expressed by antigen presenting cells (macrophages, dendritic cells) and other various cells (e.g., airway epithelial cells) are innate immune sensors, which recognize microbial pathogen-associated molecular patterns (bacteria, fungi, and virus structures) as well as endogenous danger molecules from host cells. TLR2 is a receptor for β-glucans of fungi and peptidoglycans of Gram-positive bacteria 19 , while TLR4 detects LPS 20 . Myeloid differentiation factor 88 (MyD88) is a downstream signalling adapter protein and essential for cytokine production in response to TLR ligands 21 . We have recently reported that MyD88 downstream of TLR2 and TLR4 is a key protein in OVA-induced exacerbation of murine lung eosinophilia by Asian sand dust 22 .
Investigation of the role of TLR2, TLR4 and MyD88 in the exacerbation of allergen-induced lung eosinophilia caused by PM2.5 was conducted to clarify the relationship between microbial elements present in PM2.5 and disease exacerbation in a mouse model of asthma. Gene deficient (TLR2 −/− , TLR4 −/− and MyD88 −/− ) and wild type (WT) BALB/c mice were intratracheally challenged with OVA and/or PM2.5 in the present study.

OVA-specific IgE and IgG1 in serum.
OVA alone caused no induction of OVA-specific IgE and IgG1.

Discussion
The results of the present study indicate that the activation of TLR signalling by PM2.5, which leads to inflammatory cytokine production, may result in activation of the innate and adaptive immunity related to Th2 responses. The results of the present study demonstrated that exposure to PM2.5 alone increased neutrophil numbers in BALF and also increased chemokine KC levels in WT mice. The human homologue of KC, IL-8, recruits and activates neutrophils 23 . However, these alterations were attenuated although not completely suppressed in TLR2 −/− , TLR4 −/− and MyD88 −/− mice. Pathologically, PM2.5 caused slight peribronchiolar inflammation in WT mice, but more subtle pathological changes were observed in the gene-deficient mice, especially in MyD88 −/− animals. These data indicate that PM2.5 is capable of activating the innate immune system in a MyD88-dependent manner. These data also suggest that contaminating trace LPS and TLR2 ligands (β-glucan, etc) in PM2.5 are strong candidates for the active agents in neutrophilic lung inflammation caused by PM2.5. A recent study has reported that PM2.5-induced acute lung inflammation in BALB/c mice might be due, in part, to the production of IL-1β. The activation of the TLR4/MyD88 signaling pathway and NOD-like receptor family, pyrin domain-containing 3 (NLRP3) might be involved in the production process 24 . In the present study, PM2.5 increased IL-1β in WT mice, TLR2 −/− , TLR4 −/− and MyD88 −/− mice. PM2.5 might cause IL-1β induction through the NLRP3 inflammasome-pathway (MyD88 independent-pathway). It is well known that the inflammasome-pathway is activated by danger signals (peptidoglycan, silica, ATP, uric acid crystal) 25 . Regarding TLR2 stimuli, hydrogen peroxide (H 2 O 2 ), which is generated during inflammation, induces nuclear translocation of NF-κB and AP-1, and phosphorylation of p38 MAPK in neonatal rat ventricular myocytes; anti-TLR2 antibody inhibits these effects. H 2 O 2 increases NF-κB activation in TLR2-overexpressing Chinese hamster ovary (CHO) fibroblasts, but not in normal or TLR4-overexpressing CHO cells 26 . As metal-containing TLR stimuli, TiO 2 nanoparticles activate TLR2 and TLR4, and induce expression of TNF-α and NF-κB in mouse hippocampal tissues 27 . Zinc and nickel administration induce inflammatory responses (ICAM-1 and IL-8 induction) in vascular endothelial cells. In these responses, TLR4 plays a dominant role in NF-κB activation by nickel, but not by zinc 28 . In the present study, these metals were included in the tested PM2.5 sample (Ti, 370 ng/mg; Ni, 72 ng/mg). From these reports, we speculate that H 2 O 2 is generated extracellularly by inflammation, Ti and Ni present in PM2.5 trigger lung inflammation via TLRs.
Regarding the relationship between particle size and inflammatory response, overall urban PM2.5-induced inflammatory responses in murine lungs, including those identified in this study, are weaker than those observed in response to coarse PM (PM2.5-PM10μm) or fine particles derived from desert dust (2.5 μm) 15,16 . We hypothesis that the differences in these inflammatory responses may be due to the differences in amounts of particle-bound microbial elements, especially LPS. However, we currently have reported that type II alveolar cells might react sensitively to oxidative stress induced by PM2.5 and caused an inflammatory response because the response was suppressed by N-acetylcystein (anti-oxidants) but not by Polymyxin B (LPS inhibitor) 29 . Therefore, the oxidative stress caused by chemical-rich PM2.5 might play an important role in murine lung inflammation in addition to LPS and metals.
In OVA-treated WT mice, PM2.5 exacerbated OVA-related lung eosinophilia along with an increase in Th2 cytokines (IL-5, and IL-13) and chemokines (eotaxin and MCP-3) and infiltration of eosinophils into the airways. Increasing proliferation of goblet cells in the airway epithelium and production of antigen-specific IgE and IgG1 in serum are exacerbated by allergic inflammatory events. These Th2-derived cytokines and chemokines are key mediators in the symptoms of asthma and are critical for the recruitment and survival of eosinophils 30 , production of antigen-specific antibodies 31 , and the production of mucous cells (e.g., goblet cells) in the bronchial epithelium 32 . All these effects were stronger in TLR2 −/− mice than in TLR4 −/− mice. In MyD88 −/− mice, this pro-inflammatory mediator-inducing ability was very weak and the lung pathology exhibited negligible effect following exposure to the sample mixtures. These results suggest that TLR2 and TLR4 signalling may be important in the exacerbation of PM2.5-induced lung eosinophilia and that MyD88 is a key adapter molecule in this event. Therefore TLR2-and TLR4-ligands, TLR stimuli (H 2 O 2 , metals) and other dinger -stimulus might trigger the exacerbation of lung eosinophilia. However, PM2.5-bound LPS may be a strong candidate for exacerbation of lung eosinophilia caused by PM2.5, indicated by the strong inhibition of effects in TLR4 −/− mice.
We have recently reported that antigen-induced allergic inflammation in murine lungs was greater in microbial element (LPS, β-glucan)-rich coarse PM than in organic chemical (PAHs)-rich PM2.5 and have suggested that microbial elements have more potent exacerbating effects on the development of lung eosinophilia than do organic chemicals contained in PM2.5 16 . We have also reported that heated PM2.5-rich dust-heated at 360 °C to exclude toxic materials, such as microbial and chemical elements-caused less effect on neutrophilic lung inflammation and OVA-induced lung esosinophilia in mice than non-heated PM2.5-rich dust 18 . Thus, PM2.5-bound toxic materials are seen to be a key factor for these lung disease enhancements. Therefore, investigation to confirm the phenomenon found by the present study, whether exposure to a mixture of heated PM2.5 and LPS or β-glucan in trace revels causes exacerbation of murine lung eosinophilia or not, is in order.

Conclusion
This study demonstrates that urban PM2.5 may exacerbate allergic inflammation in the murine lung via a TLR2/ TLR4/MyD88-signaling pathway. PM2.5-bound trace microbial elements, such as LPS may be a strong candidate for exacerbation of murine lung eosinophilia. The results of the present study suggest that inhalation of urban PM2.5 is a significant risk factor for inflammatory and allergic lung diseases.

Methods
Sample collection of PM2.5. PM2.5 samples collected between Jan 20 and Jan 25, 2015 were obtained from China Medial University, Shenyang, China. In China, a massive haze event (density ranges of PM2.5 = 129-291 µg/m 3 ) appeared at this time. The experimental air samples were trapped in a four-stage multi-nozzle cascade impactor (MCI) (Tokyo Dylec Co., Tokyo, Japan). An MCI was used to measure the levels of size-classified mass and elemental concentrations of PM2.5 accordance with a previously reported method [15][16][17][18] . The MCI consisted of three stages with 12-orifice and back-up stage. The PM2.5 was trapped directly on materials placed behind the jet-nozzles of an MCI set at 20 l/min airflow. The PM2.5 collected was transferred to a sterilized dry bottle, which was placed in a germ free desiccator and stored at −30 °C for further use.
Particle size. Figure 5 shows the particle size of the PM2.5 analyzed by KEYENCE all-in-one BZ-9000 fluorescence microscope (Osaka, Japan). A total of 3159 particles were counted. The median diameter of the PM2.5 particles was 0.94 ± 0.65 μm (M ± SD). The peak of the size distribution was at 0.5~1.0 μm.

Analysis of microbial elements in particles.
Kinetic assays (Seikagaku Corp., Tokyo, Japan) were applied to measure microbial elements: Endospec ES tested MK for LPS activity and Fungitec G tested MK for β-glucan. Briefly, one ml water suspension of PM2.5 (5 mg) was allowed to stand on the bench top at room temperature for 2 h. Subsequently, supernatants recovered were examined for LPS and β-glucan concentrations using Pyro Color-MP Chromogenic Diazo-Coupling Kit (Associates of Cape Cod. Inc., MA, USA).

Analysis of polycyclic aromatic hydrocarbons (PAHs) in particles.
A previously reported method [15][16][17][18] was used to analyze PAHs. Briefly, ultrasonic extraction was performed to extract the air samples trapped on the Teflon filter. The samples were extracted twice with a 20 ml portion of dichloromethane at 15 °C. The filtrate obtained from the extract using a No 5 C filter paper was allowed to stand to dryness in the dark, yielding solid residual substances.
The residual substances dissolved in acetonitrile (0.5 ml) were analyzed for PAHs using a Hitachi Model 600 HPLC (Hitachi, Japan). The HPLC was equipped with a Model L-7485 fluorescence detector (Hitachi, Japan) and a 4.0 mmϕ × 250 mm column packed with Wakosil-II 5C 18HG (Waka Pure Chemicals Industry, Ltd., Osaka, Japan). An acetonitrile/water (80/20, v/v) solution was used as a mobile phase at 1.5 ml/min. Identification of unknown PAHs was performed by comparing the HPLC retention time and fluorescence/excitation spectra to those of authentic PAHs 33 from Supelco (Bellefonte, PA, USA) and Aldrich Chemical Co., Inc. (Milwaukee, WI, USA)/Tridom Chemical Inc (Hauppauge, NY, USA).
Animals. Homozygous TLR2, TLR4 and MyD88 knockout mice and WT mice (BALB/c parental strain, males) were purchased from Oriental BioService Japan, Inc. (Kyoto, Japan) at 8 weeks of age. Mice were fed with a commercially obtained diet CE-2 (CLEA Japan, Inc., Tokyo) and given water ad libitum. They were placed in plastic cages lined with soft wood chips and kept in a well-controlled room-temperature, 23 °C; humidity, 55-70%, 12 h/12 h light/dark cycle. The conditions used for the present study are in compliance with the U.S. National Institutes of Health Guidelines for the use of experimental animals. The Animal Care and Use Committee at Oita University of Nursing and Health Sciences (Oita, Japan) also approved the method used.