TRPC6-dependent Ca2+ signaling mediates airway inflammation in response to oxidative stress via ERK pathway

Ozone (O3) plays an extremely important role in airway inflammation by generating reactive oxygen species (ROS) including hydrogen peroxide, then promoting redox actions and causing oxidative stress. Evidences indicate that TRPC6 (canonical transient receptor potential channel 6) is a redox-regulated Ca2+ permeable nonselective cation channel, but its role in the setting of oxidative stress-related airway inflammation remains unknown. Here, we found that both TRPC6−/− mice and mice pretreated with SAR7334, a potent TRPC6 inhibitor, were protected from O3-induced airway inflammatory responses. In vitro, both knockdown of TRPC6 expression with shRNA and TRPC6 blockage markedly attenuated the release of cytokines IL-6 and IL-8 induced by O3 or H2O2 in 16HBE cells (human bronchial epithelial cell line). Treatment with O3 or H2O2 enhanced TRPC6 protein expression in vivo and vitro. We also observed that TRPC6-dependent increase of intracellular Ca2+ concentration ([Ca2+]i) was triggered by H2O2, which consisted of the release from intracellular calcium store and the influx of extracellular Ca2+ and could be further strengthened by 6-h O3 exposure in both 16HBE cells and HBEpiCs (primary human bronchial epithelial cells). Moreover, we confirmed that the activation of MAPK signals (ERK1/2, p38, JNK) was required for the inflammatory response induced by O3 or H2O2 while only the phosphorylation of ERK pathway was diminished in the TRPC6-knockdown situation. These results demonstrate that oxidative stress regulates TRPC6-mediated Ca2+ cascade, which leads to the activation of ERK pathway and inflammation and could become a potential target to treat oxidative stress-associated airway inflammatory diseases.


Introduction
Abnormal airway inflammation resulting from exposure to various oxidizing ambient pollutants is one of the most common and significant pathogenesis for numerous respiratory diseases, including asthma, chronic obstructive pulmonary disease (COPD), and lung cancer. Ambient pollutants, such as inhalable dusts, particulate matter (PM), tobacco smoke and ozone (O 3 ), have strong ability to generate reactive oxygen species (ROS), accelerate redox actions and trigger oxidative stress. As the first line of defense and major target of inhaled harmful environmental pollutants, bronchial epithelium produces a series of proinflammatory molecules and recruits inflammatory cells into interstitium and airways after suffering oxidative stress, which could further aggravate airway inflammation. However, the molecular mechanisms by which oxidative air pollutants trigger pulmonary inflammation are still elusive.
Ca 2+ , an essential secondary messenger relevant to a variety of cellular processes, plays a key role in mediating airway inflammatory responses. Rises in [Ca 2+ ] i in pulmonary cells are essential for the activation of inflammatory signal transduction proteins and transcriptions factors [1][2][3] . Dysregulation of [Ca 2+ ] i homeostasis in bronchial epithelia contributes to pulmonary disease [4][5][6] . In addition, ROS has been found to be responsible for the activity of various calcium channels 7 . TRPM2, a plasma membrane Ca 2+ -permeable channel, mediates ROSinduced chemokine production in monocytes 8 . Compared with WT mice, TRPM2 −/− mice exhibits enhanced gastric inflammation after infecting with Helicobacter pylori, which is owing to intracellular calcium overloading and augmented oxidative stress 9 . These findings suggest that the abnormality of [Ca 2+ ] i suffered from ROS in pulmonary cells may be involved in airway inflammation.
TRPC6, a Ca 2+ -permeable non-selective cation channel of the canonical transient receptor potential (TRPC) family, is widely expressed in a number of tissues including brain, heart, lung, ovary, kidney, and vascular tissues 10 . Consistent with its broad expression in lungs, including bronchial epithelial cells, alveolar macrophages and pulmonary vasculature [11][12][13] , TRPC6 contributes to pulmonary disorders, such as cystic fibrosis, asthma, pulmonary hypertension, COPD, lung edema, and lung fibrosis 11,14,15 . Via analyzing the TRPC6 gene promoter of pulmonary artery smooth muscle cells from patients with idiopathic pulmonary arterial hypertension (IPAH), three single-nucleotide polymorphisms are identified and one of them are found to increase basal gene promoter activity, which may link abnormal transcription of TRPC6 to the activation of NF-κB and lead to upregulated risk of IPAH 16 . The expression of TRPC6 mRNA in alveolar macrophages isolated from COPD patients is significantly more than healthy controls 12 . Particularly, as a modulator of membrane calcium currents, TRPC6 is newly considered as an essential element in the regulation of inflammatory response 17 . TRPC6 channels have been reported to regulate CXCR2-related chemotaxis via mediating calcium supply 18 . After the activation of TLR4 and generation of DAG, TRPC6-dependent Ca 2+ influx into endothelial cells is triggered and cooperated in endotoxin-induced lung inflammation 19 . Moreover, growing evidence points out that TRPC6 acts as a redoxrelated channel, while the definite relation between TRPC6 and ROS seems to be affected by cell specific difference [20][21][22][23] . Recently, we reported that TRPC6 is a key element in the regulation of adhesion of neutrophils to bronchial epithelial cell with O 3 exposure 24 , while the role and regulatory mechanisms of TRPC6 channel in oxidative stress-induced airway inflammation are still unclear.
Here, we investigated the relevance of TRPC6 in O 3induced airway inflammation in mice and inflammatory response in bronchial epithelial cells. We further explored the involved underlying mechanisms to extrapolate the potential of TRPC6 as target to treat oxidative stressassociated airway inflammation.

TRPC6 is required for O 3 -induced airway inflammatory responses in mice
We sought to investigate the function of TRPC6 in O 3 -induced airway inflammatory response by employing TRPC6 −/− mice and SAR7334, a TRPC6-selective inhibitor. As shown in Figs. 1, 2, both TRPC6 −/− mice and mice pretreated with SAR7334 failed to respond to O 3 exposure entirely as exhibiting mild airway inflammatory response. TRPC6-deficiency as well as SAR7334 significantly inhibited O 3 -induced inflammatory cell recruitment in BAL fluid, as reflected by the reduced numbers of neutrophils, macrophages and lymphocytes but not eosinophils, compared with that in WT + O 3 group or PBS + O 3 group (Figs. 1a, 2a). Not only TRPC6 −/− mice but also mice pretreated with SAR7334 had lower total protein, IL-6, IL-8, and TNF-α content in BAL fluid than WT mice had after O 3 exposure (Figs. 1b, c, 2b, c). The O 3 -induced increased lung inflammation scores and inflammatory changes of lung sections were also significantly inhibited by TRPC6deficiency or SAR7334 (Figs. 1d, e, 2d, e). Given the above, TRPC6 contributes to the development of O 3 -induced airway inflammation in mice.
Ca 2+ signal and TRPC6 is required for oxidative stressinduced inflammatory responses in human bronchial epithelial cells Ahead of studying the effect of Ca 2+ and TRPC6 in oxidative stress-induced inflammatory response, we applied O 3 (100 ppb) exposure on 16HBE cells and found that it made no difference to cell viability when it lasted for ≤12 h (Fig. S1a). The releases of IL-6 and IL-8 increased after exposure for 6 h and further augmented till 24 h post-exposure, but the release level of TNF-α remained unchanged (Fig. 3a). The levels of O 3 -induced production of IL-6 and IL-8 were significantly attenuated by removal of extracellular Ca 2+ (Fig. 3b), which indicated that the influx of extracellular Ca 2+ was a key step in O 3induced releases of inflammatory cytokines. We surmised that TRPC6-mediated O 3 -induced influx of extracellular Ca 2+ as it is a potent monitor of membrane calcium currents. TRPC6 mRNA and protein expression in 16HBE cells was strikingly reduced after transducing with TRPC6 shRNA (shTRPC6) while that in the non-silenced negative control (NC) shRNA group had no difference with that in the control group (Fig. S2.). In parallel, Larixyl Acetate (LA), a potent and specific blocker of TRPC6 channels 25 , was administrated at 1, 5, 10 μM concentrations. Importantly, O 3 -induced releases of IL-6 and IL-8 were significantly reduced in shTRPC6-treated or LAtreated cells (Fig. 3c, d), suggesting an involvement of TRPC6 in O 3 -induced production of IL-6 and IL-8 in 16HBE cells.
Hydrogen peroxide (H 2 O 2 ), an oxidant generated during exposure to ambient oxidizing pollutants 26 , is deemed as an intermediate capable of exerting some of its biological effects owing to its diffusibility over membranes and longer half-life than most other ROS 27 . Thus, H 2 O 2 was used in current study to further explore the mechanism regarding O 3 -induced inflammatory response. The viability of 16HBE cells was not affected Fig. 1 Effect of TRPC6-deficiency on O 3 -induced airway inflammation. WT and TRPC6 −/− mice were exposed to O 3 (1 ppm) for 3 h every other day (day 1, 3, 5).The mice were anesthetized 24 h after the last exposure. a-c Total white blood cell counts (Total WBC), macrophage counts (Mac), neutrophil counts (Neu), lymphocyte counts (Lym), eosinophil counts (Eos) (a), total protein content (b) and the release of inflammatory mediators IL-6, IL-8, TNF-α (c) in BAL fluid of different groups were compared. d Inflammation scores in air control and O 3 -exposed mice. e Representative histological sections of mouse lungs (H&E staining) after exposure to O 3 or air. Black arrows: inflammatory changes. Scale bar: 100 μm. Results are presented as mean ± SEM, n = 8. *P < 0.05 or **P < 0.01 compared with WT + Air group, ## P < 0.01 compared with WT + O 3 group. Fig. 2 Effect of TRPC6-blockage on O 3 -induced airway inflammation in mice. WT mice were exposed to O 3 (1 ppm) for 3 h every other day (day 1, 3, 5). Mice received PBS or SAR7334 by oral gavage 4 h before exposure. The mice were anesthetized 24 h after the last exposure. a-c Total white blood cell counts (Total WBC), macrophage counts (Mac), neutrophil counts (Neu), lymphocyte counts (Lym), eosinophil counts (Eos) (a), total protein content (b) and the release of inflammatory mediators IL-6, IL-8, TNF-α (c) in BAL fluid of different groups were compared. d Inflammation scores in air control and O 3 -exposed mice. e Representative histological sections of mouse lungs (H&E staining) after exposure to O 3 or air. Black arrows: inflammatory changes. Scale bar: 100 μm. Results are presented as mean ± SEM, n = 8. *P < 0.05 or **P < 0.01 compared with PBS + Air group, ## P < 0.01 compared with PBS + O 3 group, && P < 0.01 compared with SAR (5 μM) + O 3 group. by H 2 O 2 treatment with ≤100 μM of dose and ≤24 h of time (Fig. S1b, c). Similar with O 3 treatment, H 2 O 2 increased the releases of inflammatory mediators IL-6 and IL-8 in a concentration-dependent and timedependent manner but not that of TNF-α (Fig. 3e). Consistently, absence of extracellular Ca 2+ , transfection with shTRPC6 or pretreatment with LA significantly reduced the release levels of IL-6 and IL-8 evoked by H 2 O 2 ( Fig. 3f-h), which indicated that Ca 2+ signal and TRPC6 contributed to H 2 O 2 -induced production of IL-6 and IL-8 in 16HBE cells. Together, these results suggest that TRPC6 mediates oxidative stress-induced inflammatory responses in human bronchial epithelial cells.

TRPC6 expression is increased by oxidative stress in vivo and vitro
Since TRPC6-mediated oxidative stress-induced inflammatory responses, we next assessed TRPC6 protein expression in lungs of O 3 -exposed mice and human bronchial epithelial cells administrated with O 3 or H 2 O 2 . Indeed, repeated O 3 exposure increased TRPC6 protein expression in lungs (Fig. 4a). Lung section IHC Fig. 3 Role of Ca 2+ and TRPC6 in oxidative stress-induced inflammatory responses in 16HBE cells. a Release levels of IL-6, IL-8, and TNF-α were detected after 16HBE cells were exposed to O 3 (100 ppb) for 0, 3, 6 h (a, Left) or cultured in fresh atmosphere for another 12, 24 h following 6-hour O 3 (100 ppb) exposure (a, Right). Data represent the mean ± SEM, n = 5. **P < 0.01 compared with the Control group. b 16HBE cells incubated with or without Ca 2+ -free bath solution containing 100 μM EGTA were exposed to O 3 (100 ppb) for 6 h and then cultured in fresh atmosphere for another 24 h. Release levels of IL-6 and IL-8 were detected. Data represent the mean ± SEM, n = 5. **P < 0.01 compared with Control group, ## P < 0.01 compared with O 3 group. c After NC or shTRPC6 infection, 16 HBE cells were exposed to O 3 (100 ppb) for 6 h and then cultured in fresh atmosphere for another 24 h. Release levels of IL-6 and IL-8 were detected. Data represent the mean ± SEM, n = 5. **P < 0.01 compared with Control group, ## P < 0.01 compared with O 3 group, && P < 0.01 compared with NC + O 3 group. Con: Control, NC : Negative Control, shT6: shRNA TRPC6. d After pretreatment with LA (0, 1, 5, 10 μM) for 1 h, 16HBE cells were exposed to O 3 (100 ppb) for 6 h and then cultured in fresh atmosphere for another 24 h. Release levels of IL-6 and IL-8 were detected. Data represent the mean ± SEM, n = 5. **P < 0.01 compared with Control group, ## P < 0.01 compared with O 3 group. e Release levels of IL-6, IL-8, and TNF-α were detected after 16HBE cells were stimulated by H 2 O 2 (0, 10, 50, 100 μM) for 12 h (a, Left) or H 2 O 2 (100 μM) for 0, 3, 6, 12, 24 h (a, Right). Data represent the mean ± SEM, n = 5. **P < 0.01 compared with Control group. f 16HBE cells incubated with or without Ca 2+ -free bath solution containing 100 μM EGTA were exposed to H 2 O 2 (100 μM) for 24 h. Release levels of IL-6 and IL-8 were detected. Data represent the mean ± SEM, n = 5. **P < 0.01 compared with Control group, ## P < 0.01 compared with H 2 O 2 group. g After NC or shTRPC6 infection, 16HBE cells were exposed to H 2 O 2 (100 μM) for 24 h. Release levels of IL-6 and IL-8 were detected. Data represent the mean ± SEM, n = 5. **P < 0.01 compared with Control group, ## P < 0.01 compared with H 2 O 2 group, && P < 0.01 compared with NC + H 2 O 2 group. Con: Control, NC: Negative Control, shTRPC6: shRNA TRPC6. h After pretreatment with or without LA (1, 5, 10 μM) for 1 h, 16HBE cells were exposed to H 2 O 2 (100 μM) for 24 h. Release levels of IL-6 and IL-8 were detected. Data represent the mean ± SEM, n = 5. **P < 0.01 compared with Control group, ## P < 0.01 compared with H 2 O 2 group. Fig. 4 Effect of oxidative stress on TRPC6 expression in lung tissue and human bronchial epithelial cells. a, b WT mice were exposed to O 3 (1 ppm) for 3 h every other day (day 1, 3, 5) and were anesthetized 24 h after the last exposure. TRPC6 expression in lung tissues from mice were analyzed with western blot (a) and immunohistochemistry-stained lung sections (b). Black arrows: epithelial cells and macrophages. Scale bar: 100 μm. Data represent the mean ± SEM, n = 8. **P < 0.01 compared with Air group. c, d TRPC6 expression in 16HBE cells was analyzed by western blot after exposure to O 3 (100 ppb) (c) or H 2 O 2 (100 μM) (d) for 0, 2, 4, 6, 8 h. Data represent the mean ± SEM, n = 5. *P < 0.05 or **P < 0.01 compared with 0 h group. e TRPC6 expression in 16HBE cells and primary HBEpiCs was analyzed by immunofluorescence after exposure to O 3 (100 ppb) or H 2 O 2 (100 μM) for 6 h. Scale bar: 10 μm. n = 5. also showed an enhanced TRPC6 protein expression after O 3 exposure, mainly localized in bronchial epithelial cells and macrophages (Fig. 4b). As shown in Fig.  4c, d, TRPC6 protein expression was increased and reached a maximum at 6 h by O 3 (100 ppb) exposure or H 2 O 2 (100 μM) stimulation in 16HBE cells. In view of reports highlighting the differences between primary cells with hTERT or viral genes transduced human cells 28,29 , primary bronchial epithelial cells (HBEpiCs) were employed in parts of our study to further confirm the results obtained from 16HBE cells. Consistently, in both 16HBE cells and HBEpiCs, the expression of TRPC6 protein located in both cytomembrane and cytoplasm was upregulated after O 3 (100 ppb) or H 2 O 2 (100 μM) exposure for 6 h (Fig. 4e, f).

TRPC6 mediates H 2 O 2 -induced increase of intracellular calcium in human bronchial epithelial cells
Previous studies have shown that oxidative stress leads to the disruption of intracellular calcium homeostasis which is capable of activating signal pathway associated with inflammatory response 7,30 . Therefore, we next aimed to study the effect of H 2 O 2 on the intracellular calcium homeostasis in bronchial epithelial cells and the role of TRPC6 in this progress. Both 16HBE cells and HBEpiCs were stimulated with different concentrations of H 2 O 2 and calcium-dependent fluorescence was monitored. Calcium imaging results showed that H 2 O 2 -induced intracellular calcium ([Ca 2+ ] i ) increase in a concentration-dependent manner (Fig. 5a, (Fig. 5a). Intriguingly, compared with 16HBE cells, HBEpiCs were more sensitive to oxidative stress that 10 μM H 2 O 2 was capable to trigger this increase (Fig. 5b). To . a, b 16HBE cells (a) and primary HBEpiCs (b) were exposed to variable concentrations of H 2 O 2 (0, 10, 100, 1000 μM). * P < 0.05 or ** P < 0.01 compared with HBSS control group. c, d 16HBE cells (c) and primary HBEpiCs (d) incubated with or without Ca 2+ -free bath solution (100 μM EGTA) were exposed to H 2 O 2 (100 μM). ** P < 0.01 compared with HBSS control group, ## P < 0.01 compared with H 2 O 2 group. e, f After pretreatment with or without Larixyl Acetate (LA, 10 μM) for 30 min, 16HBE cells (e) and primary HBEpiCs (f) were exposed to H 2 O 2 (100 μM). * P < 0.05 or ** P < 0.01 compared with HBSS control group, # P < 0.05, ## P < 0.01 compared with H 2 O 2 group. g After NC or shTRPC6 infection, 16HBE cells were exposed to H 2 O 2 (100 μM). ** P < 0.01 compared with HBSS control group, ## P < 0.01 compared with H 2 O 2 group, && P < 0.01 compared with NC + H 2 O 2 group. NC: Negative Control , shTRPC6: shRNA TRPC6. Each trace represents the mean from three independent experiments that were derived from 20 to 40 cells in each single experiment.
analyze the source of increased [Ca 2+ ] i , Ca 2+ -free buffer (containing 5 mM EGTA) was applied. Notably, the initial increase of [Ca 2+ ] i (phase Ι) induced by H 2 O 2 (100 μM) was not influenced while the late [Ca 2+ ] i increase (phase II) was significantly inhibited (Fig. 5c, d), suggesting that 100 μM H 2 O 2 -induced increase of [Ca 2+ ] i was composed by intracellular calcium store release (phase Ι) and extracellular Ca 2+ influx (phase II), mainly the latter. Moreover, both the phase Ι and phase II of the H 2 O 2induced [Ca 2+ ] i increase were attenuated by pretreatment with the TRPC6 inhibitor LA (10 μM) and the inhibition of phase II was more salient (Fig. 5e, f) We then speculated that O 3 exposure made the cells more sensitive to oxidative stress which resulted in more serious inflammatory responses and oxidative injury. As shown in Fig. 6a (Fig. 6c). These results suggest that TRPC6 has a major role in O 3 -induced augmenting sensitivity to oxidative stress in bronchial epithelial cells.
Then we sought to further address the source of 6-h O 3 exposure-induced augment of [Ca 2+ ] i increase. By incubating cells with Ca 2+ -free bath solution containing 100 μM EGTA, we found that O 3 exposure made no difference to H 2 O 2 -induced release of intracellular calcium storage (Fig. 6d, e) but enhanced the influx of extracellular Ca 2+ when the intracellular calcium storage was depleted by pretreatment with thapsigargin (Tg, 2 μM) (Fig. 6f, g). Therefore, O 3 -amplified sensibility to oxidative stress via TRPC6 is mainly dependent on enhancing the influx of extracellular Ca 2+ in the bronchial epithelial cells.

MAPK signal pathway contributes to oxidative stressinduced inflammatory response in bronchial epithelial cells
It has been established that MAPK (ERK, p38, JNK) was involved in oxidative stress-induced inflammatory response in lungs 31 . Thus we investigated whether the inflammatory response in this research was mediated by MAPK signal pathway. Results unveiled that after exposure to O 3 (100 ppb) or H 2 O 2 (100 μM) for 15 or 30 min, the phosphorylation levels of ERK1/2, p38 and JNK signaling pathways were significantly increased in 16HBE cells (Fig. 7a-c). Pretreatment with PD98059 (ERK inhibitor), SB203580 (p38 inhibitor) or SP600125 (JNK inhibitor) inhibited O 3 or H 2 O 2 -augmented release of inflammatory factors IL-6 and IL-8 in 16HBE cells in a dose-dependent manner (Fig. 7d-f), suggesting these MAPK signals participated in O 3 or H 2 O 2 -induced inflammatory response. These data demonstrate that MAPK signal pathway is responsible for oxidative stress-induced inflammatory response in bronchial epithelial cells.

TRPC6 is required for the oxidative stress-induced activation of ERK pathway
We next investigated the role of TRPC6 in oxidative stress-induced activation of MAPK pathways. Deficiency of TRPC6 by shTRPC6 significantly reversed O 3 or H 2 O 2 -induced increase of phosphorylation levels of ERK but made no difference to those of p38 and JNK (Fig. 8a,  b), indicating that TRPC6-mediated oxidative stressinduced activation of ERK pathway rather than p38 or JNK in bronchial epithelial cells. Additionally, pretreatment with MAPK inhibitors (PD98059, SB203580, or SP600125) did not affect the H 2 O 2 -induced [Ca 2+ ] i increase (Fig. 8c), which suggested that TRPC6-mediated [Ca 2+ ] i increase was not dependent on the MAPK pathways. Taken together, these experiments suggest that ERK acts as the downstream of TRPC6 in oxidative stress-induced inflammatory response.

Discussion
The major findings of our study demonstrated for the first time that TRPC6 acted as an oxidative stress sensor in bronchial epithelium and mediated oxidants-induced inflammatory responses via activating ERK. Exposure to oxidizing air pollutants (e.g. O 3 ) and the following increase of ROS in lungs (e.g. H 2 O 2 ) are key causative factors in the development of chronic inflammatory respiratory diseases. Here we found that O 3 exposure to mice motivated severe airway inflammation and potentiated the expression of TRPC6 protein in lungs, especially in bronchial epithelium and alveolar macrophages. It is well known that exposure to O 3 induces oxidative injury to the respiratory tract, causes inflammatory responses and triggers clinical symptoms of a series of chronic respiratory disease such as asthma, bronchitis and COPD 32,33 . However, the underlying mechanisms of O 3induced oxidative injury have not yet been fully defined. TRPC6, a lipid-dependent membrane protein acting as non-selective cation channels conducting Na + and Ca 2+ , is highly expressed in the lung and most studied in pulmonary diseases among TRPC channels 34,35 . We previously reported that TRPC6 channels contribute to LPSinduced inflammatory response in human bronchial epithelial cells, which implies that TRPC6 may be a therapeutic target in bronchial epithelial inflammation 36 . It has been well described in many cell types and tissues that the activity of TRPC6 channel is redox-sensitive, while it seems to have different phenotype according to different cell types 23,37 . In podocytes, HEK 293T cells and vascular myocytes, ROS not only activates TRPC6 already in the plasma membrane but also upregulates the expression of TRPC6 in the surface [38][39][40][41][42] . Some other studies show that ROS decreases the activity and abundance of TRPC6 protein in mesangial cells 20,43 . However, it has been rarely described whether ROS activate TRPC6 channel, the underlying mechanisms of the following Ca 2+ transportation and the further effects in respiratory system. Here, we found that O 3 exposure led to the increase of TRPC6 protein expression in mice lungs (Fig. 4a, b). H 2 O 2 , a diffusible and ubiquitous second messenger, contributes to the pathogenesis of several respiratory disease and their exacerbations 44 . Generated in lungs during oxidative pollutants exposure and more stable to readily penetrate the cells, H 2 O 2 is adopted in many studies to further investigate the mechanism related to ambient oxidants-induced injury. We found that both O 3 and H 2 O 2 enhanced TRPC6 expression in human bronchial epithelial cells (Fig. 4c-e). TRPC6 −/− mice or mice pretreated with TRPC6 inhibitor SAR7334 exposed to O 3 revealed attenuated recruitment of neutrophils, macrophages and lymphocytes into airway, release of inflammatory factor IL-6, IL-8 and TNF-α in BAL fluid and damage of the lungs (Figs. 1, 2). We also confirmed that activation of TRPC6 was required for the release of cytokines after stimulating with O 3 or H 2 O 2 in HBECs (Fig. 3). Although we did not detect the levels of ROS in vivo after O 3 exposure, it is conceivable that oxidative stress generated by O 3 activates TRPC6 channels and upregulates the expression of TRPC6, which leads to the disruptions of intracellular Ca 2+ homeostasis and triggers inflammatory response based on the results in vitro (Figs.  3, 5, 6). Surprisingly, the levels of released TNF-α increased after O 3 exposure in BAL fluid (Figs. 1c, 2c) but it remained unchanged after stimulating with O 3 or H 2 O 2 in epithelial cells (Fig. 3a, e). We speculated that TNF-α was secreted by other cells but not bronchial epithelial cells.
Although the functional significance of TRPC6mediated Ca 2+ transport system in health and disease has gained widespread attention, its activation mechanism is not completely elucidated yet. H 2 O 2 has been reported to behave pathophysiologically relevantly with a concentration above 100 μM while a much lower concentration (10 μM) displays a nearly maximal activating effect on TRPC6 channel in HEK293T cells 42 . In the present study, we found that H 2 O 2 was able to cause concentration-dependently increase of [Ca 2+ ] i in 16HBE cells and primary HBEpiCs (Fig. 5a, b), which was mainly resulted from the activation of TRPC6 (Fig. 5e-g). Interestingly, H 2 O 2 -induced increase of [Ca 2+ ] i occurred under the concentration of 10 μM in primary HBEpiCs while that occurred under 100 μM in 16HBE cells (Fig. 5a,  b), suggesting primary HBEpiCs were more sensitive to oxidative stress than16HBE cells were.
Intracellular Ca 2+ signal is the most universal and versatile mechanism regulating a wide range of physiological and pathophysiological processes. Store-operated Ca 2+ entry and receptor-operated Ca 2+ entry are pharmacologically and molecularly distinctive Ca 2+ pathways in non-excitable cells. Hence we aimed to illuminate the source of H 2 O 2 -induced increase of [Ca 2+ ] i in human bronchial epithelial cells. In this study we presented that H 2 O 2 triggered [Ca 2+ ] i increase in two ways: release from intracellular Ca 2+ stores and Ca 2+ influx into the cells (Fig. 5c, d). To date, although ample evidence suggests that TRPC6 is a DAG-sensitive receptor-operated Ca 2+ entry channel 45 , studies about the source of TRPC6-mediated Ca 2+ entry generate discordant findings 46 . Indeed, we found that TRPC6-mediated [Ca 2+ ] i increase after H 2 O 2 exposure consisted of the release of intracellular Ca 2+ stores and the influx of extracellular Ca 2+ since both phase I and phase II of H 2 O 2 -induced [Ca 2+ ] i increase were inhibited under the situation of TRPC6 blockage or deficiency (Fig. 5e-g). Moreover, the profound augment of H 2 O 2 -induced influx of extracellular Ca 2+ by O 3 exposure was mainly dependent on TRPC6 channels as the potentiated phase II was abolished by TRPC6 blockage or deficiency (Fig. 6a-c). According to our finding that TRPC6 channels mediated the release of intracellular Ca 2+ stores, we speculated that TRPC6 channels participated in store-operated Ca 2+ entry although the detailed process still needs further exploration.
MAPK signal pathway is implicated in inflammatory response induced by oxidative stress 31 . Furthermore, previous studies show relationships between the activation of TRPC6 and MAPK signals, which seemed to be context-dependent 47,48 . Here, we found that treatments with H 2 O 2 or O 3 evoked the release of inflammatory factors IL-6 and IL-8 via activating MAPKs (ERK, p38, JNK) (Fig. 7). In addition, deficiency of TRPC6 by shTRPC6 significantly reversed the oxidative stressinduced phosphorylation of ERK but made no effect to that of p38 or JNK (Fig. 8a, b). However, pretreatments with MAPKs inhibitors (PD98059, SB203580, SP600125) did not influence the H 2 O 2 -induced increase of [Ca 2+ ] i in 16HBE cells (Fig. 8c), excluding the possibility that MAPK signal pathway acted as the upstream of TRPC6. Therefore, it is suggested that in human bronchial epithelial cells oxidative stress-aroused inflammatory response was mediated by TRPC6 through regulating [Ca 2+ ] i increase and subsequently activating ERK signal pathway.
In summary, this is the first report demonstrating that TRPC6 is an oxidative stress-sensitive channel which can further mediate inflammatory response via ERK pathway in bronchial epithelial cells. Our result provides a mechanistic understanding of how oxidizing air pollutants lead to airway inflammation. Moreover, we revealed a critical role of intracellular calcium homeostasis in the pathogenesis of such diseases and proposed that targeting TRPC6 might provide a novel therapeutic approach to prevent and treat oxidative stress-induced airway inflammation. Total RNAs were obtained from lungs or cultured cells with the TRIzol reagent (15596018; Invitrogen, USA). Real-time PCR was performed with SYBR Premix Taq Kit (DRR081, TaKaRa, CHN) and analyzed with ABI PRISM 7000 Sequence Detection System (Applied Biosystems, USA). All primers were purchased from Invitrogen Corporation (as illustrated in Table 1). The qPCR results were presented as threshold cycle (Ct) value and the relative mRNA quantification was determined using the 2-ΔΔct method GAPDH as the endogenous control and normalized to a control group.

O 3 exposure
Experiments of O 3 exposure in vivo were performed in reference to similar exposure study 50 . Mice were randomly assigned into different groups (eight mice per group) and placed awake in whole-body Plexiglas exposure chamber (0.55 m wide, 0.75 m long, 0.65 m high) to be exposed to O 3 (1 ppm) for 3 h. Another identical exposure chamber was applied for air exposure. Exposure was repeated every other day (day 1, 3, and 5) to optimize airway inflammation and did not induce any weight change in the mice (unpublished data). Mice received SAR7334 (HY-15699; MedChem Express, USA), a TRPC6-selective inhibitor, by oral gavage 4 h before O 3 exposure, as reported by Maier and colleagues that pharmacologically effective concentrations of SAR7334 reached optimum 4 hours after oral administration and maintained for several hours 51 . No animals were excluded from the analysis and no blinding was carried out for animal experiments. Cells were exposed to O 3 (100 ppb) with 5% CO 2 humidified air at 37°C in the incubator. During exposure, cells were placed on the 3D rocking platform and tilted gently to an angle of 10°from the horizontal to each quarter to ensure direct contact with O 3 . At the end of the exposure, cells were relocated to another incubator with fresh atmosphere. Assays were performed 0-, 12-or 24hour after the exposure. Experiments of O 3 exposure in vitro were done referred to similar exposure study 52 . Before O 3 exposure, cells were pre-treated with TRPC6selective inhibitor Larixyl Acetate (LA) (02730595; Sigma-Aldrich, CHN), ERK inhibitor PD98059 (9900; CST, CHN), p38 inhibitor SB203580 (5633; CST, CHN) or JNK inhibitor SP600125 (8177; CST, CHN) for 30 or 60 min. O 3 were generated by HAILEA model HLO-800 ozonizers (HAILEA, CHN) and its concentration within the chamber or incubator was monitored over the exposing period by ambient-air O 3 motors (model 106 L; T2B, USA). The mean concentration of O 3 within the chambers remained 1 ± 0.05 ppm or 100 ± 10 ppb within the incubator during the exposure.

Measurement of inflammation by collection and analysis of BAL fluid or supernatants of cells
Mice were anesthetized by intraperitoneal injection of pentobarbital sodium (50 mg/kg) 24 h after the last exposure. To collect bronchoalveolar lavage (BAL) fluid, the left bronchus was clamped, the trachea cannulated and then the right lung lavaged three times slowly with 0.6 mL ice-cold phosphate-buffered saline (PBS). BAL fluids were centrifuged to isolate cells from samples. The supernatants were collected and stored at −80°C till analysis. To quantify total cell numbers, cell pellets were resuspended with 200 μl PBS and multiplied hemacytometer cell counts excluding red blood cells. Differential cell counts stained by Wright-Giema stain set (D10; The secretion levels of IL-6, IL-8 and TNF-α in the mentioned supernatants as well as cell medium were detected with commercially available enzyme-linked immunosorbent assay (ELISA) kits (NeoBioscience Technology, CHN) according to the instruction manual. Total protein content was determined using a Pierce BCA Protein Assay kit (23225; Thermo Fisher Scientific, USA). The results were expressed as pg/mg protein.

Histological analysis
Following BAL fluid collection, the non-lavaged lungs were cut out and immediately inflated in fresh 4% paraformaldehyde buffer. Paraffin blocks were prepared from dehydrated tissues and histological sections (5 μm) were stained with hematoxylin and eosin (H&E) for evaluation with a light microscopic (Olympus BX51, JPN).
The severity of lung inflammation in peribronchial and perivascular in H&E sections was scored on a scale ranging from 0 to 3 where 0 means no inflammation; 1 means mild inflammation with only a few inflammatory cells around bronchial or vascular wall and in alveolar space; 2 means moderate inflammation with patchy inflammatory cells infiltration or localized inflammation around bronchial or vascular wall and in alveolar space and less than one-third of lung cross-sectional area involved; 3 means severe inflammation with diffuse inflammatory cells infiltration and more than one-third of the lung area involved. The scale was made referred to similar study 53 .
Fluorescence measurement of intracellular free calcium ([Ca 2+ ] i ) 16HBE cells and HBEpiCs grown on confocal dishes were washed three times with fresh HBSS and 200 μl HBSS was added per dish. The fluorescence intensity of Fluo-4 (5 μM, Molecular Probes Eugene, USA) in the cells was recorded by laser scanning confocal microscopy (Leica TCS SP8, GER). Detailed steps were performed as previously described 36 . H 2 O 2 was used to stimulate the cells when the baseline was stable. The [Ca 2+ ] i was expressed as a pseudo-ratio value (F/F 0 ) of the actual fluorescence intensity (F) divided by the average baseline fluorescence intensity (F 0 ). The Ca 2+ -free bath solution contained 100 μM EGTA and no CaCl 2 . Data from 20 to 40 cells were compiled from a single run, and at least three independent experiments were conducted.

Analysis of the expression of TRPC6 protein in the cells by immunofluorescence
The cells cultured on confocal dishes were fixed with 4% formaldehyde in 0.1 M PBS for 15 min and washed three times with 0.1 M PBS. Further details were performed as previously described 36 6. The cells were incubated with primary specific antibodies against TRPC6 (ACC-017; Alomone Labs, ISR) at 4°C overnight and rewarmed for 30 min subsequently. Following 3 washes with PBS, the cells were incubated with fluorescein isothiocyanate (FITC)-conjugated secondary antibody (Life Technology, USA) for another 1 h. The nuclei were stained by the fluorescent DNA-binding dye 136 4′, 6-diamidino-2-phenylindole dihydrochloride (DAPI) (Roche, CHN) for 15 min. Fluorescence pictures were taken under a confocal laser scanning microscopy (SP8; Leica TCS, GER).

Statistical analysis
Statistical analysis was performed with SPSS 13.0 software. Data expressed as mean ± SEM represented at least five independent experiments. Statistical significance was determined using Student's test or one-way analysis of variance followed by ANOVA and post hoc Bonferroni or Dunnett T3 test. Differences were considered statistically significant when the probability value <0.05 or <0.01.