Effects of Exposure to Incense Smoke on Airway Function: An in Vitro and in Vivo Study

Recent clinical studies have suggested that inhalation of incense smoke (IS) may result in impaired lung function and asthma. However, there is little experimental evidence to link IS with airway dysfunction. Using mouse and cell culture models, we evaluated the effects of IS exposure on airway function, such as airway hyperresponsiveness (AHR), expression of multiple epithelial tight junction (TJ)- and adherens junction-associated mRNAs and proteins in the lungs, and the barrier function of bronchial epithelial cells assessed by transepithelial electronic resistance (TEER). Exposure of BALB/c mice to IS increased AHR and inammatory macrophage recruitment to BALF; reduced claudin-1, -2, -3, -7, -10b, -12, -15, and -18, occludin, zonula occludens-1 [ZO-1], and E-cadherin mRNA expression; and caused discontinuity of claudin-2 and ZO-1 protein immunostaining in lung tissue. IS extract dose-dependently decreased TEER and increased reactive oxygen species production in bronchial epithelial cell cultures. Treatment with N-acetyl-l-cysteine, but not glucocorticosteroids or long-acting β 2 -agonists, prevented the detrimental effects of IS. IS exposure can be problematic for respiratory health, as evidenced by AHR, increased recruitment of inammatory macrophages and disruption of TJ proteins in the lung, and damage to epithelial barrier integrity. However, antioxidants may be useful for the treatment of IS-induced airway dysfunction. cell American Type Manassas, were in Dulbecco’s modied Eagle’s medium/F-12 (Thermo Scientic) with 10% foetal bovine serum and 1% penicillin-streptomycin and maintained at 37 in a humidied of 5% CO 2 in air. The methods for ALI cultures are described in the Supplementary Methods. Preliminary experiments demonstrated that transepithelial electronic resistance of the cells cultured under ALI conditions reached a plateau at about day 8 post-seeding and then decreased to about half-maximal levels at day 21 11


Introduction
Burning of incense in temples and homes is a common religious and cultural practice in many Asian and Middle Eastern countries. Incense smoke (IS) contains particulate matter of varying sizes; gases such as carbon monoxide, nitrogen dioxide and sulfur dioxide; and volatile organic compounds such as benzene, aldehydes and polycyclic aromatic hydrocarbons 1 . Compared with cigarette smoking, incense burning has been reported to generate larger quantities of particulate matter of ≤ 2.5 µm diameter and particles that remain in the air for hours 2,3 . Increasing evidence suggests that ambient air pollution from IS can cause health problems, especially airway dysfunction. Recent studies have reported that indoor exposure to IS increases the risk of wheezing and asthma and is associated with impaired lung function in adolescents [4][5][6][7] . Despite a number of clinical studies, however, there is little experimental evidence that inhalation of IS causes airway dysfunction.
Bronchial epithelial cells form a barrier against a wide range of inhaled substances and are at the front line of mucosal innate immunity. The epithelial barrier function is maintained by apical junctional protein complexes, composed of apical tight junctions (TJs) and underlying adherens junctions (AJs), that form between neighbouring cells 8 . TJ-associated proteins, which include members of the occludin, claudin, and junctional adhesion molecule families, and AJ-associated proteins such as E-cadherin are the major constituents of junctional complexes in the bronchial epithelium 9 . Members of the zonula occludens (ZO) protein family act as scaffolds that link the intracellular domains of TJ and AJ-associated proteins with the cytoskeleton. There is increasing evidence that junctional protein expression and airway barrier function are reduced in patients with asthma 10 . We recently reported that exposure of human bronchial epithelial cells to cigarette smoke in vitro disrupted epithelial barrier function and simultaneously downregulated the expression of multiple TJ and AJ-associated proteins 11 . However, the effects of inhalation of harmful substances on multiple junctional protein expression in animal models is unknown.
In this study, we evaluated the effect of a single exposure of mice to IS on airway responsiveness (AHR), in ammation and multiple TJ and AJ-associated protein expression in the lung, and we additionally analysed the effect of exposure to IS extract (ISE) on epithelial barrier function in human bronchial epithelial cells in air-liquid interface (ALI) cultures. We also investigated whether treatment with glucocorticosteroids (GCSs), long-acting β 2 -agonists (LABAs), or the antioxidant N-acetyl-l-cysteine (NAC) could protect against IS-induced airway dysfunction in vitro and in vivo.

Mice
Six-week-old female BALB/c mice were purchased from Japan SLC (Shizuoka, Japan) and housed under speci c pathogen-free conditions. The study protocol was approved by The Kyushu University Animal Care and Use Committee (A19-200-0). All experiments were performed in accordance with our institutional guidelines and carried out in compliance with the ARRIVE guidelines.

Mouse experiments
The most commonly used incense sticks in Japan were used for the study (0.4 g/stick, Nippon Kodo brand). Mice were randomly assigned to two or three groups depending on the experimental protocol: (i) unexposed, (ii) exposed to a high dose of IS (IS high ), or (iii) exposed to a low dose of IS (IS low ). The mouse groups were housed separately to avoid cross-exposure to IS. Mice were placed in 44-L chambers and exposed to IS generated by burning of 3.2 g (IS high ) or 1.6 g (IS low ) of incense sticks for 60 min. Incense sticks were burnt in a separate 8-L chamber connected by a tube to the 44-L exposure chamber, and IS was drawn into the exposure chamber with fresh air at 4 L/min. Mice from the unexposed group were maintained for the same period in fresh air. The ratio of the exposure chamber volume to the mean body weight of mice was designed to recapitulate the ratio of a standard living room volume to the standard body weight of humans.
In some experiments, mice received intraperitoneal injections of NAC (320 mg/kg, Sigma-Aldrich, St. Louis, MO) at 6 h before and 6 h after the 1-h exposure to IS. NAC was dissolved in phosphate buffer saline (PBS) and neutralised to pH 7.4 with NaOH. ISE preparation ISE for in vitro experiments was prepared by bubbling IS from 1.6 g of incense sticks (Nippon Kodo brand) through 20 mL of culture medium. After adjustment of the pH to 7.4, the ISE was sterile-ltered (0.22-µm pore size, 33-mm diameter Millex GV; Merck Millipore, Billerica, MA). This solution was considered to be 100% ISE and was diluted in medium containing 10% foetal bovine serum before use. ISE preparations were standardized based on absorbance at 320 nm and were freshly prepared for each experiment.

Measurement of AHR
Airway responsiveness was measured according to our previously described protocol 12 . Brie y, mice were anesthetized with a mixture of ketamine and sodium pentobarbital intraperitoneally, and then their tracheas were cannulated via tracheostomy. Mice were ventilated mechanically (tidal volume, 0.3 ml; frequency, 120 breaths/min) after a paralytic agent was administered. The airway opening pressure was measured with a differential pressure transducer and recorded continuously. Stepwise increase in acetylcholine dose (1.25 to 20 mg/ml) were given with an ultrasonic nebulizer (NE-U07; OMRON Co., Kyoto, Japan) for one minute. The data were expressed as the provocative concentration 200 (PC 200 ), the concentration at which airway pressure was 200% of its baseline value. PC 200 was calculated by loglinear interpolation for individual animals as described previously 12  Collection of bronchoalveolar lavage uid (BALF), ow cytometry analysis and enzyme-linked immunosorbent assay (ELISA) After measurement of AHR, mice were euthanised by administration of pentobarbital. For collection of BALF, the lungs were gently lavaged twice with 1 mL of 0.9% saline via a tracheal cannula. BALF was centrifuged at 250 × g for 10 min and the supernatants were stored at − 80 °C until analysed. Total and differential cell counts in BALF were performed as described previously 13 .

Fluorometric TUNEL assay
Apoptotic cells in lung tissue sections were detected using a DeadEnd™ Fluorometric TUNEL System (Promega, Madison, WI) according to the manufacturer's instructions and observed by confocal laser microscopy (LSM700; Zeiss, Jena, Germany). Negative and positive control slides were prepared by omitting the TdT enzyme from the nucleotide mix and by treating tissue sections with DNase I, respectively.
Quantitative reverse-transcription PCR (qRT-PCR) Total RNA was extracted from mouse lungs, reverse-transcribed, and subjected to qPCR as described in the Supplementary Methods. Primer sequences are provided in Supplementary Table S1. Immuno uorescence staining Immuno uorescence staining were performed according to modi ed protocol as described previously 14 . Brie y, freshly isolated lungs were washed in PBS and then lung tissue was embedded using optimal cutting temperature compound and frozen. Four µm tissue sections on microscope slides were xed with cold methanol for 10 min at -20 °C and then blocked with PBS containing 1% BSA for 30 min at room temperature. The tissue sections were incubated with each primary antibody prepared in PBS containing 1% BSA at 4 °C overnight, followed by incubation with Alexa Fluor 488-conjugated goat anti-rabbit IgG antibody (diluted 1:500; Abcam, Cambridge, UK), Alexa Fluor 568-conjugated goat anti-rat IgG antibody (diluted 1:500; Abcam) and nuclear staining with 4′,6-diamidino-2-phenylindole (DAPI). Images of the stained tissue sections were obtained with a confocal laser microscope (LSM700; Zeiss). The primary antibodies were as follows: rabbit anti-claudin-1 polyclonal antibody (

Measurement of TEER
Bronchial epithelial cell layer integrity was evaluated by TEER measurements using a Millicell-ERS 2 V-Ohmmeter (Millipore Co., Bedford, MA). Medium was added to the apical chamber 1 h prior to TEER measurement. The electrode was soaked in 70% ethanol and rinsed with culture medium prior to use. TEER was calculated by the following Eq. 1 6 : TEER (Ωcm 2 ) = (R sample -R blank ) × effective membrane area (cm 2 ).

Viability assay
Cell viability was assessed according to our previously described protocol 11 . After 24 h of ISE exposure, the medium in the apical chamber was removed and Calu-3 cells were washed with PBS, detached with trypsin-EDTA, and stained with 0.4% trypan blue solution. Non-viable cells were counted in a LUNA™ Automated Cell Counter (Logos Biosystems, Annandale, VA).

Reactive oxygen species (ROS) assay
Calu-3 cells were grown in monolayer culture to approximately 50% con uency, incubated with 1 mM NAC or vehicle for 2 h and then incubated with or without 50% ISE for 1 h. Positive control cells were exposed to 200 µM of the ROS inducer pyocyanin (10 mM stock in anhydrous dimethylformamide; Enzo Life Sciences, Farmingdale, NY). Total ROS and superoxide production were evaluated using a ROS-ID® Total ROS/Superoxide detection kit (Enzo Life Sciences) according to the manufacturer's instructions. The cells were then visualised by uorescence microscopy (BZ-X800; KEYENCE, Osaka, Japan).

Statistical analysis
Unless otherwise stated, data are expressed as the mean ± standard error (SEM). The Mann-Whitney Utest was used to compare data between two groups and one-or two-way analysis of variance (ANOVA) followed by Tukey's multiple comparisons test were used to compare three or more datasets. Statistical analyses were conducted with Prism 8 software (GraphPad Software, San Diego, CA). Differences were considered statistically signi cant at p < 0.05.

IS exposure induces airway hyperresponsiveness and in ammation in mouse lungs
To determine the effects of a single exposure period to IS on airway function, groups of mice were exposed to fresh air (unexposed) or high or low doses of IS for 1 h, and AHR, recruitment of in ammatory cells to BALF, and apoptosis of lung cells were assessed over the following 24 h. IS exposure increased AHR, as re ected by PC 200 values, in an acetylcholine dose-dependent manner (Fig. 1a). The number of macrophages, but not neutrophils, lymphocytes or eosinophils, in BALF was also signi cantly increased in IS-exposed compared with unexposed mice (Fig. 1b), and ow cytometry analysis of lung-derived cells showed that IS exposure speci cally increased the Ly-6G low /Ly-6C high population of in ammatory macrophages (Fig. 1c). However, IL-6, IL-1β and TNF-α concentrations in BALF from both unexposed and IS high -exposed mice were below the assay levels of detection ( Supplementary Fig. S1). Analysis of apoptosis in lung sections using a uorometric TUNEL assay revealed no increase in cell apoptosis in the lungs of IS-exposed compared with unexposed mice (Fig. 1d).

Expression of claudin-2 and ZO-1 protein in mouse lungs is disrupted by IS exposure
To determine whether disruption of TJ-and AJ-associated genes was also observed at the protein level, we performed immuno uorescence staining of lung sections from unexposed and IS high -exposed mice. Confocal microscopy revealed predominant staining of claudin-1, claudin-10 and E-cadherin in the bronchial epithelium of unexposed mice, while claudin-2, claudin-15, occludin and ZO-1 proteins were expressed in both bronchial epithelial and alveolar cells (Fig. 3). Notably, lung sections from IS high mice showed discontinuous or reduced staining of claudin-2 and ZO-1 at 24 h after IS exposure, indicating disruption of TJs (Fig. 3).

IS impairs bronchial epithelial barrier integrity
TEER measurements and permeability assays were performed to assess the effects of IS on epithelial barrier function. Calu-3 sub-bronchial epithelial cells were differentiated in ALI cultures for 9 days, and then incubated with 0%, 25%, 50%, or 75% ISE concentrations for an additional 24 h. Treatment with ISE caused a dose-dependent reduction in TEER that was maintained for 24 h post-exposure (Fig. 4a). In contrast to the signi cant reduction in TEER induced by 50% and 75% ISE (Fig. 4b), the permeability and viability of Calu-3 cells was signi cantly decreased at 24 h only by exposure to 75% ISE (Fig. 4c, d). Thus, treatment with 50% ISE impaired epithelial barrier integrity without in uencing cell permeability or viability.

ISE-induced impairment of bronchial epithelial barrier integrity is not affected by GCS or LABA treatment
The bene ts of inhaled GCSs and LABAs in asthma are widely recognized. We previously demonstrated that GCSs protect against epithelial barrier dysfunction induced by cigarette smoke extract 11 . Therefore, we investigated the effects on ISE-induced TEER reduction in Calu-3 cells after pretreatment with vehicle, 10 nM GCSs (FP or BUD), and/or 10 nM LABAs (SAL or FOR) for 2 h before exposure to 50% ISE (plus GCS/LABA) for an additional 24 h. However, we found no effects of GCSs and LABAs, alone or in combination, on ISE-induced reduction in TEER (Fig. 5a-c).
The antioxidant NAC protects against ISE-induced impairment of bronchial epithelial barrier integrity Immuno uorescence staining of Calu-3 cells con rmed that generation of total ROS, including superoxide, could be detected within 1 h of exposure to 50% ISE, and pretreatment of cells with 1 mM NAC for 2 h before ISE exposure diminished ROS generation (Fig. 6a). Similarly, NAC pretreatment strongly suppressed the decrease in TEER observed at 12 and 24 h after Calu-3 cells exposure to 50% ISE (Fig. 6b), suggesting that ROS production was directly involved in IS-induced disruption of epithelial barrier function. Treatment of ISE-exposed or unexposed cells with NAC alone did not affect cell viability ( Supplementary Fig. S2).
NAC treatment prevents disruption of TJ-associated proteins in mouse lungs exposed to IS To determine whether NAC can also protect lungs against IS exposure in vivo, we examined TJ-associated protein expression in lung tissues. Immuno uorescence staining of claudin-2 and ZO-1 was performed on lung sections collected at 24 h after exposure. As noted earlier, IS high exposure caused discontinuous or attenuated staining of claudin-2 and ZO-1 in both bronchial epithelial and alveolar cells (Fig. 7). Although 24 h of NAC treatment alone did not affect the pattern or intensity of protein expression in unexposed mice, NAC treatment effectively prevented the IS high -induced disruption of claudin-2 and ZO-1 immunostaining patterns (Fig. 7).

NAC treatment prevents IS-induced AHR and in ammation in mouse lungs
Finally, we examined the effects of NAC on IS-induced AHR and accumulation of in ammatory macrophages in BALF of mice at 24 h after IS exposure. NAC treatment of IS high mice reversed the effect of IS on acetylcholine-induced AHR to the same levels detected in unexposed mice (Fig. 8a). Similarly, NAC treatment completely abrogated both the increase in macrophage abundance and the proportion of in ammatory macrophages in BALF after exposure to IS high (Fig. 8b, c).

Discussion
In the present study, we showed that a single exposure of mice to IS aggravated AHR, provoked an in ux of in ammatory macrophages into the lungs, and disrupted the expression and location of TJ-associated proteins in bronchial epithelium. In vitro experiments con rmed that exposure of bronchial epithelial cells to ISE induced ROS production and dose-dependently reduced TEER through a mechanism resistant to clinically relevant concentrations of GCSs and/or LABAs. Finally, NAC treatment ameliorated IS-induced effects on AHR, macrophage recruitment and claudin-2 and ZO-1 expression in the murine airway and reversed the effect of ISE on TEER in vitro.
The results in the present study indicate that oxidative stress may be responsible for IS-induced respiratory complications by inducing AHR, disassembly of TJ proteins, and epithelial barrier integrity.
Previous reports suggest that increased ROS may be associated with aggravated AHR and impaired epithelial barrier function. Using animal models, ROS has been shown to contribute directly to AHR via damage to oxidant-sensitive β-adrenergic receptors and increases in vagal tone, and this effect was enhanced when the epithelium was injured [17][18][19] . Studies of cigarette smoke exposure in bronchial epithelium suggest that ROS production after exposure may induce disassembly of TJ proteins and impair epithelial barrier function through epidermal growth factor receptor (EGFR)-extracellular signalregulated kinase 1/2 signaling pathway 20,21 . Furthermore, Heijink et al. reported that the protective effect of GCSs against cigarette smoke induced-epithelial barrier dysfunction was mediated by affecting EGFRdownstream target glycogen synthase kinase-3β 22 . We also showed that cigarette smoke-induced reduction in TEER (approximately 35% compared with untreated control) was signi cantly attenuated in Calu-3 cells treated with 10 nM BUD (15.6% TEER reduction compared with untreated control) or 10 nM FP (9.7% TEER reduction compared with untreated control) 11 . However, in the present study, IS-induced reduction in TEER (approximately 35% compared with untreated control) was affected by neither BUD nor FP. Together with those observations, our present results suggest that IS-induced ROS may impair epithelial barrier integrity through a different signal transduction pathway activated by cigarette smoke exposure and IS-induced barrier dysfunction might be resistant to treatment with inhaled GCSs.
In this study, we showed that IS exposure disrupted claudin-2 and ZO-1 and simultaneously downregulated the gene expression of multiple TJ-associated proteins in the lung. Bronchial epithelium from patients with asthma has also been shown to exhibit an irregular ZO-1 staining pattern and reduced barrier function that is further compromised by exposure to the Th2 cytokines IL-4 and IL-13 23,24 .
Although claudin-2 is expressed in healthy human bronchiolar and alveolar cells, its precise function in lung physiology is unknown 25 . A limitation of our study is that the consequences of claudin-2 and ZO-1 disruption by IS exposure remain unclear. Further studies are needed to clarify the impact of these changes on the maintenance of airway responsiveness and epithelial barrier function.
In conclusion, our results suggest that inhalation of IS might be harmful to respiratory health, as evidenced by AHR, increased recruitment of in ammatory macrophages and disruption of TJ-associated proteins in the lung, and damage to epithelial barrier integrity. However, treatment with NAC reversed many of the detrimental effects of IS exposure, suggesting that antioxidants may be bene cial for the treatment of IS-related airway dysfunction. Figure 2 Effects of IS exposure on TJ-and AJ-associated gene expression in mouse lungs. qRT-PCR analyses of the indicated mRNAs in lung tissues was performed at up to 24 h after a 1-h exposure to fresh air (unexposed) or IShigh. mRNA levels were normalised to GAPDH mRNA levels. Data represent the mean ± SEM (n=6-9 per group) and are representative of at least two independent experiments. *p<0.05, **p<0.01, by two-way ANOVA. IS, incense smoke. IS exposure-induced changes in TJ and AJ-associated protein expression in mouse lungs. Confocal immuno uorescence microscopy of ZO-1 (red) and additional AJ and TJ-associated proteins (green) in lung sections was performed at 24 h after a 1-h exposure to fresh air (unexposed) or IShigh. DAPI staining of nuclei is shown in blue. Scale bar, 100 μm. Results are representative of at least two independent experiments. DAPI, 4′,6-diamidino-2-phenylindole; IS, incense smoke. Effects of GCS and/or LABA treatment on ISE-induced reduction in TEER in Calu-3 cells. Cells were cultured under ALI conditions and pretreated with vehicle, 10 nM GCS and/or 10 nM LABA for 2 h prior to addition of vehicle or 50% ISE for 24 h. a. Cells incubated with or without 10 nM of the GSCs FP and BUD.

Figure 6
Effects of NAC on ISE-induced ROS generation and TEER reduction in Calu-3 cells. Cells were pretreated with 1 mM NAC or vehicle for 2 h and then incubated with vehicle or 50% ISE. a. Detection of total ROS