Analysis of Acrolein Exposure Induced Pulmonary Response in Seven Inbred Mouse Strains and Human Primary Bronchial Epithelial Cells Cultured at Air-Liquid Interface

Background Acrolein is a major component of environmental pollutants, cigarette smoke, and is also formed by heating cooking oil. We evaluated the interstrain variability of response to subchronic inhalation exposure to acrolein among inbred mouse strains for inflammation, oxidative stress, and tissue injury responses. Furthermore, we studied the response to acrolein vapor in the lung mucosa model using human primary bronchial epithelial cells (PBEC) cultured at an air-liquid interface (ALI) to evaluate the findings of mouse studies. Methods Female 129S1/SvlmJ, A/J, BALB/cByJ, C3H/HeJ, C57BL/6J, DBA/2J, and FVB/NJ mice were exposed to 1 part per million (ppm) acrolein or filtered air for 11 weeks. Total cell counts and protein concentrations were measured in bronchoalveolar lavage (BAL) fluid to assess airway inflammation and membrane integrity. PBEC-ALI models were exposed to acrolein vapor (0.1 and 0.2 ppm) for 30 minutes. Gene expression of proinflammatory, oxidative stress, and tissue injury-repair markers was assessed (cut off: ≥2 folds; p < 0.05) in the lung models. Results Total BAL cell numbers and protein concentrations remained unchanged following acrolein exposure in all mouse strains. BALB/cByJ, C57BL/6J, and 129S1/SvlmJ strains were the most affected with an increased expression of proinflammatory, oxidative stress, and/or tissue injury markers. DBA/2J, C3H/HeJ, A/J, and FVB/NJ were affected to a lesser extent. Both matrix metalloproteinase 9 (Mmp9) and tissue inhibitor of metalloproteinase 1 (Timp1) were upregulated in the strains DBA/2J, C3H/HeJ, and FVB/NJ indicating altered protease/antiprotease balance. Upregulation of lung interleukin- (IL-) 17b transcript in the susceptible strains led us to investigate the IL-17 pathway genes in the PBEC-ALI model. Acrolein exposure resulted in an increased expression of IL-17A, C, and D; IL-1B; IL-22; and RAR-related orphan receptor A in the PBEC-ALI model. Conclusion The interstrain differences in response to subchronic acrolein exposure in mouse suggest a genetic predisposition. Altered expression of IL-17 pathway genes following acrolein exposure in the PBEC-ALI models indicates that it has a central role in chemical irritant toxicity. The findings also indicate that genetically determined differences in IL-17 signaling pathway genes in the different mouse strains may explain their susceptibility to different chemical irritants.


Introduction
Acrolein (2-propenal) is a highly volatile and reactive α-βunsaturated aldehyde that is primarily used as an intermediate in chemical manufacturing [1]. It is also produced during endogenous oxidative processes and is a major bioactive component of environmental pollutants such as automobile exhaust, biomass fuel, burning of wood and plastics, and smog. Acroleins are also formed by heating cooking oil and fat (above 300°C) during domestic cooking [2]. Acrolein is an important constituent of mainstream cigarette smoke with concentrations about 90 ppm [3]. The major effects of inhalation exposure to acrolein in humans and animals result in eye, nose, and throat irritation. Nasal irritation, activation of sensory nerves in the nasal mucosa, increaseexposure (0.31-1.7 ppm) in rodent studies [2].
Several in vivo studies have demonstrated that acute exposure to 0.2 to 6 ppm acrolein causes pulmonary edema, stimulates sensory nerves and airway cell proliferation, and diminishes pulmonary defenses against bacterial and viral infection [4]. Leikauf et al. [5] performed a study exposing 40 inbred mouse strains to 10 ppm acrolein to induce acute lung injury [1]. The study demonstrated remarkable variation in survival times among the mouse strains (<20 h: 13 strains; 20-30 h: 20 strains; >30 h: 7 strains) following acrolein-induced acute lung injury. The study was extended to identify genes that may render susceptibility or resistance to acrolein-induced acute lung injury by comparing the polar strains. Acute lung injury is a sporadic event, and case-based observation suggests great variability in individual susceptibility; i.e., individuals with the same severity score can have markedly different clinical outcomes thereby suggesting a high degree of genetic predisposition [5].
Borchers et al. [6] examined the effect of 0.01-100 nM acrolein on mucin gene expression in cultured human airway epithelial cells. Increased mucin (MUC5AC) transcript levels following a 4 h exposure to acrolein in cultured human airway epithelial cells were reported. Inflammation, oxidative stress, and tissue destruction are considered as primary adverse outcome pathways of most chronic lung diseases (McGuinness and Sapey [7]). Therefore, in this study, we sought to investigate variability in pulmonary response to subchronic acrolein exposure among seven inbred mouse strains measured as oxidative stress, proinflammatory response, and tissue injury. The study was aimed at investigating acrolein as an irritant model substance with a plausible genetic susceptibility contribution. The survival time of the seven inbred mouse strains selected for our study was spread across the survival times of the 40 inbred mouse strains (female) observed by Leikauf et al. [5]. together with easy availability and common use in respiratory research. Based on our findings of altered lung interleukin-17B (IL-17B) transcript expression in acrolein-exposed mice, we further investigated the IL-17 pathway genes in the human primary bronchial epithelial cells (PBEC) cultured at an air-liquid interface (ALI) following 0.1 and 0.2 parts per million (ppm) acrolein exposure.

Methods
2.1. Animals. Female 129S1/SvlmJ (stock number 002448), A/J (stock number 000646), BALB/cByJ (stock number 001026), C3H/HeJ (stock number 000659), C57BL/6J (stock number 000664), DBA/2J (stock number 000671), and FVB/NJ (stock number 001800) mice (age: 10 weeks) were obtained from the Jackson Laboratory (Bar Harbor, ME, USA), housed under specific pathogen-free conditions at the animal facility of the New York University, School of Medicine (NY, USA), and quarantined for at least 2 weeks before study initiation. Food and water were provided ad libitum except during exposure. The study was approved by the Institutional Animal Care and Use Committee, New York University, School of Medicine (NY, USA). We used only female mice for our study to minimize the animal numbers and for comparative analyses with other studies [5].

Mouse Procedures
2.2.1. Acrolein Exposure. Mice (n = 5/group/strain) were exposed to filtered air or acrolein vapor via whole-body exposure in 1.3 M 3 stainless steel inhalation exposure chambers for 6 h/day, 4-5 days/week for a total of 11 weeks. Acrolein vapor was generated by passing charcoal and HEPA-filtered air over acrolein (Sigma Chemical Co.) in a glass flask. The chamber concentration of acrolein was monitored on an hourly basis on each exposure day (target concentration of 1 ppm) with a Miran 1A single-beam infrared spectrometer (Foxboro Analytical, Foxboro, MA). The actual chamber concentration was 1:03 ± 0:03 ppm (mean ± SD). The 1 ppm dosage was a high dose to start with and nonlethal for a repeated exposure study, and in addition, it was an order of magnitude lower than that used in acute lung injury exposure studies.

Bronchoalveolar Lavage (BAL).
At 24 h after the final exposure, mice were euthanized by intraperitoneal injections of ketamine HCl (100 mg/kg; Vetalar, Fort Dodge Laboratories, Fort Dodge, IA) and sodium pentobarbital (175 mg/kg; Sleepaway, Fort Dodge Laboratories), and the posterior abdominal aorta was severed. The lungs of each mouse were lavaged two times with 1.2 ml of Dulbecco's phosphatebuffered saline without Ca or Mg (pH 7.2-7.4, 37°C; Gib-coBRL, Life Technologies, Grand Island, NY). The recovered BAL fluid was immediately placed on ice (4°C). The lavage fluid was then centrifuged (500 x g, 8 min, 4°C), and the supernatant was pipetted off. The total protein concentration in the supernatant was measured using an assay kit that follows the method of Bradford (1976) [8] (bovine serum albumin standard, 550 nm; Pierce, Rockford, IL). The total protein concentration in BAL fluid was used as an indicator of changes in lung permeability and injury [9]. Total cell counts were determined with a hemacytometer.

Air-Liquid Interface (ALI) Model Development
The human airway mucosa model was developed using PBEC from 3 individual (N = 3) donors and cultured at ALI as previously described [10][11][12]. All procedures performed in this study were in accordance with the approval of the Ethics Committee of Karolinska Institutet, Stockholm.
The PBEC that we have used in this study are well characterized and used in several previous studies [13,14]. PBEC from passages 3 or 4 were grown under submerged conditions in Petri dishes and were cultured up to 80% confluency at 37°C in a humidified atmosphere of 5% CO 2 using a serumfree keratinocyte medium (KSFM; Gibco, USA) supplemented with 5 ng/ml epidermal growth factor (EGF; Gibco, USA), 50 μg/ml bovine pituitary extract (BPE; Gibco, USA), and 20 U/ml pen/strep (Gibco, USA). The medium was changed every second day. Cells from the Petri dishes were trypsinized, resuspended, and seeded on precoated 0.4 μm semiporous transwell inserts (BD Falcon™, USA), placed in twelve well plates at a density of 0.1 million cells/well. The inserts were maintained in submerged conditions with 1 ml complete KSFM on both (inner and outer) sides of inserts.
Once cells reached 90% confluence, the inserts were turned upside down and put in a sterile Petri dish to add MRC-5 cells (Medical Research Council cell strain 5, fibroblasts derived from human lung tissue) in a complete Dulbecco's Modified Eagle's Medium to the outer or basal side of the insert membrane [10]. After 30 minutes of incubation, the inserts were again placed in the plate with 1 ml of complete KSFM per well, on both inner (apical) and outside (basal side) of the insert. The next day, the models were air-lifted by removing the medium and adding 870 μl coculture medium (i.e., complete KSFM with 6 μg/ml CaCl 2 in double-distilled water (ddH 2 O), 15 ng/ml ethanolamine in ddH 2 O, and 10-5 M retinoic acid) basal or outside the insert. The cells were incubated at 37°C with 5% CO 2 for at least two weeks till they began to differentiate into different cell types: mucus-producing cells, ciliated cells and club cells, and basal cells. Three technical replicates (n = 3/donor/acrolein concentration) were developed from each donor and were exposed for 30 minutes to different concentrations of acrolein vapor separately (see below). Finally, the expression of genes within the IL-17 pathway was analyzed at 24 h postexposure to acrolein.
3.1. Acrolein Exposure of the ALI Model. Three technical replicates developed from each donor were exposed for 30 minutes to clean air (sham), 0.05, 0.1, 0.2, and 0.5 ppm (0, 0.1, 0.2, 0.5, and 1.1 mg/m 3 ) acrolein using our in-house developed exposure systems as described in previous studies [10,12]. The actual acrolein concentrations in the ALI chamber air were monitored by gas chromatography (five times during 30 minutes), as described previously [10], and were on average 17-25% higher than the target concentrations.
Cell viability of PBEC was tested at 24 h after exposure by trypan blue staining (200 μl of 1 : 1 in PBS diluted 0.4% trypan blue solution for 1 minute), washing by PBS, and followed by bright-field microscopy [10]. At 0.5 ppm, cell viability was 70% ± 5% (mean ± SD) [10]. Since none of the other doses showed any cytotoxicity [2], further analysis was carried out in 0.1 and 0.2 ppm acrolein-exposed samples compared to sham. Based on the findings of altered Il-17B regulation in the mouse lungs following subchronic acrolein exposure, we explored the IL-17 pathway in acroleinexposed PBEC. Transcript expression of IL-17A, IL-17B, IL-17C, IL-17D, IL-17E, IL-1B, IL-22, RAR-related orphan receptor A (RORA), and signal transducer and activator of transcription 3 (STAT3) was analyzed 24 h after the 30 min exposure to clean air (sham), 0.1 and 0.2 ppm acrolein. Forward and reverse primer sequences are provided in supplementary table S1.

Statistical Analyses.
The results are generally expressed as medians and interquartile ranges (25th-75th percentiles) normalized with its own sham (control). For the in vivo study, the Friedman test was used to analyze differences between sham (clean air: control) and acrolein-exposed animals of each strain. Additionally, the interstrain differences of in vivo studies were compared by Kruskal-Wallis followed by Mann-Whitney. Similarly, for in vitro study, the Friedman test was used to analyze differences between sham (clean air: control) and different concentrations of acrolein used to expose the in vitro ALI model, followed by the Wilcoxon signed-rank t test as a post hoc test. A p value < 0.05 was considered as statistically significant. All data were analyzed using the STATISTICA13 software (StatSoft, Inc., Uppsala, Sweden). The statistical analysis of transcript expression was carried out using the dCT values.

BAL Analysis.
Total BAL cell counts and protein concentrations showed no changes in the acrolein-exposed mice compared to their respective strain control exposure (supplementary Table S2 and Table S3). 1(e)) was also detected in C3H/HeJ compared to a comparatively low oxidative stress response in DBA/2J and FVB/NJ strains (Figure 1(a)). Interestingly, the oxidative stress marker Ho1 did not show any significant alteration in any of the investigated seven inbred mouse strains exposed to subchronic doses of acrolein (Figure 1(d)). The A/J and FVB/NJ strains exhibited a medium proinflammatory response as observed by~2-fold increase of Tnf, Cxcl1, Cxcl2, and IL-6 transcripts (Figures 2(a)-2(e)). C3H/HeJ and DBA/2J mice did not show any alteration in these proinflammatory markers.

Transcript Expression in
To determine the effect of acrolein exposure on proteinase and antiproteinase homeostasis, measurements of the  )). Mmp9 was increased in BALB/cByJ, C3H/HeJ, and DBA/2J by~2-fold (Figure 3(a)), and a corresponding >3-fold increase in Timp1 was detected in the lungs of C3H/HeJ and DBA/2J but not in BALB/cByJ mice (Figure 3(b)). Timp1 expression was also increased by~2fold in FVB/NJ mice. Basal lung transcript levels of some selected oxidative stress, proinflammatory, and extracellular matrix markers among the seven inbred mouse strains are shown in Table 1. None of the genes exhibited any statistically significant difference in their transcript expression among the sham-exposed groups. The interstrain difference in transcript expression of the investigated genes following acrolein exposure is shown in Figures 1-3. . Beta actin (Actb) served as the housekeeping control gene. Data are presented as an expression in the exposed group relative to age and strain-matched sham-exposed group (median and 25th-75th percentiles; n = 5 mice/strain/stage); white bars: sham, gray bars: acrolein exposed. * p value < 0.05 and * * p value < 0.01; sham vs. exposed group of each strain. # p value < 0.05, exposed BALB/cByJ vs. other exposed strains like DBA/2J, C3H/HeJ, and A/J. ǂ p value < 0.05 and ǂ ǂ p vaue < 0.01; exposed C57BL/6J vs. exposed strains like DBA/2J, C3H/HeJ, and FVB/NJ. + p value < 0.05, exposed 129S1/SvlmJ vs. exposed FVB/NJ. x p value < 0.05, exposed C3H/HeJ vs. exposed A/J vs. FVB/NJ.
In spite of similar basal lung transcript levels of the investigated oxidative stress, proinflammatory, and extracellular sham (clean air); light gray bars: 0.1 ppm exposed; dark gray bars: 0.2 ppm acrolein exposed. Table 2: Summary of transcript level changes in oxidative stress, proinflammatory, and tissue injury markers in lung tissue collected from seven mouse strains after subchronic inhalation exposure to acrolein. Numbers indicate the fold change level compared to controls exposed to filtered air.
10 BioMed Research International matrix marker genes in sham (control group), acrolein exposure resulted in different response patterns of those genes among the seven mouse strains (Figures 1-3). These interstrain differences in responses suggest that acrolein-induced lung injury may be driven by genetic susceptibility. Increased lung Nfkb1, Sod3, Gpx1, and Gpx3 transcript expression levels strongly support the onset of oxidative stress due to subchronic acrolein exposure (Figure 1). Several other studies have also demonstrated that reactive oxygen species (ROS) and Tnf can increase the activation of Nfkb1 [17][18][19]. Increased Nfkb1 lung transcripts in the acroleinexposed BALB/cByJ, C57BL/6J, and 129S1/SvlmJ mice may cause substantial proinflammatory reactions (Figure 1(e)). The pulmonary Nfkb1 signaling pathway serves as a critical modulator for airway hyperresponsiveness to inhaled agents [20], and Nfkb1 is a central transcription factor for the production of numerous inflammatory cytokines (Figure 1(e)). Nfkb1 is also considered as a generic marker of toxic stress which may explain the different patterns of lung Nfkb1 expression among the mouse strains compared to other oxidative stress markers [21]. Cheng et al. [22] demonstrated that Nfkb1 signaling in nonimmune cells is a critical determinant for the pulmonary response to injurious stimuli and has a great impact on a number of key biological processes, including host defense mechanisms. Different groups have [20,22] reported that the activation of Nfkb1 in airway epithelial cells of mice leads to the increased expression of several cytokines and chemokines such as IL-6, Cxcl1, Cxcl2, and IL-17. Our findings in the lungs of subchronically acrolein-exposed mice are consistent with these studies. It is well established that Nfkb1 is involved in IL-6 production [23,24], and therefore, our findings of simultaneously increased lung IL-6 levels in the same mouse strains with increased lung Nfkb1 are consistent (Figures 1 and 2). Since IL-17 can also induce IL-6 expression, the pattern of transcript expression in the lungs of acrolein-exposed mouse strains correlates ( Figure 2) and supports the validity of the mouse models.
An imbalance of proteinases and antiproteinases and especially Mmp and Timp (Figures 3(a) and 3(b)) is an important feature of the pathogenesis of chronic lung diseases [25,26], and their expression is regulated by various inflammatory markers such as Tnf, IL-6, Cxcl1 [27,28], and Cxcl2. Therefore, the findings of the present study, i.e., increased Mmp9 along with unaltered Timp1 transcript expression in the acrolein-exposed BALB/cByJ strain, indicate susceptibility to acrolein (Figures 3(a) and 3(b)). Thus, low levels of Timp1 can cause increased sensitivity to acrolein and other chemical irritants. However, in the more resistant strains (DBA/2J, C3H/HeJ, and FVB/NJ), both Mmp9 and Timp1 were upregulated suggesting a level protease-antiprotease balance to counter tissue injury (Figures 3(a) and 3(b)). Upregulation of IL-17B in the lungs of the more susceptible strains (i.e., BALB/cByJ, C57BL/6J, and 129S1/SvlmJ, Figure 2(d)) is an important observation. The IL-17 pathway has been recognized as a proinflammatory cytokine pathway involved in chronic airway inflammation among asthmatics [29,30]. Data suggesting a role for IL-17 in tobacco smoke-induced lung diseases, including COPD, has also emerged [31,32]. Acrolein is a major component of tobacco smoke. It is plausible that the increased transcript expression of IL-17B detected in mouse lungs may be due to the stimulation of resident CD4+ cells by proinflammatory cytokines such as IL-6, Cxcl1, Cxcl2, and Tnf as described by Wang et al. [30]. This led us to investigate the IL-17 pathway genes in more detail in the PBEC-ALI model following exposure to 0.1 and 0.2 ppm acrolein.
Increased expression of IL-17A, C, and D in the bronchial mucosa model following 0.1 ppm acrolein exposure (Figures 4(a)-4(e)) indicates the plausible airway inflammation that may be caused by low-dose acrolein exposures relevant for the onset of chronic lung diseases such as asthma and COPD. Increased expression of IL-1B, IL-22, and RORAv ( Figures 5(a)-5(c)) following acrolein exposure further demonstrates IL-17 as a candidate pathway for low-dose acrolein exposure-mediated airway effects. It is plausible that IL-17 is indirectly triggered by cells damaged by acrolein and IL-17 may have a central role in the response to chemical irritants. Mechanistic studies focused on the role of IL-17 to other aldehydes will enhance our understanding in this aspect.
To summarize, our study utilized 7 inbred mouse strains (female) that are commonly used in respiratory research and demonstrated that subchronic exposure to low-dose (1 ppm) acrolein does not result in transient neutrophilia and lung injury. However, it does induce significant and strainselective changes in the expression of several oxidative stress, proinflammatory, and tissue injury/repair markers leading to the identification of sensitive and resistant mouse strains for use in genetic susceptibility studies. Consistent findings of a simultaneous increase in lung Nfkb1, Il6, and IL-17B within the same acrolein-exposed mouse strains support the validity of the models. The findings also indicate that genetically determined differences in IL-17 signaling pathway genes in the different mouse strains may explain their susceptibility to different chemical irritants. On the basis of our mouse studies, we have further investigated the effects of low-dose acrolein exposure on PBEC cultured at ALI and thus identified IL-17 as a candidate pathway for low-dose acroleininduced effects in human primary bronchial epithelial cells cultured at an air-liquid interface. Moreover, our findings implicate that low-dose acrolein exposure impairs the innate immune response in the airways which may in turn result in a predisposition to chronic lung diseases such as asthma and COPD. Therefore, it is plausible that IL-17 plays a central role in chemical irritants and it will be important to elucidate if the IL-17 pathway has a specific role in acrolein-mediated toxicity through mechanistic and functional studies.

Data Availability
All data required to comprehend the manuscript have been provided in the manuscript's main body and supplementary material. All raw data are available from the corresponding author.

Ethical Approval
The study was approved by the ethics committee at Karolinska Institutet, Stockholm, Sweden, and by the Institutional Animal Care and Use Committee, New York University, School of Medicine (NY, USA).