Background
Cigarette smoke (CS) produces oxidant and reactive oxygen species in the lungs [
1] and the unfavorable impacts of CS represent a primary cause of respiratory diseases, such as chronic obstructive pulmonary disease (COPD) and a leading contributor to preventable morbidity and mortality [
2]. Preventing CS-related diseases can be best achieved by avoiding smoking initiation [
3]; however, when the smoking habit has already started, smoking cessation considerably decreases numerous adverse health effects [
4]. Intervention by smoking cessation during the early stages of COPD can decrease the rate of lung-function loss [
5]. Moreover, CS-induced oxidant burden can irreversibly change endogenous antioxidant protection in airways. CS-derived oxidants injure airway epithelial cells by affecting lipid-membrane, carbohydrate, protein, and DNA structures and lead to the persistent inflammation evident in COPD development [
6].
In addition to its status as a leading cause of mortality worldwide, COPD is also the only major cause of death that continues to increase annually [
2]. Two main features of COPD include progressive limitation of expiratory airflow, which is not completely reversible, and an abnormal inflammatory response in the lungs [
7]. Hogg et al. [
8] suggested that the initial locus of inflammation in COPD is the small airways represented by membranous (diameter, < 2 mm), terminal and respiratory bronchioles. In smokers, the small airways show structural abnormalities, with or without COPD. The relationship between inflammatory mucus exudates from airway lumen and COPD severity was investigated by analyzing lung tissue from patients with all clinical stages of COPD [
9], revealing that small-airway obstruction associated with COPD is related to thickening of the airway wall as a consequence of the remodelling process marked by mucociliary clearance-component malfunction and tissue repair related to innate-immune mechanisms, finally resulting in acculumation of lumen inflammatory exudates [
9,
10]. A more recent study from the same group, based on multidetector computed tomography obtained from patients at various stages of COPD, showed that narrowing and loss of small conducting airways occurs before the emphysematous changes associated with COPD [
11]. However, the genetic and biochemical alterations involved in responses to CS exposure by airway epithelial cells from different sites in the lungs have not been fully investigated.
In this study, we investigated the response to CS in two commercially available human airway epithelial cell lines [small airway epithelial cells (SAECs) and normal human bronchial epithelial cells (NHBEs)]. SAECs were isolated from the distal portion of a healthy lung in the 1-mm bronchiole area, and NHBEs were derived from the epithelial lining of airways above the bifurcation of the lung (according to documentation from Lonza, Basel, Switzerland). We explored the genetic differences between SAECs and NHBEs during the initial phase of CS-exposure using microarray analysis and performed in vivo smoking experiments using a mouse model to confirm that cells from different lung sites respond differently to CS exposure.
Methods
CS extract (CSE) was prepared using commercially marketed Peace unfiltered cigarettes (29 mg of tar and 2.5 mg of nicotine per cigarette; Japan Tobacco, Tokyo, Japan). The smoke from one cigarette was bubbled through 15 mL of sterilized purified water, and this solution passed through a 0.22-μm pore filter (Pall Corporation, Port Washington, NY, USA), to generate 100% CSE. Pilot cell-growth experiments showed that concentrations of ≤2.5% CSE were not cytotoxic; therefore, cultured cells were exposed to 2.5% CSE.
Cell culture
SAECs and NHBEs were purchased from Lonza. Three batches of each cell type, each originally isolated from three different healthy never-smoker donors, were used. SAECs were cultured with small airway epithelial cell growth medium (Lonza), and NHBEs were cultured with bronchial epithelial cell growth medium (Lonza), according to the manufacturer’s instructions (SAECs and NHBEs were seeded at 2500 and 3500 cells/cm2, respectively, in 25 cm2 plastic flasks). Both cell types were used at passages three to six and assayed at the same passages for direct comparisons. After reaching approximately 80% confluence, cells were treated with or without 2.5% CSE until assayed.
RNA extraction and real-time quantitative polymerase chain reaction (qPCR)
Total RNA was extracted using miRNeasy mini kits (Qiagen, Hilden, Germany). For cDNA synthesis, total RNA was transcribed using ReverTra Ace qPCR RT master mix with gDNA remover (Toyobo, Osaka, Japan). Real-time qPCR for cyclooxygenase (COX)-2 mRNA detection was performed using Thunderbird SYBR qPCR mix (Toyobo), an ABI 7500 Fast real-time PCR instrument (Applied Biosystems, Foster City, CA, USA), and COX-2-specific primers [(5′ → 3′) Fw: CTGGCGCTCAGCCATACAG; Rv: CCGGGTACAATCGCACTTATACT; Thermo Fisher Scientific, Waltham, MA, USA] according to manufacturer instructions. The expression of β-actin was measured using specific primers [(5′ → 3′) Fw: CTCTTCCAGCCTTCCTTCCT; Rv: AGCACTGTGTTGGCGTACAG; Thermo Fisher Scientific] as an internal control.
Microarray analysis
Two different batches of SAECs and NHBEs were used for microarray analysis. RNA samples (2 μg) were labeled using the low input Quick Amp labeling kit (Agilent Technologies, Waldbronn, Germany). Labeled RNA was hybridized to an oligonucleotide microarray (Whole Human Genome 4× 44 K; Agilent Technologies) at 60 °C for 17 h. After washing with reagents from the gene expression wash buffer kit (Agilent Technologies) and drying, slides were scanned with an Agilent microarray scanner and analyzed using Feature Extraction software (v9.5.1; Agilent Technologies). Normalization was performed using global normalization methods. Data from the microarray analysis have been deposited in the Gene Expression Omnibus database (accession ID: GSE107200). Differentially expressed genes were defined based on unpaired
t-test
p values < 0.01 and fold change ≥2. Gene-set enrichment analysis (GSEA) of differentially expressed genes was conducted using the DAVID functional annotation tool (
https://david.ncifcrf.gov/home.jsp). Gene sets from the Kyoto Encyclopedia of Genes and Genomes (KEGG;
http://www.genome.jp/kegg/) and the Gene Ontology (GO) biological process database (
http://www.geneontology.org/) were used. Pathways and ontology terms with a
p < 0.05 were considered significant. The Benjamini-Hochberg procedure was used in the analysis to control the false discovery rate.
Western blot analysis
Whole-cell lysates were prepared, and the total protein content of each measured using a BCA kit (Thermo Fisher Scientific). Protein aliquots (10 μg) were separated by electrophoresis on 10% sodium dodecyl sulfate polyacrylamide gels. After transferring the proteins to Immobilion-P membranes (Merck Millipore Corporation, Darmstadt, Germany), membranes were blocked with 5% skim milk in Tris-buffered saline with Tween-20 and allowed to react with an anti-COX-2 (1:400; Santa Cruz Biotechnology, Dallas, TX, USA) and anti-β-actin (1:10,000; Wako, Tokyo, Japan) monoclonal antibodies at 4 °C overnight. Membranes were then incubated with the appropriate horseradish-peroxidase-conjugated secondary antibody (1:20,000; Sigma-Aldrich, St. Louis, MO, USA), followed by detection with the Clarity western enhanced chemiluminescence substrate (Bio-Rad Laboratories, Hercules, CA, USA). Western blot signals were captured using an ImageQuant LAS-4000 mini fluorescence imager (GE Healthcare, Pittsburgh, PA, USA) and quantified using Multi Gauge image analysis software (Fujifilm Corporation, Tokyo, Japan).
In vivo exposure to CS and preparation of lung sections
C57BL/6 J male mice (3 months old;
n = 6) were purchased from Oriental Yeast Co., Ltd. (Tokyo, Japan) and used for the smoking experiment. The procedures for animal experimentation were approved by the Animal Care and Use Committee of Juntendo University School of Medicine (Tokyo, Japan). Mice were maintained in a limited-access barrier facility and housed in a humidity (55% ± 10%) and temperature (24 °C ± 2 °C) controlled room under 12-h light/dark cycles. Mice were fed with standard commercial chow (CRF-1; Oriental Yeast) and provided water ad libitum. Mice were exposed to CS using Peace unfiltered cigarettes and a tobacco smoke-inhalation experimental system for small animals (Model SIS-CS; Shibata Scientific Technology, Tokyo, Japan) [
12‐
15]. The experimental settings were as follows: 15 mL stroke volume, six puffs per minute, and 3.5% CS diluted with compressed air (the mass concentration of total particulate matter; 1430 ± 100 mg/m
3). Mice were exposed to either diluted CS or fresh air (as a control,
n = 3 each) for 30 min/day over 5 days.
Mice were sacrificed at day 5 and their lungs subsequently processed as described previously [
13‐
15]. Briefly, mice were anesthetized and then sacrificed by exsanguination of the left atrium with perfusion of phosphate-buffered saline (PBS) through the main pulmonary artery. Lungs were inflated and fixed by intratracheal instillation of 10% buffered formalin (pH 7.4; Wako) at a constant pressure of 25 cm H
2O for 48 h. Frontal plane sections of the lungs were taken at the depth of the hilum, followed by further sectioning of each block into two equal-sized pieces, and from the frontal plane, and embedded in paraffin.
Immunohistochemistry
Paraffin-embedded lung sections (4 μm) were deparaffinized in xylene and rehydrated in ethanol. Heat-induced epitope retrieval (10 min at 125 °C) was performed in citrate buffer (pH 6.0; LSI Medience Corporation, Tokyo, Japan). After blocking with 2% bovine serum albumin in PBS, tissue sections were incubated at 4 °C overnight with anti-COX-2 antibody (1:300; BD Biosciences, San Jose, CA, USA). Tissue sections were then incubated with biotin-labeled anti-mouse IgG antibody (1:300; Agilent Technologies) and labeled with avidin (1:50; Vector Laboratories, Burlingame, CA, USA) for 40 min at between 25 °C and 28 °C. Signals were detected using hydrogen peroxide and 3,3-diaminobenzidine tetrahydrochloride. Tissue sections were counterstained with hematoxylin and dehydrated in xylene. The ratio of positively immuno-stained nuclei to the total count of nuclei present in a field at 400× magnification was determined in 10 different lung areas per mouse.
Statistical analysis
Data are expressed as the mean ± standard error of the mean (SEM) and were analyzed using GraphPad Prism 6 (GraphPad Software, San Diego, CA, USA). Analysis of variance was performed using a non-parametric Kruskal–Wallis test. When applicable, the Mann–Whitney U test was used for comparisons between groups. Differences were considered significant at p < 0.05.
Discussion
This study demonstrated that small and large airways differed in their initial responses to CS exposure. Microarray analysis revealed that small airways showed higher susceptibility to CS compared with large airways and displayed enhanced expression of genes associated with inflammation-related pathways, including TNF-signaling. Among TNF-pathway related genes, PTGS2, also known as COX-2, showed the greatest difference in expression levels, with higher CSE-induced increases in both mRNA and protein expression in SAECs compared with NHBEs. In vivo short-duration smoking experiments also showed that initial COX-2 expression occurred in lung small-airways. This is the first study clarifying initial genetic and biochemical alterations associated with inflammatory responses to CS exposure in the small and large airways.
CS is the most typical source of exogenous oxidants [
18], and oxidants derived from CS can injure airway epithelial cells, leading to persistent inflammation. Oxidative stress and persistent inflammation combine to result in increases in airway free radicals and amplification of the expression of proinflammatory genes that contribute to the release of inflammatory proteins, thereby promoting inflammatory cell infiltration [
6]. The present study demonstrated that the expression of proinflammatory genes, and initiation of signaling associated with their related pathways, were more greatly enhanced in SAECs relative to NHBEs following a comparatively short period of CSE exposure. Further, we identified initial induction of
COX-2 expression in the small airway of CS-exposed mice. These results suggest that SAECs represent the initial targets of CS and play an important role in the development of CS-mediated inflammation, which might contribute to CS-related respiratory diseases, such as COPD.
Accumulating evidence suggests that the small airway is the initial locus of inflammation in COPD and plays a critical role in disease development [
19]. Based on clinical findings, even patients with mild COPD already show some pathological abnormalities in the small-airway area, including active inflammation and obstruction [
9]. Based on the results of multidetector computed-tomography, McDonough et al. [
11] reported that small-airway injury precedes the development of centrilobular emphysema, and another study of pathology in smokers who underwent surgical resection showed that the number of neutrophils and mast cells infiltrating the small airway was higher than that in the large airway [
20]. In terms of genetic analysis, Mercer et al. [
21] performed microarray analyses using CSE-exposed SAECs and showed differential regulation of 425 genes involved in various biological processes, including immune function, signal transduction, apoptosis, the cell cycle, cell proliferation, and antioxidation. The results of the present study were consistent with the findings of Mercer et al. [
21], and further demonstrate that CSE exposure affects gene groups related to diverse biological processes in SAECs, including the inflammatory response, wound healing, aging, apoptosis, cell differentiation, and neutrophil chemotaxis (Table
2). Importantly, these findings extend previous results by demonstrating that SAECs exhibit greater alterations in gene regulation and a higher susceptibility to CSE exposure than NHBEs.
A recent study by Yang et al. demonstrated that smoking induced
distal-to-proximal repatterning of the adult human small airway epithelium, which is a pathologic feature of COPD [
22] . The authors showed that the proximal airway epithelium is enriched for molecular features related to oxidative stress, xenobiotic metabolism, and nicotine degradation, suggesting that the proximal airway epithelium can more robustly tolerate oxidants and toxins compared with the small airway epithelium. In the current study, the GSEA in NHBEs indicated that a smaller number of pathways and GO terms were overrepresented following CSE exposure compared to SAECs (Tables
3,
4); however,
GPX-2 [glutathione peroxidase] was one of the enriched genes in the arachidonic acid metabolism pathway, according to KEGG analysis (Table
3). Furthermore, GO enrichment analysis showed that several
AKR [aldo-keto reductase] family members, which encode enzymes that detoxify oxidative stress-induced carbonyl proteins, were enriched (Table
4). These findings are consistent with those of Yang et al. [
22], showing that
GPX-2 is a proximal airway epithelial signature gene. Interestingly, Yang et al. [
22] also showed that the expression of proximal signature genes was increased in the small airway epithelium of healthy smokers and COPD patients compared with nonsmokers. They named this phenomenon, which may be relevant to the pathogenesis of small airway disorder in COPD,
distal-to-proximal repatterning. Here, we investigated the initial response towards CS-exposure to determine whether the small-airway is the initial site of the CS-induced lung damage; however, investigations during longer periods of CSE exposure will likely be fruitful beyond the current study.
Here, we focused on COX-2 as an important factor that initiates an inflammatory response to CSE exposure in the lung. COX-2 is a key mediator of inflammation that catalyzes the transformation of arachidonic acid to thromboxanes and PGs, such as prostaglandin E
2 (PGE
2) [
23]. COX-2 is also believed to be associated with COPD pathogenesis [
24]. A previous animal study showed that treatment with a COX-2 inhibitor attenuated PGE
2 synthesis and protected against the development of CS-induced pulmonary emphysema [
25]. Our previous study, using primary lung fibroblasts from patients with COPD and smokers without COPD (controls), demonstrated that following stimulation with inflammatory cytokines, fibroblasts from COPD subjects produced higher PGE
2 levels relative to those from control subjects, and that elevated PGE
2 levels were the result of an increase in COX-2 expression due to alterations in mRNA stability caused by weak induction of microRNA-146a [
17]. More recently, Zago et al. demonstrated that expression of RelB, a component of the non-canonical NF-κB pathway, and the aryl hydrocarbon receptor, are both reduced in fibroblasts from COPD patients and regulate COX-2 protein levels by altering mRNA stability [
26‐
29]. These findings suggest that the regulation of COX-2 is critical to controlling abnormal inflammation associated with COPD pathogenesis, and that its targeting could be a promising strategy for COPD therapeutics. Interestingly, a recent study suggested that mesenchymal stem cells attenuated airway inflammation and emphysema in CS-exposed rats through the downregulation of COX-2 and PGE
2 [
30].
This study had several limitations. First, we did not use primary cells from COPD patients to evaluate the direct contribution of the current findings to the disease condition. Second, the SAECs and NHBEs used in the current study were not from matched donors. Instead, we confirmed COX-2 expression in the lungs of CS-exposed mice as an experimental model of COPD. Thirdly, we used different media approved by the supplier for SAECs and NHBEs. Their components are almost identical; however, there is a possibility that they influenced the current findings. Additionally, we did not perform air-liquid interface (ALI) culture to promote differentiation. The phenotypes of each cell type observed under non-differentiating conditions could be affected by contamination with different cell types. Therefore, differentiating SAECs and NHBEs using ALI culture system first and then exposing to CSE will make experimental findings more applicable to human airways. However, our findings were consistent with those from a previous study using ALI culture [
22], indicating that commercially available cells cultured using conventional methods can be suitable for these types of experiment. Lastly, we did not define the mechanisms by which inflammation spreads from the initial site to the whole lung, leading to development of pulmonary emphysema. Further investigation of the transition of COX-2 expression in the airways, following various durations of CS exposure, and the interactions between small-airway cells and other cells will be beneficial.