Background
Reflux-aspiration, the regurgitation of gastric material up the gastro-oesophageal tract and its subsequent penetration into the lower airways, is associated with respiratory disorders such as idiopathic pulmonary fibrosis, chronic obstructive pulmonary disorder and bronchiolitis obliterans syndrome in lung transplant patients [
1‐
4]. It is also a common cause of respiratory morbidity and mortality in children and adults with severe neurodisability (sND) [
5‐
7].
Aspiration episodes can occur acutely as bolus events leading to acute lung injury and pneumonia, or chronically, through repeated aspiration of small volumes of gastric aspirate leading to chronic respiratory symptoms [
8]. There is also emerging consensus that “silent reflux” with chronic microaspiration of aerosolised, non-acidic or weakly acidic, gastric material is a clinical entity [
9,
10]. Supporting evidence comes from oesophageal and tracheal pH measurements in patients with suspected reflux-aspiration and chronic respiratory disease, in whom tracheal pH rarely drops below pH 5.0 even with reflux episodes [
7,
11‐
14]. It is speculated that this relatively mild airway acidification leads to infection, persistent inflammation, and irreversible lung damage over time. In bronchoalveolar lavage fluid (BAL) from children with sND at high risk of chronic microaspiration, we have recently shown bacterial airway colonisation with oral commensals associated with elevated levels of IL-8 and marked airway neutrophilia [
15]. In vitro and in vivo studies have previously shown that airway epithelial cells (AECs) exposed to acid produce pro-inflammatory and chemoattractant cytokines leading to leukocyte infiltration/activation in the airways, but most of these studies involved the administration of low pH (i.e. pH < 3.0) boluses [
16‐
20]. As such, they are unlikely to reflect the scenario in those patients described above with chronic reflux-aspiration.
In this study, we have documented BAL pH in a group of children at high risk of reflux-aspiration. We then investigated the effects of mild acidification (pH range 7.4 – 5.5) on inflammatory responses of airway epithelial cells (AECs). These studies have led us to explore the consequences of acidification in the context of infection, through exposure of AECs to mild acidification followed by challenge with lipopolysaccharide (LPS), a major component of the outer membrane of Gram-negative bacteria.
Methods
Patients
BAL was collected from children admitted to Alder Hey Children’s Hospital in Liverpool, UK between October 2009 and September 2011 as part of a larger study into airway pathology in children with sND. The Liverpool Paediatric Research Ethics Committee reviewed and approved the study and consent procedure (REC Reference number: 09/H1002/58). Informed written parental consent was obtained. Children over the age of 2 years with central ND (not neuromuscular disease) who were non-ambulant (Gross Motor Function Classification IV-V) were recruited at a time of respiratory stability when admitted for elective surgical procedures, or at a time of respiratory deterioration whilst ventilated on the paediatric intensive care unit (PICU).
Sample collection and processing
Non-bronchoscopic BAL samples were collected according to European Respiratory Society (ERS) guidelines following induction of anaesthesia [
21]. Within 15 min of collection, BAL sample pH was measured using a pH meter (Philips PW 9418, Netherlands), calibrated prior to each reading.
Cell culture and experimental conditions
BEAS-2B bronchial epithelial cells were used as a model of healthy airway epithelial cells and grown in BEGM media (Lonza, Belgium) on flasks coated with 1 % PureCol® collagen (Nutacon, Netherlands) in HEPES buffered saline solution under humidified conditions at 37 °C and 5 % CO2. All experiments were carried out between passage 3–5. The same passage of cells was used to test each experimental condition. Each biological replicate was carried out on the subsequent passage. Cells were seeded at 3×103 or 2×104 cells per well in 96-well and 24-well plates respectively, and grown to 70 % confluence having being fed 24 h before each experiment. Cells were incubated for 4, 8, 16 or 24 h at varying pH. Media pH was adjusted to pH7, pH6.5, pH6 and pH5.5 through the addition of 1 M HCl. Cells were stimulated with Escherichia coli LPS (Sigma, UK) at 5 μg/ml for 4 or 16 h in acidic media or for 24 h in normal pH media (pH7.4). For some experiments, cells were treated with the intracellular protein transport inhibitor, Brefeldin A (eBioscience, UK), for 1 h before addition of LPS at pH7.4. This acts as a positive control for cytosolic protein retention following cell activation. Images of cells were taken using a phase-contrast microscope with a DFC420 camera (Leica, Germany).
Preparation of whole cell lysates
At the end of each time point, media was removed from cells and centrifuged. Cell-free supernatant was stored for future analysis at −30 °C. Cells were twice washed with ice cold sterile PBS and lysed using Cytobuster™ Protein Extraction Reagent (Merck Millipore, Germany) following the manufacturer’s instructions. Whole cell lysate was stored at − 30 °C for future analysis.
Cytokine mRNA expression
IL-6 and IL-8 was measured by quantitative real-time PCR (qPCR). RNA was extracted from cells using the RNeasy MiniKit (Qiagen, Netherlands) following the manufacturer’s instructions. Reverse transcription was performed using a High Capacity cDNA Reverse Transcription Kit (Applied Biosystems, UK) and qPCR was performed using TaqMan primer probe assays: IL-6, Hs00985639_m1; IL-8, Hs00174103_m1; L32, Hs00388301_m1; β-actin, Hs99999903_m1 (Life Technologies, USA). Ribosomal protein L32 and β-actin were used as internal standards [
22,
23]. Expression was measured in duplicate and was calculated using the comparative C
T method [
24].
Cytokine protein measurement
Intracellular and extracellular IL-6 and IL-8 protein expression was quantified in whole cell lysate and culture supernatant by ELISA (R&D Systems, USA). Intracellular cytokine concentration was normalised to the total protein concentration of whole cell lysate as measured by BCA protein assay (Pierce, UK). Cytokine and total protein concentrations were measured in duplicate. Interleukin protein stability at pH ≥ 5.5 was confirmed by spike retrieval assay.
Cell viability measurement
Cell viability was measured using an MTT assay (Life Technologies, USA). MTT is a tetrazolium dye, taken up by live cells and therein reduced to a purple insoluble product. This reaction can be quantified and used as a measure of metabolic activity and an indicator of cell viability. Following incubation under experimental conditions, cell media was replaced with fresh BEGM (pH7.4). MTT dye solution was added to the normal pH media in each well for 4 h at 37 °C and 5 % CO2. All but 25 μl media was then removed and cells incubated for 10 min at 37 °C with 50 μl DMSO to solubilize the cytosolic formazan product. Each well was mixed thoroughly by pipetting and the plate was read at 540 nm. Cell viability was expressed as a percentage of the control OD value. Viability was measured in triplicate on a 96-well plate.
Statistical analysis
StatsDirect 2.7.9 (StatsDirect Ltd, UK) was used for statistical analysis of experimental data. BAL pH was analysed by Mann-Whitney test. All qPCR data was analysed by Kruskal-Wallis one-way analysis of variance followed by Conover-Inman pairwise comparison. ELISA data was analysed by Kruskal-Wallis one-way analysis of variance followed by Conover-Inman pairwise comparison. MTT assay data was analysed by one-way ANOVA followed by Dunnett’s multiple comparison test. Values are presented as mean ± SEM. Statistical significance was defined as p < 0.05.
Discussion
Chronic microaspiration of gastric reflux has been implicated in the pathogenesis of multiple respiratory diseases, yet our understanding of how aerosolized refluxate affects the airways remains limited. In this manuscript we report that BAL pH ranges varied between pH5.0–7.7 in individuals at high risk of aspiration. Consequently, we investigated inflammatory responses of AECs to a weakly acidic environment. We found that prolonged exposure resulted in both cytotoxicity but did not induce inflammatory cytokine expression. However, exposure for shorter periods (4 h), while not significantly affecting cytotoxicity, did blunt epithelial responses to LPS, an effect not associated with increased intracellular cytokine retention. Suppression of epithelial inflammatory responses without significant cytotoxicity continued to be seen when we “shock treated” cells with acid prior to stimulation with LPS at normal pH.
We based our in vitro experimental design on evidence from combined oesophageal and tracheal pH monitoring studies in individuals with gastro-oesophageal reflux-associated asthma, stroke and cystic fibrosis which rarely report drops in pH below pH5.0 [
7,
11,
14]. Studying a pH range of pH7.4–5.5 was further supported by the pH range of BAL collected from children at high risk of aspiration lung disease undergoing routine elective surgery. Furthermore, a recent study investigating reflux-aspiration in critically ill children found that the majority of reflux episodes were weakly acidic (pH4-7) [
31]. Our BAL pH data implies that there is variability in the airway pH of individuals at high risk of aspiration. As there is no definitive test for reflux-aspiration [
32], this variability may be explained by the fact that not all patients measured are aspirating their refluxate. We accept that in our cohort, airway acidification may not have solely been due to reflux-aspiration. Lower airway inflammation, so often seen in these children [
15], is characterised by an acidic microenvironment believed to be a consequence of neutrophil infiltration and bacterial metabolism [
33‐
36]. Of note in our study, BAL pH in those children with ND who were mechanically ventilated for a respiratory exacerbation (PICU-ND), was generally more alkaline (
p = 0.06), a finding that probably reflects lower carbon dioxide levels in the airways of these patients. Additionally, we have previously reported that BAL from PICU-ND patients had higher protein concentrations than BAL from elective patients [
15]. It is known that high concentrations of proteins, such as albumin, can buffer pH therefore it is possible that this may contribute to the higher pH of PICU-ND BAL.
Previous studies in mice have reported the effects of acid aspiration on pulmonary inflammatory cytokine secretion and neutrophil recruitment [
16‐
20]. These studies are probably not physiologically relevant in the context of chronic microaspiration of aerosolized gastric fluid, as the observed inflammation and neutrophilia was due to acute tissue injury caused by the direct administration of very low pH solutions (pH1.25) into the airways. In contrast, we have studied prolonged exposure to a weakly acidic environment, and found that while it does cause some cell death, it also suppresses inflammatory cytokine production. This suppression may be influenced by the reduction in viable cells following 24 h incubation in weakly acidic media, although it is surprising that this cytotoxicity did not induce an inflammatory response. Given that airway epithelium is regularly exposed to a variety of inhaled exogenous stimuli, there will undoubtedly be a necessity for a degree of stimulatory tolerance. It is conceivable that AECs are relatively refractory to ‘mild’ changes in pH within their microenvironment [
37], and certainly capable of recovering from such insults.
Repeated respiratory tract infections are common in individuals who aspirate [
6], with bronchiectasis a common end stage pathology [
38]. Weakly acidic conditions can blunt inflammatory responses to infection exhibited by macrophages [
25,
26,
28,
36,
39]. Furthermore, a reduction in airway surface liquid pH in the CF pig model reduces bacterial killing, a finding attributed to loss of antimicrobial peptide activity [
40]. As AECs play an important role in regulating local immunity through secretion of chemoattractants such as IL-8 in response to infection [
37], we hypothesized that weakly acidic conditions might reduce these responses. Consequently, we investigated whether AEC ability to express inflammatory/chemotactic proteins in response to LPS stimulation was altered in an acidic environment. LPS-induced cytokine expression was reduced when AECs were incubated for 4 h at weakly acidic pH, with cells maintaining viability. Expression was further reduced on 16 h exposure to low pH, but was associated with a significant decrease in cell viability.
There is little information in the published literature about duration of microaspiration episodes. We postulated that maximum duration of airway exposure to refluxate might occur at night-time and be up to 8 h. We therefore investigated whether discrete ‘shock treatments’ of 4 or 8-hours weakly acidic pH exposure might influence subsequent airway inflammatory responses to endotoxin. When cells were viewed by microscopy immediately following ‘shock’ treatment, some cells had detached (particularly those incubated at pH5.5), implying a degree of cell death. However, following subsequent 24-hour incubation at pH7.4 with LPS, all cultures appeared macroscopically similar to those maintained in pH7.4 media throughout, with cell viability well preserved. Again, we observed a significant reduction in LPS-induced cytokine protein production following weakly acidic shock treatment and subsequent prolonged exposure to endotoxin at normal pH. It is possible that similar exposure in individuals who micro-aspirate might compromise their response to infection through reduced cytokine production diminishing immune cell recruitment.
Interestingly, the metabolic rate of AECs “shocked” at pH5.5 and measured by MTT assay, was higher than cells maintained at higher pH (including those at pH7.4). Given that this pH obviously causes some cellular injury/detachment, it may be that those cells that survive this “shock” treatment increase their rate of proliferation in an effort to heal the monolayer. It is possible that this type of repeated injury could occur in individuals that chronically aspirate, resulting in loss of epithelial tissue and potentially leading to airway remodelling, a finding commonly observed in patients suspected of reflux-aspiration [
4].
Whilst there are limitations to this primarily in vitro study, the experimental design we used was more physiologically relevant in the context of chronic microaspiration of gastric refluxate than previously published work examining neat administration of hydrochloric acid. We accept that our in vitro monolayer culture model has a number of limitations including the lack of cellular differentiation and that our proxy aspirate lacked other important components of gastric juice such as pepsin that could also affect AEC function. Nevertheless, this work is novel and should be informative for future investigations into this much under-researched respiratory pathology.
Abbreviations
AEC, airway epithelial cell; BA, Brefeldin A; BAL, bronchoalveolar lavage fluid; BCA, bicinchoninic acid; BEGM, bronchial epithelial cell growth medium; DMSO, dimethyl sulfoxide; HCl, hydrochloric acid; IL-6, interleukin 6; IL-8, interleukin 8; LPS, lipopolysaccharide; MTT, 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide; qPCR, quantitative real-time PCR; sND, severe neurodisability; TNF-α, tumour necrosis factor α