Abstract
Asthmatic smokers have poor symptom control and accelerated decline in lung function. A reduced ratio of matrix metalloproteinase (MMP)-9/tissue inhibitors of metalloproteinases (TIMPs) in nonsmokers with asthma has been implicated in airway remodelling. We tested the hypothesis that sputum MMP-9 activity/TIMPs ratios are reduced in smokers compared with never-smokers with asthma and are associated with reduced lung function and altered computed tomography (CT) measures of airway wall dimensions.
Lung function, airway dimensions by CT, and induced sputum concentrations (and activity) of MMP-9 and TIMP-1 and -2 were measured in 81 asthmatics and 43 healthy subjects (smokers and never-smokers). Respiratory epithelial MMP9 and TIMP mRNA was quantified in 31 severe asthmatics and 32 healthy controls.
Sputum MMP-9 activity/TIMP-1 and TIMP-2 ratios, and nasal epithelial MMP9/TIMP1 and MMP9/TIMP2 expression ratios were reduced in smokers with asthma compared with never-smokers with asthma. Low sputum ratios in asthmatic smokers were associated with reduced post-bronchodilator forced expiratory volume in 1 s (FEV1), FEV1/forced vital capacity ratio and segmental airway lumen area.
The association of a low sputum MMP-9 activity/TIMP-1 ratio with persistent airflow obstruction and reduced CT airway lumen area in smokers with asthma may indicate that an imbalance of MMP-9 and TIMPs contributes to structural changes to the airways in this group.
Abstract
In asthmatic smokers, a low sputum MMP-9 activity/TIMP-1 is associated with spirometric and CT airway narrowing http://ow.ly/wnbAh
Introduction
A considerable proportion of patients with asthma, particularly those with severe disease [1], or who are cigarette smokers [2] have poorly controlled asthma, which can be associated with corticosteroid resistance [3], altered patterns of airway inflammation [4], airway remodelling [4] and persistent airflow obstruction [5]. The mechanisms leading to corticosteroid insensitivity and remodelling in asthma are poorly understood [3, 4].
Matrix metalloproteinases (MMPs) are zinc-dependent neutral endopeptidases that form a family of extracellular matrix (ECM) proteolytic enzymes. Activity of MMPs is regulated by specific protease inhibitors called tissue inhibitors of metalloproteinases (TIMPs). Inappropriate expression and activity of several MMPs, including MMP-9, and imbalances in the MMP-9/TIMP ratio have been implicated in the pathogenesis of asthma [6–12]. In nonsmokers with asthma a reduced sputum MMP-9/TIMP ratio is associated with more severe airway obstruction [13, 14] and with computed tomography (CT) measures of increased airway wall area and thickness [14] as well as with increased CT lung scan abnormalities [15]. Based on these findings it has been proposed that high concentrations of TIMP-1 in relation to MMP-9 can lead to increased ECM deposition and increased myofibroblast proliferation, through TIMP-1 induced cell growth, and that a low sputum MMP-9/TIMP ratio may be a biomarker of airway remodelling in asthma [10–12]. Studies of asthma have recruited predominantly never-smokers; much less is known about MMP-9 activity and TIMP in smokers with asthma. Acute exposure of bronchial epithelial cells from patients with asthma to cigarette smoke in vitro reduces the MMP-9/TIMP-1 ratio suggesting that an interaction between asthma and cigarette smoking may potentiate structural changes to the airways [16]. We tested the hypothesis that MMP-9 activity/TIMP ratios are reduced in sputum from smokers with asthma compared with never-smokers with asthma and are associated with reduced lung function and altered CT measures of airway wall dimensions.
Materials and methods
Subjects
Participants were recruited with mild, moderate and severe persistent asthma (Global Initiative for Asthma classification) [17]; participants with asthma included both current smokers and never-smokers, and healthy smokers and never-smokers. All subjects were on stable medication and had had no exacerbation of disease for 4 weeks. Smokers were defined as individuals with ≥10 pack-years who currently smoke five or more cigarettes per day. Asthma subjects had to demonstrate either reversibility of forced expiratory volume in 1 s (FEV1) or airway hyperresponsiveness. The West Glasgow Research Ethics Committee approved the study and all patients gave written informed consent.
Study design
The study was designed to examine the relationship between sputum MMP-9 activity/TIMP-1 and TIMP-2 ratios and the expression of MMP-9/TIMP-1 and MMP-9/TIMP-2 with sputum cytology, lung function and CT measures of airway dimensions in smokers and never-smokers with asthma. It was a cross-sectional study, performed in subjects with asthma and healthy controls, as part of a chronic obstructive pulmonary disease and asthma biomarker study [18].
Measurements
Additional details of the measurements can be found in the online supplementary material.
Spirometry was performed according to American Thoracic Society guidelines [19] and airway hyperresponsiveness to methacholine was measured. Reversibility was defined as >12% and 200 mL improvement in FEV1 following salbutamol.
Exhaled nitric oxide fraction was measured (Niox Flex; Aerocrine, Solna, Sweden) at 50 mL·s−1 (FeNO50) in concordance with standardised guidelines [20].
Lung volumes and diffusing capacity of the lung for carbon monoxide (DLCO) were performed using the body box technique (Zan 500 Body Plethysmography; nSpire Health Limited, Hertford, UK).
Sputum induction was performed as previously described [18].
CT scans of the chest were performed at full inspiration using 16-slice Brightspeed and 64-slice Lightspeed (GE Healthcare, Milwaukee, WI, USA) with the following parameters: 120 KV, 100 mA·s, collimation 1 mm, reconstruction slice thickness 0.65 mm, reconstruction slice separation 0.5 mm and pitch of one. The data were reconstructed with a CHEST filter. All scans were evaluated centrally at the University of Edinburgh (Edinburgh, UK). Airway dimensions were measured using the software Pulmonary Workstation 2.0 (VIDA Diagnostics, Iowa City, IA, USA) to plot an airway path from which airway profiles were generated on cross-sections orthogonal to the airway path. Airway dimensions were measured in the right lower lobe posterior basal segmental bronchus (designated RB10). The following CT airway values were obtained: RB10 % wall area; RB10 wall thickness (mm); RB10 lumen area (mm2).
MMP-9 assays were carried out using the MMP-9 enzyme immunoassay and MMP-9 activity kits (catalogue numbers DMP900 and F9M00, respectively; R&D Systems Europe Ltd, Abingdon, UK). TIMPs were quantified by using fluorokine multiplex assay (LKT003; R&D Systems Europe Ltd).
MMP-9 (Affymetrix number 203936_s_at), TIMP-1 (201666_at) and TIMP-2 (231579_s_at) mRNA expression levels were analysed using Affymetrix U133+2 chips (Affymetrix, Santa Clara, CA, USA) and quantified by global normalisation to sputum cell and nasal respiratory epithelial cell samples obtained from healthy never-smokers (n=15), healthy smokers (n=13), smokers with asthma (n=15) and never-smokers with asthma (n=15). Samples were selected by asthma severity and quality of RNA. Details of RNA isolation and quality are included in the online supplementary material.
Analysis
Continuous variables were summarised as median (interquartile range). Their comparison between different patient groups was by Wilcoxon test or Kruskal–Wallis test. Categorical variables were summarised as frequencies and percentages per category and were compared using Fisher’s exact test. Linear regression models were used to predict each marker from disease group, smoking status and additional covariates, testing for any group by smoking status interaction. Associations of log-transformed MMP-9/TIMP ratios with sputum neutrophil percentage and CT RB10 airway dimension measurements were assessed using the Pearson correlation coefficient with bootstrap 95% confidence intervals. The analyses were exploratory for the purpose of hypothesis generation, and statistical significance was accepted if p<0.05. All analyses were carried out in R version 2.15.0 (The R Project for Statistical Computing. www.r-project.org).
Results
Demographics and baseline characteristics
81 asthmatics (35 smokers and 46 never-smokers) and 43 healthy subjects (19 smokers and 24 never-smokers) took part in the study. The asthma and healthy control groups were matched for age and smoking histories (table 1). None of the patients were on oral corticosteroids or anti-interleukin-5 therapy. The never-smokers with asthma were similar to the smokers with asthma in terms of age, duration of asthma, body surface area, atopic status, spirometry, dose of inhaled corticosteroid, treatment with montelukast, induced sputum total cell counts, eosinophil and neutrophil proportion, and CT measurements of airway wall thickness, but had a higher DLCO (86% versus 77% predicted, p=0.002) and FeNO50 (23 ppb versus 9 ppb, p<0.001).
Sputum MMP-9/TIMP activity and concentration ratios
The ratio of sputum MMP-9 activity to sputum TIMP-1 or sputum TIMP-2 was lower in smokers with asthma compared with never-smokers with asthma (p=0.002 and p=0.001 respectively), and in healthy smokers compared with healthy never-smokers (p=0.018 and p=0.024, respectively) (fig. 1 and table 2). The difference in these ratios was due to elevated sputum TIMP concentrations in smokers compared to never-smokers (in both asthma and healthy controls) (table 2). There was no significant difference between groups according to asthma severity (table 3).
a) Sputum matrix metalloproteinase (MMP)-9 activity/tissue inhibitor of metalloproteinase (TIMP)-1 concentration ratio, b) MMP-9 activity/TIMP-2 concentration ratio, c) MMP-9 protein concentration/TIMP-1 concentration ratio, and d) MMP-9 protein concentration/TIMP-2 concentration ratio in patients with asthma and healthy controls. The ratio of sputum MMP-9 activity with sputum TIMP-1 concentration or sputum TIMP-2 concentration was lower in smokers with asthma compared with never-smokers with asthma (p=0.002 and p=0.001, respectively) and in healthy smokers compared with healthy never-smokers (p=0.018 and p=0.024, respectively).
The ratios of sputum MMP-9 concentration to sputum TIMP-1 or sputum TIMP-2 were similar between smokers with asthma and never-smokers with asthma and between both asthma groups and healthy never-smokers and smokers, respectively (fig. 1 and table 2). There was no significant difference between groups according to asthma severity (table 3).
Linear regression analyses of interaction demonstrated that neither age nor dose of inhaled corticosteroid had a significant effect on MMP-9/TIMP ratios and that smoking had a significant effect on MMP-9 activity/TIMP-1 (r= -0.738 (95% CI -1.276– -0.200), p=0.008) and on MMP-9 activity/TIMP-2 ratios (r= -0.518 (95% CI -0.960– -0.077), p=0.022).
Nasal respiratory epithelial cell and sputum cell MMP-9/TIMP expression ratios
Median (interquartile range) nasal respiratory epithelial cell MMP9/TIMP1 and MMP9/TIMP2 expression ratios were lower in smokers with severe asthma compared with never-smokers with severe asthma (0.01 (0.00–0.03) versus 0.05 (0.04–0.09) ×10−3 relative fluorescence units (RFU), p<0.001 and 0.02 (0.01–0.06) versus 0.10 (0.05–0.14) ×10−3 RFU, p<0.001, respectively). Ratios were also lower in healthy smokers compared with healthy never-smokers (p<0.001 for both ratios) (online supplementary table E1). Sputum cell MMP9/TIMP1 and TIMP2 expression ratios were higher in smokers with severe asthma compared with never-smokers with severe asthma (0.19 (0.09–0.27) versus 0.13 (0.06–0.18), p<0.001 and 0.27 (0.12–0.38) versus 0.13 (0.05–0.18), p=0.046, respectively) (online supplementary table E1).
Correlation of sputum MMP-9 protein and enzyme activity and TIMP concentrations with lung function, sputum neutrophil percentage and CT scan measures of airway dimensions
Lung function
In smokers with asthma sputum MMP-9 protein activity/TIMP-1 and TIMP-2 ratios were positively associated with post-bronchodilator FEV1 % pred (r=0.359 (95% CI 0.065–0.599) and 0.334 (0.008–0.587), respectively) and FEV1/forced vital capacity (FVC) ratio (0.365 (0.070–0.626) and 0.386 (0.105–0.614), respectively) (table 4). There were no significant associations with DLCO, except for sputum MMP-9/TIMP-2 ratio in never-smokers with asthma (-0.344 (-0.621– -0.028)).
Sputum neutrophil percentage
In never-smokers with asthma, MMP-9 activity/TIMP-1 and MMP-9 concentration/TIMP-1 were positively associated with sputum neutrophil percentage (0.373 (95% CI 0.020–0.697) and 0.579 (0.251–0.797), respectively) (table 4). Associations of MMP-9 concentration and activity/TIMP-2 were not significant in never-smokers with asthma, nor were any associations with sputum neutrophil percentage significant in smokers with asthma.
CT measure of airway wall dimensions
Sputum MMP-9 activity/TIMP-1 and TIMP-2 and sputum MMP-9 protein concentration/TIMP-1 and TIMP-2 ratios were positively associated with RB10 airway lumen area in smokers with asthma (r=0.450 (95% CI 0.138–0.717), 0.410 (0.102–0.668), 0.499 (0.189–0.728) and 0.320 (0.000–0.579), respectively) (fig. 2 and table 4). Sputum MMP-9/TIMP ratios were not significantly associated with RB10 airway wall thickening. There were no significant associations between the ratios and RB10 CT measures in never-smokers with asthma. RB10 airway lumen area in smokers with asthma and never-smokers with asthma did not correlate with body surface area (r=0.208, p=0.170 and r=0.066, p=0.680, respectively).
Association between a) sputum matrix metalloproteinase (MMP)-9 activity/tissue inhibitors of metalloproteinase (TIMP)-1 concentration ratio, b) MMP-9 activity/TIMP-2 concentration ratio, c) MMP-9 protein concentration/TIMP-1 concentration ratio, and d) MMP-9 protein concentration/TIMP-2 concentration ratio, with computed tomography (CT) airway lumen area in smokers with asthma. The ratios of sputum MMP-9 activity/TIMP-1 concentration, MMP-9 activity/TIMP-2 concentration, sputum MMP-9 protein concentration/TIMP-1 concentration, and MMP-9 protein concentration/TIMP-2 concentration correlated with right bronchial division 10 (RB10) airway lumen area in smokers with asthma (r=0.450 (95% CI 0.138–0.717), r=0.410 (0.102–0.668), r=0.499 (0.189–0.728) and r=0.320 (0.000–0.579), respectively).
Discussion
A reduced ratio of MMP-9/TIMPs in nonsmokers with asthma has been implicated in airway remodelling [8, 10–12]. We sought to determine if the ratio of sputum MMP-9/TIMPs is associated with lung function or CT measures of airway dimensions in smokers and never-smokers with asthma. Particular strengths of our study were recruitment of smokers with asthma, who account for up to a third of all adults with asthma, and yet in whom there are limited published data describing the balance between MMP-9 and TIMPs, or CT imaging to assess airway wall dimensions. We also describe the use of functional assays for MMP-9 activity, which adds information about the proportion of the enzyme that becomes activated.
Our main new findings were that sputum MMP-9 activity/TIMP ratios were reduced in smokers with asthma compared with never-smokers with asthma and that reduced ratios in smokers with asthma were associated with reduced post-bronchodilator FEV1/FVC ratio and reduced airway lumen area as measured using CT. These findings suggest that airway narrowing is affected by the functional balance between MMP-9 and its inhibitors, which appears crucial in the dynamic process of airway remodelling and collagen deposition in asthma [8, 9, 21]. We also found reduced ratios of MMP9/TIMP1 and MMP9/TIMP2 mRNA expression levels in nasal respiratory epithelium in the subset of smokers with severe asthma, suggesting that these changes occurred in the sub-epithelial compartment where MMPs have to be activated and their interactions with TIMPs are locally regulated [22] and where the airway remodelling in asthma is most damaging.
Acute exposure of bronchial epithelial cells from patients with asthma to cigarette smoke in vitro reduces the MMP-9/TIMP-1 ratio [16]. Based on this finding, it was postulated that a reduction in the MMP-9/TIMP ratio may contribute to remodelling in smokers with asthma due to increased deposition of ECM and airway wall thickness [16]. However, our findings did not support this hypothesis, in that we could not demonstrate any association between MMP-9/TIMP ratio and increased CT airway wall thickening in smokers with asthma. Instead, an important finding in the smokers with asthma group were the positive correlations between sputum MMP-9 activity/TIMP-1 and TIMP-2, MMP-9 concentration/TIMP-1 and TIMP-2 ratios, and CT measurements of segmental airway lumen area. The link between reduced MMP-9/TIMP ratios and a reduced airway lumen in smokers with asthma is not explained by our data. Lower sputum MMP-9 activity/TIMP-1 and TIMP-2 ratios were associated with reduced FEV1 % pred in smokers with asthma, an association reported by others in nonsmokers with asthma [13, 14, 23]. In addition, we showed that lower ratios were associated with reduced post-bronchodilator FEV1/FVC, implying persistent airflow obstruction in subjects with lower ratios. Possibly a low MMP-9/TIMP ratio is associated with the bronchoconstrictor effects of cigarette smoke, increased secretions or epithelial inflammation, although the latter might also be expected to increase airway wall thickness. Body surface area could influence airway dimensions, but we found no association with airway lumen area. The excess of TIMP over MMP-9 may be an exaggerated attempt to protect the bronchi against metalloproteinase degrading activity, leading to increased extracellular matrix deposition [12].
Unlike the results obtained by Vignola et al. [8] in mild, steroid naïve asthmatics, we did not observe a difference in asthmatic nonsmokers compared with healthy controls, possibly as the laboratory techniques were different and majority of our subjects were on inhaled corticosteroids. Inhaled corticosteroids have been shown to reduce the expression of MMP-9 and increase the expression of TIMP-1 in bronchial biopsy tissue from patients with asthma [24]. In the present study, the dose of inhaled corticosteroid did not have a significant effect on MMP-9 or TIMP measurements in sputum.
We found that the proportion of sputum neutrophils correlated positively with sputum MMP-9 activity/TIMP-1 and MMP-9 concentration/TIMP-1 ratios in never-smokers with asthma, but not in smokers with asthma. Previous studies using sputum and bronchoalveolar lavage samples have reported a similar association in nonsmokers with asthma [13, 14, 25]. MMP-9 and TIMP-1 are known to be secreted by a variety of cells, including bronchial epithelial cells, eosinophils, mast cells and alveolar macrophages, and these may contribute more, relative to neutrophils, in orchestrating the changes in MMP-9 and TIMP-1 levels in smokers with asthma.
We recognise that there were limitations in the study, such as the fact that we did not perform bronchial biopsies to confirm the sputum, nasal and CT findings, and that the cross-sectional study design precludes collection of data on longitudinal changes to the balance between sputum MMP-9 and TIMPs. The inclusion of different severities of asthma reduces the numbers for conclusive statistical sub-analysis; however, it provides useful pilot data for larger studies of the effects of asthma severity.
In conclusion, these results show, for the first time, that the balance of MMP-9 and TIMPs is lower in smokers with asthma compared with never-smokers with asthma. A reduced sputum MMP-9 activity/TIMP ratio is associated with worse post-bronchodilator FEV1 and FEV1/FVC ratio and decreased sub-segmental CT measurements of airway lumen in smokers with asthma. This result suggests that an imbalance of MMP-9 and TIMPs may contribute to structural changes to the airways in this group.
Acknowledgments
The authors would like to thank Joyce Thompson, Jane Lafferty, Maureen Brannigan (University of Glasgow, Glasgow, UK) and Kathleen Masterson (NHS Greater Glasgow and Clyde) for sample collection, and Lisa Jolly and Iona Donnelly (University of Glasgow) for processing the samples and sputum cytology.
Footnotes
This article has supplementary material available from erj.ersjournals.com
Support statement: This work was funded by an award (INF-GU-090) from the Translational Medicine Research Collaboration, a consortium made up of the Universities of Glasgow, Edinburgh, Aberdeen and Dundee and the four associated NHS Health Boards (Greater Glasgow and Clyde, Lothian, Grampian, and Tayside), Scottish Enterprise and Pfizer (formerly Wyeth) and supported financially by NHS Research Scotland (NRS), through the Scottish Primary Care Research Network.
Conflict of interest: Disclosures can be found alongside the online version of this article at erj.ersjournals.com
- Received June 28, 2013.
- Accepted April 21, 2014.
- ©ERS 2014
References
