Abstract
Rationale: Chronic obstructive pulmonary disease (COPD) has a genetic component, explaining susceptibility. Tumor necrosis factor (TNF)-α polymorphisms have been associated with COPD, but it is unclear if genotype influences clinical phenotype, protein expression, and bioactivity.
Objectives: To determine if a functional polymorphism was important by assessing TNF-α expression and activity and its association with clinical severity over time.
Methods: Patients with COPD with rs361525 polymorphism were matched to patients with COPD without rs361525 polymorphism. TNF-α, its antagonists, and downstream mediators were measured in plasma and sputum. To determine TNF-α bioactivity, IL-8 secretion from primary bronchial epithelial cells (PBECs) was measured, and neutrophil migration was assessed using sputum from both subject groups in the presence and absence of TNF-α antibody. Subjects were followed annually and compared.
Measurements and Main Results: Patients with polymorphism had more chronic bronchitis, a lower body mass index, and a greater annual decline in FEV1 than patients with COPD without rs361525 polymorphism. TNF-α concentrations were 100-fold higher in airway secretions from the patients with the rs361525 polymorphism, with no difference in TNF-α antagonists. Their lung secretions contained more IL-8 and myeloperoxidase, consistent with downstream inflammation. Sputum from patients with rs361525 polymorphism induced greater secretion of IL-8 from PBECs and increased neutrophil migration. These effects could be abrogated by TNF-α antibody, demonstrating the bioactivity of TNF-α in lung secretions from this group.
Conclusions: This TNF-α polymorphism is associated with clinical features of disease including progression. There is clear evidence of TNF-α overexpression and bioactivity with neutrophilic inflammation. The polymorphism is likely to be a factor that influences a COPD disease phenotype and its progression.
It is accepted that chronic obstructive pulmonary disease has a genetic component, explaining susceptibility and there has been interest in functional polymorphisms of the TNF-α gene. Studies to date have focused upon clinical associations and systemic measures of TNF-α with contradictory results.
This study finds that TNF-α polymorphism genotype influences clinical phenotype (with a greater incidence of chronic bronchitis, a faster decline in FEV1 and a lower body mass index). The genotype is also associated with greater TNF-α expression in lung secretions and increased downstream signaling in vivo and bioactivity in vitro, suggesting that the polymorphism is likely to be a factor that influences chronic obstructive pulmonary disease phenotype and its progression.
Chronic obstructive pulmonary disease (COPD) is a group of diseases defined by predominantly irreversible airflow obstruction that includes bronchitis, emphysema, and small airways disease. It is widely accepted that COPD is driven by inflammation, but, despite its high prevalence and vigorous research activity, few genetic associations (other than α1-antitrypsin deficiency) have been found that affect the function of genes thought to be central to the pathophysiology of the inflammatory disease. Where such functional abnormalities have been found, clinical evidence of the defect and its sequelae have yet to be identified.
Tumor necrosis factor (TNF)-α is thought to be an important proadhesive cytokine in COPD. Moderately increased levels of TNF-α have been reported in serum and bronchial secretions in some (3) but not all (4, 5) studies of patients with COPD, although there seem to be clear increases during exacerbations (6). TNF-α has also been associated with the systemic manifestations of COPD, including a low body mass index (BMI) (7), perhaps via leptin (8), and abnormal resting energy expenditure (9). Mouse models with an inducible TNF-α gene construct have shown that overexpression of TNF-α is associated with the development of emphysema and a general increase in lung inflammation (10), probably by inducing matrix metalloproteinase production (11). TNF-α receptor knock-out mice demonstrate reduced smoking or elastase-induced emphysema in comparison with the wild type (12, 13). However, a recent trial of anti-TNF therapy did not show clear benefit in a general cohort of patients with COPD (14), perhaps implying that this cytokine is not important in COPD unless it is present in concentrations that are outside usual regulatory constraints.
There have been a number of studies examining the effects of polymorphisms of the TNFA gene in COPD, although results have been contradictory. For example, in a pilot study of patients with COPD and alpha-1–antitrypsin deficiency, the rs361525 polymorphism was associated with increased sputum concentrations of TNF-α (15) and chronic bronchitis (16). In a different study, a number of TNF-α polymorphisms were found to be associated with COPD, including rs361525 (17). However, there have been other studies where an association with lung disease was not found (18, 19). Most of the large population studies of TNF-α polymorphisms have not included rs361525 and instead have focused on other polymorphisms (20, 21). The association differences may be explained by the low numbers included in many of the studies, racial variation in gene expression, and differences in specific patient characteristics (21).
Although there have been no large population studies, the minor allele frequency of rs361525 in Caucasians is believed to range from 3 to 6% (22, 23), and studies in COPD have found minor allele frequencies of 5 to 10% (15, 21). Given the estimated minor allele frequency of the rs361525 polymorphism and the heterogenicity of COPD as a condition, population studies are likely to be prohibitive due to the large numbers that would need to be included. No studies have gone beyond characterizing COPD phenotype to understand the functional impact of the polymorphism on the disease process.
We report a functional polymorphism of the TNF-α gene that has been associated with chronic bronchitis in COPD (16). We focused upon the rs361525 polymorphism due to its association with chronic bronchitis and COPD. Previous studies have found no increase in plasma expression in the presence of this polymorphism (24), but we have demonstrated its overexpression in airway secretions and an increase in its functional activity that influences neutrophilic responses in COPD and therefore is likely to play a role in the clinical phenotype and its progression. Some of the results of these studies have been previously reported in the form of an abstract (25).
All subjects provided written informed consent after appropriate ethical approval was obtained to conduct the studies included in this work.
Subjects with COPD and the rs361525 polymorphism were identified from genetic studies of 450 individual patients recruited to a COPD database. Of the 450 patients tested, 12 were heterozygous for the rs361525 polymorphism, 10 of whom were daily sputum producers. No patients were homozygous.
The polymorphism group consisted of subjects with moderate to severe COPD defined by GOLD criteria (26). They were exsmokers between 50 and 80 years of age. All patients had been clinically stable for at least 8 weeks before sample collection with no changes in medication during this time. Alternative and concomitant lung disease was excluded clinically, physiologically, and by high-resolution computed tomography. Subjects had no other significant medical conditions and were matched for lung function, smoking status and exposure, age, sex, absence of airway colonization, and treatment to patients without the polymorphism who were also regular sputum producers. Control patients with COPD were selected after screening the 131 patients with COPD in the database with chronic bronchitis but without the rs361525 polymorphism, on an individual basis, to provide the best possible match for each individual patient with the polymorphism. This extensive matching was performed because it is known that inflammatory burden alters with differing disease severity assessed by FEV1, with smoking exposure, the presence of bacteria, and therapeutic treatments.
DNA extraction was performed using a modified Nucleon Bacc II kit (Tepnel Life Science, Manchester, UK) and quantified using Picogreen (Molecular Probes Inc., Paisley, UK). Genotyping was performed using TacMan genotyping technologies (Applied Biosystems, Carlsbad, CA) as described previously (16).
Serial plasma samples were collected from 12 patients with the polymorphism for 3 days and were used to measure TNF-α and the TNF-α antagonists (TNFsR1 and TNFsR2). The mean result per subject was used in further analysis.
Serial spontaneous sputum samples were collected from the 10 sputum producers over 4 hours (from rising) after the mouth-washing procedures to minimize salivary contamination as described previously (27). The samples were aliquoted, and the first sample was ultracentrifuged (50,000 × g for 90 min at 4°C) to prepare a sol phase sample to determine mediator levels of TNF-α and TNFsR1 and TNFsR2, IL-8, and myeloperoxidase (MPO). Samples were stored in aliquots at −70°C until analyzed. The second aliquot was used for quantitative microbiological analysis. All subjects produced a sputum sample on each of the three visits.
Mediators were measured using Enzyme Amplified Sensitivity Immunoassay (R&D Systems, Abingdon, UK) and are expressed in molar concentrations except for MPO, which is expressed as mg/ml. All assays were validated as described previously (28) to determine their working range, the variability of mediator measurements, and spike recovery (29). The TNF-α assay measured free and bound cytokine. Differential cell counts were not available because the samples were historical and of limited volume. However, MPO was used as a surrogate marker for neutrophils because it has been shown to correlate well with neutrophil counts in previous studies (30, 31).
To assess the potential for TNF-α bioactivity, primary bronchial epithelial cells (PBECs) (Lonza Group Ltd, Switzerland) from healthy donors were cultured to confluence in 6-well culture plates (Invitrogen, UK). Cell viability was assessed using trypan blue exclusion, and culture plates were included in experiments if viability exceeded 95%. Cells were stimulated with TNF-α (250 pM) and pooled individual sputum samples from subjects with and without the polymorphism in the presence and absence of TNF-α monoclonal antibody (1.6 μM) (R&D Systems, UK). Cells were incubated in 5% CO2 for 12 hours, and the supernatant was harvested for quantification of IL-8, providing cell viability remained above 95%. Each experiment was repeated in triplicate. The mean value was obtained for each experiment and used in analysis.
To assess the chemotactic potential of sputum samples from both patient groups, migratory studies were performed. Neutrophils were isolated from the whole blood of six healthy control subjects as described previously (32). The neutrophils (> 95% pure, > 97% viable, by exclusion of trypan blue) were resuspended in RPMI 1640 medium (Flow Laboratories, Rickmansworth, UK) containing 0.15% BSA (Sigma-Aldrich, Poole, UK) at concentrations of 106 cells per ml. The chemotaxis assay was performed using a modified Dunn Chamber (Weber Scientific International Ltd, Teddington, UK) as described previously (33, 34). In brief, peripheral neutrophils (250 μl) were placed on a coverslip coated with 7.5% albumin (Sigma-Aldrich) and incubated for 20 minutes at 20°C. The coverslip was placed inverted on the viewing field of the modified Dunn chamber, and 75 μl of negative or positive control, TNF-α, or pooled sputum sample was added to the chemoattractant well.
Neutrophil migration was studied with a negative control (buffer), IL-8 10 nM (positive control), TNF-α (250 pM) with and without TNF-α antibody (1.6 μM). The migratory effect of pooled sputum from patients with the rs361525 polymorphism (incubated with and without TNF-α antibody) was compared with pooled sputum from patients without the polymorphism, again with and without the TNF-α antibody. Each sputum pool was made using 20 μl of each sample from each patient, and experiments were repeated six times using a different donor's neutrophils, with the mean value used for subsequent analysis. Previous time course experiments confirmed that all migration studies could be performed without significant differences in migratory dynamics due to the experimental process (data not shown).
Migration was studied using time-lapse recordings obtained using a Zeiss (Welwyn Garden City, UK) Axiovert 100 inverted microscope with 20 slides captured using Improvision OpenLab software (PerkinElmer, Waltham, USA). The java software ImageJ (Wayne Rasband, NIH, Bethesda, MD) was used to analyze 10 randomly chosen cell tracks. All analysis was performed by a single analyst who was blinded to the experimental conditions. Migration was assessed as follows: (1) Cell speed (measured as the distance traveled between two frames, referred to as random speed of movement or chemokinesis), (2), cell velocity (speed in a consistent direction, referred to as chemotaxis), (3) directional persistence (a measure of the continuity of cell orientation over time, calculated by the cosine of the angle between directions in consecutive frames) was assessed such that cells that tend to move in a straight line, or execute slow changes in direction, have a high persistence value, whereas cells that move randomly and rapidly change direction do not; and (4) chemotactic index (a measure of the accuracy of the cell's directional orientation) was calculated as the cosine of the angle between the cell's direction and the orientation of the chemoattractant gradient.
All mediators and antibodies were supplied by R&D systems (UK). All studies, including those using chemoattractants and antibodies, were performed after relevant dose and time response experiments and repeated with the TNF-α antibody alone and a matched irrelevant antibody to ensure that antibody alone did not effect IL-8 secretion or chemotaxis.
Subjects were reviewed annually for 3 years in the stable state. On each review, exacerbation frequency and treatments were documented, postbronchodilation lung function was performed, and plasma and sputum samples were collected. Serial lung physiology was used to measure decline.
Data are expressed as mean + SD unless otherwise indicated. Mediator expression and migratory parameters were compared (after log conversion if necessary) using a t test, and differences were considered to be significant at P < 0.05. Subject numbers were defined by the prevalence of the polymorphism; however, power calculations performed on pilot studies in usual COPD determined that the number available provided 80% power to detect a 30% difference in TNF-α, a 25% difference in TNFsR1, TNFsR2, IL-8, and MPO and a 25% difference in chemokinesis and chemotaxis in a two-by-two ANOVA (the primary analytic outcome), with P = 0.05. Decline in physiology was determined using the regression line for annual measurements. Statistical analysis was performed using statistical software SPSS (Chicago, IL).
Subject characteristics are summarized in Table 1. All patients had COPD (characterized by nonreversible airflow obstruction) and a significant smoking history. Patients were matched for age, sex, smoking exposure and status, severity of airflow obstruction, current therapy, and absence of airway colonization. No patients were receiving oral prednisolone (although all were receiving inhaled steroids, according to guidelines recommendations), and all were clinically stable with no exacerbations in the 8 weeks before the study. There was no evidence of bronchiectasis in either patient group. Bacterial studies of sputum were negative in all patients, and none was currently taking antibiotics. Interestingly, 83% of patients initially found to be positive for the polymorphism (10 of the 12 found to be rs361525 positive) met the MRC definition for chronic bronchitis, whereas only 31% of patients without the polymorphism had chronic bronchitis (131 out of 438 records checked). Using a χ2 test, the number of patients with chronic bronchitis in the polymorphism group was significantly greater (P < 0.0001) than would be predicted in this group (3.8), demonstrating that the polymorphism was associated with a chronic bronchitis phenotype.
COPD + rs361525 | COPD − rs361525 | |
|---|---|---|
| n | 10 | 10 |
| Age, yr | 54 (39–72) | 53 (38–76) |
| Male, n (%) | 8 (80) | 9 (90) |
| BMI | 18 (15–21)† | 24 (19–27) |
| Smoking history | 100% previous | 100% previous |
| Pack-years | 27 (12–49) | 22 (12–34) |
| FEV1, percent predicted | 30 (13–58) | 27 (13–47) |
| FEV1/FVC ratio | 35 (21–58) | 28 (19–50) |
| Inhaled steroids | 10 (100%) | 10 (100%) |
| Exacerbations per year | 2.1 (0–3) | 1.9 (0–4) |
Patients with the rs361525 genotype also had a lower BMI than matched patients without the rs361525 genotype (median BMI, 18 vs. 24; median difference, 5.2; P = 0.04), but there were no other differences in smoking exposure, physiology, exacerbation frequency, treatments, or microbial colonization (only one patient was colonized in each group). Patients with the polymorphism had a greater annual decline in absolute FEV1 and the percent predicted FEV1 for age, race, and sex over four data points (absolute decline in FEV1: 73 ± 27 ml/yr vs. 26 ± 15 ml/yr; P = 0.03 and decline in percent predicted FEV1 3.5 ± 0.7 vs. 0.7 ± 0.7; P = 0.01, respectively). There were no differences in the decline of total lung capacity or measures of gas transfer. Exacerbation frequency, bacterial colonization, and BMI did not change during the study period.
Three sequential sputum and plasma samples were collected from each individual each year, and the mean result per person was used for further analysis. Plasma samples were collected from all patients (including the two patients with the polymorphism but without chronic bronchitis); however, sputum sample collection and subsequent cellular studies were performed using the samples of the 10 patients with chronic bronchitis and the matched control subjects. There were no significant differences in daily 4-hour sputum sample volume between groups. There were no differences in TNF-α concentrations or concentrations of its naturally occurring antagonists in plasma samples from patients with or without the polymorphism (TNF-α concentrations: 0.15 ± 0.04 pM vs. 0.12 ± 0.05 pM; the antagonist TNFsR1: 21.2 ± 1.3 pM vs. 20.6 ± 1.3 pM; and the antagonist TNFsR2: 32.7 ± 2.0 pM vs. 25.4 ± 1.6 pM, respectively).
At the first visit, TNF-α concentrations were significantly higher in sputum samples from patients with COPD with the rs361525 polymorphism compared with those without, by a ratio of 100 to 1 (mean, 130 ± 39 pM vs. 1·33 ± 0.44 pM; P = 0.00016). There was significantly less TNFsR1 in the sputum samples from patients with COPD with the rs361525 polymorphism compared with those without (mean 65 ± 11 pM vs. 127 ± 23 pM; P = 0.04), but no differences in concentrations of TNFsR2 (mean 68 ± 21 pM vs. 25 ± 6 pM; P = 0.1). These data are summarized in Figure 1.
Figure 1. The mean concentrations of tumor necrosis factor (TNF)-α and the TNF-α antagonists TNFsR1 and TNFsR2 in spontaneous sputum from patients with and without the rs361525 genotype and chronic obstructive pulmonary disease. Each data point represents one subject's data (logged). The mean is represented by the horizontal bar. GG indicates data from subjects without the polymorphism; AG indicates data from subjects with the polymorphism. The significance of differences between groups (p) is shown.
[More] [Minimize]The sputum samples collected from patients with the rs361525 genotype contained significantly higher concentrations of IL-8 (mean 47 ± 13 nM vs. 2.57 ± 0.8 nM; P < 0.0001) and MPO (mean, 1.69 ± 0.3 mg/ml vs. 0.17 ± 0.04 mg/ml; P = 0.0001). These data are summarized in Figure 2. This suggested there was more downstream inflammation present in sputum samples taken from patients with the polymorphism because TNF-α stimulates the release of the chemoattractant IL-8 and MPO reflects subsequent neutrophil recruitment and activation. The variability concentrations of TNF-α, IL-8, and MPO are shown in Table E1 of the online supplement.
Figure 2. The mean concentrations of IL-8 and myeloperoxidase (MPO) in spontaneous sputum from patients with and without the rs361525 genotype and chronic obstructive pulmonary disease. Each data point is the average result from one patient (logged). The overall mean is represented by the horizontal bar. GG indicates data from subjects without the polymorphism; AG indicates data from subjects with the polymorphism. The significance of differences between groups (p) is shown.
[More] [Minimize]To confirm whether TNF-α present in sputum was biologically active, confluent primary bronchial epithelial cells (PBECs) were stimulated with sputum alone, sputum that had been preincubated with TNF-α antibody, TNF-α alone, and TNF-α preincubated with its monoclonal antibody. A pooled sputum sample from each individual was formed by using 100 μl from each of the three sequential samples. Experiments were repeated three times per subject.
Quiescent PBECs released 0.36 ± 0.03 nM IL-8 over 12 hours, and stimulation with 250 pM TNF-α increased this to 6.7 ± 0.8 nM. The addition of TNF-α antibody had little effect on the quiescent PBECs (IL-8 secretion after 12 h, 0.25 ± 0.01 nM) but reduced the TNF-α response to baseline (0.2 ± 0.03 nM).
Sputum from subjects with the polymorphism increased IL-8 secretion to 3.7 ± 0.4 nM (P < 0.0001), and this was reduced (P = 0.004) by the TNF-α antibody to 1.8 ± 0.2 nM, indicating that about 50% of the IL-8 produced by the lung secretions was TNF-α dependent. In comparison, sputum from patients without the polymorphism induced less IL-8 secretion (0.95 ± 0.2 nM) (mean difference, 2.73 nM; P = 0.001). Furthermore, the addition of TNF-α antibody had no effect on the result for sputum alone (0.59 ± 0.07 nM; P = 0.1) (Figure 3), suggesting that the TNF-α in these samples was not inducing downstream inflammation.
Figure 3. The effects of preincubation with the tumor necrosis factor (TNF)-α monoclonal antibody on IL-8 secretion from primary bronchial epithelial cells (PBECs) stimulated with sputum. Each column is the mean IL-8 secretion from PBECs after 24 hours of incubation with the experimental condition given (with standard error bars), repeated three times and averaged for each sample. The addition of TNF-α antibody significantly decreased IL-8 secretion compared with TNF-α alone (*P < 0.0001) and sputum (sol) from patients with the rs361525 genotype (**P = 0.004) but not patients without the polymorphism. GG indicates data from subjects without the polymorphism; AG indicates data from subjects with the polymorphism.
[More] [Minimize]To assess the chemotactic potential of sputum samples from both patient groups and the role of TNF-α, migratory studies were performed using isolated peripheral neutrophils. Neutrophils migrated to shallow gradients of TNF-α or to pooled sputum with and without TNF-α antibody. Each experimental condition was studied on six separate occasions.
Neutrophils moved with increased chemokinesis (random movement) in gradients of TNF-α alone compared with the negative control (mean speed, 4.0 ± 0.2 μm/min vs. 2.4 ± 0.09 μm/min; P < 0.0001). The effect of pure TNF-α was fully abrogated with the addition of TNF-α antibody but not with the matched irrelevant antibody. However, TNF-α did not alter the velocity (directional speed or chemotaxis), persistence of direction, or accuracy of migration compared with the negative control.
There was more neutrophil migration in the presence of pooled sputum samples from patients with the polymorphism than from those without in terms of random speed of migration (chemokinesis: mean, 5.74 ± 0.2 μm/min vs. 4.57 ± 0.1 μm/min, respectively: P = 0.0001) and directed movement toward the sputum (chemotaxis: mean, 1.59 ± 0.3 μm/min vs. 0.78 ± 0.2 μm/min, respectively; P = 0.004). Migratory persistence (a measure of directional change) and overall accuracy (measured using vector analysis) were not different between the samples. Figure 4 demonstrates examples of tracks of neutrophils migrating toward sputum from subjects with or without the polymorphism. Figure 5 summarizes the overall migratory chemokinesis and chemotaxis to sputum in the presence and absence of the TNF-α antibody. In addition, significantly more neutrophils (assessed by direct counting) migrated to the polymorphism sputum compared with the normal sputum (expressed as a percentage of all neutrophils: 93 ± 6% vs. 70 ± 13; P = 0.03).
Figure 4. Typical migratory tracks from neutrophils in the presence of shallow gradients of sputum from patients with and without the TNF-α polymorphism. GG indicates data from a patient without the polymorphism; AG indicates data from a patient with the polymorphism. Each figure is a still from a migratory video. The large black arrow represents the chemotactic gradient. The small black arrows indicate the migratory pathway of one cell. The arrow begins where the cell started, and the arrowhead indicates where the cell was at the end of filming. Ten pathways are shown. Migration to rs361525 sputum had greater speed (5.7 μm/min vs. 4.6 μm/min) and greater velocity (1.6 μm/min vs. 0.7 μm/min) when compared with control sputum.
[More] [Minimize]
Figure 5. Differences in neutrophil chemokinesis and chemotaxis when migrating in shallow gradients of sputum from subjects with and without the rs361525 polymorphism. Each column is the overall mean migratory parameter (chemokinesis or random speed of movement and chemotaxis or directed speed of movement) with standard error bars shown. Negative control indicates migratory parameters when neutrophils were in the presence of RPMI (the control buffer). GG indicates data from pooled sputum from subjects without the polymorphism; AG indicates data from pooled sputum from subjects with the polymorphism. Each experiment was repeated six times, and the average data were used. The effects of preincubation with the tumor necrosis factor (TNF)-α antibody are shown. Sputum from patients with the polymorphism induced more neutrophil migration in terms of random chemokinesis (P = 0.0001) and directed chemotaxis (P = 0.004). Chemokinesis was significantly reduced by the addition of TNF-α antibody in the rs361525 sputum (P = 0.001), but the reduction in chemotaxis was not statistically significant (P = 0.09).
[More] [Minimize]The mean migratory parameters before and after addition of TNF-α antibody are shown in Table 2. TNF-α antibody caused a significant reduction in chemokinesis (random movement) in response to samples with the polymorphism from 5.74 ± 0.2 μm/min to 4.9 ± 0.2 μm/min (P = 0.001) but caused no change in chemotaxis (directed movement), directional persistence, or accuracy. TNF-α antibody did not alter any migratory parameter of neutrophils to sputum from patients without the rs361525 polymorphism.
Speed | Velocity | Persistence | Accuracy | |||||
|---|---|---|---|---|---|---|---|---|
| Negative control | 2.4 ± 0.2 | 0.02 ± 0.01 | 0.11 ± 0.06 | 0.01 ± 0.02 | ||||
| GG | ||||||||
| Sputum | 4.6 ± 0.1 | 0.8 ± 0.2 | 0.24 ± 0.2 | 0.14 ± 0.02 | ||||
| Sputum plus TNF-α antibody | 4.4 ± 0.2 | 0.6 ± 0.3 | 0.22 ± 0.2 | 0.09 ± 0.02 | ||||
| AG | ||||||||
| Sputum | 5.7 ± 0.2 | 1.6 ± 0.3 | 0.32 ± 0.2 | 0.2 ± 0.02 | ||||
| Sputum plus TNF-α antibody | 4.9 ± 0.2† | 1·4 ± 0.3 | 0·20 ± 0·3 | 0·2 ± 0·02 | ||||
TNF-α concentrations in sputum remained 2 log higher in the polymorphism group compared with patients with COPD without the polymorphism throughout the study period (Table E1).
Although TNF-α has been implicated in the pathophysiology of COPD, these data provide unique and compelling evidence of the role of a recognized polymorphism associated with COPD. The polymorphism was associated with recognized markers of disease progression, including chronic sputum production, a reduced BMI, and faster decline in FEV1, emphasizing its clinical relevance. Furthermore, protein expression was significantly higher in lung secretions, with evidence of downstream inflammation in vivo and in vitro bioactivity.
The polymorphism occurs in the promoter region of the gene and is known to increase TNF-α gene transcription in vitro (35); however, its functional role in disease has been unclear. rs361525 has been associated with a number of unrelated inflammatory conditions, including COPD, in some but not all studies, graft versus host disease (36), psoriasis (37), and lymphoma (38). It is unlikely that the polymorphism alone causes these diverse conditions, but it instead may predispose toward inflammation with an altered innate and adaptive immune response in the presence of other environmental and genetic factors or triggers, which leads to disease.
These data show that patients with the polymorphism have a 2 log higher TNF-α concentration in their sputum compared with a well-matched group without the polymorphism, with no associated systemic rise in TNF-α.
Although absolute concentrations in individual samples are variable, we collected and combined sequential samples (n = 3) for each patient to markedly reduce this effect (39). Furthermore, the comparable levels of the antagonists between both patient groups indicate that this difference in TNF-α is not related to sample collection or nonspecific dilutional factors (e.g., contaminating saliva). Although the antagonists would tend to block TNF-α effects, the dynamics of interaction, including binding affinities, the active half life of TNF-α, and dissociation constants (40–42), would still favor TNF-α bioavailability in the polymorphic group where the TNF-α and antagonists in sputum were approximately equimolar. This is supported by the demonstration that one of the downstream TNF signals (IL-8) is also increased in the samples from patients with the polymorphism. This chemokine is a major neutrophil chemoattractant, and the higher levels of MPO (a marker of neutrophil recruitment and activation) in the samples indicate a further downstream effect.
The TNF-α, IL-8 axis was investigated further by the cellular experiments. Sputum contains several factors with the ability to increase IL-8 production by epithelial cells (43). Sputum from patients without the polymorphism induced an increase in IL-8 production that was not abrogated by TNF-α antibody, which is in keeping with previous studies of IL-8 production from A549 cells after stimulation with sputum with and without appropriate concentrations of TNF-α antibody (44). However, not only did the samples with the polymorphism stimulate a far greater IL-8 response, but the TNF-α antibody abrogated this by approximately 50%, indicating a significant TNF drive.
Based upon studies of pure TNF-α (data not shown), the reduction of IL-8 secretion in the presence of antibody (to 1.88 nM) suggests that the sputum samples contained, on average, approximately 75 pM of active TNF-α. This is less than the average measured mean amount (130 pM), suggesting that only a proportion of the TNF-α is biologically active, with the remainder being inhibited probably by the antagonists (TNFsR1 and TNFsR2) because the ELISA used in this study measures active and bound inactive TNF-α. These data therefore provide a mechanism to explain (at least in part) the higher IL-8 in the sputum samples, driven by active TNF-α.
In vitro migratory studies with airway secretions indicate that IL-8 provides at least 30% of the neutrophilic migratory drive of sputum in COPD (45). The presence of the polymorphism with higher subsequent IL-8 production would be expected to increase neutrophil recruitment and activation. Although we have been unable to quantify the neutrophil numbers in these secretions, MPO (a recognized surrogate for PMN number and activation [46]) was increased in the polymorphism samples, supporting this process. Furthermore, our in vitro studies confirmed that polymorphism sputum samples caused a greater number of neutrophils to migrate than samples from subjects without the polymorphism.
The pooled polymorphism samples induced more neutrophil chemotaxis and chemokinesis than pooled samples from patients without the polymorphism. Dose–response experiments with the modified Dunn chamber have shown that IL-8 is predominantly chemotactic until concentrations are in excess of 100 nM (data not shown), whereas TNF-α is predominantly chemokinetic even at low concentrations (47). To assess this further we undertook studies of neutrophil migration, and the reduction in chemokinesis after the addition of TNF-α antibody suggests that this cytokine may predominantly drive chemokinesis rather than chemotaxis or migratory accuracy as these parameters were less effected. Potentially, removing the influence of excess TNF-α may positively effect neutrophil migration by decreasing random movement (which would be associated with more tissue damage by the process of quantum proteolysis [48–50]) without altering the neutrophils ability to migrate accurately (a necessary neutrophil function to reach the site of inflammation or infection). Increased neutrophil chemokinesis with impaired chemotaxis has been associated with worse outcomes in pulmonary infections (51, 52), suggesting that this may be a detrimental effect of the polymorphism in the patients studied here.
The overall implications of our findings have their basis in neutrophil-mediated lung damage. The polymorphism is associated with increased bioactive TNF-α. leading to increased IL-8 production, neutrophil migration, activation, and more random cell movement. Neutrophil activation would lead to subsequent neutrophil degranulation and the release of neutrophil proteinases. Neutrophil proteinases are known to cause all of the pathological features of COPD, including epithelial damage, impaired ciliary beat frequency, mucous gland hyperplasia, and mucus secretion (46). The polymorphism would therefore enhance these effects and could explain the relationship to chronic bronchitis in usual COPD and α1-antitrypsin deficiency (16), the increased incidence of daily mucus production, and the faster decline in FEV1 seen in the patients studied here.
TNF-α effects neutrophil function by priming cells, increasing adhesion molecule interactions (53), and arresting the cell at sites of inflammation (54). TNF priming also increases neutrophil degranulation (55) and superoxide production (56) in the presence of other inflammatory stimuli. In subjects with an excess of TNF-α, these effects may be amplified, especially if there is an ongoing inflammatory drive (such as inflammatory bowel disease or COPD) leading to increased tissue damage. These factors make anti-TNF strategies appealing, and they are already in use in debilitating conditions such as Crohn's disease and rheumatoid arthritis (where systemic levels of TNF-α are high), with evidence of clinical efficacy (57). Trials of anti-TNF therapy have also been undertaken in asthma with some evidence of efficacy (58, 59), although there have been concerns about subsequent infection risk and safety (60). This could be a treatment strategy in the subgroup of patients with COPD and excess pulmonary TNF-α, perhaps in the form of an inhaled agent, to decrease systemic side effects.
Increased bronchitic symptoms are associated with worse outcomes in COPD (61), which highlights the potential clinical importance of this polymorphism. Our data also suggest that the polymorphism is associated with a faster subsequent decline in FEV1, which is in keeping with a greater inflammatory burden and increased neutrophil presence; however, the numbers included in this study precluded a detailed regression analysis that would be required to robustly determine if this is a true association.
Patients in the polymorphism and nonpolymorphism groups had similar values of FEV1 at baseline as a direct result of the extensive matching process because we wished to reduce any differences in COPD characteristics that might otherwise cause inflammation. This seems counterintuitive if the polymorphism influences the decline in FEV1. However, although the patients were matched at baseline, this may not reflect the influences of factors before this point when smoking habit, exacerbation type and frequency, treatment, and development of chronic bronchitis may have differed. After the matching point, all things were equal, suggesting that the subsequent decline was related to the polymorphism alone.
The BMI of subjects with the polymorphism was lower than those without, and, because TNF-α has also been implicated in muscle wasting (7, 8), it is likely that this extrapulmonary effect is also a consequence of increased TNF-α bioavailability associated with the polymorphism. At first sight, the lack of increase in plasma TNF-α in these patients argues against this axis. However, plasma has a large volume of distribution, and detectable TNF in this compartment would be decreased by rapid binding to tissue receptors, dissociating measurement from biological effect.
The pathogenesis of COPD is poorly understood. Genetic and environmental factors are key because only 20% of smokers develop significant disease, and it is likely that differing combinations of genetic susceptibility dictate disease development phenotype and progression. Responses to generic treatments have been disappointing, but it may be possible to develop specific treatments based upon the knowledge of genetic determinants and their biochemical and physiological outcomes (pharmacogentics). If the results described herein represent a TNF-α dependent clinical phenotype, it may be that it also represents an anti–TNF-α therapy responsive population. Although the incidence of this polymorphism is low in the patient cohort studied here, it is more common than α1-antitrypsin deficiency in usual COPD. Further studies in carefully defined populations are indicated to confirm this hypothesis.
The authors thank Mrs. Diane Griffiths and Mr. Ross Edgar for help with patient recruitment and clinical data collection.
| 1. | Riordan JR, Rommens JM, Kerem B, Alon N, Rozmahel R, Grzelczak Z, Zielenski J, Lok S, Plavsic N, Chou JL, et al. Identification of the cystic fibrosis gene: cloning and characterization of complementary DNA. Science 1989;245:1066–1073. |
| 2. | Jeppsson JO. Amino acid substitution glu-lys in alpha 1 antitrypsin piz. FEBS Lett 1976;65:195–197. |
| 3. | Hacievliyagil SS, Gunen H, Mutlu LC, Karabulet AB, Temel I. Association between cytokines in induced sputum and severity of chronic obstructive pulmonary disease. Respir Med 2006;100:846–854. |
| 4. | Kuschner WG, D'Alessandro A, Wong H, Blanc PD. Dose dependent cigarette smoking-related inflammatory changes in healthy adults. Eur Respir J 1996;9:1989–1994. |
| 5. | Sapey E, Ahmad A, Bayley D, Newbold P, Snell N, Rugman P, Stockley RA. Imbalances between interleukin-1 and tumor necrosis factor agonists and antagonists in stable COPD. J Clin Immunol 2009;29:508–516. |
| 6. | Aaron SD, Angel JB, Lunau M, Wright K, Fex C, Le Saux N, Dales RE. Granulocyte inflammatory markers and airway infection during acute exacerbation of chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2001;163:349–355. |
| 7. | Di Francia M, Barbier D, Mege JL, Orehek J. Tumour necrosis factor alpha levels and weight loss in chronic obstructive pulmonary disease. Am J Respir Crit Care Med 1994;150:1453–1455. |
| 8. | Calikoglu M, Sahin G, Unlu A, Ozturk C, Tamer L, Ercan B, Kanik A, Atik U. Leptin and TNF alpha levels in patients with chronic obstructive pulmonary disease and their relationship to nutritional parameters. Respiration 2004;71:45–50. |
| 9. | Nguyen LT, Bedu M, Caillaud D, Beaufrere B, Beaujon G, Vasson M, Coudert J, Ritz P. Increased resting energy expenditure is related to plasma TNF alpha concentration in stable COPD patients. Clin Nutr 1999;18:269–274. |
| 10. | Vuillemenot BR, Rodriguez JF, Hoyle GW. Lymphoid tissue and emphysema in the lungs of transgenic mice inducibly expressing TNF alpha. Am J Respir Crit Care Med 2004;30:438–448. |
| 11. | Wright JL, Tai H, Wang R, Wang X, Churg A. Cigarette smoke up-regulates pulmonary vascular matrix metalloproteinases via TNF alpha signaling. Am J Physiol Lung Cell Mol Physiol 2007;292:L125–L133. |
| 12. | Churg A, Wang RD, Tai H, Wang X, Xie C, Wright JL. Tumour necrosis factor alpha drives 70% of cigarette smoke-induced emphysema in the mouse. Am J Respir Crit Care Med 2004;170:492–498. |
| 13. | Lucey EC, Keane J, Kuang PP, Snider GL, Goldstein RH. Severity of elastase induced emphysema decreased in TNF alpha and IL-1 beta receptor deficient mice. Lab Invest 2002;82:79–85. |
| 14. | Rennard SI, Fogarty C, Kelson S, Long W, Ramsdell J, Allison J, Mahler D, Saadeh C, Siler T, Snell P, et al. The safety and efficacy of infliximab in moderate to severe chronic obstructive pulmonary disease. Am J Respir Crit Care Med 2007;175:926–934. |
| 15. | Wood AM, Gough SC, Stockley RA. Sputum TNFA levels are associated with TNFA genotype [abstract]. Am J Respir Crit Care Med 2009;179:A2942. |
| 16. | Wood AM, Simmonds MJ, Bayley DL, Newby PR, Gough SC, Stockley RA. The TNF alpha gene relates to clinical phenotype in alpha-1-antitryspin deficiency. Respir Res 2008;9:52–57. |
| 17. | Gingo MR, Silveira LJ, Miller YE, Friedlander AL, Cosgrove GP, Chan ED, Maier LA, Bowler RP. Tumour necrosis factor polymorphisms are associated with COPD. Eur Respir J 2008;31:1005–1012. |
| 18. | Higham MA, Pride NB, Alikhan A, Morrell NW. Tumour necrosis factor alpha gene promotor polymorphism in chronic obstructive pulmonary disease. Eur Respir J 2000;21:444–449. |
| 19. | Ferrarotti I, Zorzetto M, Beccaria M, Gile LS, Porta R, Ambrosino N, Pignatti, P.F., Cerveri, I, Pozzi, E., Luiseeti, M. Tumour necrosis factor family genes in a phenotype of COPD associated with emphysema. Eur Respir J 2003;21:444–449. |
| 20. | Hersh CP, Demeo DL, Lange C, Litonjua AA, Reilly JJ, Kwiatowski D. Attempted replication of reported chronic obstructive pulmonary disease candidate gene associations. Am J Respir Cell Mol Biol 2005;33:71–78. |
| 21. | Castaldi PJ, Cho MH, Cohn M, Langerman F, Moran S, Tarragona N, Moukhachen H, Venugopal R, Hasimja D, Kao E, et al. The COPD genetc association compendium: a comprehensive online database of COPD genetic associations. Hum Mol Genet 2010;19:526–534. |
| 22. | Simmonds MJ, Heward JM, Howson JMM, Foxall H, Nithiyananthan R, Franklyn JA, Gough SCL. A systematic approach to the assessment of known TNF alpha polymorphisms in graves disease. Genes Immun 2004;5:267–273. |
| 23. | NCBI SNP database. Accessed January 2010. Available from: www.ncbi.nlm.gov/projects/SNP/spr-ref.cgi. 2010 Accessed January 2010. |
| 24. | Bayley JP, Ottenhoff THM, Verweij CL. Is there a future for TNF promoter polymorphisms? Genes Immun 2004;5:315–329. |
| 25. | Sapey E, Wood AM, Ahmad A, Griffith D, Edgar E, Stockley RA. Imbalances in TNF in patients with a TNF polymorphism and a COPD phenotype. Eur Respir J 2009;81:850. |
| 26. | Pauwels RA, Buist AS, Calverley PM, Jenkins CR, Hurd SS, Committee GS. Global strategy for the diagnosis, management and prevention of chronic obstructive pulmonary disease. NHLBI/WHO global initiative for chronic obstructive lung disease (GOLD) workshop summary. Am J Respir Crit Care Med 2001;163:1256–1276. |
| 27. | Sapey E, Bayley D, Ahmad A, Stockley RA. The validation of assays used to measure biomarkers in exhaled breath condensate. Eur Respir J 2008;32:1408–1409. |
| 28. | Woolhouse IS, Bayley DL, Stockley RA. Effect of sputum processing with dithiothreitol on the detection of inflammatory mediators in chronic bronchitis and bronchiectasis. Thorax 2002;57:667–671. |
| 29. | Stockley RA, Bayley D. Validation of assays for inflammatory mediators in sputum. Eur Respir J 2000;15:778–781. |
| 30. | Gompertz S, Bayley D, Hill SL, Stockley RA. Relationship between airway inflammation and the frequency of exacerbations in patients with smoking related COPD. Thorax 2001;56:36–41. |
| 31. | Gompertz S, O'Brien C, Bayley D, Hill SL, Stockley RA. Changes in bronchial inflammation during acute exacerbations of chronic bronchitis. Eur Respir J 2001;17:1112–1119. |
| 32. | Mikami M, Llewellyn-Jones CG, Bayley D, Stockley RA. The chemotactic activity of sputum from patients with bronchiectasis. Am J Respir Crit Care Med 1998;157:723–728. |
| 33. | Zicha D, Dunn GA, Brown AF. A new direct-viewing chemotaxis chamber. J Cell Sci 1991;99:769–775. |
| 34. | Andrew N, Insall R. Chemotaxis in shallow gradients is mediated independently of ptdins-3-kinase by biased choices bewteen random protrusions. Nat Cell Biol 2007;9:193–200. |
| 35. | Bayley JP, de Rooij H, van den Elsen PJ, Huizinga TW, Verweij CL. Functional analysis of linker-scan mutants spanning the -376, -308, -244 and -238 polymorphic sites of the TNF alpha promoter. Cytokine 2001;14:316–323. |
| 36. | Do V, Tsuneto LT, Sossai CR, Lieber SR, Marques SB, Vigorito AC, Aranha FJ, de Brito Eid KA, Oliveria GB, Miranda EC, et al. IL2 and TNFA gene polymorphisms and the risk of graft-versus-host disease after allogenic naematopoietic stem cell transplantation. Scand J Immunol 2007;66:703–710. |
| 37. | Nedoszytko B, Szczerkowski-Dobosz A, Zablotna M, Glen J, Rebala K, Roskiewicz J. Associations of promoter region polymorphism in the tumour necrosis factor alpha gene and early onset psoriasis vulgaris in a northern Polish population. Br J Dermatol 2007;157:165–167. |
| 38. | Wang SS, Cerhan JR, Hartge P, Davis S, Cozen W, Severson RK, Chatterjee N, Yeager M, Chanock SJ, Rothman N. Common genetic variants in proinflammatory and other immunoregulatory genes and risk for non-hodgkin lymphoma. Cancer Res 2006;66:9771–9780. |
| 39. | Sapey E, Bayley D, Ahmad A, Newbold P, Snell N, Stockley RA. Inter-relationships between inflammatory markers in stable COPD patients with bronchitis; the intra and inter patient variability. Thorax 2008;63:493–499. |
| 40. | Grell M, Wajant H, Zimmerman G, Scheurich P. The type 1 receptor is the high affinity receptor for soluble tumour necrosis factor. Proc Natl Acad Sci USA 1998;95:570–575. |
| 41. | Medvedev AE, Espevik T, Ranges G, Sundan A. Distinct roles of the two tumour necrosis factor (TNF) receptors in modulating TNF and lymphotoxin alpha effects. J Biol Chem 1996;271:9778–9784. |
| 42. | Beutler B, Milsark IW, Cerami A. Calchectin/tumour necrosis factor: production, distribution and metabolic fate in vivo. J Immunol 1985;135:3972–3977. |
| 43. | Johnston WW. The cytopathology of the respiratory tract: a review. Am J Pathol 1976;84:372–424. |
| 44. | Ahmad A, Bayley DL, He S, Stockley RA. Myeloid related protein 8/14 stimulates interleukin-8 production in airway epithelial cells. Am J Respir Cell Mol Biol 2003;29:523–530. |
| 45. | Woolhouse IS, Bayley D, Stockley RA. Sputum chemotactic activity in chronic obstructive pulmonary disease: effect of alpha 1 anti-trypsin deficiency and the role of leukotriene b4 and interleukin 8. Thorax 2002;57:709–714. |
| 46. | Stockley RA. Neutrophils and the pathogenesis of COPD. Chest 2002;121:151S–155S. |
| 47. | Smart SJ, Casale TB. TNF alpha induced transendothelial neutrophil migration is IL8 dependent. Am J Physiol 1994;266:L238–L245. |
| 48. | Liou TG, Campbell EJ. Quantum proteolysis resulting from release of single granules by human neutrophils: a novel, nonoxidative mechanism of extracellular proteolytic activity. J Immunol 1996;157:2624–2631. |
| 49. | Campbell EJ, Campbell MA, Boukedes SS, Owen CA. Quantum proteolysis by neutrophils: implications for pulmonary emphysema in alpha 1 anti trypsin deficiency. J Clin Invest 1999;104:337–344. |
| 50. | Cepinskas G, Sandig M, Kvietys PR. Paf induced elastase dependent neutrophil transendothelial migration is associated with the mobilisation of elastase to the neutrophil surface and localised to the migrating front. J Cell Sci 1999;112:1937–1945. |
| 51. | Niwa Y, Kasama T, Miyachi Y, Kanoh T. Neutrophil chemotaxis, phagocytosis and parameters of reactive oxygen species in human aging: cross sectional and longitudinal studies. Life Sci 1989;44:1655–1664. |
| 52. | Egger G, Aigner R, Glasner A, Hofer HP, Mitterhammer H, Zelzer S. Blood polymorphonuclear leukocyte migration as a predictive marker for infections in severe trauma: comparison with various inflammatory parameters. Intensive Care Med 2004;30:331–334. |
| 53. | Lauterbach M, O'Donnell P, Asano K, Mayadas T. Role of TNF priming and adhesion molcules in neutrophil recruitment to intravascular immune complexes. J Leukoc Biol 2008;83:1423–1430. |
| 54. | Sheppard FR, Kelher MR, Moore EE, McLaughlin MJ, Banerjee A, Silliman CC. Structural organizaton of the neutrophil naph oxidase: phosphorylation and translocation durng priming and activation. J Leukoc Biol 2005;78:1025–1042. |
| 55. | Ogilvie AC, Hack CE, Wagstaff J, van Mierlo GJ, Erenberg AJ, Thomsen LL, Hoekman K, Rankin EM. IL-1beta does not cause neutrophil degranulation but does lead to IL-6, IL-8 and nitrite/nitrate release when used in pateints with cancer. J Immunol 1996;156:389–394. |
| 56. | Berkow RL, Wang D, Larrick JW, Dodson RW, Howard TH. Enhancement of neutrophil superoxide production by pre-incubation with recombinant human tumour necrosis factor. J Immunol 1987;139:3788–3791. |
| 57. | Eiger A, Sinha B, Hartmann G, Endres S. Taming TNF: strategies to restrain this pro-inflammatory cytokine. Immunol Today 1997;18:487–492. |
| 58. | Howarth PH, Babu KS, Arshad HS, Lau L, Buckley M, McConnell W, Beckett P, Al Ali M, Chauhan A, Wilson SJ, et al. Tumour necrosis factor as a novel therapeutic target in symptomatic corticosteroid dependent asthma. Thorax 2005;60:1012–1018. |
| 59. | Morjaria JB, Chauhan A, Babu KS, Polosa R, Davies DE, Holgate ST. The role of a soluble TNF alpha receptor fusion protein (etanercept) in corticosteroid refractory asthma: a double-blinded randomised placebo controlled trial. Thorax 2008;63:584–591. |
| 60. | Oliveria GB, Polosa R. Etanercept in chronic severe asthma. Thorax 2006;61:640. |
| 61. | Vestbo J, Prescott E, Lange P. Association of chronic mucus hypersecretion with FEV1 decline and chronic obstructive pulmonary disease morbidity. Copenhagen city heart study group. Am J Respir Crit Care Med 1996;153:1530–1535. |
