INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The inflammatory process in the bronchial tree is complex. It involves the release of many mediators including chemoattractants, and cytokines that regulate the adhesion molecules, the processes of cell migration, and their activation and degranulation. The summation of these processes in chronic destructive lung diseases such as chronic bronchitis and emphysema is the release of the neutrophil proteinase elastase (1). This enzyme is a major common mediator of many of the pathological changes seen in chronic lung disease including epithelial damage (2), reduced ciliary beat frequency (3), mucus gland hyperplasia (4), mucus secretion (5), and inactivation of many of the critical lung host defenses (6, 7). However, in order to have these effects, the elastase released from the activated neutrophil has to overcome the naturally occurring inhibitors such as secretory leukoprotease inhibitor (SLPI) and ₁-antitrypsin.

There have been many studies of the effects of elastase in isolation in vivo and, in addition, a multitude of in vitro experiments exploring the role of this mediator. However, there have been very few studies assessing the complex interplay of inflammatory cells and appropriate mediators in established lung disease. Recent limited studies have indicated that neutrophils in the airway of patients with chronic obstructive pulmonary disease (COPD) are increased (8, 9) and that these cells or their product myeloperoxidase (MPO) is related to the presence of a single chemoattractant, interleukin-8 (IL-8) (8- 10). Although these studies have suggested that IL-8 may be the major factor influencing neutrophil recruitment, there are clearly other chemoattractants that may also play a role (11). This concept is of importance because in vitro studies have indicated that chemoattractants may interact in an additive way (11) and in addition there may be a hierarchical response influencing the cell and its migration pattern (12).

Furthermore, studies of the relationship between chemoattractants and neutrophils may be further complicated, although most of this concept is based on in vitro studies. Neutrophil activation can lead to the release of the chemoattractants IL-8 (13) and leukotriene B₄ (LTB₄) (14) that could result in further neutrophil recruitment. Release of elastase from the neutrophil may stimulate epithelial cells to produce more IL-8 (15) and at the same time reduce production of its own natural inhibitor, SLPI (16) thereby perpetuating its function. Furthermore, impairment of host defenses by elastase could facilitate bacterial colonization, and endotoxin release from bacteria also provides a further potential mechanism for neutrophil recruitment by stimulating production of IL-8 by epithelial cells (17).

The inflammatory process itself has been shown to be associated with worsening lung function (18) although separation of cause or effect has not been possible. Nevertheless marked bronchial inflammation is a feature of patients with bronchiectasis with (19) and without (20) cystic fibrosis (CF). Whereas the former patient group often has severe airflow limitation, this is not always the case in non-CF bronchiectasis (21, 22) suggesting that the association is not simple.

The purpose of the present study was to assess the interrelationship of the inflammatory markers in patients with a wide spectrum of neutrophil influx. In particular we wished to assess the relationships between neutrophil influx as reflected by the sputum MPO concentration and the chemoattractants (IL-8 and LTB₄). In addition, we wanted to relate the concentrations of active neutrophil elastase (NE) to the chemoattractants as well as its own inhibitor SLPI, and bronchial protein leak (sputum/serum albumin ratio). Finally, we wanted to determine whether there was a relationship between FEV₁ (percentage of predicted) and inflammation as reflected by MPO or protein leak.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patient Selection

Patients with established bronchial disease and a history of chronic bronchitis, as defined by the Medical Research Council (MRC) (23), were studied in the stable clinical state. Patients who had an infection within the preceding 2 mo of the study were excluded.

Sputum and Serum Processing

Sputum was collected over a 4-h period, from rising, into sterile containers from all patients in a stable clinical state. Sputum was ultracentrifuged at 50,000 × g for 90 min at 4° C; the sol phase was removed and stored at 70° C until analyzed. Venous blood was collected into a plain vacutainer tube, allowed to clot, centrifuged at 1,500 g for 10 min at 25° C, and the serum removed and stored at 70° C until analyzed.

Measurement of Sputum MPO

Sputum MPO activity was used as an assessment of neutrophil influx and measured as described previously (24). The MPO concentration was derived from the standard curve using a single preparation of lysed neutrophils (freeze/thaw). Results for individual samples were obtained from this curve by interpolation and expressed as arbitrary units/ml. The interassay coefficient of variation (standard deviation/ mean [%]) was less than 10%.

Measurement of NE Activity from Sputum Sol Phase

NE standard was purified from empyema pus according to the method of Baugh and Travis (25). The activity of the NE standard was determined by active site titration using published kinetic constants (26). NE activity present in the samples was measured spectrophotometrically using the synthetic substrate methoxysuccinyl-ala-ala-pro-val-paranitroanilide (MeOSAAPVpNa) (Sigma-Aldrich Company Ltd., Poole, Dorset, UK). Aliquots of 20 µl of standard or sample were added to wells of a microtiter plate (Life Technologies, Ltd., Paisley, UK), followed by MeOSAAPVpNa (0.3 mM) in phosphate buffer (0.2 M, pH 8.0). The reaction was allowed to continue for 1 h at 37° C and then stopped by the addition of 200 µl of 1 N acetic acid. The absorbance was read at 410 nM and the activity derived from the standard curve by interpolation. The interassay coefficient of variation was less than 1% with a lower limit of detection of 1 nM. Samples with activity below 1 nM were considered to be 0.8 nM for statistical purposes.

Measurement of IL-8, LTB₄, and SLPI from Sputum Sol Phase

Untreated sputum sol phase was obtained as described previously and assayed using commercially available specific ELISAs: IL-8 and SLPI (R&D Systems Europe Ltd., Abingdon, UK); LTB₄ (Amersham International plc, Buckinghamshire, UK). The lower limit of detection for these assays was 0.008 nM for IL-8, 0.01 µM for SLPI, and 0.17 nM for LTB₄. All three of the commercial assays had an interassay coefficient of variation < 10% and resulted in greater than 85% recovery when sputum samples were spiked with the appropriate pure reagent.

Sputum and Serum Albumin

Albumin in the sputum sol phase and serum was measured by radial immunodiffusion using a commercially available kit (The Binding Site Limited, Birmingham, UK). Five microliters of standard or sample was added to each well, and the plate was incubated for 72 h at 25° C. The ring diameters were then measured using an eye piece graticule and the albumin concentration present in the samples calculated by interpolation from the standard curve. The interassay coefficient of variation was less than 5% for both sputum and serum. The results were expressed as a sol phase/serum ratio (%) to determine the degree of protein leak.

Statistical Analysis

Values are reported as mean (± standard error). Correlations between the inflammatory markers were assessed by the Spearman correlation coefficient (2-tailed).

The Mann-Whitney U test for nonpaired data (2-tailed) was used to compare results between samples grouped by their elastase activity. Samples were subdivided into four groups: those with no detectable elastase activity (Group A); those with small amounts of elastase activity (1 to 50 nM) as Group B; those with a moderate level of elastase activity (50 to 100 nM) as Group C; and those with high levels of elastase activity (> 100 nM) as Group D. A value of p < 0.05 was considered to be significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The patients who entered the study included 101 samples from 55 subjects with chronic bronchitis (without ₁-antitrypsin deficiency), 61 samples from 40 patients with chronic bronchitis with homozygous (PiZ) ₁-antitrypsin deficiency, and 64 samples from 43 patients with idiopathic bronchiectasis diagnosed by high-resolution computed tomography or bronchogram. Table 1 reveals the characteristics of the study population and the major results section reflects analysis of all samples as individual data points.

MPO and Elastase Activity

MPO activity was detectable in all but one sample and 147 samples (65.0%) had detectable elastase activity > 1 nM. Figure 1 shows the correlation between MPO and elastase activity (r = 0.68, n = 226, p < 0.001).

View larger version (17K): [in this window] [in a new window]   Figure 1.   The relationship between MPO (arbitrary units/ml) and elastase activity (µM) is shown on logarithmic scales. Each circle represents the data from individual samples (r = 0.68, n = 226, p < 0.001).

Chemoattractants and MPO

All samples (226) contained detectable amounts of both chemoattractants (range = 0.05 to 87.38 nM for IL-8 and 0.34 to 263.34 nM for LTB₄). Figures 2 and 3 show the relationship between the levels of the chemoattractants IL-8 and LTB₄ and neutrophil influx (MPO). The overall data revealed a positive correlation between both chemoattractants and MPO activity (IL-8; r = 0.52, p < 0.001 and LTB₄; r = 0.41, p < 0.001). Similar correlations were present even when the outlying samples (MPO > 10 arbitrary units/ml) are excluded (see Figures 2 and 3).

View larger version (17K): [in this window] [in a new window]   Figure 2.   The relationship between MPO (arbitrary units/ml) and the chemoattractant IL-8 (nM) is shown (r = 0.52, p < 0.001). Each circle represents the result from a single sample (open circles represent samples with an MPO value < 10 arbitrary units/ml; closed circles represent samples with an MPO value > 10 arbitrary units/ml). After excluding these latter samples the relationship persisted (r = 0.40, p < 0.001).

View larger version (17K): [in this window] [in a new window]   Figure 3.   The relationship between MPO (arbitrary units/ml) and the chemoattractant LTB4 (nM) is shown (r = 0.41, p < 0.001). Each circle represents the result from a single sample (open circles represent samples with an MPO value < 10 arbitrary units/ml; closed circles represent samples with an MPO value > 10 arbitrary units/ml). After excluding these latter samples the relationship was retained (r = 0.43, p < 0.001).

Chemoattractants and Elastase Activity

The overall data revealed a positive correlation between both IL-8 and LTB₄ and elastase activity (IL-8 r = 0.55, p < 0.001 and LTB₄ r = 0.41, p < 0.001). In view of the wide scatter of elastase values (range, 0 to 8,564 nM) we further subdivided samples into four groups based on activity: those with no detectable elastase activity were classified as Group A (n = 79); those with small amounts of elastase activity (1 to 50 nM) as Group B (n = 92); those with a moderate level of elastase activity (50 to 100 nM) as Group C (n = 14); and those with high levels of elastase activity (> 100 nM) as Group D (n = 41). The mean (SE) levels of free elastase in each of the groups were: zero for Group A; 21.7 ± 1.4 nM for Group B; 71.1 ± 4.7 nM for Group C; and 1,739.7 ± 330.4 nM for Group D.

Figure 4 demonstrates the average IL-8 levels (± SE) and LTB₄ for the four sample groups (A-D). Low concentrations of IL-8 and LTB₄ were found in patients with no detectable elastase (IL-8 5.61 nM ± 0.67; LTB₄ 9.09 nM ± 1.47). However, in the presence of elastase activity (Group B) greater amounts of these chemoattractants were found (p < 0.001 for IL-8 and p < 0.005 for LTB₄). The highest levels were found in patients in Group D (elastase activity > 100 nM) with an average IL-8 concentration of 25.61 nM ± 2.48 and 36.51 nM ± 8.80 for LTB₄ (both p < 0.001) compared with Group A. 

View larger version (12K): [in this window] [in a new window]   Figure 4.   The relationship between the chemoattractants (IL-8 [nM] and LTB4 [nM]) and elastase activity (nM) is shown. Samples are grouped into those with no elastase activity, elastase activity between 1 to 50 nM, elastase activity between 50 to 100 nM, and those with elastase activity greater than 100 nM. The histograms represent the mean ± SE bar. The asterisk indicates values that are significantly greater than samples with no elastase activity (*p < 0.005; **p < 0.001).

Elastase Activity, SLPI, and Protein Leakage

SLPI was detectable in all sputum samples (range, 0.04 to 14.26 µM). There was a negative correlation between the SLPI concentration and the activity of NE (r = 0.49, p < 0.001). Separation of samples into their elastase ranges indicated that SLPI concentrations were unaltered in samples with absent or low concentrations of elastase (Groups A and B). However, the average concentration was reduced (p < 0.005) from 3.78 µM ± 0.51 (Group A) to 1.35 µM ± 0.41 in samples where elastase activity ranged from 50 to 100 nM (Group C). Furthermore a greater reduction (p < 0.001) was found in samples containing > 100 nM elastase activity where the average SLPI concentration was 0.66 µM ± 0.16. These results are summarized in Figure 5.

View larger version (9K): [in this window] [in a new window]   Figure 5.   The relationship between SLPI levels (mean ± SE) and elastase activity (nM) is shown. The asterisk indicates group values that are significantly greater than those with no elastase activity (*p < 0.005; **p < 0.001).

Protein leakage (as assessed by the sputum/serum albumin ratio) showed a positive correlation with elastase activity (r = 0.36, p < 0.001). In samples where elastase activity was undetectable (Group A) the degree of leakage was low (1.02 ± 0.14%). Protein leakage remained low in Group B (1.08 ± 0.10%), but increased in samples with elastase activity > 50 nM (Group C: 2.02 ± 0.51%, p < 0.05). Protein leakage reached a maximum (2.54 ± 0.42%) in samples containing greater than 100 nM elastase (Group D) (p < 0.005 compared with Group A).

Effect of Treatment

Patients with chronic lung disease are frequently treated with a variety of agents that may affect lung inflammation. No patients had taken oral steroids or antibiotic therapy within the preceding 2 mo of sample collection. There were 90 of the 226 samples, however, from patients on inhaled steroids with or without theophylline therapy. Exclusion of these samples did not alter the results seen for the group as a whole (data not shown).

Influence of Replicate Samples

Some patients provided more than one sample (usually two). These were treated as individual data points (see METHODS). Nevertheless, because it is possible that this approach could bias the interrelationships, we reanalyzed the data for single samples from each patient and this had no effect on the correlation or statistics (data not shown).

Relationship with FEV₁ (Percentage of Predicted)

There was no correlation between the severity of lung function and neutrophil influx (MPO) or protein leakage for the group as a whole. Subset analysis of the individual patient groups was undertaken because the data would be skewed by the patients with idiopathic bronchiectasis who usually have well-preserved lung function despite marked bronchial inflammation.

When the patient groups were studied independently there was still no correlation between FEV₁ (percentage of predicted) and either neutrophil influx or protein leakage in the patients with bronchiectasis or ₁-antitrypsin deficiency. In patients with chronic bronchitis, however, there was a negative correlation between FEV₁ (percentage of predicted) and both MPO and protein leakage (MPO, r = 0.37, p < 0.001; protein leakage, r = 0.42, p < 0.01). This relationship was retained when the analysis was confined to an average sample result from each patient (MPO, r = 0.31, p < 0.005; protein leakage, r = 0.54, p = 0.01). The data for protein leakage are summarized in Figure 6.

View larger version (9K): [in this window] [in a new window]   Figure 6.   The relationship between FEV1 (percentage of predicted) and protein leak (sol/serum albumin ratio [%]) is shown for patients with chronic bronchitis without 1-antitrypsin deficiency. The closed circles represent mean sample data for single patients (r = 0.54, p = 0.01).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

The current data spans a wide range of MPO activity used as a marker of the presence of bronchial neutrophils as suggested previously (8, 9, 27). Activation of these cells (as would occur during the migration process) leads to elastase release (28). The data show a good correlation between the MPO activity and elastase activity over the wide range, although this is less apparent at the lower end where many samples with detectable MPO had no detectable elastase activity. This is to be expected because the presence of bronchial inhibitors such as SLPI and ₁-antitrypsin would be expected to inhibit the low concentrations of elastase released when neutrophil numbers are low. In the present study we did not measure total elastase immunologically. This reflects the suggestion by previous workers (29), but also the fact that in vitro studies have suggested that the effects of elastase on inflammation are dependent on its activity.

Neutrophil recruitment is dependant upon the chemoattractants being released. Studies have shown that IL-8 and LTB₄ (11) are the major chemoattractants in patients with bronchial disease and indeed both correlate with the MPO activity as indicated here. The source of these two chemoattractants, however, is unknown. For instance, endotoxin and tumor necrosis factor-alpha (TNF-) (both of which are likely to be present in the airway) are known to increase IL-8 production by epithelial cells (17). In addition, IL-8 is stored within the specific granules of a neutrophil and is also released from this cell on activation (30). Finally, it has been suggested that elastase itself when released from the neutrophil can lead to IL-8 production by bronchial epithelial cells (15). Whatever the mechanism it would therefore be predicted that IL-8 levels should correlate not only with neutrophils (as indicated by the MPO concentration) but also with the elastase activity. Although the IL-8 levels measured here do correlate with MPO, the relationship is less robust than that for MPO and elastase activity, suggesting that the relationship is not a simple cause and effect but probably reflects multiple sources of IL-8. However, even though there is a highly statistical relationship between IL-8 and MPO and elastase, examination of the raw data suggests this may not be a simple linear relationship. Indeed the data would suggest that the relationship maybe curvilinear with a threshold effect leading to rapidly increasing concentration of neutrophil MPO once the IL-8 level exceeds approximately 10 nM (see Figure 2).

The relationship with LTB₄, on the other hand, seems to be much more straightforward; although there was a wide scatter of values for both MPO and LTB₄, the relationship tends to follow a general linear trend. This suggests that the neutrophil itself may be the source of both proteins.

It has been shown that LTB₄ and IL-8 are additive in their effect on neutrophil recruitment. It may well be therefore that at low levels of IL-8, recruitment of small numbers of neutrophils leads to release of LTB₄ from the activated cell. This may thereafter add to the chemoattractant gradient recruiting more cells in a linear fashion. However, once IL-8 levels exceed 10 nM the inflammatory process may become excessive leading to much greater neutrophil recruitment (as indicated by MPO) than would be expected by reviewing the LTB₄ data. This concept would be consistent with the potential curvilinear relationship between IL-8 and MPO as well as the data points for MPO displaced from the LTB₄ curve (see Figure 3). Again this concept would be consistent with a hierarchical organization of chemoattractants (12).

Alternatively it is possible that other chemoattractants are involved in samples where neutrophil recruitment (as indicated by MPO) is excessive. The solid symbols in Figures 2 and 3 represent samples where MPO exceeds 10 arbitrary units/ml. These samples appear displaced from the relationship of both chemoattractants to MPO, and this complex relationship needs further study. However, preliminary data indicated that of the 10 samples colonized with Pseudomonas aeruginosa, seven were in this high MPO group of 17 samples. It therefore remains possible that the colonizing organism may also influence cell migration.

Studies of the relationships of the inflammatory process to elastase are worthy of further comment. When released from the neutrophil, this enzyme would be rapidly inhibited by both ₁-antitrypsin and, in the airway, more particularly by SLPI. Our studies showed that there were low levels of SLPI associated with high levels of elastase activity. In vitro studies have indicated that elastase can adversely affect SLPI production by epithelial cells (16); thus, it would be expected that an inverse relationship should exist. However, the data as presented here indicate that SLPI concentration does not decrease until the elastase activity of the samples is in excess of 50 nM. The relationship between SLPI and elastase therefore is not a simple linear one and the data presented here may indicate a threshold at which significant interference with epithelial cell metabolism or epithelial damage occurs resulting in a reduction in SLPI secretion. At present whether this relationship reflects cause or effect is not known and interpretation will have to await intervention studies using specific antielastases. If the reduction in SLPI production is merely a reflection of the elastase activity in the airway, the introduction of an effective antielastase would be expected to lead to an acute rise in SLPI concentration. However if significant airway cell damage is responsible for the reduction in SLPI, a longer period of repair would be necessary before SLPI concentrations return to normal.

There is, however, a similar relationship between the elastase activity and inflammation in the airway as reflected in leakage of serum albumin. Protein leakage increased only when free elastase activity exceeded 50 nM, and further increased when the elastase activity exceeded 100 nM. The exact mechanism for protein leakage is currently unknown. Previous studies in patients with bronchiectasis have shown antibiotic therapy reducing the bacterial load and hence neutrophil influx, leading to a rapid reduction in protein leakage as sputum purulence and elastase activity disappear (31). The rapidity of this response would suggest that airway leakage is not a direct effect of epithelial cell destruction. The leakage may therefore reflect an effect of inflammation on the tight junction between epithelial cells (32) or the simultaneous release of vasoactive mediators in the airway. Once again interpretation of the responses will await studies with effective antielastases.

The relationship between elastase and IL-8 is of further interest. As indicated previously, studies have suggested that IL-8 release by epithelial cells can be induced by the presence of free elastase (15). However, because both mediators can also be derived from the neutrophil (although from different cell granules), it would be expected that the two would be interrelated. The data provided here show that the relationship is present although weak, and it is therefore unlikely that these two proteins reflect a cause and effect. Also, as indicated previously, the relationship between IL-8 and MPO may not be linear across a wide range of inflammatory situations unlike previous more limited studies (8). However, it is also possible that the presence of additional (as yet uncharacterized) factors may play a role when inflammation is excessive.

Finally Keatings and Barnes showed that increased neutrophil influx was associated with worse lung function in patients with COPD (18) although separation of cause or effect has not been possible. Similarly, the current study showed that both increased neutrophil influx (as assessed by the concentration of MPO) and increased protein leak were associated with worse lung function in patients with chronic bronchitis without ₁-antitrypsin deficiency. However there was no association in the other two patient groups (those with ₁-antitrypsin deficiency or bronchiectasis) although it should be noted that all patients with ₁-antitrypsin deficiency had an FEV₁ < 56% predicted which would tend to minimize the range of inflammation that might potentially be expected. Thus, in this group there was a limited spectrum of disease being assessed. The association is therefore complex and further studies are required to determine whether this observation in COPD represents a true cause and effect.

In summary, therefore, we have assessed the interrelationship between the neutrophil and some of its products and the inflammatory process in patients with a wide range of bronchial disease. Although relationships can be demonstrated, it is clear that these are complex, and understanding the interplay of various mediators will require the development of specific antagonists and appropriately designed intervention studies.