INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Lung transplantation has become a therapeutic option for various end-stage pulmonary vascular and parenchymal disorders. Median survival at 3 yr after transplantation is currently estimated at 50% (1). Long-term survival is limited primarily by chronic allograft rejection and infections (1). Although several risk factors for chronic rejection have been proposed (2), the pathophysiology is currently uncertain. Most research has concentrated on the role of lymphocytes and immunologic mechanisms in the development of chronic graft rejection. The contribution of other inflammatory cells such as neutrophils and macrophages and their products has largely been ignored.

Activated neutrophils and macrophages release matrix metalloproteases (MMPs), a group of enzymes with the combined ability to degrade virtually all known components of the extracellular matrix (ECM). In addition to neutrophils and macrophages, MMPs are products of fibroblasts (5), endothelial cells (6), and smooth muscle cells (7). As a group, MMPs are key modulators of ECM homeostasis contributing to normal physiologic processes such as growth, cell migration, inflammation, and wound healing (8). Their contribution to pathologic conditions, including various pulmonary disorders, remains the subject of intensive investigation. Previous reports have linked unregulated proteolytic activity to the pathogenesis of bronchiectasis (9), cystic fibrosis (10), emphysema (11), pulmonary fibrosis (12), and sarcoidosis (13). A common factor in all these conditions seems to be the presence of chronic inflammation, resulting in augmented protease activity and subsequent pulmonary parenchymal destruction and/or fibrosis.

Although often referred to collectively as collagenases, three classes of MMPs have been identified. These have been grouped somewhat arbitrarily according to their substrate specificity (14). The three classes include the collagenases, with specificity for types I, II, and III collagen; the gelatinases, with specificity for gelatin (denatured collagen) and type IV collagen; and the stromelysins, with specificity for proteoglycans and laminin. All share major structural and functional domains, including a zinc-dependent catalytic site (14). Their proteolytic activity, in the absence of an inflammatory stimulus, is controlled at various intracellular and extracellular levels. When these regulatory mechanisms break down, however, the potential for destruction of any ECM scaffold, including lung parenchyma, exists.

MMP-2 and MMP-9 are members of the gelatinase class with relative molecular weights in their latent forms of 72 and 92 kD, respectively (14). They have the ability to digest not only type IV collagen, the major component of basement membranes, but also act sequentially with interstitial collagenase (type I/III collagenase) to cause degradation of interstitial collagen, the major structural protein of the pulmonary parenchyma (15).

In this study, we investigated the cellular profile and activity of gelatinases in bronchoalveolar lavage fluid (BALF) from clinically stable lung transplant recipients, i.e., those without evidence of disease in the lung allograft. We hypothesize that a chronic inflammatory state is present after lung transplantation, even in the absence of known inflammatory stimuli. This results in the up-regulation of proteolytic activity, especially gelatinases capable of degrading basement membranes. These proteases may contribute to the structural changes associated with chronic allograft rejection after lung transplantation.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Bronchoscopy in Lung Transplant Recipients

For purposes of diagnosing occult infection or rejection, all lung transplant (LT) recipients at St. Louis University Hospital have undergone routine surveillance fiberoptic bronchoscopic examination weekly for the first month postoperatively, at 3 mo, 6 mo, and 12 mo, and then yearly thereafter. Bronchoscopy was performed according to methods previously described (16). For purpose of bronchoalveolar lavage (BAL), the bronchoscope was advanced and wedged in the right middle lobe or lingula of the transplanted lung. Three 50-ml aliquots of sterile nonbacteriostatic 0.9% NaCl were instilled and then gently aspirated via the wedged bronchoscope. The collected aliquots were pooled and placed on ice. All BALF specimens from LT recipients were evaluated to exclude active infection. This assessment included routine stains (gram stain, silver stain, Calcofluor, acid fast stain, and Papanicolau smear) and cultures (routine bacterial, acid fast bacterial, and viral, including shell vial culture for cytomegalovirus). Secretions were collected by a protected-specimen brush for purposes of quantitative bacteriologic specimens. Four to eight transbronchial biopsy specimens were taken from the affected segment of the transplanted lung for pathologic examination, with fluoroscopic confirmation. Specimens were obtained from the right middle lobe or lingula of the transplanted lung. Tissue specimens were fixed and sections stained with hematoxylin and eosin, Gomori's methenamine-silver, and carbolfuchsin. The biopsy was then assigned a histologic grade by pathologists according to criteria established by the International Society of Heart and Lung Transplantation (17).

Each BALF sample was catalogued using a grading scale based on the presence or absence of infection and histologic abnormalities of concurrently obtained transbronchial biopsies (Table 1). The scale used by our institution is a modified version of the histologic grading system originally established by the International Society of Heart and Lung Transplantation (17). For the purpose of this study, only BALF samples of clinically stable LT recipients were evaluated. Clinically stable patients were defined as those whose lung biopsies showed grade 0 histology (normal pulmonary parenchyma without evidence of mononuclear infiltration, hemorrhage, or necrosis) and whose stains and cultures were free of active infection. This resulted in the inclusion of cell counts from 49 BALF samples (23 LT recipients).

Bronchoscopy in Heart Transplant Recipients

Both heart transplant (HT) and LT recipients were maintained on similar immunosuppressive regimens for the prevention of rejection. The effects of these immunosuppressive agents, i.e., cyclosporine, azathioprine, and prednisone, on the number and degree of activation of inflammatory cells in BALF and their products including proteolytic enzymes are unknown. BALF specimens from 11 asymptomatic HT recipient volunteers were included to control for any alterations in BALF total white blood cell (WBC) counts, differential cell counts, and enzyme activity that may result from chronic treatment with these immunosuppressive agents. BAL in the HT group was performed in the same manner as described above, with the sample being taken from the right middle lobe. All BALF samples were processed in the same manner as that described for LT recipients except that no microbiologic or histologic studies were performed in HT or normal volunteers (NL).

Normal Subjects

Thirty-one healthy nonsmoking volunteers without known pulmonary disease underwent bronchoscopic evaluation. BAL was performed according to methods described above, with BALF samples taken from the right middle lobe and processed in the same manner.

BALF Cell Counts and Differentials

Total WBC counts were performed on a portion of all BALF specimens (uncentrifuged) using a hematocytometer. Pooled BALF (100 µl) was mixed with an equal volume of 0.2% bovine serum albumin in 0.9% saline solution and centrifuged (40 × g for 2 min). The preparations were then stained with Diff-Quik (Fisher Scientific, St. Louis, MO), and differential cell counts were performed on a total of 300 cells. The final remaining portion of BALF was centrifuged (400 × g for 15 min). The cell-free supernatant was aliquoted and stored at 70° C until analyzed.

Gelatinolytic Activity

To measure the presence of type IV collagenolytic/gelatinolytic activity, six randomly selected BALF specimens from each of the three groups (NL, HT, and LT) were subjected to sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) substrate zymography according to Banda and colleagues (18). There was an approximately 70% return of BALF from the total amount of NSS instilled into all patients evaluated. All samples were concentrated fivefold using a microconcentrator tube (Centricon-10; Amicon, Beverly, MA). Final protein concentrations were determined using the Bicinchoninic acid (BCA) assay (Pierce, Rockford, IL). Concentrated samples were then incubated (10 min) with 1 mM 4-aminophenyl-1-mercuric acetate (APMA; Sigma Chemical Co., St. Louis, MO) to activate latent enzymes (18). Samples were added to buffer (50 mM Tris [pH 6.8], 10% glycerol, 0.005% bromophenol blue, and 0.5% SDS) at a ratio of 3:1. A volume containing 30 µg total protein was loaded onto a 10% discontinuous SDS-polyacrylamide gel co-polymerized with 0.1% gelatin (Sigma), and proteins were resolved electrophoretically. The proteins were allowed to renature by removal of SDS during two 15-min washes with 2.5% Triton X-100 (Fisher Scientific). The gels were then incubated in substrate buffer containing 50 mM Tris, 0.15 M sodium chloride, 10 mM calcium chloride, 1 µM zinc chloride, and 0.02% sodium azide (60 h at 37° C). The reaction was stopped by staining the gels with 0.1% Coomassie blue R250 (Sigma) in 40% methanol and 10% acetic acid. The gels were destained for 30 min with 40% methanol and 10% acetic acid, which revealed enzyme activity as clear bands against a blue background. A positive control lane containing 0.3 µg purified Clostridium histolyticum collagenase (Worthington Biochemical Corp., Freehold, NJ) was also included. High and low molecular weight standards (Bio-Rad Laboratories, Richmond, CA) were run concurrently and used to determine enzyme molecular weights by determining their relative mobility. To further confirm the identity of the enzyme(s), zymograms of representative BALF samples from each of the three groups were incubated with either EDTA (20 mM), a metalloprotease inhibitor, or phenylmethylsulfonyl fluoride (PMSF; 2 mM), a serine protease inhibitor, added to the substrate buffer.

The gelatinolytic activity of cell-free BALF from LT recipients relative to peripheral neutrophils, and alveolar macrophages from a normal volunteer and a LT recipient, was also evaluated. Neutrophils were isolated from the blood of a normal donor by density gradient centrifugation using NIM isolation media according to the manufacturer's instructions (Cardinal Associates, Sante Fe, NM). Alveolar macrophages (> 90% pure) were collected from BALF by centrifugation (400 × g for 15 min). Cells (1 × 10⁶ cells/ml) were subjected to hypotensive lysis and sonication (80 s for the macrophages and 40 s for the neutrophils) to achieve complete disruption of cells. Debris-free samples (30 µg protein/lane) were loaded onto a single gel for comparison of relative collagenolytic activity. A positive control lane containing 0.3 µg purified C. histolyticum collagenase was also included.

Enzyme activity was quantified by measuring the intensity of the negative bands using a laser densitometer (ImageQuant; Molecular Dynamics, Sunnyvale, CA). With ImageQuant Software (Molecular Dynamics), intensity of the bands was converted to a linear graph whose peaks corresponded to areas of proteolytic activity (negative bands). Area under the curve at each band of activity was then calculated and reported as arbitrary units (A.U.).

Statistical Analysis

All data is reported as the mean ± SEM unless otherwise indicated. Differences among means were determined by one-way ANOVA for statistical analysis with homogenous variances. Significance of differences between groups was analyzed using the Tukey-Kramer test for multiple comparisons. If Bartlett's test showed the presence of significant variation among standard deviations, the Kruskal-Wallis nonparametric ANOVA was performed followed by Dunn's multiple comparison test. Correlations were determined using Spearman's correlation analysis for nonparametric data, with a two-tailed t test used to determine significance. All statistical analysis was performed using the InStat software program (GraphPad Software, San Diego, CA). A value of p < 0.05 was considered significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Patient Demographics

All evaluable BALF specimens were included in analysis of cell profiles (49 BALF specimens from 23 LT recipients). Demographic data for the LT recipients are provided in Table 2. Six specimens were then randomly selected from each of the three groups (NL, HT, LT) for zymographic analysis. In the LT group whose BALF samples were analyzed by zymography, indications for transplantation included emphysema related to ₁-protease inhibitor deficiency (two patients), smoking-induced chronic obstructive pulmonary disease (one patient), sarcoidosis (one patient), and congenital heart anomaly with Eisenmenger's syndrome (two patients). Three patients each underwent single lung transplantation and heart-lung transplants. BALF specimens in the LT group were collected a mean of 939 d after transplant (range: 216 to 2,221 d). All were maintained on immunosuppressive regimens that included cyclosporine, azathioprine, and prednisone (dosage range: 5 to 12.5 mg). None of the evaluated LT patients were current smokers.

Of the HT recipients whose samples were subjected to zymography, two patients never smoked, four were previous smokers with a mean smoking history of 24 pack-years (range: 15 to 30 pack-years), one of whom had quit but then resumed and was smoking approximately one-half pack per day at time of bronchoscopy. All patients were free of cardiac rejection as documented clinically and/or by concurrent endomyocardial biopsy.

Standard immunosuppressive regimens for both LT and HT recipients included maintenance therapy with cyclosporine, azathioprine, and prednisone. One HT recipient was also on methotrexate for previously diagnosed acute rejection. In both transplant groups, cyclosporine was dosed on an individual basis to maintain a steady-state level of 100 to 200 ng/ml.

BALF Cell Counts

BALF cell counts from all evaluable specimens revealed a significantly greater total WBC count in the LT group compared with the NL group (274 ± 30 versus 128 ± 8 cells/µl BALF, respectively; Figure 1a). The difference in total WBC count between the HT control group (202 ± 67 cells/µl BALF) and the LT group was not significant. The percentage of neutrophils (2 ± 0 versus 4 ± 1 versus 15 ± 3%, NL versus HT versus LT, respectively; Figure 2a) and total neutrophil count (2 ± 0 versus 4 ± 1 versus 55 ± 16 cells/µl BALF, respectively; Figure 1b) present in the BALF samples from LT recipients were significantly higher than in the BALF samples from NL control subjects. Although both the percent neutrophils and total neutrophils tended to be higher in the LT group as compared with the HT group, the differences were not statistically significant. The percent macrophages (80 ± 2 versus 80 ± 4 versus 67 ± 3%; Figure 2b) and total macrophages (103 ± 8 versus 160 ± 49 versus 170 ± 19 cells/µl BALF, NL versus HT versus LT, respectively; Figure 1c) did not differ significantly among the three subject groups. Insets in both Figures 1 and 2 represent the respective total number of leukocytes, neutrophils, and macrophages as well as percent neutrophils and macrophages from the subset of BALF samples analyzed by zymography (n = 6 in all groups). Comparing the zymography group (insets) to the entire population studied, similar results for total and differential cell counts were seen in both groups.

View larger version (26K): [in this window] [in a new window]   Figure 1.   Comparison of (a) total WBC count, (b) total neutrophil count, and (c) total macrophage count in BALF of the three populations studied (n = 31, 11, and 49, normal control subjects [NL] versus heart transplant recipients [HT] versus lung transplant recipients [LT], respectively). Inset in each graph represents the respective BALF cell counts of the subset of samples analyzed by zymography (n = 6 in all groups). *p < 0.05 versus NL.

View larger version (31K): [in this window] [in a new window]   Figure 2.   Comparison of (a) percent neutrophils and (b) percent macrophages in BALF among the three groups studied. Inset in each graph represents the respective percentage of cells found in the subset of samples analyzed by zymography. *p < 0.05 versus NL.

Gelatinolytic Activity

Three bands of gelatinolytic activity were visualized in one or more of the samples analyzed, corresponding to enzymes of molecular weights of 225, 92, and 72 kD. As shown in Figure 3, the 92 kD band was the most prominent band in all samples. There was an absence of proteolytic activity at 72 kD in all of the normal control subjects and two of six HT control subjects. Interestingly, the 72 kD activity was found in all LT recipients.

View larger version (34K): [in this window] [in a new window]   Figure 3.   Gelatin zymography of BALF samples from normal control subjects (NL), heart transplant recipients (HT), and lung transplant recipients (LT). Representative BALF samples from each of the three groups were electrophoresed on a 10% SDS-polyacrylamide gel impregnated with 0.1% gelatin. Gels were then allowed to incubate in substrate buffer for 60 h. Gelatin degradation was detected by negative staining. Apparent molecular weights were calculated using concurrently run, known molecular weight standards.

Quantitative measurements by densitometry confirmed the visual findings, with proteolytic activity reported as arbitrary units of area (A.U.) within the negative bands. As is seen in Figure 4, activities of the 225 kD (0.62 ± 0.14 versus 0.59 ± 0.25 versus 1.50 ± 0.28 A.U., NL versus HT versus LT, respectively), 92 kD (4.87 ± 0.14 versus 4.60 ± 1.72 versus 18.79 ± 4.92 A.U., NL versus HT versus LT, respectively), and 72 kD (0.0 versus 0.76 ± 0.58 versus 5.87 ± 1.75 A.U., NL versus HT versus LT, respectively) regions were all significantly greater in the LT recipients than in NL and HT control subjects, suggesting an up-regulation of the gelatinase activity at these molecular weights. Activity at the 92 kD band accounted for the majority of the total proteolytic activity detected in each of the three groups (89%, 77%, and 72%, NL, HT, and LT, respectively). Activity of all three bands was very similar in both control populations.

View larger version (22K): [in this window] [in a new window]   Figure 4.   Densitometric analysis of proteolytic activity detected by zymography. Densitometry was performed by laser scanning a representive section of each lane containing BALF samples from either normal control subjects (NL), heart transplant recipients (HT), or lung transplant recipients (LT). Three negative bands representing activity at relative molecular weights of 225, 92, and 72 kD were detected and the results are shown in panels a, b, and c, respectively.

Enzymes other than the metalloproteases, namely elastase, present in the lung are also reported to be capable of digesting gelatin (19). To confirm that the activity observed on the zymograms was due to one or more metalloprotease(s) and not elastase, representative BALF samples from all three groups were run with the addition of EDTA (a metalloprotease inhibitor) or PMSF (a serine protease inhibitor) to the substrate buffer. As seen in Figure 5, gelatinolytic activity was absent in the presence of EDTA. The addition of PMSF did not affect enzyme activity, suggesting that the proteolysis demonstrated was due to one or more metalloprotease(s).

View larger version (45K): [in this window] [in a new window]   Figure 5.   Representative lanes from 10% SDS-polyacrylamide gels impregnated with 0.1% gelatin and incubated in the presence of either: (a) EDTA (a metalloprotease inhibitor) or (b) PMSF (a serine protease inhibitor). Lanes pictured contain BALF samples from lung or heart-lung transplant recipients.

The collagenolytic activity of blood neutrophils and alveolar macrophages relative to the collagenolytic activity found in cell-free BALF was evaluated to assure that the amount of collagenase detected in BALF was not the result of enzyme "leak" during BALF processing. As seen in Table 3, gelatinolytic activity in peripheral neutrophils is approximately fifty fold that found in alveolar macrophages and in cell-free BALF supernatants. Gelatinolytic activity in alveolar macrophages from normal subjects is threefold higher than in macrophages from LT recipients, but in cell-free BALF, gelatinolytic activity is higher in LT recipients than in normal subjects.

Statistical Correlation between Alveolar Cell Numbers and Gelatinase Activity

Neutrophils and macrophages are potential sources of the gelatinases assayed for in this study. Neutrophils secrete 92 kD gelatinase (20), whereas macrophages produce and secrete both 72 and 92 kD gelatinase (21). An increased presence of either cell should result in the up-regulation of enzyme activity. Therefore, in an attempt to determine the cellular source of the enzymes detected by zymography, a correlation analysis was done between the detected activity of each band versus the neutrophil and macrophage counts and percentages. As seen in Figure 6, a positive correlation existed between total number of neutrophils per microliter of BALF for each of the three bands detected (correlation coefficient of 0.55, 0.68, and 0.75 for the 225, 92, and 72 kD bands, respectively). No correlation was seen with the macrophage counts (data not shown).

View larger version (11K): [in this window] [in a new window]   Figure 6.   Correlation between total neutrophil count and proteolytic activity at (a) 225 kD, (b) 92 kD, and (c) 72 kD bands. r Values represent Spearman's correlation coefficient for nonparametric data and were significant for all comparisons (p < 0.01).

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Chronic allograft dysfunction has become the major factor in limiting the benefits seen after lung transplantation. Evidence of chronic rejection is found in up to 50% of all LT recipients (22). Pathologically, chronic rejection is characterized by an accumulation of submucosal scar tissue along with an inflammatory cell infiltration of the peribronchiolar interstitium (17). This inflammatory process can include a variety of leukocytes, including lymphocytes, monocytes, plasma cells, and neutrophils. Immunohistochemical staining of allograft biopsy specimens with evidence of chronic rejection has shown a decrease in the amount of type IV collagen that is normally predominant in the basement membrane of the interstitium as well as an accumulation of type III collagen, instead of the normally predominant type I collagen (23). The presence of an inflammatory infiltrate accompanied by collagen destruction and abnormal deposition led us to hypothesize that metalloprotease activity is altered in LT recipients and contributes to the structural changes associated with long-term complications after lung transplantation.

A unique feature of this study is that samples were collected from clinically stable transplant recipients after the period of time when acute post-transplantation reaction would be expected. In this study, an increase in total number of leukocytes as well as proportion of neutrophils was present in BALF from clinically stable LT recipients compared with normal control subjects (Figures 1 and 2). BALF cell profiles from both the normal and HT control populations did not differ significantly from previously established normal values for healthy nonsmoking volunteers (24). There was a significant increase in proteolytic activity (Figure 4) accompanying the increase in inflammatory cells. Three enzyme species with relative molecular weights of 225, 92, and 72 kD accounted for the proteolytic activity in LT recipients.

The finding of an increased number of alveolar neutrophils in the LT recipients has previously been reported (25). These studies, however, included specimens collected immediately after transplant (< 4 wk) or during episodes of acute rejection and/or infection when the presence of increased number of BALF neutrophils would be expected. Maurer and coworkers (25) found an increase in neutrophils (10.9 ± 3.4%) in BALF from four heart-lung and 9 LT recipients collected at 3 and 6 mo after transplantation. Neutrophil numbers were higher in patients that went on to develop bronchiolitis obliterans. Information on the presence of rejection or infection at the time of BALF collection was not provided, so the increase in neutrophils could have been explained by the presence of active allograft disease. The finding of an increase in BALF neutrophils and its association with chronic rejection after heart-lung transplantation has also been reported by Haslam and associates (26). These investigators found a significant increase in neutrophil numbers in pediatric heart-lung transplant recipients up to 50 mo after transplantation. Clinical correlations demonstrated that neutrophil numbers were higher in patients with bacterial infections and patients who developed chronic rejection. Neutrophil counts also showed an inverse correlation with lung function. Another study on the presence of BALF neutrophils and their relationship to bronchiolitis obliterans was recently reported by DiGiovine and colleagues (27). They also detected a significant increase in neutrophil numbers along with an increase in levels of the neutrophil chemoattractant, interleukin-8. The interleukin-8 level was positively correlated with the increase in neutrophil numbers in LT recipients that developed bronchiolitis obliterans.

The observed increase in the total number of leukocytes and number of neutrophils (Figure 1) in clinically stable LT recipients is indicative of chronic inflammation. Bronchoalveolar neutrophils are known to be present in various interstitial lung diseases that, like chronic lung allograft rejection, are associated with both inflammation and with changes in the lung parenchymal collagen composition. In idiopathic pulmonary fibrosis, an increased number of alveolar neutrophils has been demonstrated during episodes of active alveolitis (28). The early stage of the adult respiratory distress syndrome is also characterized by the presence of a marked increase in alveolar neutrophils (29). Both idiopathic pulmonary fibrosis and the adult respiratory distress syndrome have been associated with an increase in metalloprotease activity, abnormal collagen deposition, and fibrosis in their late stages (12, 30).

The gelatinases are released in latent forms and are known to undergo extracellular processing to achieve their active form. Under nonreducing conditions, Hibbs and associates (20) found gelatinase activity from human neutrophils at three different molecular weights, 225, 130, and 92 kD. Immunologically, all three protein bands were recognized by polyclonal antibodies prepared to both 92 and 130 kD proteins (20). This suggests that the 225 kD band detected by zymography in our study may represent an aggregate form of the 92 kD gelatinase.

Neutrophils are the most likely source of the proteolytic activity detected in BALF of the transplant recipients. There are several pieces of evidence to support this assumption. First, neutrophils secrete 92 kD gelatinase (20). Second, by zymography, 92 kD gelatinase (and its 225 kD form) accounted for 78% of the total proteolytic activity detected in LT recipients. Third, the number of neutrophils present in the BALF samples from LT recipients showed a positive correlation to the gelatinolytic activity. The increase in neutrophil numbers, however, cannot explain the increase in activity of the 72 kD gelatinase.

The source of the increased 72 kD gelatinase activity in this study is not certain but two possibilities exist: (1) fibroblasts and other connective tissue cells are the most abundant source of this enzyme (5) and (2) macrophages, including alveolar macrophages, produce and secrete 72 kD gelatinse (21). Welgus and colleagues found a marked up-regulation of 72 kD gelatinolytic activity from macrophages in cell culture when stimulated by lipopolysaccharide (21). This level of 72 kD gelatinase activity was, however, much less than that attributable to similarly stimulated fibroblast cell cultures. Furthermore, because macrophage numbers in LT and HT recipients were equivalent, it is less likely that alveolar macrophages account for the increase in 72 kD gelatinolytic activity. Although there is no histologic evidence of fibrosis at the time of transbronchial biopsy in these clinically stable LT recipients, it is known to be a finding in chronic rejection. The presence of a chronic inflammatory state in lung allografts could theoretically lead to activation of either fibroblasts or macrophages, thereby increasing production and secretion of 72 kD gelatinase without necessarily increasing the total number of cells.

Regardless of the cellular source of the gelatinases, the observed increase in gelatinolytic activity in the BALF of clinically stable LT recipients may result from increased production and secretion or decreased presence of inhibitors. The tissue inhibitors of metalloproteases (TIMPs) are primary inhibitors of MMP-2 and MMP-9 (72 and 92 kD gelatinase, respectively). They are commonly secreted by macrophages and neutrophils complexed with the gelatinases (31, 32). The cosecretion of TIMPs with the MMPs by inflammatory cells therefore makes it less likely that a decrease in inhibition of MMPs is a cause for the increased gelatinolytic activity in the BALF of LT recipients. Rather, increased production or increased secretion of the gelatinases is more likely. As shown in Table 3, LT alveolar macrophages contain less gelatinolytic activity than NL alveolar macrophages despite the finding of increased gelatinolytic activity in the cell-free BALF of LT recipients compared with NL control subjects. This supports the contention that the increased gelatinolytic activity observed in LT recipients is the result of increased secretion rather than increased production.

The effect of immunosuppressive agents on activity of metalloproteases is variable. Glucocorticoid agents have been shown to inhibit the production of collagenase by cells cultured from synovectomy specimens of patients with rheumatoid arthritis (33). Others have reported that MMP-2 and MMP-9 gelatinase activity in dermal blisters was not affected in patients treated with systemic prednisone (34). Cyclosporine has been shown to inhibit the production of interstitial collagenase in a monocytic cell line after stimulation by activated T cells (35). To our knowledge, there is no data available detailing the influence of these immunosuppressive agents on proteolytic activity in BALF. By zymography, no difference in activity between healthy volunteer and HT control samples was found (Figure 4). Yet, a significant increase in the proteolytic activity of samples from clinically stable LT recipients was observed. This observation suggests that the immunosuppression required for the maintenance of graft function did not significantly affect the protease activity in BALF and cannot explain the up-regulation of gelatinolytic activity seen in the LT group.

In this study, we have confirmed that an inflammatory state, as manifested by increased numbers of leukocytes and neutrophils, exists within the lungs of clinically stable LT recipients. Further, we have demonstrated an up-regulation of gelatinolytic activity, not accounted for on the basis of immunosuppression. The presence of this increased proteolytic activity indirectly suggests the presence of activated inflammatory cells. Others have shown that immunohistochemical evidence of intercellular collagen matrix remodeling occurs in chronic lung allograft rejection and that an increase in alveolar neutrophils is related to chronic rejection. Collectively, these results indicate that up-regulated proteolytic activity in clinically stable LT recipients may have a role in chronic rejection after transplantation.