Introduction

Abstract Introduction Materials and Methods Results Discussion References

Although multiple organ systems are involved in the pathology of cystic fibrosis (CF), deterioration in pulmonary function is responsible for much of the morbidity and mortality from this disease (2). CF lung disease is characterized by airway obstruction and bronchiectasis, with copious mucopurulent secretions, chronic infection, and chronic inflammation. Pulmonary inflammation is pervasive. Infiltrated neutrophils and other indicators of inflammation have been documented in CF patients with good pulmonary function and even in infants and very young children with CF (3).

In the chronically inflamed lungs of CF patients, the activity of the neutrophil's abundant serine proteases, elastase (neutrophil elastase [NE]), cathepsin-G, and proteinase-3 is turned upon lung tissue. NE in particular is well characterized for its contribution to lung destruction. By virtue of its broad range of substrate specificity, NE degrades elastin and other structural proteins (8) and impairs the antibacterial functions of macrophages and neutrophils. NE compromises phagocytosis of antibody-coated bacteria, the dominant macrophage pathway, by proteolysis of antibody molecules (9, 10), and phagocytosis of complement-opsonized bacteria, the dominant neutrophil pathway, by proteolysis of complement components (11). NE also exacerbates other manifestations of airway dysfunction by increasing mucin release from tracheal epithelial cells (12- 14) and stimulating production of neutrophil chemoattractants including interleukin (IL)-8 (15). As a result, lung structure and function decline inexorably over the lifetime of the patient.

Although natural inhibitors of neutrophil proteases are normally found in lung fluids and their levels even increase in the CF lung, protection is inadequate against the overwhelming burden of NE and other proteases present as a result of the neutrophil-dominated inflammation (3, 16). Naturally occurring inhibitors of NE include ₁-antitrypsin (₁-AT), secretory leukocyte protease inhibitor (SLPI), elafin, and monocyte/neutrophil elastase inhibitor (M/NEI). ₁-AT, the 50-kD prototype of serpin proteins, is found at high concentrations in plasma, from which it enters tissue sites and lung epithelial lining fluid (17). SLPI, a 12-kD polypeptide, is produced by cells of mucosal surfaces (18, 19), including bronchial epithelial cells (20), and was recently described in neutrophils (21). Elafin, a 7-kD polypeptide, is found in skin (22) and in bronchial secretions (23). M/NEI, a 42-kD serpin molecule, is found in the cells (neutrophils, monocytes, and macrophages) that accumulate at inflammatory sites (24), where it is thought to function in regulating the activity of NE, cathepsin-G, and proteinase-3 (27, 28).

One approach to intervention in this lung-destruction process in CF is to supply inhibitors to suppress protease activity. One could, for example, supply a natural inhibitor of NE in sufficient quantity to create a balance between extracellular proteases and protease inhibitors (see DISCUSSION). A rat model of lung injury induced by instilled protease is useful for evaluating the ability of individual antiproteases to restore this balance. Previous studies (29, 30) have shown that pure human NE or CF airway secretions (CF sol) instilled into the lungs of rodents initiate both injury and inflammation, thereby approximating the inflammatory environment of the human lung in CF. The hemorrhagic component of injury in the rat model, because it can be completely obliterated by pretreating NE or CF sol with irreversible synthetic inhibitors of NE, provides a clear readout of NE-induced damage (29, 30).

In the present study we examined the ability of recombinant M/NEI (rM/NEI) to protect lungs against injury by NE and other proteases present in CF airway secretions. Gram quantities of active rM/NEI were prepared. The capacity of pure rM/NEI to inhibit NE in the complex biologic milieu of CF inflammatory exudate and to protect the lungs in a rat injury model was examined. In addition, we used fluorochrome-labeled rM/NEI to examine the distribution and clearance of rM/NEI instilled into rat lungs.

	Materials and Methods

Abstract Introduction Materials and Methods Results Discussion References

rM/NEI Expression and Purification

rM/NEI was produced in Sf 9 insect cells infected by recombinant baculovirus AcNPV-M/NEI (27). The cells were grown in 15-liter flasks, and rM/NEI was purified from cell lysates by anion exchange and molecular-sieve chromatography (J. Cooley and colleagues, unpublished data). The protein concentration of pure rM/NEI was determined by amino acid composition analysis (Molecular Biology Core Facility, Dana Farber Cancer Institute, Boston, MA). Endotoxin levels, measured with Limulus amebocyte lysate reagents (BioWhittaker, Inc., Walkersville, MD), were < 0.25 units (30 pg)/ml. rM/NEI for lung injury studies was stored at ~ 1 mg/ml in 20 mM Tris-HCl buffer, pH 7.4; 500 mM NaCl; 2 mM mercaptoethanol; and 2 mM ethylenediamine tetraacetic acid (EDTA) in aliquots at 80°C. rM/NEI for enzyme assay was dialyzed against phosphate-buffered saline (PBS) prior to storage. The preparations were 72-92% active on the basis of stoichiometry of inhibition of pure NE (described subsequently); the concentrations reported in figure legends represent active rM/NEI.

Anti-M/NEI Antiserum

A New Zealand female rabbit was immunized by subcutaneous injection of 60 µg rM/NEI that had been harvested from 6-d media of infected insect cells (27), partly purified by thiopropyl-sepharose chromatography (25), denatured with 0.1% sodium dodecyl sulfate (SDS) at 100°C for 2 min, and emulsified with complete Freund's adjuvant. The rabbit was injected subcutaneously, 2 and 4 wk later, with 20 µg of similarly denatured M/NEI emulsified in incomplete Freund's adjuvant. Antiserum was collected 1-4 wk thereafter.

Fluorescent Conjugation of rM/NEI

rM/NEI was conjugated with the fluorophore Texas Red-X (TR), using the Texas Red kit (Molecular Probes, Eugene, OR). rM/NEI (1-1.4 mg/ml) was incubated with Texas Red for 1 h at pH 8.5 (31). Unreacted dye was inactivated with hydroxylamine, and the rM/NEI-Texas Red conjugate (rM/NEI-TR) was purified with microfuge spin columns. Conjugation was verified by thin-layer chromatography, and the integrity of rM/NEI-TR was confirmed by Coomassie blue-stained SDS-polyacrylamide gel electrophoresis (PAGE) bands. The activity of rM/NEI was, however, substantially decreased by conjugation of the fluorochrome moiety; the NE-inhibitory activity of rM/NEI-TR was 10-20% that of unlabeled rM/NEI (data not shown).

NE

Human NE purified from sputum (32) was obtained from Elastin Products Co. (Owensville, MO) and stored as a 2-mg/ml stock solution in pyrogen-free saline at 70°C.

CF Sol Preparation

Sputum was collected as part of routine therapy from adult patients with advanced CF who were hospitalized for acute pulmonary exacerbation of the disease (Children's Hospital, Boston, MA). All patients were receiving antibiotic therapy. Some were also receiving recombinant deoxyribonuclease (DNase) (Genentech, South San Francisco, CA). Sputum was stored on ice for up to 6 h, until fractionated into the aqueous sol and gel fractions by ultracentrifugation at 50,000 × g for 90 min (30). Seventeen sol preparations (CF sol) were combined to obtain 106 ml of pooled CF sol with 11.5 µM NE activity for in vitro studies; 23 sol preparations were combined to obtain 58 ml of pooled CF sol with 4.0 µM NE activity for lung injury studies.

NE Activity

NE amidolytic activity was measured kinetically through changes in OD at 410 nm and ~ 22°C with 0.8 mM N-methoxysuccinyl-Ala-Ala-Pro-Val-p-nitroanilide (Sigma Chemical Co., St. Louis, MO) in 20 mM Tris-HCl, pH 7.4; 500 mM (or 1,000 mM) NaCl; and 0.01% polyethylene glycol (PEG) at ~ 22°C. The apparent NE content of CF sol was determined by measuring the activity of the sol relative to a standard curve of pure NE.

NE-Inhibitory Activity

To measure its NE-inhibitory activity, rM/NEI was incubated with NE in PBS with 0.05% Tween-20 (Pierce, Rockford, IL) for 5 min at 37°C, and was then diluted for measurement of amidolytic activity as described previously.

rM/NEI Complex Formation Assay

rM/NEI was incubated with protease (NE or CF sol) for 3 min at 37°C. To stop the reaction, diisopropylfluorophosphate (DIFP) was added to a concentration of 2 mM and incubation continued for 2 min. The samples were solubilized and examined with SDS-PAGE or immunoblotting.

SDS-PAGE and Immunoblotting

Protein samples were analyzed through SDS-PAGE (33) under reducing conditions on 10% polyacrylamide gels (80 × 70 × 1 or 1.5 mm) (Novex, San Diego, CA) containing 420 mM Tris-HCl buffer, pH 8.6, with a pH 8.3 running buffer (33). The polypeptides were either stained with Coomassie blue or transferred at 80 mA for 16 h to nitrocellulose, which was blocked at ~ 22°C with 20% milk solids in PBS and 0.05% Tween-20 (PBS-Tween). The blot was incubated with rabbit anti-M/NEI antiserum (1:1,000 dilution) in PBS-Tween-0.1% milk for 2 h, washed, and incubated for 1 h in PBS-Tween-0.1% milk containing ¹²⁵I-labeled goat antibodies (0.3 µg/ml) to rabbit immunoglobulin G. The M/NEI bands were detected with the Storm 860 Phosphor Imager (Molecular Dynamics, Sunnydale, CA).

Treatment of Rats for Study of Lung Injury

Harlan Sprague-Dawley female rats (~ 200-250 g) (Charles River Laboratories, Wilmington, MA) were treated with pure NE or CF sol diluted in pyrogen-free saline (or with pyrogen-free saline as control) by intratracheal instillation under halothane anesthesia (34), using a miniature atomizing cannula (Penn Century, Philadelphia, PA). This device, which is fabricated of stainless steel, has an 0.0047-in inner diameter and was used to instill liquids (NE, CF sol, rM/NEI, and combinations thereof) as a mist directly into the trachea. The volume instilled was 150 µl per 100 g body weight (gbw), except for CF sol, which was introduced as two sequential 150 µl/100 gbw instillations separated by 15 min.

Instillation of NE or CF sol followed two treatment protocols: (1) instillation of NE or CF sol alone or after incubation in vitro with rM/NEI (17.5 µM) at 37°C for 15 min, and (2) instillation with rM/NEI from 1 to 24 h prior to instillation of NE or CF sol.

Rats were killed 4 h after instillation of NE or CF sol by intraperitoneal injection of sodium pentobarbital, and were exsanguinated. To prepare bronchoalveolar lavage fluid (BALF), the lungs were lavaged with 12 aliquots of 3 ml each of isotonic saline (34). Cells were pelleted at 300 × g from pooled bronchoalveolar lavage (BAL) washes 1 and 2. The supernatant was clarified at 14,600 × g for 30 min and analyzed for its albumin content by measuring binding of bromcresol green (35). A cell pellet prepared from pooled washes 3-12 was combined with the cell pellet of washes 1 and 2 and analyzed for hemoglobin content with Drabkin's reagent (36) after lysing red blood cells with 0.015% Brij.

Treatment of Rats for Histologic Examination

Lung tissue from female rats instilled with NE with and without rM/NEI pretreatment, as described previously, was also subjected to histologic examination. The animals were killed from 5 min to 24 h after instillation of NE, and freshly excised tissue was fixed in 2% paraformaldehyde, dehydrated with absolute ethanol, and processed for paraffin embedding. Deparaffinized sections were stained with hematoxylin and eosin and examined by light microscopy. Color micrographs of stained lung sections were recorded with a Sony DCX-755 camera (Sony Electronics, Fort Myers, FL) mounted on a Nikon Microphot-Fx upright microscope (Nikon, Inc., Melville, NY).

Instillation of rM/NEI-TR for Study of Distribution and Clearance

Male Sprague-Dawley rats were lightly anesthetized by intraperitoneal injection of sodium methohexital (Brevital) (0.5-0.6 ml/animal). The animals were instilled through the trachea, as described previously, with rM/NEI-TR diluted in pyrogen-free saline (20-60 µg in 150 µl per 100 gbw). At 5 min, 1 h, 4 h, and 24 h after instillation, rats were killed by intraperitoneal injection of a pentobarbital overdose (5 mg/100 gbw) and lung specimens were prepared for microscopy.

Confocal Microscopy and Imaging of rM/NEI-TR

Freshly excised lung specimens from rats instilled with rM/ NEI-TR were embedded in OCT, a water-soluble embedding medium (Miles, Inc., Elkhart, IN), cryofixed in liquid nitrogen, cryosectioned (IEC Minatome; International Equipment Co., Needham, MA), and vapor fixed with 2% paraformaldehyde. To assess distribution of rM/NEI-TR, airways of cryofixed lungs were studied with a Sarastro 2000 confocal microscope (Molecular Dynamics, Sunnyvale, CA). Micrographs were recorded with two imaging modes: transmitted light to illustrate airways and lung architecture, and fluorescent confocal imaging to illustrate rM/NEI-TR. To correct for background autofluorescence, images of collagen and elastic fibers were recorded in a separate channel (540 ± 30 nM) and subtracted from the images used to detect rM/NEI-TR (> 595 nM).

Statistical Analysis

Measurements are presented as means ± SEM, with n  4 for all groups. The effects of rM/NEI mixed with NE instillate on BALF hemoglobin and albumin were analyzed by two-way analysis of variance (ANOVA), using Tukey's test for multiple comparisons. For all other treatment groups, data were analyzed by one-way ANOVA, using Dunnett's method for multiple comparisons. Differences were considered significant when P < 0.05.

	Results

Abstract Introduction Materials and Methods Results Discussion References

rM/NEI Inhibits NE Activity In Vitro

Gram quantities of rM/NEI were prepared with a baculovirus expression system (see MATERIALS AND METHODS), and SDS-PAGE showed that the preparations were > 95% homogeneous (Figure 1, inset). The ability of the rM/NEI to completely inhibit the amidolytic activity of 5.8 µM pure NE is shown in Figure 1A. rM/NEI was also found to inhibit the equal levels of NE amidolytic activity present in CF sol (Figure 1B). The rM/NEI concentrations of 3.1 and 3.4 µM required for 50% inhibition for pure NE and NE activity in CF sol, respectively, were both close to the theoretically predicted stoichiometric value. For complete inhibition, however, higher rM/NEI levels were required for CF sol than for the pure protease. Whereas 90% inhibition of the activity of the pure NE required 5.9 µM rM/NEI, 90% inhibition of equivalent activity in CF sol required 9.2 µM rM/ NEI (compare Figures 1A and 1B) (see DISCUSSION). The finding of complete inhibition of NE activity in CF sol predicts that rM/NEI will be inhibitory of NE within CF airways.

View larger version (16K): [in this window] [in a new window]   Figure 1.   Inhibition by rM/NEI of the NE amidolytic activity of pure human NE and CF sol. Pure NE (A) or CF sol (B) was combined with an equal volume of rM/NEI, incubated at 37°C for 3 min, and diluted for assay of Ala-Ala-Pro-Val-p-nitroanilide cleaving activity. The concentration of pure NE was chosen to replicate the NE-amidolytic activity of the CF sol (5.8 µM after 1:1 dilution). The abscissa shows the concentration of rM/NEI during the reaction with protease. The inset shows the preparation of rM/NEI (1.5 µg) examined with SDS-PAGE and Coomassie blue staining.

Since M/NEI, a serpin superfamily protein, functions by forming a 1:1 stable complex with protease (25), we examined whether rM/NEI forms this complex with protease within CF sol. As previously demonstrated, the product of the reaction of rM/NEI (42 kD) and NE (26 kD) was detected on Coomassie blue-stained SDS-PAGE as a 66-kD SDS-stable complex (Figure 2A). The multiplicity of proteins in CF sol made Coomassie blue staining uninformative for analysis of the sol reaction, and immunoblotting with a rabbit anti-M/NEI antiserum was used instead to detect M/NEI and its complex. Staining of the unreacted CF sol revealed the presence of a 66-kD apparent complex of M/NEI-protease, as well as larger amounts of 42-kD M/NEI, which is probably inactive (see DISCUSSION). Addition of increasing amounts of rM/NEI to the sol and incubation for 3 min generated increasing amounts of the 66-kD apparent rM/NEI-protease complex (Figure 2B), suggesting that rM/NEI has the capacity to complex with protease in the microenvironment of CF airways.

View larger version (60K): [in this window] [in a new window]   Figure 2.   Complex formation of rM/NEI with pure NE and with protease in CF sol. Pure NE (A) or CF sol (B) was combined with varying concentrations of rM/NEI and incubated at 37°C for 3 min. Protease action was stopped with DIFP and the reaction analyzed through SDS-PAGE under reducing conditions. (A) A Coomassie blue-stained gel with rM/NEI at 42 kD; NE, which stains poorly, indicated by an arrow on the left at 26 kD; and the 66-kD complex formed on incubation of 5.8 µM NE with the indicated concentrations of M/NEI. (B) An immunoblot in which rabbit anti-M/NEI antibodies stain M/NEI (42 kD) as well as the complex formed when rM/NEI reacts with protease(s) in CF sol. CF sol was 5.8 µM in NE activity during the assay.

rM/NEI Blocks Lung Injury Induced by NE or CF Sol

We evaluated the ability of rM/NEI to mitigate NE-induced lung injury. In a rat model of lung injury, instilled human NE induces a dose-dependent hemorrhage and increase in epithelial permeability. In vitro incubation of NE with rM/ NEI before instillation effectively abolished the ability of NE to induce both hemorrhage (P < 0.001) and increased epithelial permeability in rat lung (P < 0.001). rM/NEI blocked both forms of injury at inhibitor-to-protease molar ratios as low as 1:1 (17.5 µM NE + 17.5 µM rM/NEI) (Figure 3). At this equimolar ratio, BALF hemoglobin was reduced by 93 ± 4.3%, from 10.9 ± 2.2 mg/animal to 0.77 ± 0.47 mg/animal. Similarly, the NE-induced increase in BALF albumin was reduced by 95 ± 5.3%.

View larger version (12K): [in this window] [in a new window]   Figure 3.   Inhibition of human NE-induced lung injury by rM/ NEI. BAL hemoglobin (A) and BAL albumin (B) were measured 4 h after instillation of 150 µl/100 gbw of NE at the indicated concentrations without (open circle) and with (solid circle) in vitro incubation with 17.5 µM rM/NEI. Values are means ± SEM with n  4 for all groups. Two-way ANOVA indicated a significant effect of rM/NEI treatment on NE-induced hemorrhage (BAL hemoglobin) (P < 0.001) and increased epithelial permeability (BAL albumin) (P < 0.001). Multiple-comparison procedures indicated significant protection by rM/NEI at NE doses of 10 µM and 17.5 µM.

Instillation of CF sol into rat lungs also induces both hemorrhage and an increase in epithelial permeability. In vitro incubation of CF sol with rM/NEI prevented hemorrhage in rat lungs (P < 0.01) (Figure 4). Preincubation with rM/NEI reduced hemorrhage by 98 ± 0.33%, from 6.56 ± 0.94 mg/animal with CF sol alone to 0.095 ± 0.022 mg/animal with CF sol plus rM/NEI. Indeed, hemoglobin levels with CF sol plus rM/NEI were statistically indistinguishable from those of sham-treated control animals (0.067 ± 0.007 mg/animal).

View larger version (24K): [in this window] [in a new window]   Figure 4.   Inhibition of CF sol-induced lung injury by rM/NEI. BAL hemoglobin (A) and BAL albumin (B) were measured 4 h after instillation of CF sol (two instillations of 150 µl/100 gbw) or CF sol incubated in vitro with 17.5 µM rM/NEI. Control animals were instilled with pyrogen-free saline. Values are means ± SEM with n  4 for all groups. One-way ANOVA indicated a significant inhibitory effect of rM/NEI on CF sol-induced hemorrhage (BAL hemoglobin) (P < 0.01). Multiple-comparison procedures indicated significant differences between the CF sol and CF sol + rM/NEI groups and between the CF sol and saline groups (P < 0.01). The effect of rM/NEI on epithelial permeability (BAL albumin) was statistically indistinguishable from that of CF sol or saline alone.

The CF sol-induced increase in epithelial permeability, however, was not significantly reduced by preincubation with rM/NEI. This latter observation suggests that components in the CF sol other than proteases inhibitable by rM/ NEI can reduce epithelial integrity.

Distribution of Fluorochrome-Labeled rM/NEI in Airways

To examine the distribution pattern of rM/NEI instillates and derive an approximation of the bioavailability and survival time of rM/NEI in the rat airway model, retention of the molecule was examined by confocal microscopy of lung tissue from animals that were instilled with rM/NEI conjugated with the fluorochrome Texas Red (rM/NEI- TR) (Figure 5). At 5 min after instillation, rM/NEI-TR was observed on airways and alveolar surfaces, and after 1 h was detected in alveolar macrophages. Importantly, rM/NEI-TR was detected on the epithelial surfaces of large airways (> 300 µm) at all time points studied (i.e., 5 min, 1 h, 4 h, and 24 h). Although fluorochrome-conjugated rM/ NEI, which retains only minimal NE-inhibitory activity (see MATERIALS AND METHODS), is an imperfect surrogate for unlabeled active rM/NEI, the findings described here suggest that instilled active M/NEI will not be rapidly cleared from the rat airway.

View larger version (202K): [in this window] [in a new window]   Figure 5.   rM/NEI distribution in rat airways. Rats were instilled with rM/NEI-TR fluorescent conjugate (150 µl/100 gbw). Brightfield images of lung tissue (grayscale) were combined with fluorescent optical sections of rM/NEI-TR (red) to reveal distribution of rM/NEI-TR among airways (La) at 5 min (A), 1 h (B), 4 h (C), and 24 h (D) after instillation. Arrowheads indicate rM/NEI-TR on airway surface; arrows indicate rM/NEI-TR on alveolar surface. Bars = 20 µm.

Pretreatment of Rats with rM/NEI Provides Protection against Lung Injury by NE

Using the rat lung injury model, we asked whether pretreatment of rats with active rM/NEI would protect them from subsequent injury by NE. Preinstillation of rats with rM/ NEI conferred protection for up to 24 h from injury caused by subsequent instillation with NE (Figure 6). Pretreatment with a single instillation of rM/NEI at 1, 4, or 24 h before instillation of NE reduced BALF hemoglobin by 98 ± 0.34%, 95 ± 0.32%, and 72 ± 12%, respectively. Epithelial integrity was also protected, as the NE-induced increase in BALF albumin over saline control levels was blocked by 69 ± 12%, 67 ± 12%, and 74 ± 9.4%, respectively.

View larger version (23K): [in this window] [in a new window]   Figure 6.   Effect of pretreatment of rats with rM/NEI on NE- induced lung injury. Rats were instilled with rM/NEI (35 µM; 150 µl/100 gbw). At 1, 4, and 24 h later, rats were challenged with NE (10.5 µM). After an additional 4 h, rats were killed and their BAL was analyzed for hemoglobin (A) and albumin (B). Values are means ± SEM with n  4 for all groups. One-way ANOVA indicated a significant protective effect of rM/NEI against NE- induced hemorrhage (BAL hemoglobin) (P < 0.001) and epithelial permeability (BAL albumin) (P < 0.01). Multiple-comparison procedures demonstrated significant protection by rM/NEI for all rM/NEI treatment groups (P < 0.05).

In conjunction with this series of experiments, we examined whether rM/NEI itself induced injury or inflammation in rat lungs. At 24 h after instillation of 17.5 µM rM/NEI, neutrophil numbers in BALF had risen only slightly, from 0.45 ± 0.55 × 10⁶/animal in saline control animals to 3.28 ± 1.9 × 10⁶/animal in rM/NEI-treated animals. Hemoglobin levels (0.081 mg/animal) were not significantly increased over control levels (0.066 mg/animal). Albumin levels rose modestly, from 130 µg/ml in controls to 366 µg/ml in treated animals. These findings show that rM/NEI itself does not elicit an appreciable inflammatory response when introduced into rat lungs.

Light Microscopy Confirms the Protective Effect of Preinstilled rM/NEI against Lung Injury by NE

The ability of rM/NEI pretreatment to protect rat lungs against NE-induced injury was further evaluated by histologic examination of lung sections. Light microscopy of airways and terminal bronchioles of animals instilled with NE alone or given rM/NEI pretreatment followed by NE instillation verified that rM/NEI inhibited NE-induced hemorrhage. Airways, terminal bronchioles, and particularly alveolar spaces of lungs of animals treated with NE alone (Figures 7A and 7C) were observed to contain numerous erythrocytes, whereas lungs of animals that received rM/NEI before instillation of NE showed fewer detectable erythrocytes in airways, terminal bronchioles, and alveolar spaces (Figures 7B and 7D).

View larger version (129K): [in this window] [in a new window]   Figure 7.   Effect of pretreatment of rats with rM/NEI on morphology of airways exposed to NE. Rats were untreated (A and C) or instilled with rM/NEI (B and D) (35 µM; 150 µl/100 gbw). At 1 h later, the rats were challenged with NE (10.5 µM). After an additional 4 h, the rats were killed and morphologic indices of lung damage were assessed in light micrographs. Bars = 100 µm in A and B; 10 µm in C and D. Symbols are La, airway lumen; Lv, blood vessel lumen; E, erythrocyte; and star, alveolar space.

The primary damage induced by NE in this model was indicated by the presence of erythrocytes. Other indices of lung damage, such as loss of epithelium or interstitial swelling, were not detectable at the 4 h time point studied.

Pretreatment of Rats with rM/NEI Provides Protection against Lung Injury by CF Sol

Treatment of rats by instillation of rM/NEI also protected against injury by subsequently instilled CF sol. Protection was evident even 24 h after instillation of rM/NEI (Figure 8). BALF hemoglobin was reduced by 96 ± 1.6%, 86 ± 11%, and 86 ± 5.7%, respectively, in rats pretreated with rM/NEI at 1, 4, or 24 h before exposure to CF sol. However, rM/NEI pretreatment did not protect rats from subsequent injury to epithelial integrity by instilled CF sol, again suggesting that components in CF airway secretions other than proteases inhibitable by rM/NEI reduce epithelial integrity.

View larger version (24K): [in this window] [in a new window]   Figure 8.   Effect of pretreatment of rats with rM/NEI on CF sol-induced lung injury. Rats were instilled with rM/NEI (35 µM; 150 µl/100 gbw). At 1, 4, and 24 h later, rats were challenged with CF sol (two instillations of 150 µl/100 gbw). After an additional 4 h, the rats were killed and their BAL was analyzed for hemoglobin (A) and albumin (B). Values are means ± SEM with n  4 for all groups. One-way ANOVA indicated a significant ability of rM/ NEI to protect against CF sol-induced hemorrhage (BAL hemoglobin) (P < 0.001). Multiple-comparison procedures indicated a significant ability of rM/NEI to protect against hemorrhage in all rM/NEI treatment groups (P < 0.05). The protective effect of rM/NEI against CF sol-induced increased epithelial permeability (BAL albumin) was not statistically significant at any rM/NEI pretreatment interval.

	Discussion

Abstract Introduction Materials and Methods Results Discussion References

The study described here shows that rM/NEI is a potential candidate for anti-NE therapy in CF. Active rM/NEI produced in a baculovirus expression system was tested for efficacy in vitro and in vivo in a rat model of lung injury. The highly purified rM/NEI was shown to effectively inhibit pure NE as well as NE in the inflammatory milieu of CF sputum sol in vitro. rM/NEI formed a complex with NE in CF sol that was detectable by immunoblotting. Importantly, rM/NEI was capable of protecting rats from lung injury induced by instilled NE or CF sol. This could be accomplished by treating the protease with rM/NEI before instillation or by treating the rats with rM/NEI before instilling protease or CF sol. Preliminary biodistribution analysis, using fluorochrome-labeled M/NEI (rM/NEI-TR), indicated that rM/NEI-TR was present in the lungs and airway surfaces 24 h after instillation. However, this finding by itself had limited predictive value, because rM/ NEI-TR retains only minimal NE-inhibitory activity, and might therefore be cleared more slowly than native M/NEI. The more relevant finding was that unlabeled rM/NEI instilled 24 h before challenge with either NE or CF sol provided significant protection against protease-induced hemorrhage, showing that active rM/NEI has a long bioavailability and slow rate of clearance from the rat lung.

Several studies have demonstrated the ability of human NE to destroy lung structure and function, particularly in CF. NE, a serine protease, is stored as active enzyme in neutrophil azurophil granules (37), together with the closely related molecules cathepsin-G (38, 39) and proteinase-3 (40), which are also potent proteases. NE, cathepsin-G, and proteinase-3 are released into phagocytic granules and extracellularly in a targeted manner when neutrophils encounter microbes or other stimuli (43). The proteases are also released when neutrophils accumulate at high density and lyse, as occurs in the lungs of even very young CF patients (3). Although substantial evidence implicates NE in the lung deterioration of CF, cathepsin-G and/or proteinase-3 may also participate in this process. Clearly, controlling NE and possibly these other proteases is a rational approach to slowing the loss of lung function in CF.

rM/NEI is potentially useful for controlling undesirable protease activity. Native M/NEI is naturally present at sites of inflammation and probably contributes to the maintenance of protease/antiprotease balance at these sites. The current studies show that rM/NEI is effective both in vitro and in vivo against NE. As an aerosol, rM/NEI could be effectively delivered to the lungs, the desired site of action, reducing the likelihood of systemic effects. For example, protective protease-dependent functions of circulating blood neutrophils would not be compromised. Indeed, since rM/NEI replicates a native protein and in aerosol form would be delivered to a native site, it is unlikely to produce adverse side effects. The potential for local delivery and low likelihood of side effects is particularly important when one considers that any antielastase agent for CF would ideally be delivered starting soon after birth, when inflammation and increased NE levels first appear, and that lifelong treatment would be continued both prophylactically and therapeutically.

The current study demonstrates that rM/NEI is an effective inhibitor of NE in inflammatory pulmonary exudates. This environment contains high levels of multiple proteases capable of degrading a wide spectrum of substrates, as well as neutrophil myeloperoxidase and reactive oxygen species capable of inactivating proteins by oxidative mechanisms. Despite these potential barriers to its function, rM/NEI inhibited CF sol NE in vitro and for at least 24 h in vivo. The in vitro finding that rM/NEI can inhibit NE activity in CF sol predicts that it will be similarly inhibitory in CF airways. The levels of rM/NEI that gave 50% inhibition of NE activity in CF sol were close to those predicted by 1:1 stoichiometry. Also noteworthy is the finding that virtually all sol NE activity could be inhibited. Levels of rM/NEI greater than stoichiometric were, however, required for complete inhibition of sol NE activity. Possible explanations for this last observation include the presence of components in the sol that hinder the interaction of NE and rM/NEI or the presence in the sol of a second rM/NEI-inhibitable protease with NE-like amidolytic activity (i.e., proteinase-3). The practical implication is that virtually complete inhibition should be achievable in CF lung fluids, but may require rM/NEI levels ~ 40-50% in excess of measured NE activity levels.

Western blotting also identified naturally occurring M/NEI as a free component of CF sol and as part of a 66-kD apparent complex with protease (Figure 2B); these molecules, which probably represent spent (inactivated) M/NEI, will be examined in future studies. Quantitation of the Western blots by phosphor-imaging suggested the presence of ~ 1-3 µM free M/NEI in CF sol; this value is, however, a preliminary approximation because the blotting antiserum had unequal staining efficiency for inactive M/NEI and active M/NEI used as standard, and because the two species did not separate on SDS-PAGE. Since the amount of 66-kD complex did not increase when NE-containing sol was incubated without added rM/NEI, it is likely that endogenous sol M/NEI is inactive as a result of the overwhelming excess of protease and/or the action of oxidants. The majority of ₁-AT and SLPI molecules recovered from CF lung fluids have also been found to be inactivated (3, 5).

In vivo studies of rM/NEI function were done with a rat model of lung injury and inflammation used extensively by our laboratory to evaluate pulmonary damage. Evaluating rM/NEI function in vivo permits study of the effects of rM/ NEI on injury occurring within the setting of the living lung, where three-dimensional biologic structures support influx and efflux of plasma components. No in vitro system can duplicate the number of components and interactions occurring in the living lung. The rat model combines advantages of two systems: the integrity and complexity of intact lungs and the ability to study the human substances NE and CF sol, which are known to be central to the development of human lung inflammatory damage.

With this model, we were able to study injury produced by human inflammatory exudate or a purified component of exudate (NE), and to measure the protection offered by a human protein, rM/NEI. rM/NEI treatment of NE or CF sol completely blocked hemorrhage induced by these materials. Epithelial integrity was protected when injury was induced by pure NE. In contrast, when CF sol was the test substance, rM/NEI did not protect against the injury to epithelial integrity, despite efficient and virtually complete prevention of the hemorrhagic component of CF sol- induced damage. When a synthetic -lactam inhibitor of NE was studied earlier in the same model, it also failed to prevent a CF sol-induced increase in epithelial permeability (30). This and the present findings suggest that the increased permeability caused by CF sol is due to components other than proteases; agents known to be present in the sol include cytokines, bacterial products, and reactive neutrophil metabolites.

The effectiveness of rM/NEI in providing significant protection against lung hemorrhage induced by NE or CF sol for 24 h probably reflects the efficient distribution of rM/NEI in the airways and its relatively slow clearance from the airway surface, as suggested by studies with fluorochrome-labeled rM/NEI. rM/NEI-TR was readily detectable 24 h after instillation, the longest time examined. The biochemical indices of lung injury represented by the hemorrhage measurements indicate that at least one-third of the rM/NEI instilled into the lung remains available and active on the airway surfaces at 24 h after instillation.

M/NEI, a native human protein found at sites of inflammation, is effective in inhibiting NE in vitro and NE-induced damage in an in vivo model of lung injury resulting from instilled human NE and CF sol. Recent studies have found markers of inflammation, such as high neutrophil counts, active NE, and IL-8 even in infant CF patients (6, 7), and the inflammatory process itself is increasingly recognized as a major cause of parenchymal damage in CF. Because rM/NEI has the ability to inhibit NE, a primary agent of human inflammatory damage, the local use of  rM/NEI applied directly to the airway as an aerosol offers promise for preventing or reducing the lung injury component of CF.