Human Neutrophil Elastase Inhibitors Research Articles

Directed evolution of a protein: selection of potent neutrophil elastase inhibitors displayed on M13 fusion phage.

Inhibitors of human neutrophil elastase were engineered by designing and producing a library of phage-displayed protease inhibitory domains derived from wild-type bovine pancreatic trypsin inhibitor and fractionating the library for binding to the target protease. The affinity of one of the engineered variants for human neutrophil elastase (Kd = 1.0 pM) is 3.6 x 10(6)-fold higher than that of the parental protein and exceeds the highest affinity reported for any reversible human neutrophil elastase inhibitor by 50-fold. Thus the display phage method has allowed us to obtain protein derivatives that exhibit greatly increased affinity for a predetermined target. The technology can be applied to design high-affinity proteins for a wide variety of target molecules.

Effect of the 7-amino substituent on the inhibitory potency of mechanism-based isocoumarin inhibitors for porcine pancreatic and human neutrophil elastases: a 1.85-A X-ray structure of the complex between porcine pancreatic elastase and 7-[(N-tosylphenylalanyl)amino]-4-chloro-3- methoxyisocoumarin.

A series of new acyl, urea, and carbonate derivatives of 7-amino-4-chloro-3-methoxyisocoumarin were synthesized and evaluated as irreversible inhibitors of human neutrophil elastase (HNE) and porcine pancreatic elastase (PPE). Inhibition of HNE is directly related to the hydrophobicity of the substituent on the 7-amino group. The N-Tos-Phe derivative (19) is the best HNE inhibitor with a second-order rate constant kobs/[I] = 200,000 M-1 s-1. The closest analogue in this series, the 3,3-diphenylpropionyl derivative 5, had a kobs/[I] = 130,000 M-1 s-1 with HNE. In contrast to the Tos-Phe derivative 19, phenylacetyl derivative 2 and carbonates 22 and 25 gave extremely stable enzyme-inhibitor complexes with deacylation half-lives longer than 48 h with both elastases. N-Phenylurea derivative 25 was the best inhibitor for PPE with a second-order rate constant kobs/[I] = 7300 M-1 s-1. The crystal structure of a complex of PPE with N-tosyl-Phe derivative 19 was determined at 1.85-A resolution and refined to a final R factor of 16.9%. The isocoumarin forms an acyl enzyme with Ser-195, while His-57 is near the inhibitor, but not covalently linked. The Tos-Phe makes a few hydrophobic contacts with the S' subsites of PPE, but appears to be interacting primarily with itself in the PPE structure. This region of HNE is more hydrophobic and modeling indicates that the inhibitor would probably make additional contacts with the enzyme.

Neutrophil elastase stimulates tracheal submucosal gland secretion that is inhibited by ICI 200,355.

To investigate whether human neutrophil elastase (HNE) stimulates airway submucosal gland secretion, we studied the effect of purified HNE on secretion of 35S-labeled macromolecules from isolated tracheal tissues from ferrets, dogs, and humans. HNE stimulated secretion in a concentration-dependent fashion, the secretory response being most pronounced in ferret tissues with a maximal response of 1,498 +/- 384% above baseline at 10(-5) M. In dog tissues, maximal secretory responses (509 +/- 169%) to HNE were much greater than to bethanechol (80 +/- 26%). Human tissues obtained several hours postmortem still responded to HNE (179 +/- 48% at 10(-5) M) significantly more than to the combination of isoproterenol, phenylephrine, and bethanechol (23 +/- 10%). Morphometric analysis of canine tracheal tissues showed degranulation of submucosal gland cells after HNE. A specific inhibitor of HNE (ICI 200,355) potently inhibited secretory responses in a concentration-dependent fashion. We suggest that HNE in airways of patients may cause hypersecretion and that treatment with ICI 200,355 may provide a strategy for therapeutic intervention.

A cytosolic inhibitor of human neutrophil elastase and cathepsin G.

The neutrophil serine proteinases elastase and cathepsin G produce connective tissue injury, the extent of which depends on the balance between these enzymes and their inhibitors. The most important of these inhibitors is alpha 1-proteinase inhibitor, a member of a superfamily of homologous proteins known as serpins. Neutrophil cytosol inhibited the activities of human neutrophil elastase and cathepsin G in a dose-dependent fashion. To demonstrate formation of an enzyme-inhibitor complex, we combined 125I-elastase or 125I-cathepsin G with neutrophil cytosol or alpha 1-proteinase inhibitor and analyzed the products by polyacrylamide gel electrophoresis. Unbound elastase and cathepsin G each migrated to an apparent molecular weight of 25 kDa. In the presence of cytosol from neutrophils both radiolabeled enzymes migrated with a relative size of 68 kDa, whereas in the presence of alpha 1-proteinase inhibitor the relative size was 85 kDa. Enzyme-inhibitor complexes were stable in sodium dodecyl sulfate at 100 degrees C but were dissociated by hydrolysis in ammonium hydroxide (1.5 mol/L) at 37 degrees C. Formation of each complex was prevented by pretreatment of elastase or cathepsin G with diisopropylfluorophosphate, indicating that the inhibitor binds to the active site of the enzyme. Exposure of either alpha 1-proteinase inhibitor or neutrophil cytosol to the myeloperoxidase-H2O2-halide system prevented complex formation, suggesting the presence of an oxidizable amino acid at the binding site of the inhibitor. By electrophoretic analysis, the molecular weight of the cytosolic inhibitor was 43 kDa and neutrophils contained approximately 1 attomol of inhibitor per cell. The isoelectric points of the elastase and cathepsin G inhibitor were 5.5-5.9 and inhibitors of the two proteinases coeluted using size exclusion chromatography. These data demonstrate that human neutrophil cytosol contains a single serpinlike protein that inhibits elastase and cathepsin G. The inhibitor may be important in protecting the intracellular environment from proteolytic injury during degranulation.

Conversion of the Alzheimer's beta-amyloid precursor protein (APP) Kunitz domain into a potent human neutrophil elastase inhibitor.

Site-specific mutagenesis techniques have been used to construct active site variants of the Kunitz-type protease inhibitor domain present in the Alzheimer's beta-amyloid precursor protein (APP-KD). Striking alteration of its protease inhibitory properties were obtained when the putative P1 residue, arginine, was replaced with the small hydrophobic residue valine. The altered protein was no longer inhibitory toward bovine pancreatic trypsin, human Factor XIa, mouse epidermal growth factor-binding protein, or bovine chymotrypsin, all of which are strongly inhibited by the unaltered APP-KD (Sinha, S., Dovey, H. F., Seubert, P., Ward, P. J., Blacher, R. W., Blaber, M., Bradshaw, R. A., Arici, M., Mobley, W. C., and Lieberburg, I. (1990) J. Biol. Chem. 265, 8983-8985). Instead, the P1-Val-APP-KD was a potent inhibitor of human neutrophil elastase, with a Ki = 0.8 nM, as estimated by the inhibition of the activity of human neutrophil elastase measured using a chromogenic substrate. It also inhibited the degradation of insoluble elastin by the enzyme virtually stoichiometrically. Replacement of the P1' (Ala) and P2' (Met) residues of P1-Val-MKD with the corresponding residues (Ser, Ile) from alpha 1-proteinase inhibitor resulted in an inactive protein, underscoring the mechanistic differences between the serpins from the Kunitz-type protease inhibitor family. These results confirm the importance of the P1 arginine residue of APP-KD in determining inhibitory specificity, and are also the first time that a single amino acid replacement has been shown to generate a specific potent human neutrophil elastase inhibitor from a human KD sequence.

Biologic Characterization of ICI 200,880 and ICI 200,355, Novel Inhibitors of Human Neutrophil Elastase

ICI 200,880 and its close structural analog, ICI 200,355, are representatives of a new chemical class of inhibitors of human neutrophil elastase (HNE). Both compounds are substituted tripeptide ketones, which demonstrated competitive kinetics versus HNE, with identical Ki values of 5.0 x 10(-10) M. The selectivity of ICI 200,880 for HNE versus a variety of enzymes ranged from 150-fold [relative to porcine pancreatic elastase (PPE)] to greater than 360,000-fold in favor of HNE. The compound effectively inhibited HNE-hydrolysis of bovine ligamentum nuchae elastin. In pharmacokinetic studies, ICI 200,880 and ICI 200,355 displayed long retention times when administered directly to the lung and were rapidly eliminated after intravenous administration. Pretreatment of hamsters with either inhibitor before intratracheal administration of HNE produced dose- and time-dependent inhibition of enzyme-induced increases in lung weight, total lavageable red cells, and total lavageable white cells. Aerosol administration of ICI 200,880 produced similar results. Subcutaneous administration of either 50 or 100 mumol/kg (twice/day) of ICI 200,880 for 14 or 28 days prevented the time-dependent increase in alveolar diameter produced by a single intratracheal dose of PPE when compound dosing was initiated 24 h after the enzyme. Treatment of hamsters with the same protocol and doses of ICI 200,880 for 8 wk prevented the destructive lesion induced by a single intratracheal dose of HNE. It is concluded that ICI 200,880 and ICI 200,355 have biochemical, pharmacokinetic, and pharmacologic profiles that make them useful therapeutic agents for understanding the role of HNE in various diseases. ICI 200,880 is presently being evaluated in humans.

ONO-5046, a novel inhibitor of human neutrophil elastase

ONO-5046, N-[2-[4-(2,2-Dimethylpropionyloxy)phenylsulfonylamino]benzoyl] aminoacetic acid, competitively inhibited human neutrophil elastase (IC50=0.044 μM, Ki = 0.2μM). It also inhibited leukocyte elastase obtained from rabbit, rat, hamster and mouse. However, ONO-5046 did not inhibit trypsin, thrombin, plasmin, plasma kallikrein, pancreas kallikrein, chymotrypsin and cathepsin G even at 100μM. In in vivo studies, ONO-5046 suppressed lung hemorrhage in hamster (ID50 = 82 μg/kg) by intratracheal administration and increase of skin capillary permeability in guinea pig (ID50 = 9.6 mg/kg) by intravenous administration, both of which were induced by human neutrophil elastase.

Biochemistry and Pharmacology of ICI 200,880, a Synthetic Peptide Inhibitor of Human Neutrophil Elastase

ICI 200,880 is representative of a new chemical class of inhibitors of human neutrophil elastase (HNE). The compound demonstrated competitive kinetics vs HNE with a Ki value of 5.0 x 10(-10) M. The selectivity of ICI 200,880 for HNE versus a variety of enzymes ranged from 150- (relative to porcine pancreatic elastase [PPE]) to greater than 360,000-fold in favor of HNE. In pharmacokinetic studies ICI 200,880 displayed a long retention time when administered directly to the lung and was rapidly eliminated when administered intravenously. Aerosol pretreatment of hamsters with ICI 200,880 before intratracheal administration of HNE produced a long-lasting inhibition of enzyme-induced increases in lung weight, total lavageable red cells, and total lavageable white cells. Subcutaneous administration of either 50 or 100 mumol/kg (twice/day) of ICI 200,880 for 14 or 28 days prevented the time-dependent increase in alveolar diameter produced by a single intratracheal dose of PPE when compound dosing was initiated 24 hours following the enzyme. Treatment of hamsters with ICI 200,880 using the same protocol and doses for 8 weeks prevented the destructive lesion induced by a single intratracheal dose of HNE. It is concluded that ICI 200,880 has biochemical, pharmacokinetic, and pharmacologic profiles that make it a useful therapeutic agent for understanding the role of HNE in various diseases. ICI 200,880 is presently being evaluated in man.

Inhibition of human neutrophil elastase by ICI 200,355

To examine the pathogenetic role of neutrophil elastase in airway hypersecretion, we have studied the novel inhibitor of this enzyme, [4-(4-bromophenylsulfonylcarbamoyl)benzoyl-L-valyl-L-proline 1 (RS)-(1-trifluroacetyl-2-methylprolyl)amide] (ICI 200, 355). This compound was a potent (K i = 0.6 ± 0.22 nM) inhibitor of human neutrophil elastase and a much weaker inhibitor of other hydrolases. ICI 200,355 also inhibited the ongoing destruction of insoluble elastin by human neutrophil elastase. ICI 200,355 produced a concentration-dependent inhibition of the secretory response induced by human neutrophil elastase (10 −8 M), with an IC 50 of 1.6 × 10 −8 M. ICI 200,355 had no effect on baseline secretion or on the secretory response to chymase, cathepsin G or Pseudomonas aeruginosa elastase. Thus, ICI 200,355 appears to be a useful tool for investigating the role of human neutrophil elastase in inflammatory disorders associated with hyperscretion, such as cystic fibrosis, chronic bronchitis, and asthma.

Biochemical and Pharmacological Characterization of ICI 186,756: A Novel, Potent, and Selective Inhibitor of Human Neutrophil Elastase

ICI 186,756 is a representative of a new chemical class of synthetic inhibitors of human neutrophil elastase (HNE). This compound demonstrated competitive inhibition of HNE with a Ki of 3.6 +/- 0.8 x 10(-9) mol/L. The selectivity of ICI 186,756 for HNE versus a variety of enzymes ranged from a minimum of 870-fold to greater than 640,000. The compound effectively inhibited hydrolysis of bovine ligamentum nuchae elastin by HNE. Pretreatment of hamsters with ICI 186,756 up to 2 h before intratracheal administration of HNE inhibited enzyme-induced increases in lung weight, total lavageable red cells, and total lavageable white cells measured 24 h after HNE administration. In contrast, similar lung effects produced by intratracheal administration of porcine pancreatic elastase (PPE) were not inhibited by ICI 186,756. Treatment of hamsters with 43 mumol/kg (sc) of ICI 186,756 for 14 or 28 days modulated the increases in alveolar diameter produced by both PPE and HNE, respectively. It is concluded that ICI 186,756 is a potent, competitive, and selective inhibitor of HNE that may be useful in understanding the role of this enzyme in animal models of various diseases. Furthermore, the maintenance or progression of emphysema-like lesions induced in hamsters by PPE do not appear to be due to the persistence of that enzyme within the lung.

Effect of human antileucoprotease on experimental emphysema

Antileucoprotease (ALP), a potent inhibitor of human neutrophil elastase (HNE), may be a modulating factor in the pathogenesis of emphysema. Investigating the clearance of intratracheally-instilled ALP in hamsters, we observed a rapid clearance from the airway lumen within 60 min, whereafter the remaining 40% slowly decreased with a calculated half-life (T0.5) of 2.8 h. Lung tissue-associated ALP showed a peak at 40 min and slowly decreased (T0.5 approximately 3 h). In vivo efficacy of ALP on HNE-induced pulmonary lesions was studied by instillation of either 365 micrograms or 730 micrograms ALP, followed after 1 h by 420 micrograms HNE. Emphysema, haemorrhage and secretory cell metaplasia (SCM) were quantified 21 days after instillations. ALP was found to be able to inhibit emphysema and haemorrhage in a dose-related way, the decrease of haemorrhage being less pronounced. SCM was minimally affected. These results show that ALP inhibits efficiently the development of HNE-induced emphysema and, to a lesser extent, haemorrhage. We speculate that tissue-associated ALP might be responsible for this protection.

The inhibition of human neutrophil elastase and cathepsin G by peptidyl 1,2-dicarbonyl derivatives

Neutrophil elastase and cathepsin G are serine proteases that can damage connective tissue and trigger other pathological reactions. Compounds containing a peptide sequence to impart specificity and bearing an α-dicarbonyl unit (α-diketone or α-keto ester) at the carboxy terminus are potent inhibitors of the neutrophil serine proteases (human neutrophil elastase: R-Val-COCH 3, K i = 0.017 μM; R-Val-COOCH 3, K i = 0.002 μM; human neutrophil cathepsin G: R-Phe-COCH 3, K i = 0.8 μM; R-Phe-COOCH 3, K i = 0.44 μM; R = N-(4-[(4-chlorophenyl)sulfonylaminocarbonyl]phenylcarbonyl)ValylProlyl).

Induction and Exacerbation of Emphysema in Hamsters with Human Neutrophil Elastase Inactivated Reversibly by a Peptide Boronic Acid

In solution, MeO-Suc-Ala-Ala-Pro-D,L-boro-Val pinacol ester (Boroval) is a highly effective but reversible inhibitor of both porcine pancreatic elastase and human neutrophil elastase (HNE) (50% inhibition with a 1.5 M ratio of Boroval to elastase). Boroval has been shown to prevent porcine-pancreatic-elastase-induced emphysema in hamsters. But with HNE-induced emphysema in hamsters, pretreatment with as much as a 170-fold M excess of Boroval, given intratracheally 1 h before 0.3 mg HNE, did not prevent emphysema. Indeed, lung volumes were larger after Boroval pretreatment than after HNE alone. Emphysema was also induced by instilling HNE that had been mixed with and inactivated by a 41-fold M excess of Boroval (a molar ratio of 42). When 0.25 or 0.5 mg of HNE were given mixed with a 41-fold M excess of Boroval, the emphysema was much more severe with the 0.5 mg dose. Two hours after instillation of 0.3 mg HNE inactivated with a 34-fold M excess of Boroval, bronchoalveolar lavage contained elastolytic activity but no evidence of hemorrhage. In contrast, hemorrhage was severe in hamsters that had been instilled with 0.3 mg HNE alone. We conclude that Boroval can enhance HNE-induced emphysema. We postulate that Boroval suppresses HNE-induced hemorrhage and the resultant influx of plasma protease inhibitors; the HNE-Boroval complex is transported into the alveolar interstitium, followed by dissociation of the inhibitor from the active site of HNE. Because of its small size, free Boroval is rapidly cleared, and the reactivated HNE attacks elastic fibers, giving rise to emphysema.

(Acyloxy)benzophenones and (acyloxy)-4-pyrones. A new class of inhibitors of human neutrophil elastase.

A series of 4-(acyloxy)- and 4,4'-bis(acyloxy)benzophenones were synthesized. Some of them, pivalates (trimethylacetates) and isobutyrates in particular, were found to be potent and selective inhibitors of human neutrophil (leukocyte) elastase. A series of 2-[(acyloxy)methyl]-5-(acyloxy)-4-pyrones were synthesized regioselectively from kojic acid. The 4-pyrones bearing a long chain acyl group at the 2-position and either pivaloyloxy or isobutyryloxy at the 5-position were potent and selective inhibitors of the human elastase. A number of analogues and derivatives in both series were synthesized in order to study the structure-activity relationship as summarized in Tables I-VI and in Tables IX and X. The inhibition was selective to human neutrophil elastase. No inhibition of porcine pancreatic elastase or bovine pancreatic chymotrypsin (Tables VII and XI) was observed. The most likely mechanism of inhibition is discussed. The implication of these findings for the treatment of rheumatoid arthritis and emphysema is outlined.

Inhibition of human neutrophil elastase by polyguanylic acid and other synthetic polynucleotides.

We have recently described the inhibition of human neutrophil elastase (HNE) by RNA purified from Streptococcus pneumoniae (1). Inhibition of this enzyme by a wide variety of polyanions has also been reported by others. Such agents include ribonucleic acids from yeast (1,2), E. coli (2), polyribosylribitol phosphate from H. influenzae capsules (1), and several sulfated polysaccharides (3-5). In addition to these naturally-occuring polyanions, synthetic RNA homopolymers were also tested previously (2).

Inhibition of Human Neutrophil Elastase by Bacterial Polyanions

We previously demonstrated that pneumococcal extracts contain a highly specific inhibitor of human neutrophil elastase (HNE). We now show that the active inhibitor in these extracts is a high-molecular-weight, heat-stable substance that appears to be RNA, since inhibitory activity of pneumococcal extracts is decreased by incubation with ribonuclease but not by incubation with deoxyribonuclease or proteinase K. Moreover, metabolically labeled ([3H]uridine) pneumococcal RNA, isolated by phenol extraction followed by ethanol precipitation, strongly inhibits HNE. Pneumococcal capsular polysaccharide, although polyanionic, is only weakly inhibitory toward HNE and is not a major source of elastase-inhibitory activity in pneumococcal extracts. On the other hand, the capsule of Haemophilus influenzae type b contains polyribosylribitol phosphate. This highly charged polyanion possesses HNE-inhibitory activity, but only under special circumstances to be discussed below. Pneumococci (type I, type II smooth, type II rough) and H. influenzae (type b) all release HNE-inhibitory activity into their culture medium during growth. By contrast, Klebsiella pneumoniae and Staphylococcus aureus release little (if any) stable HNE-inhibitory activity during growth. We propose that some bacterial pneumonias may spare host tissue because polyanions released by the invading microorganisms (e.g. RNA from autolysing pneumococci) inhibit elastase released from inflammatory neutrophils and thereby modulate accompanying tissue proteolysis. Pneumonias caused by microorganisms that do not release stable polyanionic inhibitors of HNE (e.g., Staphylococcus and Klebsiella) may be correspondingly more injurious to the lung.

Inhibition of human neutrophil elastase by polyguanylic acid and other synthetic RNA homopolymers.

Human neutrophils contain large amounts of a neutral serine protease, human neutrophil elastase (HNE), which has been implicated as a mediator of acute and chronic lung injury. We found that this enzyme is effectively inhibited, at physiological ionic strength, by several synthetic non-base-paired polyribonucleotides. Among the most active of these is polyguanylic acid (poly G). Inhibitory activity is greatest with high-molecular-weight poly G fractions, but poly G fractions even as low as 60K Mr (app) are effective. Both amidolysis of synthetic elastase substrates, such as succinyl-ala-ala-ala-p-nitroanilide, and proteolysis of elastin are blocked. Poly G inhibits elastin proteolysis even when subsequently added to mixtures of elastin and HNE that have first been preincubated together for 10 min. Under these conditions, polyribosylribitol phosphate, a polyanion derived from Haemophilus influenzae capsular polysaccharide, is not inhibitory. Complex formation between HNE and poly G is dependent on ionic rather than covalent interactions, since it is blocked by 0.6 M NaCl but not by inactivation of the enzyme's catalytic-site serine residue with diisopropylfluorophosphate. However, nonspecific ionic interactions alone cannot explain complex formation, since pancreatic elastase and cathepsin G, an even more basic serine protease from human neutrophils, do not form complexes with poly G, even at low ionic strength. Moreover, in the presence of the amphiphiles taurocholic acid and glycocholic acid, HNE is much less effectively blocked by poly G. Peptide chloromethyl ketone-inactivate HNE (which has its extended substrate-binding pocket occupied by the peptidyl inactivator) also fails to form complexes with poly G. These results indicate that HNE may utilize both hydrophobic and ionic binding sites to couple with poly G, and suggest that these sites may be close to or within the extended substrate-binding pocket of the enzyme.

SC-39026, a specific human neutrophil elastase inhibitor

SC-39026, (±) 2-chloro-4-(1-hydroxyoctadecyl)benzoic acid, inhibits human neutrophil elastase with an IC 50 of 0.5 uM (K I of 1.5 uM). Its inhibition of elastase is reversible and noncompetitive at low concentrations (0.5–1.25 uM). Inhibition is “mixed” at higher inhibitor concentrations. SC-39026 is inactive against hog pancreatic elastase, bovine α-chymotrypsin and Pseudomonas aeruginosa elastase, but does inhibit human neutrophil cathepsin G with an IC 50 of approximately 2.5 uM. Neutrophil elastases isolated from rat, hamster, rabbit and hog are also inhibited by SC-39026.

Development of an assay for in vivo human neutrophil elastase activity. Increased elastase activity in patients with alpha 1-proteinase inhibitor deficiency.

Leukocyte extracts contain enzymes that digest fibrinogen and release a fibrinopeptide A-containing fragment. This study was undertaken to identify the responsible proteinase and to characterize the fibrinopeptide A-containing fragment so that it could be used as an index of enzyme activity. Both the fibrinogenolytic activity and the release of the fibrinopeptide A-containing fragment mediated by the leukocyte extracts were shown to be due to human neutrophil elastase (HNE) by the following criteria: activity was completely blocked by a specific HNE inhibitor or by adsorbing HNE from the extracts with a monospecific antibody and reconstitution with purified HNE restored the ability to release the fibrinopeptide A-containing fragment. This fragment was not released by a variety of other proteinases or by HNE-inhibitor complexes indicating that, at least with respect to the enzymes tested, it is a specific product of HNE and its release requires the free enzyme. By separating the products of HNE digestion of fibrinogen using high performance liquid chromatography, identifying the immunoreactive fractions and subjecting them to amino acid analysis, the fragment was identified as A alpha 1-21, indicating an HNE cleavage site at the Val(A alpha 21)-Glu(A alpha 22) bond. The mean plasma A alpha 1-21 level was markedly higher in patients with alpha 1-proteinase inhibitor deficiency as compared to healthy controls (0.2 nM vs. 7.9 nM; P less than 0.0001), consistent with increased in vivo HNE activity in these individuals.

Subcellular localization and further characterization of a new elastase inhibitor from pneumococci.

Streptococcus pneumoniae contains an inhibitor of human neutrophil elastase. The agent does not inhibit other proteases, including neutrophil cathepsin G and pancreatic elastase. It is active in the presence of insoluble elastin as well as synthetic elastase substrates. The inhibitor is present in the pneumococcal cell membrane. [125I]elastase binding studies and inhibition experiments with intact bacterial autoplasts suggest that this agent has its elastase-binding site(s) exposed on the outside of the bacterial cell membrane. Native and randomized membrane vesicles also show equal inhibitory activity. Active inhibitor can be solubilized from pneumococcal membranes by treatment with a dipolar ionic detergent and can then be reconstituted, in active form, within artificial liposomes. Complex formation between the neutrophil elastase inhibitor and neutrophil elastase may involve noncovalent associations. Although elastase containing a covalently bound substrate analog no longer binds the pneumococcal inhibitor, the present study shows that complex formation is nevertheless independent of neutrophil elastase catalytic activity. Specific inhibitor activity and inhibitor release during bile salt-stimulated autolysis are greater in a nonnecrotizing pneumococcal strain (type I) than they are in a necrotizing strain (type III) or in Klebsiella pneumoniae. These results may help explain the frequent resolution of some pneumococcal pneumonias, despite the presence in the early pneumonic exudate of many neutrophils containing an elastolytic protease capable of injuring lung connective tissue.