Pseudomonas aeruginosa Elastase Does Not Inactivate

Transcription

1 INFECTION AND IMMUNITY, Dec. 1989, p /89/ $02.00/0 Copyright 1989, American Society for Microbiology Vol. 57, No. 12 Psedomonas aerginosa Elastase Does Not Inactivate ot1-proteinase Inhibitor in the Presence of Lekocyte Elastase MARC PADRINES AND JOSEPH G. BIETH* Laboratoire d'enzymologie, Institt National de la Sante et de la Recherche Medicale U 237, Universite Lois Paster de Strasborg, Illkirch, France Received 16 Jne 1989/Accepted 21 Agst 1989 Psedomonas aerginosa elastase rapidly inactivates a1-proteinase inhibitor by splitting its Pro-357-Met-358 peptide bond. The present stdy was aimed at testing whether this reaction takes place in the presence of lekocyte elastase. To this end we added a1-proteinase inhibitor to a mixtre of the two elastases, and we performed the following assays: (i) measrement of the residal lekocyte elastase activity, (ii) sodim dodecyl slfate-polyacrylamide gel electrophoresis, and (iii) immnoassay of the lekocyte elastase-al-proteinase inhibitor complex. These experiments were done with varios concentrations of the three proteins. All experiments gave the same reslt: lekocyte elastase was flly inhibited by ac-proteinase inhibitor in the presence of P. aerginosa elastase even when the bacterial enzyme was 10-fold more concentrated than the netrophil enzyme. We also measred the initial rate of the P. aerginosa elastase-catalyzed inactivation of a1-proteinase inhibitor as a fnction of the inhibitor concentration. The kcaitkm vale derived from this experiment was 9 x 104 M-l s-, a vale mch lower than the rate constant for the lekocyte elastase-inhibitor association (k.., 1.7 x 107 M-1 s-1). This rationalizes the above reslts. In conclsion, when ao-proteinase inhibitor is faced with its target enzyme, lekocyte elastase, it will perform its physiologic antielastase fnction even if the bacterial elastase is present in excess. Psedomonas aerginosa is an opportnistic pathogen that may case fatal infections in vlnerable hosts. Severe lng infections cased by this bacterim have been reported in immnodeficient patients, in children with cystic fibrosis, and in postoperative patients sing respirators. Patients sffering from brns may also ndergo severe wond infections. P. aerginosa infections of the hman cornea, althogh not common, sally reslt in loss of vision. Overwhelming septicemia may complicate the local infection (17). Most P. aerginosa strains secrete an elastase (PsE) and two other proteinases (30). PsE was isolated and extensively stdied by Morihara and his grop (15, 19, 22, 23). It has a moleclar mass of abot 40 kilodaltons (kda) and a pi of 5.9. It is a metalloenzyme with one Zn2+ atom per molecle. It solbilizes hman lng elastin 10-fold faster than does hman lekocyte elastase (10). It also degrades other matrix proteins, sch as type III and IV collagens and laminin (11, 12) ṖsE is thoght to be an important virlence factor of the bacterim. Some psedomonas septicemias are, for instance, accompanied by severe elastolysis of arterial elastin (24). Intraplmonary injection of elastase cases intra-alveolar hemorrhage and necrosis in the animal (9). The enzyme also cases dermonecrosis in experimental models, sggesting a pathogenic role in wond and skin infections (15). PsE frther contribtes to the psedomonas virlence by impairing host defense mechanisms. It degrades components of the immne system sch as immnogloblins A and G and a variety of complement components, and it inactivates hman airway lysozyme. In addition, it inactivates a nmber of hman protein proteinase inhibitors, inclding a1-proteinase inhibitor (capi), Cl inhibitor, al-antichymotrypsin, and mcs proteinase inhibitor (for a review, see reference 4). Inactivation of o1pi reslts from the cleavage of a single peptide bond at the Pro-357-Met-358 link of the inhibitor * Corresponding athor (21). The latter therefore acts as a macromoleclar sbstrate of the bacterial elastase. Morihara and co-workers (22) showed that this enzyme-sbstrate reaction is very fast; at a PsE-to-a1PI molar ratio of 0.01, more than 90% of the inhibitor is inactivated within 1 h. They did not, however, determine the kinetic constants of this enzymatic reaction. a1pi is a serine proteinase inhibitor present in the intraand extravasclar space. Its high association rate constant with lekocyte elastase sggests a physiological antielastase fnction in plasma and in tisses. Netrophils recrited at sites of infection may release their elastase dring phagocytosis or cell death. Therefore, if the inactivation of a1pi by PsE also occrs in vivo, it may lead to ncontrolled tisse damage cased by both bacterial and netrophil elastases. The present stdy demonstrates that a1pi is not inactivated by PsE in the presence of netrophil elastase, a reslt consistent with the finding that the rate constant for the inactivation of o1pi by PsE is mch lower than that for the association of the inhibitor with netrophil elastase. MATERIALS AND METHODS Enzymes, sbstrates, inhibitor, and bffer. Hman plasma aopi was isolated and tested for prity and activity as described previosly (7). All experiments were rn at 25 C in 0.1 M HEPES (N-2-hydroxyethylpiperazine-N'-2-ethaneslfonic acid)-0.3 M NaCl, ph 7.4, a soltion that will be called the bffer throghot this paper. Hman lekocyte elastase (HLE) was isolated from prlent sptm by the method of Martodam and associates (18). The enzyme migrated as a single protein band on sodim dodecyl slfatepolyacrylamide gel electrophoresis and was active-site titrated (27) with acetyl-ala2-pro-azaala-p-nitrophylester (Enzyme System Prodcts, Livermore, Calif.) Porcine pancreatic trypsin came from Choay (Paris, France) and was active-site titrated (8) with p-nitrophenyl-p'-ganidinobenzoate (Sigma, Paris, France). PsE (Nagase Co., Kyoto, Japan) was electrophoretically homogeneos. No active-site

2 3794 PADRINES AND BIETH titrant is available for this enzyme. We therefore sed a moleclar weight of 39,500 and an E19m at 280 nm of 14.5 (23) to calclate the molarities of enzyme soltions. Stock soltions of methoxysccinyl-ala2-pro-val-p-nitroanilide and benzoyl-arg-p-nitroanilide (both obtained from Bachem, Bbendorf, Switzerland) were prepared in dimethylformamide. Kinetics of the interaction of PsE and HLE with ajpi. Reaction mixtres containing 10 nm PsE and varios concentrations of a1pi were incbated at ph 7.4 and 25 C. At selected times, samples were withdrawn and dilted into bffered phosphoramidon soltions to inhibit PsE (20). HLE was then added to these soltions to measre the concentration of active a1pi. The final concentrations of phosphoramidon and HLE (aopi) were 100,M and 100 nm, respectively. After 10 min, the residal HLE activity was determined as indicated in the legend to Fig. 2. The rate constant for the association of HLE with a1pi (kass) was measred nder second-order conditions as described by Bran et al. (6). Immnoassay of the HLE-aoPI complex. We sed a commercially available immnosorbent assay (Merck, Darmstadt, Federal Repblic of Germany) specific for the HLEot1PI complex. Free elastase and a1pi do not interfere with this test (25). HLE and PsE were added simltaneosly to a1pi dissolved in the bffer (the final concentrations are given in Table 1). After 2 min at 25 C, the immnoassay was done by following the instrctions of the manfactrer. All other experimental details are given in the figre legends. RESULTS Kinetics of the interaction of PsE and HLE with a1pi. Morihara and co-workers (21, 22) have shown that ot1pi is inactivated by catalytic amonts of PsE and that this inactivation is de to a proteolytic cleavage of a single peptide bond of the inhibitor. a1pi ths behaves like a sbstrate of PsE. We have postlated that this enzyme-sbstrate interaction follows classical Michaelis-Menten kinetics, its initial rate (v) being given by V = kcat [E.] [So]I(Km + [Sj]) (1) where [E.] and [S.] are the total PsE and a1pi concentrations, respectively, and kcat and Km are the trnover nmber and Michaelis constant, respectively. We have measred v for a series of a1pi concentrations by determining initial rates of a1pi inactivation as shown in Fig. 1A. A plot of v verss ot1pi concentration gave a straight line (Fig. 1B), indicating that the highest sbstrate concentration sed was below Km (2, 3). Eqation 1 therefore simplifies into V = kcat [E.] [So]/Km (2) The only constant which may ths be determined is kcatlkm, also referred to as the proteolytic coefficient. This parameter, derived from the slope of the crve shown in Fig. 1B, was fond to be 9 x 104 M-1 S-1. Eqation 2 integrates into [S] = [S.] e-k' (3) where [S] is the concentration of active cx,pi present at any time t and k = kcat [EoI/Km. We have selected a nmber of inactivation experiments for which the reaction was followed for more than two half-lives (i.e., more than 75% inactivation of a,pi). Semilogarithmic plots of [S] verss t gave straight lines, indicating that the decay of active aotpi is t 0.5 0) Iw w- c 4-- C c ID OL 0 S m i n t e s INFECT. IMMUN C( 1 PI (,UM ) FIG. 1. Kinetic parameters for the inactivation of a1pi by PsE at ph 7.4 and 25 C. (A) Kinetics of inactivation of 1,M ot1pi by 10 nm PsE. The concentration of remaining active ot,pi was measred with HLE as described in Materials and Methods. The initial rate of inactivation was estimated as the slope of the visally drawn tangent (----) to the inactivation crve. For each a,pi concentration, three inactivation experiments were rn, and the initial rates were averaged. (B) Initial rate of ot1pi inactivation as a fnction of a,pi concentration. The concentration of PsE was 10 nm throghot. indeed first order and therefore confirming the validity of eqation 2. The kcatlkm vales derived from these plots were in fair agreement with those obtained from the slope of the crve shown in Fig. 1B. For instance, at 0.5,M ot,pi, kcatlkm vales of 8 x 104, 10 X 104, and 17 x 104 M-1 S-1 were obtained in three separate experiments. On the other hand, with both 1 and 5,M a,pi, identical constants (9 x 104 M-1 S-1) were obtained. Altogether, these data show that the kcat/km ratio can be considered accrate. Competition between PsE and HLE for the reaction with a1pi. Reaction of HLE with an eqimolar concentration of acpi reslted in a very fast inhibition of elastase activity (Fig. 2), in agreement with earlier findings (1). PsE did not significantly decrease this rate of inhibition, provided that the molar ratio of PsE to HLE was lower than 25 (Fig. 2) and that the two elastases were added simltaneosly to acpi (separate experiments showed that addition of PsE to aopi before HLE completely abolished the anti-hle activity of a,pi). HLE therefore competes favorably with PsE for the reaction with ot,pi. For higher PsE concentrations, the rate of HLE inhibition decreased sbstantially, bt there was still 40% HLE inhibition after 120 s for a 100-fold molar excess of PsE over HLE (Fig. 2).

3 VOL. 57, J J_ i i seconds FIG. 2. Effect of PsE on the rate of inhibition of 10 nm HLE by 10 nm a1pi. The two elastases were added simltaneosly to a1pi dissolved in the bffer at 25 C. At selected times, the residal HLE activity was measred by adding 10 R of 20 mm methoxysccinyl- Ala2-Pro-Val-p-nitroanilide to 990 RI1 of reaction mixtre and recording the rate of sbstrate hydrolysis at 410 nm. Symbols: 0, no Pse and 10 nm or 50 nm PsE (the datm points for these three experiments were identical within + 5%); *, 250 nm PsE; A, 1 F.M PsE. PsE inactivates a1pi by splitting the Pro-357-Met-358 bond of the inhibitor. The 4-kDa split carboxy-terminal fragment dissociates from the rest of the molecle nder denatring conditions. Ths, PsE-inactivated a1pi migrates faster than native a1pi in sodim dodecyl slfate-polyacrylamide gel electrophoresis (21). We sed this property to frther stdy the competition between PsE and HLE for the reaction with a1pi. Electrophoresis of mixtres formed of increasing concentrations of HLE and constant eqimolar concentrations of PsE and ot1pi (Fig. 3) showed that if a1pi is faced with an eqimolar concentration of HLE and PsE, it almost flly transforms into a 80-kDa protein, the characteristic sodim dodecyl slfate-stable enzyme-inhibitor complex (29). In contrast, if PsE is in excess over HLE, most a1pi is transformed into its 49-kDa inactive derivative. Some of it, however, forms a complex with HLE, as shown, for FIG. 3. Sodim dodecyl slfate-polyacrylamide gel electrophoresis of mixtres of ca1pi, PsE, and HLE. Lanes: I to 5, 0.5 p.m olxpi pls 0.5 p.m PsE pls 0.01, 0.05, 0.1, 0.5, and 1 p.m HLE, respectively; 6, 0.5 p.m ajpi pls 0.5 p.m PsE; 7, 0.5 p.m aopi; 8, 0.5 pm ajpi pls 0.5 pm HLE. The two elastases were added simltaneosly to acpi dissolved in the bffer. After 2 min at 25 C, the reaction mixtres as well as the marker proteins were treated, electrophoresed, and stained as described by Padrines et al. (26). The 80-, 53-, 49-, and 37-kDa bands correspond to HLE-aoPI complex, native a1pi, proteolysed ot1pi, and PsE, respectively. Nmbers on the left indicate kilodaltons. TABLE 1. PsE-a1PI INTERACTION 3795 Concentrations of immnoreactive HLE-ct1PI complex in mixtres of ct1pi, HLE, and PsE Concn (nm) of: HLE-a1PI complex concn PsE HLE a1pi (nm [% of HLE concn]) (100) (94) (80) (60) (100) (100) instance, by the faint 80-kDa band observed with the mixtre containing 0.2 mol of HLE per mol of PsE (Fig. 3, lane 3). These experiments ths confirm that HLE favorably competes with PsE for the reaction with o1pi. Complex formation between o1pi and HLE was also demonstrated by sing an enzyme-linked immnosorbent assay specific for the HLE-ot1PI complex (25). The experiments were done by sing variable concentrations of a1pi, HLE, and PsE. Table 1 shows that in the presence of eqimolar concentrations of PsE and HLE, the latter was flly bond to a1pi. On the other hand, sbstantial binding still occrred when PsE was in excess over HLE. The reslts of the above experiments clearly demonstrate that HLE is inhibited by a1pi in the presence of PsE even if the latter is in excess and that this inhibition is de to the formation of a sodim dodecyl slfate-stable complex that may be assessed immnologically. These data sggest that the rate of HLE inhibition by a1pi (vinh) is higher than the rate of a1pi inactivation by PsE (Vinac). These rates are given by Vinh = kars [HLE] [a1pi] (4) Vinac = kcat [PsE] [a,pip]/km (5) The second-order association rate constant kars was fond to be 1.7 x 107 M-1 s-1 in or bffer at 25 C (data not shown), a vale that is abot 200-fold higher than kcatlkm. If eqimolar amonts of HLE and PsE are mixed with a1pi, the initial rate of HLE inhibition is ths 200-fold higher than the initial rate of a1pi inactivation by PsE. This rationalizes or competition data. Inflence of PsE on the inhibition of trypsin by ajpi. To confirm the importance of kass, we have done competition experiments with porcine pancreatic trypsin, which reacts abot 100-fold slower with ajpi than does HLE (1). Figre 4 confirms that a1pi is a slow-binding inhibitor of trypsin. The kass calclated from this crve was 4 x 104 M-1 s-1. In the presence of an eqimolar concentration of PsE and trypsin, the rate of inhibition of the latter was significantly impaired, as was expected from the finding that the vale of kass is abot half that of kcatikm. DISCUSSION The proteolytic inactivation of a1pi and other plasma proteinase inhibitors is well docmented. In most cases, the inactivation reslts from the cleavage of a single peptide bond in the active-site region of the inhibitor (4). These target proteins may ths be considered macromoleclar sbstrates whose rate of proteolysis may be expressed in terms of the classical Michaelis-Menten rate eqation. This has been done very recently by Jordan and co-workers (14), who measred kcat and Km for the inactivation of antithrom-

4 3796 PADRINES AND BIETH 0-1- o 100 z cr 5 CK CK min te s FIG. 4. Effect of PsE on the rate of inhibition of 0.1 pm porcine pancreatic trypsin by 0.1 p.m acpi. The experiment was done as indicated in the legend of Fig. 1 except that the residal trypsin activity was measred by adding 10 of 50 mm benzoyl-arg-pnitroanilide to 990 p.l of reaction mixtre. Symbols: 0, no PsE;*, 0.1 p.m PsE. bin III by HLE in the presence of heparin. They reported a kcatlkm of 3 x 106 M-1 s-1 for their system. Or paper provides a frther example of sch a kinetic analysis. Lack of satration kinetics preclded the separate determination of the two parameters. The kcatlkm ratio for the cleavage of ot1pi by PsE cold, however, be determined accrately. This constant relates to a single catalytic event, since cleavage of a single peptide bond in cx1pi cases the inactivation of the inhibitor (21). Its vale (9 x 104 M-1 S-1) is 10-fold lower than that for the cleavage of the Gly-Le bond of carbobenzoxy-gly-le-phe, the best synthetic sbstrate of PsE (19). The proteolytic coefficient kcatlkm is a sefl parameter since it allows the initial rate of inhibitor inactivation to be predicted for any concentration of proteinase and inhibitor (see eqation 5). Also, knowledge of both kcatlkm and kass, the rate constant for the inhibition of the target proteinase, helps in anticipating the extent of inhibitor inactivation and/or complex formation if the latter is mixed with both types of proteinases (see eqations 4 and 5). The finding that kcat1km for the PsE-a1PI system is 200-fold lower than kas, for the HLE-a1PI pair clearly rationalizes the observation that HLE is flly inhibited by a1pi in the presence of an excess of PsE. On the other hand, the low kass for the porcine pancreatic trypsin-a1pi association (4 x 104 M-1 s 1) flly explains why this enzyme-inhibitor binding is strongly impaired by PsE. Morihara et al. (22) fond that the inhibition of bovine pancreatic trypsin by a1pi is only slightly decreased by PsE. This reslt may be explained by the fact that the kass for the a1pi-bovine trypsin association is abot threefold higher than that for a1pi-porcine trypsin interaction (1). Frthermore, or concept helps in predicting the inflence of PsE on other physiologically important proteinase-a1pi interactions. For instance, hman lekocyte cathepsin G (kawss 4 x 105 M-1 S-1 [1]) and hman pancreatic elastase (kass 5 x 105 M- S-1 [16]) will probably be inhibited by a1pi in the presence of PsE, whereas hman pancreatic trypsin (kass, 104 M-1 S-1 [1]) will not. In a previos article, we have shown that HLE is inhibited by mcs proteinase inhibitor in the presence of PsE (28) despite the inactivation of the free inhibitor by the bacterial elastase (13). This reslt may again be explained by the low rate of mcs proteinase inhibitor inactivation by PsE (13, 28) compared with the high kass of the HLE-inhibitor association (5). The discovery of the inactivation of a1pi by PsE led to the proposal that PsE may promote the tisse-damaging action of HLE by lowering the antielastase barrier at sites of infection (22). Or reslts show that while this hypothesis is probably valid dring the early stages of psedomonal infection, it no longer holds once massive netrophil inflx and HLE release takes place. For instance, lng secretions from P. aerginosa-infected patients contain 50- to 100-fold more HLE than PsE (28). Under these conditions, PsEindced a1pi inactivation obviosly does not occr. ACKNOWLEDGMENT INFECT. IMMUN. This work was spported in part by a grant from the Association Francaise de Ltte contre la Mcoviscidose. LITERATURE CITED 1. Beatty, K., J. G. Bieth, and J. Travis Kinetics of association of serine proteinases with native and oxidized ot1-proteinase inhibitor and otl-antichymotrypsin. J. Biol. Chem. 255: Bender, M. L., and F. J. Kezdy Mechanism of action of proteolytic enzymes. Ann. Rev. Biochem. 34: Bieth, J. G Elastases: strctre, fnction and pathological role, p In L. Robert, G. M. Collin-Lapinet, and J. G. Bieth (ed.), Frontiers of matrix biology, vol. 6. S. Karger, Basel. 4. Bieth, J. G Elastases: catalytic and biological properties, p In R. P. Mecham (ed.), Reglation of matrix accmlation. Academic Press, Inc., New York. 5. Bodier, C., and J. G. Bieth Mcs proteinase inhibitor: a fast-acting inhibitor of lecocyte elastase. Biochim. Biophys. Acta 995: Bran, N. J., J. M. Bodner, D. G. Virca, G. Metz-Virca, R. Maschler, J. G. Bieth, and H. P. Schnebli Kinetic stdies on the interaction of eglin c with hman lecocyte elastase and cathepsin G. Biol. Chem. Hoppe-Seyler 368: Brch, M., and J. G. Bieth Inflence of elastin on the inhibition of lekocyte elastase by a1 proteinase inhibitor and bronchial inhibitor. Biochem. J. 238: Chase, T., and E. Shaw p-nitrophenyl-p'-ganidinobenzoate HCl: a new active site titrant for trypsin. Biochem. Biophys. Res. Commn. 29: Gray, L., and A. Kreger Microscopic characterization of rabbit lng damage prodced by Psedomonas aerginosa proteases. Infect. Immn. 23: Hamdaoi, A., F. Wnd-Bisseret, and J. G. Bieth Fast solbilization of hman lng elastin by Psedomonas aerginosa elastase. Am. Rev. Respir. Dis. 135: Heck, L. W., K. Morihara, and D. R. Abrahamson Degradation of solble laminin and depletion of tisse-associated basement membrane laminin by Psedomonas aerginosa elastase and alkaline protease. Infect. Immn. 54: Heck, L. W., K. Morihara, W. B. McRae, and E. J. Miller Specific cleavage of hman type III and IV collagens by Psedomonas aerginosa elastase. Infect. Immn. 51: Johnson, D. A., B. Carter-Hamm, and W. M. Dralle Inactivation of hman bronchial mcosal proteinase inhibitor by Psedomonas aerginosa elastase. Am. Rev. Respir. Dis. 126: Jordan, R. E., R. M. Nelson, J. Kilpatrick, J. 0. Newgren, P. C. Esmon, and M. A. Fornel Inactivation of hman antithrombin by netrophil elastase. Kinetics of the heparin-dependent reaction. J. Biol. Chem. 264: Kawaharajo, K., J. Y. Homma, Y. Aoyama, K. Okada, and K. Morihara Effects of protease and elastase from Psedomonas aerginosa on skin. Jpn. J. Exp. Med. 45:79-88.

5 VOL. 57, Larent, P., and J. G. Bieth Kinetics of the inhibition of free and elastin-bond hman pancreatic elastase by al-proteinase inhibitor and a2-macrogloblin. Biochim. Biophys. Acta 994: Li, P. V Clinical manifestations of infection and crrent therapy, p In R. G. Doget (ed.), Psedomonas aerginosa. American Press, New York. 18. Martodam, R. R., R. J. Bagh, D. Y. Twmasi, and I. E. Liener A rapid procedre for the large scale prification of elastase and cathepsin G from hman sptm. Prep. Biochem. 9: Morihara, K., and H. Tszki Psedomonas aerginosa elastase: affinity chromatography and some properties as a metallo netral proteinase. Agric. Biol. Chem. 39: Morihara, K., and H. Tszki Phosphoramidon as an inhibitor of elastase from Psedomonas aerginosa. Jpn. J. Exp. Med. 48: Morihara, K., H. Tszki, M. Harada, and T. Iwata Prification of hman plasma a1 proteinase inhibitor and its PsE-a1PI INTERACTION 3797 inactivation by Psedomonas aerginosa elastase. J. Biochem. 95: Morihara, K., H. Tszki, and K. Oda Protease and elastase of Psedomonas aerginosa: inactivation by hman plasma al-proteinase inhibitor. Infect. Immn. 24: Morihara, K., H. Tszki, T. Oka, H. Inoie, and M. Ebata. 1%5. Psedomonas aerginosa elastase. Isolation, crystallization and preliminary characterization. J. Biol. Chem. 240: Mll, J. D., and W. S. Callahan The role of elastase of Psedomonas aerginosa in experimental infections. Exp. Mol. Pathol. 4: Nemann, S., N. Heinrich, G. Gnzer, and H. Lang Enzyme-linked immnoassay for elastase in hman plasma. J. Clin. Chem. Clin. Biochem. 19: Padrines, M., M. Schneider-Pozzer, and J. G. Bieth Inhibition of netrophil elastase by o1-proteinase inhibitor oxidized by activated netrophils. Am. Rev. Respir. Dis. 139: Powers, J. C., R. Boone, D. L. Caroll, B. F. Gpton, C. M. Kam, N. Nishino, M. Sakamoto, and P. M. Thy Reaction of azapeptides with hman lekocyte elastase and porcine pancreatic elastase. J. Biol. Chem. 259: Tornier, J. M., J. Jacqot, E. Pchelle, and J. G. Bieth Evidence that Psedomonas aerginosa elastase does not inactivate the bronchial inhibitor in the presence of lekocyte elastase: stdies with cystic fibrosis sptm and with pre proteins. Am. Rev. Respir. Dis. 132: Travis, J., and G. Salvesen Hman plasma proteinase inhibitors. Ann. Rev. Biochem. 52: Wretlind, B., and 0. R. Pavlovskis Psedomonas aerginosa elastase and its role in Psedomonas infections. Rev. Infect. Dis. 5:S998-S1004. Downloaded from on April 3, 2019 by gest