Kinetics of the inhibition of neutrophil proteinases by recombinant elafin and pre-elafin (trappin-2) expressed in Pichia pastoris Marie-Louise Zani 1 , Shila M. Nobar 2 , Sandrine A. Lacour 1 , Soazig Lemoine 3 , Christian Boudier 2 , Joseph G. Bieth 2 and Thierry Moreau 1 1 INSERM U618, University Franc¸ ois Rabelais, Tours, France; 2 Laboratory of Enzymology, INSERM U392, University Louis Pasteur, Faculty of Pharmacy, Illkirch, France; 3 Laboratory of Marine Biology, Universite ´ Antilles-Guyane, Campus de Fouillole, Pointe a ` Pitre, Guadeloupe, France Elafin and its precursor, trappin-2 or pre-elafin, are specific endogenous inhibitors of human neutrophil elastase and proteinase 3 but not of cathepsin G. Both inhibitors belong, together with secretory leukocyte protease inhibitor, to the chelonianin family of canonical protease inhibitors of serine proteases. A cDNA coding either elafin or its precursor, trappin-2, was fused in frame with yeast a-factor cDNA and expressedinthePichia pastoris yeast expression system. Full- length elafin or full-length trappin-2 were secreted into the culture medium with high yield, indicating correct processing of the fusion proteins by the yeast KEX2 signal peptidase. Both recombinant inhibitors were purified to homogeneity from concentrated culture medium by one-step cationic exchange chromatography and characterized by N-terminal amino acid sequencing, Western blot and kinetic studies. Both recombinant elafin and trappin-2 were found to be fast- acting inhibitors of pancreatic elastase, neutrophil elastase and proteinase 3 with k ass values of 2–4 · 10 6 M )1 Æs )1 , while dissociation rate constants k diss were found to be in the 10 )4 s )1 range, indicating low reversibility of the complexes. The equilibrium dissociation constant K i for the interaction of both recombinant inhibitors with their target enzymes was either directly measured for pancreatic elastase or calculated from k ass and k diss values for neutrophil elastase and pro- teinase 3. K i values were found to be in the 10 )10 molar range and virtually identical for both inhibitors. Based on the kinetic parameters determined here, it may be concluded that both recombinant elafin and trappin-2 may act as potent anti-inflammatory molecules and may be of thera- peutic potential in the treatment of various inflammatory lung diseases. Keywords: elafin; enzyme kinetics; neutrophil proteinases; Pichia pastoris; serine protease inhibitor. Inflammatory lung diseases such as chronic obstructive pulmonary disease, emphysema, acute respiratory distress syndrome or cystic fibrosis have been known for a long time to be the consequence of a protease-antiprotease imbalance. The massive accumulation of stimulated polymorpho- nuclear neutrophils (PMNs) at the site of inflammation is accompanied by degranulation and/or lysis of these inflam- matory cells resulting in the extracellular release of a variety of hydrolases and oxidases, as well as reactive oxygen or nitrogen species and antibacterial peptides. More specifi- cally, three serine proteases including human leukocyte elastase, cathepsin G and proteinase 3, are simultaneoulsy released at high concentrations as active enzymes from azurophilic granules of activated polymorphonuclear neu- trophils where they are stored at concentrations reaching millimolar range [1,2]. All three of these serine proteinases participate in the destruction of lung tissues by degrading numerous extracellular matrix proteins such as elastin, type III, IV and VI collagens, fibronectin, laminin, etc. [1,3]. In addition, these proteases stimulate mucous secretion by submucosal gland serous cells and goblet cells and also promote the synthesis of inflammatory cytokines, and therefore have a major role in perpetuating the inflammatory state. Though other degrading proteases including metalloproteases may be released from neutrophils [e.g. MMP-8 (neutrophil collagenase) and MMP-9 (92 kDa gelatinase)], it is thought that serine proteases of neutrophil origin have the greatest contribution to the protease- antiprotease imbalance observed in lung inflammation [3]. In normal lung, the proteolytic activity of extracellular neutrophil serine proteinases is efficiently regulated by at least three natural protease inhibitors present in the lung fluid, namely a 1 -proteinase inhibitor (a 1 -PI, also known as Correspondence to T. Moreau, INSERM U618, University Franc¸ ois Rabelais, 2bis Bd Tonnelle ´ , 37032 Tours Cedex, France. Fax: + 33 247 366 046, Tel.: + 33 247 366 177, E-mail: moreaut@univ-tours.fr Abbreviations: a 1 -PI, a 1 -proteinase inhibitor; SLPI, secretory leuko- cyte proteinase inhibitor; HNE, human neutrophil elastase; PR3, human neutrophil proteinase 3; PPE, porcine pancreatic elastase; Suc- (Ala) 3 -p-NA, succinyl-Ala-Ala-Ala-p-nitroanilide; MeO-Suc-(Ala) 2 - Pro-Val-p-NA, methoxysuccinyl-Ala-Ala-Pro-Val-p-nitroanilide; MeO-Suc-Lys-(pico)-Ala-Pro-Val-TBE, methoxysuccinyl-Lys- (2-picolinoyl)-Ala-Pro-Val-thiobenzyl ester; rec-elafin, recombinant elafin; rec-trappin-2, recombinant trappin-2; E, enzyme; I, inhibitor; S, substrate. Enzymes: human neutrophil elastase (HNE; EC 3.4.21.37); human neutrophil proteinase 3 (PR3; EC 3.4.21.76); porcine pancreatic elastase (PPE; EC 3.4.21.36). (Received 13 January 2004, revised 1 March 2004, accepted 8 April 2004) Eur. J. Biochem. 271, 2370–2378 (2004) Ó FEBS 2004 doi:10.1111/j.1432-1033.2004.04156.x a 1 -antitrypsin), a member of the serpin superfamily and two canonical, small inhibitors, secretory leukocyte proteinase inhibitor (SLPI) and elafin. In acute or chronic inflamma- tion, the imbalance is in favor of proteases which widely overwhelm the inhibitory capacity of lung fluid. The biological significance of this control mechanism has been highlighted by the observation that the development of emphysema in certain patients was related to an hereditary deficiency of a 1 -PI, the major elastase inhibitor [4]. In cystic fibrosis, another inflammatory lung disease, high levels of active neutrophil elastase, cathepsin G and proteinase 3 are usually found in pulmonary secretions and have been correlated with the severity of the disease [5,6]. In addition to participating to lung destruction, these proteases exhibit various deleterious effects which contribute to maintaining an inflammatory state and favor the persistence of microbial infections. Taken together, these observations suggest that increasing serine protease inhibitor levels in lungs, e.g. by aerosol administration, would be beneficial to limit the inflammation and therefore the progression of the disease. Indeed, aerosol administration of recombinant SLPI to patients with cystic fibrosis has been shown to markedly decrease the level of active neutrophil elastase and the number of neutrophil at the inflammatory sites due to the reduction of elastase-induced secretion of IL-8 [7–9]. A similar decrease in elastase levels was observed when a 1 -PI was given in aerosol form to cystic fibrosis patients [10]. While development programs for recombinant SLPI have been stalled, highly purified a 1 -PI produced in transgenic animals has been obtained in huge quantities by pharma- ceutical companies (PPL Therapeutics and Bayer), allowing this molecule to enter clinical trials for its potential use as a protein-based drug for cystic fibrosis. As an alternative to a 1 -PI, other neutrophil elastase inhibitors are currently under development [11–13] but, like a1-PI, they target only elastase and not the similar neutrophil proteases, cathep- sin G or proteinase 3. We hypothesized that elafin and/or its precursor, trappin-2 or pre-elafin, might have interesting therapeutic potential due to their capacity to inhibit elastase and proteinase 3. Trappin-2 is a nonglycosylated 114 amino acid protein comprising (a) an N-terminal domain (38 residues) containing several repeated motifs with the consensus sequence Gly-Gln-Asp-Pro-Val-Lys or cemen- toin domain [14] that can anchor the whole molecule by transglutaminase-catalyzed cross-links and (b) a C-terminal four-disulphide domain (56 residues) or whey acidic protein corresponding to elafin, that is homologous to SLPI. Elafin has been shown to be present in lung secretions [15,16] or human epithelia [17] where it is proteolytically released from its precursor trappin-2 by one or several unknown protease(s). To further characterize the maturation of elafin from trappin-2 and to compare the antiproteolytic activity of both inhibitors, we have expressed them in the Pichia pastoris expression system. Using a genetic construct consisting of the yeast a-factor signal sequence, stable transformants were obtained which secrete full-length elafin or full-length trappin-2 in the culture media. Production of elafin or trappin-2 using this expression system allows the rapid purification of large amounts of recombinant inhibitors which may be used for further in vitro characterization and evaluation of their therapeutic potential. Experimental procedures Materials Human neutrophil elastase (HNE; EC 3.4.21.37) and human neutrophil proteinase 3 (PR3; EC 3.4.21.76) were obtained from Athens Research and Technology (Athens, USA). Porcine pancreatic elastase (PPE; EC 3.4.21.36) was purified as described previously [18]. The concentrations of active enzymes were measured according to published methods [19,20]. All the enzyme or inhibitor concentrations mentioned in this article refer to active protein concen- trations. Succinyl-Ala-Ala-Ala-p-nitroanilide [Suc-(Ala) 3 - p-NA], methoxysuccinyl-Ala-Ala-Pro-Val-p-nitroanilide [MeO-Suc-(Ala) 2 -Pro-Val-p-NA] and methoxysuccinyl- Lys-(2-picolinoyl)-Ala-Pro-Val-thiobenzyl ester [MeO-Suc- Lys-(pico)-Ala-Pro-Val-TBE] were from Bachem. The cDNA coding full-length trappin-2 was a kind gift of J. Schalkwijk (University of Nijmegen, the Netherlands). The pPIC9 vector was from Invitrogen (Groningen, the Netherlands) and restriction enzymes were from Life Technologies. Oligonucleotides The following primers (Genset) were used for PCR amplifications. Triplets correspond to amino acids; restriction sites are underlined. Primer 1; 5¢-CGA CTC GAG AAA AGA GCT GTC ACG GGA GTT CCT-3¢, restriction site XhoI. This primer fuses the trappin-2 mature protein immediately downstream of the a-peptide sequence. Primer 2; 5¢-CGA CTC GAG AAA AGA GCG CAA GAG CCA GTC AA-3¢, restriction site XhoI. This primer fuses the elafin mature protein immediately downstream of the a-peptide sequence. Primer 3; 5¢-CGA GCGGCCGCCCCTC TCA CTG GGG AAC-3¢, restriction site NotI. This primer corres- ponds to the common C-terminal portion of elafin and trappin-2. Primers 1 and 2 fuse the trappin-2 and elafin mature protein, respectively, immediately downstream of the a-peptide sequence and downstream of the Lys-Arg dipep- tide sequence which is removed by the yeast KEX2 protease (Pichia pastoris Expression Kit manual, Invitrogen, Groningen, the Netherlands). Cloning of elafin and trappin-2 cDNA into pPIC9 Using the trappin-2 cDNA cloned into pGE-SKA-B/K (20 ng) as a template, PCR amplification was run for 30 cycles of 10 s at 94 °C, 30 s at 55 °Cand45sat68°Cwith primer combination 1 & 3 or 2 & 3. All the reactions were performed using 1.5 pmol of each primer, 20 nmol of each dNTP and 1 U Taq/Pwo polymerase (Expand High Fidelity PCR system, Roche). Amplified fragments were digested with XhoIandNotI, and cloned into the pPIC9 vector. The constructs containing the yeast a-peptide cDNA sequence fused to the mature elafin (pPIC9-elafin) or trappin-2 (pPIC9-trappin-2) cDNA sequence, were checked for the absence of mutations in the coding sequence by sequencing using an ABI PRISM A310 nucleotide sequencer (PE Biosystems, Courtabeuf, France). Ó FEBS 2004 Inhibitory activity of recombinant elafin and trappin-2 (Eur. J. Biochem. 271) 2371 Expression in Pichia pastoris About 10 lg of recombinant elafin or trappin-2 constructs, previously linearized with SalI, were electroporated (ECM399 electroporator, BTX Technologies, Hawthorne, NY, USA) into P. pastoris strain GS115 (his4) competent cells (Invitrogen). His + transformants were selected and screened for elafin or trappin-2 production in small-scale experiments. For the purification of large amounts of recombinant elafin or trappin-2, positives clones were grown in 2 L buffered minimal glycerol-complex medium (BMGY) at 29 °C for two days, harvested and suspended in 300 mL buffered minimal methanol-complex medium (BMMY) containing 1% (v/v) methanol to induce inhibitor production. The supernatant (about 300 mL) was collected after three (trappin-2) or seven (elafin) days of growth at 29 °C with constant methanol concentration (1%) and concentrated 30-fold using a 3 kDa cutoff YM3 ultrafiltra- tion membrane (Millipore, Paris, France). Purification of secreted elafin and trappin-2 Concentrated supernatants containing secreted elafin or trappin-2 were dialysed over a PD10 Pharmacia column against 25 m M sodium phosphate, pH 6.0 (equilibration buffer). Dialysed supernatant (200 lL) was then loaded onto a mono SÒ column HR 5/5 (0.5 · 5 cm) equilibrated with equilibration buffer using a Pharmacia FPLC chro- matographic system. The column was washed with 6 mL of equilibrium buffer to eliminate unbound proteins. Bound elafin and bound trappin-2 were eluted at a flow rate of 1mLÆmin )1 with a linear NaCl gradient of 0–0.2 M in equilibration buffer for 12 min and with a linear NaCl gradient of 0–0.5 M for 21 min, respectively. Absorbance was monitored at 220 nm. The protein content of each peak was analyzed using high resolution Tricine SDS/PAGE gels according to Scha ¨ gger & von Jagow [21]. After several runs performed using the conditions described above, fractions containing elafin or trappin-2 were pooled, concentrated by ultrafiltration with a YM3 membrane (Millipore) and stored at )70 °C until further use. The N-terminal sequence of the purified recombinant proteins was checked using an automated amino acid sequencer (Applied Biosystems 477A) associated with an online model 120A analyzer for the identification of phenylthiohydantoine derivatives. Western blot analysis was performed using a goat anti- elafin polyclonal antibody (Tebu-Bio SA, Le Perray en Yvelines, France) according to the procedure described by Zani et al. [22]. Kinetic measurements Stock solutions of Suc-(Ala) 3 -p-NA were prepared in N-methyl pyrrolidone. Other substrates and dithiodipyri- dine (Sigma) were prepared in dimethylformamide. Organic solvent final concentration was 1% (v/v). All kinetic measurements were carried out at 25 °Cin0.05 M Hepes 0.1 M NaCl, a solution referred to as Ôthe bufferÕ. Substrate breakdown was monitored by following the changes of absorbance at 410 or 324 nm for para- nitroanalide or thiobenzylester derivatives, respectively. When the later substrate was used, 3 m M dithiodipyridine was present in the reaction mixtures to assess the release of benzylthiol. Measurement of the active rec-elafin and rec-trappin-2 concentration. Recombinant elafin (rec-elafin) and recom- binant trappin-2 (rec-trappin-2) preparations were active site titrated using HNE. Reaction mixtures (990 lL) containing constant amounts of enzyme (0.3 l M )and increasing quantities of inhibitor were allowed to incubate for 15 min in the thermostated cell holder of a computerized Uvikon 943 spectrophotometer (Kontron Instruments, Trappes, France) before measurement of the residual enzymatic activity by addition of 10 lL of a 100 m M Suc- (Ala) 3 -p-NA stock solution. Product release was continu- ously recorded until a constant rate of paranitroaniline production was reached (2–4 min), indicating that enzyme (E), inhibitor (I), substrate (S), and their complexes are in thermodynamic equilibrium. The active concentration of both recombinant inhibitors was deduced from the volume of inhibitor necessary to totally inhibit the enzyme assuming a 1 : 1 binding stoichiometry as suggested previously [23,24]. Thespecificactivityofrecombinantelafinandtrappin-2 (ratio of active inhibitor vs. protein content) was found to be about 95% as inferred from active site titration experiments and determination of the protein content by the Bradford method. Determination of the equilibrium dissociation constant K i for the interaction between PPE and rec-elafin or rec- trappin-2. Equilibrium dissociation constants governing the interaction between PPE and rec-elafin or rec-trappin- 2 were determined using titration experiments. Increasing concentrations (2–25 n M ) of each inhibitor were reacted in 990 lL mixtures with 10 n M elastase for 20 min, a time sufficient to ensure full enzyme–inhibitor association under the present experimental conditions as checked by prelim- inary experiments. The residual enzyme activity was measured as mentioned above. To check the competitive nature of the inhibition, 10 n M PPE was reacted with 10 n M rec-elafin or rec-trappin-2 in a total volume of 990 lL. After 20 min, 10 lLofeither20m M or 200 n M Suc-(Ala) 3 -p-NA was added to measure the residual enzyme activity. Controls without inhibitor were run in parallel. Association kinetics. The reactions between rec-elafin or rec-trappin-2 and PR3 or HNE were investigated using the progress curve method [25]. At time zero, one volume of inhibitor + substrate solution was rapidly mixed with one volume of enzyme solution in the thermostated observation cell of a stopped flow apparatus (SFM3, Bio-Logic, Claix, France). Product formation was continuously recorded. Data acquisition and analysis were performed with the BIOKINE software available from the manufacturer. All experiments were done under pseudo-first order conditions, that is, with [I] 0 ¼ 10 · [E] 0 . Kinetics of HNE and PR3 inhibition were studied in the presence of 1.56 m M MeO-Suc-(Ala) 2 -Pro-Val-p-NA and 0.15 m M MeO-Suc-Lys-(pico)-Ala-Pro-Val-TBE, respect- ively, using rec-elafin concentrations varying from 0.6 to 0.9 l M (HNE inhibition) and from 0.8 to 2.0 l M (PR3 inhibition) or rec-trappin-2 concentrations varying from 0.75 to 0.9 l M (HNE inhibition) and from 0.8 to 1.6 l M (PR3 inhibition). 2372 M L. Zani et al.(Eur. J. Biochem. 271) Ó FEBS 2004 Dissociation kinetics. Enzyme–inhibitor complexes were obtained by reacting 1 l M rec-elafin or rec-trappin-2 with the same concentration of HNE or PR3. At time 0, 10 lL of enzyme–inhibitor complex solution was mixed in a spectrophotometer cuvette with 990 lLof1.88m M MeO- Suc-(Ala) 2 -Pro-Val-p-NA or 0.20 m M MeO-Suc-Lys-(pico)- Ala-Pro-Val-TBE in the buffer. The substrate cleavage was continuously monitored until a constant rate of product formation was reached. Results Expression and purification of recombinant elafin and trappin-2 The elafin cDNA and trappin-2 cDNA were cloned into the yeast expression vector pPIC9, allowing the production of both recombinant proteins in the P. pastoris expression system. The cloning strategy was designed so that mature proteins were secreted in the culture supernatant. Because both proteins have a high proportion of basic residues with predicted pI values of 8.51 for elafin and 9.15 for trappin-2 ( COMPUTE PI / MW program at http://www.expasy.org), no tag was introduced for further purification of each molecule by cation-exchange chromatography. The engineered construct contained the yeast a-peptide directly upstream the N-terminus of either elafin or trappin-2 with a slight modification of the linker region between the a-peptide and the full-length protein. The EAEA sequence which corres- ponds to the yeast STE13 protease cleavage site was removed so that the KR dipeptide was now directly upstream of the mature elafin or trappin-2 allowing their release by the yeast KEX2 protease. Induction of protein expression for seven days with methanol of positive yeast clones expressing elafin resulted in a major form with a molecular mass of 7 kDa consistent with mature elafin as assessed by SDS/PAGE under reducing conditions and Western blot analysis (not shown). Those conditions were retained for large-scale production of recombinant elafin. Culture of clones expressing trappin-2 in the same conditions followed by SDS/PAGE analysis of secreted proteins in supernatants revealed the presence of three pro- teins at 15, 13 and 11 kDa (Fig. 1), two of which (15 kDa and 13 kDa) were immunoreactive with antibodies directed against elafin (not shown). N-terminal sequence analysis indicated that full-length trappin-2 correspon- ded to the 15 kDa form whereas the 13 kDa protein was a clipped form of trappin-2 (partial sequence: GQDKVKAQE) resulting from a cleavage at the K32- G33 sequence. The nonimmunoreactive 11 kDa protein was believed to be a non related yeast protein and was not further characterized. To limit the appearance of the 13 kDa clipped form of trappin-2 for the large-scale production of trappin-2, the duration of fermentation was reduced to three days. Under these conditions, no other proteins except the 15 kDa form corresponding to mature trappin-2 were detected in the supernatant by SDS/PAGE analysis (Fig. 1). Recombinant elafin and trappin-2 were purified from yeast culture supernatants by cation-exchange chromato- graphy as described in Experimental procedures. For both recombinant proteins, the elafin-immunoreactive material was recovered in a single major peak (Fig. 2). An aliquot from the main peak was analyzed by high-resolution Tricine SDS/PAGE which revealed a single protein of about 7 kDa and 12 kDa for elafin in reducing and nonreducing conditions, respectively, and 12 kDa (reduced) and 15 kDa (nonreduced) for trappin-2, suggesting apparent homogeneity of the purified proteins (Fig. 2). N-terminal sequence analysis confirmed the identity of full-length elafin (AQEPVKGPVS) and full-length trappin-2 (AVTGVPVKGQ). Determination of the equilibrium dissociation constant K i for the interaction between PPE and rec-trappin-2 or rec-elafin The equilibrium dissociation constant K i for the interaction of pancreatic elastase with recombinant elafin and trappin-2 was determined directly by adding substrate to an equilib- rium mixture of protease and inhibitor, and measuring spectrophotometrically the rate of release of the reaction product. The concentration of both enzyme and inhibitor was low enough to obtain a concave inhibition curve [25] when incubating PPE with rec-elafin (Fig. 3). The best estimates of K i(app), the substrate-dependent K i was obtained by non linear regression analysis of the data based on the following equation [25]: a ¼ 1 Àð½E 0 þ½I 0 þ K iðappÞ ÞÀ ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð½E 0 þ½I 0 þ K iðappÞ Þ 2 À 4½E 0 ½I 0 q 2½E 0 ð1Þ where a is the relative steady state rate and K i(app) ¼ K i (1 + [S] 0 /K m ). The competitive nature of the inhibition was ascertained by measuring the fractional activity a for an equimolar mixture of enzyme and inhibitor using two different substrate concentrations as described in Experi- mental procedures. For both inhibitors, a was found to be substrate-dependent, indicating competitive inhibition. K i was calculated from K i(app) using K m ¼ 1.1 m M [26]. Fig. 1. Evolution of rec-trappin-2 production by Pichia pastoris as a function of the duration of fermentation. Aliquots of concentrated supernatants of rec-trappin-2-secreting P. pastoris cultures were ana- lyzedbyhighresolutionSDS/PAGEandstainedwithCoomassie BrilliantBlueafter0,1,2,3,4,5,6,7and10daysoffermentation (lanes d0, d1, d2, d3, d4, d5, d6, d7 and d10, respectively). Three days of fermentation (d3) were found to be optimum for rec-trappin-2 production before unwanted proteolysis appeared, and were therefore retained for large-scale production. Ó FEBS 2004 Inhibitory activity of recombinant elafin and trappin-2 (Eur. J. Biochem. 271) 2373 Rec-trappin-2 gave a similar inhibition curve from which the K i could be derived (not shown). The values of K i are giveninTable1. Measurement of k ass and k diss for the interaction of rec-elafin or rec-trappin-2 with HNE and PR3 Linear inhibition curves were obtained when reacting increasing amounts of each inhibitor with HNE and PR3, even when using enzyme concentrations as low as 10 n M , indicating that rec-trappin-2 and rec-elafin bind both proteinases too tightly to allow the direct measurement of the equilibrium constant K i . This latter was thus calculated from the association and dissociation rate constants k ass and k diss . Fig. 2. Purification and SDS/PAGE analysis of rec-elafin and rec- trappin-2. Aliquots (200 lL) of concentrated supernatants of rec-ela- fin- or rec-trappin-2-secreting P. pastoris cultures were loaded onto a cationic exchange Mono S column. After extensive washing to remove unbound proteins, bound material was eluted with a linear NaCl gradient ( ). Fractions containing purified rec-elafin (A) or purified rec-trappin-2 (B) corresponding to the major peak (shaded area) were pooled and stored at )70 °C before use. (C) High resolution Tricine SDS/PAGE analysis of purified elafin and purified trappin-2 under nonreducing conditions (–b) or reducing conditions (+b). Molecular masses of the protein standards are shown on the left. Fig. 3. Inhibition of the enzyme activity of pancreatic elastase by rec- elafin. Constant amounts of PPE (10 n M ) were incubated for 20 min with increasing concentrations (0–2.7 · 10 )8 M ) of rec-elafin. The residual enzymatic activity (e) was measured using Suc-(Ala) 3 -p-NA (1 m M ) as a substrate and plotted as a function of inhibitor concen- tration. K i(app) was calculated by nonlinear regression analysis (Results) using these experimental points. The theoretical curve (––) generated using K i(app) ¼ 1.4 n M was superimposed onto the experi- mental data. A similar curve was obtained with rec-trappin-2. 2374 M L. Zani et al.(Eur. J. Biochem. 271) Ó FEBS 2004 The progress curve method was used to follow the time course of HNE and PR3 inhibition. The reagent concen- trations were chosen to yield both easily detectable signals and to avoid significant substrate depletion during the acquisition time. Because enzyme and inhibitor were reacted under pseudo-first order conditions, the concentration of product vs. time is given by the following equation [25]: ½P¼v s t þ v z À v s k ð1 À e Àkt Þð2Þ where [P] is the product concentration at any time t, v z is the rate of substrate breakdown at t ¼ 0 and vs. the steady state rate. The best estimates of k, the apparent pseudo-first order rate constant for the approach to the steady state, v z and v s were obtained by non linear regression analysis of the progress curves based on Eqn (2). HNE and PR3 inhibition was analysed by assuming that E and I react according to a bimolecular and reversible mechanism as described in Scheme I. Hence, k ass , k diss and K i may be deduced from the following relationships [25]: k ¼ k ass ½I 0 1 þ½S 0 =K m þ k diss ð3Þ k diss ¼ kv s =v z ð4Þ K i ¼ k diss =k ass ð5Þ Kinetics for the association of HNE and Pr3 with rec-elafin or rec-trappin-2 were studied as described in Experimental procedures. We observed good fits of the experimental data to the theoretical curves generated using the best estimates of k, indicating that enzyme inhibition was satisfactorily described by Eqn (2) (not shown). Also, k was proportional to [I] 0 . Typical values of k were 0.33 ± 0.02 s )1 and 0.19 ± 0.02 s )1 for the association of HNE with 0.9 l M rec-elafin and rec-trappin-2, respectively, 0.16 ± 0.01 s )1 for the reaction of PR3 with 0.8 l M rec-elafin and 0.19 ± 0.02 s )1 for the inhibition of PR3 by 1.6 l M rec- trappin-2. Accurate values of k diss could not be calculated using Eqn (4) because of the almost complete inhibition of HNE and PR3 once the steady state was reached. For this reason, the dissociation rate constant was independ- ently obtained from further experiments. Figure 4A shows the kinetics of product accumulation following the dilution of an aliquot of preformed HNE–rec-elafin complex into substrate solution. Complex dissociation was triggered by both high dilution (100-fold) and high substrate concentration ([S] 0 ¼ 13.4 K m ). The concentra- tion of the latter was appropriate to ensure both sufficient dissociation (Scheme I) and continuous enzyme detection without significant decrease of its concentration during the experiment. The experimental data were used to calculate the derivative curve (Fig. 4B) representing the concentra- tion of free enzyme vs. time. Free enzyme was almost absent at t ¼ 0 and its concentration increased up to a steady state level corresponding to 17% of the total enzyme present in the reaction mixture (1.7 n M ), indica- ting that E, I and S were in thermodynamic equilibrium with their complexes. A similar procedure was used to study the dissociation kinetics of 1 l M HNE–rec-trappin-2, PR3–rec-elafin and PR3–rec-trappin-2 complexes. Their 100-fold dilution into the appropriate substrate solutions yielded 12%, 43% and 46% of total enzyme release, respectively. As neither free enzyme nor free inhibitor were present to a significant extent at t ¼ 0, the rate of complex dissociation, that is, the rate of enzyme release, is given by: À d½EI dt ¼ d½E dt ¼ k ass ½EIÀk diss ½E½Ið6Þ which integrates into Eqn (7) [27]: ½E¼ ½E e Àðe fk diss tð2½EI 0 À½E e Þ=½E e g À 1Þ e fk diss tð2½EI 0 À½E e Þ=½E e g À½E e =½EIþ1 ð7Þ where [E] e and [E] are the concentrations of free HNE or PR3 at equilibrium and at any time t, respectively, and [EI] 0 and [EI] are the initial concentration of complex and its Scheme 1. Table 1. Equilibrium and rate constants for the inhibition of neutrophil elastase, proteinase 3 and pancreatic elastase by recombinant elafin and recombinant trappin-2. Methods and experimental conditions are described in Experimental procedures. Values are given as means ± SEM. ND, not determined. Enzyme Rec-elafin Rec-trappin-2 k ass ( M )1 Æs )1 ) k diss (s )1 ) K i ( M ) k ass ( M )1 Æs )1 ) k diss (s )1 ) K i ( M ) Neutrophil elastase (3.7 ± 0.1) 10 6 (3.2 ± 0.1) 10 )4 (0.8 ± 0.05) 10 )10a (3.6 ± 0.5) 10 6 (1.1 ± 0.2) 10 )4 (0.3 ± 0.1) 10 )10a Proteinase 3 (3.3 ± 0.03) 10 6 (4 ± 0.3) 10 )4 (1.2 ± 0.1) 10 )10a (2 ± 0.1) 10 6 (3.7 ± 1.1) 10 )4 (1.8 ± 0.6) 10 )10a Pancreatic elastase ND ND (7.5 ± 1.5) 10 )10 ND ND (3.2 ± 0.8) 10 )10 a Calculated as the k diss /k ass ratio. Ó FEBS 2004 Inhibitory activity of recombinant elafin and trappin-2 (Eur. J. Biochem. 271) 2375 concentration at any time t, respectively. Figure 4B shows the theoretical curve calculated by non linear regression analysis of the data using Eqn (7) and using the best estimate of k diss governing the dissociation of the HNE–rec- elafin complex. Good fits were also obtained for the three other enzyme–inhibitor pairs indicating that the kinetics of enzyme release satisfactorily agrees with Eqn (7). Compar- ison of the k diss values found using this procedure (Table 1) with values of k reported above show that k diss is 400–1700- fold lower than k and may therefore be neglected in Eqn (3). The average association rate constant k ass for each enzyme- inhibitor pair were therefore calculated from k ass ¼ k (1 + [S] 0 /K m )/[I] 0 and [S] 0 ¼ 1.56 m M , K m ¼ 0.14 m M for the HNE-MeO-Suc-(Ala) 2 -Pro-Val-p-NA system [28] and [S] 0 ¼ 0.15 m M , K m ¼ 0.01 m M for the PR3-MeO- Suc-Lys-(pico)-Ala-Pro-Val-TBE pair [29]. These k ass values are reported in Table 1. Discussion The control of the excessive proteolytic activity of HNE has long been recognized to be crucial to avoid degradation of the lung parenchyma in many inflammatory lung diseases. As a consequence, lung therapies based on the inhibition of HNE have lead to intensive research on the development of HNE inhibitors, either as recombinant proteins or synthetic small-molecule inhibitors [13]. However, there is concern now that other neutrophil-derived proteases, namely cath- epsin G and proteinase 3, might have similar deleterious effects as HNE, hence the necessity to design inhibitors able to target all three neutrophil-derived serine proteases. In our efforts to evaluate the therapeutic potential of recombinant genetically modified protease inhibitors derived from nat- ural inhibitors, we report here on the biosynthetic produc- tion of elafin and its precursor, trappin-2. The cDNA coding either elafin or trappin-2 was cloned into the yeast expression vector pPIC9, allowing the production of both inhibitors in the P. pastoris system. The cloning strategy was designed so that mature elafin or mature trappin-2 were secreted in the culture supernatant. No tag to facilitate the purification of the expressed proteins was introduced because both proteins were predicted to be mainly basic, allowing further purification with cation-exhange chroma- tographic procedures. While the level of elafin production increased up to seven days of fermentation with no apparent modifications of the protein, the expression of trappin-2 was found to become sensitive to unwanted proteolysis as fermentation duration increased. A clipped form of trappin-2 resulting from a cleavage C-terminal to Lys32 appeared together with the full-length trappin-2 after three days of fermentation. Such a proteolytic susceptibility after lysyl residues was observed by Bourbonnais et al.[30] who expressed trappin-2 in Saccharomyces cerevisae.Clea- vage after Lys14 and Lys36 in the so-called cementoin domain of trappin-2 was attributed unambiguously to yapsin-1, an aspartic plasma membrane protease active within the periplasmic space. Though the cleavage sites in trappin-2 were different in the two yeast expression systems, we can hypothesize that yapsin-like enzyme(s) are also involved in the non specific degradation of heterologous proteins expressed in P. pastoris. Reducing the fermentation to a maximum of three days for yeast clones expressing trappin-2 was found to suppress the apparition of the 13 kDa clipped form of trappin-2 at the cost of a somewhat lower protein concentration. Using the culture conditions described above, we purified about 15 mgÆL )1 of each recombinant inhibitor from the yeast culture media. Using shake-flask culture conditions which give expression levels typically low relative to what is obtainable in fermenter cultures [31], we found that the amount of elafin and trappin-2 produced in our system was higher than that reported for trappin-2 expressed in similar conditions in the yeast S. cerevisiae system (2–3 mgÆL )1 ) [30]. Though the range of expression yields is variable from one protein to another, our study confirms that the P. pastoris system allows the production of heterologous proteins at a high concentration level. In addition, considering the ease by which the protein production can be scaled up from shake- flask to fermentation conditions [31], P. pastoris is a system of choice to produce large amounts of therapeutic proteins. Fig. 4. Dissociation kinetics of HNE–rec-elafin complexes. (A) Time course of p-nitroaniline release resulting from the hydrolysis of MeO- Suc-(Ala) 2 -Pro-Val-p-NA by HNE released from its complex with rec-elafin. Complexes were first formed by incubating equimolar concentrations (10 )6 M ) of enzyme and inhibitor. Dissociation of the complexes was induced by dilution in a concentrated substrate (1.88 m M ) solution. (B) Kinetics of elastase release calculated from (A) as described in Results. The theoretical curve (––) superimposed onto the experimental data was calculated using Eqn (7) and k diss ¼ 3.2 10 )4 s )1 . 2376 M L. Zani et al.(Eur. J. Biochem. 271) Ó FEBS 2004 N-terminal sequencing and Western blot analysis showed that recombinant elafin and trappin-2 are identical to the natural proteins. In addition, enzyme kinetics showed that the K i of PPE–rec-elafin complex is very close to that reported for natural elafin [32]. Also, the kinetic constants k ass , k diss and K i for the interaction of rec-elafin with HNE and PR3 are of the same order of magnitude as those reported for chemically synthesized elafin by Ying & Simon [23,24]. The P. pastoris expression system described here therefore yields a protein structurally and functionally identical to natural elafin. Litterature lacks information on the kinetic parameters describing the interaction of trappin-2 with HNE and PR3. Based on the kinetic parameters determined here, the most important result of our investigation is that elafin and trappin-2 have very close inhibitory capacities. This means that the N-terminal cementoin domain of trappin-2 has little or no influence on the reactive inhibitory site of elafin. However, it is noteworthy that trappin-2, but not elafin, has been shown to significantly reduce a HNE-induced experimental lung hemorrhage in hamsters [33] or a lipolysaccharide-induced acute lung inflammation in mice [34]. This has been attributed to the unique capacity of the cementoin domain to be cross- linked to extracellular matrix proteins through the cata- lytic action of tissue transglutaminase(s) [33,34]. In this context, it will be especially interesting to evaluate the inhibitory properties of bound trappin-2, as this covalent linking may increase significantly the bioavailability of such an inhibitor at the site of inflammation, e.g. in the case of therapeutic administration, as well as providing a source of inhibitory elafin. Knowledge of the kinetic parameters characterizing a protease–inhibitor interaction and of the in vivo concen- tration of an inhibitor is necessary to evaluate whether such an inhibitor may control the activity of its target enzyme(s) [25,35]. From the kinetic constants determined here and from the in vivo concentration of elafin estimated to be in the range 1.5–4.5 l M in bronchial secretions of normal patients [16,24], we can conclude that both elafin and its precursor are fast-acting inhibitors of HNE and PR3 with a delay time for total inhibition of a few milliseconds (d(t) ¼ 5/k ass Æ[I] 0 [35]). The second conclusion is that both inhibitors will exhibit a pseudo-irreversible behaviour because the [I] 0 /K i ratio of about 15–45 · 10 3 is greater than 10 3 [25]. Altogether, our results clearly demonstrate for the first time that, in vitro, trappin-2 and elafin exhibit a similar and potent inhibitory capacity towards HNE and PR3, strongly suggesting that boosting elafin or trappin-2 level by an aerosol administration would be beneficial in the treatment of inflammatory lung diseases. Acknowledgements This study was supported by the French cystic fibrosis association ÔVaincre la MucoviscidoseÕ. We thank Dr Antoine Touze ´ for nucleotide sequencing, Miche ` le Brillard-Bourdet for N-terminal protein sequen- cing and Prof. Joost Schalkwijk and Dr Patrick Zeeuwen for their kind gift of trappin-2 cDNA. We also thank Dr Fre ´ de ´ ric Delamotte and Prof. Francis Gauthier for valuable discussions. Shila M. Nobar holds afellowshipfromÔVaincre la MucoviscidoseÕ. References 1. Campbell, E.J., Campbell, M.A. & Owen, C.A. (2000) Bioactive proteinase 3 on the cell surface of human neutrophils: quantifi- cation, catalytic activity, and susceptibility to inhibition. J. Immunol. 165, 3366–3374. 2. Campbell, E.J., Silverman, E.K. & Campbell, M.A. (1989) Elas- tase and cathepsin G of human monocytes. Quantification of cellular content, release in response to stimuli, and heterogeneity in elastase-mediated proteolytic activity. J. Immunol. 143, 2961–2968. 3. Owen, C.A. & Campbell, E.J. (1999) The cell biology of leukocyte- mediated proteolysis. J. Leukoc. Biol. 65, 137–150. 4. Perlmutter, D.H. & Pierce, J.A. (1989) The alpha 1-antitrypsin gene and emphysema. Am. J. Physiol. 257, 147–162. 5. Birrer, P., McElvaney, N.G., Rudeberg, A., Sommer, C.W., Liechti-Gallati, S., Kraemer, R., Hubbard, R. & Crystal, R.G. (1994) Protease-antiprotease imbalance in the lungs of children with cystic fibrosis. Am. J. Respir. Crit. Care Med. 150, 207–213. 6. Doring, G. (1994) The role of neutrophil elastase in chronic inflammation. Am.J.Respir.Crit.CareMed.150, S114–S117. 7. McElvaney, N.G., Nakamura, H., Birrer, P., Hebert, C.A., Wong, W.L.,Alphonso,M.,Baker,J.B.,Catalano,M.A.&Crystal,R.G. (1992) Modulation of airway inflammation in cystic fibrosis. In vivo suppression of interleukin-8 levels on the respiratory epithelial surface by aerosolization of recombinant secretory leu- koprotease inhibitor. J. Clin. Invest. 90, 1296–1301. 8. Vogelmeier, C., Gillissen, A. & Buhl, R. (1996) Use of secretory leukoprotease inhibitor to augment lung antineutrophil elastase activity. Chest 110, 261S–266S. 9. Vogelmeier, C., Hubbard, R.C., Fells, G.A., Schnebli, H.P., Thompson, R.C., Fritz, H. & Crystal, R.G. (1991) Anti-neutrophil elastase defense of the normal human respiratory epithelial surface provided by the secretory leukoprotease inhibitor. J. Clin. Invest. 87, 482–488. 10. McElvaney, N.G., Hubbard, R.C., Birrer, P., Chernick, M.S., Caplan, D.B., Frank, M.M. & Crystal, R.G. (1991) Aerosol alpha 1-antitrypsin treatment for cystic fibrosis. Lancet 337, 392–394. 11. Delacourt, C., Herigault, S., Delclaux, C., Poncin, A., Levame, M., Harf, A., Saudubray, F. & Lafuma, C. (2002) Protection against acute lung injury by intravenous or intratracheal pre- treatment with EPI-HNE-4, a new potent neutrophil elastase inhibitor. Am. J. Respir. Cell. Mol. Biol. 26, 290–297. 12. Bradbury, J. (2001) Neutrophil elastase inhibitors for cystic fibrosis. Drug Discov. Today 6, 441–442. 13. Tremblay, G.M., Janelle, M.F. & Bourbonnais, Y. (2003) Anti- inflammatory activity of neutrophil elastase inhibitors. Curr. Opin. Invest. Drugs 4, 556–565. 14. Nara, K., Ito, S., Ito, T., Suzuki, Y., Ghoneim, M.A., Tachibana, S. & Hirose, S. (1994) Elastase inhibitor elafin is a new type of proteinase inhibitor which has a transglutaminase-mediated anchoring sequence termed ÔcementoinÕ. J. Biochem. 115, 441–448. 15. Sallenave, J.M. & Silva, A. (1993) Characterization and gene sequence of the precursor of elafin, an elastase-specific inhibitor in bronchial secretions. Am. J. Respir. Cell. Mol. Biol. 8, 439–445. 16. Tremblay, G.M., Sallenave, J.M., Israel-Assayag, E., Cormier, Y. & Gauldie, J. (1996) Elafin/elastase-specific inhibitor in broncho- alveolar lavage of normal subjects and farmer’s lung. Am. J. Respir. Crit. Care Med. 154, 1092–1098. 17. Pfundt, R., van Ruissen, F., van Vlijmen-Willems, I.M., Alke- made, H.A., Zeeuwen, P.L., Jap, P.H., Dijkman, H., Fransen, J., Croes, H., van Erp, P.E. & Schalkwijk, J. (1996) Constitutive and inducible expression of SKALP/elafin provides anti-elastase def- ense in human epithelia. J. Clin. Invest. 98, 1389–1399. 18. Shotton, D.M. (1970) Elastase. Methods Enzymol. 19, 113–140. 19. Boudier, C. & Bieth, J.G. (1992) The proteinase: mucus proteinase inhibitor binding stoichiometry. J. Biol. Chem. 267, 4370–4375. Ó FEBS 2004 Inhibitory activity of recombinant elafin and trappin-2 (Eur. J. Biochem. 271) 2377 20. Rao, N.V., Wehner, N.G., Marshall, B.C., Gray, W.R., Gray, B.H. & Hoidal, J.R. (1991) Characterization of proteinase-3 (PR-3), a neutrophil serine proteinase. Structural and functional properties. J. Biol. Chem. 266, 9540–9548. 21. Scha ¨ gger, H. & von Jagow, G. (1987) Tricine-sodium dodecyl sulfate-polyacrylamide gel electrophoresis for the separation of proteins in the range from 1 to 100 kDa. Anal. Biochem. 166, 368–379. 22. Zani, M., Brillard-Bourdet, M., Lazure, C., Juliano, L., Courty, Y., Gauthier, F. & Moreau, T. (2001) Purification and char- acterization of active recombinant rat kallikrein rK9. Biochim. Biophys. Acta 1547, 387–396. 23. Ying, Q.L. & Simon, S.R. (1993) Kinetics of the inhibition of human leukocyte elastase by elafin, a 6-kilodalton elastase-specific inhibitor from human skin. Biochemistry 32, 1866–1874. 24. Ying, Q.L. & Simon, S.R. (2001) Kinetics of the Inhibition of Proteinase 3 by Elafin. Am. J. Respir. Cell Mol. Biol. 24, 83–89. 25. Bieth, J.G. (1995) Theoretical and practical aspects of proteinase inhibition kinetics. Methods Enzymol. 248, 59–84. 26. Bieth, J.G., Spiess, B. & Wermuth, C.G. (1974) The synthesis and analytical use of a highly sensitive and convenient substrate of elastase. Biochem. Med. 11, 350–357. 27. Capellos, C. & Bielski, B.H.J. (1980) Kinetic Systems. Mathema- tical Description of Chemical Kinetics in Solution. Krieger Pub- lication Co., Huntington, NY. 28. Nakajima, K., Powers, J.C., Ashe, B.M. & Zimmerman, M. (1979) Mapping the extended substrate binding site of cathepsin G and human leukocyte elastase. Studies with peptide substrates related to the alpha 1-protease inhibitor reactive site. J. Biol. Chem. 254, 4027–4032. 29. Duranton, J. & Bieth, J.G. (2003) Inhibition of Proteinase 3 by alpha1-Antitrypsin In Vitro Predicts Very Fast Inhibition In Vivo. Am. J. Respir. Cell Mol. Biol. 29, 57–61. 30. Bourbonnais, Y., Larouche, C. & Tremblay, G.M. (2000) Production of full-length human pre-elafin, an elastase specific inhibitor, from yeast requires the absence of a functional yapsin 1 (Yps1p) endoprotease. Protein Expr. Purif. 20, 485–491. 31. Cereghino, G.P., Cereghino, J.L., Ilgen, C. & Cregg, J.M. (2002) Production of recombinant proteins in fermenter cultures of the yeast Pichia pastoris. Curr. Opin. Biotechnol. 13, 329–332. 32. Wiedow,O.,Schroder,J.M.,Gregory,H.,Young,J.A.&Chris- tophers, E. (1990) Elafin: an elastase-specific inhibitor of human skin. Purification, characterization, and complete amino acid sequence. J. Biol. Chem. 265, 14791–14795. 33. Tremblay, G.M., Vachon, E., Larouche, C. & Bourbonnais, Y. (2002) Inhibition of human neutrophil elastase-induced acute lung injury in hamsters by recombinant human pre-elafin (trappin-2). Chest 121, 582–588. 34. Vachon, E., Bourbonnais, Y., Bingle, C.D., Rowe, S.J., Janelle, M.F. & Tremblay, G.M. (2002) Anti-inflammatory effect of pre- elafin in lipopolysaccharide-induced acute lung inflammation. Biol. Chem. 383, 1249–1256. 35. Bieth, J.G. (1984) In vivo significance of kinetic constants of pro- tein proteinase inhibitors. Biochem. Med. 32, 387–397. 2378 M L. Zani et al.(Eur. J. Biochem. 271) Ó FEBS 2004 . Kinetics of the inhibition of neutrophil proteinases by recombinant elafin and pre -elafin (trappin-2) expressed in Pichia pastoris Marie-Louise. Expression Kit manual, Invitrogen, Groningen, the Netherlands). Cloning of elafin and trappin-2 cDNA into pPIC9 Using the trappin-2 cDNA cloned into pGE-SKA-B/K (20