Discriminating between the activities of human cathepsin G and chymase using fluorogenic substrates Brice Korkmaz 1,2 , Gwenhael Je ´ got 1,2 , Laurie C. Lau 3 , Michael Thorpe 4 , Elodie Pitois 1,2 , Luiz Juliano 5 , Andrew F. Walls 3 , Lars Hellman 4 and Francis Gauthier 1,2 1 Unite ´ INSERM U-618 ‘Prote ´ ases et Vectorisation pulmonaires’, Tours, France 2 Universite ´ Franc¸ois Rabelais de Tours, France 3 Immunopharmacology Group, Sir Henry Wellcome Laboratories, Southampton General Hospital, UK 4 Department of Cell and Molecular Biology, The Biomedical Center, Uppsala University, Sweden 5 Departamento de Biofı ´ sica, Escola Paulista de Medicina, Universidade Federal, Sa˜o Paulo, Brazil Introduction Cathepsin G (CG) (EC 3.4.21.20) and chymase (EC 3.4.21.39) are monomeric chymotrypsin-like serine pro- teases that display a high degree of sequence similarity and highly similar substrate specificity [1–3]. They are located predominantly in the primary granules of neu- trophils and mast cells, respectively, although CG may also be found in mast cells [4]. The understanding of their distinctive roles in inflammatory events involving both neutrophils and mast cells can represent a chal- lenge as a result of their closely-related substrate specificities. No substrate has been identified to date that allows differentiation of their activities when both Keywords cathepsin G; chymase; FRET substrate; kinetics; mast cell; serine protease Correspondence B. Korkmaz, Unite ´ INSERM U-618 ‘Prote ´ ases et Vectorisation pulmonaires’, Universite ´ Franc¸ois Rabelais de Tours, 37032 Tours, France Fax: +33 2 47 36 60 46 Tel: +33 2 47 36 62 53 E-mail: brice.korkmaz@inserm.fr (Received 4 April 2010, revised 11 May 2011, accepted 16 May 2011) doi:10.1111/j.1742-4658.2011.08189.x Cathepsin G (CG) (EC 3.4.21.20) and chymase (EC 3.4.21.39) are two clo- sely-related chymotrypsin-like proteases that are released from cytoplasmic granules of activated mast cells and ⁄ or neutrophils. We investigated the potential for their substrate-binding subsites to discriminate between their substrate specificities, aiming to better understand their respective role dur- ing the progression of inflammatory diseases. In addition to their prefer- ence for large aromatic residues at P1, both preferentially accommodate small hydrophilic residues at the S1¢ subsite. Despite significant structural differences in the S2¢ subsite, both prefer an acidic residue at that position. The Ala226 ⁄ Glu substitution at the bottom of the CG S1 pocket, which allows CG but not chymase to accommodate a Lys residue at P1, is the main structural difference, allowing discrimination between the activities of these two proteases. However, a Lys at P1 is accommodated much less effi- ciently than a Phe, and the corresponding substrate is cleaved by b2-tryp- tase (EC 3.4.21.59). We optimized a P1 Lys-containing substrate to enhance sensitivity towards CG and prevent cleavage by chymase and b2- tryptase. The resulting substrate (ABZ-GIEPKSDPMPEQ-EDDnp) [where ABZ is O-aminobenzoic acid and EDDnp is N-(2,4-dinitrophenyl)-ethy- lenediamine] was cleaved by CG but not by chymase and tryptase, with a specificity constant of 190 m M )1 Æs )1 . This allows the quantification of active CG in cells or tissue extracts where it may be present together with chym- ase and tryptase, as we have shown using a HMC-1 cell homogenate and a sputum sample from a patient with severe asthma. Abbreviations ABZ, O-aminobenzoic acid; ACT, antichymotrypsin; CG, cathepsin G; CMK, chloromethyl ketone; EDDnp, N-(2,4-dinitrophenyl)- ethylenediamine; FRET, fluorescence resonance energy transfer; HNE, human neutrophil elastase; PR3, proteinase 3; Z, benzyloxycarbonyl. FEBS Journal 278 (2011) 2635–2646 Journal compilation ª 2011 FEBS. No claim to original French government works 2635 proteases are present. Moreover, CG is weaker than chymase at hydrolyzing most substrates currently employed to quantify their activity and, accordingly, this has hampered studies of their enzymatic properties [4,5]. CG and chymase genes are located on chromosome 14 together with the genes of granzymes B and H [6]. The two proteases are synthesized as a prepro-protein, containing a peptide signal, a prodipeptide and a C-terminal propeptide [7]. Mast cell chymase and CG convert angiotensin I to the vasoactive peptide angio- tensin II in human tissues [8], and this reaction may be important in the progression to heart failure [9] and aortic stenosis [10]. Both proteases can also convert the CXC chemokine connective tissue-activating pep- tide III into active chemokine neutrophil-activating peptide 2 through limited proteolysis [11], and both are secretagogues for cultured serous cells [12]. Mast cell chymase and CG can also inactivate hepatocyte growth factor [13] and both can degrade connective tis- sue components such as fibronectin and vascular endo- thelial cadherin [14]. A close relationship between CG and chymase is highlighted by the recent development of a dual inhibitor, the administration of which has been reported to be efficacious in the treatment of lung inflammation in animal models [15]. The selective presence of CG in neutrophils confers a destructive role on this protease with respect to the degradation of pathogens within the phagolysosomes [16]. CG may also be secreted on neutrophil activa- tion, and may remain associated with the neutrophil membrane as an active protease [17]. Soluble and membrane-bound extracellular CG may participate in the regulation of inflammatory processes through the processing of chemokines ⁄ cytokines and activation of specific cell surface receptors [16,18]. This protease is also likely to contribute to the proteolysis of con- nective tissue components in chronic inflammatory disease [19]. Measuring protease-specific activities in situ is criti- cal for the understanding of their distinctive functions, as well as for the design of drugs that may be able to regulate their activity. Fluorescence resonance energy transfer (FRET) substrates have proven to be valuable alternatives to classical chromogenic and fluorogenic substrates, both in terms of specificity and sensitivity. This is because FRET substrates allow an investigation of protease specificity on both sides of the cleavage site, unlike peptides with 4-nitroanilides, peptide thiob- enzyl esters, 4-methyl-7-coumarylamide or naphthyla- mides, which release chromophores or fluorophores from the C-terminus [20,21]. Moreover, FRET substrates are particularly appropriate for a kinetic investigation of neutrophil serine proteases because these proteases have an extended binding site on both the S and S¢ sides, as shown by X-ray analysis of the complex with inhibitors [22]. Furthermore, synthesis of FRET substrates does not require sophisticated chemi- cal procedures and may be applied readily in the routine measurement of proteolytic activity in biologi- cal fluids or in fractionated cell suspensions [20]. We and others have previously developed FRET substrates that are sufficiently sensitive to measure subnanomolar concentrations of human neutrophil elastase (HNE) ( EC 3.4.21.37) and proteinase 3 (PR3) (EC 3.4.21.76) and CG [20,21,23]. However, to date, no in depth investigation of the S¢ specificity of CG has been carried out that could aid the understanding of its pathophysiological function, and distinguish its activity from that of mast cell chymase. Ultimately, a better knowledge of CG specificity should help in the devel- opment of a selective inhibitor of therapeutic interest. Results and Discussion The crystal structure of CG in complex with the pept- idyl phosphonate inhibitor Suc-Val-Pro-Phe P (OPh) 2 exhibited the characteristic fold of chymotrypsin-like serine proteases and was very similar to that of human chymase [1]. Preferential accommodation of a large hydrophobic residue in the S1 subsite of the two proteases is a result of the absence of a disulfide bond between Cys191 and Cys220, which is conserved in the neutrophil serine proteases HNE and PR3. The presence of a Glu at position 226 at the bottom of the CG S1 subsite explains the accommodation of a posi- tively-charged P1 residue [1,24]. Similar to other chymotrypsin-like serine proteases, CG and chymase preferentially accommodate a Pro at P2, and most of the commonly used chromogenic and fluorogenic substrates contain the Pro-Phe pair at P2–P1 [25,26]. A prolyl residue at the P2 position allows a change in the substrate chain as it threads through the active site, leading to an optimal positioning of the scissile bond in the active site [25]. Lys192 in CG and chymase has been suggested to favour interaction with a negatively- charged P3 residue [1]. These observations explain the very similar substrate specificity of CG and chymase with both synthetic and natural substrates, although CG generally cleaves synthetic substrates more slowly than do chymase and chymotrypsin-like proteases [5]. The S¢ specificity of both CG and chymase is less well documented than S specificity; thus, a better knowl- edge of the combination between S and S¢ specificities could help to distinguish between the specificities of the two proteases. Cathepsin G versus chymase specificity B. Korkmaz et al. 2636 FEBS Journal 278 (2011) 2635–2646 Journal compilation ª 2011 FEBS. No claim to original French government works S1¢ specificity of CG and chymase The crystallographic data reported by Hof et al. [1] indicate that the side-chain of Arg41 located on the 30S insertion loop in CG projects from the molecular surface to the east of the active site in accordance with the standard orientation (Fig. 1). Thus, the S1¢ pocket in CG appears as a narrow crevice stabilized by the Cys42-Cys58 disulfide bridge that defines the 30S loop in both CG and chymase. The S1¢ pocket is bordered by His57 of the catalytic triad, Ser40 and Arg41, whose flexibility allows it to be close to both the S1¢ and S2¢ subsites (Fig. 1). Interestingly, an Arg residue at position 41 is specific to CG and is shared only by human and chimpanzee CG, suggesting a recent appearance over the course of evolution (not shown). In chymase, as in many other serine proteases, residue 41 is a Phe but, unlike other serine proteases, it projects from the surface of the molecule and is proxi- mal to the substrate P2¢ side-chain [27]. Thus, the chymase S1¢ pocket on the top of the Cys42-Cys58 loop is bordered by His57 to the west and the aliphatic part of the Lys40 side-chain to the east [2] (Fig. 1). As a result, the P1¢ specificity of chymase could be different from that of CG on account of the Lys40 in chymase helping to accomodate a negatively-charged P1¢ residue. This could explain why chymase is more efficient than CG at inactivating bradykinin (RPPGFSPFRCOO ) ) upon cleavage of the C-terminal F–R bond [28] and it is likely that the Lys40 in chym- ase will form an electrostatic interaction with the negatively-charged carboxyl group of bradykinin. To confirm this hypothesis, we raised two FRET sub- strates that had either an Asp or an Arg at P1¢. The peptidyl backbone of these substrates was that of a previously described FRET substrate: ABZ-GIA- TFCMLMPEQ-EDDnp (substrate 1) [where ABZ is O-aminobenzoic acid and EDDnp is N-(2,4-dinitrophe- nyl)-ethylenediamine], which was derived from the inhibitory loop sequence of serpinB1 (previously called Arg143 Arg143 Lys40 Lys40 Phe41 P1 P2 P3 P4 S1’ S2’ Arg143 Arg143 Lys217 Lys192 Lys192 Arg217 Arg41 Arg41 Ser40 P1 P2 P3 P4 S1’ S2’ Cathepsin G Chymase Ser40 Arg143 Arg41 Lys192 Lys192 Phe41 His57 Ile99 Ile99 Cys42 Ser40 His57 Arg143 Lys40 Cathepsin G Chymase S1’ S2’ S2’ A B S1’ Fig. 1. Structural differences between CG and chymase. (A) The solvent accessible surface based on the atom coordinates of CG (1CGH) [1] and chymase (1PJP) [2] is coloured to show positive (blue) and nega- tive (red) electrostatic potentials. The irre- versible phosphonate inhibitors Suc-Val-Pro- Phe P -(OPh) 2 and Suc-Ala-Ala-Pro-Phe-chlo- romethylketone complexed to CG and to chymase, respectively, are shown as cyan stick models. The serine of the catalytic triad is yellow. (B) Ribbon plot of CG and chymase in irreversible complexes with syn- thetic inhibitors showing ball-and-stick mod- els for the seven residues located in the vicinity of the active site. The molecular sur- faces were generated using YASARA software (http://www.yasara.org). B. Korkmaz et al. Cathepsin G versus chymase specificity FEBS Journal 278 (2011) 2635–2646 Journal compilation ª 2011 FEBS. No claim to original French government works 2637 monocyte neutrophil elastase inhibitor) and can be cleaved at the F–C bond by CG and by chymase [29,30]. We found that the specificity constants, k cat ⁄ K m , for cleavage by CG and by chymase of ABZ- GIATFDMLMPEQ-EDDnp (substrate 2) and ABZ- GIATFRMLMPEQ-EDDnp (substrate 3) were similar in the 2 · 10 2 mm )1 Æs )1 range (Table 1), indicating that the Lys40 in chymase does not act as a discriminating structural determinant of P1¢ specificity. The two best substrates developed previously for CG, and which are also cleaved by chymase, differ mainly in the size of the P1¢ residue. One is derived from the antichymotrypsin (ACT) sequence (ABZ- TPFSGQ-EDDnp) and bears a Ser at P1¢ [31] and the other, from a CG-cleaved sequence in protease-acti- vated receptor-1, PAR-1 (ABZ-EPFWEDQ-EDDnp), bears a Trp at this position [31,32]. Because of the small size of the S1¢ pocket in CG, we hypothesized that small-sized residues are preferred by CG and that they could possibly help to discriminate between CG and chymase. We introduced either a Ser or a Trp residue at P1¢ in substrate 1 to obtain substrate 4 (ABZ-GI- ATFSMLMPEQ-EDDnp) and substrate 5 (ABZ-GI- ATFWMLMPEQ-EDDnp) and tested these substrates with CG and chymase. As expected, cleavage sites iden- tified by HPLC fractionation of the proteolysis prod- ucts remained unchanged after the P1 Phe residue, and a Trp residue at P1¢ significantly decreased the k cat ⁄ K m ; however, this result was obtained for both proteases (Table 1), which strongly suggests that S1¢ subsites in CG and chymase are too closely related structurally to allow discrimination between these two proteases. S2¢ specificity of CG and chymase Crystallographic data show that the S2¢ subsite of CG is highly polar as a result of the presence of three posi- tively-charged residues: Arg41, Arg143 and Lys192 [1] (Fig. 1). In chymase, the Arg ⁄ Phe substitution at posi- tion 41 projects the Phe side-chain into the active site cleft, resulting in partial obstruction at the bottom of the S2¢ subsite. However, the crystal structure of chym- ase also indicates that the orientation of Arg143 in chymase differs from that in CG and is more proximal to the S2¢ subsite. This probably explains why, despite the Arg ⁄ Phe substitution, chymase accomodates a neg- atively-charged P2¢ residue, as shown using a phage display random nonapeptide library (Fig. 1) [33–35]. Thus, CG and chymase could accommodate a negative P2¢ residue, although via a different mechanism that involves Arg41 and Lys192 in CG and Arg143 and Lys192 in chymase. We have tested the influence of negative and positive residues at P2¢ in the serpinB1- derived FRET substrate to possibly take advantage of this different mechanism for discriminating between the two proteases. We observed a significant increase in specificity constant value using ABZ-GIATFCD- LMPEQ-EDDnp (substrate 6) compared to substrate 1 and a significant decrease in this rate constant using ABZ-GIATFCRLMPEQ-EDDnp (substrate 7) but, again, similar results were obtained with both chymase and CG. Nevertheless, this demonstrates the impor- tance of the S2¢ subsite for both proteases, and also that Arg143 in chymase has a function similar to that of Arg41 in CG (Table 1). This finding is in agreement with our observation that mouse CG, in which Arg41 is replaced by an Ala residue, cleaves substrates 6 and 7 at the same rate [24]. Thus, despite the significantly different structure of their S2¢ subsite, CG and chym- ase have a similar preference for negatively-charged P2¢ residues. We have previously shown that PR3 and HNE poorly accommodate a Pro at P2¢, which empha- sizes the importance of the S2¢ subsite in neutral serine proteases [36]. Unlike PR3 and HNE, CG accommo- dates a P2¢ prolyl residue, as shown using substrate 8 (ABZ-GIATFCPLMPEQ-EDDnp), that is cleaved approximately twice as fast as control substrate 1 (Table 1). Again, however, the same result was obtained with chymase, further confirming the similar specificity of these two proteases. Table 1. Influence of residues at P1, P1¢ and P2¢ on the specificity of CG and chymase as deduced from the specificity constant k cat ⁄ K m with FRET substrates derived from the serpinB1 and ACT- reactive site loops. Values (m M )1 Æs )1 ) are the mean of ‡ 3 experi- ments. The error for k cat ⁄ K m is < 15%. The arrow indicates cleav- age sites by CG and chymase. NSH, no significant hydrolysis. Number Substrates k cat ⁄ K m CG Chymase Derived from SERPINB1 S1¢ specificity 1 ABZ-GIATFCMLMPEQ-EDDnp 263 a 238 2 ABZ-GIATFDMLMPEQ-EDDnp 175 221 3 ABZ-GIATFRMLMPEQ-EDDnp 162 203 4 ABZ-GIATFSMLMPEQ-EDDnp 217 209 5 ABZ-GIATFWMLMPEQ-EDDnp 69 96 S2¢ specificity 6 ABZ-GIATFCDLMPEQ-EDDnp 817 1035 7 ABZ-GIATFCRLMPEQ-EDDnp 68 144 8 ABZ-GIATFCPLMPEQ-EDDnp 560 287 Derived from ACT S1 specificity 9 ABZ-TPFSALQ-EDDnp 153.8 b 136 10 ABZ-TPKSALQ-EDDnp 8.1 b NSH 11 ABZ-TPWSALQ-YNO 2 69 97 a Value from Korkmaz et al. [29]. b Value from Re ´ hault et al. [25]. Cathepsin G versus chymase specificity B. Korkmaz et al. 2638 FEBS Journal 278 (2011) 2635–2646 Journal compilation ª 2011 FEBS. No claim to original French government works S1 specificity of CG and chymase The dual specificity of CG for cleaving after large hydrophobic or positively-charged residues has been explained by the presence of a Glu residue at position 226 at the bottom of the S1 pocket [1,24]. This idea has received support using mouse CG that has an Ala at position 226 and does not cleave P1- Lys containing substrates [24] and, more recently, as a result of a phylogenetic analysis of mammalian CGs [37]. Human chymase also has an Ala residue at position 226 and this could be exploited to raise a specific CG substrate (Fig. 2A). However, the specificity constant for the reaction between CG and a P1 Lys-containing substrate is far lower than that of the corresponding substrate with a Phe at P1 [25]. The presence of an Ala residue at position 226 in chymase also makes the S1 subsite wider, and this could favour the accommodation of a P1 Trp residue by chymase, as recently shown using a phage-displayed selection of peptides susceptible to chymase cleavage [34]. We compared the hydrolysis by CG and chymase of ABZ-TPFSALQ-EDDnp (substrate 9), ABZ-TP KSALQ-EDDnp (substrate 10) and ABZ-TPWSALQ- YNO 2 (substrate 11) (Table 1). As expected, CG and chymase prefered a Phe at P1 (substrate 9), although both also accommodated a Trp in their S1 subsite and only CG cleaved the P1Lys-containing substrate (Table 1). However, this occurred at a very low rate, in accordance with previous findings [25]. Because no other subsites from S2 to S3¢ in CG and chymase demonstrated a specificity that would allow discrimina- tion between the two proteases, we next attempted to improve the specificity constants of P1 Lys-containing substrates, aiming to measure subnanomolar amounts of CG specifically. Design of specific and sensitive substrates for CG and chymase A first step was to improve the k cat ⁄ K m value of P1 Phe-containing substrates before substituting the P1-Phe by Lys. Accordingly, we started from our most sensitive but not specific CG ⁄ chymase FRET substrate ABZ-GIATFCDLMPEQ-EDDnp (substrate 6) and replaced the Thr residue by Pro [ABZ-GIAP FCDLM- PEQ-EDDnp (substrate 12)], aiming to prevent cleav- age at the C–D bond by HNE and PR3 with a Pro at P3 [38,39] and to improve cleavage by CG and chym- ase, although the latter prefers aliphatic residues at P2 [34]. The Pro-Phe pair at P2–P1 is present in most of Cathepsin G Chymase S1 S1 Arg41 Lys192 Gln192 Arg143 Asp143 Asp147 Glu217 Arg217 S2’ S2’ S3 Cathepsin G P3 β2-Tryptase (monomer) A B Fig. 2. Structural differences between CG, chymase and b2-tryptase. (A) Ribbon plot of CG and chymase in a complex with syn- thetic inhibitors. The irreversible phospho- nate inhibitors Suc-Val-Pro-Lys P -(OPh) 2 and Suc-Ala-Ala-Pro-Phe-chloromethyl ketone in a complex with cathepsin G and to chym- ase, respectively, are shown as cyan stick models. Glu 226 and Ala 226 residues at the bottom of the S1 subsite are shown in green. (B) Electrostatic surface potential of human CG and b2-tryptase [50]. Solvent- accessible surfaces with a positive electro- static potential are shown in dark blue, and these with a negative electrostatic potential are shown in red. The serine of the catalytic triad is shown in yellow. The molecular sur- faces were generated using using YASARA software (http://www.yasara.org). B. Korkmaz et al. Cathepsin G versus chymase specificity FEBS Journal 278 (2011) 2635–2646 Journal compilation ª 2011 FEBS. No claim to original French government works 2639 the commonly used chromogenic and fluorogenic substrates of CG that are also cleaved by chymase [31]. As expected, k cat ⁄ K m of substrate 12 was increased significantly using CG and chymase, and was resistant to HNE cleavage (Table 2). However, this substrate was still cleaved by PR3 at the C–D bond (Table 2). Total resistance to PR3 hydrolysis was obtained by substituting Ser for Cys in substrate 12 as a result of the higher electronegative charge of the O atom of the Ser side-chain compared to that of the sulfur atom in the Cys side-chain; P3¢ Leu for Pro because a Pro is not well accommodated by the PR3 S2¢ subsite [36]; and Ala for Glu at P3 because this improves interaction with Lys192 at the S3 subsite of CG. The resulting substrate (ABZ-GIEPFSDPMPEQ- EDDnp (substrate 13) fulfils most of the requirements for CG, as well as for chymase cleavage (i.e. a nega- tively-charged residue at P3 and P2¢, a Pro-Phe pair at P2–P1, and a Ser and a Pro at P1¢ and P3¢, respec- tively), and this represents one of the most sensitive substrates to have been reported for these two prote- ases (Table 2). Finally, substituting Phe for Lys in this optimized substrate [ABZ-GIEPKSDPMPEQ-EDDnp (substrate 14)] totally abolished cleavage by chymase, at the same time as maintaining specificity constant in the 10 5 m )1 Æs )1 range (i.e. sufficiently high to allow specific measurements of nanomolar concentrations of CG) (Table 2). As expected, HPLC analysis showed that CG cleaved this substrate at the K–S bond (Fig. 3A). Furthermore, this substrate was not hydro- lyzed by b2-tryptase (EC 3.4.21.59), despite the Lys at P1 that is a preferential cleavage site for trypsin-like proteases (Fig. 3B). This was a result of the presence of negatively-charged residues at P2 and P2¢ that are not accommodated within the b2-tryptase active site because of the presence of Asp147 and Asp143 within the S2 and the S2¢ subsites, respectively [40,41] (Fig. 2B). We cannot exclude the possibility, how- ever, that trypsin-like protease(s) other than tryptase are present in cells, tissues or biological fuids, such as lung secretions and skin exudates, where CG and chymase have been identified as critical pathophysio- logical actors. Trypsin-like activities, however, could be easily detected using broad spectrum inhibitors such as leupeptin or N-tosyl-l-lysine chloromethyl ke- tone that do not affect chymotrypsin-like proteases. Nevertheless, we used a lysate of cells from a mast cell line and also sputum from a patient with severe asthma to measure hydrolysis of the newly-described substrate. Measurement of CG activity in a mast cell line extract and in sputum Mast cells contain substantial amounts of a variety of proteases, including chymase, tryptase, carboxypepti- dase A3 and dipeptidyl peptidase I (cathepsin C), that participate in host defence and homeostasis [3]. The qualitative and quantitative importance of CG or a CG-like protease in mast cells and mast cell lines remains unclear because the substrate specificity of CG is close to that of chymase [42] and the corre- sponding mRNA has not been detected in the cell extracts [43]. We used a mast cell line (HMC-1) extract to measure CG activity using ABZ-GIEPKSDPM- PEQ-EDDnp (substrate 14) and evaluate its concentra- tion in comparison with that of chymase. Accordingly, we compared the rate of hydrolysis of the specific CG substrate and a CG ⁄ chymase substrate by the cell extract, as well as by purified CG and chymase. Opti- mized kinetic conditions were first determined to ensure that both substrates were cleaved at approxi- mately V max . We measured CG activity in the HMC-1 cell line, which confirms previous results obtained using a specific trypsin-like fluorophosphonate probe [44]. We ensured that the activity measured with ABZ-GIEPKSDPMPEQ-EDDnp was only a result of CG by adding the irreversible chloromethylketone inhibitor Z-GLF-CMK (where Z is benzyloxycarbonyl and CMK is chloromethyl ketone), which specifically targets chymotrypsin-like proteases. Full inhibition was obtained under these conditions, confirming Table 2. Specificity constant k cat ⁄ K m for the hydrolysis of the FRET substrates derived from serpinB1 by CG, chymase, HNE and PR3. Values (m M )1 Æs )1 ) are the means of ‡ 3 experiments. The error for k cat ⁄ K m is < 15%. NSH, no significant hydrolysis. Number Substrates derived from serpinB1 k cat ⁄ K m CG Chymase HNE PR3 6 ABZ-GIATFCDLMPEQ-EDDnp 817 1035 1100 2230 12 ABZ-GIAPFCDLMPEQ-EDDnp 2054 1963 < 1 488 13 ABZ-GIEPFSDPMPEQ-EDDnp 1700 1648 NSH NSH 14 ABZ-GIEPKSDPMPEQ-EDDnp 190 NSH NSH NSH Cathepsin G versus chymase specificity B. Korkmaz et al. 2640 FEBS Journal 278 (2011) 2635–2646 Journal compilation ª 2011 FEBS. No claim to original French government works the specific role of CG in cleavage (Fig. 4A). We checked that this inhibitor did not alter cleavage by the cell lysate of the trypsin-like substrate ABZ- TPRSALQ-EDDnp at the R–S bond (not shown). We also found that chymase activity was only twice as high as that of CG in HMC-1 cells, in accordance with preliminary observations made using MC TC mast cells [4]. We also measured the hydrolysis of ABZ-GIEP FSDPMPEQ-EDDnp (substrate 13) and ABZ-GIEP KSDPMPEQ-EDDnp (substrate 14) by a sample of whole sputum from a patient with severe asthma. Both substrates were rapidly cleaved at a single site identi- fied at the F-S bond and the K–S bond, respectively, by HPLC analysis (Fig. 4B). Cleavage was completely abolished after incubation with the chymotrypsin- like-specific Z-GLF-CMK inhibitor, which clearly demonstrates that no trypsin-like protease cleaved substrate 14 in the sputum (not shown). However, the resulting EDDnp-containing fragments from CG ABZ-GIEPKSDPMPEQ-EDDnp Fluorescence 10 12 14 16 18 20 22 24 0 50 100 150 200 250 10 12 14 16 18 20 22 24 0 100 200 300 400 500 600 2000 1000 0 0 200 400 600 800 1000 1200 1400 Elution time (min) Absorbance 220 nm 320 nm 360 nm SDPMPEQ-EDDnp SDPMPEQ-EDDnp ABZ-GIEPF ABZ-GIEPK Time (s) ABZ-GIEPKSDPMPEQ-Y Cathepsin G + ABZ-GIEPKSDPMPEQ-Y + Tryptase ABZ-TPKSALQ-EDDnp + Tryptase NO2 NO2 ABZ-GIEPFSDPMPEQ-EDDnp A B Fig. 3. Hydrolysis of ABZ-GIEPFSDPMPEQ- EDDnp and ABZ-GIEPKSDPMPEQ-EDDnp by CG. (A) Demonstration of identical cleav- age sites within the two substrates as visu- alized by reverse-phase HPLC and recording at 360 nm of the EDDnp-containing frag- ments. (B) Control experiment showing no cleavage of the Lys-containing CG substrate ABZ-GIEPKSDPMPEQ-Y NO2 (20 lM)by b2-tryptase (10 )7 M final concentration) but a rapid cleavage of ABZ-TPKSALQ-EDDnp (20 l M)by10 )9 M b2-tryptase. Hydrolysis of ABZ-GIEPKSDPMPEQ-Y NO2 (20 lM)by 10 )9 M CG is shown for comparison. Assays were carried out at 37 °Cin50m M Hepes buffer (pH 7.4), 100 m M NaCl, 0.01% Igepal CA-630 (v ⁄ v). B. Korkmaz et al. Cathepsin G versus chymase specificity FEBS Journal 278 (2011) 2635–2646 Journal compilation ª 2011 FEBS. No claim to original French government works 2641 hydrolysis (SDPMPEQ-EDDnp) were sequentially degraded in a time-dependent manner. This could be a result of the presence of amino peptidase activity(ies) in asthma sputum, although further work is required using larger numbers of sputum samples to confirm this hypothesis. Time (s) Elution time ( min ) FluorescenceAbsorbance ABZ-GIEPKSDPMPEQ-EDDnp (substrate 14) ABZ-GIEPKSDPMPEQ-EDDnp (substrate 14) ABZ-GIEPFSDPMPEQ-EDDnp (substrate 13) 0 200 400 600 800 1000 1200 1400 0 100 000 200 000 300 000 400 000 500 000 + HMC-1 cells lysate + Asthma sputum + Asthma sputum + [HMC-1 cells lysate + Z-GLF-CMK] Time (s) Fluorescence (10 x) 0 400 800 1200 1600 2000 600 500 400 300 200 100 0 + substrate 14 + ABZ-GIEPFSY Purified chymase -3 10 12 14 16 18 20 22 24 0 50 100 150 200 250 300 350 10 12 14 16 18 20 22 24 0 50 100 150 200 250 300 350 220 nm 320 nm 360 nm SDPMPEQ-EDDnp ABZ-GIEPF ABZ-GIEPK B A SDPMPEQ-EDDnp NO2 Fig. 4. Hydrolysis of the CG substrate by a cell line extract and by a biological sample. (A) Monitoring of ABZ-GIEPKSDPMPEQ-EDDnp hydrolysis by a HMC-1 mast cell lysate before and after incubation with the chymotrypsin-like protease inhibitor Z-GLF-CMK (3 m M final con- centration). The total inhibition observed in the presence of inhibitor indicates that the cleavage of the P1 Lys-containing substrate was a result of CG. The insert shows the peptidase activity of purified chymase on a polyvalent substrate and its inability to cleave substrate 14 under the same experimental conditions. (B) Hydrolysis of ABZ-GIEPFSDPMPEQ-EDDnp and ABZ-GIEPKSDPMPEQ-EDDnp by sputum from a patient with severe asthma as visualized by reverse-phase HPLC and recording at 360 nm for the EDDnp-containing fragments. Identical cleavage sites are observed within the two substrates but their cleavage was abolished after previous incubation with Z-GLF-CMK (not shown), indicating that only CG was involved in these cleavages. Further degradation of the EDDnp-containing fragment, most probably by aminopeptidase activity present in the sputum, is observed for both peptides. Assays were carried out at 37 °Cin50m M Hepes buffer (pH 7.4), 100 m M NaCl, 0.01% Igepal CA-630 (v ⁄ v). Cathepsin G versus chymase specificity B. Korkmaz et al. 2642 FEBS Journal 278 (2011) 2635–2646 Journal compilation ª 2011 FEBS. No claim to original French government works The reason why two closely-related proteases such as chymase and CG are co-stored within the same cell type remains unclear. Mast cells are involved in a variety of biological functions [45,46] and are mediated by a range of potent mediators and proteases of differ- ent specificities whose roles require clarification. Using a specific CG substrate such as that described in the present study should help to define the roles of these two proteases in diseases associated with mast cell activation and facilitate the development of specific inhibitors that could control their activity. Materials and methods Materials Purified CG (EC 3.4.21.20), HNE (EC 3.4.21.37) and ACT were obtained from Biocentrum (Krakow, Poland). Purified PR3 ( EC 3.4.21.76) and b2-tryptase (EC 3.4.21.59) were provided by Athens Research & Technology Inc. (Athens, GA, USA) and Merck (Nottingham, UK), respectively. Ige- pal CA-630 was obtained from Sigma (St Louis, MO, USA). Z-GLF-CMK was obtained from Enzyme System Products (Livermore, CA, USA). N,N-dimethylformamide and acetonitrile were obtained from Merck (Darmstad, Germany). Electrophoresis chemicals were obtained from Bio-Rad (Marnes-la-Coquette, France). All other chemical reagents were of analytical grade. Design and synthesis of quenched fluorescent substrates Quenched fluorogenic substrates were either obtained from Genecust-Europe (Dudelange, Luxembourg) or prepared by solid phase synthesis with Fmoc methodology [47]. Sub- strate purity was checked by MS (TofSpec-E; Micromass, Manchester, UK) and by reversed-phase chromatography on a C18 column. The purified ABZ-peptidyl-EDDnp con- centration was determined by measuring A 365 with e 365 = 17 300 m )1 Æcm )1 for EDDnp [where ABZ is O-am- inobenzoic acid and EDDnp is N-(2,4-dinitrophenyl)-ethy- lenediamine]. Stock substrate solutions (2–5 mm) were prepared in 30% (v ⁄ v) N,N-dimethylformamide and diluted to 0.5 with 50 mm Hepes buffer (pH 7.4). Enzyme assays HNE, PR3 and CG were titrated with a 1 -proteinase inhibi- tor, as described previously [48]. Recombinant chymase, produced and activated as described previously [34], was titrated with ACT, the titre of which had been determined by titration with CG. Assays were carried out at 37 °Cin 50 mm Hepes buffer (pH 7.4), 100 m m NaCl and 0.01% Igepal CA-630 (v ⁄ v) for CG; in 0.1 m Tris ⁄ HCl (pH 8.0) and 50 mm Hepes (pH 7.4) for chymase; and in 750 mm NaCl and 0.05% Igepal CA-630 (v ⁄ v) for HNE and PR3. The hydrolysis of ABZ-peptidyl-EDDnp substrates was monitored by measuring fluorescence at k ex = 320 nm and k ex = 420 nm in a Hitachi F-2000 spectrofluorometer (Hit- achi, Tokyo, Japan). Specificity constants (k cat ⁄ K m ) were determined under first-order conditions, using a substrate concentration far below the estimated K m as described pre- viously [31]. HMC-1 cells, kindly provided by Dr J. H. Butterfield (Mayo Clinic, Rochester, MN, USA) were cultured as described previously [42]. Suspensions of 30–60 million cells were lysed in 2 mL of NaCl ⁄ P i supplemented with 1% Ige- pal CA-630 (v ⁄ v). Proteolytic activity was measured at 37 °C using 50 lL of the cell lysates with ABZ-GIE- PFSDPMPEQ-EDDnp (25 lm) or ABZ-GIEPKSDPM- PEQ-EDDnp (25 lm) and 5 lL of cell lysate with ABZ-TPRSALQ-EDDnp (25 lm) in a total volume of 70 lL using a microplate fluorescence reader (Spectra Max Gemini; Molecular Devices, Sunnyvale, CA, USA) under continuous stirring. A sample of induced sputum from a patient with severe asthma was kindly provided by Dr Peter H. Howarth (University of Southampton, Southampton, UK). Written informed consent was obtained from the patient from whom the sputum sample was obtained. Chromatographic procedures and analysis of peptide products Once the enzyme–substrate reaction was complete, the reac- tion medium was incubated with four volumes of absolute ethanol for 15 min on ice and centrifuged at 13 000 g for 10 min. The supernatant containing the hydrolysis products was recovered, air-dried under vacuum and dissolved in 200 lL of 0.0075% trifluoroacetic acid (v ⁄ v). Hydrolysis fragments were fractionated by reversed-phase HPLC and eluted peaks were monitored at three wavelengths (220, 320 and 360 nm) simultaneously, which allowed direct identifi- cation of EDDnp-containing peptides before sequencing or MS analysis to identify cleavage sites. Nomenclature The nomenclature used for the individual amino acid resi- dues (e.g. P2, P1, P1¢,P2¢, etc.) of a substrate and corre- sponding residues of the enzyme subsites (e.g. S2, S1, S1¢, S2¢, etc.) follows that of Schechter and Berger [49]. Acknowledgements This work was supported by ‘Region Centre’ and the ‘Fonds Europe ´ en de De ´ veloppement Re ´ gional’ (Projet INFINHI) and Agence Nationale pour la Recherche (project ANR-07-PHYSIO-029-01). The authors thank B. Korkmaz et al. Cathepsin G versus chymase specificity FEBS Journal 278 (2011) 2635–2646 Journal compilation ª 2011 FEBS. No claim to original French government works 2643 Miche ` le Brillard-Bourdet for sequence analyses; Chris- tophe Epinette and Lise Vanderlynden for technical support; Dr Peter H. Howarth, University of South- ampton, for providing a sputum sample; and the ‘Plate-forme d’Analyse Inte ´ grative des Biomarqueurs’ for MALDI-TOF MS analyses. References 1 Hof P, Mayr I, Huber R, Korzus E, Potempa J, Travis J, Powers JC & Bode W (1996) The 1.8 A crystal structure of human cathepsin G in complex with Suc- Val-Pro-PheP-(OPh)2: a Janus-faced proteinase with two opposite specificities. EMBO J 15, 5481–5491. 2 Pereira PJ, Wang ZM, Rubin H, Huber R, Bode W, Schechter NM & Strobl S (1999) The 2.2 A crystal structure of human chymase in complex with succinyl- Ala-Ala-Pro-Phe-chloromethylketone: structural expla- nation for its dipeptidyl carboxypeptidase specificity. J Mol Biol 286, 163–173. 3 Trivedi NN & Caughey GH (2010) Mast cell peptidases: chameleons of innate immunity and host defense. Am J Respir Cell Mol Biol 42, 257–267. 4 Schechter NM, Irani AM, Sprows JL, Abernethy J, Wintroub B & Schwartz LB (1990) Identification of a cathepsin G-like proteinase in the MCTC type of human mast cell. J Immunol 145, 2652–2661. 5 Schechter NM, Choi JK, Slavin DA, Deresienski DT, Sayama S, Dong G, Lavker RM, Proud D & Lazarus GS (1986) Identification of a chymotrypsin-like protein- ase in human mast cells. J Immunol 137, 962–970. 6 Fellows E, Gil-Parrado S, Jenne DE & Kurschus FC (2007) Natural killer cell-derived human granzyme H induces an alternative, caspase-independent cell-death program. Blood 110, 544–552. 7 Salvesen G, Farley D, Shuman J, Przybyla A, Reilly C & Travis J (1987) Molecular cloning of human cathepsin G: structural similarity to mast cell and cytotoxic T lymphocyte proteinases. Biochemistry 26, 2289–2293. 8 Reilly CF, Tewksbury DA, Schechter NM & Travis J (1982) Rapid conversion of angiotensin I to angiotensin II by neutrophil and mast cell proteinases. J Biol Chem 257, 8619–8622. 9 Jahanyar J, Youker KA, Loebe M, Assad-Kottner C, Koerner MM, Torre-Amione G & Noon GP (2007) Mast cell-derived cathepsin g: a possible role in the adverse remodeling of the failing human heart. J Surg Res 140, 199–203. 10 Helske S, Syvaranta S, Kupari M, Lappalainen J, Laine M, Lommi J, Turto H, Mayranpaa M, Werkkala K, Kovanen PT et al. (2006) Possible role for mast cell- derived cathepsin G in the adverse remodelling of ste- notic aortic valves. Eur Heart J 27, 1495–1504. 11 Schiemann F, Grimm TA, Hoch J, Gross R, Lindner B, Petersen F, Bulfone-Paus S & Brandt E (2006) Mast cells and neutrophils proteolytically activate chemokine precursor CTAP-III and are subject to counterregula- tion by PF-4 through inhibition of chymase and cathep- sin G. Blood 107, 2234–2242. 12 Sommerhoff CP, Nadel JA, Basbaum CB & Caughey GH (1990) Neutrophil elastase and cathepsin G stimulate secretion from cultured bovine airway gland serous cells. J Clin Invest 85, 682–689. 13 Raymond WW, Cruz AC & Caughey GH (2006) Mast cell and neutrophil peptidases attack an inactivation segment in hepatocyte growth factor to generate NK4- like antagonists. J Biol Chem 281, 1489–1494. 14 Mayranpaa MI, Heikkila HM, Lindstedt KA, Walls AF & Kovanen PT (2006) Desquamation of human coronary artery endothelium by human mast cell prote- ases: implications for plaque erosion. Coron Artery Dis 17, 611–621. 15 Maryanoff BE, de Garavilla L, Greco MN, Haertlein BJ, Wells GI, Andrade-Gordon P & Abraham WM (2009) Dual inhibition of cathepsin G and chymase is effective in animal models of pulmonary inflammation. Am J Respir Crit Care Med 181, 247–253. 16 Pham CT (2006) Neutrophil serine proteases: specific regulators of inflammation. Nat Rev Immunol 6, 541– 550. 17 Owen CA & Campbell EJ (1999) The cell biology of leukocyte-mediated proteolysis. J Leukoc Biol 65 , 137–150. 18 Meyer-Hoffert U (2009) Neutrophil-derived serine pro- teases modulate innate immune responses. Front Biosci 14 , 3409–3418. 19 Korkmaz B, Horwitz M, Jenne DE & Gauthier A (2010) Neutrophil elastase, proteinase 3 and cathepsin G as therapeutic targets in human diseases. Pharmacol Rev 62, 726–759. 20 Carmona AK, Juliano MA & Juliano L (2009) The use of fluorescence resonance energy transfer (FRET) pep- tides for measurement of clinically important proteolytic enzymes. An Acad Bras Cienc 81, 381–392. 21 Korkmaz B, Attucci S, Juliano MA, Kalupov T, Jourdan ML, Juliano L & Gauthier F (2008) Measuring elastase, proteinase 3 and cathepsin G activities at the surface of human neutrophils with fluorescence resonance energy transfer substrates. Nat Protoc 3, 991–1000. 22 Korkmaz B, Moreau T & Gauthier F (2008) Neutrophil elastase, proteinase 3 and cathepsin G: physicochemical properties, activity and physiopathological functions. Biochimie 90, 227–242. 23 Lesner A, Wysocka M, Guzow K, Wiczk W, Legowska A & Rolka K (2008) Development of sensitive cathepsin G fluorogenic substrate using combinatorial chemistry methods. Anal Biochem 375, 306–312. Cathepsin G versus chymase specificity B. Korkmaz et al. 2644 FEBS Journal 278 (2011) 2635–2646 Journal compilation ª 2011 FEBS. No claim to original French government works [...]... fluoridetreated human chymase at 1.9 A Biochemistry 36, 14318–14324 28 Reilly CF, Schechter NB & Travis J (1985) Inactivation of bradykinin and kallidin by cathepsin G and mast cell chymase Biochem Biophys Res Commun 127, 443–449 29 Korkmaz B, Attucci S, Hazouard E, Ferrandiere M, Jourdan ML, Brillard-Bourdet M, Juliano L & Gauthier F (2002) Discriminating between the activities of human neutrophil elastase and. .. cathepsin G, and proteinase 3: evidence for distinct cleavage sites within the aminoterminal domain of the thrombin receptor Blood 89, 1944–1953 33 Bastos M, Maeji NJ & Abeles RH (1995) Inhibitors of human heart chymase based on a peptide library Proc Natl Acad Sci USA 92, 6738–6742 34 Andersson MK, Enoksson M, Gallwitz M & Hellman L (2009) The extended substrate specificity of the human mast cell chymase. .. activity-based probes for trypsin-family serine proteases Bioorg Med Chem Lett 16, 2882–2885 Bischoff SC (2007) Role of mast cells in allergic and non-allergic immune responses: comparison of human and murine data Nat Rev Immunol 7, 93–104 Bradding P, Walls AF & Holgate ST (2006) The role of the mast cell in the pathophysiology of asthma J Allergy Clin Immunol 117, 1277–1284 Hirata IY, Cezari MHS, Nakaie... claim to original French government works 2645 Cathepsin G versus chymase specificity fluorogenic protease substrates: solid phase synthesis and fluorescence spectroscopy of peptides containing ortho-aminobenzoyl ⁄ dinitrophenyl groups as donor-acceptor pairs Lett Pept Sci 1, 299–308 48 Serveau C, Moreau T, Zhou GX, ElMoujahed A, Chao J & Gauthier F (1992) Inhibition of rat tissue kallikrein gene family... protease with Cathepsin G versus chymase specificity 35 36 37 38 39 40 41 42 43 44 45 46 47 well-defined substrate recognition profile Int Immunol 21, 95–104 Andersson MK, Thorpe M & Hellman L (2010) Arg143 and Lys192 of the human mast cell chymase mediate the preference for acidic amino acids in position P2¢ of substrates FEBS J 277, 2255–2267 Korkmaz B, Attucci S, Moreau T, Godat E, Juliano L & Gauthier... Gauthier F (2004) Design and use of highly specific substrates of neutrophil elastase and proteinase 3 Am J Respir Cell Mol Biol 30, 801–807 Raymond WW, Trivedi NN, Makarova A, Ray M, Craik CS & Caughey GH (2010) How immune peptidases change specificity: cathepsin g gained tryptic function but lost efficiency during primate evolution J Immunol 185, 5360–5368 Segal DM, Powers JC, Cohen GH, Davies DR & Wilcox... Juliano L, Hazouard E, Girardin C, Brillard-Bourdet M, Rehault S, Anthonioz P & Gauthier F (2002) Measurement of free and membranebound cathepsin G in human neutrophils using new sensitive fluorogenic substrates Biochem J 366, 965–970 32 Renesto P, Si-Tahar M, Moniatte M, Balloy V, Van Dorsselaer A, Pidard D & Chignard M (1997) Specific inhibition of thrombin-induced cell activation by the neutrophil proteinases... Guyot N, Juliano L, Moreau T, Belaaouaj A & Gauthier F (2009) Structural characterization of mouse neutrophil serine proteases and identification of their substrate specificities: relevance to mouse models of human inflammatory diseases J Biol Chem 284, 34084–34091 ´ 25 Rehault S, Brillard-Bourdet M, Juliano MA, Juliano L, Gauthier F & Moreau T (1999) New, sensitive fluorogenic substrates for human cathepsin. .. neutrophil elastase and proteinase 3 using serpin-derived fluorogenic substrates J Biol Chem 277, 39074–39081 30 Korkmaz B, Hajjar E, Kalupov T, Reuter N, BrillardBourdet M, Moreau T, Juliano L & Gauthier F (2007) Influence of charge distribution at the active site surface on the substrate specificity of human neutrophil protease 3 and elastase A kinetic and molecular modeling analysis J Biol Chem 282, 1989–1997... for human cathepsin G based on the sequence of serpin-reactive site loops J Biol Chem 274, 13810–13817 26 Tanaka T, Minematsu Y, Reilly CF, Travis J & Powers JC (1985) Human leukocyte cathepsin G Subsite mapping with 4-nitroanilides, chemical modification, and effect of possible cofactors Biochemistry 24, 2040– 2047 27 McGrath ME, Mirzadegan T & Schmidt BF (1997) Crystal structure of phenylmethanesulfonyl . Discriminating between the activities of human cathepsin G and chymase using fluorogenic substrates Brice Korkmaz 1,2 , Gwenhael Je ´ got 1,2 , Laurie C. Lau 3 , Michael. G Chymase Ser40 Arg143 Arg41 Lys192 Lys192 Phe41 His57 Ile99 Ile99 Cys42 Ser40 His57 Arg143 Lys40 Cathepsin G Chymase S1’ S2’ S2’ A B S1’ Fig. 1. Structural differences between CG and chymase. (A) The solvent accessible surface based on the atom coordinates of CG (1CGH) [1] and. commonly used chromogenic and fluorogenic substrates of CG that are also cleaved by chymase [31]. As expected, k cat ⁄ K m of substrate 12 was increased significantly using CG and chymase, and was resistant