Granzyme M Is a Regulatory Protease That Inactivates Proteinase Inhibitor 9, an Endogenous Inhibitor of Granzyme B
2004; Elsevier BV; Volume: 279; Issue: 52 Linguagem: Inglês
10.1074/jbc.m411482200
ISSN1083-351X
AutoresSami Mahrus, Walter Kisiel, Charles S. Craik,
Tópico(s)Research on Leishmaniasis Studies
ResumoGranzyme M is a trypsin-fold serine protease that is specifically found in the granules of natural killer cells. This enzyme has been implicated recently in the induction of target cell death by cytotoxic lymphocytes, but unlike granzymes A and B, the molecular mechanism of action of granzyme M is unknown. We have characterized the extended substrate specificity of human granzyme M by using purified recombinant enzyme, several positional scanning libraries of coumarin substrates, and a panel of individual p-nitroanilide and coumarin substrates. In contrast to previous studies conducted using thiobenzyl ester substrates (Smyth, M. J., O'Connor, M. D., Trapani, J. A., Kershaw, M. H., and Brinkworth, R. I. (1996) J. Immunol. 156, 4174–4181), a strong preference for leucine at P1 over methionine was demonstrated. The extended substrate specificity was determined to be lysine = norleucine at P4, broad at P3, proline > alanine at P2, and leucine > norleucine > methionine at P1. The enzyme activity was found to be highly dependent on the length and sequence of substrates, indicative of a regulatory function for human granzyme M. Finally, the interaction between granzyme M and the serpins α1-antichymotrypsin, α1 -proteinase inhibitor, and proteinase inhibitor 9 was characterized by using a candidate-based approach to identify potential endogenous inhibitors. Proteinase inhibitor 9 was effectively hydrolyzed and inactivated by human granzyme M, raising the possibility that this orphan granzyme bypasses proteinase inhibitor 9 inhibition of granzyme B. Granzyme M is a trypsin-fold serine protease that is specifically found in the granules of natural killer cells. This enzyme has been implicated recently in the induction of target cell death by cytotoxic lymphocytes, but unlike granzymes A and B, the molecular mechanism of action of granzyme M is unknown. We have characterized the extended substrate specificity of human granzyme M by using purified recombinant enzyme, several positional scanning libraries of coumarin substrates, and a panel of individual p-nitroanilide and coumarin substrates. In contrast to previous studies conducted using thiobenzyl ester substrates (Smyth, M. J., O'Connor, M. D., Trapani, J. A., Kershaw, M. H., and Brinkworth, R. I. (1996) J. Immunol. 156, 4174–4181), a strong preference for leucine at P1 over methionine was demonstrated. The extended substrate specificity was determined to be lysine = norleucine at P4, broad at P3, proline > alanine at P2, and leucine > norleucine > methionine at P1. The enzyme activity was found to be highly dependent on the length and sequence of substrates, indicative of a regulatory function for human granzyme M. Finally, the interaction between granzyme M and the serpins α1-antichymotrypsin, α1 -proteinase inhibitor, and proteinase inhibitor 9 was characterized by using a candidate-based approach to identify potential endogenous inhibitors. Proteinase inhibitor 9 was effectively hydrolyzed and inactivated by human granzyme M, raising the possibility that this orphan granzyme bypasses proteinase inhibitor 9 inhibition of granzyme B. Cytotoxic lymphocytes, which include cytotoxic T cells and natural killer (NK) 1The abbreviations used are: NK, natural killer; ACT, α1-antichymotrypsin; DTT, dithiothreitol; PI, proteinase inhibitor; ACC, 7-amino-4-carbamoylmethylcoumarin; PNGase F, N-glycosidase F; AMC, 7-amino-4-methylcoumarin; pNA, p-nitroanilide; Suc, succinyl; MES, 2-(N-morpholino)ethanesulfonic acid; HPLC, high pressure liquid chromatography; SI, stoichiometry of inhibition; MALDI-TOF, matrix-assisted laser desorption ionization time-of-flight. 1The abbreviations used are: NK, natural killer; ACT, α1-antichymotrypsin; DTT, dithiothreitol; PI, proteinase inhibitor; ACC, 7-amino-4-carbamoylmethylcoumarin; PNGase F, N-glycosidase F; AMC, 7-amino-4-methylcoumarin; pNA, p-nitroanilide; Suc, succinyl; MES, 2-(N-morpholino)ethanesulfonic acid; HPLC, high pressure liquid chromatography; SI, stoichiometry of inhibition; MALDI-TOF, matrix-assisted laser desorption ionization time-of-flight. cells, recognize and kill host cells infected with intracellular pathogens such as viruses and certain types of bacteria. Death of target cells is predominantly mediated through granule exocytosis. In this process lysosome-like vesicles whose principal components are perforin and a family of serine proteases known as the granzymes are vectorially secreted toward the target cell. Perforin then facilitates entry of granzymes into the target cell, whereupon key protein substrates such as caspases are cleaved to induce death (2Russell J.H. Ley T.J. Annu. Rev. Immunol. 2002; 20: 323-370Crossref PubMed Scopus (806) Google Scholar). In addition to being important mediators of immunity, some granzymes may also represent potential drug targets for treatment of autoimmune disorders (3Tak P.P. Spaeny-Dekking L. Kraan M.C. Breedveld F.C. Froelich C.J. Hack C.E. Clin. Exp. Immunol. 1999; 116: 366-370Crossref PubMed Scopus (135) Google Scholar).The five known human granzymes have varied primary substrate specificities. Granzymes A and K cleave after basic residues (4Beresford P.J. Kam C.M. Powers J.C. Lieberman J. Proc. Natl. Acad. Sci. U. S. A. 1997; 94: 9285-9290Crossref PubMed Scopus (92) Google Scholar, 5Wilharm E. Parry M.A. Friebel R. Tschesche H. Matschiner G. Sommerhoff C.P. Jenne D.E. J. Biol. Chem. 1999; 274: 27331-27337Abstract Full Text Full Text PDF PubMed Scopus (65) Google Scholar); granzyme B cleaves after aspartic acid (6Poe M. Blake J.T. Boulton D.A. Gammon M. Sigal N.H. Wu J.K. Zweerink H.J. J. Biol. Chem. 1991; 266: 98-103Abstract Full Text PDF PubMed Google Scholar); granzyme H cleaves after aromatic residues (7Edwards K.M. Kam C.M. Powers J.C. Trapani J.A. J. Biol. Chem. 1999; 274: 30468-30473Abstract Full Text Full Text PDF PubMed Scopus (88) Google Scholar), and granzyme M cleaves after long aliphatic residues such as methionine, norleucine, and leucine (1Smyth M.J. O'Connor M.D. Trapani J.A. Kershaw M.H. Brinkworth R.I. J. Immunol. 1996; 156: 4174-4181PubMed Google Scholar). Granzyme M is unique among the family because it is specifically expressed in NK cells and thus may have evolved to serve a specialized function in innate immunity (8Sayers T.J. Brooks A.D. Ward J.M. Hoshino T. Bere W.E. Wiegand G.W. Kelly J.M. Smyth M.J. Kelley J.M. J. Immunol. 2001; 166: 765-771Crossref PubMed Scopus (83) Google Scholar). Several studies have demonstrated that granzymes A and B serve as important effectors of lymphocyte cytotoxicity by inducing nuclear and non-nuclear damage in target cells (2Russell J.H. Ley T.J. Annu. Rev. Immunol. 2002; 20: 323-370Crossref PubMed Scopus (806) Google Scholar). In contrast, far less is known about the orphan human granzymes H, K, and M. Evidence has been presented for granzyme K and a chymase-like serine protease from human NK cells, presumably granzyme H, contributing to induction of target cell damage (9Shi L. Kam C.M. Powers J.C. Aebersold R. Greenberg A.H. J. Exp. Med. 1992; 176: 1521-1529Crossref PubMed Scopus (419) Google Scholar, 10Woodard S.L. Jackson D.S. Abuelyaman A.S. Powers J.C. Winkler U. Hudig D. J. Immunol. 1994; 153: 5016-5025PubMed Google Scholar). It has also been demonstrated recently (11Kelly J.M. Waterhouse N.J. Cretney E. Browne K.A. Ellis S. Trapani J.A. Smyth M.J. J. Biol. Chem. 2004; 279: 22236-22242Abstract Full Text Full Text PDF PubMed Scopus (110) Google Scholar) that treatment of target cells with purified granzyme M and sublytic quantities of perforin leads to lysis.Toward gaining a better understanding of the molecular function of granzyme M and its NK cell specificity, biochemical characterization of this enzyme is presented here with respect to its substrate specificity and interaction with three serpin macromolecular inhibitors. To obtain protease that is free from contaminating proteolytic activity, human granzyme M was recombinantly expressed in the yeast Pichia pastoris and purified to homogeneity. The primary (P1) 2Nomenclature for the substrate amino acid preference is Pn, Pn-1,..., P2, P1, P1′, P2′,..., Pm-1′, Pm′, with amide bond hydrolysis occurring between P1 and P1′. The corresponding enzyme binding sites are denoted by Sn, Sn-1,..., S2, S1, S1′, S2′,..., Sm-1′, Sm′ (34Schechter I. Berger A. Biochem. Biophys. Res. Commun. 1968; 32: 898-902Crossref PubMed Scopus (337) Google Scholar). 2Nomenclature for the substrate amino acid preference is Pn, Pn-1,..., P2, P1, P1′, P2′,..., Pm-1′, Pm′, with amide bond hydrolysis occurring between P1 and P1′. The corresponding enzyme binding sites are denoted by Sn, Sn-1,..., S2, S1, S1′, S2′,..., Sm-1′, Sm′ (34Schechter I. Berger A. Biochem. Biophys. Res. Commun. 1968; 32: 898-902Crossref PubMed Scopus (337) Google Scholar). and extended (P4–P2) substrate specificity of the recombinant enzyme was first characterized by using positional scanning libraries of fluorogenic tetrapeptide coumarin substrates (12Harris J.L. Backes B.J. Leonetti F. Mahrus S. Ellman J.A. Craik C.S. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 7754-7759Crossref PubMed Scopus (469) Google Scholar). Single substrates were then used for validation of library results and a more detailed kinetic characterization. Finally, in an attempt to explore how the activity of human granzyme M is controlled under physiological conditions, interaction of the enzyme with the serpins α1-antichymotrypsin (ACT), α1-proteinase inhibitor (α1PI), and proteinase inhibitor 9 (PI9) was characterized. These three serpins are known to inhibit other leukocyte proteases that display hydrophobic primary specificity such as neutrophil elastase, cathepsin G, proteinase 3, and chymase (13Potempa J. Korzus E. Travis J. J. Biol. Chem. 1994; 269: 15957-15960Abstract Full Text PDF PubMed Google Scholar, 14Dahlen J.R. Foster D.C. Kisiel W. Biochim. Biophys. Acta. 1999; 1451: 233-241Crossref PubMed Scopus (20) Google Scholar) and are thus good candidate physiological inhibitors of human granzyme M.EXPERIMENTAL PROCEDURESMaterials—Unless otherwise stated, all chemicals were purchased from Sigma. Oligonucleotide primers were synthesized on an Applied Biosystems Expedite DNA synthesizer (Foster City, CA). Molecular weight markers for gel electrophoresis, restriction enzymes, and N-glycosidase F (PNGase F) were purchased from New England Biolabs (Beverly, MA) and used according to the manufacturer's instructions. Human granzyme B was a generous gift from Dr. Nancy Thornberry (Merck). Human granzyme A was recombinantly expressed and purified as described previously (15Bell J.K. Goetz D.H. Mahrus S. Harris J.L. Fletterick R.J. Craik C.S. Nat. Struct. Biol. 2003; 10: 527-534Crossref PubMed Scopus (57) Google Scholar). Human PI9 and anti-PI9 antibodies were prepared and purified as described previously (16Sprecher C.A. Morgenstern K.A. Mathewes S. Dahlen J.R. Schrader S.K. Foster D.C. Kisiel W. J. Biol. Chem. 1995; 270: 29854-29861Abstract Full Text Full Text PDF PubMed Scopus (76) Google Scholar, 17Annand R.R. Dahlen J.R. Sprecher C.A. De Dreu P. Foster D.C. Mankovich J.A. Talanian R.V. Kisiel W. Giegel D.A. Biochem. J. 1999; 342: 655-665Crossref PubMed Scopus (97) Google Scholar). Human ACT and α1PI were purchased from Calbiochem. The p-nitroanilide (pNA) substrates Suc-AAPL-pNA, Suc-AAPn-pNA, Suc-AAPM-pNA, and Suc-AAPK-pNA, and the 7-amino-4-methylcoumarin (AMC) substrates Ac-AAPA-AMC, Suc-AAPF-AMC, and Ac-IEPD-AMC were purchased from Bachem (Torrance, CA).Heterologous Expression of Human Granzyme M—The cDNA encoding mature human granzyme M was amplified from I.M.A.G.E. clone 112558 and subcloned into the yeast expression vector pPICZα A (Invitrogen). The resulting construct permitted the sequence of mature human granzyme M to immediately follow the Kex2 signal cleavage site of the Saccharomyces cerevisiae α-factor secretion signal. The vector was linearized with SacI and transformed into the X33 strain of Pichia pastoris. Clones with the integrated human granzyme M cDNA were selected by resistance to Zeocin™ (Invitrogen) and were used to inoculate 1-liter shaker flask cultures.Purification of Recombinant Human Granzyme M—After 3 days of induction with methanol, the conditioned media from the shaker flask cultures was isolated and loaded onto an SP-Sepharose cation exchange column (Amersham Biosciences). The column was washed with 4 column volumes of 50 mm MES, pH 6.0, 50 mm NaCl, and bound protein was eluted with 4 column volumes of 50 mm MES, pH 6.0, 1 m NaCl. Eluted protein was concentrated, exchanged into 50 mm MES, pH 6.0, 50 mm NaCl, and loaded onto a Mono-S cation exchange fast protein liquid chromatography column (Amersham Biosciences). The column was washed with 4 column volumes of 50 mm MES, pH 6.0, 50 mm NaCl, and bound protein was eluted with 50 column volumes of a linear gradient of 0.32 m NaCl to 0.48 m NaCl in 50 mm MES, pH 6.0. Fractions containing human granzyme M eluted between 0.38 and 0.45 m NaCl. These were pooled, concentrated, and exchanged into 50 mm MES, pH 6.0, 50 mm NaCl. Purity of the preparation was assessed by SDS-PAGE and Coomassie Brilliant Blue staining, followed by densitometric analysis with a MultiImage™ Light Cabinet (Alpha Innotech, San Leandro, CA).Preparation of Anti-human Granzyme M Antibodies—Surface loops of human granzyme M with low similarity to those of the other four human granzymes were identified, and the sequence analysis software package MacVector (Oxford Molecular, Madison, WI) was used to rank their antigenic potential. Peptides corresponding to the amino acid sequence of two loops were synthesized using Fmoc (N-(9-fluorenyl)methoxycarbonyl) chemistry on an Applied Biosystems automated peptide synthesizer (Foster City, CA), purified by reversed-phase HPLC, and characterized by MALDI-TOF mass spectrometry. Each peptide was conjugated to keyhole limpet hemocyanin and used to immunize rabbits for polyclonal antiserum production (Covance Corp., Richmond, CA). Antibodies raised against the peptide corresponding to residues 98–109 of human granzyme M demonstrated good reactivity and selectivity against the recombinant enzyme and were used for subsequent studies.Immunoblot Analysis—Protein samples were separated by SDS-PAGE and transferred to a nitrocellulose membrane (Schleicher & Schuell). The membrane was blocked in TBST (Tris-buffered saline with 0.1% Triton X-100) containing 5% nonfat dry milk, washed with TBST, incubated in a dilution of anti-human granzyme M antibodies or anti-human PI9 antibodies in TBST containing 5% nonfat dry milk, washed with TBST, incubated in a dilution of horseradish peroxidase-conjugated goat anti-rabbit antibody (Bio-Rad) in TBST containing 5% nonfat dry milk, and washed with TBST once again. Antibody-bound protein bands were then detected by enhanced chemiluminescence (Amersham Biosciences).Active Site Titration of Human Granzyme M and Other Proteases— Bovine trypsin was active site-titrated with 4-methylumbelliferyl p-guanidinobenzoate and was then used to determine the concentration of a solution of the macromolecular serine protease inhibitor ecotin using Suc-AAPK-pNA as the substrate. Ecotin was then used to active site-titrate human granzyme M and bovine chymotrypsin using Suc-AAPL-pNA as the substrate, human neutrophil elastase and porcine pancreatic elastase using Ac-AAPA-AMC as the substrate, and human neutrophil cathepsin G using Suc-AAPF-AMC as the substrate.Positional Scanning Synthetic Combinatorial Libraries—The preparation and characterization of the P1-diverse, P1-Leu, and P1-Met libraries of 7-amino-4-carbamoylmethylcoumarin (ACC) substrates used in this study are described elsewhere (12Harris J.L. Backes B.J. Leonetti F. Mahrus S. Ellman J.A. Craik C.S. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 7754-7759Crossref PubMed Scopus (469) Google Scholar, 18Dauber D.S. Ziermann R. Parkin N. Maly D.J. Mahrus S. Harris J.L. Ellman J.A. Petropoulos C. Craik C.S. J. Virol. 2002; 76: 1359-1368Crossref PubMed Scopus (53) Google Scholar). Screening of these libraries was also carried out as described previously. Briefly, ∼10-9 mol of each well of the P1-Leu or P1-Met stock libraries was added to 57 or 60 wells, respectively, of a 96-well Microfluor plate (Dynex Technologies, Chantilly, VA), and 10-10 mol of each well of the P1-diverse stock library was added to 20 wells of a 96-well Microfluor plate. Final concentration of each substrate in the assay was ∼0.1 μm for the P1-Leu and P1-Met libraries and ∼0.01 μm for the P1-diverse library. Assays were initiated by addition of ∼500 nm enzyme and were conducted at 30 °C in assay buffer containing 100 mm HEPES, pH 7.4, 200 mm NaCl, 0.01% Tween 20, and 1% Me2SO. Hydrolysis of substrates was monitored fluorimetrically with an excitation wavelength of 380 nm and an emission wavelength of 460 nm on a Spectramax Gemini microtiter plate reader (Molecular Devices, Sunnyvale, CA).Synthesis of Single Coumarin Substrates—Single ACC substrates were prepared as described previously (12Harris J.L. Backes B.J. Leonetti F. Mahrus S. Ellman J.A. Craik C.S. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 7754-7759Crossref PubMed Scopus (469) Google Scholar), purified by reversed-phase HPLC, and characterized by MALDI-TOF mass spectrometry.Single Substrate Kinetics—Enzyme activity was monitored at 25 °C in assay buffer containing 100 mm HEPES, pH 7.4, 200 mm NaCl, and 0.01% Tween 20. Human granzyme M concentration in assays ranged from 10 nm to 2 μm and substrate concentration ranged from 2 μm to 3 mm. Substrate stock solutions were prepared in Me2SO, and final Me2SO concentrations in assays never exceeded 2% because higher concentrations were found to be detrimental to enzyme activity. Hydrolysis of ACC substrates was monitored as described for library assays, and hydrolysis of pNA substrates was monitored spectrophotometrically at 405 nm on a Molecular Devices UVmax microtiter plate reader (Molecular Devices, Sunnyvale, CA).Analysis of Serpin-Granzyme M Complex Formation by SDS-PAGE— The concentration of human ACT and human α1PI were calculated using ϵ280 = 3.9 × 104 and 2.8 × 104m-1 cm-1, respectively. The concentration of human PI9 was determined by the Bradford assay (Bio-Rad). ACT, α1PI, or PI9 were incubated at 25 °C with 1 μm human granzyme M at a 2:1 serpin to protease molar ratio in 50 μl of assay buffer for 24 h. Assay buffer was the same as described above for substrate kinetics, but 1 mm DTT was added for reactions between PI9 and granzyme M to prevent inactivation of the serpin following oxidation of cysteine 342 in the reactive center loop. After the incubation, all samples were deglycosylated using PNGase F and analyzed by SDS-PAGE followed by Coomassie Brilliant Blue staining.Titration of Granzyme M with Serpins—200 nm human granzyme M was incubated at 25 °C with 0–200 nm ACT in assay buffer for 24 h, 0–200 nm α1PI in assay buffer for 96 h, and with 0–8 μm PI9 in assay buffer supplemented with 1 mm DTT for 24 h. The remaining enzymatic activity after these incubation times was determined using 100 μm Ac-KVPL-ACC. Substrate hydrolysis was monitored as described for library assays.Serpin Kinetics—The inhibition of human granzyme M by ACT, α1PI, and PI9 was characterized by monitoring the hydrolysis of 1 mm Ac-KVPL-ACC by 2 nm protease in the presence of varying serpin concentrations in assay buffer or assay buffer supplemented with 1 mm DTT for reactions between PI9 and granzyme M. Substrate hydrolysis was monitored as described for library assays over the course of 2 h. Serpin concentrations ranged from 0.48 to 7.6 μm for ACT, 0.45 to 7.2 μm for α1PI, and 2.3 to 37.2 μm for PI9. Data from substrate hydrolysis progress curves were then fit to Equation 1 by using the program KaleidaGraph™ (Synergy Software, Reading, PA), P=vs+(vi−vs)(1−e−k′t)k′ (Eq. 1) where P represents arbitrary fluorescence units at time t; vi is the initial velocity; vs is the steady-state velocity, and k′ is an apparent first-order rate constant (19Morrison J.F. Trends Biochem. Sci. 1982; 7: 102-105Abstract Full Text PDF Scopus (494) Google Scholar). Along with inhibitor concentrations, these apparent rate constants were then fit to Equation 2, k′=koff[1+1Ki(1+[S]/Km)] (Eq. 2) where k′ is the apparent first-order rate constant; [S] is the substrate concentration; Km is the Michaelis constant for interaction of the substrate with the protease, and Ki = koff/kon is the overall inhibition constant (19Morrison J.F. Trends Biochem. Sci. 1982; 7: 102-105Abstract Full Text PDF Scopus (494) Google Scholar). The second-order association rate constants kon = ka for inhibition of human granzyme M by serpins were derived from the Ki and koff values obtained from this secondary fit.PI9 Inactivation—All reactions were carried out at 25 °C in 10 μl of activity buffer supplemented with 1 mm DTT. 0.5 μm human granzyme B was incubated for 1 h with 1 μm PI9 or 1 μm PI9 that had been preincubated for 1 h at a concentration of 2 μm with 1 μm human granzyme M. These two samples and controls for each individual protein were then diluted hundredfold into water, and 7.5 μl of these diluted samples were used for SDS-PAGE and subsequent immunoblot analysis with anti-human PI9 antibodies. PI9 inactivation was also verified by using 200 μm Ac-IEPD-AMC to measure the relative activity of 10 nm human granzyme B following a 30-min incubation with 2 μm PI9, 2 μm PI9 and 10 nm human granzyme M, or 2 μm PI9 and 10 nm human granzyme M that had been preincubated for 30 min before addition of human granzyme B. Substrate hydrolysis was monitored as described for library assays in assay buffer supplemented with 1 mm DTT.RESULTSExpression and Purification of Human Granzyme M—The gene for mature human granzyme M was subcloned from I.M.A.G.E. clone 112558 into pPICzα A for expression in the methylotropic yeast P. pastoris. The nucleic acid sequence of this gene was found to be identical to that first reported for human granzyme M (20Smyth M.J. Sayers T.J. Wiltrout T. Powers J.C. Trapani J.A. J. Immunol. 1993; 151: 6195-6205PubMed Google Scholar). The protein was expressed as a C-terminal fusion to the S. cerevisiae α-factor signal sequence, allowing for purification of the mature enzyme from the media following secretion and cleavage of the signal sequence by Kex2. Following induction with methanol, purification of mature human granzyme M from conditioned media was carried out using SP-Sepharose and Mono-S cation exchange chromatography. Typical yields of purified protein were between 0.2 and 1.0 mg/liter of culture.SDS-PAGE followed by Coomassie Brilliant Blue staining indicated the recombinant enzyme was differentially glycosylated, migrating as a doublet between 33 and 48 kDa, and as a broad smear of hyperglycosylated enzyme between 48 and 83 kDa (Fig. 1A, lane 2). Glycosylation of human granzyme M in P. pastoris is expected because the enzyme contains three potential N-linked glycosylation consensus sites. Treatment with PNGase F caused all bands to decrease in apparent molecular weight and to coalesce at 26 kDa (Fig. 1A, lane 3). Densitometric analysis of this deglycosylated band indicated that the enzyme was greater than 98% pure. A polyclonal antibody raised against a peptide corresponding to residues 98–109 of human granzyme M recognized all glycosylated species and the single deglycosylated species of the recombinant enzyme (Fig. 1B, lanes 1 and 2).P1 Substrate Specificity—The primary specificity of purified recombinant human granzyme M was profiled using a positional scanning synthetic combinatorial library of tetrapeptide coumarin substrates referred to as the P1-diverse library (12Harris J.L. Backes B.J. Leonetti F. Mahrus S. Ellman J.A. Craik C.S. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 7754-7759Crossref PubMed Scopus (469) Google Scholar). As described previously, this library is of general structure Ac-XXXP1-ACC and is composed of 20 wells in which the P1 position is sequentially fixed as 1 of 20 different amino acids (excluding cysteine and including the methionine isostere norleucine), although P4, P3, and P2 are randomized in all wells, resulting in 8,000 substrates/well. Human granzyme M only displayed activity with P1-Leu, P1-Nle, and P1-Met in the P1-diverse library, in agreement with previous studies conducted with thiobenzyl ester substrates, purified native rat enzyme from RNK-16 cells (21Smyth M.J. Wiltrout T. Trapani J.A. Ottaway K.S. Sowder R. Henderson L.E. Kam C.M. Powers J.C. Young H.A. Sayers T.J. J. Biol. Chem. 1992; 267: 24418-24425Abstract Full Text PDF PubMed Google Scholar), and supernatant from COS cells transiently transfected with human and mouse granzyme M (1Smyth M.J. O'Connor M.D. Trapani J.A. Kershaw M.H. Brinkworth R.I. J. Immunol. 1996; 156: 4174-4181PubMed Google Scholar, 22Kelly J.M. O'Connor M.D. Hulett M.D. Thia K.Y. Smyth M.J. Immunogenetics. 1996; 44: 340-350Crossref PubMed Scopus (26) Google Scholar). In contrast to prior studies, human granzyme M displayed an extremely pronounced primary specificity for leucine, with relative activities of 100% for P1-Leu, 46% for P1-Nle, and 16% for P1-Met (Fig. 2A). When purified recombinant human granzyme M was assayed with p-nitroanilide substrates containing Leu, Nle, or Met at P1, activity was observed with all three substrates, but a clear preference for P1-Leu was apparent once again (Fig. 2B). In relative terms, the enzyme displayed 100% activity with Suc-AAPL-pNA, 52% activity with Suc-AAPn-pNA, and 37% activity with Suc-AAPM-pNA.Fig. 2P1 substrate specificity of human granzyme M. A, profile with the P1-diverse library, where the y axis represents activity relative to the P1 leucine well, and the x axis represents the positioned P1 amino acid (with norleucine represented by n). Data represents the average from three separate experiments. B, second-order rate constants for the hydrolysis of p-nitroanilide substrates with leucine, norleucine, or methionine at P1 by human granzyme M. Data represent the average, and error bars represent the S.D. from two separate experiments.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Extended Substrate Specificity—The P4-P2 substrate specificity of human granzyme M was profiled using two different positional scanning synthetic combinatorial libraries of tetrapeptide coumarin substrates referred to as the P1-Leu library and the P1-Met library (12Harris J.L. Backes B.J. Leonetti F. Mahrus S. Ellman J.A. Craik C.S. Proc. Natl. Acad. Sci. U. S. A. 2000; 97: 7754-7759Crossref PubMed Scopus (469) Google Scholar). The P1-Leu library is composed of three sub-libraries of general structures Ac-XXP2L-ACC, Ac-XP3XL-ACC, and Ac-P4XXL-ACC. Each of these is composed of 19 wells in which the P2, P3, or P4 position is sequentially fixed as 1 of 19 different amino acids (excluding cysteine and methionine and including the methionine isostere norleucine), whereas the remaining positions are randomized, resulting in 361 substrates/well. The P1-Met library is set up in an analogous manner, with the only difference being that, in contrast to the P1-Leu library, methionine is also present at randomized and fixed positions, resulting in 400 substrates/ well. Profiling of human granzyme M using the P1-Leu library revealed narrow specificity at P2, with a preference for proline over alanine, broad specificity at P3 with a preference for the aromatic residues phenylalanine and tyrosine, and intermediate specificity at P4 with a preference for lysine and the methionine isostere norleucine (Fig. 3). To validate these results and explore whether the specificity at P4-P2 is dependent on the identity of the P1 residue, human granzyme M was also profiled using the P1-Met library. Results at P2 and P4 with the P1-Met library were generally consistent with P1-Leu library results. Notably, specificity at P3 was different, with a preference for valine and glutamate instead of aromatic residues (Fig. 3).Fig. 3P4-P2 substrate specificity of human granzyme M. Profiles with the P1-Leu and P1-Met libraries, where the y axis represents activity relative to the P2 proline well, and the x axis represents the positioned P4, P3, or P2 amino acid (with norleucine represented by n). All rates can be normalized to the single well exhibiting highest activity in each library because a single experiment simultaneously collects data on all three subsites. Data represent the average from two separate experiments.View Large Image Figure ViewerDownload Hi-res image Download (PPT)Analysis of Substrate Specificity with Single Substrates— Because the positional scanning libraries described above are composed of pools of substrates, activities reflect the average preferred residues at a particular site. Quantitative kinetic analysis to address issues such as cooperativity can only be carried out with single, purified substrates. To validate the results obtained using the positional scanning libraries and to analyze the extended substrate specificity of human granzyme M in more detail, a series of single ACC substrates was prepared and assayed (Table I). Out of all tested permutations of residues found to be preferred based on the P1-Leu and P1-Met library profiles, the enzyme exhibited the highest preference for Ac-KVPL-ACC, with a kcat/Km of 1,900 m-1 s-1. In contrast, the enzyme showed no detectable activity with Ac-GRLL-ACC, a substrate with sub-optimal P4-P2 residues, indicating the importance of primary sequence recognition for human granzyme M. Shortening the optimal tetrapeptide su
Referência(s)