Selective Inhibition of Fibroblast Activation Protein Protease Based on Dipeptide Substrate Specificity
2006; Elsevier BV; Volume: 281; Issue: 11 Linguagem: Inglês
10.1074/jbc.m511112200
ISSN1083-351X
AutoresConrad Yap Edosada, Clifford Quan, Christian Wiesmann, Thuy Tran, Dan Sutherlin, Mark Reynolds, J. Michael Elliott, Helga Raab, Wayne J. Fairbrother, Beni B. Wolf,
Tópico(s)Neuropeptides and Animal Physiology
ResumoFibroblast activation protein (FAP) is a transmembrane serine peptidase that belongs to the prolyl peptidase family. FAP has been implicated in cancer; however, its specific role remains elusive because inhibitors that distinguish FAP from other prolyl peptidases like dipeptidyl peptidase-4 (DPP-4) have not been developed. To identify peptide motifs for FAP-selective inhibitor design, we used P2-Pro1 and acetyl (Ac)-P2-Pro1 dipeptide substrate libraries, where P2 was varied and substrate hydrolysis occurs between Pro1 and a fluorescent leaving group. With the P2-Pro1 library, FAP preferred Ile, Pro, or Arg at the P2 residue; however, DPP-4 showed broad reactivity against this library, precluding selectivity. By contrast, with the Ac-P2-Pro1 library, FAP cleaved only Ac-Gly-Pro, whereas DPP-4 showed little reactivity with all substrates. FAP also cleaved formyl-, benzyloxycarbonyl-, biotinyl-, and peptidyl-Gly-Pro substrates, which DPP-4 cleaved poorly, suggesting an N-acyl-Gly-Pro motif for inhibitor design. Therefore, we synthesized and tested the compound Ac-Gly-prolineboronic acid, which inhibited FAP with a Ki of 23 ± 3 nm. This was ∼9- to ∼5400-fold lower than the Ki values for other prolyl peptidases, including DPP-4, DPP-7, DPP-8, DPP-9, prolyl oligopeptidase, and acylpeptide hydrolase. These results identify Ac-Gly-BoroPro as a FAP-selective inhibitor and suggest that N-acyl-Gly-Pro-based inhibitors will allow testing of FAP as a therapeutic target. Fibroblast activation protein (FAP) is a transmembrane serine peptidase that belongs to the prolyl peptidase family. FAP has been implicated in cancer; however, its specific role remains elusive because inhibitors that distinguish FAP from other prolyl peptidases like dipeptidyl peptidase-4 (DPP-4) have not been developed. To identify peptide motifs for FAP-selective inhibitor design, we used P2-Pro1 and acetyl (Ac)-P2-Pro1 dipeptide substrate libraries, where P2 was varied and substrate hydrolysis occurs between Pro1 and a fluorescent leaving group. With the P2-Pro1 library, FAP preferred Ile, Pro, or Arg at the P2 residue; however, DPP-4 showed broad reactivity against this library, precluding selectivity. By contrast, with the Ac-P2-Pro1 library, FAP cleaved only Ac-Gly-Pro, whereas DPP-4 showed little reactivity with all substrates. FAP also cleaved formyl-, benzyloxycarbonyl-, biotinyl-, and peptidyl-Gly-Pro substrates, which DPP-4 cleaved poorly, suggesting an N-acyl-Gly-Pro motif for inhibitor design. Therefore, we synthesized and tested the compound Ac-Gly-prolineboronic acid, which inhibited FAP with a Ki of 23 ± 3 nm. This was ∼9- to ∼5400-fold lower than the Ki values for other prolyl peptidases, including DPP-4, DPP-7, DPP-8, DPP-9, prolyl oligopeptidase, and acylpeptide hydrolase. These results identify Ac-Gly-BoroPro as a FAP-selective inhibitor and suggest that N-acyl-Gly-Pro-based inhibitors will allow testing of FAP as a therapeutic target. Tumor-associated stromal cells can promote epithelial tumorigenesis (1Bhowmick N.A. Moses H.L. Curr. Opin. Genet. Dev. 2005; 15: 97-101Crossref PubMed Scopus (373) Google Scholar, 2Joyce J.A. Cancer Cell. 2005; 7: 513-520Abstract Full Text Full Text PDF PubMed Scopus (469) Google Scholar), suggesting that stromal proteins may represent novel therapeutic targets. Fibroblast activation protein (FAP), 2The abbreviations used are: FAP, fibroblast activation protein; Ac, acetyl; AFC, 7-amino-4-trifluoromethylcoumarin; AMC, 7-amino-4-methylcoumarin; AMCC, 7-amino-4-methyl-3-carbamoylmethylcoumarin; BoroPro, prolineboronic acid; APH, acylpeptide hydrolase; DPP, dipeptidyl peptidase; Z, benzyloxycarbonyl. a transmembrane serine peptidase, is one potential target because it is highly expressed by stromal fibroblasts in most epithelial cancers (3Garin-Chesa P. Old L.J. Rettig W.J. Proc. Natl. Acad. Sci. U. S. A. 1990; 87: 7235-7239Crossref PubMed Scopus (534) Google Scholar, 4Rettig W.J. Garin-Chesa P. Healey J.H. Su S.L. Ozer H.L. Schwab M. Albino A.P. Old L.J. Cancer Res. 1993; 53: 3327-3335PubMed Google Scholar, 5Scanlan M.J. Raj B.K. Calvo B. Garin-Chesa P. Sanz-Moncasi M.P. Healey J.H. Old L.J. Rettig W.J. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 5657-5661Crossref PubMed Scopus (460) Google Scholar, 6Niedermeyer J. Scanlan M.J. Garin-Chesa P. Daiber C. Fiebig H.H. Old L.J. Rettig W.J. Schnapp A. Int. J. Cancer. 1997; 71: 383-389Crossref PubMed Scopus (70) Google Scholar, 7Dolznig H. Schweifer N. Puri C. Kraut N. Rettig W.J. Kerjaschki D. Garin-Chesa P. Cancer Immun. 2005; 5: 1-10Google Scholar). Increased FAP expression in tumors correlates with increased FAP activity relative to normal tissues (8Park J.E. Lenter M.C. Zimmermann R.N. Garin-Chesa P. Old L.J. Rettig W.J. J. Biol. Chem. 1999; 274: 36505-36512Abstract Full Text Full Text PDF PubMed Scopus (441) Google Scholar, 9Huber M.A. Kraut N. Park J.E. Schubert R.D. Rettig W.J. Peter R.U. Garin-Chesa P. J. Investig. Dermatol. 2003; 120Abstract Full Text Full Text PDF PubMed Scopus (133) Google Scholar), and FAP overexpression promotes tumorigenesis in xenograft models (10Cheng J.D. Dunbrack R.L. Valianou M. Rogarto A. Aplaugh R.K. Weiner L.M. Cancer Res. 2002; 62: 4767-4772PubMed Google Scholar, 11Huang Y. Wang S. Kelly T. Cancer Res. 2004; 64: 2712-2716Crossref PubMed Scopus (97) Google Scholar, 12Cheng J.D. Valianou M. Canutescu A.A. Jaffe E.K. Lee H.-O. Wang H. Lai J.H. Bachovchin W.W. Weiner L.M. Mol. Cancer Ther. 2005; 4: 351-360Crossref PubMed Scopus (134) Google Scholar). This effect requires catalytically active FAP (12Cheng J.D. Valianou M. Canutescu A.A. Jaffe E.K. Lee H.-O. Wang H. Lai J.H. Bachovchin W.W. Weiner L.M. Mol. Cancer Ther. 2005; 4: 351-360Crossref PubMed Scopus (134) Google Scholar), suggesting that FAP activity promotes tumor growth and that FAP inhibition may have therapeutic value. FAP belongs to the prolyl peptidase family, which comprises serine proteases that typically cleave peptide substrates after a proline residue. This family has been implicated in several diseases, including diabetes, cancer, and mood disorders (13Polgar L. Cell. Mol. Life Sci. 2002; 59: 349-362Crossref PubMed Scopus (274) Google Scholar, 14Rosenblum J.S. Kozarich J.W. Curr. Opin. Chem. Biol. 2003; 7: 496-504Crossref PubMed Scopus (268) Google Scholar), and includes dipeptidyl peptidase-4 (DPP-4), DPP-7, DPP-8, DPP-9, prolyl oligopeptidase, acylpeptide hydrolase, and prolyl carboxypeptidase. These proteases differ in structure at the N terminus, but each has a C-terminal αβ-hydrolase domain that contains the catalytic Ser, Asp, and His residues. FAP, like its most closely related family member, DPP-4, is a type II transmembrane protein; both have a short cytoplasmic tail, a transmembrane domain, and a β-propeller domain containing several sites of N-linked glycosylation (5Scanlan M.J. Raj B.K. Calvo B. Garin-Chesa P. Sanz-Moncasi M.P. Healey J.H. Old L.J. Rettig W.J. Proc. Natl. Acad. Sci. U. S. A. 1994; 91: 5657-5661Crossref PubMed Scopus (460) Google Scholar, 15Aertgeerts K. Levin I. Shi L. Snell G.P. Jennings A. Prasad G.S. Zhang Y. Kraus M.L. Salakian S. Sridhar V. Wijnands R. Tennant M.G. J. Biol. Chem. 2005; 280: 19441-19444Abstract Full Text Full Text PDF PubMed Scopus (202) Google Scholar, 16Rasmussen H.B. Branner S. Wiberg F.C. Wagtmann N. Nat. Struct. Biol. 2003; 10: 19-25Crossref PubMed Scopus (372) Google Scholar, 17Thoma R. Loffler B. Stihle M. Huber W. Ruf A. Henning M. Structure. 2003; 11: 947-959Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar, 18Engel M. Hoffmann T. Wagner L. Wermann M. Heiser U. Kiefersauer R. Huber R. Bode W. Demuth H.-U. Brandstetter H. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 5063-5068Crossref PubMed Scopus (292) Google Scholar, 19Aertgeerts K. Ye S. Tennant M.G. Kraus M.L. Rogers J. Sang B.-C. Skene R.J. Webb D.R. Prasad G.S. Protein Sci. 2004; 13: 1-10Crossref PubMed Scopus (2) Google Scholar, 20Aertgeerts K. Ye S. Shi L. Prasad G.S. Witmer D. Chi E. Sang B.-C. Wijnands R. Webb D.R. Swanson R.V. Protein Sci. 2004; 13: 145-154Crossref PubMed Scopus (61) Google Scholar). Crystallographic data for FAP and DPP-4 show that the β-propeller has important substrate binding sites and suggest that this domain precludes access of large substrates to the αβ-hydrolase domain (15Aertgeerts K. Levin I. Shi L. Snell G.P. Jennings A. Prasad G.S. Zhang Y. Kraus M.L. Salakian S. Sridhar V. Wijnands R. Tennant M.G. J. Biol. Chem. 2005; 280: 19441-19444Abstract Full Text Full Text PDF PubMed Scopus (202) Google Scholar, 16Rasmussen H.B. Branner S. Wiberg F.C. Wagtmann N. Nat. Struct. Biol. 2003; 10: 19-25Crossref PubMed Scopus (372) Google Scholar, 17Thoma R. Loffler B. Stihle M. Huber W. Ruf A. Henning M. Structure. 2003; 11: 947-959Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar, 18Engel M. Hoffmann T. Wagner L. Wermann M. Heiser U. Kiefersauer R. Huber R. Bode W. Demuth H.-U. Brandstetter H. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 5063-5068Crossref PubMed Scopus (292) Google Scholar, 19Aertgeerts K. Ye S. Tennant M.G. Kraus M.L. Rogers J. Sang B.-C. Skene R.J. Webb D.R. Prasad G.S. Protein Sci. 2004; 13: 1-10Crossref PubMed Scopus (2) Google Scholar, 20Aertgeerts K. Ye S. Shi L. Prasad G.S. Witmer D. Chi E. Sang B.-C. Wijnands R. Webb D.R. Swanson R.V. Protein Sci. 2004; 13: 145-154Crossref PubMed Scopus (61) Google Scholar, 21Hiramatsu H. Yamamotoa A. Higashiyama Y. Fukushima C. Shima H. Sugiyama S. Inaka K. Shimizu R. Biol. Chem. 2004; 385: 561-564Crossref PubMed Scopus (61) Google Scholar). DPP-4 regulates biological processes by cleaving regulatory peptides of < ∼10 kDa, including glucagon-like peptide-1, glucose-dependent insulinotropic peptide, and stromal-derived factor-1 (22Mentlein R. Regul. Pept. 1999; 85: 9-24Crossref PubMed Scopus (1157) Google Scholar, 23Augustyns K. Van der Veken P. Senten K. Haemers A. Curr. Med. Chem. 2005; 12: 971-998Crossref PubMed Scopus (80) Google Scholar, 24Zhu L. Tamvakopoulos C. Xie D. Dragovic J. Shen X. Fenyk-Melody J. Schmidt K. Bagchi A. Griffin P. Thornberry N. Roy R. J. Biol. Chem. 2003; 278: 22418-22423Abstract Full Text Full Text PDF PubMed Scopus (213) Google Scholar, 25Lambeir A.-M. Proost P. Durinx C. Bal G. Senten K. Augustyns K. Scharpe S. Damme J. De Meester I. J. Biol. Chem. 2001; 276: 29839-29845Abstract Full Text Full Text PDF PubMed Scopus (241) Google Scholar). DPP-4 cleaves these peptides via a well characterized dipeptidyl peptidase activity that removes P2-Pro1 or P2-Ala1 dipeptides, (P2 represents any amino acid) from the N terminus of the substrate. DPP-4 similarly cleaves P2-Pro1-based synthetic peptides and does so with high catalytic efficiency and broad specificity (25Lambeir A.-M. Proost P. Durinx C. Bal G. Senten K. Augustyns K. Scharpe S. Damme J. De Meester I. J. Biol. Chem. 2001; 276: 29839-29845Abstract Full Text Full Text PDF PubMed Scopus (241) Google Scholar, 26Leitig B. Pryor K. Wu J. Marsilio F. Patel R. Craik C. Ellman J. Cummings R. Thornberry N. Biochem. J. 2003; 371: 525-532Crossref PubMed Google Scholar). In contrast, endogenous peptide substrates of FAP are not known, and the activity of the protease against synthetic substrates remains poorly characterized. FAP also cleaves proteins such as gelatin (8Park J.E. Lenter M.C. Zimmermann R.N. Garin-Chesa P. Old L.J. Rettig W.J. J. Biol. Chem. 1999; 274: 36505-36512Abstract Full Text Full Text PDF PubMed Scopus (441) Google Scholar, 27Ghersi G. Dong H. Goldstein L. Yeh Y. Hakkinen L. Larjava H. Chen W.-T. J. Biol. Chem. 2002; 277: 29231-29241Abstract Full Text Full Text PDF PubMed Scopus (141) Google Scholar) and α2-antiplasmin (28Lee K. Jackson K. Christiansen V. Chung K. McKee P. Blood. 2004; 103: 3783-3788Crossref PubMed Scopus (113) Google Scholar), suggesting that the two proteases cleave distinct substrates. The dipeptidase inhibitor Val-BoroPro shows efficacy in tumor models (12Cheng J.D. Valianou M. Canutescu A.A. Jaffe E.K. Lee H.-O. Wang H. Lai J.H. Bachovchin W.W. Weiner L.M. Mol. Cancer Ther. 2005; 4: 351-360Crossref PubMed Scopus (134) Google Scholar, 29Adams S. Miller G. Jesson M. Watanabe T. Jones B. Wallner B. Cancer Res. 2004; 64: 5471-5480Crossref PubMed Scopus (155) Google Scholar) that correlates with FAP inhibition (12Cheng J.D. Valianou M. Canutescu A.A. Jaffe E.K. Lee H.-O. Wang H. Lai J.H. Bachovchin W.W. Weiner L.M. Mol. Cancer Ther. 2005; 4: 351-360Crossref PubMed Scopus (134) Google Scholar). However, because Val-BoroPro also inhibits DPP-4, -7, -8, and -9 (30Flentke G. Munoz E. Huber B. Plau A. Kettner C. Bachovchin W.W. Proc. Natl. Acad. Sci. U. S. A. 1991; 88: 1556-1559Crossref PubMed Scopus (185) Google Scholar, 31Coutts S. Kelly T. Snow R. Kennedy C. Barton R. Adams J. Krolikowski D. Freeman D. Campbell S. Ksiazek J. Bachovchin W.W. J. Med. Chem. 1996; 39: 2087-2094Crossref PubMed Scopus (127) Google Scholar, 32Shreder K. Wong M. Corral S. Yu Z. Winn D. Wu M. Hu Y. Nomanbhoy T. Alemayehu S. Fuller S. Rosenblum J.S. Kozarich J.W. Bioorg. Med. Chem. Lett. 2005; 15: 4256-4260Crossref PubMed Scopus (16) Google Scholar), its mechanism of action remains unclear. To overcome this lack of selectivity, we sought here to define the substrate specificity of FAP with the goal of identifying a FAP-selective motif for inhibitor design. We identified N-acyl-Gly-Pro dipeptides as FAP-selective substrate motifs and synthesized a representative boronic acid inhibitor, Ac-Gly-BoroPro. This inhibitor showed ∼9- to ∼5400-fold selectivity for FAP inhibition relative to other prolyl peptidases, suggesting that N-acyl-Gly-Pro-based inhibitors will aid in testing whether FAP is a valid therapeutic target. Materials—Ala-Pro-7-amino-4-trifluoromethylcoumarin (AFC), Phe-Pro-AFC, Gly-Pro-AFC, Ile-Pro-AFC, acetyl (Ac)-Gly-Pro-AFC, and Lys-Ala-AFC were from Enzyme Systems Products. Benzyloxycarbonyl (Z)-Gly-Pro-7-amino-4-methylcoumarin (AMC), Gly-Ala-AMC, Ac-Ala-AMC, and amino acid derivatives were from Bachem. An Ac-P2-Pro1-AFC substrate library, where P2 was varied with all amino acids (except Cys and Trp), was custom synthesized by Enzyme Systems Products. N-substituted-Gly-Pro-7-amino-4-methyl-3-carbamoylcoumarin (AMCC) substrates and a P2-Pro1-AMCC substrate library were prepared essentially as described by Maly et al. (33Maly D. Leonetti F. Backes B. Dauber D. Harris J. Craik C. Ellman J. J. Org. Chem. 2020; 67: 910-915Crossref Scopus (127) Google Scholar) with the exception that 7-amino-4-methyl-3-coumarinylacetic acid was used as the fluorophore. Acetylated substrates were prepared by treating peptides on resin with acetic anhydride in 10% triethylamine/dichloromethane until the resin was negative to the Kaiser ninhydrin test (34Sarin V. Kent S. Tam J. Merrifield R. Anal. Biochem. 1981; 117: 147-157Crossref PubMed Scopus (999) Google Scholar). Formylated substrates were prepared as described (35Fields G. Fields C. Petefish J. Van Wart H. Cross T. Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 1384-1388Crossref PubMed Scopus (61) Google Scholar). Ac-Gly-Proline boronic acid (BoroPro) was synthesized as described by Gibson et al. (36Gibson F. Singh A. Soumeillant M. Manchand P. Humora M. Kronenthal D. Org. Proc. Res. Dev. 2002; 6: 814-816Crossref Scopus (20) Google Scholar) except that Ac-Gly was substituted for Val, making deprotection unnecessary. Substrates and inhibitors were purified by reverse phase chromatography and verified by matrix-assisted laser-desorption ionization mass spectrometry. N-glycanase was from Sigma. Protease Cloning and Expression—cDNAs encoding the extracellular domains of FAP (amino acids 38–760) and DPP-4 (amino acids 39–766) were generated by polymerase chain reaction (PCR) using Quick Clone cDNA library (Stratagene) as a template. PCR products were TA-cloned into pGemT (Promega) and confirmed by DNA sequencing. Confirmed cDNAs were then subcloned into pFLAG-CMV1 (Sigma) for expression as N-terminal FLAG-tagged proteins. Plasmids containing full-length DPP-7, DPP-8, DPP-9, prolyl oligopeptidase, and acylpeptide hydrolase were obtained from Origene and used as templates to generate pFLAG-CMV1 expression constructs encoding each protease as above. These constructs encoded amino acids 26–492 of DPP-7, 2–883 of DPP-8, 2–864 of DPP-9, 2–710 of prolyl oligopeptidase, and 2–732 of acylpeptide hydrolase. For protein production, we transfected 293 cells with plasmids encoding proteases using calcium phosphate and purified proteins from serum-free conditioned medium by affinity chromatography with M2-anti-FLAG resin (Sigma). Proteins were >95% pure as determined by SDS-PAGE with Coomassie Blue staining, with the exception of DPP8, which had a purity of ∼70%. Protein concentrations were determined by the bichinconic acid method (Bio-Rad). Gel Filtration Chromatography and Light Scattering Analysis—To calculate the stoichiometry of purified FAP and DPP-4, we measured the molecular mass of each protease in solution using multiangle light scattering in combination with gel filtration chromatography and interferometric refractometry. This method allows accurate determination of the molecular mass of a protein based on protein concentration, refractive index, and the degree of light scattering (37Mogridge J. Methods Mol. Biol. 2004; 261: 113-118PubMed Google Scholar). Proteases (50 μg) in 50 mm Tris (pH 7.4), and 100 mm NaCl were loaded onto a Shodex KW-803 gel filtration column (flow rate 0.5 ml/min). The column was developed using an Agilent 1100 HPLC system with a MiniDawn 3-angle light scattering detector (Wyatt Technology) and an OPTILAB DSP interferometric refractometer (Wyatt Technology) in line. ASTRA software was used for molecular mass calculations. Protease Assays—Protease activity was monitored continuously using a SpectraMax M2 microplate reader (Molecular Devices) in the kinetic mode. Assays were conducted at 23 °C in 50 mm Tris (pH 7.4), 100 mm NaCl, 1 mm EDTA. The excitation/emission wavelengths were 360/460 nm for AMC/AMCC substrates and 400/505 nm for AFC substrates. Standard curves of the appropriate fluorescent product versus concentration were used to convert relative fluorescence units to moles of product produced. Generally, kinetic constants (kcat, Km) were determined with initial rate (v0) measurements, using substrate concentrations in the range of 0.1–5 Km and protease concentrations as indicated in the figure legends. Kinetic parameters were calculated from Michaelis-Menten plots (v0 versus [S]) with nonlinear regression analysis using GraphPad software. kcat values were calculated assuming that each protease was 100% active. When saturating amounts of substrate could not be achieved, catalytic efficiencies (kcat/Km) were determined under pseudo-first order conditions ([S] <<the estimated Km) and fit to the following equation: Ln[St/S0] =-kobst, where St is the concentration of substrate remaining at time t, S0 is the initial substrate concentration, and kobs is the apparent first order substrate cleavage constant equal to (kcat/Km) × ET, the total enzyme concentration (38Copeland R. Enzymes, A Practical Introduction to Structure, Mechanism, and Data Analysis. 2nd. John Wiley and Sons, Inc., New York2000: 136Google Scholar). The linear relationship of Ln[St/S0] versus time allows calculation of kcat/Km by dividing the slope of the plot by ET. Inhibition Kinetics—Ki values for inhibition of proteases by Ac-Gly-BoroPro were determined using the method of progress curves for analysis of tight binding competitive inhibitors (39Henderson P.J. Biochem. J. 1972; 127: 321-333Crossref PubMed Scopus (479) Google Scholar, 40Nicklin M. Barrett A. Biochem. J. 1984; 223: 245-253Crossref PubMed Scopus (142) Google Scholar). Briefly, proteases were added to a reaction mixture containing inhibitor and substrate (Ac-Ala-AMC for acylpeptide hydrolase, Z-Gly-Pro-AMC for prolyl oligopeptidase, and Ala-Pro-AFC for all others) in assay buffer at 23 °C. Protease activity was followed continuously as described above to monitor time-dependent inhibition. Reactions contained inhibitor concentrations at least 20-fold greater than protease concentrations, such that the protease-inhibitor complex does not significantly deplete the free inhibitor. Data were plotted as v0/vi - 1 versus [I], where v0 is the rate of substrate hydrolysis in the absence of inhibitor, vi is the steady state rate of substrate hydrolysis in the presence of inhibitor, and [I] is the concentration of Ac-Gly-BoroPro. Plots of v0/vi - 1 versus [I] were linear, and the apparent inhibition constant, Kapp, was determined from the reciprocal of the slope. Ki, the true equilibrium inhibition constant, was determined according to the following relationship: Ki = Kapp/(1+ [S]/Km), where [S] is the concentration of substrate used in the assay and Km is the Michaelis constant for substrate cleavage. Protease Expression and Characterization—DPP-4 exists in serum as a soluble glycoprotein beginning at residue 39 (41Durinx C. Lambeir A.-M. Bosmans E. Falmagne J.-B. Berghmans R. Haemers A. Scharpe S. De Meester I. Eur. J. Biochem. 2000; 267: 5608-5613Crossref PubMed Scopus (230) Google Scholar). We therefore expressed and purified recombinant Ser-39-DPP-4 and an analogous soluble FAP molecule beginning at amino acid Thr-38. When analyzed by SDS-PAGE under reducing conditions, FAP migrated with an apparent molecular mass of 97 kDa (Fig. 1A), whereas DPP-4 migrated with a molecular mass of 105–115 kDa (Fig. 1B). These molecular masses are 15–20 kDa greater than expected based on primary amino acid sequence and decreased upon treatment with N-glycanase (not shown), indicating that each protease is N-glycosylated. To further characterize each protease, we determined molecular mass in solution using multiangle light scattering in combination with gel filtration chromatography and interferometric refractometry. This analysis showed that FAP exists predominantly as a dimer with a molecular mass of 200 ± 15 kDa (Fig. 1A). Small amounts of monomeric (elution volume 9.0 ml) and multimeric (elution volume <8.0 ml) FAP were also observed. The predominant elution peak of DPP-4 had a molecular mass of 220 ± 15 kDa (Fig. 1B), indicating a dimeric composition. The dimeric nature of our soluble protease preparations is consistent with the dimeric composition of FAP and DPP-4 crystal structures (15Aertgeerts K. Levin I. Shi L. Snell G.P. Jennings A. Prasad G.S. Zhang Y. Kraus M.L. Salakian S. Sridhar V. Wijnands R. Tennant M.G. J. Biol. Chem. 2005; 280: 19441-19444Abstract Full Text Full Text PDF PubMed Scopus (202) Google Scholar, 16Rasmussen H.B. Branner S. Wiberg F.C. Wagtmann N. Nat. Struct. Biol. 2003; 10: 19-25Crossref PubMed Scopus (372) Google Scholar, 17Thoma R. Loffler B. Stihle M. Huber W. Ruf A. Henning M. Structure. 2003; 11: 947-959Abstract Full Text Full Text PDF PubMed Scopus (214) Google Scholar, 18Engel M. Hoffmann T. Wagner L. Wermann M. Heiser U. Kiefersauer R. Huber R. Bode W. Demuth H.-U. Brandstetter H. Proc. Natl. Acad. Sci. U. S. A. 2003; 100: 5063-5068Crossref PubMed Scopus (292) Google Scholar, 19Aertgeerts K. Ye S. Tennant M.G. Kraus M.L. Rogers J. Sang B.-C. Skene R.J. Webb D.R. Prasad G.S. Protein Sci. 2004; 13: 1-10Crossref PubMed Scopus (2) Google Scholar), suggesting that they are structurally intact. Dipeptide Substrate Specificity—Although FAP cleaves certain proline-containing DPP-4 substrates (8Park J.E. Lenter M.C. Zimmermann R.N. Garin-Chesa P. Old L.J. Rettig W.J. J. Biol. Chem. 1999; 274: 36505-36512Abstract Full Text Full Text PDF PubMed Scopus (441) Google Scholar, 15Aertgeerts K. Levin I. Shi L. Snell G.P. Jennings A. Prasad G.S. Zhang Y. Kraus M.L. Salakian S. Sridhar V. Wijnands R. Tennant M.G. J. Biol. Chem. 2005; 280: 19441-19444Abstract Full Text Full Text PDF PubMed Scopus (202) Google Scholar), the full spectrum of dipeptidase specificity of FAP remains undefined. Therefore, to better understand the specificity of FAP, we examined the activity of the protease against a P2-Pro1-AMCC dipeptide substrate library (Fig. 2), where P2 was varied with all amino acids (except Cys and Trp) and amide bond hydrolysis occurs between the P1 Pro and the fluorogenic leaving group, AMCC. Even though DPP-4 can cleave dipeptides with a P1 Ala residue, we limited the P1 position of our library to Pro because FAP showed little activity against P2-Ala dipeptide substrates in preliminary studies. The dipeptide cleavage profile of the library provides information regarding protease S2 subsite specificity. As shown in Fig. 2A, FAP preferred Ile, Pro, and Arg in the P2 position and showed little activity against dipeptides with Asp, His, or Asn in the P2 position. DPP-4 readily cleaved all substrates in the library, indicating broad specificity at the S2 subsite (Fig. 2B). Kinetic Constants for Cleavage of Dipeptide Substrates—To extend the results obtained with the dipeptide substrate library, we determined kinetic parameters for cleavage of commercially available dipeptide substrates by FAP (Table 1). The catalytic efficiency (kcat/Km) for substrate cleavage was greatest with Ile-Pro-AFC, followed by Ala-Pro-, Gly-Pro-, and Phe-Pro-AFC, consistent with the dipeptide substrate library results (Fig. 2). With the exception of Ile-Pro-AFC, these differences reflect differences in kcat values, as the Km for each substrate was ∼250 μm. The greater catalytic efficiency observed for Ile-Pro-AFC hydrolysis was due to both kcat and Km effects as the observed Km was ∼2.5-fold lower than the other P2-Pro1-AFC substrates. FAP showed markedly less activity against P2-Ala1-based substrates. GA-AMC was not cleaved, and the catalytic efficiency for Lys-Ala-AFC cleavage was 400–1000-fold less than the catalytic efficiency for cleavage of P2-Pro1-based peptides (Table 1), indicating that FAP prefers Pro in the P1 position.TABLE 1Kinetic constants (average ± S.E., n =3) for hydrolysis of dipeptide substrates by FAP and DPP-4 Substrate cleavage was at 23 °C in 50 mm Tris-HCl, 100 mm NaCl, 1 mm EDTA, pH 7.4.ProteaseSubstrateKmkcatkcat/Kmμms-1m-1 s-1FAPAP-AFCaAmino acids are listed in single-letter code244 ± 2814.2 ± 0.95.8 × 104FP-AFC245 ± 221.1 ± 0.14.5 × 103GP-AFC248 ± 125.6 ± 0.22.3 × 104IP-AFC106 ± 66.9 ± 0.36.5 × 104GA-AMCNCbNC, no cleavage up to 1 mm substrateNCbNC, no cleavage up to 1 mm substrateNCbNC, no cleavage up to 1 mm substrateKA-AFC189 ± 200.01 ± 0.00153DPP-4AP-AFC16 ± 345.6 ± 1.72.9 × 106GP-AFC76 ± 10121 ± 51.6 × 106a Amino acids are listed in single-letter codeb NC, no cleavage up to 1 mm substrate Open table in a new tab For comparison, we determined kinetic constants for cleavage of Ala-Pro-AFC and Gly-Pro-AFC by DPP-4 (Table 1). The catalytic efficiency for Ala-Pro-AFC hydrolysis was greater than that for Gly-Pro-AFC, consistent with the dipeptide library. Strikingly, the catalytic efficiencies for dipeptide hydrolysis by DPP-4 were consistently ∼100-fold greater than observed with FAP, reflecting both an increase in kcat and decrease in Km. N-acetyl-dipeptide Substrate Specificity—Given reports suggesting that FAP has endopeptidase activity (8Park J.E. Lenter M.C. Zimmermann R.N. Garin-Chesa P. Old L.J. Rettig W.J. J. Biol. Chem. 1999; 274: 36505-36512Abstract Full Text Full Text PDF PubMed Scopus (441) Google Scholar, 28Lee K. Jackson K. Christiansen V. Chung K. McKee P. Blood. 2004; 103: 3783-3788Crossref PubMed Scopus (113) Google Scholar), we next examined the endopeptidase specificity of the protease. For this, we assayed FAP against an Ac-P2-Pro1-AFC substrate library in which the acetyl group forms an amide bond with the P2 amino acid, thereby mimicking an endopeptidase substrate. Thus, in contrast with the P2-Pro1 library, which contains a free N terminus, the acetylated library is N-terminal blocked. FAP showed a marked preference for a P2 Gly residue in the acetylated library, having little activity against all other substrates in the library (Fig. 2C). Strikingly, DPP-4 had little activity against Ac-Gly-Pro-AFC and all other substrates in the acetylated library (Fig. 2D). These data suggest that DPP-4 has limited endopeptidase activity and that FAP endopeptidase activity is restricted to Gly-Pro-containing substrates. Kinetic Analysis of Gly-Pro- and Ac-Gly-Pro-AFC Hydrolysis—In light of the Ac-P2-Pro1-AFC library results, we performed kinetic analyses with Ac-Gly-Pro-AFC and Gly-Pro-AFC substrates as their hydrolysis rates may be directly compared. FAP cleaved each substrate with similar efficiency as shown in Fig. 3A. FAP showed a moderate increase in Km (330 ± 30 μm) for Ac-Gly-Pro-AFC relative to Gly-Pro-AFC (Table 1); however, a concomitant increase in kcat (7.7 ± 0.2 s-1) for the acetylated substrate resulted in an equivalent catalytic efficiency for both substrates. By contrast, DPP-4 only showed significant activity against Gly-Pro-AFC (Fig. 3B). The catalytic efficiency for cleavage of Ac-Gly-Pro-AFC by DPP-4 was equivalent to 36 ± 3 m-1 s-1 when determined under pseudo-first order conditions. This value is over four orders of magnitude lower than that for cleavage of Gly-Pro-AFC (Table 1), indicating a marked preference for the substrate with a free N terminus. Cleavage of N-substituted-Gly-Pro-AMCC Substrates—To test whether FAP could cleave other N-substituted-Gly-Pro-based substrates, we synthesized N-blocked dipeptide substrates (N-methyl-, formyl-, succinyl-, benzyloxycarbonyl- (Z-), biotinyl-Gly-Pro-AMCC) and a peptide modeled after the FAP cleavage site in α2-antiplasmin (Ac-Thr-Ser-Gly-Pro-AMCC) (28Lee K. Jackson K. Christiansen V. Chung K. McKee P. Blood. 2004; 103: 3783-3788Crossref PubMed Scopus (113) Google Scholar). With the exception of succinyl-Gly-Pro-AMCC, FAP cleaved all N-substituted-Gly-Pro-AMCC substrates at 35–165% of the rate for Gly-Pro-AMCC hydrolysis (Fig. 4B), indicating that th
Referência(s)