Portal de Periódicos da CAPES

The Mechanism by Which Heparin Promotes the Inhibition of Coagulation Factor XIa by Protease Nexin-2

1997; Elsevier BV; Volume: 272; Issue: 42 Linguagem: Inglês

10.1074/jbc.272.42.26139

1083-351X

Yan Zhang, Joseph M. Scandura, William E. Van Nostrand, Peter N. Walsh,

Blood Coagulation and Thrombosis Mechanisms

Previous kinetic studies have shown that protease nexin-2 is a potent, reversible, and competitive inhibitor of factor XIa. Here we show that high molecular weight heparin potentiates the ability of protease nexin-2 to inhibit factor XIa with a parabolic concentration dependence, predominantly because of an increase of the association rate constant with little perturbation of the dissociation rate constant. No effect on factor XIa inhibition by protease nexin-2 was observed with heparin preparations of 6–22 saccharide units (0.1 nm–10 μm), whereas heparin preparations with 32–64 saccharide units potentiated factor XIa inhibition by protease nexin-2 in a size- and concentration-dependent manner. We propose a model wherein heparin exerts this effect by providing a template for the assembly of factor XIa-protease nexin-2 complexes, and only heparin polymers consisting of greater than 32 saccharide units (M r ∼10,000) are sufficiently long to provide a template to which factor XIa and protease nexin-2 molecules can bind simultaneously. Heparin-mediated enhancement of factor XIa inhibition by protease nexin-2 was partially abrogated by high molecular weight kininogen, suggesting that high molecular weight kininogen may play a role in regulating factor XIa activity. Previous kinetic studies have shown that protease nexin-2 is a potent, reversible, and competitive inhibitor of factor XIa. Here we show that high molecular weight heparin potentiates the ability of protease nexin-2 to inhibit factor XIa with a parabolic concentration dependence, predominantly because of an increase of the association rate constant with little perturbation of the dissociation rate constant. No effect on factor XIa inhibition by protease nexin-2 was observed with heparin preparations of 6–22 saccharide units (0.1 nm–10 μm), whereas heparin preparations with 32–64 saccharide units potentiated factor XIa inhibition by protease nexin-2 in a size- and concentration-dependent manner. We propose a model wherein heparin exerts this effect by providing a template for the assembly of factor XIa-protease nexin-2 complexes, and only heparin polymers consisting of greater than 32 saccharide units (M r ∼10,000) are sufficiently long to provide a template to which factor XIa and protease nexin-2 molecules can bind simultaneously. Heparin-mediated enhancement of factor XIa inhibition by protease nexin-2 was partially abrogated by high molecular weight kininogen, suggesting that high molecular weight kininogen may play a role in regulating factor XIa activity. Protease nexin-2 (PN-2) 1The abbreviations used are: PN-2, protease nexin-2; FXIa, factor XIa; HK, high molecular weight kininogen; Boc-EAR-AMC, Boc-Glu(OBzl)-Ala-Arg-7-amino-4-methylcoumarin; S-2366, pyro-Glu-Pro-Argp-nitroaniline. 1The abbreviations used are: PN-2, protease nexin-2; FXIa, factor XIa; HK, high molecular weight kininogen; Boc-EAR-AMC, Boc-Glu(OBzl)-Ala-Arg-7-amino-4-methylcoumarin; S-2366, pyro-Glu-Pro-Argp-nitroaniline.is a ∼120-kDa soluble fragment of Alzheimer amyloid β-protein precursor isoforms containing the 56-amino acid Kunitz-type (or Kunin) 2The term Kunin refers to a family of proteins having structural homology with aprotinin and is preferred over the designation Kunitz-type (4Salvesen G. Pizzo S.V. Hemostasis and Thrombosis: Basic Principles and Clinical Practice. J. B. Lippincott, Philadelphia1994Google Scholar). 2The term Kunin refers to a family of proteins having structural homology with aprotinin and is preferred over the designation Kunitz-type (4Salvesen G. Pizzo S.V. Hemostasis and Thrombosis: Basic Principles and Clinical Practice. J. B. Lippincott, Philadelphia1994Google Scholar). protease inhibitor domain (1Oltersdorf T. Fritz L.C. Schenk D.B. Lieberburg I. Johnson-Wood K.L. Beattie E.C. Ward P.J. Blacher R.W. Dovey H.F. Sinha S. Nature. 1989; 341: 144-147Crossref PubMed Scopus (367) Google Scholar, 2Van Nostrand W.E. Wagner S.L. Suzuki M. Choi B.H. Farrow J.S. Geddes J.W. Cotman C.W. Cunningham D.D. Nature. 1989; 341: 546-549Crossref PubMed Scopus (347) Google Scholar, 3Kitaguchi N. Takahashi Y. Tokushima Y. Shiojiri S. Ito H. Nature. 1988; 331: 530-532Crossref PubMed Scopus (884) Google Scholar, 4Salvesen G. Pizzo S.V. Hemostasis and Thrombosis: Basic Principles and Clinical Practice. J. B. Lippincott, Philadelphia1994Google Scholar). PN-2 has been found to be an abundant platelet α-granule protein that is secreted from platelets upon their activation with physiological agonists (5Bush A.I. Martins R.N. Rumble B. Moir R. Fuller S. Milward E. Currie J. Ames D. Weidemann A. Fischer P. Multhaup G. Beyreuther K. Masters C.L. J. Biol. Chem. 1990; 265: 15977-15983Abstract Full Text PDF PubMed Google Scholar, 6Smith R.P. Higuchi D.A. Broze Jr., G.J. Science. 1990; 248: 1126-1128Crossref PubMed Scopus (253) Google Scholar, 7Van Nostrand W.E. Schmaier A.H. Farrow J.S. Cunningham D.D. Science. 1990; 248: 745-748Crossref PubMed Scopus (264) Google Scholar, 8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar). Factor XIa (FXIa) is a homodimeric serine protease (∼160 kDa) involved in the intrinsic pathway of blood coagulation. We have previously conducted detailed kinetic analyses to characterize the inhibition of FXIa by PN-2 using a fluorogenic substrate that is ∼1,000-fold more sensitive for FXIa than the most sensitive and specific chromogenic substrates (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar). The mechanism of FXIa inhibition by PN-2 is best described as a simple, one-step binding interaction during which PN-2 reversibly inhibits FXIa. Our studies have shown that PN-2 is a slow, tight binding inhibitor of FXIa, with an association rate constant (k on) of 2.1 ± 0.2 × 106m−1 s−1, a dissociation rate constant (k off) of 8.5 ± 0.8 × 10−4 s−1, and aK i of 400 pm, a value in good agreement with previous reports (6Smith R.P. Higuchi D.A. Broze Jr., G.J. Science. 1990; 248: 1126-1128Crossref PubMed Scopus (253) Google Scholar, 7Van Nostrand W.E. Schmaier A.H. Farrow J.S. Cunningham D.D. Science. 1990; 248: 745-748Crossref PubMed Scopus (264) Google Scholar, 8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar). Further, we found that high molecular weight kininogen (HK) protects FXIa from inhibition by PN-2 with a saturable concentration dependence (the EC50 for HK was essentially identical to theK d for the binding interaction between FXIa and HK) and that this protection results from the decreased ability of PN-2 to inhibit the FXIa·HK complex (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar). Zinc ions, which are known to affect several functions of both FXIa (9Van Nostrand W.E. Thromb. Res. 1995; 78: 43-53Abstract Full Text PDF PubMed Scopus (31) Google Scholar, 10Greengard J.S. Heeb M.J. Ersdal E. Walsh P.N. Griffin J.H. Biochemistry. 1988; 25: 3884-3890Crossref Scopus (86) Google Scholar) and PN-2 (11Bush A.I. Multhaup G. Moir R.D. Williamson T.G. Small D.H. Rumble B. Pollwein P. Beyreuther K. Masters C.L. J. Biol. Chem. 1993; 268: 16109-16112Abstract Full Text PDF PubMed Google Scholar), are known to increase the affinity of the FXIa·PN-2 complex (9Van Nostrand W.E. Thromb. Res. 1995; 78: 43-53Abstract Full Text PDF PubMed Scopus (31) Google Scholar, 12Komiyama Y. Murakami T. Egawa H. Okubo S. Yasunaga K. Murata K. Thromb. Res. 1992; 66: 397-408Abstract Full Text PDF PubMed Scopus (20) Google Scholar). We found that Zn2+ was able to abolish the protection afforded to FXIa by HK and hypothesized the existence of a zinc-dependent conformation of HK which does not obstruct the interaction between FXIa and PN-2 (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar). Several studies have shown that the capacity of PN-2 to inhibit FXIa is potentiated by heparin (6Smith R.P. Higuchi D.A. Broze Jr., G.J. Science. 1990; 248: 1126-1128Crossref PubMed Scopus (253) Google Scholar, 13Van Nostrand W.E. Wagner S.L. Farrow J.S. Cunningham D.D. J. Biol. Chem. 1990; 265: 9591-9594Abstract Full Text PDF PubMed Google Scholar), a negatively charged glycosaminoglycan that is synthesized by mast cells and is similar to the heparan sulfate glycosaminoglycans expressed on the surface of endothelial cells. In the present studies, we have used highly purified heparin fractions to perform experiments designed to elucidate the mechanism by which heparin enhances the interaction between FXIa and PN-2. Moreover, because Zn2+ ions potentiate FXIa inhibition by PN-2 (9Van Nostrand W.E. Thromb. Res. 1995; 78: 43-53Abstract Full Text PDF PubMed Scopus (31) Google Scholar,12Komiyama Y. Murakami T. Egawa H. Okubo S. Yasunaga K. Murata K. Thromb. Res. 1992; 66: 397-408Abstract Full Text PDF PubMed Scopus (20) Google Scholar) and because previous studies have shown that HK and Zn2+ ions regulate the inhibition of FXIa by PN-2 secreted from activated platelets (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar), we were prompted to investigate the effects of HK and Zn2+ ions on the heparin-potentiated inhibition of FXIa by PN-2. PN-2 was purified from fibroblast culture media using techniques of heparin affinity chromatography and immunoaffinity chromatography as described previously (7Van Nostrand W.E. Schmaier A.H. Farrow J.S. Cunningham D.D. Science. 1990; 248: 745-748Crossref PubMed Scopus (264) Google Scholar). Human FXIa was purchased from Hematologic Technologies, Inc. (Essex Junction, VT). Human HK was from Enzyme Research Laboratories, Inc. (South Bend, IN). Fluorescein mono-p-guanidinobenzoate hydrochloride, fluorescein, and unfractionated heparin from bovine lung (averageM r = 10,000) were purchased from Sigma. Purified heparin fractions consisting of 6, 10, 14, 18, 22, 32, 48 and 64 saccharide units (1.74, 2.97, 4.2, 5.43, 7, 10, 15, and 20 kDa, respectively) were purified and kindly donated by Dr. Steven T. Olson (University of Illinois, Chicago) and Dr. Ingemar Björk (Uppsala Biomedical Center, Uppsala, Sweden). Their preparation and characterization are described elsewhere (14Lindahl U. Thunberg L. Backstrom G. Riesenfeld J. Nordling K. Björk I. J. Biol. Chem. 1984; 259: 12368-12374Abstract Full Text PDF PubMed Google Scholar, 15Olson S.T. J. Biol. Chem. 1988; 263: 1698-1708Abstract Full Text PDF PubMed Google Scholar). The fluorogenic substrate Boc-EAR-AMC was from Peptide International, Inc. (Louisville, KY). The chromogenic substrate S-2366 was purchased from Chromogenix (Mölndal, Sweden). All other reagents were analytical grade or the best quality commercially available. The active site concentration of purified FXIa was titrated with the sensitive burst titrant fluorescein mono-p-guanidinobenzoate hydrochloride (16Melhado L.L. Peltz S.W. Leytus S.P. Mangel W.F. J. Am. Chem. Soc. 1982; 104: 7299-7306Crossref Scopus (54) Google Scholar) as described previously (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar). The concentration of PN-2 reactive sites was determined by titration with active site-titrated trypsin or FXIa as described (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar). We have shown previously (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar) that in the absence of heparin PN-2 inhibits FXIa via a "one-step" mechanism as follows, E+S⇆KmE·S→kcatE+P+Ikon⇵koffE·I MECHANISM1where E represents the enzyme FXIa, I represents the inhibitor PN-2, and k on andk off are the forward and the reverse rate constants for the inhibition, respectively. We assumed that in the presence of heparin, this mechanism was also valid, and we show evidence to support this assumption (see "Results"). Progress curve measurements were performed both in the absence and the presence of unfractionated heparin or purified heparin fractions. Cleavage of the fluorescent substrate Boc-EAR-AMC (17Kawabata S. Miura T. Morita T. Kato H. Fujikawa K. Iwanaga S. Takada K. Kimura T. Sakakibara S. Eur. J. Biochem. 1988; 172: 17-25Crossref PubMed Scopus (230) Google Scholar) by FXIa in the presence of varying concentrations of PN-2 was monitored in a Bowman series 2 spectrofluorometer (SLM Aminco, Urbana, IL) with excitation and emission wavelengths of 375 and 445 nm, respectively (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar). All reactions were conducted at 37 °C in 2 ml of Hepes-Tyrode's buffer (126 mm NaCl, 2.7 mm KCl, 1 mmMgCl2·6H2O, 0.4 mmNaH2 PO4·H2O, 5.6 mmdextrose, 15 mm Hepes, 0.1% bovine serum albumin, pH 7.4, at 37 °C) and underwent continuous gentle stirring throughout the measurement period. Hydrolysis of the fluorescent substrate Boc-EAR-AMC (25 μm final) by FXIa (20 pm final active site concentration) was recorded at 15-s intervals for 10 min before and up to 1 h after the addition (<20 μl) of PN-2 (0–2 nm final). Alternatively, FXIa and PN-2 were preincubated for 1 h at 37 °C before being diluted 100-fold into substrate-containing buffer. The change in relative fluorescence after dilution was monitored for 1 h at 15-s intervals. Data recorded after the addition of PN-2 to FXIa represent association progress curves, whereas data monitored after the dilution of preformed FXIa·PN-2 complexes represent dissociation progress curves. The rate constants (k on andk off) were obtained by nonlinear, least squares fits to Equation 2 of the association and dissociation progress curve data measured at multiple PN-2 concentrations. Equation 2 results from the solution of the nonlinear, separable differential equation, Equation 1, and is valid for reactions proceeding according to a one-step mechanism (described below) even when the concentration of inhibitor changes dramatically during the experimental time course (i.e. under true second-order conditions). In some cases, when data were collected under pseudo first-order conditions, the progress curve data were analyzed with the equations of Cha (18Cha S. Biochem. Pharmacol. 1975; 24: 2177-2185Crossref PubMed Scopus (446) Google Scholar) as described previously (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar). ςEIςt=kon·(ET−EI)·(IT−EI)−koff·EIEquation 1 When the enzyme concentration is low enough to ensure that the substrate concentration remains essentially constant during the measurement period (i.e. S t ≈S0), as it did during our experiments using 20 pm FXIa, Equation 2 is the closed form, explicit solution to Equation 1. P(t)=kcatS0Km+S0·ET+b+γ2·t+Ikon,app·lnQ0−exp(−kon,app·γ·t)Q0−1Equation 2 In Equation 2, P(t) is the fluorescent product generated with time. K m andk cat are the classic Michaelis-Menten kinetic parameters for the fluorogenic substrate, and S0 is the initial substrate concentration. Other parameters used in this equation are defined as follows, kon,app=kon1+S0/KmEquation 3 b=−(IT+ET+koff/kon,app)Equation 4 γ=b2−4·IT·ETEquation 5 Q0=2(EI)0+b−γ2(EI)0+b+γEquation 6 where k on and k offare the forward and reverse rate constants governing the interaction between enzyme and inhibitor, and IT andE T are the total inhibitor and enzyme concentrations, respectively. (EI)0 is the enzyme-inhibitor complex present at the start of the monitored reaction. For association experiments (EI)0 = 0. For dissociation experiments, (EI)0=12·ET+IT+koffkon·D−ET+IT+koffkon·D2−4·ET·ITEquation 7 where D is the factor by which the preincubated enzyme-inhibitor mixture is diluted into the final, monitored reaction. The K i was either calculated as the ratio of the reverse and forward rate constants determined from progress curve analysis or obtained from equilibrium measurements as described below. In the equilibrium method, increasing amounts of PN-2 (0–10 nm in the absence of heparin and 0–1 nm in the presence of heparin) were incubated with a constant amount of FXIa (125 pm) at 37 °C for 1 h to allow the establishment of an equilibrium between the inhibitor and enzyme. After this preincubation step, a small volume of substrate S-2366 (0.5 mm) was added, and the residual FXIa activity was determined by measuring the initial velocity of substrate cleavage for 10 min. This time period was sufficiently short, and the substrate concentration was sufficiently low to ensure that the equilibrium was not perturbed during the measurement period. Measurements were made in the presence of varying concentrations of unfractionated heparin or purified heparin fractions. The measurements were also made in the presence of HK (50–500 nm) or ZnCl2 (25 μm) with and without unfractionated heparin (0.1 μg/ml) or the 64 saccharide heparin fraction (100 nm). After equilibration, the inhibition constant, K i, was obtained by the least squares fit to Equation 8 of the ratio of the velocity of substrate cleavage measured at varying inhibitor concentrations, V I, to the velocity measured in the absence of inhibitor, V 0 (3Kitaguchi N. Takahashi Y. Tokushima Y. Shiojiri S. Ito H. Nature. 1988; 331: 530-532Crossref PubMed Scopus (884) Google Scholar). VIV0=1−(ET+IT+Ki)−(ET+IT+Ki)2−4·IT·ET2ETEquation 8 We have measured the equilibrium inhibition constant (K i ) of FXIa inhibition by PN-2 in the presence of varying concentrations of unfractionated heparin and well characterized, purified heparin subfractions (14Lindahl U. Thunberg L. Backstrom G. Riesenfeld J. Nordling K. Björk I. J. Biol. Chem. 1984; 259: 12368-12374Abstract Full Text PDF PubMed Google Scholar, 15Olson S.T. J. Biol. Chem. 1988; 263: 1698-1708Abstract Full Text PDF PubMed Google Scholar). We found that high molecular weight heparin (fractions containing 32–64 saccharide units) potentiated FXIa inhibition by PN-2 with a parabolic concentration dependence, whereas heparin fractions with a shorter length (fractions containing 6–22 saccharide units) were unable to affect the PN-2/FXIa interaction (Fig.1). Each of the active heparin subfractions possessed a broad optimal concentration range occurring between 0.01 and 1 μm. Moreover, the extent to which each active heparin species was able to potentiate the ability of PN-2 to inhibit FXIa was directly related to its strand length. Our previous studies (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar) demonstrated that in the absence of heparin the kinetics of FXIa inhibition by PN-2 conforms to a one-step second-order reaction mechanism. Heparin does not appear to affect this mechanism. We analyzed pseudo first-order data obtained in the presence of an optimal concentration of fractionated heparin (100 nm, 64 saccharide units) using methods described previously (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar) and equations developed by Cha (18Cha S. Biochem. Pharmacol. 1975; 24: 2177-2185Crossref PubMed Scopus (446) Google Scholar) and extended by Morrison and Walsh (19Morrison J.F. Trends Biochem. Sci. 1982; 7: 102-105Abstract Full Text PDF Scopus (492) Google Scholar, 20Morrison J.F. Walsh C.T. Adv. Enzymol. 1988; 61: 201-301PubMed Google Scholar). Consistent with the reaction proceeding via a simple one-step mechanism, the pseudo first-order rate constant observed,k obs, was linearly dependent upon the inhibitor concentration (Fig. 2). We also found that the stoichiometry of the FXIa/PN-2 interaction is not changed by the presence of heparin (data not shown). To investigate the mechanism by which heparin increases the affinity of the interaction between PN-2 and FXIa, progress curve measurements were performed and the kinetic rate constants ascertained. The forward (k on) and reverse (k off) rate constants were obtained by fitting families of association (e.g. Fig. 3) or dissociation (e.g. Fig. 4) progress curves simultaneously to the appropriate form of Equation 2. Rate constants derived from the data shown in Fig. 3 demonstrate that the increased potency with which PN-2 inhibits FXIa in the presence of an optimal concentration (100 nm) of long chain heparin molecules (64 saccharide units) is the result of an increased forward rate constant (k on increased from 3.2 to 43 × 106m−1 s−1 in the presence of heparin); the reverse rate constant,k off, was not significantly perturbed (k off was 0.0013 s−1 in the absence of and 0.0014 s−1 in the presence of heparin). The rate constants derived from analysis of dissociation progress curves were very similar to those obtained from the association progress curves and confirm that heparin, when sufficiently long, brings about an increase in the second-order association rate constant (k on increased from 2.0 to 27 × 106m−1 s−1) but has little discernible effect on the reverse rate constant (k off = 8.5 and 7.8 × 10−4 in the absence and presence of heparin, respectively) (see TableI for average values).Figure 4Dissociation progress curves of the inhibition of FXIa by PN-2 in the presence and absence of heparin.Equal volumes of FXIa and PN-2 were preincubated in the absence (panel A) or presence (panel B) of heparin fractions containing 64 saccharide units (100 nm) at 37 °C for 1 h. The mixture was then diluted 100-fold into substrate (Boc-EAR-AMC, 25 μm)-containing buffer. The final concentration of FXIa was 20 pm. Final PN-2 concentrations are shown on the right side of eachpanel. Data were fitted in Equation 2 as described under "Experimental Procedures."View Large Image Figure ViewerDownload Hi-res image Download (PPT)Table IRate constants of FXIa inhibition by PN-2ConditionAssociation rate constant (k on)1-aSecond-order rate constants were obtained by computational analysis of progress curve data as shown in Fig. 3 using Equation 2. K i ,calc was calculated from the relationship K i =k off /k on.Dissociation rate constant (k off)1-aSecond-order rate constants were obtained by computational analysis of progress curve data as shown in Fig. 3 using Equation 2. K i ,calc was calculated from the relationship K i =k off /k on.K i, calc1-aSecond-order rate constants were obtained by computational analysis of progress curve data as shown in Fig. 3 using Equation 2. K i ,calc was calculated from the relationship K i =k off /k on.K i, meas1-bK i ,meas was determined directly with equilibrium measurement.m −1 s −1s −1pmpmNo heparin3.2 × 106 ± 0.21.3 × 10−3 ± 0.3410 ± 20370 ± 45ZnCl2 alone1-c25 μm ZnCl2.6.8 × 106 ± 0.49.4 × 10−4 ± 0.5140 ± 6150 ± 10HK alone1-d250 nm HK.1.8 × 106 ± 0.31.3 × 10−3 ± 0.2710 ± 46740 ± 23ZnCl2 + HK5.3 × 106 ± 0.68.7 × 10−4 ± 0.4170 ± 5160 ± 7UF Heparin1-eUnfractionated heparin (∼10,000 Da), 0.1 μg/ml.2.4 × 107 ± 0.18.4 × 10−4 ± 0.535 ± 1026 ± 5F Heparin1-fFractionated heparin consisting of 64 saccharide units (100 nm).4.3 × 107 ± 0.51.4 × 10−3 ± 0.233 ± 225 ± 3 with ZnCl29.2 × 106 ± 0.58.9 × 10−4 ± 0.297 ± 2278 ± 13 with HK1.1 × 107 ± 0.71.2 × 10−3 ± 1.0100 ± 15140 ± 21 with ZnCl2 + HK3.4 × 106 ± 1.27.6 × 10−4 ± 0.8220 ± 30160 ± 101-a Second-order rate constants were obtained by computational analysis of progress curve data as shown in Fig. 3 using Equation 2. K i ,calc was calculated from the relationship K i =k off /k on.1-b K i ,meas was determined directly with equilibrium measurement.1-c 25 μm ZnCl2.1-d 250 nm HK.1-e Unfractionated heparin (∼10,000 Da), 0.1 μg/ml.1-f Fractionated heparin consisting of 64 saccharide units (100 nm). Open table in a new tab We conducted experiments similar to those presented in Fig. 3 to ascertain the effect of increasing concentrations of each heparin subfraction (containing 6, 10, 14, 18, 22, 32, 40, and 64 saccharide units) on the kinetics of FXIa inhibition by PN-2 (see Fig.5). As shown in Fig. 5, heparin strands affect the forward rate constant in a manner that is both concentration- and chain length-dependent but do not significantly affect the reverse rate constant at any concentration. As would be expected from the results presented in Fig. 1, only the heparin fractions greater than 32 saccharide units in length bring about an increase in the forward rate constant. This effect is maximal for the longest heparin molecules (64 saccharide units) and has a distinctive parabolic concentration dependence. In the absence of heparin, HK and Zn2+ ions affect the inhibition of FXIa by PN-2 in opposite ways (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar): HK partially protects FXIa from PN-2, whereas Zn2+ ions enhance the ability of PN-2 to inhibit it. We also found (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar) (Fig. 6) that only the effect on Zn2+ is seen when both HK and Zn2+are present: Zn2+ completely abrogates the effect of HK in the absence of heparin. In the present study, we have investigated the effect of HK and Zn2+ in both the absence and presence of heparin using the same kinetic approach used in their absence (TableI). Determination of association and dissociation rate constants (TableI) shows that Zn2+ ions, like heparin, enhance the inhibitory effect of PN-2 on FXIa (i.e. decrease theK i ) by increasing the association rate constant without affecting the dissociation rate constant. Similarly, the protective effect of HK is caused by its capacity to decrease the association rate constant without affecting the dissociation rate constant, and as predicted from previous studies (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar), Zn2+completely abrogates the protective effect of HK when both are present. Contrary to what was seen in the absence of heparin, in its presence, zinc ions have very little effect on the rate constants with which FXIa is inhibited by PN-2; the effect of heparin predominates (Table I). Nonetheless, even in the presence of heparin, HK induces a saturable but partial reversal of the effect of heparin onk on. This effect of HK is saturated at its plasma concentration (HK ∼640 nm in plasma) where the ability of heparin to potentiate PN-2 is approximately halved (see Fig.6 and Table I). Because HK also binds to heparin, a plausible mechanism by which HK could reverse the effect of heparin would be by displacing FXIa and PN-2 from their heparin binding sites. As shown in Fig.7, it is unlikely that this mechanism pertains because the optimal heparin concentration is not shifted to a higher value when HK is present. In any event, the protective effect of HK is canceled when Zn2+ is added to the reaction mixture even in the presence of heparin (Fig. 6 and Table I). The goal of this study was to determine the effect of heparin on the kinetics of FXIa inhibition by PN-2. Using computational methods to analyze progress curve data, we obtained reliable estimates of the individual rate constants governing the reaction between FXIa and PN-2 in the presence of heparin. We found that 1) the overall reaction mechanism can be well described by a simple second-order model; 2) the increased inhibitory potency of PN-2 in the presence of heparin results entirely from an increase in the forward rate constant and not from a decrease in the reverse rate constant; 3) the heparin effect is concentration-dependent, showing little effect at both low and high heparin concentrations and possessing an optimal concentration at which maximum potentiation is observed; and 4) only heparin preparations at least 32 saccharide units in length are able to affect the kinetics of FXIa inhibition by PN-2. We believe that heparin brings about its effect by providing a template sufficiently long (at least 32 saccharide units) to facilitate the rapid assembly of a high affinity protease-inhibitor complex. Our "template model," similar to that of the antithrombin III/thrombin interaction in the presence of heparin (21Griffith M.J. J. Biol. Chem. 1982; 257: 7360-7365Abstract Full Text PDF PubMed Google Scholar, 22Danielsson Å. Raub E. Lindahl U. Björk I. J. Biol. Chem. 1986; 261: 15467-15473Abstract Full Text PDF PubMed Google Scholar, 23Nesheim M. Blackburn M.N. Lawler C.M. Mann K.G. J. Biol. Chem. 1986; 261: 3214-3221Abstract Full Text PDF PubMed Google Scholar, 24Olson S. Björk I. Semin. Thromb. Hemostasis. 1994; 20: 373-409Crossref PubMed Scopus (126) Google Scholar, 25Holmer E. Kurachi K. Söderstrom G. Biochem. J. 1981; 193: 395-400Crossref PubMed Scopus (210) Google Scholar), suggests that a heparin-bound FXIa molecule can rapidly and efficiently react with PN-2 molecules bound to the same heparin strand. This model provides a qualitative explanation for all of the observed effects of heparin on the interaction between FXIa and its most potent inhibitor, PN-2. 1) Our data show that high affinity interactions of FXIa and PN-2 with heparin occur more rapidly than the interaction between FXIa and PN-2, as evidenced by an increased second-order rate constant (k on) without any change in the first-order dissociation rate constant. 2) The parabolic concentration profile observed for heparin is also consistent with the template model. At low heparin concentrations, there are not enough heparin strands to saturate all protein binding sites, and the rate of FXIa inhibition is determined mainly by the interaction between the free proteins, whereas excessively high concentrations of heparin saturate FXIa and PN-2 molecules individually, thus decreasing the likelihood of their binding to the same heparin strand. Only at the optimal heparin concentration is there the highest probability of FXIa and PN-2 molecules being bound to the same heparin strand. 3) This model can also explain why only heparin strands longer than a discrete threshold are able to potentiate the ability of PN-2 to inhibit FXIa. Preparations less than 32 saccharide units in length are too short to provide a template to bind both FXIa and PN-2 molecules simultaneously (Figs. 1 and 5). We have observed, as predicted by the model, that as the length of the heparin strand increases, so does its ability to potentiate PN-2. The fact that both the concentration and size of heparin molecules are critical to the degree to which it can enhance the inhibitory capacity of PN-2 supports the view that one heparin strand interacts with both FXIa and PN-2 molecules. Both heparin and Zn2+ ions have been found to bind to PN-2 molecules, to enhance the ability of PN-2 to inhibit FXIa, and to regulate several functions of PN-2 (6Smith R.P. Higuchi D.A. Broze Jr., G.J. Science. 1990; 248: 1126-1128Crossref PubMed Scopus (253) Google Scholar, 11Bush A.I. Multhaup G. Moir R.D. Williamson T.G. Small D.H. Rumble B. Pollwein P. Beyreuther K. Masters C.L. J. Biol. Chem. 1993; 268: 16109-16112Abstract Full Text PDF PubMed Google Scholar). Heparin and Zn2+ bind to different regions of PN-2, thereby permitting complementary effects (9Van Nostrand W.E. Thromb. Res. 1995; 78: 43-53Abstract Full Text PDF PubMed Scopus (31) Google Scholar). However, this was not observed in our present studies. Although Zn2+ alone induced an ∼3-fold increase in k on and heparin alone increasedk on ∼10-fold (Table I), there was little additional effect when both Zn2+ and heparin were present together. Our recent (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar) and present studies show that HK also regulates the inhibition of FXIa by PN-2. HK alone protects FXIa from inactivation by PN-2 (Fig. 6). In the presence of heparin, HK can also partially protect FXIa from PN-2 (Fig. 6), but this protective effect is superimposed upon the more significant potentiation of PN-2 by heparin. Both in the presence and absence of heparin, the effect of HK is saturable and likely results from the binding of HK to FXIa (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar). Zn2+ both potentiates PN-2 and reduces the ability of HK to protect FXIa from inhibition by PN-2. In the presence of both Zn2+ ions and heparin, the protective effect of HK is abrogated. Because there are high affinity heparin binding sites on all three proteins, PN-2 (6Smith R.P. Higuchi D.A. Broze Jr., G.J. Science. 1990; 248: 1126-1128Crossref PubMed Scopus (253) Google Scholar), HK (26Björk I. Olson S.T. Sheffer R.G. Shore J.D. Biochemistry. 1989; 28: 1213-1221Crossref PubMed Scopus (42) Google Scholar), and FXIa 3D. Ho, K. Badellino, F. A. Baglia, and P. N. Walsh, manuscript in preparation. and because FXIa forms a noncovalent complex with both PN-2 (5Bush A.I. Martins R.N. Rumble B. Moir R. Fuller S. Milward E. Currie J. Ames D. Weidemann A. Fischer P. Multhaup G. Beyreuther K. Masters C.L. J. Biol. Chem. 1990; 265: 15977-15983Abstract Full Text PDF PubMed Google Scholar, 6Smith R.P. Higuchi D.A. Broze Jr., G.J. Science. 1990; 248: 1126-1128Crossref PubMed Scopus (253) Google Scholar, 7Van Nostrand W.E. Schmaier A.H. Farrow J.S. Cunningham D.D. Science. 1990; 248: 745-748Crossref PubMed Scopus (264) Google Scholar, 8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar) and HK (27Thompson R.E. Mandle R. Kaplan A.P. J. Clin. Invest. 1977; 60: 1376-1380Crossref PubMed Scopus (198) Google Scholar), it is possible that HK itself or a binary complex formed between FXIa and HK on a heparin template could compete with FXIa and PN-2 on a heparin strand. The results presented in Fig. 7 strongly argue against such a hypothesis: if HK protected FXIa from PN-2 in the presence of heparin by displacing FXIa or PN-2 from heparin binding sites, the optimal heparin concentration would have to be shifted rightward, and this was not the case (Fig. 7). Certain comparisons can be made between the PN-2/FXIa interaction and the antithrombin III/thrombin interaction (21Griffith M.J. J. Biol. Chem. 1982; 257: 7360-7365Abstract Full Text PDF PubMed Google Scholar, 22Danielsson Å. Raub E. Lindahl U. Björk I. J. Biol. Chem. 1986; 261: 15467-15473Abstract Full Text PDF PubMed Google Scholar, 23Nesheim M. Blackburn M.N. Lawler C.M. Mann K.G. J. Biol. Chem. 1986; 261: 3214-3221Abstract Full Text PDF PubMed Google Scholar, 24Olson S. Björk I. Semin. Thromb. Hemostasis. 1994; 20: 373-409Crossref PubMed Scopus (126) Google Scholar, 25Holmer E. Kurachi K. Söderstrom G. Biochem. J. 1981; 193: 395-400Crossref PubMed Scopus (210) Google Scholar). The antithrombin III/thrombin reaction is accelerated dramatically by heparin, which provides a template for the antithrombin III·thrombin·heparin ternary complex to form. A change in antithrombin III conformation has an added effect on rate enhancement. In contrast to antithrombin III, PN-2 is an excellent FXIa inhibitor even in the absence of heparin (k on = 3.2 × 106m−1 s−1,k off = 1.3 × 10−3s−1). Our kinetic analysis of the heparin-enhanced PN-2/FXIa interaction showed only a 10-fold increase in rate, in contrast to the 2,000–4,000-fold increase in the rate of thrombin inhibition by antithrombin III brought about by heparin. At least 18 saccharide units are required for heparin to promote rate enhancements for inhibition of thrombin, FXIa, or FIXa by antithrombin III (22Danielsson Å. Raub E. Lindahl U. Björk I. J. Biol. Chem. 1986; 261: 15467-15473Abstract Full Text PDF PubMed Google Scholar, 25Holmer E. Kurachi K. Söderstrom G. Biochem. J. 1981; 193: 395-400Crossref PubMed Scopus (210) Google Scholar). For the PN-2/FXIa interaction, a heparin strand containing at least 32 saccharide units was both necessary and sufficient to provide this template. This size template may be required to accommodate both the FXIa molecule (160 kDa) and the PN-2 molecule (120 kDa), whereas a smaller heparin template may be sufficient to accommodate the smaller antithrombin III molecule (58 kDa). Because PN-2 is secreted from activated platelets (5Bush A.I. Martins R.N. Rumble B. Moir R. Fuller S. Milward E. Currie J. Ames D. Weidemann A. Fischer P. Multhaup G. Beyreuther K. Masters C.L. J. Biol. Chem. 1990; 265: 15977-15983Abstract Full Text PDF PubMed Google Scholar, 6Smith R.P. Higuchi D.A. Broze Jr., G.J. Science. 1990; 248: 1126-1128Crossref PubMed Scopus (253) Google Scholar, 7Van Nostrand W.E. Schmaier A.H. Farrow J.S. Cunningham D.D. Science. 1990; 248: 745-748Crossref PubMed Scopus (264) Google Scholar) at the site (i.e. platelet membranes) of assembly of the protease-cofactor-substrate complexes that lead to the generation of thrombin and hemostatic thrombi, it is ideally situated to regulate FXIa activity in the vicinity of platelet thrombi. Platelet activation and the secretion of PN-2 from platelets limit the lifetime of FXIa within the locus of activated platelets (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar). In the presence of platelets, as in the purified system, HK protects FXIa from inactivation by PN-2, and Zn2+ ions potentiate FXIa inhibition by PN-2 (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar). However, when both HK and Zn2+ are present together with platelets, it is the protective effect of HK which predominates and prolongs the lifetime of FXIa after platelet activation (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar). In our present study, we found that exogenous heparin does not promote FXIa inhibition by PN-2 secreted from activated platelets (data not shown), probably as a consequence of the neutralization of heparin by platelet factor 4, which is also secreted, along with PN-2, from the α-granules of activated platelets (28Broekman M.J. Handin R.I. Cohen P. Br. J. Haematol. 1975; 31: 51-55Crossref PubMed Scopus (59) Google Scholar, 29Handin R.I. Cohen H.J. J. Biol. Chem. 1976; 251: 4273-4282Abstract Full Text PDF PubMed Google Scholar, 30Lane D.A. Pejler G. Flynn A.M. Thompson E.A. Lindahl U. J. Biol. Chem. 1986; 261: 3980-3986Abstract Full Text PDF PubMed Google Scholar). Thus, the major regulator of FXIa within the vicinity of activated platelet thrombi is PN-2, a potent, slow, tight binding, reversible inhibitor of FXIa which is much less effective in inhibiting FXIa bound to the platelet surface in the presence of HK (8Scandura J.M. Zhang Y. Van Nostrand W.E. Walsh P.N. Biochemistry. 1997; 36: 412-420Crossref PubMed Scopus (40) Google Scholar, 10Greengard J.S. Heeb M.J. Ersdal E. Walsh P.N. Griffin J.H. Biochemistry. 1988; 25: 3884-3890Crossref Scopus (86) Google Scholar, 31Walsh P.N. Sinha D. Kueppers F. Blankstein K.B. Seaman F.S. J. Clin. Invest. 1987; 80: 1578-1586Crossref PubMed Scopus (22) Google Scholar). Nonetheless, PN-2 may be much more effective in inhibiting FXIa bound to heparin-like molecules such as heparan sulfate on the surface of nonthrombogenic cells such as endothelial cells. The relative contributions of other dominant FXIa inhibitory serine protease inhibitors found in plasma such as antithrombin III (32Beeler D. Marcum J.A. Schiffman S. Rosenberg R.D. Blood. 1986; 67: 1488-1492Crossref PubMed Google Scholar), α1-protease inhibitor, α2-antiplasmin, and C1 esterase inhibitor (33Scott C.F. Schapira M. James H.L. Cohen A.B. Colman R.W. J. Clin. Invest. 1982; 69: 844-852Crossref PubMed Scopus (97) Google Scholar) will have to be reexamined in comparison with PN-2 to resolve the question of which regulatory protein predominates under various physiological and pathophysiological conditions. We are grateful to Patricia Pileggi for assistance in manuscript preparation. We are grateful to Dr. Steven Olson and Dr. Ingemar Björk for generously providing purified heparin fractions.