Chemical modification of bovine prothrombin fragment 1 in the presence of Tb3+ ions

The formaldehyde-morpholine method for the conversion of gamma-carboxyglutamyl (Gla) residues to gamma-methyleneglutamyl (gamma-MGlu) residues has been applied to the modification of bovine prothrombin fragment 1. In the absence of Tb3+ ions or at Tb3+ ion concentrations of 2 Km app and 25 Km app the action of 10,000-fold molar excess of formaldehyde and morpholine, pH 5.0, converts the 10 Gla residues of the protein into 10 gamma-MGlu residues. Modification of the protein using the same conditions but increasing the Tb3+ concentration to 100 Km app provided a homogeneous protein containing 3 gamma-MGlu and 7 Gla residues, bovine 3 gamma-MGlu-fragment 1. The modified protein binds the same number of Ca2+ ions (6-7) as bovine fragment 1. However, the positive cooperatively associated with Ca2+ binding is abolished and the overall affinity for Ca2+ ions is reduced. Fluorescence titrations of 3 gamma-MGlu-fragment 1 using either Ca2+ or Mg2+ ions indicate that the modified protein retains a fluorescence quenching behavior similar to that of the native protein. The modified protein does not bind to phosphatidylserine/phosphatidylcholine vesicles in the presence of Ca2+ ions. Thus the metal ion-induced fluorescence transition exhibited by the bovine protein appears to be a necessary but not sufficient condition for phospholipid binding.     

Influence of specific gamma-carboxyglutamic acid residues on the integrity of the calcium-dependent conformation of human protein C.

The concentration of Ca2+ that produced 50% of the saturable intrinsic fluorescence change (C50) of wild-type (wt) recombinant (r) human protein C (PC) was 0.40 mM. The C50 for Ca2+ increased < 2.5-fold for the following r-PC variants (Gla is gamma-carboxyglutamic acid): [Gla6-->Asp]r-PC, [Gla7-->Asp]r-PC, [Gla14-->Asp]r-PC, [Gla19-->Asp]r-PC, or [Gla25-->Asp]r-PC, and approximately 4-6-fold for [Gla20-->Asp]r-PC and [Gla29-->Asp]r-PC. Much more dramatic increases in the C50 for Ca2+ were observed for [Gla16-->Asp]r-PC (> 75-fold) and [Gla26-->Asp]r-PC (ca. 30-fold). A substantially larger maximum fluorescence change (> 3-fold) as compared to that for wtr-PC, was also found in the case of the Ca2+/[Gla16-->Asp]r-PC complex, suggesting that the final Ca(2+)-induced conformation for this variant is dissimilar to that for wtr-PC and the above mutants. When a mutation was constructed at Arg15 ([Arg15-->Leu]r-PC), a residue conserved in all Gla-containing coagulation proteins, no fluorescence alteration occurred upon addition of Ca2+. The C50 for Ca2+ for promotion of the binding of the Ca(2+)-dependent, Gla-domain-directed, conformational monoclonal antibodies, JTC-1 and JTC-3, to wtr-PC was 3.0 and 4.0 mM, respectively. A similar C50 value was found for [Gla25-->Asp]r-PC. In the case of each antibody, approximately 4-6-fold higher C50 values for Ca2+ were found for the mutants; [Gla14-->Asp]r-PC, [Gla19-->Asp]r-PC, and [Gla29-->Asp]r-PC. Ca2+ did not promote binding of either of these antibodies to the following variants; [Gla6-->Asp]r-PC, [Gla7-->Asp] r-PC, [Arg15-->Leu]r-PC, [Gla16-->Asp]r-PC, [Gla20-->Asp]r-PC, and [Gla26-->Asp]r-PC. The results of this study suggest that adoption of the Ca(2+)-dependent conformation of PC is greatly dependent upon the presence of specific essential Gla residues, particularly those, namely Gla16 and Gla26, shown in the crystal structure of the prothrombin Gla domain/Ca2+ complex to be involved with coordination of Ca2+ ions not exposed to the surface. Of similar importance is Arg15. On the other hand, Gla residues at positions 14 and 19 are much less important in directing this same conformation. This finding is readily reconciled with the above crystal structure, which shows that these latter 2 residues are mainly responsible for coordination of a surface-exposed Ca2+ that is present at the end of the Ca(2+)-ion channel.     

A metal ion-binding site in the kringle region of bovine prothrombin fragment 1.

45Ca(II) binding studies (equilibrium dialysis) on the kringle domain of bovine prothrombin fragment 1 were conducted using a mixture of peptides (residues 43-156 and 46-156) resulting from limited alpha-chymotryptic hydrolysis of fragment 1. Analysis of the Scatchard plot of these data indicates a single, low affinity Ca(II)-binding site to be present. Similar results were obtained from studies on the decarboxylated fragment 1 derivative, 10-gamma-MGlu-fragment 1. Acetylation of bovine fragment 1 in the absence of Ca(II) or Mg(II) ions results in the loss of the metal ion-promoted quenching of the intrinsic Trp fluorescence of the protein and the Ca(II)-mediated binding to phosphatidylserine/phosphatidylcholine (PS/PC) vesicles. The acetylation of the NH2 alpha-group of Ala-1 has been shown (Welsch, D. J., and Nelsestuen, G. L. (1988) Biochemistry 27, 4946-4952) to abolish the PS/PC binding property of fragment 1. The present study demonstrates that acetylation of a second site possibly Ser-79 or Thr-81 using the conditions described in the preceding paper results in loss of both the fluorescence transition and the Ca(II)-mediated PS/PC binding of the resulting protein derivative. Removal of the O-acetyl group at the Ser-79/Thr-81 site is accomplished by aminolysis with 0.2 M hydroxylamine, pH 10, 50 degrees C; the fluorescence transition is partially restored. PS/PC binding is partially restored if the NH2 alpha-group of Ala-1 is trinitrophenylated but is not restored if the NH2 alpha-group of Ala-1 is acetylated. We conclude that the Ser-79/Thr-81 site may represent a portion of the metal ion-binding site within the kringle domain of fragment 1. Occupancy of this site by a Ca(II) ion appears to be important in the binding of the protein to PS/PC vesicles.     

Modifications of bovine prothrombin fragment 1 in the presence and absence of Ca(II) ions. Loss of positive cooperativity in Ca(II) ion binding for the modified proteins.

Chemical modification of bovine prothrombin fragment 1 according to the procedure of D. J. Welsch and G. L. Nelsestuen (1988) [Biochemistry 27, 4946-4952 and ealier papers] provided a series of fragment 1 derivatives in which various nitrogen-containing side chains were N-acetylated and/or N-2,4,6-trinitrophenylated. In addition the des-[Ala-1,Asn-2]- and des-[Ala-1,Asn-2,Lys-3]-fragment 1 derivatives were prepared by limited enzymatic hydrolysis of fragment 1 using cathepsin C and plasmin, respectively. Quantitative studies on the Ca(II) binding of these proteins have been accomplished using 45Ca(II) equilibrium dialysis. Binding of these fragment 1 derivatives to phosphatidylserine/phosphatidylcholine (PS/PC) vesicles (25:75) in the presence of Ca(II) ions has been studied using the light-scattering technique. Acylation of the 5 lysine residues of fragment 1 by the action of acetic anhydride (500-fold molar excess) in the presence of 75 mM Ca(II), pH 8.0, results in loss of positive cooperativity in Ca(II) binding (Scatchard plot) and an increase in the number of Ca(II) ions bound. The Ca(II)-dependent PS/PC binding of the acylated protein is reduced. Removal of 2 and 3 residues from the amino terminus likewise leads to loss of positive cooperativity in Ca(II) binding and reduced binding affinity to PS/PC vesicles. The important role of the amino-terminal 1-10 sequence is discussed. We conclude that positive cooperativity in Ca(II) binding is not a prerequisite for the Ca(II)-dependent binding of bovine prothrombin fragment 1 to PS/PC vesicles.     

Proteolytic formation and properties of a fragment of protein C containing the gamma-carboxyglutamic acid rich domain and the EGF-like region

The function of the epidermal growth factor (EGF) like domains in the vitamin K dependent plasma proteins is largely unknown. In order to elucidate the function of these domains in protein C, we have devised a method to isolate the EGF-like region from the light chain connected to the NH2-terminal region, containing the gamma-carboxyglutamic acid (Gla) residues. This was accomplished by tryptic cleavage of protein C that had been reversibly modified with citraconic anhydride to prevent cleavage at the lysine residue (in position 43) that is located between the two regions. The isolated fragment consists of residues 1-143 from the light chain of protein C connected by a disulfide bond to residues 108-131 from the heavy chain. Upon Ca2+ binding to the isolated Gla-EGF fragment from bovine protein C, the tryptophan fluorescence emission was quenched in a manner indicating binding to at least two classes of binding sites. These were presumably the Gla-independent Ca2(+)-binding site located in the EGF-like region and the lower affinity sites in the Gla region. A comparison with the tryptophan fluorescence quenching that occurred upon Ca2+ binding to the separately isolated EGF-like and Gla regions suggested that the EGF-like region influenced the structure and Ca2+ binding of the Gla region. The isolated Gla-EGF fragment functioned as an inhibitor of the anticoagulant effect of activated protein C in a clotting assay, whereas no inhibition was observed with either the Gla region or the EGF-like region.     

Structural and functional characteristics of activated human factor IX after chemical modification of gamma-carboxyglutamic acid residues

Activated human factor IX (factor IXa) was treated under mildly acidic conditions with a mixture of formaldehyde and morpholine. This reagent has been shown to react preferentially with gamma-carboxyglutamyl (Gla) residues and to convert these residues to gamma-methyleneglutamyl residues (Wright, S.F., Bourne, C.D., Hoke, R.A., Koehler, K.A., and Hiskey, R.G. (1984) Anal. Biochem. 139, 82-90). The modified enzyme was evaluated for coagulant activity and calcium-dependent fluorescence quenching. [14C]Formaldehyde was employed to allow quantitation of the modification and to facilitate localization of the modified residues in the primary structure of factor IXa. In the presence of the [14C]formaldehyde/morpholine reagent, factor IXa rapidly lost coagulant activity, which corresponded to incorporation of radiolabel. Examination of the relationship between protein modification (radiolabel incorporation) and the loss of coagulant activity suggested that modification of 1 mol of Gla/mol of factor IXa results in complete loss of factor IXa coagulant activity. Primary structure analysis of the radioactivity labeled factor IXa suggested that modification of any one of 11 Gla residues was responsible for the loss of coagulant activity. In the presence of calcium, modified factor IXa exhibited a smaller Gla-dependent decrease in protein fluorescence than native factor IXa, but the Gla-independent fluorescence change was the same for both proteins. It therefore appears that the Gla domain of factor IXa must be completely intact for the enzyme to undergo a functionally important calcium-dependent conformational change necessary for coagulant activity.     

The mechanism of action of GTP on Ca2+ efflux from rat liver microsomal vesicles.

1. Guanosine 5'-[gamma-thio]triphosphate (GTP[S]), if added before GTP, blocks both Ca2+ efflux promoted by GTP and the effect of GTP on enhancement of inositol 1,4,5-triphosphate (IP3)-promoted Ca2+ release from preloaded microsomal vesicles. If, however, GTP[S] is added after GTP, it does not reverse the Ca2+ efflux promoted by GTP, nor does it inhibit IP3-promoted Ca2+ release. 2. The effect of GTP in enhancing IP3-promoted Ca2+ release is maintained after washing the microsomal vesicles free of added GTP. After this treatment, enhancement of IP3-promoted Ca2+ efflux can be observed in the absence of poly(ethylene glycol). 3. Electron microscopy shows that during GTP treatment of microsomal vesicles there is rapid production of very large vesicular structures, apparently produced by fusion of smaller vesicles. 4. Light-scattering changes are detectable during the fusion process. 5. Both Ca2+ efflux promoted by GTP and the enhancement of IP3-promoted Ca2+ release seen in the presence of GTP can probably be attributed to GTP-dependent vesicle fusion.     

Comparison of lipid binding and kinetic properties of normal, variant, and gamma-carboxyglutamic acid modified human factor IX and factor IXa

The abilities of normal and three abnormal factor IXa molecules to activate factor X and to bind to phospholipid membranes have been compared to define the contributions of protein-lipid interactions and factor IXa light chain-heavy chain interactions to the functioning of this protein. The abnormal proteins studied had altered amino acid residues in their light chains. The heavy-chain regions, containing the active site serine and histidine residues, were normal in the abnormal proteins on the basis of titration by antithrombin III. The binding constants (Kd) for normal (N), variant [Chapel Hill (CH) and Alabama (AL)], and gamma-carboxyglutamic acid (Gla) modified (MOD) factors IX and IXa to phosphatidylserine (PS)/phosphatidylcholine (PC) small, unilamellar vesicles (SUV) were measured by 90 degrees light scattering. The Kd values for factor IXN binding were quite sensitive to the PS content of the membrane but less sensitive to Ca2+ concentrations between 0.5 and 10 mM. The zymogen and activated forms of both normal and abnormal factor IX bound with similar affinities to PS/PC (30/70) SUV. In the cases of factor IXaN and factor IXaAL, but not factor IXaCH or factor IXaMOD, irreversible changes in scattering intensity suggested protein-induced vesicle fusion. Since the activation peptide is not released from factor IXaCH, the normal interaction of factor IXa with a membrane must require the release of the activation peptide and the presence of intact Gla residues. The rate of factor X activation by normal and abnormal factor IXa was obtained by using a chromogenic substrate for factor Xa in the presence of PS/PC (30/70) SUV and 5 mM Ca2+.     

Calcium-dependent alpha-helical structure in osteocalcin

Osteocalcin is an abundant Ca2+-binding protein of bone containing three residues of vitamin K dependent gamma-carboxyglutamic acid (Gla) among its 49 (human, monkey, cow) or 50 (chicken) amino acids. Gla side chains participate directly in the binding of Ca2+ ions and the adsorption of osteocalcin to hydroxylapatite (HA) surfaces in vivo and in vitro. Osteocalcin exhibits a major conformational change when Ca2+ is bound. Metal-free chicken osteocalcin is a random coil with only 8% of its residues in the alpha helix as revealed by circular dichroism. In the presence of physiological levels of Ca2+, 38% of the protein adopts the alpha-helical conformation with a transition midpoint at 0.75 mM Ca2+ in a rapid, reversible fashion which (1) requires an intact disulfide bridge, (2) is proportionally diminished when Gla residues are decarboxylated to Glu, (3) is insensitive to 1.5 m NaCl, and (4) can be mimicked by other cations. Tyr fluorescence, UV difference spectra, and Tyr reactivity to tetranitromethane corroborate the conformational change. Homologous monkey osteocalcin also exhibits Ca2+-dependent structure. Integration of predictive calculations from osteocalcin sequence has yielded a structural model for the protein, the dominant features of which include two opposing alpha-helical domains of 9-12 residues each, connected by a bea turn and stabilized by the Cys23-Cys29 disulfide bond. Cation binding permits realization of the full alph a-helical potential by partial neutralization of high anionic charge in the helical domains. Periodic Gla occurrence at positions 17, 21, and 24 has been strongly conserved throughout evolution and places all Gla side chains on the same face of one alpha helix spaced at intervals of approximately 5.4 A, closely paralleling the interatomic separation of Ca2+ in the HA lattice. Helical osteocalcin has greatly increased affinity for HA; thus, the Ca2+-induced structural transition may perform an informational role related to bone metabolism.     

The role of hydrophobic interactions in the phospholipid-dependent activation of protein kinase C.

The effects of hydrophobic interaction on the activation of Ca2+-stimulated phospholipid-dependent protein kinase (protein kinase C), isolated from mouse brain, by phosphatidylserine (PS) and diacylglycerol (DAG) or phorbol 12-myristate 13-acetate were studied. To maintain bilayer structure during assay conditions, phosphatidylcholine was added to the PS vesicles. The vesicular structure of all types of PS was confirmed by freeze-fracture electron microscopy. The PS-dependent activation of purified protein kinase C from mouse brain is affected by the fatty acid composition of PS: an inverse relationship between the unsaturation index of PS (isolated from bovine heart, bovine spinal cord or bovine brain) and the ability to activate protein kinase C was demonstrated. In highly saturated PS lipid dispersions, only slight additional activation of protein kinase C by DAG was found, in contrast with highly unsaturated PS lipid dispersion, where DAG increased protein kinase C activity by 2-3-fold at optimal PS concentrations. We quantified the formation of the protein kinase C-Ca2+-PS-phorbol ester complex by using [3H]phorbol 12,13-dibutyrate [( 3H]PDBu). The efficiency of complex-formation, determined as the amount of [3H]PDBu bound, is not affected by variations in the hydrophobic part of PS. These results indicate a role of the hydrophobic part of the activating phospholipid in the activation mechanism of protein kinase C and in the action of cofactors.     

Membrane-bound orientation and position of the synaptotagmin C2B domain determined by site-directed spin labeling

Site-directed spin labeling is used to determine the orientation and depth of insertion of the second C2 domain from synaptotagmin I (C2B) into membrane vesicles composed of phosphatidylcholine (PC) and phosphatidylserine (PS). EPR line shapes of spin-labeled mutants located with the Ca(2+)-binding loops of C2B broaden in the presence of Ca(2+) and PC/PS vesicles, indicating that these loops undergo a Ca(2+)-dependent insertion into the membrane interface. Power saturation of the EPR spectra provides a position for each spin-labeled site along the bilayer normal, and these EPR-derived distance constraints, along with a high-resolution structure of the C2B domain, are used to generate a model for the domain orientation and position at the membrane interface. Our data show that the isolated C2B domain from synaptotagmin I penetrates PC/PS membranes, and that the backbone of Ca(2+)-binding loops 1 and 3 is inserted below the level of a plane defined by the lipid phosphates. The side chains of several loop residues are within the bilayer interior, and both Ca(2+)-binding sites are positioned near a plane defined by the lipid phosphates. A Tb(3+)-based fluorescence assay is used to compare the membrane affinity of the C2B domain to that of the first synaptotagmin C2 domain (C2A). Both C2A and C2B bind PC/PS (75:25) membrane vesicles with a micromolar lipid affinity in the presence of metal ion. These results indicate that C2A and C2B have a similar membrane affinity and position when bound to PC/PS (75:25) membrane vesicles. EPR spectroscopy indicates that the C2B domain has different interactions with PC/PS membranes containing 1 mol % phosphatidylinositol 4,5-bisphosphate.     

Ca2+-binding proteins from bovine brain including a potent inhibitor of protein kinase C.

We have previously described the use of Ca2+-dependent hydrophobic-interaction chromatography to isolate the Ca2+ + phospholipid-dependent protein kinase (protein kinase C) and a novel heat-stable 21 000-Mr Ca2+-binding protein from bovine brain [Walsh, Valentine, Ngai, Carruthers & Hollenberg (1984) Biochem. J. 224, 117-127]. The procedure described for purification of the 21 000-Mr calciprotein to electrophoretic homogeneity has been modified to permit the large-scale isolation of this Ca2+-binding protein, enabling further structural and functional characterization. The 21 000-Mr calciprotein was shown by equilibrium dialysis to bind approx. 1 mol of Ca2+/mol, with apparent Kd approx. 1 microM. The modified large-scale purification procedure revealed three additional, previously unidentified, Ca2+-binding proteins of Mr 17 000, 18 400 and 26 000. The 17 000-Mr and 18 400-Mr Ca2+-binding proteins are heat-stable, whereas the 26 000-Mr Ca2+-binding protein is heat-labile. Use of the transblot/45CaCl2 overlay technique [Maruyama, Mikawa & Ebashi (1984) J. Biochem. (Tokyo) 95, 511-519] suggests that the 18 400-Mr and 21 000-Mr Ca2+-binding proteins are high-affinity Ca2+-binding proteins, whereas the 17 000-Mr Ca2+-binding protein has a relatively low affinity for Ca2+. Consistent with this observation, the 18 400-Mr and 21 000-Mr Ca2+-binding proteins exhibit a Ca2+-dependent mobility shift on sodium dodecyl sulphate/polyacrylamide-gel electrophoresis, whereas the 17 000-Mr Ca2+-binding protein does not. The amino acid compositions of the 17 000-Mr, 18 400-Mr and 21 000-Mr Ca2+-binding proteins show some similarities to each other and to calmodulin and other members of the calmodulin superfamily; however, they are clearly distinct and novel calciproteins. In functional terms, none of the 17 000-Mr, 18 400-Mr or 21 000-Mr Ca2+-binding proteins activates either cyclic nucleotide phosphodiesterase or myosin light-chain kinase, both calmodulin-activated enzymes. However, the 17 000-Mr Ca2+-binding protein is a potent inhibitor of protein kinase C. It may therefore serve to regulate the activity of this important enzyme at elevated cytosolic Ca2+ concentrations.     

Phospholipid binding properties of bovine prothrombin peptide residues 1-45

The present study investigates the unique contribution of the NH2-terminal 33 residues of prothrombin, the gamma-carboxyglutamic acid (Gla) domain, to the Ca(II) and phospholipid-binding properties of prothrombin. Two Gla domain peptides, 1-42 and 1-45, produced by chymotryptic cleavage of prothrombin fragment 1 (residues 1-156 of the amino terminus of bovine prothrombin) and isolated by anion-exchange chromatography were utilized to characterize the Gla domain of prothrombin. This investigation utilized several experimental approaches to examine the properties of the Gla domain peptides. These studies were somewhat hampered by the metal ion-induced insolubility of the peptides. However, the 1-45 peptide was specifically radioiodinated, which facilitated the study of this peptide at low concentrations. In contrast to prothrombin fragment 1, the intrinsic fluorescence of both 1-42 and 1-45 was not quenched upon the addition of 1 mM Ca(II) or any concentration of Mg(II). Equilibrium dialysis studies revealed that the 1-42 peptide bound three Ca(II) ions noncooperatively, whereas fragment 1 binds seven Ca(II) ions in a positive cooperative manner. Ca(II)-promoted conformational changes are observed by comparison of electrophoretic mobility changes in the presence of increasing Ca(II) concentrations. Prothrombin, fragment 1, and the Gla domain peptides 1-42 and 1-45 exhibited similar electrophoretic mobility behavior in the presence of Ca(II) ions. The radiolabeled 1-45 peptide was found to comigrate with phospholipid vesicles on gel permeation chromatography in the presence of Ca(II). Fragment 1 was shown to inhibit this Ca(II)-dependent phospholipid binding of 1-45, demonstrating that the 1-45 peptide does possess the necessary phospholipid-binding structure. Furthermore, a metal ion-dependent conformational monoclonal antibody, F9.29, was inhibited from binding fragment 1 by the 1-42 peptide.     

Biological activities of bovine cardiac-muscle troponin C C-terminal peptide (residues 84-161).

1. Bovine cardiac-muscle troponin C was digested at cysteine residues 35 and 84, and the C-terminal peptide (residues 84-161) was isolated. 2. The C-terminal peptide contains two Ca2+-binding sites. These sites bind Ca2+ with a binding constant of 2.0 X 10(8) M-1. In the presence of 2 mM-Mg2+ the binding constant for Ca2+ is decreased to 3.7 X 10(7) M-1. The corresponding constants for native troponin C are 5.9 X 10(7) M-1. and 2.9 X 10(7) M-1 respectively. 3. Electrophoretic mobility of the C-terminal peptide is increased in the presence of 0.1 mM-CaCl2 as compared with the mobility in the presence of 2mM-EDTA. The same phenomenon was observed when electrophoresis was performed in the presence of 6 M-urea or 0.1% sodium dodecyl sulphate. 4. When saturated with Ca2+, the C-terminal peptide forms complexes with bovine cardiac-muscle troponin I both in the absence and in the presence of 6 M-urea. This complex is dissociated on removal of Ca2+. 5. The data suggest that the C-terminal peptide of troponin C contains two Ca2+/Mg2+-binding sites and interacts with troponin I. Thus, despite the 30% difference in amino acid composition, the properties of bovine cardiac-muscle troponin C C-terminal peptide are similar to those of rabbit skeletal-muscle troponin C C-terminal peptide.     

A single amino acid replacement results in the Ca2+-induced self-assembly of a helical conantokin-based peptide

Conantokins are short (17-27 amino acid residues), gamma-carboxyglutamate (Gla)-rich peptide components of the venoms of marine snails of the genus Conus. They display high apo and/or Ca(2+)-induced helicity and act as potent and selective inhibitors of the N-methyl-d-aspartate receptor (NMDAR). We have previously established that one of the conantokins, conantokin-G (con-G), self-associates in the presence of Ca(2+) with high specificity for antiparallel chain orientation [Dai, Q., Prorok, M., and Castellino, F. J. (2004) J. Mol. Biol. 336, 731-744]. The dimerization appears to be driven by interhelical Ca(2+) coordination between the following residue pairings: Gla(3)-Gla(14)('), Gla(7)-Gla(10)('), Gla(10)-Gla(7)('), and Gla(14)-Gla(3)('). A second member of the conantokin family, conantokin-T (con-T), shares sequence identity with con-G at 8 of 21 amino acids, including 4 Gla residues. These similarities notwithstanding, several primary and secondary structural differences exist between con-T and con-G. Particularly notable is that con-T contains a Lys, rather than a Gla, at position 7. Moreover, unlike con-G, con-T does not undergo Ca(2+)-triggered self-assembly. In the present study, sedimentation equilibrium ultracentrifugation is employed to demonstrate that a single amino acid replacement analogue of con-T, con-T[K7gamma], assumes a dimeric superstructure in the presence of Ca(2+) at pH values consistent with the ionization of Gla carboxylate groups. Furthermore, HPLC-monitored thiol-disulfide folding and rearrangement assays with Cys-containing con-T variants suggest that the relative chain alignment preference in the noncovalent complex is antiparallel. Our results suggest that interchain Ca(2+) coordination in con-T[K7gamma] is incumbent upon an "i, i + 4, i +7, i +11" arrangement of Gla residues, as occurs in native con-G.     

Interaction of human gamma-carboxyglutamic acid domainless prothrombin with phospholipids

The conversion of prothrombin into thrombin occurs at the surface of stimulated platelets. In order to see the influence of gamma-carboxyglutamic acid in the interaction of prothrombin with phospholipid, we investigated the direct interaction of the peptide 1-41 and prothrombin (des 1-44) with phospholipid monolayers of various compositions. Adsorption of the labeled proteins was determined by surface radioactivity measurements. Penetration of the proteins into the lipid layers was inferred from capacitance variation of the monolayer, measured by a.c. polarography. Prothrombin (des 1-44) was found both to adsorb and to penetrate the lipid monolayers, in the presence and in the absence of Ca++. The effects are higher on 100% PS than on 25% PS. This protein was also found to increase the permeability of vesicles containing 25% PS to T1+ ions, in the presence and in the absence of Ca++. Comparison with prothrombin shows that Gla residues are clearly involved in the interaction at 25% PS; nevertheless, the peptide 1-41 does not penetrate. A model of interaction of prothrombin with phospholipid, including both adsorption of prothrombin by Gla residues and its penetration by another domain, is proposed.     

ENDOR studies of VO2+: probing protein-metal ion interactions in nephrocalcin.

Nephrocalcin inhibits the growth of calcium oxalate monohydrate crystals in the mammalian kidney. Isoforms A and B contain three equivalents of gamma-carboxyglutamic acid (Gla) residues implicated in Ca2+-binding and exhibit strong inhibitor properties and high Ca2+-binding affinity (Kd approximately 10(-8) M). Isoforms C and D lack these properties and exhibit low Ca2+-binding affinity (Kd approximately 10(-6) M). With VO2+ as a structural probe, electron paramagnetic resonance (EPR) studies of the Ca2+-binding sites of isoforms B and D showed that VO2+ binds competitively with a metal ion:protein stoichiometry of 4:1. EPR spectral parameters of the VO2+ ion were consistent with only equatorial oxygen-donor ligands. EPR and angle-selected electron nuclear double resonance (ENDOR) spectra showed two equatorially positioned, metal coordinating waters in isoform D while in isoform B no ligands undergoing hydrogen exchange were found. Since isoform D showed no evidence for axially coordinated water, similarly to isoform B, it is likely that the protein residues occupying the axial sites are identical in both proteins. ENDOR spectra of VO2+-complexes of isoforms B and D were compared to spectra of the VO2+-complex with alpha-ethylmalonic acid (EMA), a molecular mimic of Gla. Spectra of the VO2+-complex of EMA showed axial water located trans to the V=O bond and outer shell water hydrogen-bonded to the vanadyl oxygen, consistent with the X-ray structure of Ca(EMA)2. We, therefore, conclude that the spatial disposition of carboxylate groups of Gla residues coordinating Ca2+ in isoforms A and B must differ from that observed in the crystal structure of Ca(EMA)2.     

Using protein design to dissect the effect of charged residues on metal binding and protein stability

Ca2+ controls biological processes by interacting with proteins with different affinities, which are largely influenced by the electrostatic interaction from the local negatively charged ligand residues in the coordination sphere. We have developed a general strategy for rationally designing stable Ca2+- and Ln3+-binding proteins that retain the native folding of the host protein. Domain 1 of cluster differentiation 2 (CD2) is the host for the two designed proteins in this study. We investigate the effect of local charge on Ca2+-binding affinity based on the folding properties and metal-binding affinities of the two proteins that have similarly located Ca2+-binding sites with two shared ligand positions. While mutation and Ca2+ binding do not alter the native structure of the protein, Ca2+ binding specifically induced changes around the designed Ca2+-binding site. The designed protein with a -5 charge at the binding sphere displays a 14-, 20-, and 12-fold increase in the binding affinity for Ca2+, Tb3+, and La3+, respectively, compared to the designed protein with a -3 charge, which suggests that higher local charges are preferred for both Ca2+ and Ln3+ binding. The localized charged residues significantly decrease the thermal stability of the designed protein with a -5 charge, which has a T(m) of 41 degrees C. Wild-type CD2 has a T(m) of 61 degrees C, which is similar to the designed protein with a -3 charge. This decrease is partially restored by Ca2+ binding. The effect on the protein stability is modulated by the environment and the secondary structure locations of the charged mutations. Our study demonstrates the capability and power of protein design in unveiling key determinants to Ca2+-binding affinity without the complexities of the global conformational changes, cooperativity, and multibinding process found in most natural Ca2+-binding proteins.     

Derivatives of blood coagulation factor IX contain a high affinity Ca2+-binding site that lacks gamma-carboxyglutamic acid

We have examined the calcium-binding properties and metal ion-dependent conformational changes of proteolytically modified derivatives of factor IX that lack gamma-carboxyglutamic acid (Gla) residues. Equilibrium dialysis experiments demonstrated that a Gla-domainless factor IX species retained a single high affinity calcium ion-binding site (Kd = 85 +/- 5 microM). Ca2+ binding to this site was accompanied by a decrease in intrinsic fluorescence emission intensity (Kd = 63 +/- 15 microM). These spectral changes were reversed upon the addition of EDTA. Titration with Sr2+ resulted in little change in fluorescence intensity below 1 mM, while titration with Tb3+ caused fluorescence changes similar to those observed with Ca2+. Tb3+ and Ca2+ appear to bind to the same site because tryptophan-dependent terbium emission was reduced by the addition of Ca2+. Similar results were obtained with a Gla-domainless factor IX species lacking the activation peptide. Gla domain-containing factor IX species exhibited fluorescence changes similar to those of the Gla-domainless proteins at low Ca2+, but an additional structural transition was found at higher Ca2+ concentrations (apparent Kd greater than 0.8 mM). Thus, the conformations of factor IX proteins are nucleated and/or stabilized by calcium binding to a high affinity site which does not contain Gla residues. The binding of Ca2+ to lower affinity Gla domain-dependent metal ion-binding sites elicits an additional conformational change. The strong similarities between these results and those obtained with protein C (Johnson, A. E., Esmon, N. L., Laue, T. M. & Esmon, C. T. (1983) J. Biol. Chem. 258, 5554-5560), coupled with the remarkable sequence homologies of the vitamin K-dependent proteins, suggest that the high affinity Gla-independent Ca2+-binding site may be a common feature of vitamin K-dependent proteins.     

Reconstitution of rabbit thrombomodulin into phospholipid vesicles

The influence of phospholipid on thrombin-thrombomodulin-catalyzed activation of protein C has been studied by incorporating thrombomodulin into vesicles by dialysis from octyl glucoside-phospholipid mixtures. Thrombomodulin was incorporated into vesicles ranging from neutral (100% phosphatidylcholine) to highly charged (30% phosphatidylserine and 70% phosphatidylcholine). Thrombomodulin is randomly oriented in vesicles of different phospholipid composition. Incorporation of thrombomodulin into phosphatidylcholine, with or without phosphatidylserine, alters the Ca2+ concentration dependence of protein C activation. Soluble thrombomodulin showed a half-maximal rate of activation at 580 microM Ca2+, whereas half-maximal rates of activation of liposome-reconstituted thrombomodulin were obtained between 500 microM Ca2+ and 2 mM Ca2+, depending on the composition (protein:phospholipid) of the liposomes. The Ca2+ dependence of protein C activation fits a simple hyperbola for the soluble activator, while the Ca2+ dependence of the membrane-associated complex is distinctly sigmoidal with a Hill coefficient greater than 2.4. In contrast, the Ca2+ dependence of gamma-carboxyglutamic acid (Gla) domainless protein C activation is unchanged by membrane reconstitution (1/2 max = 53 +/- 10 microM) and fits a simple rectangular hyperbola. Incorporation of thrombomodulin into pure phosphatidylcholine vesicles reduces the Km for protein C from 7.6 +/- 2 to 0.7 +/- 0.2 microM. Increasing phosphatidylserine to 20% decreased the Km for protein C further to 0.1 +/- 0.02 microM. Membrane incorporation has no influence on the activation of protein C from which the Gla residues are removed proteolytically (Km = 6.4 +/- 0.5 microM). The Km for protein C observed on endothelial cells is more similar to the Km observed when thrombomodulin (TM) is incorporated into pure phosphatidylcholine vesicles than into negatively charged vesicles, suggesting that the protein C-binding site on endothelial cells does not involve negatively charged phospholipids. In support of this concept, we observed that prothrombin and fragment 1, which bind to negatively charged phospholipids, do not inhibit protein C activation on endothelial cells or TM incorporated into phosphatidylcholine vesicles, but do inhibit when TM is incorporated into phosphatidylcholine:phosphatidylserine vesicles. These studies suggest that neutral phospholipids lead to exposure of a site, probably on thrombomodulin, capable of recognizing the Gla domain of protein C.