Measurement of active coagulation factors in Autoplex-T with colorimetric active site-specific assay technology.

Autoplex-T is a partially activated prothrombin complex concentrate used primarily for the treatment of patients expressing factor VIII inhibitors. While Autoplex-T has a demonstrated record of clinical effectiveness, the procoagulant composition of this material has not been reported. This absence of composition data is a reflection of the lack of techniques appropriate for accurately measuring an individual protease such as factor IXa in complex mixtures of similar proteases. The development of Colorimetric Active Site-Specific ImmunoAssay technology (CASSIA) has permitted the accurate analysis of the coagulant enzymes present in Autoplex-T. Ten lots of Autoplex-T were reacted with both biotinylated phenylalanylprolylarginine chloromethylketone and biotinylated glutamylglycylarginine chloromethylketone. Only activated forms of the clotting factors present in Autoplex-T react with the peptide chloromethylketones and were thus separated from the other proteins present in Autoplex-T by adsorption onto streptavidin. The individual proteins bound to streptavidin were then detected with specific antibodies. Mean results from the analysis of ten lots of Autoplex-T (mean values) are as follows: factor Xla, 5.9 nM or 0.8 microg/ml; factor Xa, 46.5 nM or 2.1 microg/ml; factor IXa, 177.8 nM or 11.7 microg/ml; factor VIIa, 68.6 nM or 3.3 microg/ml and factor IIa, 5.3 nM or 0.2 microg/ml. These results are discussed with respect to the mechanism of action of Autoplex-T in the treatment of factor VIII inhibitor patients.

Tertiary structure of the hepatic cell protein fibrinogen in fluid revealed by atomic force microscopy.

Fibrinogen participates in important cellular physiological processes, such as cell adhesion and blood clotting. Although the primary and secondary structures of fibrinogen are known, its tertiary structure is yet to be determined. In attempts to understand the tertiary structure of this important hydrated cellular and plasma membrane protein, the present study using atomic force microscopy was carried out. The techniques presented in this manuscript may also be applicable to enhance the imaging of live cells as well as their subcellular components. The authors have imaged fibrinogen by Tapping Mode atomic force microscopy in fluid. Purified human fibrinogen, together with 15-nm colloidal gold particles serving as an internal calibration standard, were adhered to a poly-L-lysine substrate on freshly cleaved mica. Atomic force microscopy images were obtained using oxide-sharpened silicon nitride probes, either unaltered or with an electron beam deposited extended tip. Although various structures were observed, the predominant forms consisted of a bi- or trinodular slightly curved linear shape. Approximately 300 of these structures were observed with six different tips (1 unaltered and 5 electron beam deposited) and their lengths and heights were analyzed. The mean length of the fibrinogen molecules was 65.8 nm and the mean height was 3.4 nm. The quantitative measurements were little influenced by the shape of the tip, whereas the sharper electron beam deposited tips seemed to produce qualitatively superior images.

Evidence that meizothrombin is an intermediate product in the clotting of whole blood.

Meizothrombin is an intermediate that is produced during the conversion of prothrombin to thrombin in systems composed of purified factor Xa and factor Va that are quantitatively assembled on an anionic phospholipid surface. The biological significance of this intermediate has recently been challenged by the apparent absence of meizothrombin during clotting of sodium citrate-anticoagulated plasma. We analyzed the formation of thrombin during coagulation of nonanticoagulated, unchilled, minimally manipulated whole blood in glass tubes. The process was stopped at 0, 3, 5, and 7 minutes by the addition of biotinylated peptidyl chloromethyl-ketone active-site labeling reagents. Plasma/serum was separated by centrifugation, and labeled species were extracted by immunoadsorption with a polyclonal anti-prothrombin antibody. The purified prothrombin-derived species were separated by SDS-polyacrylamide gradient gel electrophoresis and visualized on a chemiluminescent avidin blot. Meizothrombin appeared as an intermediate product of this reaction and persisted with some increase through the 7-minute time point. We also observed incorporation of the active-site label into a species of lower molecular weight consistent with the B1 chain of beta- and/or gamma-thrombin. These degraded forms of thrombin have not been previously demonstrated in a biologically relevant preparation. Our data clearly establish the generation of meizothrombin as an intermediate product of thrombin generation during whole-blood clotting. The data also represent the first experimental evidence for the generation of beta- and gamma-thrombin in a biologically relevant environment and time scale.

Phosphorylation of factor Va and factor VIIIa by activated platelets.

Platelet activation leads to the incorporation of 32[PO4(2-)] into bovine coagulation factor Va and recombinant human factor VIII. In the presence of the soluble fraction from thrombin-activated platelets and (gamma-32P) adenosine triphosphate, radioactivity is incorporated exclusively into the M(r) = 94,000 heavy chain (H94) of factor Va and into the M(r) = 210,000 to 90,000 heavy chains as well into the M(r) = 80,000 light chain of factor VIII. Proteolysis of the purified phosphorylated M(r) = 94,000 factor Va heavy chain by activated protein C (APC) gave products of M(r) = 70,000, 24,000, and 20,000. Only the intermediate M(r) = 24,000 fragment contained radioactivity. Because the difference between the M(r) = 24,000 and M(r) = 20,000 fragments is located on the COOH-terminal end of the bovine heavy chain, phosphorylation of H94 must occur within the M(r) = 4,000 peptide derived from the carboxyl-terminal end of H94 (residues 663 through 713). Exposure of the radioactive factor VIII molecule to thrombin ultimately resulted in a nonradioactive light chain and an M(r) = 24,000 radioactive fragment that corresponds to the carboxyl-terminal segment of the A1 domain of factor VIII. Based on the known sequence of human factor VIII, phosphorylation of factor VIII by the platelet kinase probably occurs within the acidic regions 337 through 372 and 1649 through 1689 of the procofactor. These acidic regions are highly homologous to sequences known to be phosphorylated by casein kinase II. Results obtained using purified casein kinase II gave a maximum observed stoichiometry of 0.6 mol of 32[PO4(2-)]/mol of factor Va heavy chain and 0.35 mol of 32[PO4(2-)]/mol of factor VIII. Phosphoamino acid analysis of phosphorylated factor Va by casein kinase II or by the platelet kinase showed only the presence of phosphoserine while phosphoamino acid analysis of phosphorylated factor VIII by casein kinase II showed the presence of phosphothreonine as well as small amounts of phosphoserine. The platelet kinase responsible for the phosphorylation of the two cofactors was found to be inhibited by several synthetic protein kinase inhibitors. Finally, partially phosphorylated factor Va was found to be more sensitive to APC inactivation than its native counterpart. Our findings suggest that phosphorylation of factors Va and VIIIa by a platelet casein kinase II-like kinase may downregulate the activity of the two cofactors.

Identification and characterization of a phospholipid-binding site of bovine factor Va.

Coagulation factor Va is a cofactor which combines with the serine protease factor Xa on a phospholipid surface to form the prothrombinase complex. The phospholipid-binding domain of bovine factor Va has been reported to be located on the light chain of the molecule and more precisely on a fragment of Mr = 30,000 which is obtained after digestion of factor Va light chain by factor Xa. This proteolytic fragment is located in the NH2-terminal part of factor Va light chain (residues 1564-1765). In order to further characterize the lipid-binding domain of bovine factor Va, isolated bovine light chain was preincubated with synthetic phospholipid vesicles (75% phosphatidylcholine, 25% phosphatidylserine) and digested with trypsin, chymotrypsin, and elastase. Two peptide regions protected from proteolytic cleavage were identified and characterized from each proteolytic digestion. A comparison of the NH2-terminal sequence and amino acid composition of the two tryptic peptides with the deduced sequence of human factor V indicates a match with residues 1657-1791 of the light chain of human factor V for one peptide and residues 1546-1656 for the other peptide. When chymotrypsin or elastase were used for digestion, the NH2-terminal sequence of one peptide showed a match with residues 1667-1797 of the light chain, while the other peptide presented an NH2-terminal sequence identical with the previously described for the bovine factor Va light chain. When these peptides were assayed for direct binding to phospholipid vesicles, only the tryptic and the chymotryptic peptides covering the middle region of the A3 domain of the bovine factor Va light chain demonstrated an ability to interact with phospholipid vesicles. Thus, knowing that the factor Xa cleavage site on the factor Va light chain is located between residues 1765 and 1766 of the light chain this lipid-binding region of the bovine factor Va is further localized to amino acid residues 1667-1765.

Factor V: a prototype pro-cofactor for vitamin K-dependent enzyme complexes in blood clotting.

The relative abundance of factor V, factor X and prothrombin has enabled detailed analyses of the prothrombinase complex. Determination of the primary structure for factor V has provided the basis for examination of structure-function relationships. The imminent in vitro expression of recombinant factor V will provide the opportunity for site-specific mutagenesis and a verification of these structure-function relationships. A comparison of the physical properties and primary structures for factors V and VIII has revealed extensive similarities in these two cofactor proteins. This observation indicates that a direct application of the technology developed for the analysis of prothrombinase will lead to an equal understanding of the factor Xase complex. Whether similar relationships exist for other blood coagulation enzyme complexes remains to be determined.

Multiple forms of rat kidney L-arginine:glycine amidinotransferase.

The relative amount of L-arginine:glycine amidinotransferase (transamidinase) protein in kidneys from rats fed a complete purified diet with and without the addition of creatine and/or glycine was determined by a monoclonal antibody-immunosorbent inhibition assay. Kidneys from the creatine-fed rats had 10% of the transamidinase activities and 78% of the monoclonal antibody immunoreactive transamidinase protein as kidneys from the control rats. An excellent correlation between transamidinase activities and protein was reported previously when the amounts of enzyme protein were determined by immunotitration with polyclonal antibodies. One possible explanation for the contrasting results was that multiple forms of transamidinase are present in rat kidneys. If so, the monoclonal antibody may have recognized forms of the enzyme that were not decreased in amounts commensurate with the decrease in enzyme activities as a result of creatine feeding. Evidence is presented in this report that multiple forms of transamidinase are present in rat kidneys. The distribution of the isoelectric points of the individual forms of transamidinase in kidneys of the control rats appeared to be dissimilar from that in the creatine-fed rats. Therefore, an alteration in the distribution of the individual forms of the enzyme may be a factor in the alteration of transamidinase activities in creatine-fed rats.

Complete cDNA and derived amino acid sequence of human factor V.

cDNA clones encoding human factor V have been isolated from an oligo(dT)-primed human fetal liver cDNA library prepared with vector Charon 21A. The cDNA sequence of factor V from three overlapping clones includes a 6672-base-pair (bp) coding region, a 90-bp 5' untranslated region, and a 163-bp 3' untranslated region within which is a poly(A) tail. The deduced amino acid sequence consists of 2224 amino acids inclusive of a 28-amino acid leader peptide. Direct comparison with human factor VIII reveals considerable homology between proteins in amino acid sequence and domain structure: a triplicated A domain and duplicated C domain show approximately equal to 40% identity with the corresponding domains in factor VIII. As in factor VIII, the A domains of factor V share approximately 40% amino acid-sequence homology with the three highly conserved domains in ceruloplasmin. The B domain of factor V contains 35 tandem and approximately 9 additional semiconserved repeats of nine amino acids of the form Asp-Leu-Ser-Gln-Thr-Thr/Asn-Leu-Ser-Pro and 2 additional semiconserved repeats of 17 amino acids. Factor V contains 37 potential N-linked glycosylation sites, 25 of which are in the B domain, and a total of 19 cysteine residues.

Identification and isolation of vitamin K-dependent proteins by HPLC.

Six of the seven known vitamin K-dependent proteins found in plasma were chromatographed on a large-pore propylsilane column using aqueous trifluoroacetic acid/acetonitrile gradients. Prothrombin and Factor VII coeluted, the others were readily resolved. The technique has been used to monitor the purification of protein C and protein S using immobilized anti-protein S. Preliminary evidence is presented which is suggestive of the existence of additional vitamin K-dependent proteins in plasma.

Monoclonal antibodies to human vitamin K-dependent protein S.

Monoclonal antibodies to human protein S have been prepared using established hybridoma technology. One antibody was isolated that binds protein S only when Ca2+ is present; others bind antigen equally well in the presence or absence of EDTA. Other antibodies display a diminished affinity for protein S in the presence of EDTA. Purified immunoglobulins from cell lines displaying Ca2+ dependence were immobilized and used to purify protein S from fractions obtained by barium precipitation of citrated plasma, ammonium sulfate fractionation, and chromatography on diethylaminoethanol (DEAE)-Sephadex and dextran sulfate agarose. Essentially homogeneous protein S was isolated from the barium-citrate-adsorbed, 35% ammonium-sulfate-soluble proteins using a totally Ca2+-dependent antibody and EDTA elution. Protein S and several substances of higher mol wt were bound directly from plasma by a partially Ca2+-dependent antibody and were eluted partially with EDTA and NaCl and finally with NaSCN. The largest and most abundant of the high mol wt materials is likely protein S-complement C4b-binding protein complex. The immunoaffinity-isolated protein S was found to be indistinguishable from conventionally isolated protein S in terms of activity, sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) mobility, and by high-performance liquid chromatography (HPLC). These studies establish reagents that can probe the structure of protein S and isolate protein S in its free and complexed forms.