Expression and purification of recombinant serine protease domain of human coagulation factor XII in Pichia pastoris.

Human coagulation factor XII, the initiating factor in the intrinsic coagulation pathway, is critical for pathological thrombosis but not for hemostasis. Pharmacologic inhibition of factor XII is an attractive alternative in providing protection from pathologic thrombus formation while minimizing hemorrhagic risk. Large quantity of recombinant active factor XII is required for screening inhibitors and further research. In the present study, we designed and expressed the recombinant serine protease domain of factor XII in Pichia pastoris strain X-33, which is a eukaryotic expression model organism with low cost. The purification protocol was simplified and the protein yield was high (~20 mg/L medium). The purified serine protease domain of factor XII behaved homogeneously as a monomer, exhibited comparable activity with the human ßFXIIa, and accelerated clot formation in human plasma. This study provides the groundwork for factor XII inhibitors screening and further research.

The autoactivation of rabbit Hageman factor.

Proteolytic cleavage and activation of isolated, single chain, zymogen Hageman factor was observed in the presence of kaolin alone. The rate of cleavage of kaolin-bound Hageman factor was enhanced 50-fold by the presence of prekallikrein and high molecular weight kininogen. The two-chain 82,000 dalton form of activated Hageman factor (alpha-HF(a)) also cleaved kaolin- bound single-chain Hageman factor in a dose-dependent manner, yielding fragments of 28,000 and, 50,000 dahons under reducing conditions. Cleavage of kaolin-bound single-chain Hageman factor was not inhibited by preincubation with diisopropylfluorophosphate (12 mM) for 10 min, but long-term incubation of Hageman factor with diisopropylfluorophosphate (up to 48 h) resulted in inhibition of cleavage of kaolin-bound Hageman factor to an extent proportional to the inhibition of procoagulant Hageman factor activity. Hageman factor cleavage was maximal when the kaolin concentration was {approximately} 10-fold greater than the Hageman factor concentration (wt:wt), and was partially inhibited by high molecular weight kininogen. Kaolin-bound Hageman factor cleaved clotting factor XI in an amount which correlated with the extent of cleavage of the Hageman factor. These findings are compatible with the concept that single-chain Hageman factor and alpha- HF(a), are both capable of cleaving and activating kaolin-bound Hageman factor and that a close molecular association of kaolin-bound Hageman factor molecules is required for this reaction.

The fibrinolytic pathway of human plasma. Isolation and characterization of the plasminogen proactivator.

The conversion of the plasminogen proactivator to plasminogen activator by activated Hageman factor or its fragments has been recognized as an essential step in the conversion of plasminogen to plasmin. The plasminogen proactivator has been completely separated from prekallikrein and pre-PTA, two other proenzyme substrates of activated Hageman factor or its fragments. Plasminogen proactivator, free of any contaminating proteins as assessed by disc gel electrophoresis or isoelectric focusing, revealed a single band with an isoelectric point of 8.9 corresponding in position to the Hageman factor activatable material eluted from replicate unstained gels. After conversion of plasminogen proactivator by Hageman factor fragments to the plasminogen activator, the active site of the plasminogen activator is not inhibited by C1INH and is thus readily distinguished from that of kallikrein or PTA. The plasminogen activator is susceptible to inactivation by DFP while the plasminogen proactivator is not, as has been the case for esterases having a serine in the active site. Its interaction with plasminogen is inhibited by epsilon-aminocaproic acid.

Partial purification and characterization of contact activation cofactor.

The contact phase of intrinsic clotting involves Factor XI, Factor XII, Fletcher factor, and a fourth activity that we call contact activation cofactor (CAC). All four of these activities are reduced or absent in Dicalite-adsorbed plasma. A modified activated partial thromboplastin time assay for CAC has been defined by using a substrate of Dicalite-adsorbed plasma combined with partially purified sources of Factors XI and XII, and Fletcher factor. The following properties of CAC in plasma have been determined by using the assay: it is stable up to 60 min at 56 degrees C; gradually loses activity at 80 degrees C; is stable between pH 6 and 9; is precipitated by ammonium sulfate between 40% and 50% saturation; is slightly adsorbed by A1(OH)3; and is eluted from DEAE-cellulose after the major protein peaks. A purification procedure has been devised that separates CAC from other known clotting factors. Isolated CAC was less stable than CAC in plasma, but in the presence of dilute human serum albumin it retained full activity for 80 min at 56 degrees C. On gel filtration CAC had an apparent mol wt of 220,000 daltons. These properties are consistent with those described for Fitzgerald factor, which further supports the conclusion that CAC and Fitzgerald factor represent the same activity. Isolated CAC promoted the generation of activated Factor XI (XIa) in a mixture containing purified Factor XI, Factor XII, and kaolin. The amount of Factor XIa generated was proportional to the amount of added CAC. No time-consuming reaction between Factor XI or Factor XII and CAC could be demonstrated.

Isolation and properties of an abnormal Hageman factor (Factor XII) molecule in a cross-reacting material-positive hageman trait plasma.

We have previously described two unrelated individuals with homozygous Hageman trait (Factor XII deficiency) whose plasmas contained nonfunctional material immunologically indistinguishable from normal Hageman factor (HF). Abnormal HF from the plasma of one these subjects has now been purified to homogeneity, as judged by sodium dodecyl sulfate-polyacrylamide gel electrophoresis, alkaline disc gel electrophoresis, and immunoelectrophoresis. Purified abnormal HF had no clot-promoting activity, but showed the same specific antigenicity as purified normal HF by an immunoassay. The abnormal HF was of a single chain polypeptide with the same molecular weight (80,000) as normal HF and was positively stained by periodic acid-Schiff reagent. Both normal and abnormal HF had similar amino acid compositions and isoelectric points (pI 6.5 approximately 7.1). When 125I-labeled abnormal HF and 131I-labeled normal HF were mixed with normal plasma and exposed to glass, both HF underwent an identical pattern of cleavage, yielding 52,000- and 30,000-mol wt fragments. Similarly, abnormal HF was fragmented by trypsin in the same way as normal HF, but no prekallikrein-activating activity was generated after cleavage. [3H]Diisopropyl phosphorofluoridate was incorporated into a 29,000-mol wt fragment of the trypsin-cleaved normal HF, but not into that of the trypsin-cleaved abnormal HF. These data suggest that the molecular defect in this abnormal HF resides at or near the active site serine residue in the 30,000-mol wt part of the molecule.

Inactivation of factor XII active fragment in normal plasma. Predominant role of C-1-inhibitor.

To define the factors responsible for the inactivation of the active fragment derived from Factor XII (Factor XIIf ) in plasma, we studied the inactivation kinetics of Factor XIIf in various purified and plasma mixtures. We also analyzed the formation of 125I-Factor XIIf -inhibitor complexes by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). In purified systems, the bimolecular rate constants for the reactions of Factor XIIf with C-1-inhibitor, alpha 2-antiplasmin, and antithrombin III were 18.5, 0.91, and 0.32 X 10(4) M-1 min-1, respectively. Furthermore, SDS-PAGE analysis revealed that 1:1 stoichiometric complexes were formed between 125I-Factor XIIf and each of these three inhibitors. In contrast, kinetic and SDS-PAGE studies indicated that Factor XIIf did not react with alpha 1-antitrypsin or alpha 2-macroglobulin. The inactivation rate constant of Factor XIIf by prekallikrein-deficient plasma was 14.4 X 10(-2) min-1, a value that was essentially identical to the value predicted from the studies in purified systems (15.5 X 10(-2) min-1). This constant was reduced to 1.8 X 10(-2) min-1 when Factor XIIf was inactivated by prekallikrein-deficient plasma that had been immunodepleted (less than 5%) of C-1-inhibitor. In addition, after inactivation in normal plasma, 74% of the active 125I-Factor XIIf was found to form a complex with C-1-inhibitor, whereas 26% of the enzyme formed complexes with alpha 2-antiplasmin and antithrombin III. Furthermore, 42% of the labeled enzyme was still complexed with C-1-inhibitor when 125I-Factor XII was inactivated in hereditary angioedema plasma that contained 32% of functional C-1-inhibitor. This study quantitatively demonstrates the dominant role of C-1-inhibitor in the inactivation of Factor XIIf in the plasma milieu.

Anaphylactic release of a prekallikrein activator from human lung in vitro.

We have demonstrated the in vitro IgE-mediated release of a prekallikrein activator from human lung. The lung prekallikrein activator was partially purified by sequential chromatography on sulfopropyl-Sephadex, DEAE-Sephacel, and Sepharose 6B. Purified human prekallikrein was converted to its active form (kallikrein) by the lung protease. The generated kallikrein was shown to be biologically active; that is, it generates bradykinin from purified human high-molecular weight kininogen and also cleaves benzoyl-propyl-phenyl-arginyl-p-nitroanilide, a known synthetic substrate of kallikrein. The lung prekallikrein activator differs from the known physiologic activators of prekallikrein (the activated forms of Hageman factor) with respect to: (a) size (it has a mol wt of approximately 175,000); (b) synthetic substrate specificity (D-propyl/phenyl/arginyl-p-nitroanilide is a substrate for the activated forms of Hageman factor, but not the lung protease); (c) antigenic specificity (an anti-Hageman factor immunoadsorbent column did not remove significant amounts of the lung protease, while it removed most of the activity of activated Hageman factor fragments); and (d) inhibition profile (the lung proteases was not inhibited by corn trypsin inhibitor). This prekallikrein activator provides a physiologic mechanism by which prekallikrein can be directly activated during IgE-mediated reactions of the lung. While the role of this lung prekallikrein activator in immediate hypersensitivity reactions and in other inflammatory processes is not clear, it does represent a first and important interface between IgE-mediated reactions and the Hageman factor-dependent pathways of the inflammatory response.

The interaction of Hageman factor and immune complexes.

The possible interaction of Hageman factor from human or rabbit plasma with a variety of immunologic reactants was studied. Evidence of an interaction was not obtained and neither binding of radiolabeled Hageman factor to immune aggregates nor depletion of the Hageman factor from the supernate was observed. Cleavage of the labeled Hageman factor molecule into its 30,000 molecular weight-active fragments was not detectable after incubation with immune complexes. Isolated Hageman factor was far more sensitive to activation than Hageman factor in plasma or serum. There was no consistent activation of isolated Hageman factor by immunologic reactants as determined by conversion of prekallikrein to its enzymatic form or by shortening of the clotting time of factor XII-deficient plasma. A variety of immunologic stimuli were tested: (a) antigen-antibody complexes in soluble or precipitated form; (b) particulate antigen-antibody complexes, i.e., zymosan-anti-zymosan in which a surface was presented for activation; (c) human IgM-IgG and IgG-IgG (rheumatoid factor) complexes; (d) immune aggregates consisting of heat or bis-diazotized benzidine-aggregated myeloma proteins of all human immunoglobulin classes and subclasses: IgG(1,2,3,4), IgA, IgD, IgM, and IgE. Absorption with immune aggregates did not reduce the quantity of Hageman factor in solution, nor was the Hageman factor bound to the precipitates. The presence of plasma or serum with immune aggregates did not generate activity of the Hageman factor. The only preparations of immunoglobulins capable of activating Hageman factor were found to be contaminated with bacteria. These bacteria, upon isolation, activated Hageman factor.

Structural changes accompanying enzymatic activation of human Hageman factor.

The structure of Hageman factor, isolated from human plasma, was analyzed before and after enzymatic activation. The purified molecule is a single polypeptide chain of 80,000 molecular weight (mol wt) sedimenting at 4.5S. An amino acid analysis has been performed. The concentration of Hageman factor in normal human plasma was found to be 29 mug/ml with variation between individuals ranging from 15 to 47 mug/ml. Treatment of the molecule with kallikrein, plasmin, or trypsin resulted in cleavage at two primary sites, yielding fragments of 52,000, 40,000, and 28,000 mol wt. No further changes occurred in the fragments with subsequent reduction. Prekallikrein-activating ability was associated exclusively with the 28,000 moiety.

The binding and cleavage characteristics of human Hageman factor during contact activation. A comparison of normal plasma with plasmas deficient in factor XI, prekallikrein, or high molecular weight kininogen.

The ability of human Hageman factor (coagulation factor XII) to bind to a glass surface and its susceptibility to limited proteolytic cleavage during the contact activation of plasma have been studied using normal human plasma and plasmas genetically deficient in factor XI, prekallikrein, or high molecular weight kininogen (HMWK). When diluted normal plasma containing (125)I-Hageman factor was exposed to a glass surface for varying times, the Hageman factor was found to bind to the surface, and within 5 min became maximally cleaved from its native 80,000 mol wt to yield fragments of 52,000 and 28,000 mol wt. Hageman factor in factor XI-deficient plasma behaved similarly. In prekallikrein-deficient plasma, the binding of Hageman factor to the glass surface occurred at the same rate as in normal plasma but the cleavage was significantly slower, and did not reach maximum until 60 min of incubation. Cleavage of Hageman factor in HMWK-deficient plasma occurred at an even slower rate, with greater than 110 min of incubation required for maximal cleavage, although the rate of binding to the glass was again the same as in normal plasma. Normal rates of cleavage of Hageman factor were observed for the deficient plasmas after reconstitution with purified human prekallikrein or HMWK, respectively. These observations suggest that normal contact activation in plasma is associated with proteolytic activation of surfacebound Hageman factor. The cleavage of the surface-bound Hageman factor molecule responsible for the formation of the 52,000-and 28,000-mol wt fragments occurred at two closely situated sites, one of which was within a disulfide loop. Cleavage at the site external to the disulfide bond resulted in the release from the surface of the 28,000-mol wt fragment. Cleavage at the site within the disulfide loop resulted in the formation of a 28,000-mol wt fragment which remained surface bound, presumably by virtue of the disulfide linkage to the larger fragment.

Studies on factor XII in porcine plasma: purification and its conversion to activated form by porcine plasma kallikrein.

When porcine plasma was subjected to four steps of ion-exchange column chromatographies, factor XII (F. XII) was separated into two fractions, F. XII-1 and F.XII-2, and about 8.5 mg of F. XII-1 and 2.3 mg of F. XII-2 were obtained from 675 ml of the plasma. The purified proteins were found to give a single band on sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE). The molecular masses of F. XII-1 and XII-2 were estimated to be about 84 kDa by SDS-PAGE, and consisted of a single polypeptide chain. However, F. XII-1 converted to active form after removing Polybrene, but F. XII-2 was not activated. Activation of F. XII-2 took place on incubation with porcine plasma kallikrein, and the activation rate was increased only in the presence of negatively charged surfaces, such as sulfatide. When the preparation was incubated with sulfatide, spontaneous activation was not observed on the condition that we used. Activation of porcine F. XII-2 by porcine plasma kallikrein was found to involve the cleavage of the peptide bond on the disulfide-bridged polypeptide chain, and further fragmentation of activated form was observed during prolonged incubation time.

Interaction of bovine factor XIIa with an inhibitor from bovine plasma.

An inhibitor of factor XIIa has been purified from bovine plasma and characterized (Thornton, R.D. and Kirby, E.P. (1987) J. Biol. Chem. 262, 12714-12721). This inhibitor interacts with XIIa to form a very stable complex with a 1:1 stoichiometry. The active site of XIIa, located on the light chain, is directly involved in the interaction, and complex formation between factor XIIa inhibitor and XIIa can be blocked by diisopropyl fluorophosphate, corn trypsin inhibitor, or the chromogenic substrate S2302. Incubation of the complex with excess XIIa does not result in cleavage of the complex. The complex does not spontaneously dissociate and is stable to boiling, SDS, thiocyanate, acid, and hydroxylamine or Tris at pH 7-10. In addition to complex formation, a cleaved form of factor XIIa inhibitor can be observed. We suggest that the inhibitor is acting as a mechanism-based inactivator, using the criteria of time-dependent inactivation under pseudo-first-order conditions, 1:1 stoichiometry, active site involvement, kinetic protection by substrate or by an active site inhibitor, and partitioning between cleavage of factor XIIa inhibitor and inactivation by complex formation.

Purified factor XII has a higher specific activity than the parent molecule in plasma.

The specific clot promoting activity of factor XII (F XII) in plasma samples from 50 healthy adults was between 30 and 48 U/mg, whereas the specific activity of purified F XII ranged from 55 to 66 U/mg. This difference was neither due to partial proteolytic activation during purification of F XII nor to the influence of plasma protease inhibitors. Purified F XII showed normal size and charge, as demonstrated by sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE) and isoelectric focusing, respectively. The increase of the specific F XII activity during the purification process mainly occurred after anion exchange chromatography on DEAE-Sephadex and after the final gel filtration step. Upon dextran sulfate activation, proteolytic cleavage of F XII and generation of kallikrein-like amidolytic activity was faster in F XII deficient plasma containing purified F XII than in F XII deficient plasma containing a corresponding amount of pooled normal plasma (NHP). The binding to kaolin was similar for both, purified F XII and plasma F XII. In conclusion, purification alters the properties of F XII in an unknown way, resulting in an increased specific clot promoting activity.

Effect of heparin on the activation of factor XII and the contact system in plasma.

We examined in purified systems and in human plasma whether heparin serves as a contact system activating compound. Purified human factor XII zymogen was not activated by heparin through an autoactivation mechanism, but was activated in the presence of purified prekallikrein. Zn2+ (12 microM) did not support autoactivation by heparin. The activation of factor XII and the contact system by heparin in plasma anticoagulated with citrate or with hirudin (not chelating ions) was examined by the cleavage of 125I-labeled factor XII and high molecular weight kininogen (HK). Heparin at 1.6 and 16 USP U/ml was not able to produce activation, in contrast to dextran sulfate (20 micrograms/ml) which supported activation of both factor XII and HK. This study indicates that heparinized plasma does not support activation of the contact system mediated through activation of factor XII. It is not expected that heparin anticoagulant therapy will contribute to activation of the contact system.

A rapid purification with high recovery of factor XII (Hageman factor) on immunoaffinity column: application to an abnormal clotting factor XII (factor XIITORONTO).

A rapid purification procedure with high recovery of blood coagulation factor XII (Hageman factor, HF) was established. Homogeneous HF was isolated in 6 days on a monoclonal antibody-immunoaffinity column chromatography followed by gel filtration. Approximately 4,300-fold purification of HF was attained with 31% yield on average (n = 4). Using this method, an abnormal HF was purified from plasma of a patient with cross-reacting material (CRM)-positive Hageman trait (factor XIITORONTO). The abnormal HF was found to be a single chain polypeptide with the same molecular weight (80,000) as the normal HF. Both abnormal and normal HF had similar amino acid compositions.

Purification of human factor XII from plasma using zinc chelate affinity chromatography.

Human factor XII (Hageman Factor) was isolated from human plasma to apparent homogeneity using a four step procedure with yields up to 30%. The method, which is more rapid than the current conventional procedures, consists of a 25-50% ammonium sulfate fractionation, two affinity chromatography steps using Zinc Chelate Sepharose, followed by gel filtration. The isolated zymogen factor XII was a single protein component when examined by SDS PAGE with a Mr of 80,000. Activation of zymogen factor XII to its active enzymatic forms by kallikrein resulted in factor XIIa and factor XIIf as observed with factor XII purified by other procedures.

Preparation of beta-XIIa (Hageman factor fragment) from human plasma.

beta-Factor XIIa (beta-XIIa, Mr approximately 30,000) was isolated from human plasma by a procedure which utilized, as an initial step, the adsorption of Factor XII to celite. Activation of Factor XII and subsequent release of beta-XIIa was brought about by the proteolytic action of co-adsorbed kallikrein. Two successive chromatographic procedures were then used to achieve a final purification of 4,420-fold over plasma and an overall field of 2.3 mg of beta-XIIa per liter.

Contact activation factors in plasma from pregnant women--increased level of an association between factor XII and kallikrein.

The plasma levels of FXII, prekallikrein (PK), and high- and low molecular weight kininogens (HK and LK) were studied in pregnant women in the last trimester and in non-pregnant controls. FXIIa and plasma kallikrein were assayed in acetone-treated citrated plasma (CPLa) with the tetrapeptide S-2222 as substrate, using soybean trypsin inhibitor and corn inhibitor to exclude kallikrein and FXIIa respectively. No difference in PK-level could be registered for the two kinds of plasma, but the level of FXII had increased to about 150% in the pregnancy plasma. No difference in HK-level was observed, whereas the LK-level was significantly higher in pregnancy plasma, about 250% and 160% in rocket immunoassay and bioassay respectively. In fractions from gel filtration of plasma acetone-activated in the presence of benzamidine (BPLa), kallikrein was assayed as S-2302 amidase, HK and LK were measured in rocket immunoassay, and HK and FXII were studied in PAGE immunoblot experiments. In contrast to previous results obtained upon gel filtration of CPLa, not only kallikrein and HK, but in addition also FXII now appeared together in the same fractions and as two separate peaks. One peak eluting in early fractions (gel mol. wt. 300-400 KD), and one late eluting peak of proteins adsorbed to the gel material. The first peak was notably marked in pregnancy plasma. The results provide support for the assumption of an association in plasma between the three contact activation factors studied.

Characterization of plasminogen activators in unstimulated and stimulated human whole saliva.

The fibrinolytic activity of saliva from healthy males was studied on plasminogen-free and plasminogen-rich fibrin plates. A cell-bound plasminogen activator in human unstimulated and stimulated whole saliva was demonstrated. The assessed fibrinolytic activities could always be quenched by incorporation into the fibrin plates of IgG antibodies raised against human two-chain tissue-plasminogen activator (t-PA), while additional experiments indicated the absence in normal human saliva of urokinase-like and factor XII-dependent plasminogen activators as well as the absence of inhibitors of fibrinolysis. Thus, t-PA is the only type of plasminogen activator in normal human saliva. The present findings seem to support our recent clinical observations of a decrease in the incidence of bleeding complications and need for replacement therapy in hemophiliacs undergoing oral surgery during local antifibrinolytic therapy with tranexamic acid. Whether the findings might also be of importance in other pathological conditions of the oral cavity, such as impaired wound healing, remains to be elucidated.

Hageman factor dependent pathway in local vascular permeability enhancement.

The possibility of an involvement of the kallikrein-kinin system in increasing vascular permeability induced by intradermal injection of the guinea pig activated Hageman factor (beta HFa) was examined. In vitro system, the kinin generation was observed in guinea pig plasma when the plasma was incubated with beta HFa. This kinin generation was dependent upon the dose of beta HFa added and upon the presence of plasma prekallikrein, since 97 percent of the kinin release was diminished by removing the prekallikrein in plasma by treatment with anti-prekallikrein antibody. These results suggested that the Hageman factor-kallikrein-kinin cascade in guinea pig was similar to those in human and bovine plasma. In the permeability experiment in vivo, a simultaneous injection of soybean trypsin inhibitor (10(-6) M), which is the inhibitor of guinea pig plasma kallikrein, inhibited the permeability response to beta HFa by more than 90 percent. The permeability response to beta HFa was attenuated 5 fold in animals depleted of the circulating plasma prekallikrein by intraarterial antibody administration. These results indicated the participation of plasma prekallikrein in the permeability reaction to beta HFa. A simultaneous injection of a kinin destructive enzyme, carboxypeptidase B, diminished the permeability reaction to beta HFa, without any inhibition of the amidolytic activity of beta HFa or plasma kallikrein. A simultaneous injection of an inhibitor of a kinin destructive enzyme (kininase II), SQ 20,881(Glu-Tyr-Pro-Arg-Pro-Gln-Ile-Pro-Pro-OH), augmented the permeability reaction to beta HFa 10 fold, without any effect on the amidolytic activity of beta HFa or plasma kallikrein.(ABSTRACT TRUNCATED AT 250 WORDS)