Antithrombin-heparin cofactor: an inhibitor of plasma kallikrein

The Inhibition of Serine Proteases by Serpins Is Augmented by Negatively Charged Heparin: A Concise Review of Some Clinically Relevant Interactions

Serine proteases are members of a large family of hydrolytic enzymes in which a particular serine residue in the active site performs an essential role as a nucleophile, which is required for their proteolytic cleavage function. The array of functions performed by serine proteases is vast and includes, among others, the following: (i) the ability to fight infections; (ii) the activation of blood coagulation or blood clot lysis systems; (iii) the activation of digestive enzymes; and (iv) reproduction. Serine protease activity is highly regulated by multiple families of protease inhibitors, known collectively as the SERine Protease INhibitor (SERPIN). The serpins use a conformational change mechanism to inhibit proteases in an irreversible way. The unusual conformational change required for serpin function provides an elegant opportunity for allosteric regulation by the binding of cofactors, of which the most well-studied is heparin. The goal of this review is to discuss some of the clinically relevant serine protease-serpin interactions that may be enhanced by heparin or other negatively charged polysaccharides. The paired serine protease-serpin in the framework of heparin that we review includes the following: thrombin-antithrombin III, plasmin-anti-plasmin, C1 esterase/kallikrein-C1 esterase inhibitor, and furin/TMPRSS2 (serine protease Transmembrane Protease 2)-alpha-1-antitrypsin, with the latter in the context of COVID-19 and prostate cancer.

A Comparison of the Oligosaccharide Structures of Antithrombin Derived from Plasma and Recombinant Using POTELLIGENT® Technology

Human antithrombin (AT) has two isoforms of which the predominant α-form is glycosylated on all four possible glycosylation sites and the lower abundant β-isoform lacks the oligosaccharide on Asn135. The main oligosaccharide structure of human AT consists of biantennary complex-type oligosaccharides lacking a core fucose. Generally, Chinese hamster ovary (CHO) cells produce recombinant human AT (rhAT) with core-fucosylated oligosaccharides. However, rhAT lacking core-fucose oligosaccharides can be produced by POTELLIGENT^® technology, which uses FUT8 knockout CHO cells in production. The rhAT has more variable glycan structures, such as tetra-antennary complex type, high-mannose type, and mannose 6-phosphate species as minor components compared to plasma-derived human AT (phAT). In addition, the site-specific glycan profile was different between two ATs. We evaluated the effect of these properties on efficacy and safety based on a comparison of rhAT made by that technology with phAT in terms of their respective oligosaccharide structures, site-specific oligosaccharide profiles, and the ratio of α- and β-forms. Although some structural differences were found between the rhAT and phAT, we concluded that these differences have no significant effect on the efficacy and safety of rhAT.

Protein adsorption on polyurethane catheters modified with a novel antithrombin-heparin covalent complex

Highly anticoagulant covalent antithrombin-heparin complex (ATH) was covalently grafted onto polyurethane catheters to suppress adsorption/activation of procoagulant proteins and enhance adsorption/activation of anticoagulant proteins for blood compatibility. Consistency of catheter coating was demonstrated using immunohistochemical visualization of ATH. The ability of the resulting immobilized ATH heparin chains to bind antithrombin (AT) from plasma, as measured by binding of (125)I-radiolabeled AT, was greater than that for commercially-available heparin-coated catheters, and much greater than for uncoated catheters. Complementary measurements of antifactor Xa (FXa) activity and plasma protein binding were also performed. Both ATH-coated and heparin-coated catheters demonstrated functional binding of exogenous AT. However, the ATH-coated catheters gave a trend towards elevated anti- FXa activities/AT binding ratios, consistent with the higher active pentasaccharide content in starting ATH. Western blot analysis of proteins adsorbed to catheters after incubation with rabbit plasma established protein binding profiles that showed AT and albumin as major plasma proteins adsorbed to ATH-coated catheters, while AT and altered forms of fibrinogen were major plasma protein species adsorbed to heparinized catheters.

Anti-thrombin is expressed in the benign prostatic epithelium and in prostate cancer and is capable of forming complexes with prostate-specific antigen and human glandular kallikrein 2

Anti-thrombin, a member of the serpin family and an inhibitor of thrombin and blood coagulation factor Xa, was recently shown to inhibit angiogenesis and tumor growth. In the present study, we examined the expression of anti-thrombin in benign and malignant prostate gland. Using immunohistochemistry, anti-thrombin was found in prostate epithelium and stroma cells. Tissue microarrays of tumors (n = 112) and three different prostate cancer cell lines (PC-3, LNCaP, and DU-145) were all positive for anti-thrombin. Abundant expression in a population of prostatic tumor cells was further evidenced by in situ hybridization experiments. The immunostaining for anti-thrombin was confined to the cytoplasm, was most intense in Gleason grade 3 tumors, and in part overlapped with that of prostate-specific antigen. Western blotting of benign and malignant tissue homogenates revealed a predominant 58-kd anti-thrombin immunoreactive component. In vitro, anti-thrombin formed complexes more readily with human kallikrein 2, particularly in the presence of heparin, and less efficiently with prostate-specific antigen. Both complexes could be recognized by polyclonal and monoclonal IgGs against anti-thrombin. We conclude that anti-thrombin is widely expressed in prostate cancer but is gradually lost in tumors of high Gleason grade. Anti-thrombin may act as a local anti-angiogenic factor, the effect of which is partially lost in poorly differentiated prostatic tumors.

Transgenically Produced Human Antithrombin: Structural and Functional Comparison to Human Plasma–Derived Antithrombin

Recombinant human antithrombin (rhAT) produced in transgenic goat milk was purified to greater than 99%. The specific activity of the rhAT was identical to human plasma–derived AT (phAT) in an in vitro thrombin inhibition assay. However, rhAT had a fourfold higher affinity for heparin than phAT. The rhAT was analyzed and compared with phAT by reverse phase high-performance liquid chromatography, circular dichroism, fluorophore-assisted carbohydrate electrophoresis (FACE), amino acid sequence, and liquid chromatography/mass spectrography peptide mapping. Based on these analyses, rhAT was determined to be structurally identical to phAT except for differences in glycosylation. Oligomannose structures were found on the Asn 155 site of the transgenic protein, whereas only complex structures were observed on the plasma protein. RhAT contained a GalNAc for galactose substitution on some N-linked oligosaccharides, as well as a high degree of fucosylation. RhAT was less sialylated than phAT and contained both N-acetylneuraminic and N-glycolylneuraminic acid. We postulate that the increase in affinity for heparin found with rhAT resulted from the presence of oligomannose-type structures on the Asn 155 glycosylation site and differences in sialylation.

Transgenically Produced Human Antithrombin: Structural and Functional Comparison to Human Plasma–Derived Antithrombin

Recombinant human antithrombin (rhAT) produced in transgenic goat milk was purified to greater than 99%. The specific activity of the rhAT was identical to human plasma–derived AT (phAT) in an in vitro thrombin inhibition assay. However, rhAT had a fourfold higher affinity for heparin than phAT. The rhAT was analyzed and compared with phAT by reverse phase high-performance liquid chromatography, circular dichroism, fluorophore-assisted carbohydrate electrophoresis (FACE), amino acid sequence, and liquid chromatography/mass spectrography peptide mapping. Based on these analyses, rhAT was determined to be structurally identical to phAT except for differences in glycosylation. Oligomannose structures were found on the Asn 155 site of the transgenic protein, whereas only complex structures were observed on the plasma protein. RhAT contained a GalNAc for galactose substitution on some N-linked oligosaccharides, as well as a high degree of fucosylation. RhAT was less sialylated than phAT and contained both N-acetylneuraminic and N-glycolylneuraminic acid. We postulate that the increase in affinity for heparin found with rhAT resulted from the presence of oligomannose-type structures on the Asn 155 glycosylation site and differences in sialylation.

Purification and characterization of an antithrombin III inactivating enzyme from the venom of the African night adder (Causus rhombeatus)

A serine proteinase was isolated from the venom of the night adder (Causus rhombeatus) by fast protein liquid chromatography (anion-exchange, gel filtration and hydrophobic interaction). The protein (termed CR-serpinase) had an estimated mol. wt of 45,500 as determined by SDS-PAGE, pI of 4.7 and a carbohydrate content of 18.9%. Incubation of CR-serpinase with purified human antithrombin III at a molar ratio of 1:66 resulted in a loss of more than 90% of the initial AT III activity within 10 min. The reaction was dependent on heparin. In SDS-PAGE inactivation of human antithrombin III was correlated with the occurrence of two cleavage products. The cleavage site in the antithrombin III molecule was determined to be Arg 393-Ser 394 by amino-terminal sequencing. CR-Serpinase had no thrombin-like activity since no fibrinogen conversion was induced and had no procoagulant activity. CR-Serpinase activity was not inhibited by antithrombin III-heparin and was not decreased by a 10-min preincubation in normal human plasma. Inactivation of antithrombin III by CR-serpinase appeared to be very specific.

The role of the endothelium in in vivo anticoagulation

Endothelial cells play an important role in the anticoagulation of circulating blood. They participate in at least four mechanisms that aid in preventing vascular occlusion by platelet plugs or fibrin deposition. These mechanisms include the fibrinolytic pathway, the enhancement of action of endogenous inhibitors of coagulation, the inhibition of cofactors required for fibrin formation, and the synthesis and release of platelet inhibitors. Thus, the endothelium can no longer be viewed merely as a passive barrier separating the blood from the interstitial matrix. Instead, it must be viewed as an active participant in the maintenance of blood fluidity and the prevention of thrombogenesis.

Normal haemostasis and its regulation

Regulation of normal haemostasis and blood flow involves complex interactions between plasma proteins and blood cells, including platelets, leukocytes and the endothelial lining of blood vessels. Thrombin acts as a pivot in the maintenance of the haemostatic balance; the vascular endothelial cell in particular limits the generation of thrombin by localisation of anticoagulant processes on its luminal membrane. The endothelial cell synthesises key molecules in this process and also binds exogenously derived molecules, as well as releasing proteins of the fibrinolysis cascade. The thromboresistance of the luminal surface is further regulated by lipoxygenase and cyclo-oxygenase metabolites of unsaturated fatty acids synthesised by the endothelial cell. In response to trauma, inflammatory reactions, normal wound healing and in association with a variety of disease states, the anticoagulant and fibrinolytic mechanisms are downregulated and the procoagulant and thrombotic mechanisms predominate with resultant generation of thrombin, fibrin clot formation and subsequent platelet adhesion and aggregation. Pro-inflammatory and prothrombotic cytokines downregulate the fibrinolytic and activated protein C pathways as well as inducing synthesis of specific procoagulant and prothrombotic mediators by platelets and leukocytes as well as endothelium.

Effect of heparin on the inhibition of the contact system enzymes

1. One can accurately predict the contribution of each inhibitor to the total inactivation of an enzyme in plasma once its pseudo-first-order reaction rate constant and concentration are known. 2. Because the mechanism of augmentation of the inactivation rate of an enzyme by ATIII occurs via formation of an ATIII-heparin complex, the degree of potentiation can be predicted by knowing the binding capacity (sites per mole) of the heparin preparation and the concentration of heparin in the reaction (to calculate the concentration of the ATIII-heparin complex). 3. The augmentation by heparin of the inactivation rate of a particular enzyme by ATIII is dependent upon the presence of other enzymes with higher kassoc, since these would strongly compete for the ATIII-heparin complex. 4. In a plasma environment, using therapeutic levels of heparin, there is no augmentation of the inactivation rate of any of the contact enzymes.

Contribution of blood and systemic circulation to the processing of pro-(atrial natriuretic factor)

Atrial natriuretic factor-(Asn1-Tyr126)-peptide, the 13.6 kDa propeptide of atrial natriuretic factor (ANF), is stored in the secretory granules of atrial cardiocytes. ANF-(Ser99-Tyr126)-peptide, the 28-amino-acid species, is the circulating form of this hormone in the rat. As the site of maturation of the prohormone is still unknown, the present study was undertaken to understand the contribution of the circulation to the maturation process of pro-ANF. 125I-ANF-(Asn1-Tyr126)-peptide was incubated with whole rat blood, plasma or serum for different time intervals, and the products were analysed. There was minimal activation of the propeptide in either whole blood or plasma. Incubation with serum, however, resulted in the formation of an 11 kDa and a 3 kDa peptide which corresponded respectively to the N-terminal and C-terminal parts of the propeptide. These results suggest that hydrolysis of the propeptide in serum is brought about by enzymes that may be stimulated during coagulation but which may not play a major role in the activation of pro-ANF in the circulation. Plasma analysis at different time intervals after prohormone injection indicated a non-specific hydrolysis of the pro-ANF molecule. The disappearance rate curves, obtained with radiolabelled pro-ANF, suggested the presence of two components with half-lives of 2.1 +/- 0.4 min and 52.5 +/- 8.4 min respectively. A metabolic clearance rate of 1.49 +/- 0.22 ml/min and an initial distribution volume of 47.4 +/- 8 ml were calculated. These results indicate that the maturation of pro-ANF to its active circulating form takes place before it is released into the circulation.

The contact activation system: biochemistry and interactions of these surface-mediated defense reactions

This review is intended to be a critical state-of-the-art overview of the activation and inhibition of the proteins (factor XII, prekallikrein, high molecular weight kininogen, and factor XI) of the contact phase of coagulation. Specifically, this review will reconsider the concept of the reciprocal activation of the proteases of the contact phase of coagulation, factor XII, and prekallikrein, in light of much recent evidence indicating that factor XII, itself, autoactivates when associated with negatively charged surfaces. In addition, the mechanisms for amplification of activation of the proteins of the contact phase of coagulation will be discussed from the pivotal role of high molecular weight kininogen, or one of its altered forms, serving as a cofactor to order the activation of the zymogens it is associated with. The role and relative importance of each of the naturally occurring plasma protease inhibitors (C1-inhibitor, alpha-2-macroglobulin, alpha-1-antitrypsin, antithrombin III, and alpha-1-antiplasmin) will be assessed as they relate to the dampening of contact phase activation. Finally, the contact phase of coagulation activation will be discussed not only as a plasma proteolytic mechanism, but also as it interacts with platelets.

Inactivation of kallikrein in human plasma

Human plasma kallikrein is inactivated by plasma protease inhibitors. This study was designed to determine the nature of these protease inhibitors and to assess their relative importance in the inactivation of kallikrein. Therefore, the kinetics of kallikrein inactivation and the formation of kallikrein inhibitor complexes were studied in normal plasma and in plasma depleted of either alpha 2-macroglobulin (alpha 2M), C1 inhibitor, or antithrombin (AT III). Prekallikrein was activated by incubation of plasma with dextran sulfate at 4 degrees C. After maximal activation, kallikrein was inactivated at 37 degrees C. Inhibition of kallikrein amidolytic activity in AT III-deficient plasma closely paralleled the inactivation rate of kallikrein in normal plasma. The inactivation rate of kallikrein in alpha 2M-deficient plasma was slightly decreased compared with normal plasma, but in contrast to normal, C1 inhibitor-deficient, and AT III-deficient plasma, no kallikrein amidolytic activity remained after inactivation that was resistant to inhibition by soybean trypsin inhibitor. Suppression of kallikrein activity in C1 inhibitor-deficient plasma was markedly decreased, and this was even more pronounced in plasma deficient in both C1 inhibitor and alpha 2M. The pseudo first-order rate constants for kallikrein inactivation in normal, AT III-deficient, alpha 2M-deficient, C1 inhibitor-deficient plasma, and plasma deficient in both alpha 2M and C1 inhibitor, were 0.68, 0.60, 0.43, 0.07, and 0.016 min-1, respectively. Sodium dodecyl sulfate gradient polyacrylamide slab gel electrophoresis showed that during inactivation of kallikrein in plasma, high-Mr complexes were formed with Mr at 400,000-1,000,000, 185,000, and 125,000-135,000, which were identified as complexes of 125I-kallikrein with alpha 2M, C1 inhibitor, and AT III, respectively. In addition, the presence of an unidentified kallikrein-inhibitor complex was observed in AT III-deficient plasma. 52% of the 125I-kallikrein was associated with C1-inhibitor, 35% with alpha 2M, and 13% with AT III and another protease inhibitor. A similar distribution of 125I-kallikrein was observed when the 125I-kallikrein inhibitor complexes were removed from plasma by immunoadsorption with insolubilized anti-C1 inhibitor, anti-alpha 2M, or anti-AT III antibodies. These results suggest that only covalent complexes are formed between kallikrein and its inhibitors in plasma. As a function of time, 125I-kallikrein formed complexes with C1 inhibitor at a higher rate than with alpha 2M. No difference was observed between the inactivation rate of kallikrein in high-Mr kininogen-deficient plasma and that in high-Mr kininogen-deficient plasma reconstituted with high-Mr kininogen; this suggests that high-Mr kininogen does not protect kallikrein from inactivation in the plasma milieu. These results have quantitatively demonstrated the major roles of C1 inhibitor and alpha 2M in the inactivation of kallikrein in plasma.

Contribution of plasma protease inhibitors to the inactivation of kallikrein in plasma

Although Cl-inhibitor (Cl-INH) and alpha(2)-macroglobulin (alpha(2)M) have been reported as the major inhibitors of plasma kallikrein in normal plasma, there is little quantitative support for this conclusion. Thus, we studied the inactivation of purified kallikrein in normal plasma, as well as in plasma congenitally deficient in Cl-INH, or artificially depleted of alpha(2)M by chemical modification of the inhibitor with methylamine. Under pseudo-first-order conditions, the inactivation rate constant of kallikrein in normal plasma was 0.60 min(-1). This rate constant was reduced to 0.35, 0.30, and 0.06 min(-1), in plasma deficient respectively in Cl-INH, alpha(2)M, or both inhibitors. Thus Cl-INH (42%) and alpha(2)M (50%) were found to be the major inhibitors of kallikrein in normal plasma. Moreover all the other protease inhibitors present in normal plasma contributed only for 8% to the inactivation of the enzyme. To confirm these kinetic results, (125)I-kallikrein (M(r) 85,000) was completely inactivated by various plasma samples, and the resulting mixtures were analyzed by gel filtration on Sepharose 6B CL for the appearance of (125)I-kallikrein-inhibitor complexes. After inactivation by normal plasma, 52% of the active enzyme were found to form a complex (M(r) 370,000) with Cl-INH, while 48% formed a complex (M(r) 850,000) with alpha(2)M. After inactivation by Cl-INH-deficient plasma, >90% of the active (125)I-kallikrein was associated with alpha(2)M. A similar proportion of the label was associated with Cl-INH in plasma deficient in alpha(2)M. After inactivation by plasma deficient in both Cl-INH and alpha(2)M, (125)I-kallikrein was found to form a complex of M(r) 185,000. This latter complex, which may involve antithrombin III, alpha(1)-protease inhibitor, and/or alpha(1)-plasmin inhibitor, was not detectable in appreciable concentrations in the presence of either Cl-INH or alpha(2)M, even after the addition of heparin (2 U/ml). These observations demonstrate that Cl-INH and alpha(2)M are the only significant inhibitors of kallikrein in normal plasma confirming previous predictions based on experiments in purified systems. Moreover, in the absence of either Cl-INH or alpha(2)M, the inactivation of kallikrein becomes almost entirely dependent on the other major inhibitor.

Protection of human plasma kallikrein from inactivation by C1 inhibitor and other protease inhibitors. The role of high molecular weight kininogen

High Mr kininogen increases the activation rate of prekallikrein by activated factor XII on a surface. The resulting serine protease, plasma kallikrein, Mr 88 000, is inhibited in plasma by C1 inhibitor, Mr 105 000. Since prekallikrein circulates in plasma with high Mr kininogen as a complex and a kallikrein-high Mr kininogen complex can be formed in purified systems, we studied whether the inhibition of kallikrein by C1 inhibitor was influenced by high Mr kininogen. With C1 inhibitor in excess, the inactivation of kallikrein followed pseudo-first-order kinetics. The second-order rate constant for the reaction was 1.7 X 10(4) M-1 s-1, and a kallikrein-C1 inhibitor complex, Mr 190 000 was identified on polyacrylamide gel electrophoresis in the presence of sodium dodecyl sulfate. Kallikrein and C1 inhibitor formed an irreversible complex without measurable prior equilibrium. The rate of this reaction was decreased by 50% in the presence of high Mr kininogen (1 unit/mL or 0.73 muM). Kinetic analysis indicated that this protection was the result of the formation of a reversible complex between kallikrein and high Mr kininogen, which had a dissociation constant of 0.75 muM. However, low Mr kininogen did not protect kallikrein from inactivation by C1 inhibitor. High Mr kininogen also protected kallikrein from inactivation by diisopropyl fluorophosphate. These findings suggest that the kallikrein-high Mr kininogen complex was formed by noncovalent interactions between the light chains of both kallikrein and high Mr kininogen.