John W. Fenton

Thrombin-Elicited Contractile Responses of Aortic Smooth Muscle:

Abstract Human α-thrombin at physiologically relevant concentrations of 0.75 to 225 nM (0.01 to 22 clotting units/ml) caused rabbit thoracic aorta to slowly contract in isolated organ baths. Near maximum contractile tension (10% below the irreversible contraction caused by 22 units/ml) was slowly generated by 75 nM α-thrombin (8 units/ml) over 18 min and was 50% that of the norepinephrine (NE) control. The initial tonus could only be regained by repeated washing and 60 min equilibration. Tissues were refractory to a second 75 nM α-thrombin challenge but responded fully to 1.0 nM NE. Conversion of human α-thrombin to nonclotting but estero/amidolytically active 7-thrombin (<0.1% clotting activity) or nitration of the enzyme (1% clotting activity) did not interfere with the contractile activity, whereas chemical conjugates of the parent enzyme at the catalytic site were inactive. Neither atropine, phentolamine, nor indomethacin blocked the thrombin-induced contractions, whereas D-600 was markedly inhibitory. Preincubation of the tissue with inactive forms of thrombin did not prevent the α-thrombin-induced response. Removal of the vascular endothelium did not prevent contraction. Aorta preparations with intact endothelium relaxed in the presence of very low concentrations (0.75 to 750 pM) of α-thrombin prior to contracting in response to higher concentrations during cumulative dose-response experiments. Our data suggest that catalytically active thrombin forms (α- or nonclotting β- and 7-thrombins) may have hemostatic functions at the vascular levels in hemostasis.

Polyethylene glycol 6,000 enhancement of the clotting of fibrinogen solutions in visual and mechanical assays

Abstract Low concentrations of 5–10 g/liter of polyethylene glycol (PEG) 6,000 markedly enhance the clotting of fibrinogen solutions with thrombin. At constant polymer and fibrinogen concentrations, clotting times are linearly related to reciprocal NIH units of thrombin, thus permitting clotting units to be measured. This synthetic polymer is essentially an inert polyether with defined properties. It is readily available, very soluble, and indefinitely stable in solution at room temperature. Compared to acacia on a weight basis, PEG 6,000 is fivefold more effective in promoting clotting.

The action of thrombin on peptide p-nitroanilide substrates: Hydrolysis of Tos-Gly-Pro-Arg-pNA and D-Phe-Pip-Arg-pNA by human α and γ and bovine α and β-thrombins

Human and bovine α-thrombins (>90% a form) with high fibrinogen clotting activities (∼3,000 U.S. units/mg protein) exhibit similar Michaelis Menten kinetics with the p-nitroanilide tripeptide substrates Tos-Gly-Pro-Arg-pNA (Chromozym-TH) and D-Phe-Pip-Arg-pNA (S-2238). The kinetic parameters at I = 0.11 M, 25°C, pH 7.8 are: (Km = 4.18 ± 0.22 and 3.61 ± 0.15 μM; kcat = 127 ± 8 and 100 ± 1 s−1) for Chromozym TH and (Km = 1.33 ± 0.07 and 1.58 ± 0.10 μM; kcat = 91.4 ± 1.8 and 98.0 ± 0.5 s−1) for S-2238 for the human and bovine enzymes, respectively. Unlike the native enzyme forms, their “non-clotting” terminal degradative forms, human γ-thrombin (∼5 units/mg) and bovine β-thrombin (∼200 units/mg), give increased values for these parameters (Km = 14.3 ± 2.4 and 14.4 ± 2.2 μM; kcat = 160 ± 9 and 124 ± 6 s−1) for Chromozym-TH; and (Km = 2.50 ± 0.36 and 2.99 ± 0.33 μM; kcat = 106 ± 3 and 106 ± 3 s−1) for S-2238. Based on these parameters, 50% degradation of human or bovine α-thrombins can be calculated to produce relatively small errors in the kinetic measurement of total thrombin concentrations (maximally 9% and 7% for Chromozym-TH; 7% and 3% for S-2238, respectively) if the kinetic parameters for all a forms are erroneously used and assays are at 150 μM substrate. This is in contrast to the large errors inherent in clotting activity measurements on thrombin mixtures. Incorporation of 1 mg/ml of polyethylene glycol 6,000 into assay solutions eliminates systematic errors otherwise caused by thrombin adsorption to surfaces and enables thrombin to be accurately assayed at concentrations <0.1 μM or 0.01 clotting unit/ml of α-thrombin.

Warfarin-stereochemical aspects of its metabolism by rat liver microsomes.

Abstract The R - and S -enantiomers of warfarin were differentially metabolized by hepatic microsomes prepared from male Wistar and Sprague-Dawley rats. These studies were not only carried out with two strains of rats but were conducted independently in two laboratories employing different techniques. Although minor differences were observed, the same stereoselectivity was found for the microsomal transformations produced by both strains. The formation of 7- and 8-hydroxywarfarin was stereoselective for the R -enantiomer and in addition this enantiomer was metabolized more rapidly than the S -enantiomer. The converse stereoselectivity was found for the process of 4′-hydroxylation in Sprague-Dawley rats but it could not be conclusively shown for Wistar rats. A Michaelis-Menten analysis of the metabolic products of R - and S -warfarin formed by liver microsomes from male Sprague-Dawley rats is reported. The K m for the processes of 6-, 7-, and 4′-hydroxylation for both isomers and the K m for 8-hydroxylation of the R -isomer were all of the order of 0.03 to 0.11 mM and were not statistically different. The K m for 8-hydroxylation of the S -isomer, 0.20 mM, was significantly greater. The K m for benzylic hydroxylation of both isomers appeared to be still greater but was less precisely determined. The V max for each of the enantiomeric pairs of products was statistically different. The kinetic data are interpreted as being inconsistent with the supposition that an arene oxide (6–7 and/or 7–8) may serve as the intermediate in the formation of 7-hydroxywarfarin from either isomer. Further, if product formation is assumed to be rate limiting, the data provide evidence for at least three distinct enzymatic processes which may or may not be distinct hemoproteins. Reduction of the side-chain ketonic function of warfarin to the corresponding diastereomeric warfarin alcohols by the 105,000 g supernatant fraction displayed both a high degree of stereoselectivity ( R -isomer) and stereospecificity ( S -reduction). This reduction was best catalyzed by NADPH rather than by NADH. Determination and quantification of the metabolic products obtained after incubation of R - and S -warfarin with the 10,000 g supernatant were consistent with the summation of those independently produced by the microsomal pellet and the 105,000 g supernatant.

Thrombin Inhibition by Hirudin: How Hirudin Inhibits Thrombin

In addition to its classical active-site regions (catalytic site and adjacent regions), alpha-thrombin has a unique anion-binding exosite, which is functionally independent of the catalytic site and is involved in fibrin(ogen) recognition. This exosite also accounts for adhesion to negatively charged surfaces (e.g., glass), binding to cell surfaces, and interactions with the anionic tail of hirudin. Hirudin (as an apolar, tridisulfide-linked core structure followed by its anionic tail) interacts with alpha-thrombin by apolar (e.g., catalytic-site and adjacent regions of thrombin), as well as by ionic binding (e.g., anion-binding exosite). Circular dichroism measurements reveal a sigmoidal nonadditivity for the hirudin tail fragments, which block fibrinogen-clotting activity without interfering with tripeptide chromogenic substrate activities. Such fragments, however, inhibit factor V activation to much lesser extents than hirudin, where factor V activation is the key step in regulating thrombin generation by hirudin or heparin/antithrombin III. Hirudin-derived antithrombotics may thus have differential modes of action in hemostasis and wound healing processes.

Opposing Actions of Thrombin and Protease Nexin‐1 on Amyloid β‐Peptide Toxicity and on Accumulation of Peroxides and Calcium in Hippocampal Neurons

Abstract: Amyloid β‐peptide (Aβ) is the principal component of neuritic plaques in the brain in Alzheimers disease (AD). Recent studies revealed that Aβ can be neurotoxic by a mechanism involving free radical production and loss of cellular ion homeostasis, thus implicating Aβ as a key factor in the pathogenesis of AD. However, other proteins are present in plaques in AD, including the protease thrombin and protease nexin‐1 (PN1), a thrombin inhibitor. We therefore tested the hypothesis that thrombin and PN1 modify neuronal vulnerability to Aβ toxicity. In dissociated rat hippocampal cell cultures the toxicity of Aβ was significantly enhanced by coincubation with thrombin, whereas PN1 protected neurons against Aβ toxicity. Aβ induced an increase in levels of intracellular peroxides and calcium. Thrombin enhanced, and PN1 attenuated, the accumulation of peroxides and calcium induced by Aβ. Taken together, these data demonstrate that thrombin and PN1 have opposing effects on neuronal vulnerability to Aβ and suggest that thrombin and PN1 play roles in the pathogenesis of neuronal injury in AD.

Analysis of the secondary structure of hirudin and the mechanism of its interaction with thrombin

Highly purified hirudin with a specific activity of 13,950 antithrombin units/mg was isolated from a commercial preparation by reversed-phase chromatography. The circular dichroism (CD) spectrum of hirudin was investigated and it was found that the spectrum cannot be accounted for solely in terms of the traditional three components of peptide backbone. It was also found that the CD spectrum of the thrombin-hirudin complex was not additive with respect to the individual spectra of thrombin and hirudin. This deviation from additivity was significant between 210 and 225 nm, indicating alterations in the secondary structures of the proteins during complex formation. When thrombin was titrated with hirudin, the spectral deviation from additivity was sigmoidal, suggesting the cooperative nature of the binding process. Gel filtration of the thrombin-hirudin mixture showed no molecular species greater than a 1:1 complex (Mr 45,500), but gel filtration of free hirudin showed a multimeric form (Mr 51,300) under the same experimental conditions. It is concluded that the cooperative nature of the binding process is due to the binding of thrombin molecules to the multimeric form of hirudin. This initial binding occurs with little or no change in the CD spectrum. In the second step, the multiple complex dissociates to form 1:1 complexes, resulting in larger conformational changes and a considerable increase in binding affinity.

The site in human antithrombin for functional proteolytic cleavage by human thrombin

Ingemar BJC)RK, Ake DANIELSSON, John W. FENTON ii+ and Hans J8RNVALL* Department of Medical and Physiological Chemistry, College of Veterinary Medicine, Swedish University of Agricultural Sciences, The Biomedical Centre, Box 57.5, S-751 23 Uppsala, *Department of Chemistry, Karolinska Institutet, S-l 04 01 Stockholm, Sweden and +Division of Laboratories and Research, New York State ~epar~~~ent of Health, Albany, NY 12201, USA

Human thrombin: Partial primary structure

Abstract Human thrombin, prepared from an extrinsically activated prothrombin, was finally chromatographed on an affinity column of p -chlorobenzylamido-ϵ-aminocaproyl agarose. After reduction and s -pyridylethylation, the A and B chains were separated. The A chain contains 36 residues and no carbohydrate and has a formula weight of 4093. Its sequence corresponds to 14–49 of the bovine A chain sequence with nine replacements. Human B chain contains 264 residues. The carbohydrate content is 3.5% neutral sugar, 2.6% glucosamine, and 1.5% sialic acid. High-speed sedimentation equilibrium in guanidine gave a minimum molecular weight of 33,500. Sequenator analysis yielded the first 50 amino-terminal residues and established the positions of three of the cyanogen bromide fragments. A fourth fragment lacked homoserine and was placed at the carboxyl-terminus. The four remaining fragments, comprising half of the B chain, were tentatively assigned positions by assuming homology with bovine thrombin and with pancreatic serine proteases. Within the B chain, the active site histidine and serine were both in highly conservative regions, and an Arg-Tyr bond has been identified as a biological degradation site. Sixty-nine percent of the primary structure of human thrombin was identified by sequenator analysis. In comparing this sequence with that reported for the bovine molecule, 26 replacements are present.

Depletion of protein kinase Cε in normal and scleroderma lung fibroblasts has opposite effects on tenascin expression

OBJECTIVE To determine whether the extracellular matrix protein tenascin-C (TN-C) is overexpressed in lung fibroblasts from systemic sclerosis (SSc) patients, the molecular mechanisms regulating TN-C secretion in SSc and normal lung fibroblasts, and how these processes might contribute to lung fibrosis in SSc patients. METHODS TN-C secretion by SSc and normal fibroblasts was compared in vivo (in bronchoalveolar lavage [BAL] fluid) and in vitro (in culture medium). The ability of thrombin to induce TN-C was confirmed at both the protein and the messenger RNA (mRNA) level. The role of protein kinase Cepsilon (PKCepsilon) in the expression of TN-C was evaluated by determining the effects of thrombin on PKCepsilon levels and by directly manipulating PKCepsilon levels via the use of antisense oligonucleotides. RESULTS BAL fluid from SSc patients contained high levels of TN-C, whereas that from normal subjects contained little or no TN-C. In vitro, SSc lung fibroblasts expressed much higher amounts of TN-C than did normal lung fibroblasts. Consistent with the idea that thrombin is a physiologic inducer of TN-C, thrombin stimulated TN-C mRNA and protein expression in both SSc and normal lung fibroblasts by a mechanism that required proteolytic cleavage of the thrombin receptor. Surprisingly, thrombin treatment and antisense oligonucleotide-mediated depletion of PKCepsilon indicated that TN-C expression is regulated via opposite signaling mechanisms in SSc and normal cells. In SSc lung fibroblasts, thrombin decreased PKCepsilon levels, and the decreased PKCepsilon induced TN-C secretion; in normal fibroblasts, thrombin increased PKCepsilon levels, and the increased PKCepsilon induced TN-C secretion. Normal and SSc lung fibroblasts also differed in the subcellular localization of PKCepsilon, both before and after thrombin treatment. CONCLUSION These studies are the first to demonstrate that thrombin is a potent simulator of TN-C in lung fibroblasts and that PKCepsilon is a critical regulator of TN-C protein levels in these cells. Furthermore, our results indicate that both the regulation of PKCepsilon levels by thrombin and the regulation of TN-C levels by PKCepsilon are defective in SSc lung fibroblasts.