Portal de Periódicos da CAPES

Molecular recognition mechanisms of thrombin

2005; Elsevier BV; Volume: 3; Issue: 8 Linguagem: Inglês

10.1111/j.1538-7836.2005.01363.x

1538-7933

James A. Huntington,

Hemophilia Treatment and Research

Journal of Thrombosis and HaemostasisVolume 3, Issue 8 p. 1861-1872 Free Access Molecular recognition mechanisms of thrombin J. A. HUNTINGTON, J. A. HUNTINGTON Department of Haematology, Cambridge Institute for Medical Research, Division of Structural Medicine, Thrombosis Research Unit, University of Cambridge, Cambridge, UKSearch for more papers by this author J. A. HUNTINGTON, J. A. HUNTINGTON Department of Haematology, Cambridge Institute for Medical Research, Division of Structural Medicine, Thrombosis Research Unit, University of Cambridge, Cambridge, UKSearch for more papers by this author First published: 08 August 2005 https://doi.org/10.1111/j.1538-7836.2005.01363.xCitations: 191 James A. Huntington, Department of Haematology, Cambridge Institute for Medical Research, Division of Structural Medicine, Thrombosis Research Unit, University of Cambridge, Wellcome Trust/MRC Building, Hills Road, Cambridge CB2 2XY, UK. Tel.: +44 (0) 1223 763230; fax: +44 (0) 1223 336827; e-mail: jah52@cam.ac.uk AboutSectionsPDF ToolsRequest permissionExport citationAdd to favoritesTrack citation ShareShare Give accessShare full text accessShare full-text accessPlease review our Terms and Conditions of Use and check box below to share full-text version of article.I have read and accept the Wiley Online Library Terms and Conditions of UseShareable LinkUse the link below to share a full-text version of this article with your friends and colleagues. Learn more.Copy URL Share a linkShare onFacebookTwitterLinked InRedditWechat Abstract Summary. Thrombin is the final protease generated in the blood coagulation cascade, and is the only factor capable of cleaving fibrinogen to create a fibrin clot. Unlike every other coagulation protease, thrombin is composed solely of its serine protease domain, so that once formed it can diffuse freely to encounter a large number of potential substrates. Thus thrombin serves many functions in hemostasis through the specific cleavage of at least a dozen substrates. The solution of the crystal structure of thrombin some 15 years ago revealed a deep active site cleft and two adjacent basic exosites, and it was clear that thrombin must utilize these unique features in recognizing its substrates. Just how this occurs is still being investigated, but recent data from thrombin mutant libraries and crystal structures combine to paint the clearest picture to date of the molecular determinants of substrate recognition by thrombin. In almost all cases, both thrombin exosites are involved, either through direct interaction with the substrate protein or through indirect interaction with a third cofactor molecule. The purpose of this article is to summarize recent biochemical and structural data in order to provide insight into the thrombin molecular recognition events at the heart of hemostasis. Multiple functions of thrombin in hemostasis Thrombin has been the focus of intense study since its discovery in the 19th century [1] and continues to command attention and research funds in the biomedical field. This is in part due to the position of thrombin at the end of the blood clotting cascade, and its unique ability to cleave fibrinogen to the polymerogenic fibrin. In the absence of efficient and timely thrombin generation, stable blood clots cannot form, resulting in hemorrhage. Conversely, unregulated thrombin activity results in dissemination of the clot beyond the site of tissue damage, causing thrombosis. Although critical, the role of thrombin in hemostasis is not limited to the cleavage of fibrinogen to fibrin. Over the years, many additional activities have been identified: it is now clear that thrombin is a major player in the early steps of blood coagulation; that it is a significant contributor to the so-called 'thrombin burst' through positive feedback mechanisms; that it functions to stabilize clots; and that it participates in attenuating its own procoagulant activity (for a recent review see Ref. [2]). The multiple functions of thrombin are summarized in Fig. 1. A fundamental question about thrombin is how a single protease domain possesses the determinants to specifically recognize such a large number of substrates; the answer requires an understanding of the unique structural features of thrombin. Figure 1Open in figure viewerPowerPoint The network of thrombin activities. Thrombin is generated from its zymogen prothrombin by cleavage at two sites by the prothrombinase complex, FVa and FXa and a negatively charged phospholipid. Initial thrombin formation is rapidly followed by the cleavage of fibrinogen and activation of platelets by cleavage of PAR1. Thrombin feeds back to stimulate its own formation by activating cofactors V and VIII, and by activating FXI, when bound to Gp1bα on the platelet surface. Stabilization of the fibrin clot is also effected by thrombin through activation of the transglutaminase FXIII, and by activating the metalloprotease TAFI when bound to TM. In addition to these procoagulant roles, thrombin is capable of undergoing a switch in specificity upon binding to TM to favor activation of the anticoagulant PC. Activated PC efficiently shuts off thrombin generation through cleavage inactivation of cofactors Va and VIIIa. Thrombin is finally inhibited and cleared from the circulation by circulating serpins AT and HCII, in a GAG-dependent fashion. Structural features of thrombin Thrombin is a trypsin-like member of the chymotrypsin family of serine proteases; meaning that its structure and catalytic mechanism are like the prototypic protease chymotrypsin, and that its preferred substrates, like trypsin, have a positively charged amino acid at the P1 position of the scissile bond (nomenclature of Schechter and Berger [3], where substrate residues are numbered from the scissile P1–P1′ bond toward the N- and C-termini, respectively). Balancing the multiple functions of thrombin in hemostasis, quite simply stated, is a matter of substrate specificity. As thrombin has lost the Gla and kringle domains, all determinants of substrate recognition are necessarily found on the catalytic domain. Thus, the solution of the first crystallographic structure of thrombin in 1989 [4] provided the framework for identifying the molecular interaction, which underlie the multiple functions of thrombin [5-7]. Thrombin is traditionally viewed in the 'standard' orientation with the active site facing and the peptide substrate running from left to right, from its N→C-termini (Fig. 2). A stereo representation is necessary to appreciate the depth of the active site cleft, which for thrombin is often referred to as a 'canyon'. The walls of the canyon are formed by the insertion loops above and below the active site. These loops are longer than those found in the parent molecule chymotrypsin, and are known as the 60 and γ-insertion loops. The 60-loop is composed of Leu60, Tyr60a, Pro60b, Pro60c, Trp60d, Asp60e, Lys60f, Asn60g, Phe60h, and Thr60i, with Asn60g glycosylated. It is evident from the sequence that the 60-loop is hydrophobic in nature, with the two consecutive prolines serving to rigidify. The 60-loop, thus provides a rigid, hydrophobic cap over the active site and mediates contacts with the hydrophobic substrate residues N-terminal to the scissile bond. In contrast, the γ-loop is more hydrophilic and flexible in nature; composed of Thr147, Trp147a, Thr147b, Ala147c, Asn147d, and Val147f. It is often incompletely modeled in crystal structures due to its inherent mobility, but, as it is adjacent to the active site cleft of thrombin, it can contact substrate residues C-terminal to the scissile bond, and can make contacts with the body of substrate proteins. In forming the canyon, the loops also serve to restrict access to the catalytic site of thrombin to proteins with long, flexible substrate loops, or to proteins with complementary surfaces which mediate contact with the loops. Figure 2Open in figure viewerPowerPoint Thrombin geography. Stereo pairs of surface representations of thrombin are shown in the standard orientation, colored according to electrostatic potential, A, (red for negative and blue for positive), or hydrophobicity (green), B. The active site is occupied in this figure by the reactive center loop of AT from P4 to P2′ (yellow rods). Substrate recognition within the active site depends on favorable interactions between the P1 residue and the deep acidic S1 pocket, and between hydrophobic residues N-terminal to P1 (often P2 proline and an aromatic residue at P4) in the hydrophobic groove known as the aryl-binding pocket. The active site cleft of thrombin is unusually deep due to the presence of the 60- and γ-insertion loops above and below the active site. Two basic exosites on the surface of thrombin have been identified as critical for substrate and cofactor recognition: the so-called anion-binding exosites I and II (ABEI and ABEII). Although the figure depicts a wide open thrombin active site, thrombin can exist in a less active, closed conformation in the absence cofactor, substrate or Na+, which co-ordinates near the site indicated. The original structure of thrombin contained the covalent inhibitor D-Phe-Pro-Arg-chloromethylketone (PPACK). This inhibitor mimicked natural substrate interactions and helped explain the preference for an arginine at the P1 position, a proline at the P2 position, and a hydrophobic (preferably aromatic) residue at P4 (the D-Phe side chain occupies the position normally occupied by the P4 of the natural l-stereoisomer). This and subsequent structures of thrombin with peptides elucidated the substrate interactions N-terminal of the scissile bond, but no information was obtained about the C-terminal interactions until much later. One thing that is clear from an analysis of thrombin substrate sequences is that only a part of the information which determines specificity is to be found there (Table 1). Exosite interactions, outside the active site, either direct or cofactor mediated, are required to accelerate the formation of, or to stabilize the initial thrombin-substrate complexes (Michaelis complexes) sufficiently so that cleavage of the peptide bond can proceed. Table 1. Natural substrate sequences of thrombin P4 P3 P2 P1 P1′ P2′ P3′ Cofactor Fibrinogen (A) Gly Gly Val Arg Gly Pro Arg None Fibrinogen (B) Phe Ser Ala Arg Gly His Arg None FV (709) Leu Gly Ile Arg Ser Phe Arg None FV (1018) Leu Ser Pro Arg Thr Phe His None FV (1545) Trp Tyr Leu Arg Ser Asn Asn None FVIII (372) Ile Gln Ile Arg Ser Val Ala None FVIII (740) Ile Glu Pro Arg Ser Phe Ser None FVIII (1689) Gln Ser Pro Arg Ser Phe Gln None FXIII Gly Val Pro Arg Gly Val Asn Fibrin PAR1 Leu Asp Pro Arg Ser Phe Leu GpIbα PAR4 Pro Ala Pro Arg Gly Tyr Pro GpIbα FXI Ile Lys Pro Arg Ile Val Gly GpIbα PC Val Asp Pro Arg Leu Ile Asp TM TAFI Val Ser Pro Arg Ala Ser Ala TM AT Ile Ala Gly Arg Ser Leu Asn Heparin HCII Phe Met Pro Leu Ser Thr Gln Heparin Two exosites of significance are the basic anion-binding exosites I and II (ABEI and II, or alternatively referred to as exosites I and II). Exosite I is adjacent to the P′ side of the active site cleft, and is also known as the fibrinogen recognition exosite. Exosite II is the more basic of the two exosites, and was originally identified as the putative heparin-binding site of thrombin. With an increasing number of studies characterizing thrombin mutants, it is becoming clear that all of the natural substrates and cofactors of thrombin utilize one or both anion-binding exosites. Thrombin specificity can thus be understood as a competition for three sites: the active site, and exosites I and II. The purpose of this article is to concisely summarize the relevant biochemical and structural data that, over the last 15 years, have combined to elucidate the complex network of thrombin molecular recognition events which lie at the heart of hemostasis. Method In dealing with thrombin mutagenesis data, point mutations known to affect thrombin's catalytic activity (i.e. two-fold or greater reduction in rate of S2238 substrate cleavage) and those seen to participate in critical structural contacts (thus not available for direct intermolecular interactions) were excluded from the list of residues potentially mediating direct substrate or cofactor interaction. It was also necessary to exclude some of the structurally defined thrombin interactions on the basis of biochemical data to the contrary. In this article, thrombin is numbered according to the chymotrypsin template scheme [4], and combined mutations are indicated by a forward slash between residues. Figures were made using MolScript [8], BobScript [9], Spock and Raster3D [10]. Fibrinogen Thrombin cleavage of fibrinogen at the N-termini of the A and B chains leads to the formation of a fibrin monomer capable of linear and lateral self-association, resulting in a fibrin clot [11]. Fibrinogen is a substrate whose recognition by thrombin is directly mediated by active site and exosite I interactions to form a Michaelis complex defined by a Km in the micromolar range [12]. Two mutagenesis studies [13, 14] identified 15 thrombin residues which potentially mediate direct contact with fibrinogen: K36, R67, H71, R73, Y76, R77a, K81, and K109/110 in exosite I; D186a/K186d near the Na+-binding region; and, K60f and R173/R175/D178 which border the active site cleft. Several structural studies have cocrystallized thrombin with fibrinopeptide A, but with mixed results [15-19]. The most complete peptide structure is of bovine thrombin with an uncleavable analog of residues 7–19 of human fibrinopeptide A [16]. The three monomers in the asymmetric unit revealed conserved contacts from P10 to P2′, with P3′ in two different orientations. The P10–P1 interactions were consistent with those seen in other structures [15, 17-19], and involve the burying the hydrophobic face of a one-turn helix into the aryl-binding pocket of thrombin (of primary importance is P9 Phe), and normal interactions of the P1 Arg in the acidic S1 pocket. In 2004, the structure of human thrombin bound to the central E domain of human fibrin was solved, revealing the predicted exosite I interaction interface [20]. Thrombin residues seen in direct contact with fibrin were F34, S36a, L65, Y76, R77a, I82, and K110. Although the interface interposes the basic exosite I of thrombin with an acidic surface of fibrin, the majority of the contacts, and presumably the major energetic contribution, derive from the apposition of hydrophobic surfaces on the two molecules (F34, L65, Y76 and I82 on thrombin and F35 and A68 on the Aα and Bβ chains of fibrin, respectively). Because the structural data agree with the mutagenesis data, a figure of the active site and exosite interactions between thrombin and fibrinogen was constructed using the structural data alone (Fig. 3C, active site contacts from 1UCY in white, and fragments of Aα, Bβ and γ chains colored magenta, cyan and yellow, respectively). It is possible that other contacts exist in the 17 residue stretch which links the C-terminus of the substrate peptide to the Aα chain of the E domain, but these are not expected to be important. The exosite interactions are predicted to be preserved in thrombin cleavage of the Bβ chain of fibrinogen [20], but no structural information is available on the active site interactions involved. Figure 3Open in figure viewerPowerPoint Crystallographic and mutagenic identification of thrombin exosite and active site interactions. Thrombin is shown in three orientations: in the standard orientation with the active site facing, center; rotated −90° with exosite I facing, left; and, rotated +90° with exosite II facing, right. The first two panels are colored according to electrostatic and hydrophobic potential, as before, to illustrate the nature of the active site and two exosites. All other panels are colored to indicate interaction surfaces on thrombin: blue indicates residues whose mutations specifically affect the rate of substrate cleavage, presumably through direct interaction; and, red indicates thrombin surfaces less than 4 Å distant from substrate or cofactor residues which have been observed crystallographically. This figure emphasizes thrombin's exosite interactions, and although it is understood that the active site is necessarily involved in all substrate recognition events, only active site interactions which have been crystallographically defined are shown. Details of individual panels are given in the appropriate section in the text. Factors V and VIII Factors V and VIII are required cofactors for the activities of the prothrombinase and Xase complexes, respectively, and their activation by thrombin thereby upregulates thrombin generation through positive feedback [2]. Both factors are cleaved in multiple places by thrombin to release the central B domains from the active components, factors Va (FVa) and VIIIa (FVIIIa) (Table 1, and for current reviews see Refs. [21, 22]). Consistent with earlier work [23], two recent studies using the same library of thrombin variants have identified several residues which potentially mediate direct thrombin contact with FV [24] and FVIII [25]: K36, H71, R73, R75, Y76, R77a, K109/110, R93/R97/E97a, R101, R233/K236/Q239, and K60f for FV; and, K36, H71, R73, R75, Y76, R101 and D186a/K186d for FVIII. Of the identified residues, K36, H71, R73, R75, Y76, R77a and K109/110 are in exosite I, and R93/R97/E97a, R101, and R233/K236/Q239 are in exosite II. All studies support the absolute involvement of exosite I in recognition FV and FVIII, but the role of exosite II in recognition of individual cleavage sites is less clear. Although no structural data exist for either active site or exosite interactions, it can be concluded from the biochemical studies that in addition to the active site, both exosites I and II are involved in thrombin recognition of FV and FVIII (Fig. 3D,E). Factor XIII Factor XIII and all subsequent thrombin substrates discussed in this article require a cofactor for effective recognition and cleavage by thrombin (Table 1). FXIII is a heterotetrameric zymogen composed of two catalytic A subunits and two carrier B subunits (for review see Ref. [26]). Thrombin cleavage at a single position in the A chains (R37) results in the release of the B-chains and exposure of the catalytic site. The activated transglutaminase, FXIIIa, covalently cross-links glutamine residues with lysines on adjacent fibrin monomers to stabilize the nascent fibrin clot. Thrombin thus plays a critical role in both activating FXIII and in creating its substrate fibrin. Not surprisingly, the cofactor which accelerates the activation of FXIII by thrombin is fibrin (80-fold acceleration) [27]. Exosite I of thrombin is thus indirectly involved in FXIII recognition through its interaction with fibrin (described above). A mutagenesis study has identified residues adjacent to the active site (R173/R175/D178) as affecting FXIII cleavage [28], but no direct exosite interactions appear to be involved in the recognition of the FXIII by thrombin. For FXIII, as with all thrombin substrates, an interaction with the active site of thrombin is a critical part of the molecular recognition event; the details of which have been revealed by a crystal structure of the FXIII activation peptide with thrombin [29]. The crystal structure showed residues P10 to P1 of the activation peptide (white rods in Fig. 3F) in a similar conformation as fibrinopeptide Aα (see above), and conserved S1 and aryl-binding pocket (P9 Val) contacts. GpIbα Glycoprotein Ibα is one of four integral membrane proteins comprising the platelet receptor complex Gp Ib-IX-V (for reviews see Refs. [30, 31]). GpIbα interacts with von Willebrand factor in a molecular event which adheres platelets to the site of tissue damage, but it also serves as a thrombin cofactor for at least three substrate cleavage events (PARs 1 and 4, and FXI, see below). Two very different crystal structures of GpIbα complexed to thrombin were reported in back-to-back articles in 2003 [32, 33]. The structures show different GpIbα interactions with both thrombin exosites I and II, and some considerable effort was made to reconcile the structures with one another [34, 35] and with the fact that solution studies conclusively limit the interaction to exosite II. This has led to some confusion in the field. However, it is clear from a careful examination of the biochemical and structural data that only the thrombin interactions involving exosite II are relevant to the cofactor activity of GpIbα. Two mutagenesis studies published in 2001 are of crucial importance in arriving at this conclusion. Wardell et al. [36] showed that exosite I variants R73E and R70E had no effect on thrombin binding to GpIbα, but that exosite II variants R93E and K236E decrease the affinity of the interaction by 10 and 25-fold, respectively. In addition, the weak ligand heparin, which binds to exosite II, decreased the affinity of the thrombin-GpIbα interaction, but the very strong exosite I ligand hirudin had no effect. The second study by De Cristofaro et al. [37] similarly found that alanine mutations in exosite I had no effect on GpIbα binding, but that those in exosite II decreased affinity appreciably. In addition, the loss of thrombin affinity for GpIbα correlated with the magnitude loss in platelet activation activity. When the two GpIbα-thrombin structures are compared with one another, that determined by Dumas et al. stands out as the most consistent with the biochemical data. The thrombin-binding region of GpIbα is composed of 10 negatively charged side chains: D269, E270, D272, D274, D277, E281, E282, D283, and two sulfated tyrosines at 276 and 279 [38, 39]. In the structure by Celikel et al., only three salt bridges between exosite II and the acidic region are found; whereas a much more extensive interaction is observed in the structure reported by Dumas et al., involving five ionic interactions. In the Dumas structure residues R126, K236, K233, R101 and K240 make salt bridges, and K235 and R93 are in proximity to contribute. The dependence of the GpIbα-thrombin interaction on ionic strength is also consistent with the Dumas structure, revealing an electrostatic contact of four-to-five ionic interactions [36] (shown as exosite II-docked white rods in Fig. 3G,H). PARs In one of the most important procoagulant actions of thrombin, platelets are activated by the cleavage of G-protein coupled, protease-activated receptors (PARs) 1 and 4 (for review see Ref. [40]). Thrombin cleaves PARs at a single site resulting in the release of the N-terminal activation peptide. PAR1 is the primary thrombin receptor on platelets, requiring picomolar thrombin concentrations for effective activation, while PAR4 cleavage is only relevant at high thrombin concentrations (>10 nm). The rapidity of PAR1 cleavage is conferred by exosite interactions. The extended substrate sequence of PAR1 contains a hirudin-like domain C-terminal to the scissile bond (DKYEPFWEDEE), absent in PAR4, which has been demonstrated by mutagenesis and competition studies to interact with exosite I [41-44]. A cocrystallization study of thrombin with the PAR1 recognition sequence has also been undertaken [45], but with mixed results. Several structures were solved, but in none was the peptide fully bound in a productive fashion. However, one structure (PDB ID code 1NRS) revealed the active site interactions from P4 to P1 (38LDPR41) and another revealed the exosite I interactions (1NRN) with the hydrophobic sequence (51KYEPFW56). At the time of the publication of this study, little was known about the P′ side interactions, but the orientation of the two fragments from the two separate structures was consistent with productive peptide–thrombin interactions. I have made a model of the entire PAR1 peptide–thrombin interaction based on the fixed position of substrate sequence LDPR from 1NRS and the exosite I-interacting residues KYEPFW from 1NRN. The conformation of the intervening sequence was taken from the structure of the antithrombin (AT) reactive center loop in the active site cleft of thrombin [46]. The conserved P1′ Ser was fixed, and the hydrophobic P2′ Leu of AT was replaced by Phe of PAR1 and then fixed. The intervening residues were fit loosely according to the AT structure and energy minimized for 200 steps using the program CNS [47]. The result is shown as white active site and exosite-bound rods in Fig. 3G, and the model is available upon request. In addition to the exosite I and active site interactions, thrombin cleavage of PAR1 (and presumably PAR4) is also exosite II dependent due to the cofactor effect of GpIbα, which accelerates the rate of reaction by six-to-seven-fold [48]. Factor XI It has only recently become clear that thrombin is the preferred physiological activator of FXI [49-52] (and reviewed in Ref. [53]), with efficient activation occuring exclusively on the surface of platelets where thrombin and FXI are colocalized through separate GpIbα interactions. One study has investigated the thrombin residues involved [54] and found direct exosite I interactions between thrombin and FXI, and a dependence of exosite II for GpIbα binding (discussed above). Exosite I residues H71, R73, R75, Y76, R77a, K110, and K109/110 are involved in direct contacts, possibly to the A1 domain of FXI [55] (Fig. 3H). Thrombomodulin Thrombomodulin (TM) [56] is a single chain glycoprotein of 557 amino acids in length which is found in high density on the surface of the endothelial cells which line blood vessels (for recent reviews see Refs. [57-59]). It is composed of five regions: (1) the N-terminal lectin-like domain; (2) six consecutive EGF-like repeats; (3) a proteoglycan-like Ser-Thr-rich region which is highly O-glycosylated; (4) a single transmembrane helix; and, (5) the C-terminal cytoplasmic tail (Fig. 4). The molecular details of the thrombin–TM interaction were revealed in 2000 by the structure of thrombin bound to a minimal TM fragment [60] containing the final three EGF-like repeats, TME456 [61]. As expected from biochemical studies [62-66], EGF repeats five and six interact directly with exosite I of thrombin, with a buried interface of ∼900 Å2 (yellow ribbon in Fig. 3I,J). While the interface interposes surfaces with complementary electrostatic properties, only one salt bridge was observed between thrombin and TM (Lys110 of thrombin and Asp461 of TM), and the majority of the energy of the interaction is provided by hydrophobic contacts. The more basic exosite II of thrombin can also be involved in the TM interaction [67]. A chondroitin sulfate (CS) moiety found on about 20% of TM molecules in the vascular endothelium and about 30%–35% of those lining the arteries [68] improves the affinity of TM for thrombin by an order-of-magnitude [69, 70]. Chondroitin 4-sulfate [71] is similar to heparin in its composition and sulfation pattern and would be predicted to bind in a similar fashion (rods in Fig. 3I,J). The avidity of the thrombin–TM interaction is great, with an overall dissociation constant in the nanomolar range [66, 72]. Such a tight interaction ensures that TM will effectively compete for any thrombin that has diffused away from the clot onto the surface of the adjacent intact endothelium. Once bound to TM, thrombin is no longer capable of cleaving any of its many procoagulant substrates [73, 74] (with the exception of the fibrinolysis inhibitor TAFI), but its ability to activate the anticoagulant protein C (PC) is enhanced by three orders-of-magnitude. Figure 4Open in figure viewerPowerPoint The mechanism by which TM confers thrombin specificity for PC. TM (yellow) is an integral membrane protein, which is expressed on the surface of endothelial cells which line the vasculature. It is a multidomain protein composed of an N-terminal lectin-like domain, six consecutive EGF-like domains, a Ser/Thr-rich proteoglycan domain, a single transmembrane helix, and a small C-terminal cytoplasmic domain. The proteoglycan domain is heterogeneous in its sulfation, so that about one third of TM molecules possess a single highly sulfated CS moiety. Thrombin (red, IIa) which diffuses away from the clot can encounter TM on the surface of the intact endothelium, where it engages the final two EGF domains of TM in a tight interaction on exosite I. Exosite II will also be occupied by the CS moiety of TM, when present, significantly strengthening the interaction between thrombin and TM. In this manner, thrombin is incapable of interacting with any of its procoagulant substrates which rely on exosite I, and to some degree, exosite II binding. Thrombin specificity for PC (green) is partially conferred by their colocalization on the endothelial cell surface through the docking of the Gla domain of PC onto its receptor EPCR; this also ensures correct positioning of the activation loop of PC relative to the active site of TM-docked thrombin. The encounter complex is further stabilized by direct contact between TM EGF domain 4 and the body of PC. The result of this quarternary complex is the activation of protein C (APC), which then shuts down thrombin generation by inactivating essential cofactors Va and VIIIa. Protein C How TM improves PC recognition by thrombin is illustrated in Fig. 4. It is clear that the tight interaction of EGF domains five and six with exosite I of thrombin effectively blocks the binding of any procoagulant substrate which depends on direct or cofactor dependent exosite I contacts (e.g. fibrinogen, FV, FVIII, FXI, FXIII, and PAR1). In addition, EGF domain four of TM directly interacts with a basic patch on PC [75], thus explaining why efficient recognition of PC