The deposition of fibrin is required to prevent bleeding after vascular injury. However, unregulated excessive clotting can result in the occlusion of the blood vessels and thrombosis. Thus, the control of procoagulant mechanisms is essential. The regulation of the coagulation occurs at multiple levels, either through specific direct inhibition of clotting enzymes or by modulation of the function of enzyme cofactors. The major physiological anticoagulant mechanisms include 1) tissue factor pathway inhibitor (TFPI), 2) the activated protein C (APC) pathway and 3) antithrombin. Inherited deficiencies of components of these pathways caused by gene defects can predispose individuals to thrombosis.

The Haemostasis group is conducting research on molecular mechanisms involved in the TFPI and APC anticoagulant pathways through a combination of a BHF-funded fellowship (Dr Josefin Ahnström) and BHF-funded or Rosetrees Trust-funded project grants. Current work focuses on identification and characterisation of protein-protein interactions between components of each of these two pathways and how/when they drive specific complex formation with their procoagulant target. Through the identification of the interaction sites between each protein and kinetic evaluation of inhibition/inactivation of important coagulation factors, we aim to further establish the physiological importance of specific anticoagulant mechanisms, and identify potentially novel mechanisms/targets for therapeutic intervention.

Figure 1: Amino acid residues directly involved in the protein S cofactor function for APC. Molecular models of the N-terminal domains of protein S (left) and protein C (right). We identified an APC variant, D36A/L38D/A39V, which had normal anticoagulant activity in the absence of protein S but could not be enhanced by protein S, suggesting these residues (highlighted in blue) might form part of a functional interaction site with protein S (Preston et al. JBC, 2006). This fits well with results from earlier studies that have suggested that the N-terminal domains of protein S, i.e. the Gla, TSR, EGF1 and EGF2 domains, contain interaction sites needed for the enhancement of APC. We confirmed the importance of the protein S Gla and EGF1 domains for the enhancement of APC. Protein S Asp95, located within the EGF1 domain, is essential for protein S to function as a cofactor for APC (Andersson et al. Blood, 2010). Adjacent residues Asp78 and Gln79 were shown to contribute to the enhancement. Also Gla36, located within the Gla domain is essential for the APC cofactor function (Ahnström et al. Blood, 2011). This residue is of additional interest because it identified a new role for a Gla residue, other than the well-known function of calcium-coordination leading to phospholipid binding. Protein S residues essential for enhancement of APC are highlighted in red.

Figure 2: A direct interaction between TFPI and protein S is crucial for protein S to function as a cofactor in the TFPI pathway.  In 2006 protein S was identified as a cofactor for TFPI in the inhibition of the initiation phase of coagulation. We demonstrated that the TFPI/protein S interaction is dependent upon Glu226 (highlighted in red) in the TFPI Kunitz domain 3 (Ahnström et al. 2012). The reciprocal interaction site is located within the LG-subunit in the SHBG-like domain of protein S (highlighted in green; Reglinska-Matveyev et al. Blood 2014). Research aiming to further pin point the TFPI interaction site on protein S, as well as to further characterise the molecular mechanisms around this interaction and enhancement of TFPI, is ongoing.

The Haemostasis group also conducts research projects on von Willebrand factor (VWF) - a key protein that mediates platelet tethering at sites of vascular injury, and ADAMTS13 - a metalloprotease that cleaves VWF and thus modulates its function. Abnormalities or deficiencies of these important proteins may result in thrombotic or haemorrhagic disorders such as thrombotic thrombocytopenic purpura and von Willebrand disease, respectively and are also risk factors for cardiovascular disease.

Current ADAMTS13 research activity includes structure/function studies that aim to delineate how ADAMTS13 interacts with and proteolyses VWF, how the high degree of substrate specificity of ADAMTS13 is conferred, and the pathogenicity of anti-ADAMTS13 autoantibodies in acquired TTP patients.

Through a combination of a BHF-funded Programme (Professor David Lane), a BHF Fellowship (Dr Jim Crawley) and several BHF-funded Project and Studentship grants, we have provided a model for the specific recognition and proteolysis of unravelled VWF by ADAMTS13 (Crawley et al Blood 2011).

Figure 3: Molecular interactions between ADAMTS13 and the unravelled VWF A2 domain. Molecular model of the ADAMTS13 N-terminal domains - metalloprotease (MP), disintegrin-like (Dis), thrombospondin type 1 repeat (TSP), cysteine-rich (Cys) and spacer domain. Underneath is the aa sequence of the unraverlled VWF A2 domain residues 1599 to 1668. The amino acids in VWF that interact with specific exosites in ADAMTS13 are highlighted.

1. VWF circulates in a globular conformation that is resistant to ADAMTS13 proteolysis because the cleavage site is buried in the folded VWF A2 domain. The folded A2 domain conformation is maintained by structural features including a vicinal disulphide bond (Luken et al Blood 2009), a Ca²⁺binding site (Lynch et al Blood 2014) and an N-linked glycan (McKinnon et al Blood 2008)
2.  ADAMTS13 also circulates in a folded form in which its C-terminal tail folds back upon the N-terminal domains limiting exposure of the spacer domain exosite. (South et al PNAS 2014)
3. When shear forces induce VWF to adopt its string-like conformation during a) secretion from the endothelium, b) passage through the high shear microvasculature, or c) at sites of vessel injury - only then may the VWF A2 domain unravel and expose the cleavage site. 
4. Interactions between the ADAMTS13 C-terminal tail with the D4-CK region of VWF opens ADAMTS13 to expose exosite(s) within its N-terminal domains (Fig 1B). (Zanardelli et al Blood 2009, South et al PNAS 2014)
5. The ADAMTS13 Spacer (Pos et al Blood 2009), Cys-rich (de Groot et al Blood 2015) and Dis domains (de Groot et al Blood 2009) sequentially interact with the unfolded A2 domain in VWF and facilitate the interaction of the Ca²⁺ and Zn²⁺-dependent MP domain (Gardner et al Blood 2009) with the target scissile bond (Tyr1605-Met1606) in VWF, enabling proteolysis to occur (de Groot et al Blood 2010, Xiang et al PNAS 2011).

Currently research through a combination of MRC and BHF funded project grants aim to uncover the molecular and structural basis of ADAMTS13 activation and function, as well as its therapeutic potential in thrombotic disease.

Plasma levels of VWF are closely dependent on its numerous O- and N-linked sugar chains. Current research projects are exploring the functional consequences of N- and O-linked glycosylation on VWF conformation and on its many interactions with other molecules. Recent work has shown that the N-linked glycans can modulate the amount of free thiols present on the surface of VWF. Free thiols have been suggested to mediate VWF self-association especially under flow conditions. We are now using a variety of flow based assays to investigate the functions of thiols in VWF interactions and consequently the underlying role of the VWF C-terminus which is largely unknown but is a region rich in cysteine residues. These flow based assays of VWF function are also being used to determine the key elements of the extracellular matrix to which VWF binds.

O-linked glycans which are clustered in groups flanking the VWF A1 domain appear to determine the stiffness of the these flanking linker regions and thus determine the ability of the A domains to interact with their partners and for A2 to be cleaved by ADAMTS13. These functional studies are supported by collaborative projects (Professor Anne Dell, Imperial College) which are producing a detailed glycan map of VWF, its modulation by ABO blood group and exploring the role of VWF in endothelial cell function.

Despite our considerable understanding of VWF and platelet function, a large proportion of mild bleeding phenotypes remain unexplained. We are therefore participating in a major NIHR -funded exome sequencing project (BRIDGE project) to discover novel genetic explanations for these disorders.

Platelets are central to normal haemostasis and defects in both platelet number and function are causes of both bleeding and thrombotic diseases in humans. Platelets are formed by megakaryocytes in the bone marrow and thereafter released into the blood. The formation and function of platelets is influenced by multiple factors, among those are cell surface receptors that transduce molecular signals. Platelets are also important mediators in inflammatory and cancer diseases. The Haemostasis group is currently exploring novel haemostatic modulators through a BHF-funded Fellowship (Isabelle Salles-Crawley) and BHF funded Project and Studentship grants.

As part of the Bloodomics consortium effort to discover novel genes linked to the development/severity of coronary artery disease and myocardial infraction, BAMBI (Bone morphogenetic protein and Activin Membrane-Bound Inhibitor), along with other novel genes was selected from a panel of transmembrane proteins that were highly expressed in megakaryocytes and HUVECS in comparison to other blood cell lineages, but hitherto unknown roles in platelet function and thrombus formation (O’Connor, Salles et al, Blood, 2009). We have recently shown that BAMBI plays a role in haemostasis and also in thrombus stability using intravital microscopy (Salles-Crawley et al, Blood 2014). BAMBI deficiency had no effect on platelet count and platelet function as well as no influence in the ability of thrombin generation in plasma (central to coagulation). We also demonstrated that BAMBI present in the endothelium rather than in platelets is associated with the thrombus instability defect. The Haemostasis lab is currently examining how BAMBI influences thrombus formation via modulating normal endothelial homeostasis. Another arm of research consists in understanding how BAMBI may influence inflammatory diseases such as atherosclerosis.