Mapping the Location of Autoinhibitory B-Domain Motifs within Factor V: New Insights Into How Factor V Activation Unfolds

Abstract 495

Blood coagulation factor V (FV) circulates as a procofactor with little or no procoagulant activity. Proteolytic removal of a large central B-domain converts FV (domain organization of A1-A2-B-A3-C1-C2) to an active cofactor for membrane-bound FXa to form prothrombinase. Recently we found that discrete, evolutionarily conserved sequences within the B-domain serve an autoinhibitory function and are necessary to maintain FV as an inactive procofactor. The autoinhibitory region within the B-domain (termed the procofactor regulatory region or PRR) is sequence specific and remarkably short (∼100 amino acids out of 836 from B-domain) consisting of basic (963–1008; basic region or BR) and acidic (1493–1537; acidic region or AR) sequences. Dismantling this region appears to be the driving force to unveil a high affinity binding site(s) for FXa and thus underpins the procofactor to cofactor transition. While the BR and AR seem to work together to provide “on-site” repression of cofactor activity, there is no structural information about the FV B-domain that could provide clues into how this critical region functions. Here we use B-domain fragments tethered to an artificial protease in order to map the position of the PRR relative to the rest of the FV molecule. A BR peptide with an introduced free Cys at position 990 was stoichiometrically modified with FeBABE (Fe-p-bromoacetamidobenzyl-EDTA). FeBABE is a protein cutting reagent with a sulfhydryl-reactive moiety for attachment to Cys and an EDTA-chelated iron atom which, when triggered with ascorbic acid and peroxide, generates hydroxyl radicals that cut peptide bonds in a sequence-independent fashion. When a protein labeled with FeBABE binds its target, the Fe-BABE moiety facilitates cutting if it is orientated correctly and near (<12 Å) the contact site of the target protein. In our experimental system, we employed a constitutively active B-domainless form of FV (FV-810) that harbors the AR but lacks the BR. Initial studies revealed that the BR peptide, without or tethered with FeBABE, binds to FV-810 with high affinity (nM) and inhibits cofactor activity. Proteolysis of FV-810 by BR-FeBABE, was visualized by Western blot, using monoclonal antibodies recognizing either the heavy or light chains. Using this approach, we consistently found that BR-FeBABE cut FV-810 predominantly at two sites. The first is at the C-terminal end of the B-domain. This region is enriched in acidic residues and represents the AR defined above. The second site of cleavage is within the heavy chain. While we cannot exclude the possibility that the site of contact is at the N-terminus of the A1 domain, based on the proposed 3D structure of FVa, it is more likely that the BR is binding near the C-terminal end of the A2 domain which is also enriched in acidic amino acids. Compared to a panel of FVa derivatives with variably truncated heavy chains, we suggest that BR-FeBABE cuts FV-810 near residues 675–685. Overall the data suggest that the B-domain must be folded in such a way as to place the BR in close spatial proximity to the AR at the C-terminal end of the B-domain and also to the C-terminal heavy chain region. Consistent with the idea that these interacting motifs work to keep FV as a procofactor by suppressing FXa binding, adding saturating amounts of FXa, but not the zymogen FX, prevented the BR-FeBABE induced cleavage of FV-810. Similarly, activated protein C (APC) which is thought to share a related binding site with FXa, also prevented BR-FeBABE from cutting FV-810. Addition of excess unlabeled BR peptide substantially reduced cutting of FV-810 by BR-FeBABE, indicating that proteolysis is not due to non-specific cleavage. Additionally, removing the basic charge of BR-FeBABE via acetylation eliminated binding to FV-810 and also any signs of specific cutting. Further, FV-derivatives harboring both the AR and BR in cis and hence not binding the BR peptide, were not cut by excess BR-FeBABE added in trans. In conclusion, these findings provide new insights into the spatial relationship between key regulatory B-domain motifs and the core heavy/light chain region. These interacting autoinhibitory motifs directly or indirectly obscure FXa binding and hence stabilize the inactive procofactor state of FV. These studies also document a remarkably powerful way to map the spatial relationship between regions of proteins and provide new information on coagulation factor binding sites.