Activated platelets support factor VIII (FVIII) activity at 100-fold lower FVIII concentration than phospholipid vesicles (PLV) but the mechanism(s) remain unknown. Therefore, we compared activity of FVIIIa vs. emicizumab using platelets and PLV. We also utilized engineered B-domain deleted human factor VIII (FVIIIBDD) with degradation-resistant primary APC cleavage sites R336Q/R562Q (FVIIIQQ) or a stabilized A2 domain, D519V/E665V (FVIIIVV), or both (FVIIIQQVV). FVIII-dependent clotting was faster with platelets than PLV. The relative platelet:PLV FVIII procoagulant activity varied from 40:1 to 2:1; the highest ratios achieved with lower FVIII concentrations. In contrast, activity of Emi was similar with platelets and PLV. Steady state kinetic experiments indicated that platelets do not convey superior Xase activity. This indicated that FVIII binding sites protect or stabilize FVIII(a). FVIIIVV and FVIIIQQVV supported equivalent clot times on platelets and PLV suggesting that platelets slow dissociation of the A2 subunit. Studies with purified proteins confirmed that FVIIIa dissociates more slowly on platelets than on PLV. Addition of APC prevented PLV-supported clotting but modestly slowed clotting with FVIIIQQ and PLV or with FVIIIBDD and platelets. This suggested that platelets protect FVIII from APC. Defined studies confirmed that platelets slowed APC degradation of FVIII(a) to about the same extent as inactivation of FVIIIQQ with PLV. Platelet microparticles slowed FVIII degradation to a degree that was intermediate between PLV and platelets. Together, these data indicate that platelets support greater activity of FVIII than PLV through stabilization against dissociation of the A2 domain and protection from degradation by APC.

Prior reports indicate that the convex membrane curvature of phosphatidylserine (PS)-containing vesicles enhances formation of binding sites for factor Va and lactadherin. Yet, the relationship of convex curvature to localization of these proteins on cells remains unknown. We developed a membrane topology model, using phospholipid bilayers supported by nano-etched silica substrates, to further explore the relationship between curvature and localization of coagulation proteins. Ridge convexity corresponded to maximal curvature of physiologic membranes (radii of 10 or 30 nm) and the troughs had a variable concave curvature. The benchmark PS probe lactadherin exhibited strong differential binding to the ridges, on membranes with 4% to 15% PS. Factor Va, with a PS-binding motif homologous to lactadherin, also bound selectively to the ridges. Bound factor Va supported coincident binding of factor Xa, localizing prothrombinase complexes to the ridges. Endothelial cells responded to prothrombotic stressors and stimuli (staurosporine, tumor necrosis factor-α [TNF- α]) by retracting cell margins and forming filaments and filopodia. These had a high positive curvature similar to supported membrane ridges and selectively bound lactadherin. Likewise, the retraction filaments and filopodia bound factor Va and supported assembly of prothrombinase, whereas the cell body did not. The perfusion of plasma over TNF-α-stimulated endothelia in culture dishes and engineered 3-dimensional microvessels led to fibrin deposition at cell margins, inhibited by lactadherin, without clotting of bulk plasma. Our results indicate that stressed or stimulated endothelial cells support prothrombinase activity localized to convex topological features at cell margins. These findings may relate to perivascular fibrin deposition in sepsis and inflammation.

The cocrystal structure of antibody 2A9 and factor VIII (fVIII), reported by Gish et al(1) in this issue of Blood, delineates an epitope on the C1 domain and novel mobility of the C2 domain, both relevant to membrane binding.

Coagulation factor (F)V is a critical procoagulant protein, functioning as a cofactor in the prothrombinase complex. However, FV is also an anticoagulant protein, serving as a cofactor for activated protein C (APC) and for tissue factor pathway inhibitor (TFPI)α. Presently, the balance and relative importance of these opposing functions of FV is not well understood. In this issue of J Thromb Haemost, Castoldi et al. investigated a patient who is homozygous for a novel mutant FV, FV Besanҫon. By causing a strong thrombotic phenotype even though the plasma has only 3% of normal FV activity, this mutant FV helps to illustrate the importance of maintaining the balance between the pro- and anticoagulant functions of FV. Studies of FV Besanҫon suggest that the dominant role of FV in normal people is anticoagulant, preventing thrombosis. Even though FV Besanҫon has less than 4% of normal FV capacity to generate thrombin, it is sufficient to prevent bleeding. However, because the anticoagulant activity is attenuated more than the procoagulant activity, the phenotype is thrombotic. As detailed below, these studies also indicate that the mechanisms involved in achieving full functionality of FV Besanҫon on activated platelets differs from phospholipid vesicles, thus implicating a role for an unidentified FV cofactor or modifier on the platelet surface. In contrast to the procoagulant role of FV, which requires proteolytic removal of the B domain, the anticoagulant properties of FV require proportions of the FV B domain to be present.1-3 In the APC pathway, full-length FV and protein S act together as synergistic cofactors in the inactivation of FVIIIa.4, 5 FV has more recently also been identified as a cofactor for TFPIα. Here, full-length, procofactor FV enhances TFPIα-mediated FXa inhibition together with protein S.2, 3 Alternatively, a partially activated FV, retaining a portion of the B domain, inhibits the prothrombinase complex through TFPIα. Likewise, the partially activated form of FVa released from platelets enhances TFPIα in the inhibition of the prothrombinase complex.1 Complete removal of the B domain, upon complete activation by thrombin, leaves activated FV with only procoagulant activity.1 Here, Castoldi et al. report a partially FV-deficient patient suffering from severe thrombophilia. The patient was found to be homozygous for a novel mutation in the FV C2-domain (Ala2086Asp, FV Besanҫon). The mutation caused significant FV deficiency (4% of normal FV levels) with in vitro studies indicating that it has impaired cellular secretion. The FV Besanҫon antigen was also less effective in the prothrombinase complex, yet the patient paradoxically suffered from thrombophilia. This could be explained, in part, by the associated reduction in TFPIα levels (~24% of normal), as is common in FV-deficient individuals.6 In fact, the authors themselves have previously shown that reduced TFPIα levels can limit the bleeding phenotype that would be expected as a result of FV-deficiency. The relatively high TFPIα levels in the patient's platelet rich plasma (PRP) compared to platelet poor plasma (PPP) are consistent with the observation that FV functions as a carrier of TFPIα in plasma, because no such carrier function is needed for platelet TFPIα.7, 8 However, while the reduction in plasma TFPIα levels could explain an absence of bleeding, it could not explain the thrombotic phenotype of the patient. A number of reported FV mutations cause thrombophilia.9, 10 FV Leiden (Arg506Glu), FV Nara (Trp1920Arg), and FV-short (missing a.a. 756–1458) are three such mutations. FV Leiden, the most common risk factor for venous thrombosis among Whites, results in APC resistance due to a mutation of one of the APC cleavage sites.9, 11 FV Leiden has also recently been shown to have reduced TFPI cofactor function in the inhibition of prothrombinase formation and activity.12 FV Nara has an amino acid substitution in the C1 domain, which results in APC resistance in association with reduced affinity toward negatively charged phospholipid membranes and, likely, a reduced affinity for protein S.13, 14 FV-short, a partially B domain deleted FV is constitutively active, predicting increased prothrombotic activity.7 However, FV-short causes a severe bleeding phenotype through increasing TFPIα levels and function.7, 15 Castoldi et al. hypothesized that, similar to these previously described FV mutations, the cause of thrombophilia in the affected patient was a result of strongly impaired anticoagulant properties of FV Besanҫon. Because recombinant FV Besanҫon was poorly expressed, the authors instead relied on FV in PPP and PRP from the patient for biochemical assays. FV Besanҫon was activated similarly to normal FV and had a relatively mild reduction in cofactor function for FXa. In contrast, the anticoagulant properties of the protein were substantially impaired. There was a degree of impairment in defined FVa and FVIIIa inactivation assays, though inadequate to explain the thrombotic phenotype. However, the plasma-based thrombin generation assays provided the convincing evidence of marked impairment of anticoagulant function. The authors show a reduced sensitivity of the patient plasma toward both TFPIα and APC, compared to plasma collected from healthy volunteers. The resistance to both APC and TFPIα was much stronger in assays performed with platelets compared to PPP supplemented with phospholipid vesicles (Figure 1). In fact, the phenotype of the FV Besanҫon mutation would appear inexplicable if only studied on phospholipid vesicles. The discrepancy is particularly clear from the experiments assessing the efficiency of the TFPI pathway. Despite only containing 24% of normal TFPIα levels, the TFPI pathway in PPP supported by vesicles remains functional. In contrast, TFPIα anticoagulant function was barely detectable in PRP, containing 50% normal TFPIα. Similarly, addition of increasing concentrations of APC to the patient's PPP supported by phospholipid vesicles showed only a modest resistance to APC, while the patient's PRP was significantly resistant to APC. In fact, the APC resistance observed in the patient's PRP was comparable to that of an individual homozygous for the FV Leiden mutation. Both FV and TFPIα are present in platelets and released upon their activation.16 However, the levels and functionality of platelet FV and TFPIα in the patient carrying the FV Besanҫon mutation have not been investigated in the present study and it is unclear how these may influence the differences in sensitivity toward TFPIα and APC in PRP compared to PPP. Furthermore, it has previously been proposed that the mode of binding of FVa to platelets differs from that to phospholipid vesicles and that platelets may express a surface protein that binds and modulates FVa activity.17 The prior reports imply that platelets modify procoagulant function of FVa. In contrast, the results presented by Castoldi et al. now implicate that an unidentified platelet modifier influences the anticoagulant activity of FV (Figure 1). The FV Besanҫon mutation is located within the FV C2 domain. The importance of the FV C-domains for binding to negatively charged phospholipid membranes is well known, as is the importance of the C2 domain specifically.18 Binding to membrane surfaces is critical for all roles of FV.14, 18 While the authors failed to demonstrate any reduction in affinity of FV Besanҫon toward negatively charged membranes in direct binding assays, the authors showed that FV Besanҫon behaved similar to FV1, a natural FV isoform with reduced affinity for negatively charged phospholipids due to a glycosylation at Asn2181 in the FV C2-domain. Biochemical assays also suggested that the FV Besanҫon mutation has a modestly reduced affinity toward phospholipid membranes. However, the relationship between membrane affinity and impaired anticoagulant function for FV Besanҫon is not yet well understood. The balance between the pro- and anticoagulant roles of normal FV are illuminated by this intriguing study. It appears that less than 5% of FV is enough for supporting the thrombin generation needed to avoid major bleeding. In contrast, the anticoagulant functions of FV appear to be much more sensitive to any reduction in plasma FV levels.3, 14, 19, 20 However, reduced levels of FV are unlikely the only cause of thrombophilia in the patient carrying the FV Besanҫon mutation, highlighting this particular residue's involvement in phospholipid binding and/or interactions with other proteins involved in the anticoagulant pathways, making FV Besanҫon an inefficient anticoagulant regulator. Further, the study also shows that the molecular mechanisms involved in FV anticoagulant functions on a platelet surface differ from those on a phospholipid vesicle, suggesting that FV is modified by an unidentified factor on platelet membranes. The authors have no conflicts of interest to disclose. J.A. and G.E.G. wrote the manuscript.

Recent reports indicate that suspended skeletal and cardiac myosin, such as might be released during injury, can act as procoagulants by providing membrane-like support for factors Xa and Va in the prothrombinase complex. Further, skeletal myosin provides membrane-like support for activated protein C. This raises the question of whether purified muscle myosins retain procoagulant phospholipid through purification. We found that lactadherin, a phosphatidyl-l-serine-binding protein, blocked >99% of prothrombinase activity supported by rabbit skeletal and by bovine cardiac myosin. Similarly, annexin A5 and phospholipase A2 blocked >95% of myosin-supported activity, confirming that contaminating phospholipid is required to support myosin-related prothrombinase activity. We asked whether contaminating phospholipid in myosin preparations may also contain tissue factor (TF). Skeletal myosin supported factor VIIa cleavage of factor X equivalent to contamination by ∼1:100 000 TF/myosin, whereas cardiac myosin had TF-like activity >10-fold higher. TF pathway inhibitor inhibited the TF-like activity similar to control TF. These results indicate that purified skeletal muscle and cardiac myosins support the prothrombinase complex indirectly through contaminating phospholipid and also support factor X activation through TF-like activity. Our findings suggest a previously unstudied affinity of skeletal and cardiac myosin for phospholipid membranes.

Platelet activation supports procoagulant activity through phosphatidylserine exposure, secretion of procoagulant factors, and receptor conformational change. For example, thrombin-stimulated platelets bind factor VIII (fVIII) via a macromolecular complex including oligomeric fibrin and the active αIIbβ3 receptor (Phillips et al, JTH 2004; Gilbert et al, Blood 2015). Thus, coagulation assays in which phospholipid vesicles are substituted for platelets do not fully emulate modulators of fVIII activity. Indeed, inhibition of platelet-supported fVIII activity by a panel of mAbs against the C2 domain was not correlated to inhibition of vesicle-dependent activity (Chatterjee et al, JTH 2020). An obstacle to adoption of a platelet-based assay for fVIII is the need for fresh platelets. Therefore we asked whether cryopreserved platelets might support fVIII activity similarly to fresh platelets. Apheresis platelets were mixed with cryopreservatives with or without calcium chelators, in various aliquot sizes, and frozen on various cooling media. Cryopreserved platelets were compared to non-preserved apheresis platelets with regard to agonist response and support of procoagulant activity. Cryopreservation resulted in an increase in subcellular debris and an unresponsive fraction of platelets with decreased forward scatter judged by flow cytometry. Optimized results were obtained when platelet rich plasma with 5% DMSO, in 1 mL aliquots was frozen on powdered dry ice, and stored at -150C. Purification of thawed platelets using a density gradient removed debris and decreased unresponsive platelets resulting in a forward and side scatter profile comparable to fresh platelets. We refer to these as cryopreserved platelets (CryoPlts). CryoPlts were compared to control and outdated apheresis platelets. As with fresh platelets, procoagulant activity of CryoPlts increased with thrombin receptor agonist peptides (TRAP) 1 & 4 and supported a log-linear relationship between time to initial fibrin strand formation and fVIII activity over a range of 0.0001 - 1 u/mL (Fig 1). Further, the degree of inhibition of fVIII activity by mAbs ESH4 and G99 against the fVIII C2 domain, was the same on control and CryoPlts, but markedly different from inhibition in an aPTT-based inhibitor assay. In contrast, outdated apheresis platelets had increased procoagulant activity, minimal agonist response and a shallow curve with varying fVIII concentration. Flow cytometry studies with lactadherin-FITC indicated that 33 ± 14% of CryoPlts had high PS exposure, and the size of this population was minimally affected by TRAP 1+4. In contrast, the main platelet population had a small, uniform, increment in PS exposure comparable to control platelets. Surprisingly, the PS-rich platelets did not significantly affect the time to fibrin formation, confirming that the viable platelets, with limited PS exposure, provide much of the support for fVIII-related procoagulant activity. Flow cytometry indicated αIIbβ3 activation (PAC1-FITC) and α-granule release (anti-P-selectin-PE) were qualitatively intact on CryoPlts, although staining was decreased 70% for PAC1 and 57% for Psel. We also tested whether CryoPlts may be utilized for evaluating response to anti-PF4-heparin antibodies, relevant to heparin-induced thrombocytopenia (HIT). We evaluated platelet response to platelet factor 4 (PF4) and a platelet-activating anti-PF4 antibody (KKO), a combination that induces activation similar to authentic autoimmune antibodies for HIT. The non-activating PF4 antibody RTO served as a negative control. Geometric mean response, corrected for background, was normalized to response to thrombin activation. Both fresh and CryoPlts responded with increases in PAC1 (Fig 2A) and anti-Psel (Fig 2B) binding in response to KKO/PF4 compared to RTO/PF4 . This data demonstrates that the qualitative αIIbβ3 and P-selectin response to HIT-like antibodies is intact. Our results demonstrate a refined cryopreservation protocol of apheresis platelets. These platelets maintain qualitative agonist responsiveness with near-normal support for factor VIII activity, suggesting that they could be used for other platelet-based laboratory or diagnostic assays. Further, our results suggest that major procoagulant activity is provided by platelets with very limited PS exposure, an area for further investigation. No relevant conflicts of interest to declare.

Factor VIII (FVIII) replacement therapy for hemophilia A is complicated by development of inhibitory antibodies (inhibitors) in ∼30% of patients. Because endothelial cells (ECs) are the primary physiologic expression site, we probed the therapeutic potential of genetically restoring FVIII expression selectively in ECs in hemophilia A mice (FVIIInull). Expression of FVIII was driven by the Tie2 promoter in the context of lentivirus (LV)-mediated in situ transduction (T2F8LV) or embryonic stem cell-mediated transgenesis (T2F8Tg). Both endothelial expression approaches were associated with a strikingly robust immune response. Following in situ T2F8LV-mediated EC transduction, all FVIIInull mice developed inhibitors but had no detectable plasma FVIII. In the transgenic approach, the T2F8Tg mice had normalized plasma FVIII levels, but showed strong sensitivity to developing an FVIII immune response upon FVIII immunization. A single injection of FVIII with incomplete Freund adjuvant led to high titers of inhibitors and reduction of plasma FVIII to undetectable levels. Because ECs are putative major histocompatibility complex class II (MHCII)-expressing nonhematopoietic, "semiprofessional" antigen-presenting cells (APCs), we asked whether they might directly influence the FVIII immune responses. Imaging and flow cytometric studies confirmed that both murine and human ECs express MHCII and efficiently bind and take up FVIII protein in vitro. Moreover, microvascular ECs preconditioned ex vivo with inflammatory cytokines could functionally present exogenously taken-up FVIII to previously primed CD4+/CXCR5+ T follicular helper (Tfh) cells to drive FVIII-specific proliferation. Our results show an unanticipated immunogenicity of EC-expressed FVIII and suggest a context-dependent role for ECs in the regulation of inhibitors as auxiliary APCs for Tfh cells.

BACKGROUND:We recently reported that factor VIII (FVIII) binds to a macromolecular complex including fibrin on thrombin-stimulated platelets and that two antibodies against FVIII diminish platelet-supported FVIII activity more than vesicle-supported activity. The C2 domain of FVIII is known to bind to phospholipid membrane and also binds fibrin.OBJECTIVES:We asked whether the degree of inhibition by anti-C2 antibodies would show differences between platelet-supported and the standard activated partial thromboplastin time (aPTT) assay.METHODS:We evaluated the inhibition by a well-defined panel of monoclonal anti-C2 domain antibodies encompassing the major epitopes of the C2 domain. Activity was measured in an activated platelet time (aPT) assay containing fresh, density gradient-purified human platelets.RESULTS:The aPT exhibited a log-linear relationship between FVIII and time to fibrin formation over a 4-log range, encompassing 0.01% to 100% plasma FVIII. Nine of 10 mAbs inhibited 89% to 96% of FVIII activity, whereas mAb F85 did not. There was no correlation between the degree of inhibition in the aPTT-based assay and the platelet assay. In particular, four mAbs did not inhibit the aPTT assay, yet inhibited 90% of platelet-based activity. Residual FVIII activity in purified-protein assays, relying on platelets, correlated with the aPT assay.CONCLUSIONS:The degree of FVIII impairment by some inhibitor antibodies is substantially different on platelet membranes vs synthetic vesicles. Thus, current inhibitor assays may underestimate the frequency of significant inhibitors, and a platelet-based assay may more accurately assess bleeding risk.

Hemostasis encompasses an ensemble of interactions among platelets, coagulation factors, blood cells, endothelium, and hemodynamic forces, but current assays assess only isolated aspects of this complex process. Accordingly, here we develop a comprehensive in vitro mechanical injury bleeding model comprising an “endothelialized” microfluidic system coupled with a microengineered pneumatic valve that induces a vascular “injury”. With perfusion of whole blood, hemostatic plug formation is visualized and “in vitro bleeding time” is measured. We investigate the interaction of different components of hemostasis, gaining insight into several unresolved hematologic issues. Specifically, we visualize and quantitatively demonstrate: the effect of anti-platelet agent on clot contraction and hemostatic plug formation, that von Willebrand factor is essential for hemostasis at high shear, that hemophilia A blood confers unstable hemostatic plug formation and altered fibrin architecture, and the importance of endothelial phosphatidylserine in hemostasis. These results establish the versatility and clinical utility of our microfluidic bleeding model.

Hemophilia A is caused by decreased or dysfunctional blood coagulation factor VIII (FVIII). Recent developments in the understanding of FVIII biology, in particular the nature of FVIII binding sites on platelets, may provide new insight into the limitations of current assays. Recent data suggest that the phospholipid vesicles, which represent nonphysiologic membranes of high phosphatidylserine (PS) content, poorly reflect functional FVIII binding sites critical to coagulation. This narrative review describes the function of FVIII in clotting and discusses our evolving understanding of FVIII binding sites and their clinical implications. Refined models of FVIII binding sites have the potential to improve FVIII assays, possibly improving bleeding risk stratification for patients with mild and moderate hemophilia A. They may also support earlier and more accurate detection of inhibitors, before they are clinically evident.

Abstract Thrombin-stimulated platelets express binding sites for factor VIII (fVIII) despite exposing insufficient phosphatidylserine (PS) to constitute these binding sites. Our previous work implicated soluble fibrin bound to the αIIbβ3 integrin as an important component of these binding sites. Further, we demonstrated that fVIII has direct, high affinity binding to fibrin, with an associated increase in the activity of the purified Xase complex (Gilbert et al, Blood 2015). However, the stoichiometry between fVIII binding sites and fibrin monomers was at least 30:1, suggesting that fVIII binds a minor fibrin(ogen) variant. We undertook experiments to determine the nature of the fVIII-binding fibrin(ogen). fVIII binding to fibrin(ogen) was evaluated with three complementary assays: i) direct binding of fVIII-fluorescein (fVIII-fluor) to soluble fibrin linked to Superose beads, ii) a competition assay in which binding of fVIII-fluor to fibrinogen competed for binding to Superose-linked von Willebrand factor (VWF) and iii) a factor Xase assay in which fibrin(ogen) enhanced activity. Results were benchmarked against fVIII-fluor binding to thrombin-stimulated platelets. When fibrinogen (depleted of plasminogen, fibronectin, and VWF) was fractionated by anion exchange chromatography, the fibrinogen γA fraction did not bind fVIII. Rather, the fVIII-binding fractions (constituting 10-15% of total fibrinogen) eluted in association with fibrinogen γ'. However, commercially available γ' fibrinogen does not bind fVIII, suggesting that additional modification is necessary or that fVIII-binding fibrinogen (fbgn*) elutes near fibrinogen γ' on the basis of acidity rather than shared structure. Mass spectroscopy analysis indicated that only traces of VWF remained in the active fractions. Addition of mAb 418, which blocks the fVIII binding site on VWF, had no significant effect on binding of fVIII-fluor to fibrin-Superose showing that residual VWF is not the source of high affinity fVIII binding. Two rabbits inoculated with fbgn*-enriched fractions developed anti-sera that block binding of fVIII to fibrin. In contrast, 5 commercially available polyclonal antibodies against fibrin(ogen) had no effect on fVIII binding, verifying that fbgn* contains a distinct epitope relevant to fVIII binding. The fbgn*-enriched fraction competed directly with VWF for fVIII binding with at least a 3-fold increase in fVIII binding sites compared to total fibrinogen. Following treatment with thrombin, the fbgn*-enriched fraction had twice as many binding sites for fVIII compared to fibrin from the starting material and a similar affinity in the direct binding assay. FVIII-fluor bound immobilized fibrin with a KD of 2-4 nM while the competition binding experiment yielded an implied KD to fibrinogen of approx. 100 nM. This suggests that thrombin cleavage of fbgn* or assembly into fibrin protofibrils or fibrils increases fVIII affinity approx. 50-fold. Fbgn* increased activity of the factor Xase complex 2-3 fold. The enhancement was maintained when fbgn* was not exposed to thrombin, indicating that the enhancement was not restricted to the higher affinity binding sites on fibrin. Plasmin degradation of fibrinogen produces five major fragments that can be separated by ion-exchange chromatography. Enhanced Xase activity and binding of digested fbgn* to fVIII in the VWF competition assay were both similar to intact fbgn*. In order to identify the active fragment, plasmin-digested fbgn* was separated using anion exchange chromatography and tested for factor VIII binding and Xase activity. Only the latest eluting peak bound to fVIII-fluor. SDS-PAGE shows that this peak consists primarily of two bands. IEF gel analysis indicates that the isoelectric points of these two bands differ compared to bands from commercially available γ' digested in the same manner. We are currently determining the identities of the unique bands. These results show that fVIII binds to a minor, acidic fibrinogen variant (fbgn*) in a manner that competes with VWF binding. Binding to fbgn* enables fVIII binding to thrombin-stimulated platelets and enhances fVIII activity in the Xase complex. Ongoing experiments are expected to identify the molecular features that distinguish fbgn* from fibrinogens γA and γ' and produce selective antibodies that will enable testing of the physiologic contribution of the fVIII-fbgn* interaction to hemostasis. Download : Download high-res image (119KB) Download : Download full-size image Figure 1 . Disclosures No relevant conflicts of interest to declare.

The major complication of hemophilia A treatment is the development of inhibitory antibodies against factor VIII (fVIII), causing severe bleeding in spite of infused fVIII. Most antibodies are targeted to the A2 or the C2 domain of fVIII and the degree of inhibition by these antibodies measured in clinical fVIII assays fails to predict the severity of bleeding risk. We investigated whether the degree of inhibition of platelet dependent fVIII activity by a panel of monoclonal anti-C2 domain mAbs deviated from the degree of inhibition in a commercial one-stage assay.

Background: Scott Syndrome is a rare bleeding disorder characterized by a defect in platelet phosphatidylserine (PS) exposure. The syndrome has recently been linked to mutations in TMEM16F. Tmem16f -/- mice were recently reported to be viable with a prolonged tail snip bleeding time but no spontaneous bleeding. We now report analysis of an additional gene targeted Tmem16f allele generated in C57BL/6 ES cells. Results: JM8 ES cell were obtained from EUCOMM, and successful Tmem16f gene targeting in intron 1 was confirmed by PCR and sequencing. Genotyping of 120 Tmem16f +/gt (+/gt) intercross progeny identified no surviving Tmem16f gt/gt (gt/gt) mice at weaning (p<0.001). However, +/gt intercrosses generated the expected Mendelian genotype ratios at both E10.5 and E17.5, with gt/gt embryo’s exhibiting no morphological abnormalities on gross or routine histologic examination. Though complete deficiency of TMEM16F is lethal in the C57BL/6J genetic background between E17.5 and birth, an F2 intercross of +/gt mice outcrossed one generation to 129x1SvJ resulted in gt/gt mice surviving to weaning, though at reduced numbers (6/75 total progeny compared to ~19 expected, p <0.002). Progeny testing of surviving gt/gt mice suggest a single autosomal dominant 129x1SvJ-associated genetic modifier. Preliminary genetic analysis of these mice appears to map this locus to the proximal region of chromosome 3. Tail bleeding times for gt/gt were >10min, whereas littermate +/gt and +/+ mice bleeding ceased at 8 ± 1 min and 6 ± 0.8 min, respectively, each significantly different than gt/gt (p<0.05). Notably, platelets from +/gt mice exhibited a trend toward reduced PS exposure, detected with FITC-labelled lactadherin, in response to PAR4 agonist peptide, whereas gt/gt mice had significantly reduced PS exposure (p < 0.05). Conclusion: These data suggest the existence of a viability-determining genetic modifier of TMEM16F in the 129x1SvJ mouse strain. Identification of the responsible gene may uncover novel functions for TMEM16F and the regulation of hemostatic function.

In this issue of Blood , [Batsuli et al][1] demonstrate that pathogenic antibodies against the factor VIII (fVIII) C1 domain are more common than previously appreciated and that they can cause bleeding through a novel mechanism.[1][2] ![Figure][3] Influence of anti-C1 domain antibodies on

Inhibitory antibodies against factor VIII (fVIII) are the major complication of hemophilia A treatment, causing severe bleeding in spite of infused fVIII. Most antibodies are directed against the A2 or the C2 domain of fVIII and the degree of inhibition in clinical fVIII assays does not predict the severity of bleeding risk. We recently discovered that fVIII binds to a macromolecular complex including low molecular weight fibrin and the αIIbβ3integrin on membranes of thrombin-stimulated platelets rather than exposed phosphatidylserine (PS). Because clinical fVIII assays utilize PS-rich vesicles to support fVIII activity we asked whether the degree of inhibition of platelet-dependent activity by anti-C2 domain mAb’s differs from inhibition on PS-rich phospholipid vesicles.

The mechanisms contributing to an increased risk of thrombosis in uremia are complex and require clarification. There is scant morphological evidence of membrane-dependent binding of factor Xa (FXa) and factor Va (FVa) on endothelial cells (EC) in vitro. Our objectives were to confirm that exposed phosphatidylserine (PS) on microparticle (MP), EC, and peripheral blood cell (PBC) has a prothrombotic role in uremic patients and to provide visible and morphological evidence of PS-dependent prothrombinase assembly in vitro. We found that uremic patients had more circulating MP (derived from PBC and EC) than controls. Additionally, patients had more exposed PS on their MPs and PBCs, especially in the hemodialysis group. In vitro, EC exposed more PS in uremic toxins or serum. Moreover, reconstitution experiments showed that at the early stages, PS exposure was partially reversible. Using confocal microscopy, we observed that PS-rich membranes of EC and MP provided binding sites for FVa and FXa. Further, exposure of PS in uremia resulted in increased generation of FXa, thrombin, and fibrin and significantly shortened coagulation time. Lactadherin, a protein that blocks PS, reduced 80% of procoagulant activity on PBC, EC, and MP. Our results suggest that PBC and EC in uremic milieu are easily injured or activated, which exposes PS and causes a release of MP, providing abundant procoagulant membrane surfaces and thus facilitating thrombus formation. Blocking PS binding sites could become a new therapeutic target for preventing thrombosis.

Thrombin-stimulated platelets expose very little phosphatidylserine (PS) but express binding sites for factor VIII (fVIII), casting doubt on the role of exposed PS as the determinant of binding sites. We previously reported that fVIII binding sites are increased three- to sixfold when soluble fibrin (SF) binds the αIIbβ3 integrin. This study focuses on the hypothesis that platelet-bound SF is the major source of fVIII binding sites. Less than 10% of fVIII was displaced from thrombin-stimulated platelets by lactadherin, a PS-binding protein, and an fVIII mutant defective in PS-dependent binding retained platelet affinity. Therefore, PS is not the determinant of most binding sites. FVIII bound immobilized SF and paralleled platelet binding in affinity, dependence on separation from von Willebrand factor, and mediation by the C2 domain. SF also enhanced activity of fVIII in the factor Xase complex by two- to fourfold. Monoclonal antibody (mAb) ESH8, against the fVIII C2 domain, inhibited binding of fVIII to SF and platelets but not to PS-containing vesicles. Similarly, mAb ESH4 against the C2 domain, inhibited >90% of platelet-dependent fVIII activity vs 35% of vesicle-supported activity. These results imply that platelet-bound SF is a component of functional fVIII binding sites.

Scott Syndrome is a rare, moderately severe bleeding disorder caused by a defect in platelet, red cell, and lymphocyte phosphatidylserine exposure. The syndrome has been linked to mutations in TMEM16F, a Ca++-activated ion channel. A moderately severe bleeding disorder in German Shephard dogs, characterized by decreased platelet phosphatidylserine exposure, has also linked to mutations in TMEM16F. TMEM16F, is a member of a recently-identified family of calcium-activated chloride channels that are also called anoctamins. Members of the family apparently serve both as ion channels and phospholipid scrambling channels. A crystal structure of one member of the TMEM16 family shows a homo-dimeric structure of an integral membrane protein with each unit containing 10 transmembrane helices. A transmembrane hydrophilic cavity lies at the interface, and contains a cation binding site, and a slot with dimensions that could accommodate acyl chains. Tmem16f-/- mice were recently reported to be viable with a prolonged tail snip bleeding time but no spontaneous bleeding. We have developed an independent gene-targeted (gt) Tmem16f allele generated in C57BL/6 ES cells and find results that both confirm prior reports and contrast with them. JM8 ES cells were obtained from EUCOMM, and Tmem16f gene targeting in intron 1 was confirmed by PCR and sequencing. Genotyping of 120 Tmem16f+/gt intercross progeny identified no surviving Tmem16fgt/gt mice at weaning (p<0.001). However, +/gt intercrosses generated the expected Mendelian genotype ratios at both E10.5 and E17.5, with gt/gt embryos. Blinded pathological evaluation of E17.5 gt/gt pups indicated reduction of ossification and angular limb changes, consistent with 2 prior reports. Thus, our results confirm previously reported bone changes, but in contrast with prior reports, indicate that Tmem16f deficiency is lethal in the C57BL/6 genetic background. An F2 intercross of +/gt mice outcrossed one generation to 129x16vJ resulted in gt/gt mice surviving to weaning, though at only 30% of the expected Mendelian frequency. Progeny analysis points to a single autosomal dominant 129x1SvJ-associated genetic modifier. Preliminary genetic analysis of these mice appears to map this locus to the proximal region of chromosome 3. Further efforts to localize the responsible gene(s) are underway. Tail bleeding times for gt/gt were >10min, whereas littermate +/gt and +/+ mice bleeding ceased at 8 ± 1 min and 6 ± 0.8 min, respectively, each significantly different than gt/gt (p<0.05). Notably, platelets from +/gt mice exhibited a trend toward reduced PS exposure, detected with FITC-labelled lactadherin, in response to PAR4 agonist peptide, whereas gt/gt mice had significantly reduced PS exposure (p < 0.05). gt/gt platelets showed a trend toward reduced PS exposure in response to A23187, as well as prolonged platelet rich plasma clotting times, and less efficient lactadherin inhibition of platelet clotting time. Our data suggest the existence of a viability-determining genetic modifier of the TMEM16F deficiency phenotype in the 129x1SvJ mouse strain. Identification of the responsible gene may uncover a novel regulator of hemostatic function. The observation that heterozygous deficiency leads to a PS exposure and hemostatic phenotype also suggests the possibility that heterozygous TMEM16F mutations may influence hemostasis or thrombosis in humans. Disclosures No relevant conflicts of interest to declare.

Scott Syndrome is a rare bleeding disorder characterized by a defect in platelet phosphatidylserine (PS) exposure. The syndrome has recently been linked to mutations in TMEM16F, a Ca++-activated ion channel. Tmem16f-/- mice were recently reported to be viable with a prolonged tail snip bleeding time but no spontaneous bleeding (Yang et al, Cell 151: 111-122; 2012). We now report analysis of an additional gene targeted Tmem16f allele generated in C57BL/6 ES cells.