Discordance between platelet‐supported and vesicle‐supported factor VIII activity in the presence of anti‐C2 domain inhibitory antibodies

Abstract The factor VIII C2 domain is a highly immunogenic domain, whereby inhibitory antibodies develop following factor VIII replacement therapy for congenital hemophilia A patients. Inhibitory antibodies also arise spontaneously in cases of acquired hemophilia A. The structural basis for molecular recognition by two classes of anti-C2 inhibitory antibodies that bind to factor VIII simultaneously has been investigated by small angle X-ray scattering and X-ray crystallography. The C2 domain/3E6 FAB/G99 FAB stable ternary complex, both in solution and in its crystalline state, illustrates that each antibody epitope resides on opposing faces of the factor VIII C2 domain. The 3E6 epitope is a classical antibody that forms direct contacts to the C2 domain at two loops consisting of Glu2181-Ala2188 and Thr2202-Arg2215, which inhibits the binding of the C2 domain to von Willebrand Factor and phospholipid surfaces. The G99 is a non-classical antibody that prevents proteolytic activation of factor VIII, and its epitope centers on Lys2227 and also makes direct contacts with loops Gln2222-Trp2229, Leu2261-Ser2263, His2269-Val2282 and Arg2307-Gln2311. Each binding interface is highly electrostatic, with positive charges present on both C2 epitopes and complementary negative charges on each antibody. A new model of phospholipid membrane association is also presented, where the 3E6 epitope faces the negatively charged membrane surface and Arg2320 is poised at the center of the binding interface. Furthermore, a 1.7 Å X-ray crystal structure of the porcine factor VIII C2 domain has also been determined, which supports the presented model for phospholipid binding. These results illustrate the complex nature of the polyclonal immune response against the factor VIII C2 domain, and further define the epitopes for both classical and non-classical inhibitory antibodies. Disclosures: No relevant conflicts of interest to declare.

Download Full-text

Antibody ESH4, Against the Factor VIII C2 Domain, Causes a Remote Change near the Factor IXa Active Site and Inhibition That Is Related to Membrane Composition

Blood ◽

10.1182/blood.v120.21.3355.3355 ◽

2012 ◽

Vol 120 (21) ◽

pp. 3355-3355

Author(s):

Valerie A Novakovic ◽

Junhong Lu ◽

Gary E. Gilbert

Keyword(s):

Factor Viii ◽

Active Site ◽

Hemophilia A ◽

Membrane Binding ◽

Bleeding Risk ◽

C2 Domain ◽

Membrane Composition ◽

Factor Viii Activity ◽

Factor Ixa ◽

Factor Viiia

Abstract Abstract 3355 Introduction: Development of antibodies against factor VIII is a common complication of therapy for hemophilia A and can cause acquired hemophilia A in previously normal subjects. Dominant epitopes for inhibitory antibodies reside on the C2 domain of factor VIII, which has been shown to be important for membrane binding. Although acquired hemophilia A is associated with a prolonged activated partial thromboplastin time (aPTT), the relative bleeding risk does not correlate well with factor VIII activity levels. Thus there is a need for basic insights that explain the discrepancies between factor VIII activity values and bleeding risk in these patients. mAb ESH4, directed against the factor VIII C2 domain, interferes with membrane binding and is a prototypic factor VIII inhibitor. ESH4 is a type II inhibitor, with residual factor VIII activity in the presence of saturating antibody concentrations. Thus, exploration of the inhibitory mechanism of ESH4 may offer insights into bleeding risk assessment of antibodies that inhibit phospholipid binding. Methods: Binding of fluorescein-labeled factor VIII to phospholipid membranes supported on glass microspheres was measured in the presence and absence of ESH4 using flow cytometry. The effect of ESH4 on factor VIII activity was measured using a two-stage amidolytic factor Xase assay and sonicated lipid vesicles with either low (4%) or high (15%) levels of PS. The factor Xase assay was also performed using platelets as the phospholipid source. Factor IXa, with its active site labeled using fluorescein-EGR chloromethyl ketone (Fl-EGRck), was mixed with sonicated phospholipid vesicles, factor VIIIa, and factor × and the anisotropy of the fluorescein molecule was measured to test the effect of ESH4 on the factor IXa active site. Results: Saturating concentrations of ESH4 inhibited 40% of factor VIII activity in a commercial aPTT assay. In a defined assay, inhibition of factor VIII activity was directly related to the phospholipid composition and concentration. In the presence of saturating phospholipid and factor X, ESH4 caused over 60% decrease in the Vmax for vesicles with either 4% or 15% PS. To determine the mechanism through which ESH4 inhibits membrane-bound factor VIII activity, we measured the fluorescence anisotropy of factor IXa-Fl-EGRck. ESH4 decreased anisotropy of the factor VIIIa-factor IXa-factor × complex from 0.280 ± 0.002 to 0.272 ± 0.002 on 15% PS vesicles and from 0.275 ± 0.002 to 0.262 ± 0.001 on 4% PS vesicles, indicating that ESH4 alters the Vmax through a change near the factor IXa active site, remote from the C2 domain-membrane interface. ESH4 decreased the apparent affinity 4-fold for membranes of 4% PS (KD = 4.8 ± 0.4 mM without and 21 ± 4 mM with ESH4) but only 2-fold on 15% PS vesicles (KD = 1.3 ± 0.2 mM without and 2.5 ± 0.5 mM with ESH4). Direct membrane binding studies of fluorescein-labeled factor VIII indicated a reduction in affinity and number of binding sites consistent with the results from the factor Xase assay. The apparent affinity for factor × in the presence of saturating phospholipid and ESH4 was higher on 15% PS vesicles (KM = 129 ± 18 nM) than on 4% PS vesicles (KM= 284 ± 30 nM). Together, these results indicate that ESH4 can decrease factor VIII activity through three mechanisms: (1) decreased membrane affinity (2) decreased activity of membrane-bound factor VIII and (3) differential affinity of the factor Xase complex for factor X. Because two of these mechanisms are influenced by membrane composition we asked whether the degree of inhibition by ESH4 might differ on platelets stimulated to different degrees. Platelets stimulated by thrombin express limited PS in a reversible manner while platelets stimulated by > 1 μM A23187 have complete PS exposure. ESH4 showed 80% inhibition of Xase activity on platelets stimulated with thrombin vs. 40% inhibition on platelets stimulated with A23187, similar to the aPTT assay. Conclusions: Our results indicate that ESH4 disruption of factor VIII C2 domain engagement with the membrane has a remote effect at the factor IXa active site. Inhibition of factor VIII activity by ESH4 is sensitive to membrane composition and concentration through two mechanisms. These results highlight the need to better understand how membrane binding activates the factor VIIIa-factor IXa complex and to develop clinical assays that measure factor VIII activity on clinically relevant membrane types and concentrations. Disclosures: No relevant conflicts of interest to declare.

Download Full-text

Factor VIII Ise (R2159C) in a Patient with Mild Hemophilia A, an Abnormal Factor VIII with Retention of Function but Modification of C2 Epitopes

Thrombosis and Haemostasis ◽

10.1055/s-0038-1656068 ◽

1997 ◽

Vol 77 (05) ◽

pp. 0862-0867 ◽

Cited By ~ 23

Author(s):

Hiroshi Suzuki ◽

Midori Shima ◽

Morio Arai ◽

Kazuhiko Kagawa ◽

Katuyuki Fukutake ◽

...

Keyword(s):

Monoclonal Antibody ◽

Monoclonal Antibodies ◽

Factor Viii ◽

Missense Mutation ◽

Hemophilia A ◽

Sscp Analysis ◽

C2 Domain ◽

Factor Viii Activity ◽

Normal Plasma ◽

Factor Viii Antigen

SummaryWe found a patient with mild hemophilia A who had no detectable factor VIII antigen (FVIII:Ag), as shown by two-site ELISA using inhibitor alloantibodies (TK). We then analyzed A2, A2/B, and C2 antigen of the patient's DDAVP-induced FVIII using several anti-FVIII monoclonal antibodies. Factor VIII activity (FVIII : C) was increased from 12 to 42 Uldl by the administration of DDAVP. The DDAVPinduced increases in the A2 and A2/B antigens were 40 and 36 Uldl, respectively. However, the increase in the C2 antigen was only 7.5 Uldl. SSCP analysis and subsequent sequencing demonstrated an Arg to Cys transition at codon 2159. The anti-FVII1:C titer of monoclonal antibody, NMC-VIII15 which recognized the C2 domain, against normal plasma was 450 Bethesda Ulmg of IgG. However, the titer against DDAVP-treated patient's plasma was only 15 Bethesda Ulmg. We also tested DDAVP-induced increase in the FVIII : Ag in another mild hemophilia A patient with the same mutation at Arg2159. Increase in his C2 antigen levels was only 19% of those in the A2 and A2/B antigen. We designate this abnormal FVIII as FVIII Ise. Our results show that a missense mutation at Arg2159 to Cys modifies the antigenicity of the C2 domain.

Download Full-text

Structure of the factor VIII C2 domain in a ternary complex with 2 inhibitor antibodies reveals classical and nonclassical epitopes

Blood ◽

10.1182/blood-2013-08-519124 ◽

2013 ◽

Vol 122 (26) ◽

pp. 4270-4278 ◽

Cited By ~ 17

Author(s):

Justin D. Walter ◽

Rachel A. Werther ◽

Caileen M. Brison ◽

Rebecca K. Cragerud ◽

John F. Healey ◽

...

Keyword(s):

Crystal Structure ◽

Factor Viii ◽

Structure Analysis ◽

Ternary Complex ◽

Crystal Structure Analysis ◽

C2 Domain ◽

Domain Antibody ◽

Key Points ◽

Inhibitory Antibodies

Key Points Antibodies against the factor VIII C2 domain inhibit procoagulant function. Crystal structure analysis of a C2 domain/antibody ternary complex describes epitopes for classical and nonclassical inhibitory antibodies.

Download Full-text

Antigenicity of putative phospholipid membrane-binding residues in factor VIII

Blood ◽

10.1182/blood.v97.1.169 ◽

2001 ◽

Vol 97 (1) ◽

pp. 169-174 ◽

Cited By ~ 42

Author(s):

Rachel T. Barrow ◽

John F. Healey ◽

Marc G. Jacquemin ◽

Jean-Marie R. Saint-Remy ◽

Pete Lollar

Keyword(s):

Factor Viii ◽

Membrane Binding ◽

C2 Domain ◽

Phospholipid Membrane ◽

C2 Domains ◽

Affinity Membrane ◽

Human Factor Viii ◽

Binding Residues ◽

Inhibitory Antibodies ◽

Positively Charged

Abstract Most inhibitory antibodies to human factor VIII (fVIII) bind to epitopes in the A2, ap-A3, or C2 domains. The anticoagulant action of antibodies to the C2 domain is due to inhibition of binding of fVIII to phospholipid. The x-ray structure of the human fVIII C2 domain shows a putative hydrophobic, 3-prong, phospholipid membrane-binding site consisting of Met2199/Phe2200, Val2223, and Leu2251/Leu2252. Additionally, Lys2227, near Val2223, is part of a ring of positively charged residues that may contribute to electrostatic interaction of fVIII with negatively charged phosphatidylserine. In this study, 8 active mutants of human fVIII (Met2199Ile, Leu2252Phe, Phe2200Leu, Val2223Ala, Lys2227Glu, Met2199Ile/Phe2200Leu, Val2223Ala/Lys2227Glu, and Met2199Ile/Phe2200Leu/Val2223Ala/Lys2227Glu), which were constructed on the basis of differences between human, porcine, murine, and canine fVIII at proposed phospholipid binding sites, were expressed. The antigenicity of the mutants toward 5 C2-specific polyclonal human antibodies was measured by using the Bethesda assay. A human monoclonal anti-C2 antibody, BO2C11, and a murine C2-specific monoclonal antibody, NMC VIII-5, were also included in the analysis. In comparison with wild-type, B-domainless fVIII, the Met2199Ile, Phe2200Leu, and Leu2252 single mutants had lower antigenicity toward most of the inhibitors. In contrast, the Val2223Ala and Lys2227Glu mutants usually showed increased antigenicity. These results suggest that C2 inhibitors frequently target the Met2199/Phe2200 and Leu2251/Leu2252 β-hairpins and are consistent with the hypothesis that these residues participate in binding to phospholipid membranes. In contrast, Val2223 and Lys2227 may oppose antibody binding sterically or through stabilization of a low-affinity membrane-binding conformation of the C2 domain.

Download Full-text

Epitope mapping of polyclonal clotting factor VIII-inhibitory antibodies using phage display

Thrombosis and Haemostasis ◽

10.1160/th03-07-0473 ◽

2004 ◽

Vol 91 (03) ◽

pp. 619-625 ◽

Cited By ~ 18

Author(s):

Christiane Mühle ◽

Stefan Schulz-Drost ◽

Alexey Khrenov ◽

Evgueni Saenko ◽

Jens Klinge ◽

...

Keyword(s):

Phage Display ◽

Factor Viii ◽

Inhibitory Activity ◽

Phage Display Library ◽

Clotting Factor ◽

C2 Domain ◽

Clotting Factors ◽

Antibody Preparation ◽

Random Peptide ◽

Inhibitory Antibodies

SummaryClotting factor VIII (fVIII)-inhibitory antibodies represent a major problem in the treatment of haemophilia A. To understand the inactivation mechanisms and to pave the way towards modifications of recombinant clotting factors that reduce their immunogenicity, the exact localization of immunodominant epitopes is required. Here, a random peptide phage display library was employed to identify epitopes of polyclonal fVIII antibodies isolated from patient’s plasma by affinity chromatography. FVIIIbinding specificity and inhibitory activity of the isolated fVIII antibodies were confirmed by ELISA and Bethesda assays. Phage selection on the individual samples yielded several phages which were displaced from binding to the respective antibody preparation by fVIII. Their homology with amino acid motifs of human fVIII and immunoprecipitation results with radioactively labelled fVIII fragments suggested putative epitopes in the A1, A2 and C1 domains of fVIII for one and in the C2 domain for another patient. Synthetic peptides corresponding to the A2, C1 and C2 domain epitopes blocked antibody binding to fVIII and partially neutralized the inhibitory activity of the respective plasma in Bethesda assays. These results provide the proof of principle that random peptide libraries can be used for the mapping of epitopes in a polyclonal antibody preparation.

Download Full-text

Measurement of factor VIII activity of B-domain deleted recombinant factor VIII

Seminars in Hematology ◽

10.1053/shem.2001.25889 ◽

2001 ◽

Vol 38 (2, Suppl 4) ◽

pp. 13-23 ◽

Cited By ~ 16

Author(s):

M. Mikaelsson ◽

U. Oswaldsson ◽

M. A. Jankowski

Keyword(s):

Factor Viii ◽

Recombinant Factor Viii ◽

Factor Viii Activity

Download Full-text

Thrombelastography during and after Elective Abdominal Surgery

Thrombosis and Haemostasis ◽

10.1055/s-0038-1646708 ◽

1978 ◽

Vol 39 (02) ◽

pp. 488-495 ◽

Cited By ~ 13

Author(s):

J M Butler

Keyword(s):

Abdominal Surgery ◽

Factor Viii ◽

Whole Blood ◽

Dominant Factor ◽

Fibrinogen Concentration ◽

Factor Viii Activity ◽

Elective Abdominal Surgery

SummaryThrombelastography has been performed on recalcified whole blood from 50 patients before, during and after elective abdominal surgery. The characteristic changes of the thrombelastographic indices r, k and mA are described.During operation r and k shortened, but no change in mA was observed. This response was in part associated with an increase in factor VIII activity. Following operation, while r time was somewhat shortened, much more marked changes in k and mA were evident. Increasing fibrinogen concentration was the dominant factor in determining the post-operative changes in the thrombelastograph.

Download Full-text

Factor VIII Inhibitor Antibodies with C2 Domain Specificity Are Less inhibitory to Factor VIII Complexed with von Willebrand Factor

Thrombosis and Haemostasis ◽

10.1055/s-0038-1650655 ◽

1996 ◽

Vol 76 (05) ◽

pp. 749-754 ◽

Cited By ~ 73

Author(s):

Suzuki Suzuki ◽

Morio Arai ◽

Kagehiro Amano ◽

Kazuhiko Kagawa ◽

Katsuyuki Fukutake

Keyword(s):

Complex Formation ◽

Factor Viii ◽

Von Willebrand Factor ◽

Domain Specificity ◽

Factor Viii Inhibitor ◽

C2 Domain ◽

Recombinant Fviii ◽

Von Willebrand ◽

A2 Domain ◽

Willebrand Factor

SummaryIn order to clarify the potential role of von Willebrand factor (vWf) in attenuating the inactivation of factor VIII (fVIII) by those antibodies with C2 domain specificity, we investigated a panel of 14 human antibodies to fVIII. Immunoblotting analysis localized light chain (C2 domain) epitopes for four cases, heavy chain (A2 domain) epitopes in five cases, while the remaining five cases were both light and heavy chains. The inhibitor titer was considerably higher for Kogenate, a recombinant fVIII concentrate, than for Haemate P, a fVIII/vWf complex concentrate, in all inhibitor plasmas that had C2 domain specificity. In five inhibitor plasmas with A2 domain specificity and in five with both A2 and C2 domain specificities, Kogenate gave titers similar to or lower than those with Haemate P. The inhibitory effect of IgG of each inhibitor plasma was then compared with recombinant fVIII and its complex with vWf. When compared to the other 10 inhibitor IgGs, IgG concentration, which inhibited 50% of fVIII activity (IC50), was remarkably higher for the fVIII/vWf complex than for fVIII in all the inhibitor IgGs that had C2 domain reactivity. Competition of inhibitor IgG and vWf for fVIII binding was observed in an ELISA system. In 10 inhibitors that had C2 domain reactivity, the dose dependent inhibition of fVIII-vWf complex formation was observed, while, in the group of inhibitors with A2 domain specificity, there was no inhibition of the complex formation except one case. We conclude that a subset of fVIII inhibitors, those that bind to C2 domain determinants, are less inhibitory to fVIII when it is complexed with vWf that binds to overlapping region in the C2 domain.

Download Full-text

Heparin, Protamine and Polybrene

Thrombosis and Haemostasis ◽

10.1055/s-0038-1656300 ◽

1965 ◽

Vol 13 (02) ◽

pp. 550-560 ◽

Cited By ~ 2

Author(s):

Anthony Britten

Keyword(s):

Factor Viii ◽

Incubation Mixture ◽

Factor V ◽

Factor Viii Activity ◽

Rapid Loss

SummaryThe effects of incubating heparin, protamine or Polybrene with plasma were studied. All three drugs cause rapid loss of factor V from decalcified plasma, while Polybrene also accelerates the loss of factor VIII activity. These changes are related to temperature, the period of incubation and the dose of the drug used, and can be partially prevented by inclusion of neutralizing doses of the appropriate antagonist in the incubation mixture.The implications of these findings are discussed.

Download Full-text