The EAHAD blood coagulation factor VII variant database

Hereditary blood coagulation factor VII (FVII) deficiency is a rare autosomal recessive bleeding disorder resulting from variants in the gene encoding FVII (F7). Integration of genetic variation with functional consequences on protein function is essential for the interpretation of the pathogenicity of novel variants. Here, we describe the integration of previous locus-specific databases for F7 into a single curated database with enhanced features. The database provides access to in silico analyses that may be useful in the prediction of variant pathogenicity as well as cross-species sequence alignments, structural information, and functional and clinical severity described for each variant, where appropriate. The variant data is shared with the F7 Leiden Open Variation Database. The updated database now includes 221 unique variants, representing gene variants identified in 728 individuals. Single nucleotide variants are the most common type (88%) with missense representing 74% of these variants. A number of variants are found with relatively high minor allele frequencies that are not pathogenic but contribute significantly to the likely pathogenicity of coinherited variants due to their effect on FVII plasma levels. This comprehensive collection of curated information significantly aids the assessment of pathogenicity.

The European Association for Haemophilia and Allied Disorders (EAHAD) Coagulation Factor Variant Databases: Important resources for haemostasis clinicians and researchers

INTRODUCTION

Advances in genomic sequencing have facilitated the sequencing of genes associated with disorders of haemostasis. The identification of variants within genes and access to curated data incorporating structural, functional, evolutionary as well as phenotypic data has become increasingly important in order to ascribe pathogenicity.

AIM

The European Association for Haemophilia and Allied Disorders (EAHAD) Coagulation Factor Variant Database Project aims to provide a single port of entry to a web-accessible resource for variants in genes involved in clinical bleeding disorders.

RESULTS

New databases have evolved from previously developed single gene variant coagulation database projects, incorporating new data, new analysis tools and a new common database architecture with new interfaces and filters. These new databases currently present information about the genotype, phenotype (laboratory and clinical) and structural and functional effects of variants described in the genes of factor (F) VII (F7), FVIII (F8), FIX (F9) and von Willebrand factor (VWF).

CONCLUSION

The project has improved the quality and quantity of information available to the haemostasis research and clinical communities, thereby enabling accurate classification of disease severity in order to make assessments of likely pathogenicity.

Measurement of Activated Factor IX in Factor IX Concentrates: Correlation with In Vivo Thrombogenicity

Current in vitro tests for thrombogenicity of FIX concentrates used for prothrombin complex concentrates (PCCs), are of little value when applied to high purity FIX (HP FIXs). In the present study, we have developed a chromogenic assay for activated FIX (FIXa) and evaluated its ability to predict in vivo thrombogenic potential of HP FIXs in a modified Wessler stasis model. Among the HP FIXs, only 1 out of 7 products had no detectable FIXa; this product also showed no in vivo thrombogenicity. In the other 6 products, FIXa content ranged from 0.15–1.2 U/1000 iu FIX, and all showed some evidence of in vivo thrombogenicity, with mean thrombus scores ranging from 0.25–4. There was a significant positive correlation (r = 0.55, p <0.02) between FIXa levels and in vivo thrombogenicity of HP FIXs. NAPTT data were not significantly correlated with the in vivo results and the TFCT also showed no direct correlation with the mean thrombus score. These results indicate that HP FIXs may still carry a small residual thrombotic risk and measurement of FIXa content of these products may be a better predictor of thrombogenicity than the current in vitro tests.

Human Tissue Factor Pathway Inhibitor Fused to CD4 Binds both FXa and TF/FVIIa at the Cell Surface

Tissue factor pathway inhibitor (TFPI) is one of the main regulators of the tissue factor (TF) pathway of coagulation. To tether human TFPI to the cell surface, full length or truncated TFPI lacking the third Kunitz domain were fused with domains three and four and the carboxy-termi- nal sequence of human CD4. Constructs were transfected into a mouse fibroblast cell line and individual clones were checked for expression using monoclonal antibodies directed against the first two TFPI Kunitz domains and against CD4. Specific human FXa binding was detected by flow cytometry using an anti-FX polyclonal antibody, and inhibition of FXa proteolytic activity was verified by chromogenic substrate assay using S-2765. In addition, TFPI-CD4-expressing cells, preincubated with FXa, specifically bound human TF-FVIIa complexes as revealed with an anti-human TF polyclonal antibody. No functional difference was observed between full length or truncated TFPI-CD4. These results demonstrate that functionally intact TFPI can be tethered to the cell surface. Genetic manipulation of, for example, endothelial cells leading to the stable expression of TFPI may inhibit the development of coronary artery heart disease following cardiac allotransplantation, and may inhibit thrombosis in the context of xenotransplantation.

In Vivo Studies of Activated Porcine Factor VIII

The haemostatic effectiveness of activated FVIII was compared to that of non-activated FVIII in a cross-over study in a canine model of haemophilia. Activation of FVIII in porcine concentrate was achieved by the addition of 3 × 10−5 IU thrombin per ml of concentrate, which gave consistent increases in 1-stage FVIII activity of 13- to 14-fold and slow decay. The haemostatic effect was monitored by measurements of the cuticle bleeding time 10 and 45 min after infusion and there were no consistent differences between the activated and non-activated concentrates. One-stage factor VIII assays on plasmas 5 min after infusion showed identical mean values for activated and non-activated concentrates, indicating that most of the higher activity observed in vitro had disappeared rapidly from the circulation.

These results suggest that controlled activation of FVIII by thrombin, which increases its activity in 1-stage assays, is unlikely to be of therapeutic benefit. For therapeutic concentrates which may contain small amounts of activated FVIII, the 1-stage assay may be an unreliable guide to their therapeutic effect.

High Prevalence of Elevated Factor VIII Levels in Patients Referred for Thrombophilia Screening: Role of Increased Synthesis and Relationship to the Acute Phase Reaction

A recent report from the Leiden Thrombophilia Survey identified high factor VIII activity levels as an independent risk factor for venous thromboembolism in a population survey. As the study measure for factor VIII was a one-stage coagulation assay, and since markers for the acute phase reaction were not assessed, it remained uncertain whether the increase was due to a constitutional increased rate of synthesis, to circulating activated factor VIII, or to an acute phase response.

We added factor VIII activity assay (FVIII:C), factor VIII antigen (FVIILAg), vWF antigen (vWF:Ag), ABO blood group, fibrinogen and C-reactive protein to our routine thrombophilia screen of patients referred because of unexplained thromboembolism.

Elevated FVIILC (>1.5 iu/ml) emerged as the single commonest abnormality detected in 25.4% of a group of 260 such patients.

FVIILC and FVIILAg were highly correlated (p = 0.003), showing that this represented a true increase in FVIII. In 4 of 46 patients this was clearly attributable to an acute phase reaction. Eleven others showed minor elevation of ESR and one of CRP. Neither FVIILC or FVIII:Ag showed significant correlation with fibrinogen, ESR or C-reactive protein by non parametric analysis.

Although there was an excess of patients with B blood group (known to be associated with FVIILC levels which are -15% higher than those in blood group 0), this could not account for the marked elevation of factor VIII observed in these patients.

We conclude that factor VIII activity assay should be a routine part of thrombophilia screening. We are investigating the cause of the increased synthesis, initially by means of family studies and linkage analysis with polymorphic markers of the FVIII locus. We postulate that it may be constitutive in some cases and in others an abnormal or exaggerated response to inflammatory stimuli.

Pharmacokinetic In Vivo Comparison Using 1-Stage and Chromogenic Substrate Assays with Two Formulations of Hemofil-M

In a study to demonstrate the safety and pharmacokinetics (half-life and recovery) of two different method M purified AHF (Hemofil-M) concentrates processed in the USA and Spain, two different methods of factor VIII assay (one-stage clotting and chromogenic) have been compared in vivo. The study was a single centre blinded, randomised, crossover study. Twelve patients with severe haemophilia A (VIII: C <2 u/dl) were divided into two subgroups of six. None had received factor VIII concentrate within 48 h preceding the study. Twenty-four pharmacokinetic studies were performed in the 12 patients. Each subgroup received two different lots of study material (US and Spanish) at a dose of 50 u/kg seven days apart. A second randomisation was nominal potency, high: 1000 u or mid: 500 u per vial. The potency label was a one-stage clotting assay using the mega I standard. A standard pharmacokinetic study was performed over 24 h and each blinded sample was analysed in duplicate by a one-stage clotting (aPTT) and a chromogenic (Chromogenix AB; CS) assay at the Royal Free and NIBSC. Pharmacokinetic modelling was performed. The mean label for Hemofil-M using the chromogenic substrate assay was 79% that using the one stage assay (Mega I standard). The recovery was 17-28% higher measured by chromogenic compared to the clotting assay. Since most clinicians use the clotting assay, potency labelling using the chromogenic assay, will overestimate predicted Hemofil-M recovery by as much as 25%.

Factor VIII Procoagulant Protein Interacts with Phospholipid Vesicles Via its 80 kDa Light Chain

In a previous report, we detailed the fractionation of polyclonal human anti-Factor VIII :C into a component directed exclusively against the phospholipid-binding site on Factor VIII (PL-site antibody) and another directed at other sites (non-PL-site antibody). The location on the F.VIII molecule of its PL-binding site has now been studied by two different methods using this fractionated 125I-labelled anti-F.VIII: C Fab’.

The first method was modified from that of Weinstein et al. (Proc Natl Acad Sci USA 1981; 78: 5137-41), involving electrophoresis of F.VIII peptide-125I-Fab‘ A/F.VIII immunocomplexes in SDS-polyacrylamide gels. PL-site antibody reacted with F.VIII peptides of apparent Mr approximately 80 kDa and sometimes 160 kDa in plasma and concentrate, but not with larger peptides. Non-PL-site antibody, however, reacted with a range of peptides of apparent Mr 90 kDa to 280 kDa. In addition, when purified F.VIII containing heavy and light chains (HC + LC), and isolated LC peptides were analysed, PL-site antibody bound to LC peptides whereas non-PL-site antibody did not.

The second method used the antibody pools in immunoradiometric assays (IRMA’s) of purified F.VIII peptides. Both labels measured similar amounts of F.VIII: Ag in a sample of purified F.VIII containing both HC and LC; on assaying an HC preparation, however, PL-site label measured only 2% of F.VIII: Ag found by non-PL-site label, indicating that PL-binding sites are absent in HC preparations.

These results indicate that F.VIII binds to PL via its 80 kDa light chain.

Characterisation of blood coagulation factor XIT475I

PCR-SSCP and DNA sequence analysis of a factor XI (FXI) deficient patient (FXI:C 39 U/dL; FXI:Ag 27 U/dL) identified a C to T transition in exon 12 of the FXI gene (F11 c.1521C>T) that predicts the substitution of Thr475 by Ile (FXIT475I) within the serine protease domain of FXI. This mutation destroys a consensus sequence for N-linked glycosylation, N473-Y-T475, known to be utilized in vivo. The FXIT475I variant was generated by site-directed mutagenesis, together with other variants that could help explain the phenotype, and recombinant FXI variants were expressed in Chinese hamster ovary cells. FXI:Ag expression was analysed by Western blot analysis, ELISA and immunocyto-chemical staining. Wild-type FXI:Ag was secreted at high levels, however the mutant (FXIT475I) was secreted very poorly. Substitution of Thr475 by Ala, Pro, Lys or Arg (all of which abolish the glycosylation consensus sequence) also severely reduced the level of secreted FXI:Ag suggesting that glycosylation at Asn473 is required for folding or secretion. Concordant with this hypothesis the conservative substitution of Thr475 by Ser (which preserves the glycosylation consensus sequence) had no effect on FXI secretion. Thr/Ser475 is highly conserved in serine protease domains but the glycosylation site (Asn473) is not. Surprisingly, substitution of Asn473 by Ala (which removes the N-linked glycosylation site) had no effect on the levels of FXI:Ag secreted. In conclusion, although the FXI-T475I mutation destroys an N-linked glycosylation consensus sequence, the cause of failure to secrete FXI is not the loss of a glycosylation site but rather a direct effect of the substitution of this highly conserved residue.

Hemophilia A Mutations within the Factor VIII A2-A3 Subunit Interface Destabilize Factor VIIIa and Cause One-stage/ Two-stage Activity Discrepancy

Thrombin-activated factor VIII (FVIIIa) is a heterotrimer with the A2 subunit in a weak ionic interaction with the A1 and A3-C1-C2 subunits. Dissociation of the A2 subunit correlates with inactivation of FVIIIa. A homology model (Blood 89:2413, 1997) of the triplicated A domains of factor VIII (FVIII) predicts a pseudo-threefold axis at the tightly packed hydrophobic core with several interdomain interactions. These lie at the interface of A1-A2, A2-A3 and A1-A3. We have previously demonstrated that hemophilia A mutations (R531H, A284E, S289L) within the predicted A1-A2 and A1-A3 interface disrupt potential intersubunit hydrogen bonds and have the molecular phenotype of increased rate of inactivation of FVIIIa due to increased rate of A2 subunit dissociation. Patients with these mutations exhibit a clinical phenotype where the FVIII activity by one-stage(1-st) assay is at least two-fold higher than by two-stage(2-st) assay. We have now also explored mutations within the predicted A2-A3 interface (N694I, R698W and R698L) that also have the phenotype of 1-st/2-st activity discrepancy. These mutations exhibit the same molecular mechanism of increased instability of FVIIIa as those mutations described along the A1-A2 and A1-A3 interfaces. This suggests that the entire tightly packed hydrophobic core within the predicted pseudo-threefold axis contributes to stabilization of FVIIIa.

An interactive mutation database for human coagulation factor IX provides novel insights into the phenotypes and genetics of hemophilia B

BACKGROUND

Factor IX (FIX) is important in the coagulation cascade, being activated to FIXa on cleavage. Defects in the human F9 gene frequently lead to hemophilia B.

OBJECTIVE

To assess 1113 unique F9 mutations corresponding to 3721 patient entries in a new and up-to-date interactive web database alongside the FIXa protein structure.

METHODS

The mutations database was built using MySQL and structural analyses were based on a homology model for the human FIXa structure based on closely-related crystal structures.

RESULTS

Mutations have been found in 336 (73%) out of 461 residues in FIX. There were 812 unique point mutations, 182 deletions, 54 polymorphisms, 39 insertions and 26 others that together comprise a total of 1113 unique variants. The 64 unique mild severity mutations in the mature protein with known circulating protein phenotypes include 15 (23%) quantitative type I mutations and 41 (64%) predominantly qualitative type II mutations. Inhibitors were described in 59 reports (1.6%) corresponding to 25 unique mutations.

CONCLUSION

The interactive database provides insights into mechanisms of hemophilia B. Type II mutations are deduced to disrupt predominantly those structural regions involved with functional interactions. The interactive features of the database will assist in making judgments about patient management.

In silico prediction of FVIII epitopes recognised by natural autoantibodies in polyvalent immunoglobulin concentrates

Inhibitory antibodies directed against blood coagulation factor VIII (FVIII) impair FVIII replacement therapy, constituting a serious complication in haemophilic and autoimmune patients. Identifying B-cell FVIII epitopes and mapping them on the molecule remain important challenges. Using a combination of different algorithms, more than 30 hypothetical linear epitopes were predicted on the FVIII molecule surface. We selected several major predicted sequences, spanning all FVIII domains, for specific antibody induction in rabbits. All peptides tested successfully induced production of specific anti-FVIII rabbit antibodies, supporting the relevance of our approach. To investigate the presence of FVIII-reactive antibodies in the healthy donor population, a pooled fraction rich in all IgG subclasses was purified on peptide-Sepharose columns. Substantial amounts of Ig, specific for each FVIII peptide, were purified with yields ranging from 8 to 223 ng/mg immunoglobulins. Our results confirm the diversity of FVIII epitopes recognised by natural human anti-FVIII autoantibodies. All IgG subclasses were found in the affinity-isolated anti-peptide material, with overrepresentation of IgG2 and IgG4. Evidence was also found for new FVIII epitopes. Five human anti-peptide preparations displayed FVIII-neutralising activity, ranging from 1.3 to 5.3 BU/mg. Although the presence of naturally occurring anti-FVIII antibodies in healthy donors has been previously described, our methodology has allowed, for the first time, a fine mapping of several inhibitory and non-inhibitory epitopes. Our observations support the hypothesis that FVIII inhibitors in haemophilia A and autoimmune disease may originate from the proliferation of natural FVIII-specific B-cell clones.

Postinjury vascular intimal hyperplasia in mice is completely inhibited by CD34+ bone marrow-derived progenitor cells expressing membrane-tethered anticoagulant fusion proteins

BACKGROUND

Coagulation proteins promote neointimal hyperplasia and vascular remodelling after vessel injury, but the precise mechanisms by which they act in vivo remain undetermined.

OBJECTIVES

This study, using an injury model in which the neointima is derived from bone marrow (BM)-derived cells, compared inhibition of tissue factor or thrombin on either BM-derived or existing vascular smooth muscle cells.

METHODS

Two transgenic (Tg) mouse strains expressing membrane-tethered tissue factor pathway inhibitor (TFPI) or hirudin (Hir) fusion proteins driven by an alpha smooth muscle actin (SMA) promoter were generated (alpha-TFPI-Tg and alpha-Hir-Tg) and the phenotype after wire-induced endovascular injury was compared with that in wild-type (WT) controls.

RESULTS

WT mice developed progressive neointimal expansion, whereas injury in either Tg was followed by repair back to a preinjured state. This was also seen when WT mice were reconstituted with BM from Tg mice but not when Tgs were reconstituted with WT BM, in which injury was followed by slowly progressive neointimal expansion. Injection of CD34+ cells from Tg mice into injured WT mice resulted in the accumulation of fusion protein-expressing cells from day 3 onwards and an absence of neointimal hyperplasia in those areas.

CONCLUSIONS

Neointimal development after wire-induced endovascular injury in mice was completely inhibited when BM-derived cells infiltrating the damaged artery expressed membrane tethered anticoagulant fusion proteins under an alpha-SMA promoter. These findings enhance our understanding of the pathological role that coagulation proteins play in vascular inflammation.

Multiple vitamin K-dependent coagulation zymogens promote adenovirus-mediated gene delivery to hepatocytes

Upon local delivery, adenovirus (Ad) serotype 5 viruses use the coxsackie and Ad receptor (CAR) for cell binding and alpha(v) integrins for internalization. When administered systemically, however, their role in liver tropism is limited because CAR-permissive and mutated viruses show similar biodistribution, a finding recently attributed to blood coagulation factor (F) IX or complement protein C4BP binding to the adenovirus fiber and "bridging" to either low-density lipoprotein receptor-related protein or heparan sulfate proteoglycans. Here, we show that hepatocyte transduction in vitro can be enhanced by the vitamin K-dependent factors FX, protein C, and FVII in addition to FIX but not by prothrombin (FII), FXI, and FXII. This phenomenon was not dependent on proteolytic activation or cell signaling activity and for FX was mediated by direct virus-factor binding. Human FX substantially enhanced hepatocyte transduction by CAR-permissive and mutated viruses in an ex vivo liver perfusion model. In vivo, global down-regulation of vitamin K-dependent zymogens by warfarin significantly diminished liver uptake of CAR-deleted Ads; however, this phenomenon was fully rescued by acute infusion of human FX. Our results indicate a common and pivotal role for distinct vitamin K-dependent coagulation factors in mediating hepatocyte transduction by adenoviruses in vitro and in vivo.

The spectrum of mutations and molecular pathogenesis of hemophilia A in 181 Portuguese patients

Disease-causing alterations within the F8 gene were identified in 177 hemophilia A families of Portuguese origin. The spectrum of non-inversion F8 mutations in 101 families included 67 different alterations, namely: 36 missense, 8 nonsense and 4 splice site mutations, as well as 19 insertions/deletions. Thirty-four of these mutations are novel. Molecular modeling allowed prediction of the conformational changes introduced by selected amino acid substitutions and their correlation with the patients' phenotypes. The relatively frequent, population-specific, missense mutations together with de novo alterations can lead to significant differences in the spectrum of F8 mutations among different populations.

Self-complementary adeno-associated virus vectors containing a novel liver-specific human factor IX expression cassette enable highly efficient transduction of murine and nonhuman primate liver

Transduction with recombinant adeno-associated virus (AAV) vectors is limited by the need to convert its single-stranded (ss) genome to transcriptionally active double-stranded (ds) forms. For AAV-mediated hemophilia B (HB) gene therapy, we have overcome this obstacle by constructing a liver-restricted mini-human factor IX (hFIX) expression cassette that can be packaged as complementary dimers within individual AAV particles. Molecular analysis of murine liver transduced with these self-complementary (sc) vectors demonstrated rapid formation of active ds-linear genomes that persisted stably as concatamers or monomeric circles. This unique property resulted in a 20-fold improvement in hFIX expression in mice over comparable ssAAV vectors. Administration of only 1 x 10(10) scAAV particles led to expression of hFIX at supraphysiologic levels (8I U/mL) and correction of the bleeding diathesis in FIX knock-out mice. Of importance, therapeutic levels of hFIX (3%-30% of normal) were achieved in nonhuman primates using a significantly lower dose of scAAV than required with ssAAV. Furthermore, AAV5-pseudotyped scAAV vectors mediated successful transduction in macaques with pre-existing immunity to AAV8. Hence, this novel vector represents an important advance for hemophilia B gene therapy.

Permanent partial phenotypic correction and tolerance in a mouse model of hemophilia B by stem cell gene delivery of human factor IX

Immune responses against an introduced transgenic protein are a potential risk in many gene replacement strategies to treat genetic disease. We have developed a gene delivery approach for hemophilia B based on lentiviral expression of human factor IX in purified hematopoietic stem cells. In both normal C57Bl/6J and hemophilic 129/Sv recipient mice, we observed the production of therapeutic levels of human factor IX, persisting for at least a year with tolerance to human factor IX antigen. Secondary and tertiary recipients also demonstrate long-term production of therapeutic levels of human factor IX and tolerance, even at very low levels of donor chimerism. Furthermore, in hemophilic mice, partial functional correction of treated mice and phenotypic rescue is achieved. These data show the potential of a stem cell approach to gene delivery to tolerize recipients to a secreted foreign transgenic protein and, with appropriate modification, may be of use in developing treatments for other genetic disorders.

Generation of a polyclonal rabbit anti-mouse tissue factor antibody by nucleic acid immunisation

Tissue factor (TF) the cellular receptor and cofactor for factor VII, initiates coagulation and has also been implicated in several coagulation-independent functions, including inflammation, angiogenesis and tumour metastasis. Investigations of TF expression in mouse models of these processes has been limited by the availability of antibodies that specifically recognise mouse TF. We have generated a rabbit polyclonal antibody to mTF by DNA immunisation. This has yielded an antiserum that recognises native mTF in immunohistochemical and flow cytometric analyses. Furthermore, the antiserum is inhibitory in coagulation assays. This antiserum will be a valuable investigative tool in the analysis of mTF expression.

Complete inhibition of acute humoral rejection using regulated expression of membrane-tethered anticoagulants on xenograft endothelium

Xenotransplantation promises an unlimited supply of organs for clinical transplantation. However, an aggressive humoral immune response continues to limit the survival of pig organs after transplantation into primates. Because intravascular thrombosis and systemic coagulopathy are prominent features of acute humoral xenograft rejection, we hypothesized that expression of anticoagulants on xenogeneic vascular endothelium might inhibit the process. Hearts from novel transgenic mice, expressing membrane-tethered fusion proteins based on human tissue factor pathway inhibitor and hirudin, respectively, were transplanted into rats. In contrast to control non-transgenic mouse hearts, which were all rejected within 3 days, 100% of the organs from both strains of transgenic mice were completely resistant to humoral rejection and survived for more than 100 days when T-cell-mediated rejection was inhibited by administration of ciclosporin A. These results demonstrate the critical role of coagulation in the pathophysiology of acute humoral rejection and the potential for inhibiting rejection by targeting the expression of anticoagulants to graft endothelial cells. This genetic strategy could be applied in a clinically relevant species such as the pig.

Permanent phenotypic correction of hemophilia B in immunocompetent mice by prenatal gene therapy

Hemophilia B, also known as Christmas disease, arises from mutations in the factor IX (F9) gene. Its treatment in humans, by recombinant protein substitution, is expensive, thus limiting its application to intermittent treatment in bleeding episodes and prophylaxis during surgery; development of inhibitory antibodies is an associated hazard. This study demonstrates permanent therapeutic correction of his disease without development of immune reactions by introduction of an HIV-based lentiviral vector encoding the human factor IX protein into the fetal circulation of immunocompetent hemophiliac and normal outbred mice. Plasma factor IX antigen remained at around 9%, 13%, and 16% of normal in the 3 hemophilia B mice, respectively, until the last measurement at 14 months. Substantial improvement in blood coagulability as measured by coagulation assay was seen in all 3 mice and they rapidly stopped bleeding after venipuncture. No humoral or cellular immunity against the protein, elevation of serum liver enzymes, or vector spread to the germline or maternal circulation were detected.

Two novel mutations in severe factor VII deficiency

We have characterized the molecular defect in two families with severe factor VII (FVII) deficiency. In family I, the proband was found to be homozygous for a novel 18 bp deletion in exon 8 (g.10896-10913del) resulting in the in-frame deletion of six amino acids in the serine protease domain. Molecular modelling suggests the deletion is likely to disrupt folding of the FVII molecule. The reduced FVII antigen (21 U/dl) and negligible activity (0.4 U/dl) in the patient's plasma indicated that the deletion affected both the secretion/stability and function of the mutant protein. In family II, the proband was found to be a compound heterozygote for a novel missense mutation (g.7884G>A; FVII G117R) in exon 5 encoding the EGF2 domain of FVII and a nonsense mutation (g.8960C>T; FVII R152X) in exon 6. Extensive sequence comparison in a wide evolutionary context suggested that the Gly117 residue is critical for structure of FVII. The grossly reduced FVII antigen (1.1 U/dl) and activity (0.4 U/dl) plasma values indicate the mutation primarily affected the folding/secretion or stability of the protein.

Hemophilia A mutations within the factor VIII A2-A3 subunit interface destabilize factor VIIIa and cause one-stage/two-stage activity discrepancy

Thrombin-activated factor VIII (FVIIIa) is a heterotrimer with the A2 subunit in a weak ionic interaction with the A1 and A3-C1-C2 subunits. Dissociation of the A2 subunit correlates with inactivation of FVIIIa. A homology model (Blood 89:2413, 1997) of the triplicated A domains of factor VIII (FVIII) predicts a pseudo-threefold axis at the tightly packed hydrophobic core with several interdomain interactions. These lie at the interface of A1-A2, A2-A3 and A1-A3. We have previously demonstrated that hemophilia A mutations (R531H, A284E, S289L) within the predicted A1-A2 and A1-A3 interface disrupt potential intersubunit hydrogen bonds and have the molecular phenotype of increased rate of inactivation of FVIIIa due to increased rate of A2 subunit dissociation. Patients with these mutations exhibit a clinical phenotype where the FVIII activity by one-stage(1-st) assay is at least two-fold higher than by two-stage(2-st) assay. We have now also explored mutations within the predicted A2-A3 interface (N694I, R698W and R698L) that also have the phenotype of 1-st/2-st activity discrepancy. These mutations exhibit the same molecular mechanism of increased instability of FVIIIa as those mutations described along the A1-A2 and A1-A3 interfaces. This suggests that the entire tightly packed hydrophobic core within the predicted pseudo-threefold axis contributes to stabilization of FVIIIa.

Use of a non-depleting anti-CD4 antibody to modulate the immune response to coagulation factors VIII and IX

The generation of antibodies to therapeutic factors VIII or IX is a major problem in the management of haemophilia and places potential limitations on the application of gene therapy. We have investigated the administration of a non-depleting anti-CD4 antibody for modulation of the immune response to human recombinant coagulation factors VIII and IX. In mice given these clotting factors, co-administration of anti-CD4 antibody significantly reduced the appearance of factor-specific antibodies. These data provide evidence that the neutralizing antibody response to exogenous coagulation factors may be controllable if non-depleting anti-CD4 antibody is co-administered at the time of initial replacement therapy.

A Tyr346-->Cys substitution in the interdomain acidic region a1 of factor VIII in an individual with factor VIII:C assay discrepancy

The interdomain acidic region a1 is a unique structural feature of coagulation factor VIII (FVIII) and may mediate the proteolytic activation of FVIII and the inactivation of FVIIIa. We report an individual with a Tyr346-->Cys substitution within region a1, who presented with a one-stage FVIII activity (FVIII:C) of 0.34 iu/ml (normal range 0.5-2.0) but normal two-stage FVIII:C and FVIII antigen values. In a factor Xa (FXa)-generation assay for FVIII in which the activation time with thrombin was varied, the variant plasma showed normal FVIII:C at both short and long activation times. However, at intermediate activation times the FXa generation of the variant plasma was less than that of normal pooled plasma. In a modified one-stage FVIII:C assay in which partially purified FVIII was activated with thrombin at low concentrations, the variant FVIII showed less activation than wild-type FVIII, although this defect corrected with increasing concentrations of thrombin. When partially purified variant FVIII was activated with a large molar excess of thrombin, the subsequent rate of decay of FVIII:C was greater for variant FVIII. The complex defects in activation and inactivation displayed by FVIII Tyr346-->Cys support the hypothesis that the a1 sequence is a key regulator of FVIII activity.

3-Dimensional structure of membrane-bound coagulation factor VIII: modeling of the factor VIII heterodimer within a 3-dimensional density map derived by electron crystallography

Despite recent studies, the organization of coagulation factor VIII (FVIII) on a phospholipid (PL) membrane is not known in detail. Thus, 2-dimensional (2D) crystals of human FVIII lacking the B domain were prepared for electron microscopy onto negatively charged PL monolayers. The 3-dimensional (3D) density map of the PL-bound FVIII protein was calculated at 1.5 nm. Existing atomic data and models for FVIII domains were fitted unambiguously within the 3D density map of the molecule. FVIII domains arrangement followed a compact spiral organization with the A3 domains in close association with the C1 and C2 domains near the PL surface. Viewed toward the membrane the A domains' heterotrimer is oriented side-on with the pseudo-3-fold axis almost parallel to the PL surface and A1 fully covering C1. The C2 domain is partially overlapped by the A2 domain of an adjacent molecule in the 2D crystal, favoring close packing. Viewed parallel to the membrane, C2 is slightly inclined to the PL surface covering an area of 12 nm(2). Four C2 loops are embedded within the lipid monolayer at about 0.7 to 1.0 nm depth. C1 forms almost a right angle with C2, its long axis nearly parallel to the membrane. The proposed structure for membrane-bound FVIII results from modeling of the FVIII domains within a 3D density map obtained from electron crystallography and accords with the main biochemical and structural information known to date. A model is proposed for FVIIIa and factor IXa assembly within the membrane-bound factor X-activating complex. (Blood. 2002;99:1215-1223)

Human thrombin and FXa mediate porcine endothelial cell activation; modulation by expression of TFPI-CD4 and hirudin-CD4 fusion proteins

Aside from their critical role in thrombosis, activated coagulation factors also have inflammatory properties and these may be important during delayed xenograft rejection (DXR). This study assessed whether porcine EC could be activated by factor Xa (FXa) and thrombin (FIIa) and whether expression of tissue factor pathway inhibitor (TFPI)-CD4 and hirudin-CD4 fusion proteins could prevent such activation. Incubation of porcine EC with human FXa and FIIa induced cell surface expression of E-selectin, VCAM and tissue factor (TF) in a time-dependent and concentration-dependent manner. In contrast, porcine EC transfected with a human TFPI-CD4 fusion protein were selectively resistant to these pro-inflammatory effects of FXa but not FIIa. Likewise, the transfectants expressing the hirudin-CD4 fusion protein were selectively resistant to the pro-inflammatory effects of FIIa but not those of FXa. When combined, the FXa and FIIa had an additive effect on the activation of control EC. In contrast, coexpression of both hirudin-CD4 and TFPI-CD4 fusion proteins completely inhibited the upregulation of VCAM with the FXa/FIIa mix. These results indicate that expression of novel anticoagulant fusion proteins on the surface of porcine EC can protect against EC activation induced by human coagulation factors FXa and FIIa. In vivo, we anticipate that expression of these fusion proteins on the endothelium of transplanted xenografts, besides preventing intravascular thrombosis, will also protect against EC activation induced by trace amounts of FIIa and FXa, thereby further protecting the grafts from DXR.

Stable recombinant expression and characterization of the two haemophilic factor VIII variants C329S (CRM(-)) and G1948D (CRM(r))

In haemophilia A, the functional defect at the molecular level of most factor VIII (FVIII) missense mutations remains unknown. Site-directed mutagenesis of B domain-deleted FVIII cDNA (FVIIISQ) was used to introduce two mutations associated with severe cross-reacting material (CRM)-negative (FVIII-C329S) or mild/moderate CRM-reduced (FVIII-G1948D) haemophilia A. Wild-type (FVIIISQ-WT) and variant FVIIISQ proteins were successfully expressed after stable transfection in Chinese hamster ovary (CHO) cells, and partially characterized at the intracellular, molecular and functional levels. Reverse transcription polymerase chain reaction analysis confirmed that both transcription and mRNA processing appeared normal in CHO cells transfected with both the wild-type and two variant constructs. In contrast to FVIIISQ-WT, immunofluorescence analysis of both CRM(-) and CRM(r) variants showed intracellular FVIII accumulation within the rough endoplasmic reticulum, suggesting secretion defects in transfected CHO cells. Immunoblot analysis of the FVIIISQ variant proteins that were secreted showed that they were expressed as mixed populations of uncleaved 170 kDa polypeptides, processed 90 kDa heavy chains and 80 kDa light chains, similar to FVIIISQ-WT. Phenotypic analysis of the B domain-deleted FVIIISQ variants expressed in CHO cells correlated well with the patients' reduced FVIII activity and, in addition, surface plasmon resonance studies demonstrated that both missense mutations were associated with increased rates of A2 domain dissociation following thrombin activation. We conclude that the mutations found are responsible for the haemophilia A phenotype, through intracellular retention and decreased stability of the active cofactor FVIIIa.

Factor VII deficiency and the FVII mutation database

Factor VII (FVII) is a zymogen for a vitamin K-dependent serine protease essential for the initiation of blood coagulation. It is synthesized primarily in the liver and circulates in plasma at a concentration of approximately 0.5 microg/ml (10 nmol/L). The FVII gene (F7) is located on chromosome 13 (13q34), consists of 9 exons, and spans approximately 12kb. It encodes a mature protein of 406 amino acids, which has an N-terminal domain (Gla) post-translationally modified by gamma-carboxylation of glutamic acid residues, two domains with homology to epidermal growth factor (EGF1 and 2), and a C-terminal serine protease domain. The single chain zymogen is activated by proteolytic cleavage at Arg152-Ile153. There are 238 individuals described in the world literature with mutations in their F7 genes (FVII mutation database; europium.csc. mrc.ac.uk). Complete absence of FVII activity in plasma is usually incompatible with life, and individuals die shortly after birth due to severe hemorrhage. The majority of individuals with mutations in their F7 gene(s), however, are either asymptomatic or the clinical phenotype is unknown. In general, a severe bleeding phenotype is only observed in individuals homozygous for a mutation in their F7 genes with FVII activities (FVII:C) below 2% of normal, however, a considerable proportion of individuals with a mild-moderate bleeding phenotype have similar FVII:C by in vitro assay. The failure of in vitro tests to differentiate between these groups may be due to lack of sensitivity in the assays to the very low amounts of FVII:C, which are sufficient to initiate coagulation in vivo. A number of polymorphisms have been identified in the F7 gene and some have been shown to influence plasma FVII antigen levels.

Hemophilia A mutations associated with 1-stage/2-stage activity discrepancy disrupt protein-protein interactions within the triplicated A domains of thrombin-activated factor VIIIa

Thrombin-activated factor VIII (FVIIIa) is a heterotrimer with the A2 subunit (amino acid residues 373-740) in a weak ionic interaction with the A1 and A3-C1-C2 subunits. Dissociation of the A2 subunit correlates with inactivation of FVIIIa. Patients with hemophilia A have been described whose plasmas display a discrepancy between their FVIII activities, where the 1-stage activity assay displays greater activity than the 2-stage activity assay. The molecular basis for one of these mutations, (ARG)531(HIS), is an increased rate of A2 subunit dissociation. Examination of a homology model of the A domains of FVIII predicted (ARG)531 to lie at the interface of the A1 and A2 subunits and stabilize their interaction. Indeed, patients with mutations either directly contacting (ARG)531 ((ALA)284(GLU), (ALA)284(PRO)) or closely adjacent to the A1-A2 interface in the tightly packed hydrophobic core ((SER)289(LEU)) have the same phenotype of 1-stage/2-stage discrepancy. The (ALA)284(GLU) and (SER)289(LEU) mutations in FVIII were produced by transfection of COS-1 monkey cells. Compared to FVIII wild-type both mutants had reduced specific activity by 1-stage clotting activity and at least a 2-fold lower activity by 2-stage analysis (COAMATIC), similar to the reported clinical data. Analysis of immunoaffinity purified (ALA)284(GLU) and (SER)289(LEU) proteins in an optical biosensor demonstrated that A2 dissociation was 3-fold faster for both FVIIIa mutants compared to FVIIIa wild-type. Therefore, these mutations within the A1 subunit of FVIIIa introduce a similar destabilization of the FVIIIa heterotrimer compared to the (ARG)531(HIS) mutation within the A2 subunit and support that these residues stabilize the A domain interface of FVIIIa.

Contact activation in shock caused by invasive group A Streptococcus pyogenes

OBJECTIVES

The aim of this study was to characterize abnormalities of coagulation in mice with experimental, invasive group A, streptococcal shock, in an attempt to explain the prolongation of the activated partial thromboplastin time identified in patients with streptococcal toxic shock syndrome.

DESIGN

A longitudinal descriptive animal model study of coagulation times and single coagulation factors in mice infected with Streptococcus pyogenes. This was followed by an experimental study to determine whether streptococci or streptococcal products could activate the human contact system in vitro.

SETTING

University infectious diseases and hemostasis molecular biology laboratories.

SUBJECTS

CD1 outbred mice.

INTERVENTIONS

None.

MEASUREMENTS AND MAIN RESULTS

Coagulation times, single factor assays, and bradykinin assays were conducted on murine plasma at different times after streptococcal infection and compared with uninfected mice. In experiments in which streptococcal products were co-incubated with human plasma, we compared coagulation times, single factor assays, and activities against a range of chromogenic substrates with control plasma. In a murine model of streptococcal necrotizing fasciitis, the activated partial thromboplastin times were significantly prolonged in infected mice compared with controls, whereas prothrombin times were normal, suggesting an isolated abnormality of the intrinsic pathway. Bleeding was not seen. Prolongation of activated partial thromboplastin time was associated with reduced factor XII and prekallikrein, whereas levels of factors VIII, IX, XI, and high molecular weight kininogen were elevated. In vitro studies suggested that streptococcal supernatants can activate prekallikrein, in addition to causing plasminogen activation through the action of streptokinase.

CONCLUSIONS

Prolongation of activated partial thromboplastin time in streptococcal toxic shock syndrome is associated with activation of the contact system, possibly contributing to the profound shock associated with streptococcal toxic shock syndrome.

Molecular analysis of the genotype-phenotype relationship in factor VII deficiency

Factor VII (FVII) deficiency is a rare haemorrhagic condition, normally inherited as an autosomal recessive trait, in which clinical presentation is highly variable and correlates poorly with laboratory phenotype. The FVII (F7) gene was sequenced in 48 unrelated individuals with FVII deficiency, yielding a total of 23 novel lesions including 15 missense mutations, 2 micro-deletions, 5 splice junction mutations and a single base-pair substitution in the 5' untranslated region. Family studies were performed in order to distinguish the contributions of individual mutant F7 alleles to the clinical and laboratory phenotypes. Specific missense mutations were evaluated by molecular modelling in the context of the FVIIa-tissue factor crystal structure. Single base-pair substitutions in splice sites and the 5' untranslated region were studied by in vitro splicing assay and luciferase reporter gene assay, respectively. All probands were also typed for four previously reported F7 polymorphisms. In the majority of cases of FVII deficiency studied here, consideration of both mutational and polymorphism data permitted the derivation of plausible explanations for the FVII activity and antigen levels measured in the laboratory. Inter-familial variation in FVII activity and the antigen levels of heterozygous relatives of probands was found to be significantly higher than intra-familial variation, consistent with the view that the nature of the F7 gene lesion(s) segregating in a given family is a prime determinant of laboratory phenotype. Although no relationship could be discerned between laboratory phenotype and polymorphism genotype, the frequencies of the A2 and M2 polymorphic alleles were significantly higher in the FVII-deficient individuals tested than in controls. This suggests that the presence of these alleles may have served to increase the likelihood of pathological F7 gene lesions coming to clinical attention.

Assay discrepancy in mild haemophilia A due to a factor VIII missense mutation (Asn694Ile) in a large Danish family

Factor VIII gene analysis in a large consanguinous Danish family comprising 24 affected males and four homozygously affected females revealed an Asn694Ile mutation within the A2 domain. The factor VIII gene mutation led to a mild haemophilia A phenotype with factor VIII function displaying discordance between one-stage clotting and chromogenic two-stage assays. In one-stage assays, values ranged from 0.05 to 0.30 IU/ml (males) and from 0.19 to 0.29 IU/ml (homozygous affected females), whereas the chromogenic two-stage assay produced values of around only 50% of the one-stage result [0. 02-0.12 IU/ml (males); 0.06-0.10 IU/ml (females)]. The differences are suggested to be caused by the effect of the mutation on the active cleaved form of the factor (F)VIII protein. As the original amino acid (Asn) is conserved in all known FVIII A2 sequences, but not in ceruloplasmin, we suggest that Asn694 is involved in an A2-specific functional role. Examination of a homology model of the A domains predicts that the Asn694Ile mutation (i) results in the loss of two potential hydrogen-bonding interactions and (ii) hampers the integration of the bulky side-chain of Ile into the A2 domain core, probably causing an altered stability and/or folding of the protein. Interestingly, the disease in this Danish family was originally proposed to be von Willebrand-Jürgens disease. However, the current study rules out the co-existence of either von Willebrand's disease or the presence of the Normandy variant of von Willebrand factor (type 2N).

Electron crystallography of human blood coagulation factor VIII bound to phospholipid monolayers

Coagulation factor VIII binds to negatively charged platelets prior to assembly with the serine protease, factor IXa, to form the factor X-activating enzyme (FX-ase) complex. The macromolecular organization of membrane-bound factor VIII has been studied by electron crystallography for the first time. For this purpose two-dimensional crystals of human factor VIII were grown onto phosphatidylserine-containing phospholipid monolayers, under near to physiological conditions (pH and salt concentration). Electron crystallographic analysis revealed that the factor VIII molecules were organized as monomers onto the lipid layer, with unit cell dimensions: a = 81.5A, b = 67.2 A, gamma = 66.5 degrees, P1 symmetry. Based on a homology-derived molecular model of the factor VIII (FVIII) A domains, the FVIII projection structure solved at 15-A resolution presents the A1, A2, and A3 domain heterotrimer tilted approximately 65 degrees relative to the membrane plane. The A1 domain is projecting on top of the A3, C1, and C2 domains and with the A2 domain protruding partially between A1 and A3. This organization of factor VIII allows the factor IXa protease and epidermal growth factor-like domain binding sites (localized in the A2 and A3 domains, respectively) to be situated at the appropriate position for the binding of factor IXa. The conformation of the lipid-bound FVIII is therefore very close to that for the activated factor VIIIa predicted in the FX-ase complex.

Crystal structure of active site-inhibited human coagulation factor VIIa (des-Gla)

Factor VIIa (FVIIa) is a crucial haemostatic protease consisting of four distinct domains termed the Gla, epidermal growth factor-1 (EGF-1), EGF-2, and protease domains (from N- to C-terminus). The crystal structure of human FVIIa inhibited at the active site with 1, 5-dansyl-Glu-Gly-Arg-chloromethyl ketone and lacking the Gla domain has been solved to a resolution of 2.28 A. The EGF-2 and protease domains were well resolved, whereas no electron density for the EGF-1 domain was observed, suggesting a flexible arrangement or disorder within the crystal. Superposition of the protease domain of the present structure with that previously resolved in the tissue factor (TF)/FVIIai complex revealed that although overall the domain structures are similar, the EGF-2 domain is rotated by 7.5 degrees relative to the protease domain on binding TF. A single cleavage in the protease domain was found, between Arg315 and Lys316 (chymotrypsin numbering 170C-170D) in a FVII-specific insertion loop: this cleavage appeared to be essential for crystallisation. Insertion of the heavy chain N-terminal Ile153 is essentially identical in the two structures, as is the geometry of the active site residues and the inhibitor C-terminal arginine residue. Some differences are seen in the cleaved loop, but changes in TF-contact residues are generally minor. This structure supports the hypothesis that TF binding enables spatial domain arrangements in the flexible FVIIa molecule necessary for procoagulant function and furthermore that active site occupancy induces FVIIa active conformation via N-terminal insertion.

Regulated inhibition of coagulation by porcine endothelial cells expressing P-selectin-tagged hirudin and tissue factor pathway inhibitor fusion proteins

BACKGROUND

Thrombotic vascular occlusion resulting in infarction occurs during hyperacute rejection of allografts transplanted into sensitized patients and remains a major problem in experimental xenotransplantation. A similar process is also found in disorders of diverse etiology including atherosclerosis, vasculitis, and disseminated intravascular coagulation.

METHODS

We have previously constructed two membrane-tethered anticoagulant fusion proteins based on human tissue factor pathway inhibitor and the leech anticoagulant hirudin and demonstrated their functional efficacy in vitro. These constructs have now been modified by the addition of a P-selectin sequence to the cytoplasmic tail to localize them in Weibel-Palade bodies. They have been transfected into Weibel-Palade body-positive endothelial cells isolated from the inferior vena cava of normal pigs.

RESULTS

In resting endothelial cells, fusion protein expression colocalized with P-selectin and was confined to Weibel-Palade bodies. These cells had a procoagulant phenotype in recalcified human plasma. However, after activation with phorbol ester the anticoagulant proteins were rapidly relocated to the cell surface where they specifically inhibited the clotting of human plasma.

CONCLUSIONS

Novel anticoagulant molecules may prove useful therapeutic agents for gene therapy in thrombotic disease and postangioplasty or for transgenic expression in animals whose organs may be used for clinical xenotransplantation. Expression in vascular endothelial cells may be regulated by inclusion of P-selectin cytoplasmic sequence, to restrict cell surface expression to activated endothelium.

Diagnostic importance of the two-stage factor VIII:C assay demonstrated by a case of mild haemophilia associated with His1954-->Leu substitution in the factor VIII A3 domain

In some families with mild haemophilia higher results are obtained for factor VIII activity (FVII:C) determined by one-stage assay than by two-stage or chromogenic assays. Amino-acid substitutions in the A1, A2 and A3 domains of factor VIII have been described in affected individuals with this phenotype. We describe a case of mild haemophilia A in which FVIII:C measured by one-stage assay was normal at 106%. However, FVIII:C levels measured by two-stage and chromogenic assays were 18% and 35% respectively. DNA analysis revealed a novel mutation in the A3 domain of factor VIII, His1954-->Leu. In a molecular model of the FVIII A domains, His1954 is placed in close proximity to two other mutations that have previously been shown also to be associated with one-stage/two-stage discrepancies. In this patient the diagnosis of haemophilia A would be missed if only the one-stage assay was used.

Crystallization and preliminary X-ray analysis of active site-inhibited human coagulation factor VIIa (des-Gla)

Human coagulation factor VIIai that lacks the Gla domain (residues 1-44) has been prepared, purified, and crystallised. First, recombinant factor VII was activated to form factor VIIa, the active site was then inhibited with 1,5-dansyl-Glu-Gly-Arg-chloromethyl ketone, and finally the Gla domain was removed by chymotryptic digestion, yielding factor VIIai (des-Gla). After further purification single crystals suitable for x-ray analysis were obtained by vapour diffusion. Crystals of factor VIIai (des-Gla) belong to the tetragonal space group P41212 or P43212 with unit cell dimensions a = b = 94.85 A, c = 114.30 A, contain one molecule per asymmetric unit, and diffract to 2.3-A resolution when exposed to synchrotron radiation.

Inhibition of tissue factor-dependent and -independent coagulation by cell surface expression of novel anticoagulant fusion proteins

BACKGROUND

Thrombotic vascular occlusion occurs in disorders of diverse etiology, including atherosclerosis, vasculitis, and disseminated intravascular coagulation. The same process results in hyperacute rejection of renal allografts transplanted into sensitized patients and remains a major problem in experimental xenotransplantation.

METHODS

We have previously described the design and expression of several genetic constructs encoding novel fusion proteins with anticoagulant properties. They are based on two naturally occurring soluble anticoagulant proteins, human tissue factor pathway inhibitor (hTFPI) and the leech protein hirudin, which act early and late in the clotting cascade, respectively. We report the expression of human hTFPI-CD4 on the surface of immortalized porcine endothelial cells (IPEC), and show that it functions across the species divide as evidenced by the binding of membrane-expressed porcine tissue factor (pTF)-human factor VIIa complexes.

RESULTS

Using a human plasma recalcification clotting assay, we distinguished between pTF-dependent and pTF-independent fibrin generation, and we have demonstrated that expression of hTFPI-CD4 on IPEC effectively prevented pTF-dependent clotting. Moreover, we show that when hTFPI-CD4 was co-expressed with the hirudin construct, the procoagulant properties of in vitro cultured, activated IPEC were almost completely abolished.

CONCLUSIONS

These results suggest that these novel anticoagulant molecules may prove useful therapeutic agents for gene therapy or for transgenic expression in animals whose organs may be used for cliniCal xenotransplantation.

Expression of hirudin fusion proteins in mammalian cells: a strategy for prevention of intravascular thrombosis

BACKGROUND

Intravascular thrombosis occurs in disorders of diverse pathogeneses, including allograft and xenograft rejection. In this in vitro study, we describe an approach for tethering the specific thrombin inhibitor hirudin to plasma membranes as part of a genetic strategy for regulating intravascular coagulation.

METHODS AND RESULTS

An HLA class I leader sequence was fused with hirudin linked to domains 3 and 4 of human CD4 and intracytoplasmic sequence from either CD4 or human P-selectin. The constructs were transfected into mouse fibroblasts, Chinese hamster ovary (CHO)-K1 cells, immortalized porcine endothelial cells (IPECs), and a pituitary secretory cell line (D16/16). Thrombin binding to the hirudin fusion proteins expressed on fibroblasts and CHO-K1 cells could be blocked by an anti-hirudin monoclonal antibody and by pretreatment of thrombin with either the synthetic tripeptide thrombin inhibitor PPACK or native hirudin. Hirudin expression significantly modified the procoagulant phenotype of IPECs in human plasma, leading to prolongation of clotting times. Hirudin-CD4-P-selectin fusion proteins accumulated in adrenocorticotropic hormone-containing granules in D16/16 cells, with no cell surface expression except on activation with phorbol ester, when hirudin relocated to the outer membrane.

CONCLUSIONS

Hirudin fusion proteins were expressed on mammalian cells, where they reduced local thrombin levels and inhibited fibrin formation. Regulated expression was achieved on activated cells by use of the cytoplasmic sequence from P-selectin. In vivo, these fusion proteins may prove useful transgenic or gene therapy agents for preventing intravascular thrombosis.

The solution structure of human coagulation factor VIIa in its complex with tissue factor is similar to free factor VIIa: a study of a heterodimeric receptor-ligand complex by X-ray and neutron scattering and computational modeling

Factor VIIa (FVIIa) is a soluble four-domain plasma serine protease coagulation factor that forms a tight complex with the two extracellular domains of the transmembrane protein tissue factor in the initiating step of blood coagulation. To date, there is no crystal structure for free FVIIa. X-ray and neutron scattering data in solution for free FVIIa and the complex between FVIIa and soluble tissue factor (sTF) had been obtained for comparison with crystal structures of the FVIIa-sTF complex and of free factor IXa (FIXa). The solution structure of free FVIIa as derived from scattering data is consistent with the extended domain arrangement of FVIIa seen in the crystal structure of its complex with sTF, but is incompatible with the bent, less extended domain conformation seen in the FIXa crystal structure. The FVIIa scattering curve is also compatible with a subset of 317 possible extended structures derived from a constrained automated conformational search of 15 625 FVIIa domain models. Thus, the scattering data support extended domain models for FVIIa free in solution. Similar analyses showed that the solution scattering derived and crystal structures of the FVIIa-sTF complex were in good agreement. An automated constrained search for allowed structures for the complex in solution based on scattering curves showed that only a small family of compact models gave good agreement, namely those in which FVIIa and sTF interact closely over a large surface area. The general utility of this approach for structural analysis of heterodimeric complexes in solution is discussed. Analytical ultracentrifugation data and the modeling of these data were consistent with the scattering results. It is concluded that in solution FVIIa has an extended or elongated domain structure, which allows rapid interaction with sTF over a large surface area to form a high-affinity complex.

Coagulation factor VII Gln100 --> Arg. Amino acid substitution at the epidermal growth factor 2-protease domain interface results in severely reduced tissue factor binding and procoagulant function

We have used recombinant mammalian expression and purification of the factor VII (FVII) variant Gln100 --> Arg (Q100RFVII) to study FVII deficiency in subjects with this mutation. Q100RFVII was secreted poorly in comparison with wild-type FVII (WTFVII) in a stable mammalian expression system, and purified variant protein was found to have undetectable clotting activity. Following activation by immobilized factor Xa, Q100RFVIIa had amidolytic activity similar to WTFVIIa in the absence of soluble tissue factor (sTF); however, unlike WTFVIIa no typical increase in activity was seen after addition of sTF. In a factor X activation assay using relipidated transmembrane truncated tissue factor (residues 1-243), Q100RFVIIa showed less than 5% of the ability of WTFVIIa to activate factor X. We performed direct binding analysis of WT and Q100RFVII/FVIIa to immobilized sTF using surface plasmon resonance, and severely reduced binding of both non-activated and activated Q100RFVII to sTF was seen, indicating a pronounced defect in tissue factor (TF) interaction with this variant. In the sTF-FVIIa crystal structure the candidate residue Gln100 is not in contact with TF but is at the epidermal growth factor 2-protease domain interface. We suggest that the mutation results in a global fold change severely reducing tissue factor interaction; mutation of FVII residues not directly involved in the interaction with TF may still result in variant FVII unable to take part in the initiation of coagulation.

Structural determination of lipid-bound human blood coagulation factor IX

Human coagulation factor IX (FIX) is a serine protease which binds to a negatively charged phospholipid surface in the presence of Ca ions (Ca2+). FIX two-dimensional (2-D) crystals were obtained by the lipid layer crystallisation technique under near physiological conditions. The 2-D projection map of the protein was calculated to a resolution of 3 nm using electron crystallographic analysis. The structural organisation of membrane-bound FIX is discussed and compared with the known X-ray crystallographic data.

The molecular basis for cross-reacting material-positive hemophilia A due to missense mutations within the A2-domain of factor VIII

Factor VIII (FVIII) is the protein defective in the bleeding disorder hemophilia A. Approximately 5% of hemophilia A patients have normal amounts of a dysfunctional FVIII protein and are termed cross-reacting material (CRM)-positive. The majority of genetic alterations that result in CRM-positive hemophilia A are missense mutations within the A2-domain. To determine the mechanistic basis of the genetic defects within the A2-domain for FVIII function we constructed six mutations within the FVIII cDNA that were previously found in five CRM-positive hemophilia A patients (R527W, S558F, I566T, V634A, and V634M) and one CRM-reduced hemophilia A patient (DeltaF652/3). The specific activity for each mutant secreted into the conditioned medium from transiently transfected COS-1 cells correlated with published data for the patients plasma-derived FVIII, confirming the basis of the genetic defect. Sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis of immunoprecipitated FVIII protein radiolabeled in COS-1 cells showed that all CRM-positive mutant proteins were synthesized and secreted into the medium at rates similar to wild-type FVIII. The majority of the DeltaF652/3 mutant was defective in secretion and was degraded within the cell. All mutant FVIII proteins were susceptible to thrombin cleavage, and the A2-domain fragment from the I566T mutant had a reduced mobility because of use of an introduced potential N-linked glycosylation site that was confirmed by N-glycanase digestion. To evaluate interaction of FVIII with factor IXa, we performed an inhibition assay using a synthetic peptide corresponding to FVIII residues 558 to 565, previously shown to be a factor IXa interaction site. The concentration of peptide required for 50% inhibition of FVIII activity (IC50) was reduced for the I566T (800 mumol/L) and the S558F (960 mumol/L) mutants compared with wild-type FVIII (> 2,000 mumol/L). N-glycanase digestion increased I566T mutant FVIII activity and increased its IC50 for the peptide (1,400 mumol/L). In comparison to S558F, a more conservative mutant (S558A) had a sixfold increased specific activity that also correlated with an increased IC50 for the peptide. These results provided support that the defects in the I566T and S558F FVIII molecules are caused by steric hindrance for interaction with factor IXa.

The factor VIII Structure and Mutation Resource Site: HAMSTeRS version 4

Since 1996 the HAMSTeRS (Haemophilia A Mutation, Search, Test and Resource Site) WWW site has provided an online resource for access to data on the molecular pathology of haemophilia A, replacing previous text editions of the Haemophilia A Database published in Nucleic Acids Research . This report describes the continued development of the site (version 4), and in particular the expansion of factor VIII (FVIII) structure-related features. Access to the mutation database itself, both for searching the listings and for submission of new mutations, is via custom-designed forms: more powerful Boolean searches of the point mutations in the database are also available. During 1997 a total of 22 novel missense mutations were reported, increasing the total number of unique variants now described to 252 (238 in exonic sequences and 14 at intronic splice junctions). Currently, a total of 586 individual reports with associated phenotypic data are available for searching by any category including phenotype. The FVIII structure section now includes a download of a FVIII A domain homology model in Protein Data Bank format and a multiple alignment of the FVIII amino-acid sequencies from four species (human, murine, porcine and canine) in addition to the virtual reality simulations, secondary structural data and FVIII animation already available. Finally, to aid navigation across this site, a clickable roadmap of the main features provides easy access to the page desired. Our intention is that continued development and updating of the site shall provide workers in the fields of molecular and structural biology with a one-stop resource site to facilitate FVIII research and education. The HAMSTeRS URL is http://europium.mrc.rpms.ac.uk