Novel insights into the regulation of coagulation by factor V isoforms, tissue factor pathway inhibitorα, and protein S.

Factor V (FV) is a regulator of both pro- and anticoagulant pathways. It circulates as a single-chain procofactor, which is activated by thrombin or FXa to FVa that serves as cofactor for FXa in prothrombin activation. The cofactor function of FVa is regulated by activated protein C (APC) and protein S. FV can also function as an anticoagulant APC cofactor in the inhibition of FVIIIa in the membrane-bound tenase complex (FIXa/FVIIIa). In recent years, it has become clear that FV also functions in multiple ways in the tissue factor pathway inhibitor (TFPI) anticoagulant pathway. Of particular importance is a FV splice variant (FV-Short) that serves as a carrier and cofactor to TFPIα in the inhibition of FXa. FV-Short is generated through alternative splicing of exon 13 that encodes the large activation B domain. A highly negatively charged binding site for TFPIα is exposed in the C-terminus of the FV-Short B domain, which binds the positively charged C-terminus of TFPIα, thus keeping TFPIα in circulation. The binding of TFPIα to FV-Short is also instrumental in localizing the inhibitor to the surface of negatively charged phospholipids, where TFPIα inhibits FXa in process that is stimulated by protein S. Plasma FV activation intermediates and partially proteolyzed platelet FV similarly bind TFPIα with high affinity and regulate formation of prothrombinase. The novel insights gained into the interaction between FV isoforms, TFPIα, and protein S have opened a new avenue for research about the mechanisms of coagulation regulation and also for future development of therapeutics aimed at modulating coagulation.

The association between apolipoprotein M and insulin resistance varies with country of birth.

BACKGROUND AND AIMS: Risk of type 2 diabetes mellitus (T2DM) differs according to ethnicity. Levels of apolipoprotein M (ApoM) have been shown to be decreased in T2DM. However, its role in different ethnicities is not known. We examined the differences in plasma ApoM levels in Swedish residents born in Iraq (Iraqis) and Sweden (Swedes) in relation to T2DM and insulin resistance (IR). METHODS AND RESULTS: Iraqis and Swedes, aged 45-65 years residing in Rosengård area of Malmö were randomly selected from census records and underwent an oral glucose tolerance test. Plasma levels of ApoM were quantified in 162 participants (Iraqis, n = 91; Swedes, n = 71) by a sandwich ELISA method. Age-, sex-, and body mass index (BMI) adjusted plasma ApoM levels differed by country of birth, with Swedes having 18% higher levels compared to Iraqis (p = 0.001). ApoM levels (mean ± SD) were significantly decreased in Swedes with T2DM (0.73 ± 0.18) compared to those with normal glucose tolerance (NGT) (0.89 ± 0.24; p = 0.03). By contrast, no significant difference in ApoM levels was found between Iraqis with T2DM (0.70 ± 0.17) and those with NGT (0.73 ± 0.13; p = 0.41). In multivariate linear regression analysis with an interaction term between IR and country of birth, low ApoM levels remained significantly associated with IR in Swedes (p = 0.008), independently of age, sex, BMI, family history of diabetes, HDL, LDL, and triglycerides, but not in Iraqis (p = 0.35). CONCLUSION: Our results show that ApoM levels differ according to country of birth and are associated with IR and T2DM only in Swedes.

Gas6 is complexed to the soluble tyrosine kinase receptor Axl in human blood.

BACKGROUND: The vitamin K-dependent Gas6 protein (product of growth arrest specific gene 6) binds to, and activates the TAM receptor tyrosine kinases Tyro3, Axl and Mer. It has been suggested that Gas6 and the TAM receptors are important for primary platelet functions, but Gas6 cannot be found in human platelets. However, Gas6 is present in human plasma at a concentration of around 0.2 nM, which is a thousand-fold lower than that of the homologous protein S. The Axl and Mer receptors can be cleaved close to the cell membrane, yielding soluble molecules consisting of the extracellular parts of the receptors. OBJECTIVE: To investigate if soluble Axl (sAxl) is present in human serum and plasma and if Gas6 circulates in complex with sAxl. METHODS: We expressed recombinant sAxl, raised antibodies and developed and validated an ELISA for Axl. Serum and plasma were analyzed using ELISAs for Gas6, Axl and sAxl-Gas6 complexes. Serum was gel filtered and fractions analyzed by the different ELISAs to determine if Gas6 in serum is free or complexed. Immunoprecipitation was used to investigate binding between Gas6 and sAxl in serum. RESULTS: sAxl is present in serum and plasma at around 0.6 nM and all Gas6 is bound to sAxl. No complexes between Gas6 and the soluble forms of Mer and Tyro3 could be detected, indicating that sAxl is the physiological binder of Gas6 in human serum. CONCLUSIONS: Gas6 in circulation is bound to sAxl suggesting circulating Gas6 to be inhibited and incapable of stimulating the TAM receptors.

The role of phospholipid transfer protein in lipoprotein-mediated neutralization of the procoagulant effect of anionic liposomes.

BACKGROUND: Serum has the ability to neutralize the procoagulant properties of anionic liposomes, with transfer of phospholipids (PLs) to both high-density lipoprotein (HDL) and low-density lipoprotein (LDL) particles. Phospholipid transfer protein (PLTP) mediates transfer of PLs between HDL and other lipoproteins and conversion of HDL into larger and smaller particles. OBJECTIVES: To examine the role of PLTP in the neutralization of procoagulant liposomes. METHODS: Procoagulant liposomes were incubated with different lipoproteins in the presence or absence of PLTP, and then tested for their ability to stimulate thrombin formation. RESULTS AND CONCLUSIONS: In the absence of added PLTP, the lipoprotein-enriched fraction, total HDL, HDL(3) and very high-density lipoprotein (VHDL) were all able to neutralize the procoagulant properties of the liposomes. In these samples, endogenous PLTP was present, as judged by Western blotting. In contrast, no PLTP was present in LDL, HDL(2) and lipoprotein-deficient serum, all of which displayed no ability to neutralize the procoagulant liposomes. The phospholipid (PL) transfer activity was dependent on both enzyme (PLTP) and PL acceptor (lipoproteins). After treatment of the VHDL fraction with antiserum against PLTP, the neutralization of procoagulant activity was reduced, but could be regained by the addition of active PLTP. The neutralizing activity was dependent on a catalytically active form of PLTP, and addition of a low activity form of PLTP had no effect. In conclusion, PLTP was found to mediate transfer of anionic PLs to HDL and LDL, thereby neutralizing the effect of procoagulant liposomes, resulting in a reduction of procoagulant activity.

An investigation of serum concentration of apoM as a potential MODY3 marker using a novel ELISA.

OBJECTIVE: To investigate the fitness of serum apolipoprotein M (apoM) concentration as a marker for maturity-onset diabetes of the young 3 (MODY3). STUDY DESIGN AND SUBJECTS: This study consisted of two parts. A family study included 71 carriers of the P291fsinsC mutation in hepatocyte nuclear factor-1alpha (HNF-1alpha) from the Finnish Botnia study, 53 of whom were diabetic, and 75 matched family controls. A second, case-control study included 24 MODY3 patients, 17 healthy MODY3 mutation carriers, 11 MODY1 patients, 18 type 2 diabetes patients and 19 healthy control individuals. Subjects in the case-control study were recruited from the Botnia study or the Clinic of Endocrinology, Malmö University Hospital. Serum apoM levels were measured using a novel ELISA based on two monoclonal apoM antibodies. RESULTS: In the family study, mean serum apoM was 10% lower in female carriers of the P291fsinsC mutation compared to the family controls (P = 0.0058), a difference which remained significant after adjustment for diabetes status. There was no observed difference between groups for men. In the case-control study, no significant difference in apoM concentration was observed between MODY3 and type 2 diabetes patients, neither before nor after adjustment for total cholesterol. CONCLUSIONS: Female carriers of the P291fsinsC mutation in HNF-1alpha displayed slightly lower apoM serum levels. This difference is too small for apoM to be reliably employed as a biomarker for HNF-1alpha mutation status.

Novel APC-cleavage sites in FVa provide insights into mechanisms of action of APC and its cofactor protein S.

BACKGROUND: Activated protein C (APC) inhibits factor Va (FVa) by cleaving at Arg306, Arg506 and Arg679. Protein S serves as cofactor, in particular for the Arg306 site, and a protein S-mediated relocation of the active site of APC closer to the membrane has been proposed as a mechanism. Recently, it was demonstrated that FVa, which was mutated at all three APC-cleavage sites (FVa-306Q/506Q/679Q), could still be cleaved by APC. These sites were close to Arg306 and Arg506 but not further defined. OBJECTIVE: To identify and characterize the additional APC-cleavage sites in FVa. METHODS: The cDNA for FV-306Q/506Q/679Q was used as a template to create FV variants with one or more possible cleavage sites being mutated. The FV variants were expressed and their sensitivity for APC characterized functionally and with Western blotting. RESULTS: The additional APC-cleavage sites were located at Lys309, Arg313, Arg316, Arg317 and Arg505. FVa-306Q/309Q/313Q/316Q/317Q/505Q/506Q/679Q (denoted 8M-FVa) was APC resistant. To investigate individual sites, they were mutated back using 8M-FV as a template. The kinetics of APC-degradation of these variants demonstrated that protein S was equally efficient in enhancing the APC effect for all the novel sites. CONCLUSIONS: Multiple APC-cleavage sites close to Arg306 and a single site close to Arg506 were identified. Protein S was equally efficient as APC cofactor for all novel sites. The stimulation by protein S of the Arg505 cleavage argues against a specific protein S-mediated stimulation of cleavage at Arg306 due to relocation of the APC active site closer to the membrane.

Apolipoprotein M predicts pre-beta-HDL formation: studies in type 2 diabetic and nondiabetic subjects.

OBJECTIVE: Studies in mice suggest that plasma apoM is lowered in hyperinsulinaemic diabetes and that apoM stimulates formation of pre-beta-HDL. Pre-beta-HDL is an acceptor of cellular cholesterol and may be critical for reverse cholesterol transport. Herein, we examined whether patients with type 2 diabetes have reduced plasma apoM and whether apoM is associated with pre-beta-HDL formation and cellular cholesterol efflux. DESIGN: In 78 patients with type 2 diabetes and 89 control subjects, we measured plasma apoM with ELISA, pre-beta-HDL and pre-beta-HDL formation, phospholipid transfer protein (PLTP) activity and the ability of plasma to promote cholesterol efflux from cultured fibroblasts. RESULTS: ApoM was approximately 9% lower in patients with type 2 diabetes compared to controls (0.025 +/- 0.006 vs. 0.027 +/- 0.007 g L(-1), P = 0.01). The difference in apoM was largely attributable to diabetes-associated obesity. ApoM was positively related to both HDL (r = 0.16; P = 0.04) and LDL cholesterol (r = 0.28; P = 0.0003). Pre-beta-HDL and pre-beta-HDL formation were not different between diabetic and control subjects. ApoM predicted pre-beta-HDL (r = 0.16; P = 0.04) and pre-beta-HDL formation (r = 0.19; P = 0.02), even independently of positive relationships with apoA-I, HDL-cholesterol and PLTP activity. Cellular cholesterol efflux to plasma was positively related to pre-beta-HDL and PLTP activity but not significantly to apoM. CONCLUSIONS: Plasma apoM is modestly reduced in type 2 diabetes. Pre-beta-HDL and pre-beta-HDL formation are positively associated with apoM, supporting the hypothesis that apoM plays a role in HDL remodelling in humans. Lower apoM may provide a mechanism to explain why pre-beta-HDL formation is not increased in type 2 diabetes despite elevated PLTP activity.

Apolipoprotein M affecting lipid metabolism or just catching a ride with lipoproteins in the circulation?

Apolipoprotein M (apoM) is a novel apolipoprotein found mainly in high-density lipoproteins (HDL). Its function is yet to be defined. ApoM (25 kDa) has a typical lipocalin ss-barrel fold and a hydrophobic pocket. Retinoids bind apoM but with low affinity and may not be the natural ligands. ApoM retains its signal peptide, which serves as a hydrophobic anchor to the lipoproteins. This prevents apoM from being lost in the urine. Approximately 5% of HDL carries an apoM molecule. ApoM in plasma (1 microM) correlates strongly with both low-density lipoprotein (LDL) and HDL cholesterol, suggesting a link to cholesterol metabolism. However, in casecontrol studies, apoM levels in patients with coronary heart disease (CHD) and controls were similar, suggesting apoM levels not to affect the risk for CHD in humans. Experiments in transgenic mice suggested apoM to have antiatherogenic properties; possible mechanisms include increased formation of pre-ss HDL, enhanced cholesterol mobilization from foam cells, and increased antioxidant properties.

Genetic loss of Gas6 induces plaque stability in experimental atherosclerosis.

The growth arrest-specific gene 6 (Gas6) plays a role in pro-atherogenic processes such as endothelial and leukocyte activation, smooth muscle cell migration and thrombosis, but its role in atherosclerosis remains uninvestigated. Here, we report that Gas6 is expressed in all stages of human and mouse atherosclerosis, in plaque endothelial cells, smooth muscle cells and macrophages. Gas6 expression is most abundant in lesions containing high amounts of macrophages, ie thin fibrous cap atheroma and ruptured plaque. Genetic loss of Gas6 does not affect the number and size of initial and advanced plaques in ApoE(-/-) mice, but alters its plaque composition. Compared to Gas6(+/+): ApoE(-/-) mice, initial and advanced plaques of Gas6(-/-): ApoE(-/-) mice contained more smooth muscle cells and more collagen and developed smaller lipid cores, while the expression of TGFbeta was increased. In addition, fewer macrophages were found in advanced plaques of Gas6(-/-): ApoE(-/-) mice. Hence, loss of Gas6 promotes the formation of more stable atherosclerotic lesions by increasing plaque fibrosis and by attenuating plaque inflammation. These findings identify a role for Gas6 in plaque composition and stability.

Apolipoprotein M: progress in understanding its regulation and metabolic functions.

ApoM is a novel apolipoprotein mainly present in high-density lipoprotein (HDL). It belongs to the lipocalin protein superfamily and may bind a small but so far unknown lipophilic ligand. It is secreted without cleavage of its hydrophobic signal peptide, which probably anchors apoM in the phospholipid moiety of plasma lipoproteins. Recent studies suggest that apoM may affect HDL metabolism and have anti-atherogenic functions. The subfraction of human HDL that contains apoM therefore protects LDL from oxidation and mediates cholesterol efflux more efficiently then HDL without apoM. In addition to hepatocytes, apoM is highly expressed in kidney proximal tubule cells. Recent data suggest that apoM is secreted into the pre-urine from the tubule cells but is normally taken up again in a megalin-dependent fashion. Further studies of mice with genetically modified apoM expression will be essential to unravel the potential roles of apoM in lipoprotein metabolism, atherosclerosis and kidney biology.

Activated protein C resistance -- in the absence of factor V Leiden -- and pregnancy.

BACKGROUND: Activated protein C (APC) resistance with or without factor V Leiden (FVL) is a major risk factor for venous thromboembolism. Many previous pregnancy studies have been focused on APC resistance caused by FVL. Very few have investigated APC resistance in the absence of FVL (APCR(FVL-)). MATERIAL AND METHODS: In a prospective study of 2480 unselected gravidae, blood was drawn in early pregnancy (mean = 12 weeks of gestation). APC resistance was analyzed by an APTT-based method (Coatest) APC-resistance) and the presence of FVL was determined by PCR. The APCR(FVL-) group had similar mean APC resistance ratio as the heterozygous carriers of FVL. The analyses were carried out no earlier than 3 months after delivery when all data were recorded. Small-for-gestational age (SGA) was used as a proxy for intrauterine growth restriction. RESULTS: When compared with the control group, women with APCR(FVL-) had no increased risk of SGA, pre-eclampsia, first trimester fetal loss or venous thromboembolism. However, they had an increased risk of second trimester fetal loss (7.3% vs. 2.7%, P = 0.01), and a tendency of being overweight (17.3% vs. 12.6%, P = 0.19) and of delivering extremely preterm (2.8% vs. 1.0%, P = 0.11). CONCLUSION: Women with APC resistance not caused by FVL were not at increased risk for SGA, pre-eclampsia, first trimester fetal loss, or abnormal blood loss. However, they showed an increased prevalence of second trimester fetal loss.

Co-segregation of the PROS1 locus and protein S deficiency in families having no detectable mutations in PROS1.

Inherited deficiency of protein S constitutes an important risk factor of venous thrombosis. Many reports have demonstrated that causative mutations in the protein S gene are found only in approximately 50% of the cases with protein S deficiency. It is uncertain whether the protein S gene is causative in all cases of protein S deficiency or if other genes are involved in cases where no mutation is identified. The aim of the current study was to determine whether haplotypes of the protein S gene cosegregate with the disease phenotype in cases where no mutations have been found. Eight protein S-deficient families comprising 115 individuals where previous DNA sequencing had failed to detect any causative mutations were analyzed using four microsatellite markers in the protein S gene region. Co-segregation between microsatellite haplotypes and protein S deficiency was found in seven of the investigated families, one family being uninformative. This suggests that the causative genetic defects are located in or close to the protein S gene in a majority of such cases where no mutations have been found.

Factor V I359T: a novel mutation associated with thrombosis and resistance to activated protein C.

We report a kindred in which two siblings suffered spontaneous venous thromboses in the second decade of life. Further investigation showed reduced coagulation factor V (FV) activity and activated protein C resistance (APCR) ratio but no other thrombophilic abnormalities. The reduction in APCR ratio persisted in a modified APCR assay in which FV activity was normalized between test and control plasmas. Analysis of the FV gene showed that the thrombotic individuals had a complex genotype that included two novel point mutations c.529G>T and c.1250T>C resulting in FV E119X and FV I359T substitutions inherited on different alleles. Individuals in the kindred with FV E119X or FV I359T substitutions alone were asymptomatic. We suggest that the FV I359T substitution confers pro-thrombotic risk and APCR, but that this is only clinically manifest when co-inherited with the FV E119X allele. The FV I359T substitution creates a new consensus sequence for N-linked glycosylation within the FV heavy chain and we speculate that this abnormal glycosylation may disrupt activated protein C-mediated proteolysis of the variant FV and FVa.

Interactions between surface proteins of Streptococcus pyogenes and coagulation factors modulate clotting of human plasma.

Invasive and toxic infections caused by Streptococcus pyogenes are connected with high morbidity and mortality. Typical symptoms of these infections are hypotension, edema formation, tissue necrosis, and bleeding disorders. Here we report that components of the coagulation system including fibrinogen, factors V, XI, and XII, and H-kininogen, are assembled at the surface of S. pyogenes through specific interactions with bacterial surface proteins. In plasma environment, absorption of fibrinogen by S. pyogenes causes a hypocoagulatory state resulting in prolonged clotting times and impaired fibrin network formation. Moreover, the binding of coagulation factors and the subsequent activation of the coagulation system at the bacterial surface lead to the formation of a fibrin network covering S. pyogenes bacteria adhering to epithelial cells. The results suggest that interactions between S. pyogenes and components of the coagulation system contribute to some of the symptoms seen in severe infections caused by this important human pathogen.

The discovery of activated protein C resistance.

Venous thrombosis is a multifactorial disease, with the pathogenesis involving genetic and environmental risk factors. The most common genetic risk factor known to date is a single point mutation in the gene of coagulation factor V (FV), which results in the replacement of Arg506 with Gln (FV Leiden). Arg506 is one of several cleavage sites in FV for anticoagulant activated protein C (APC) and the mutation results in the loss of the cleavage site. Via a complicated series of reactions, this results in impaired APC-mediated degradation of both FVa and FVIIIa. The associated hypercoagulable condition, which causes a lifelong increased risk of thrombosis, is known as APC resistance. APC resistance was discovered in my laboratory in the late 1980s and we published the first report almost exactly 10 years ago. This started an avalanche of research activities and several thousand articles have since been published on this topic. Analyses for APC resistance and FV Leiden have made their way into clinical medicine and are now performed routinely all over the world. I have been asked to write a personal historical annotation about the discovery of APC resistance, the early research activities and the rapid progress in this field.

Molecular recognition in the protein C anticoagulant pathway.

The protein C (PC) anticoagulant system provides specific and efficient control of blood coagulation. The system comprises circulating or membrane-bound protein components that take part in complicated multimolecular protein complexes being assembled on specific cellular phospholipid membranes. Each of the participating proteins is composed of multiple domains, many of which are known at the level of their three-dimensional structures. The key component of the PC system, the vitamin K-dependent PC, circulates in blood as zymogen to an anticoagulant serine protease. Activation is achieved on the surface of endothelial cells by thrombin bound to the membrane protein thrombomodulin. The endothelial PC receptor binds the Gla domain of PC and stimulates the activation. Activated PC (APC) modulates the activity of blood coagulation by specific proteolytic cleavages of a limited number of peptide bonds in factor (F)VIIIa and FVa, cofactors in the activation of FX and prothrombin, respectively. These reactions occur on the surface of negatively charged phospholipid membranes and are stimulated by the vitamin K-dependent protein S. Regulation of FVIIIa activity by APC is stimulated not only by protein S but also by FV, which, like thrombin, is a Janus-faced protein with both pro- and anticoagulant potential. However, whereas the properties of thrombin are modulated by protein-protein interactions, the specificity of FV function is governed by proteolysis by pro- or anti-coagulant enzymes. The molecular recognition of the PC system is beginning to be unravelled and provides insights into a fascinating and intricate molecular scenario.

Anti-factor V auto-antibody in the plasma and platelets of a patient with repeated gastrointestinal bleeding.

Development of autoantibody against coagulation factor V (FV) is a rare clinical condition with hemorrhagic complications of varying severity. The aim of this study was to establish the pathomechanism of an acquired FV deficiency and characterize the FV inhibitor responsible for the clinical symptoms. A 78-year-old female was admitted to hospital with severe gastrointestinal bleeding. General clotting tests and determination of clotting factors were performed by standard methods. FV antigen and FV containing immune complexes were measured by ELISA. The FV molecule was investigated by Western blotting and by sequencing the f5 gene. The binding of patient's IgG to FV and activated FV (FVa) was demonstrated in an ELISA system and its effect on the procoagulant activity of FVa was tested in clotting tests and in a chromogenic prothrombinase assay. Localization of the epitope for the antibody was performed by blocking ELISA. FV activity was severely suppressed both in plasma and platelets. FV antigen levels were normal by ELISA using polyclonal anti-FV antibody or monoclonal antibody against the connecting region of FV, but depressed when HV1 monoclonal antibody against the C2 domain in the FV light-chain was used as capture antibody. The FV molecule was found intact. An IgG reacting with both FV and FVa was present in the patient's plasma and its binding to FV was inhibited by HV1 antibody. FV-containing immune complexes were detected in the patient's plasma and platelet lysate. The patient's IgG inhibited the procoagulant function of FVa. An anti-FV IgG was present in the patient's plasma and platelets. The autoantibody reacted with an epitope in the C2 domain of FV light chain and neutralized the procoagulant function of FVa.

Bordetella pertussis binds to human C4b-binding protein (C4BP) at a site similar to that used by the natural ligand C4b.

Human complement regulators are important targets for pathogenic microorganisms. In one such interaction, Bordetella pertussis binds human C4b-binding protein (C4BP), a high-molecular-weight plasma protein that acts as inhibitor of the classical pathway of complement activation. At least two different B. pertussis surface components, one of which is the virulence factor filamentous hemagglutinin (FHA), contribute to the binding. We used a set of C4BP mutants and monoclonal antibodies to characterize the region in C4BP that binds B. pertussis and analyzed the salt sensitivity of the interaction. These studies indicated that positively charged residues at the interface between complement control protein modules 1-2 in the C4BP alpha-chain are important for binding, and that the site in C4BP that binds B. pertussis is very similar, but not identical, to the C4b-binding site. Bacteria-bound C4BP retained its complement regulatory function and B. pertussis selectively bound C4BP in human plasma, indicating that binding occurs also in vivo. Together, these findings indicate that B. pertussis exploits a site in C4BP, resembling that used by the natural ligand C4b.