Matthew F. Whelihan

Citrate anticoagulation and the dynamics of thrombin generation

Summary.  Background: Sodium citrate has been used as an anticoagulant to stabilize blood and blood products for over 100 years, presumably by sequestering Ca++ ions in vitro. Anticoagulation of blood without chelation can be achieved by inhibition of the contact pathway by corn trypsin inhibitor (CTI). Objective: To evaluate the influence of citrate anticoagulation on the performance of blood, platelet‐rich and platelet‐poor plasma assays. Methods: Blood was anticoagulated in three ways: by collection into citrate, CTI and citrate with CTI. Plasma was prepared using each anticoagulation regimen. Functional analyses included calibrated automated thrombography, thromboelastography, plasma clotting, the synthetic coagulation proteome and platelet aggregation. Coagulation reactions were initiated with tissue factor–phospholipid and Ca++ (when indicated).Results: In all cases, citrate anticoagulation resulted in reaction dynamics significantly altered relative to blood or plasma stabilized with CTI alone. Subsequent experiments showed that calcium citrate itself impairs coagulation dynamics.Conclusion: Coagulation analyses using blood that has been exposed to citrate and recalcified do not yield reliable depictions of the natural dynamics of blood coagulation processes.

Prothrombin activation in blood coagulation: the erythrocyte contribution to thrombin generation

Prothrombin activation can proceed through the intermediates meizothrombin or prethrombin-2. To assess the contributions that these 2 intermediates make to prothrombin activation in tissue factor (Tf)-activated blood, immunoassays were developed that measure the meizothrombin antithrombin (mTAT) and α-thrombin antithrombin (αTAT) complexes. We determined that Tf-activated blood produced both αTAT and mTAT. The presence of mTAT suggested that nonplatelet surfaces were contributing to approximately 35% of prothrombin activation. Corn trypsin inhibitor-treated blood was fractionated to yield red blood cells (RBCs), platelet-rich plasma (PRP), platelet-poor plasma (PPP), and buffy coat. Compared with blood, PRP reconstituted with PPP to a physiologic platelet concentration showed a 2-fold prolongation in the initiation phase and a marked decrease in the rate and extent of αTAT formation. Only the addition of RBCs to PRP was capable of normalizing αTAT generation. FACS on glycophorin A-positive cells showed that approximately 0.6% of the RBC population expresses phosphatidylserine and binds prothrombinase (FITC Xa·factor Va). These data indicate that RBCs participate in thrombin generation in Tf-activated blood, producing a membrane that supports prothrombin activation through the meizothrombin pathway.

The role of the red cell membrane in thrombin generation

Red blood cells have historically been viewed as innocent bystanders in the process of blood coagulation and thrombin generation; however a century of clinical evidence linking red blood cells to thrombosis suggests the contrary. In this brief review, the biochemical evidence for red blood cell involvement in thrombin generation is evaluated. It is concluded that in addition to platelets, red blood cells actively participate in thrombin generation. A sub-fraction of red blood cells express phosphatidylserine on their surface and unlike platelets, red blood cells produce thrombin through the meizothrombin pathway, which has interesting consequences in the context of clot formation and stabilization.

Coagulation procofactor activation by factor XIa.

Summary.  Background: In the extrinsic pathway, the essential procofactors factor (F) V and FVIII are activated to FVa and FVIIIa by thrombin. In the contact pathway and its clinical diagnostic test, the activated partial thromboplastin time (APTT) assay, the sources of procofactor activation are unknown. In the APTT assay, FXII is activated on a negatively charged surface and proceeds to activate FXI, which activates FIX upon the addition of Ca2+. FIXa feeds thrombin generation through activation of FX. FIXa is an extremely poor catalyst in the absence of its FVIIIa cofactor, which, in the intrinsic FXase complex, increases FXa generation by ∼ 107. One potential APTT procofactor activator in this setting is FXIa. Objective: To test the hypothesis that FXIa can activate FVIII and FV. Methods: Recombinant FVIII and plasma FV were treated with FXIa, and the activities and integrities of each procofactor were measured using commercial clotting assays and sodium dodecylsulfate polyacrylamide gel electrophoresis (SDS‐PAGE). Results: Kinetic analyses of FXIa‐catalyzed activation and inactivation of FV and FVIII are reported, and the the timing and sites of cleavage are defined. Conclusions: FXIa activates both procofactors at plasma protein concentrations, and computational modeling suggests that procofactor activation during the preincubation phase of the APTT assay is critical to the performance of the assay. As the APTT assay is the primary tool for the diagnosis and management of hemophilias A and B, as well as in the determination of FVIII inhibitors, these findings have potential implications in the clinical setting.

Thrombin generation and bleeding in haemophilia A.

Summary.  Haemophilia A displays phenotypic heterogeneity with respect to clinical severity. The aim of this study was to determine if tissue factor (TF)‐initiated thrombin generation profiles in whole blood in the presence of corn trypsin inhibitor (CTI) are predictive of bleeding risk in haemophilia A. We studied factor(F) VIII deficient individuals (11 mild, 4 moderate and 12 severe) with a well‐characterized 5‐year bleeding history that included haemarthrosis, soft tissue haematoma and annual FVIII concentrate usage. This clinical information was used to generate a bleeding score. The bleeding scores (range 0–32) were separated into three groups (bleeding score groupings: 0, 0 and ≤9.6, >9.6), with the higher bleeding tendency having a higher score. Whole blood collected by phlebotomy and contact pathway suppressed by 100 μg mL−1 CTI was stimulated to react by the addition of 5 pm TF. Reactions were quenched at 20 min by inhibitors. Thrombin generation, determined by enzyme‐linked immunosorbent assay for thrombin–antithrombin was evaluated in terms of clot time (CT), maximum level (MaxL) and maximum rate (MaxR) and compared to the bleeding score. Data are shown as the mean±SD. MaxL was significantly different (P < 0.001) between the groups: 504 ± 114, 315 ± 117 and 194 ± 91 nm; with higher thrombin concentrations in the groups with lower bleeding scores. MaxR was higher in the groups with a lower bleeding score; 97 ± 51, 86 ± 60 and 39 ± 16 nm min−1 (P = 0.09). No significant difference was detected in CT among the groups, 5.6 ± 1.3, 4.7 ± 0.7 and 5.6 ± 1.3 min. Our empirical study in CTI‐inhibited whole blood shows that the MaxL of thrombin generation appears to correlate with the bleeding phenotype of haemophilia A.

The influence of prophylactic factor VIII in severe haemophilia A

Summary.  Haemophilia A individuals displaying a similar genetic defect have heterogeneous clinical phenotypes. Our objective was to evaluate the underlying effect of exogenous factor (f)VIII on tissue factor (Tf)‐initiated blood coagulation in severe haemophilia utilizing both empirical and computational models. We investigated twenty‐five clinically severe haemophilia A patients. All individuals were on fVIII prophylaxis and had not received fVIII from 0.25 to 4 days prior to phlebotomy. Coagulation was initiated by the addition of Tf to contact‐pathway inhibited whole blood ± an anti‐fVIII antibody. Aliquots were quenched over 20 min and analyzed for thrombin generation and fibrin formation. Coagulation factor levels were obtained and used to computationally predict thrombin generation with fVIII set to either zero or its value at the time of the draw. As a result of prophylactic fVIII, at the time of the blood draw, the individuals had fVIII levels that ranged from <1% to 22%. Thrombin generation (maximum level and rate) in both empirical and computational systems increased as the level of fVIII increased. FXIII activation rates also increased as the fVIII level increased. Upon suppression of fVIII, thrombin generation became comparable in both systems. Plasma composition analysis showed a negative correlation between bleeding history and computational thrombin generation in the absence of fVIII. Residual prophylactic fVIII directly causes an increase in thrombin generation and fibrin cross‐linking in individuals with clinically severe haemophilia A. The combination of each individual’s coagulation factors (outside of fVIII) determine each individual’s baseline thrombin potential and may affect bleeding risk.

The contribution of red blood cells to thrombin generation in sickle cell disease: meizothrombin generation on sickled red blood cells.

Homozygous sickle-cell disease (HbSS; SCD) is associated with vaso-occlusive manifestations of varying severity [1]. Pain crises are generally believed to result from obstruction of the microvasculature secondary to adhesion of red blood cells (RBCs) and other cellular elements, and decreased deformability of hypoxia-induced sickled RBCs, with ensuing activation of coagulation and inflammatory pathways [2, 3]. One of the proposed contributors to thrombosis in SCD is the loss of normal phospholipid asymmetry on sickled RBCs due to the repeated process of hypoxia-induced sickling and unsickling [4]. The exposure of anionic phosphatidylserine (PS) on the outer membrane supports the assembly of enzymatic clotting reactions, leading to a sub-population of RBCs with a prothrombotic phenotype [5, 6]. We recently demonstrated that in healthy individuals, a subpopulation (∼0.5%) of PS-expressing RBCs contribute a significant fraction (∼40%) of the total thrombin generating potential of blood [7]. Prothrombin activation requires factor Xa to perform two proteolytic cleavages at Arg 271 and Arg 320 to release the active α-thrombin (αIIa) product [8]. Depending on the order of proteolysis, prothrombin activation can occur via two possible intermediates; meizothrombin (mIIa), an active enzyme; or prethrombin-2 (pre2), a non-enzymatic intermediate [9]. Unlike platelets, which support thrombin generation exclusively through the pre2 intermediate [10], this subpopulation of procoagulant RBCs supports prothrombin activation via the mIIa intermediate in a manner similar to that on synthetic phospholipids [7]. mIIa is of interest because it exhibits the anticoagulant functions of αIIa towards protein C activation, while lacking any significant activity towards procoagulant substrates like fibrinogen, FV and platelets [11]. Due to the markedly enhanced PS expression by sickled RBCs, we therefore hypothesized that the total αIIa generation potential, as well as mIIa production, would be significantly increased in the whole blood of SCD patients compared to healthy controls. To test this hypothesis, we recruited 7 outpatients (4 female and 3 male, age 28-51) with HbSS, in their non-crisis, “steady states,” and 6 healthy African-American controls (3 female and 3 male age 24-38). None of the patients or controls was being treated currently with anticoagulant or anti-platelet therapy, and all SCD patients were at least 3 months remote from red cell transfusion or hospital admission for pain crisis. We utilized our immunoassays that are capable of selective quantitation of αIIa-antithrombin (αTAT) and mIIa-antithrombin (mTAT) to measure the relative production of the two species produced in SCD vs. control individuals after tissue factor (TF)-initiated coagulation. To examine the thrombin generation potential of the HbSS cohort vs. that of the control group, whole blood was subjected to a 5 pM TF stimulus in the presence of 0.1 mg/mL of corn trypsin inhibitor. Quenched time course samples were subsequently analyzed using αTAT and mTAT ELISAs. Figure 1A displays the time course data for αTAT generation in the HbSS and control cohorts. The control cohort clotted on average at 3.8±0.2 min (mean±SEM), generated αTAT at a rate of 63±3.9 nM/min and reached a maximum αTAT level of 502±14 nM. Unexpectedly, clot time (3.98±0.2 min) and the maximum level (515±49 nM) of αTAT generated in the HbSS cohort were similar to that observed in the controls while the rate of αTAT generation at 74.1±7.9 nM/min was only 15% higher (P>0.05). Figure 1 αTAT and mTAT ELISA analyses of HbSS and Control groups. Figure 1B displays the time course data for mTAT generation. As reported previously [7], the observed mTAT levels were significantly less than αTAT due to the lability of the mIIa intermediate. The control cohort generated mTAT at an average maximum rate of 1.02±0.13 nM/min and reached an average maximum level of 6.5±0.52 nM. The HbSS cohort generated mTAT 33% faster (1.51±0.12 nM/min) and reached a maximum level of mTAT (10.1±0.44 nM) that was 36% higher than that observed in the control group. Thus, unlike the case with αTAT, a significant difference (mTAT max level P<0.001, mTAT max rate P=0.02) in the mTAT generation profiles between the two subject groups was observed. The expression of PS on RBCs was detected by the binding of FITC-labeled bovine lactadherin using flow cytometry. Consistent with our previous report [7], the control cohort displayed low levels (0.59±0.19%) of PS-expressing RBCs, whereas the HbSS cohort displayed a wide range (2-26%) of PS-expressing RBCs with a mean of 6.4±6.1%, consistent with previous reports [6, 12]. The number of PS-expressing cells relative to the individual hematocrits in the HbSS and control groups were examined and compared to the rates of mTAT generation in the time course samples. Six of the 7 HbSS patients and 5 of the 6 controls were analyzed. Although these data were not statistically significant, the control cohort showed a trend towards a positive correlation of 0.74 (P=0.09) and the HbSS cohort also showed a positive correlation of 0.55 (P=0.11). RBC microparticles (MPs) were also measured in both cohorts and on average, the SCD cohort showed double the level of RBC MPs observed in the controls (1116±1079/μL vs 677±230/μL respectively). However, no correlation between the rate of either αTAT or mTAT formation and the RBC microparticle counts in either cohort was observed. In addition to the hemolytic and vaso-occlusive complications experienced by patients with SCD, activation of coagulation remains a critical component of their pathology [13]. Previous studies in SCD have generally been performed using TF-initiated thrombin generation in platelet-poor plasma, and have failed to agree on whether the principal indices of thrombin generation are abnormal [13]. While the lack of consensus on this issue may be partly explained by varying pre-analytical and analytical conditions, this experimental system fails to take into account the important potential role that red cells and other circulating hematopoietic cells play in the hypercoagulable state associated with SCD. We were therefore somewhat surprised to see that in our minimally modified whole blood model, the overall extent of αTAT generation did not differ significantly between the two groups. However, in view of the large (∼50%) reduction in RBC number in SCD (2.50 ± 0.73 × 106/μL in SCD vs. 4.93 ± 0.56 × 106/μL in controls) the erythrocyte contribution to overall thrombin generation is probably much greater than we have observed in healthy individuals. In support of this hypothesis, the rate of mTAT generation and the maximum mTAT level attained in the HbSS cohort were 36% greater than that observed in controls. As previously reported, maximum mTAT levels in both groups were approximately 40-fold lower than the corresponding αTAT levels, most likely due to the labile nature of the Arg 155 and Arg 271 bonds in mIIa species in the clotting blood of a closed system which favors the formation of α-thrombin [14]. Furthermore, since factor Xa is the limiting component of the prothrombinase complex, more phospholipid surface does not necessarily equate to more thrombin generation. However, although limited by the small study size, the increase in PS expression observed in HbSS RBC populations appeared to be directly proportional to mTAT production, in agreement with our previous report [7]. Thus, the higher rate of mIIa generation observed in HbSS is in support of our previous work implicating PS-expressing RBC membranes as the primary source of meizothrombin production in TF-activated whole blood. The mTAT generation therefore appears to represent a useful probe of red cell-dependent thrombin generation in pathologic states as well as in healthy controls. Future studies will be aimed at measuring the specific contribution of RBCs to overall thrombin generation in both ‘steady state’ and during crises, as well as the effect of hematocrit and RBC expression of PS in other hemolytic anemias.

Platelets do not express the oxidized or reduced forms of tissue factor.

BACKGROUND Expression of tissue factor (TF) antigen and activity in platelets is controversial and dependent upon the laboratory and reagents used. Two forms of TF were described: an oxidized functional form and a reduced nonfunctional form that is converted to the active form through the formation of an allosteric disulfide. This study tests the hypothesis that the discrepancies regarding platelet TF expression are due to differential expression of the two forms. METHODS Specific reagents that recognize both oxidized and reduced TF were used in flow cytometry of unactivated and activated platelets and western blotting of whole platelet lysates. TF-dependent activity measurements were used to confirm the results. RESULTS Western blotting analyses of placental TF demonstrated that, in contrast to anti-TF#5, which is directed against the oxidized form of TF, a sheep anti-human TF polyclonal antibody recognizes both the reduced and oxidized forms. Flow cytometric analyses demonstrated that the sheep antibody did not react with the surface of unactivated platelets or platelets activated with thrombin receptor agonist peptide, PAR-1. This observation was confirmed using biotinylated active site-blocked factor (F)VIIa: no binding was observed. Likewise, neither form of TF was detected by western blotting of whole platelet lysates with sheep anti-hTF. Consistent with these observations, no FXa or FIXa generation by FVIIa was detected at the surface of these platelets. Similarly, no TF-related activity was observed in whole blood using thromboelastography. CONCLUSION AND SIGNIFICANCE Platelets from healthy donors do not express either oxidized (functional) or reduced (nonfunctional) forms of TF.

Activated protein C inhibitor for correction of thrombin generation in hemophilia A blood and plasma1.

Summary.  Background: Replacement therapy for hemophilic patient treatment is costly, because of the high price of pharmacologic products, and is not affordable for the majority of patients in developing countries. Objective: To generate and evaluate low molecular weight agents that could be useful for hemophilia treatment. Methods: Potential agents were generated by synthesizing specific inhibitors [6‐(Lys‐Lys‐Thr‐[homo]Arg)amino‐2‐(Lys[carbobenzoxy]‐Lys[carbobenzoxy]‐O‐benzyl)naphthalenesulfonamide] (PNASN‐1)] for activated protein C (APC) and tested in plasma and fresh blood from hemophilia A patients. Results: In the activated partial thromboplastin time‐based APC resistance assay, PNASN‐1 partially neutralized the effect of APC. In calibrated automated thrombography, PNASN‐1 neutralized the effect of APC on thrombin generation in normal and congenital factor VIII‐deficient plasma (FVIII:C < 1%). The addition of PNASN‐1 to tissue factor‐triggered (5 pm) contact pathway‐inhibited fresh blood from 15 hemophilia A patients with various degrees of FVIII deficiency (FVIII:C < 1–51%) increased the maximum level of thrombin generated from 78 to 162 nm, which approached that observed in blood from a healthy individual (201 nm). PNASN‐1 also caused a 47% increase in clot weight in hemophilia A blood. Conclusions: Specific APC inhibitors compensate to a significant extent for FVIII deficiency, and could be used for hemophilia treatment.

Thrombin generation and fibrin clot formation under hypothermic conditions: an in vitro evaluation of tissue factor initiated whole blood coagulation

BACKGROUND Despite trauma-induced hypothermic coagulopathy being familiar in the clinical setting, empirical experimentation concerning this phenomenon is lacking. In this study, we investigated the effects of hypothermia on thrombin generation, clot formation, and global hemostatic functions in an in vitro environment using a whole blood model and thromboelastography, which can recapitulate hypothermia. METHODS Blood was collected from healthy individuals through venipuncture and treated with corn trypsin inhibitor, to block the contact pathway. Coagulation was initiated with 5pM tissue factor at temperatures 37°C, 32°C, and 27°C. Reactions were quenched over time, with soluble and insoluble components analyzed for thrombin generation, fibrinogen consumption, factor (f)XIII activation, and fibrin deposition. Global coagulation potential was evaluated through thromboelastography. RESULTS Data showed that thrombin generation in samples at 37°C and 32°C had comparable rates, whereas 27°C had a much lower rate (39.2 ± 1.1 and 43 ± 2.4 nM/min vs 28.6 ± 4.4 nM/min, respectively). Fibrinogen consumption and fXIII activation were highest at 37°C, followed by 32°C and 27°C. Fibrin formation as seen through clot weights also followed this trend. Thromboelastography data showed that clot formation was fastest in samples at 37°C and lowest at 27°C. Maximum clot strength was similar for each temperature. Also, percent lysis of clots was highest at 37°C followed by 32°C and then 27°C. CONCLUSIONS Induced hypothermic conditions directly affect the rate of thrombin generation and clot formation, whereas global clot stability remains intact.