Evidence Suggestive of Activation of the Intrinsic Pathway of Blood Coagulation After Injection of Factor Xa/Phospholipid Into Rabbits
Abstract The present study was carried out to extend an earlier observation from this laboratory that mean plasma factor X levels fell by about 15% after the injection into rabbits of a formed factor Xa/phospholipid complex that caused only minimal intravascular coagulation. We have now injected rabbits with formulations of factor Xa/phospholipid that caused considerable intravascular coagulation, as documented by substantial falls in fibrinogen, factor V, and factor VIII and a fall in plasma prothrombin activity of about 15% to 20% of the initial level. Mean plasma factor X activity fell by about 30% of the initial level. Factors participating in the intrinsic coagulation pathway—XII, XI, and IX—all fell by about 50% after injection of a complex made with 16.3 pmol factor Xa and 80 nmol phospholipid per 1 kg body wt and by about 35% after injection of a complex made with 32.6 pmol factor Xa and 40 nmol phospholipid per 1 kg body wt. In contrast, total plasma factor VII activity did not change, and specific plasma factor VIIa levels, which were lower than those measured in human plasma, did not rise after injection of factor Xa/phospholipid. The data are compatible with the hypothesis that factor Xa/phospholipid–induced generation of thrombin in vivo leads to factor XII–dependent activation of the intrinsic pathway of coagulation that results in significant activation of factor X. Further testing of this hypothesis appears warranted.
- Received August 12, 1994.
- Accepted October 24, 1994.
Injection of a mixture of factor Xa and procoagulant phospholipid can induce disseminated intravascular coagulation in rabbits, dogs, and nonhuman primates1 2 3 4 ; shorten the cuticle bleeding time of factor VIII–deficient hemophilic dogs3 ; and cause platelets to be sequestered transiently in the microvasculature of rats and rabbits.5 These phenomena can be accounted for by the availability of an injected factor Xa/phospholipid complex to provide a surface for the rapid intravascular assembly of the factor Xa/factor Va/phospholipid complex (prothrombinase) that catalyzes the conversion of prothrombin to thrombin.
In an earlier study,6 we reported that immunodepleting rabbits of tissue factor pathway inhibitor (TFPI) failed to potentiate the minimal degree of intravascular coagulation induced in control animals by a low dose of a formed factor Xa/phospholipid complex. Because factor Xa/phospholipid–induced coagulation should bypass the activation of factor X, it surprised us in this study to observe a mean fall of plasma factor X activity of about 15% of the preinjection value in both control rabbits and rabbits immunodepleted of TFPI.
Our earlier findings are compatible with the hypothesis that factor Xa/phospholipid–triggered coagulation can initiate a reaction or reactions in vivo that activate a supplemental mechanism for initiating coagulation. As one way of testing this hypothesis, we have now measured the effect of the injection of a factor Xa/phospholipid complex on plasma levels of the clotting factors known to be required for activation of factor X in both the intrinsic and tissue factor (TF)–dependent pathways of coagulation.
Phospholipids from bovine brain, bovine serum albumin (BSA), and rabbit brain thromboplastin were from Sigma Chemical Co, bovine thrombin and fibrinogen calibration standards from Baxter Diagnostics Inc, and activated partial thromboplastin (aPTT) reagents from Organon Teknika or Helena Laboratories. Rabbit albumin and octyl glucoside were purchased from Calbiochem. Chromozym X was obtained from Boehringer Mannheim, and radioisotopes were purchased from Amersham. Other chemicals, reagent grade or better, were purchased from Sigma or from Fisher Scientific.
Human factors IX and X were purified from plasma as described.7 Factor Xa was prepared by activating factor X with insolubilized Russell’s viper venom.7 Tritiated human factors IX and X were prepared by the general technique of van Lenten and Ashwell,8 as described previously.9
Rabbit factor XI was isolated from purchased pooled rabbit plasma (Pel Freeze Biologicals) essentially as described.10 Rabbit factor XIa was prepared by incubation of the purified protein with insolubilized trypsin as described.11 A sample of the purified factor XI was radiolabeled with Na(125I) using Iodogen according to the manufacturer’s bulletin (Pierce). The purified protein was also used as antigen to prepare a specific neutralizing goat anti-rabbit factor XI IgG.
Mixed phospholipid vesicles containing 60% phosphatidylcholine and 40% phosphatidylserine (PCPS) were prepared by dialysis using octyl glucoside as detergent.12
Complexes of Factor Xa and PCPS
Factor Xa was added to a mixture in a final 1.2-mL volume of sterile saline that contained PCPS vesicles, rabbit albumin 25 μg/mL, and CaCl2 5 mmol/L. The amounts of factor Xa and PCPS were varied to yield two formulations: (1) complexes made with 32.6 pmol factor Xa and 40 nmol PCPS per 1 kg rabbit body wt and (2) complexes made with 16.3 pmol factor Xa and 80 nmol PCPS per 1 kg rabbit body wt. Control mixtures were also prepared to contain 16.3 pmol Xa per 1 kg rabbit body wt without phospholipid vesicles or 80 nmol PCPS per 1 kg rabbit body wt without factor Xa. All mixtures were kept for 5 to 10 minutes at room temperature before injection.
Substrate Plasmas for Coagulation Assays
Human factor–deficient substrate plasmas were collected from human donors with specific hereditary coagulation factor deficiencies or were prepared by immunodepleting normal plasma obtained from volunteers. Rabbit plasmas deficient in factor IX, X, or VII were prepared by immunodepleting plasma freshly obtained from healthy rabbits, as described previously.13 Factor XI–deficient rabbit plasma was prepared by immunoabsorbing rabbit plasma with immobilized goat anti-rabbit factor XI IgG.
One-Stage Clotting Assays
An automated coagulometer was used for clotting factor assays (MLA Electra 700 or Lancer Coagulizer II). Test plasmas were diluted in cold Tris-buffered saline (TBS; 0.05 mol/L Tris-HCl, pH 7.5, 0.15 mol/L NaCl) containing 1 mg/mL BSA (TBS/BSA) and were clotted in duplicate. A pooled rabbit plasma prepared from plasma obtained from six or seven healthy young rabbits was used as the reference plasma. It was arbitrarily assigned a value of 1 U/mL for all assay procedures.
Fibrinogen was measured by the method of Clauss.14 Prothrombin was determined in a one-stage clotting assay by using a rabbit prothrombin-deficient substrate prepared by mixing equal parts of rabbit serum and barium-adsorbed rabbit plasma and a 1/20 dilution of the rabbit test plasma. Factor V was measured in a one-stage clotting assay using a human factor V–deficient substrate plasma, with rabbit brain thromboplastin as the activator, and a 1/200 dilution of the rabbit test plasma. Factor VIII was measured in a one-stage aPTT assay using a human factor VIII–deficient substrate plasma and a 1/25 dilution of the rabbit test plasma. Details of these assays were provided earlier.13
Factor X was measured in a one-stage assay in which 100 μL of an equal-part mixture of rabbit plasma immunodepleted of factor X and barium-adsorbed rabbit plasma was incubated for 3 minutes at 37°C with 100 μL of a 1/20 dilution of the rabbit plasma test sample and 100 μL of rabbit brain thromboplastin, and clotting was triggered by the addition of 100 μL of 35 mmol/L CaCl2. Factor IX was measured in a one-stage aPTT assay by using an equal-part mixture of human factor IX–deficient substrate plasma and barium-adsorbed rabbit plasma or an equal-part mixture of rabbit plasma immunodepleted of factor IX and barium-adsorbed ox plasma as the substrate. The substrate plasma (100 μL) was incubated with 50 μL of a 1/15 dilution of rabbit test plasma and 100 μL aPTT reagent for 5 minutes at 37°C, and clotting was triggered by 100 μL of 30 mmol/L CaCl2. Factor XI, factor XII, and high-molecular-weight kininogen (HMWK) were measured in one-stage aPTT assays in which equal parts of human deficient substrate plasmas, dilutions of the rabbit test plasmas (1/25 for the factor XI assay, 1/15 for the factor XII assay, and 1/20 for the HMWK assay), and aPTT reagent were incubated at 37°C for 5 minutes and clotted with an equal part of 35 mmol/L CaCl2. Factor XI levels were also measured on some samples with a rabbit plasma immunodepleted of factor XI as the substrate. The latter and the HMWK assays were clotted manually.
Factor VII levels were determined in a one-stage clotting assay by using an equal-part mixture of rabbit plasma immunodepleted of factor VII and barium-adsorbed rabbit plasma, a 1/200 dilution of rabbit test plasma, and rabbit brain thromboplastin. Details were provided earlier.13 Factor VIIa levels in rabbit plasma were determined in an assay by using a recombinant truncated soluble human TF preparation15 kindly provided by Dr James Morrissey (Oklahoma Medical Research Foundation). The assay procedure was as described by Wildgoose et al,16 except that the soluble TF concentration in the TF/calcium reagent was increased to 100 nmol/L. Clotting times were converted to factor VIIa concentration from a reference curve prepared with dilutions of from 0.6 pmol/L to 0.6 nmol/L of recombinant human factor VIIa (Novo-Nordisk). In this modified assay, a mean value of 30±22 pmol/L of factor VIIa was obtained for samples from 10 healthy human subjects. Repeated assay of the same pooled plasma from six rabbits gave a factor VIIa that varied from 4 to 10 pmol/L. Four plasma samples obtained over 45 minutes from one healthy rabbit each gave a value of 7 pmol/L factor VIIa activity.
TFPI Activity Assay
TFPI was measured in a two-stage capacity assay in which the ability of residual factor VIIa/TF to catalyze activation of factor X was determined after an initial 30-minute incubation of a dilute rabbit plasma test sample with a limiting concentration of purified rabbit TF,17 a saturating concentration of factor VIIa, a low concentration of factor Xa, and CaCl2. Factor X was then added, and its activation was measured in an amidolytic assay. Details of the assay were described earlier.6
Hematocrit, white blood cell (WBC) counts, and platelet counts were determined with a Coulter ST counter (Coulter Electronics Inc) in the hematology laboratory of the University of California, San Diego, Medical Center.
Female New Zealand rabbits (≈2 kg) were used for these studies in protocols approved by the Animal Subjects Committee of the University of California, San Diego. Rabbits were given a bolus injection into a marginal ear vein of 1 mL sterile saline containing a formed factor Xa/PCPS complex, factor Xa alone, or PCPS alone. The Table⇓ summarizes the number of rabbits providing data for analysis in each treatment group.
Blood samples were obtained from the marginal ear veins of the opposite ear by a technique described in detail earlier13 in which a 23-gauge needle is inserted and the drops of blood are collected into a tube containing a buffered citrate anticoagulant. Samples (1.5 mL for coagulation factor assays and 0.5 mL for measurement of hematocrit, WBC count, and platelet count) were obtained before injection and at time points thereafter for up to 90 minutes. Plasma samples were prepared without delay and stored frozen at −80°C for subsequent coagulation factor analyses, as described earlier.13
Rectal temperature, measured as a sensitive indicator of exposure to material containing endotoxin, did not change significantly in any rabbit.
In Vitro Activation Studies
Reaction mixtures containing 50% rabbit plasma or buffer (TBS/BSA), 32.6 or 65.2 nmol/L factor Xa, 160 or 80 μmol/L PCPS, 10 mmol/L CaCl2, and 5 μg/mL of either 3H-human factor IX or 3H-human factor X were incubated at 37°C in plastic tubes, and subsamples removed over 120 minutes were monitored for activation peptide release.11 In positive control mixtures, rabbit factor XIa (2.9 nmol/L), or factor VIIa/TF complexes made with 0.5 nmol/L rVIIa, and 0.1 nmol/L rabbit TF were substituted for factor Xa/PCPS.
Activation of factor XI was examined by incubating 125I-rabbit factor XI (41 nmol/L) at 37°C with 32.6 nmol/L factor Xa, 160 μmol/L PCPS, and 10 mmol/L CaCl2 in 80% rabbit plasma immunodepleted of factor XI or in TBS/BSA buffer supplemented with rabbit HMWK (60 nmol/L). After 30 minutes, samples were added to SDS-PAGE sample buffer containing mercaptoethanol, subjected to reduced SDS-PAGE, and examined by radioautography for evidence of molecular cleavage of the radiolabeled factor XI. In positive control activation mixtures, factor Xa/PCPS was replaced: in a buffer system with 25 nmol/L trypsin or 5 U/mL thrombin plus dextran sulfate and in a plasma system by kaolin powder (40 mg/mL).
Values for clotting factors, TFPI, and hematologic parameters are reported for each group as mean±SEM and are expressed as a percentage of the value determined in the sample obtained before injection of the test material (zero time sample).
Evidence of Factor Xa/PCPS–Induced Intravascular Coagulation
Intravenous injection of complexes containing factor Xa and PCPS triggered an immediate burst of intravascular coagulation that appeared to cease after about 10 minutes. Thus, plasma levels of fibrinogen, factor V, factor VIII, and prothrombin had already fallen substantially in a sample drawn within 5 minutes after injection, fell further within 10 minutes after injection, and failed to change significantly in samples drawn over the next 50 minutes (Fig 1⇓). The formulation containing increased PCPS but less factor Xa (80 nmol/kg PCPS with 16.3 pmol/kg factor Xa) initiated a greater fall in factor V, factor VIII, and fibrinogen than did the combination containing less PCPS and more factor Xa (40 nmol/kg PCPS with 32.6 pmol/kg factor Xa). In contrast, mean prothrombin levels fell to the same extent in rabbits receiving each of the two formulations (Fig 1C⇓).
Control rabbits received an intravenous injection of either 80 nmol/kg PCPS alone or 16.3 pmol/kg factor Xa alone. The data from control groups were virtually identical; therefore, the control data have been plotted as a single line in Figs 1⇑ and 2⇓. For clarity of presentation, when the same mean data points were obtained for two animal groups and were therefore plotted along a single line, the point symbols for each group were alternated along the line (eg, see the plot of the control rabbits in Fig 1C⇑ and the experimental groups in Fig 2B⇓).
Factor Xa/PCPS–Induced Falls in Mean Plasma Factor X, XII, XI, and IX Activities
Injection of factor Xa/PCPS resulted in a persistent fall in mean plasma factor X activity to about 70% of the value before injection. The extent of the fall was the same for the two formulations (Fig 2B⇑). Plasma factor XII, XI, and IX activities also fell after injection of factor Xa/PCPS. Mean activities fell by approximately 50% of initial levels in the animals receiving the complexes formed with 16.3 pmol factor Xa/80 nmol PCPS per 1 kg body wt and by about 30% to 40% in rabbits receiving complexes formed with 32.6 pmol factor Xa/40 nmol PCPS per 1 kg body wt (Fig 2A⇑, 2C⇑, and 2D⇑). A fall in HMWK coagulant activity of 35% was also found on assay of serial plasma samples from one rabbit receiving the latter formulation.
The values for factors IX, XI, and XII in Fig 2⇑ were obtained in assays using human deficient substrates and between 1/15 and 1/25 dilutions of rabbit test plasma. Rabbit plasma factor V and VIII activities are higher than human plasma factor V and VIII activities, and factor V and VIII activities fell significantly in animals given 16.3 pmol factor Xa/80 nmol PCPS per 1 kg body wt. Therefore, assays for factor IX and factor XI were repeated using rabbit immunodepleted substrates on plasma samples from one rabbit given each formulation. Similar data were obtained with either human or rabbit substrates. For example, the data for a rabbit given 16.3 pmol factor Xa/80 nmol PCPS per 1 kg body wt were as follows: for factor XI, a fall after injection of the complex to about 50%, as measured with either human or rabbit factor XI–deficient substrate; for factor IX, a fall to about 40%, as measured with human factor IX–deficient substrate, and to about 30%, as measured with rabbit factor IX–deficient substrate. These data validated for us the specificity of assay results with human substrates and 1/15 to 1/25 dilutions of rabbit test plasma samples.
Effects of Administration of Factor Xa/PCPS on Factor VII
Factor VII activity was measured in plasma samples from two rabbits given the more procoagulant dose of factor Xa/PCPS (16.3 pmol factor Xa/80 nmol PCPS per 1 kg body wt). In contrast to the falls observed in the intrinsic pathway factors, plasma factor VII activity did not change after the injection of factor Xa/PCPS (Fig 3⇓).
A specific assay for VIIa activity independent of native plasma factor VII15 16 was used to measure plasma factor VIIa levels in sequential samples from two other rabbits given 16.3 pmol/kg factor Xa and 80 nmol/kg PCPS. The low basal levels of plasma factor VIIa fell further after the injection: in one rabbit, from 8 pmol/L before the injection to less than 3 pmol/L; in the second rabbit, from 10 to 6 pmol/L. Similar results were obtained when samples from one rabbit were assayed in a modified assay in which additional factor V was supplied by diluting the test samples in adsorbed ox plasma. In a further experiment using this modified assay, factor VIIa activity was also found to fall in plasma samples from a rabbit in which intravascular coagulation was induced by a 4-hour infusion of 10.6 pmol/kg of reconstituted purified rabbit TF (data not shown).
Effects of Administration of Factor Xa/PCPS on Plasma TFPI Activity
TFPI activities were measured in sequential plasma samples obtained from rabbits receiving each formulation of factor Xa/PCPS and for control rabbits receiving only factor Xa alone or PCPS alone (Fig 5⇓). In confirmation of our previous data,6 plasma TFPI activity rose by about 25% immediately after the injection of factor Xa/PCPS and then fell progressively in later samples. In contrast, values for plasma TFPI did not change following the injection of either factor Xa or PCPS alone (Fig 5⇓).
Effects of Injection of Factor Xa/PCPS on Platelet and WBC Counts
Mean platelet and WBC counts fell to about two thirds of the starting values at 10 minutes after the injection of the more procoagulant formulation of factor Xa/PCPS (16.3 pmol factor Xa/80 nmol PCPS per 1 kg body wt) and then began to rise gradually (Fig 4⇓). In contrast, platelets fell only minimally in rabbits given the less procoagulant formulation of factor Xa/PCPS, and WBC counts did not differ from those of the controls.
Evaluation of Ability of Factor Xa/PCPS Complexes to Initiate Measurable Activation of Factor XI, IX, or X In Vitro
125I–rabbit factor XI, added either to rabbit plasma immunodepleted of factor XI or to TBS/BSA buffer containing rabbit HMWK, was not cleaved after 30 minutes of incubation with a mixture that contained final concentrations of 32.6 nmol/L factor Xa, 160 μmol/L PCPS, and 10 mmol/L CaCl2. Autoradiographs of reduced SDS gels of samples removed after incubation revealed no change in the radioactivity profile of the 125I–factor XI. In contrast, positive control mixtures in which 125I–factor XI was incubated with kaolin in plasma or with trypsin or thrombin/dextran in buffer revealed molecular cleavage of the 125I–rabbit factor XI into the heavy and light chains of the activated molecule.
3H–human factor IX or X (5 μg/mL) was incubated with either TBS/BSA buffer or rabbit plasma immunodepleted of rabbit factor IX or rabbit factor X containing final concentrations of 10 mmol/L CaCl2 alone (control) or 10 mmol/L CaCl2 with either 32.6 nmol/L factor Xa and 160 μmol/L PCPS or 65.2 nmol/L factor Xa and 80 μmol/L PCPS. For 3H–factor IX in the buffer system at 120 minutes, data were as follows: In the mixture containing 65.2 nmol/L factor Xa and 80 μmol/L PCPS, activation peptide release of about 10% trichloracetic acid (TCA)–soluble counts over the control; in the mixture containing 32.6 nmol/L factor Xa and 160 μmol/L PCPS, activation peptide release of about 6% TCA–soluble counts over the control. For 3H–factor IX in the plasma system at 120 minutes, there was 4% release of TCA-soluble counts for both experimental mixtures, which was the same percent release of TCA-soluble counts as measured in the buffer system for the control mixture. Similar data were obtained with 3H–factor X, ie, evidence of molecular cleavage in mixtures containing factor Xa/PCPS in the buffer system but no evidence of molecular cleavage in the presence of 50% plasma.
The experiments described were prompted by an earlier observation of an unexpected fall in mean plasma factor X activity of about 15% of initial activity in rabbits given an intravenous injection of a factor Xa/PCPS complex that produced only minimal evidence of intravascular coagulation.6 In the present study, more procoagulant formulations of factor Xa/PCPS complexes were used in the expectation that they might induce greater falls in plasma factor X activity. One formulation, a complex made by doubling the amount of factor Xa used in the earlier study to 32.6 pmol/kg and of PCPS to 40 nmol/kg, induced modest falls in plasma fibrinogen, factor V, and factor VIII. The other, a complex made with the same amount of factor Xa as in our earlier study (16.3 pmol/kg) but with an increase of PCPS to 80 nmol/kg, induced more extensive falls in plasma fibrinogen, factor V, and factor VIII (Fig 1⇑). The intravenous injection of either of these complexes caused mean plasma factor X activity to fall by about 30% of the initial level (Fig 2B⇑) compared with the fall of about 15% of the initial level observed in the earlier study.
An accelerated clearance of plasma factor X bound to phospholipid was ruled out as a cause for the fall because factor X activity failed to fall in control animals infused with 80 nmol/kg PCPS in the absence of factor Xa (Fig 2B⇑). Each factor Xa/PCPS formulation caused a similar 15% to 20% fall in plasma prothrombin activity, which exceeded the barely measurable fall in mean prothrombin activity observed in the earlier study.6 If one assumes that the similar falls in plasma prothrombin activity obtained with each formulation mean that each formulation catalyzed generation of similar amounts of thrombin, then the greater falls in fibrinogen, factor V, and factor VIII levels observed after injection of the complexes made with a higher amount of PCPS (Fig 1⇑) remain unexplained. Nevertheless, we felt overall that the data were compatible with the hypothesis that factor Xa/PCPS–induced generation of thrombin in vivo initiates the subsequent events that cause plasma factor X activity to fall.
Therefore, in additional experiments we tested the ability of a factor Xa/PCPS complex to activate factors XI, IX, and X in vitro. Molecular cleavage indicative of activation could not be demonstrated when 125I–factor XI was incubated with 32.6 nmol/L factor Xa, 160 μ/L PCPS, and calcium ions either in 80% rabbit plasma immunodepleted of factor XI or in buffer. Moreover, molecular cleavage of factor IX or X could not be demonstrated by an increase in TCA-soluble counts when rabbit plasma containing added human 3H–factor IX or X was incubated with a concentration of factor Xa and PCPS calculated to be at least in 100-fold excess over the plasma concentration expected after injection of 1 mL of a mixture containing 32.6 pmol factor Xa, 160 nmol PCPS, and 10 mmol/L CaCl2.
In vivo experiments were also carried out so that we could look for indirect evidence that factor Xa/PCPS–catalyzed generation of thrombin in vivo might trigger feedback activation of the coagulation reactions that led to consumption of factor X. Because plasma factor X fell without delay after injection of factor Xa/PCPS (Fig 2B⇑) and because in our earlier study the fall was not enhanced in rabbits immunodepleted of TFPI,6 the possibility of feedback activation stemming from a thrombin-induced generation of intravascular TF could be ruled out. Therefore, we focused primarily on measuring the effect of injection of factor Xa/PCPS on plasma activities of factor XII, HMWK, factor XI, and factor IX, assuming that falls in their levels would represent indirect evidence of feedback activation of the intrinsic pathway of coagulation.
Each of these plasma activities fell promptly after injection of factor Xa/PCPS, and reduced activities persisted throughout the experimental period (Fig 2⇑). As with the falls in fibrinogen, factor V, and factor VIII (Fig 1⇑), the falls in activity of factors XII, XI, and IX were more pronounced after injection of the formulation with less factor Xa but more PCPS (Fig 2A⇑, 2C⇑, and 2D⇑).
It is recognized that alternative explanations for these data have not been ruled out, eg, the possibility that proteolytic enzymes released from WBC after generation of thrombin18 could have degraded factor XII, HMWK, factor XI, and factor IX, with a resultant fall in their activities. It is reassuring in this regard that when complexes made with 32.6 pmol factor Xa and 40 nmol PCPS/kg were injected, WBC counts did not fall beyond the minimal fall observed in control animals (Fig 5⇓).
Nevertheless, one must note that in a preliminary communication,19 a marked fall in factor IX activity after injection into chimpanzees of a higher dose of factor Xa/PCPS has been attributed to proteolysis of factor IX secondary to the release of elastase from WBC. However, in that communication, factor XI activity was reported not to fall, which differs from the data reported here. In a second preliminary communication,20 the same investigative group reported that plasma factor VIIa increased by 10-fold after injection into chimpanzees of a lower dose of factor Xa/PCPS, a dose similar to the less procoagulant formulation used in our study. In our experiments, total plasma factor VII coagulant activity did not change after injection of factor Xa/PCPS into rabbits (Fig 3⇑). Moreover, the low basal levels of plasma factor VIIa (compared with human plasma factor VIIa) did not rise but were reduced further in serial plasma samples obtained from two rabbits injected with 16.3 pmol factor Xa and 80 nmol PCPS/kg.
Why our results in the rabbit differ from those reported in chimpanzees is not clear. Possibilities include a difference in the response to an injection of factor Xa/PCPS in different species, in the preparations of factor Xa and PCPS used, and in the stability of activated factor VII in plasma of different species.
In our earlier study,6 plasma TFPI activity promptly increased and then gradually fell when rabbits were injected with a factor Xa/PCPS preparation that induced minimal intravascular coagulation. The same pattern was observed after injection of the more procoagulant factor Xa/PCPS formulations used in the present study (Fig 5⇑). The findings are reminiscent of the transient increase of plasma TFPI activity observed after rabbits were given an intravenous injection of heparin,21 which has been attributed to release of TFPI from glycosaminoglycans on the luminal surface of vascular endothelium.
If one accepts that the present data represent evidence that a factor Xa/PCPS–induced intravascular coagulation can lead to back-activation of factor XII and consequent triggering of the intrinsic pathway of coagulation, then additional comments are pertinent. Such back-activation of factor XII is highly unlikely to be unique to factor Xa/PCPS–induced intravascular coagulation. Indeed, whereas it is now abundantly clear that endotoxin-induced intravascular coagulation is TF-dependent,22 23 24 it is less well appreciated that rabbit factor XII levels also fall after intravenous injection of endotoxin. It was shown many years ago that treatment with warfarin prevented not only intravascular coagulation after endotoxin but also the fall in plasma factor XII activity.25 26 If TF-induced intravascular coagulation can also lead to the back-activation and consequent consumption of plasma factor XII, then this previously puzzling observation is understandable.
Finally, the lack of clinical bleeding in patients with hereditary factor XII deficiency establishes that supplementing TF-induced coagulation by a factor XII–dependent back-activation of the intrinsic pathway of coagulation is not required for normal hemostasis. However, one should not conclude that reinforcing intravascular coagulation induced by other mechanisms as a consequence of a feedback activation of factor XII that initiates the intrinsic coagulation reactions would lack significance for the pathogenesis of thrombotic disorders. If further data confirm the hypothesis that the intravascular generation of thrombin can lead to the activation of factor XII, then studies of its pathophysiological significance in animal models of thrombosis should be of considerable interest.
This work was supported by grant HL-27234 from the National Heart, Lung, and Blood Institute. We thank Dr James Morrissey of the Oklahoma Medical Research Foundation for a generous gift of recombinant human soluble tissue factor. Steve Maki, Thomas Toneff, and Kathy Donnelly provided technical assistance in these experiments.
Giles AR, Nesheim ME, Hoogendoorn H, Tracy PB, Mann KG. The coagulant-active phospholipid content is a major determinant of in vivo thrombogenicity of prothrombin complex (factor IX) concentrates in rabbits. Blood. 1982;59:401-407.
Giles AR, Nesheim ME, Mann KG. Studies of factors V and VIII: C in an animal model of disseminated intravascular coagulation. J Clin Invest. 1984;74:2219-2225.
Hoogendoorn H, Nesheim ME, Giles AR. A qualitative and quantitative analysis of the activation and inactivation of protein C in vivo in a primate model. Blood. 1990;75:2164-2171.
Richardson M, Toh CH, Tinlin S, Giles AR. Platelet sequestration in the pulmonary microcirculation in the rat and rabbit following thrombin generation in vivo. Thromb Haemost. 1993;69:abstract 79.
Warn-Cramer BJ, Rapaport SI. Studies of factor Xa/phospholipid–induced intravascular coagulation in rabbits: effects of immunodepletion of tissue factor pathway inhibitor. Arterioscler Thromb. 1993;13:1551-1557.
van Lenten L, Ashwell G. Studies on the chemical and enzymatic modification of glycoproteins: a general method for the tritiation of sialic acid-containing glycoproteins. J Biol Chem. 1971;246:1889-1894.
Usharani P, Warn-Cramer BJ, Kasper CK, Bajaj SP. Characterization of three abnormal factor IX variants (Bm Lake Elsinore, Long Beach, and Los Angeles) of hemophilia B. J Clin Invest. 1985;75:76-83.
Bajaj SP. Cooperative 2+Ca binding to human factor IX. J Biol Chem. 1982;257:4127-4132.
Zivelin A, Rao LVM, Rapaport SI. Mechanism of the anticoagulant effect of warfarin as evaluated in rabbits by selective depression of individual procoagulant vitamin K-dependent clotting factors. J Clin Invest. 1993;92:2131-2140.
Morrissey JH, Macik BG, Neuenschwander PF, Comp PC. Quantitation of activated factor VII levels in plasma using a tissue factor mutant selectively deficient in promoting factor VII activation. Blood. 1993;81:734-744.
Wildgoose P, Nemerson Y, Hansen LL, Nielsen FE, Glazer S, Hedner U. Measurement of basal levels of factor VIIa in hemophilia A and B patients. Blood. 1992;80:25-28.
Plow EF. Leukocyte elastase release during blood coagulation: a potential mechanism for activation of the alternative fibrinolytic pathway. J Clin Invest. 1982;69:564-572.
Hoogendoorn H, Weitz JI, Giles AR. Evidence for the generation of elastase activity in a primate model of disseminated intravascular coagulation (DIC). Thromb Haemost. 1993;69:2088. Abstract.
Giles AR, Hoogendoorn H, Tinlin S, Morrissey JH. Activation of F. VII in vivo following the infusion of a combination of F. Xa and phosphatidylcholine/phosphatidylserine (PCPS) vesicles at a dosage that bypasses F. VIII and normalizes hemostatic plug formation in hemophilic dogs. Thromb Haemost. 1993;69:2739. Abstract.
Warr TA, Rao LVM, Rapaport SI. Disseminated intravascular coagulation in rabbits induced by administration of endotoxin or tissue factor: effect of anti-tissue factor antibodies and measurement of plasma extrinsic pathway inhibitor activity. Blood. 1990;75:1481-1489.
Sandset PM, Warn-Cramer BJ, Rao LVM, Maki SL, Rapaport SI. Depletion of extrinsic pathway inhibitor (EPI) sensitizes rabbits to disseminated intravascular coagulation induced with tissue factor: evidence supporting a physiologic role for EPI as a natural anticoagulant. Proc Natl Acad Sci U S A. 1991;88:708-712.
Lerner RG, Rapaport SI, Spitzer JM. Endotoxin-induced intravascular clotting: the need for granulocytes. Thromb Haemost. 1968;20:430-443.