Figure 1.
High throughput 384-well plate measurement of thrombin production in human blood.
(A) Experimental protocol. Calcium and the fluorogenic thrombin substrate (SIIa) were added to the 384-well plate on a Thermo Multidrop. The plate was placed on a Perkin-Elmer Janus where various concentrations of each individual species were added to each well. After the blood was drawn, the plate was moved to a Perkin-Elmer Evolution P3 where the blood was added to all wells simultaneously (t = 0). The plate was read in a Thermo-Electron Fluoroskan where the fluorescence was measured for 4 hr. The time from vein to first measurement was under 5 min. (B) Initiation Time. The time required to reach 5% conversion of the fluorogenic substrate was set as the initiation time (Ti). This metric correlated well with ∼10 nM TAT and preceded a burst of thrombin and a maximization of the second derivative of fluorecense. Relative prolongation or reductions in Ti were used to quantify coagulation initiation.
Figure 2.
Schematic of the Platelet-Plasma model.
(A) Wiring diagram of the Platelet-Plasma model. Blue highlighted portions represent additions to the Hockin-Mann model [18]. (B) Phosphatidylserine exposure measured by fold increase in annexin V binding was obtained from published values [60] and are shown in blue circles. The maximum platelet activation state attainable at a given thrombin concentration (εmax) was obtained by fitting a hill function to this data (green line). (C) The instantaneous platelet activation status (ε) approaches its maximum attainable value (εmax) on a time scale consistent with the time it takes for the platelet to mobilize intracellular calcium. Shown are ε transient profiles at various values of εmax. (See text for complete mathematical descriptions of εmax and ε).
Table 1.
Reactions used in the Platelet-Plasma model.
Table 2.
Initial conditions of species in the Platelet-Plasma model.
Figure 3.
Validation of experimental protocol.
(A) A titration of the fluorogenic substrates Boc-VPR-MCA (blue circle) and Z-GGR-MCA (green square) with 0 added TF showed a mild inhibitory influence of Z-GGR-MCA. *s indicates statistically significant difference (p<0.05) between initiation times detected with the two different substrates. (B) A titration of the fluorogenic substrates Boc-VPR-MCA and Z-GGR-MCA with 1 pM added TF showed inhibitory influence of both substrates where Boc-VPR-MCA was found to be the better substrate to detect initiation and exhibited little inhibition at 10 µM concentration. (C) TAT formation with 0 added TF, in the absence and presence of fluorogenic substrates showed less inhibitory influence of Boc-VPR-MCA on initiation defined by a burst in TAT compared to Z-GGR-MCA. Absolute [TAT] after initiation is decreased in the presence of either substrate. (D) TAT formation with 1 pM added TF, in the absence and presence of fluorogenic substrates showed decreased [TAT] during the propagation phase of coagulation in the presence of either substrate. Initiation detected by TAT correlated well with Ti determined by our fluorogenic assay. In panels C and D, * indicates [TAT] significantly greater than baseline levels (p<0.05) and # indicates statistically significant differences compared to no substrate (blue). Experiments in panels A, B, C and D were carried out with blood from the same phlebotomy. (E) TF titration done in blood anticoagulated with CTI alone (green), Citrate + CTI (red) and Citrate alone (blue). Dashed lines indicate controls with no added TF. No significant difference was detected in titrations done with and without citrate, showing no evidence of inhibition by the anticoagulant. Effects of the contact factor pathway were apparent only below 100 fM added TF.
Figure 4.
Coagulation initiation in the absence of externally added TF.
Blood drawn into 50 µg/ml CTI and without added TF, will still reproducibly clot in ∼75 minutes. Prior activation of platelets with CVX will lower initiation time to ∼20 minutes. Shown are the multiple replicates tested under the same conditions with the same phlebotomy.
Figure 5.
Evaluating mechanisms that could lead to initiation of clotting of blood drawn into CTI without exogenous TF addition.
(A) To evaluate the effect of phlebotomy, experiments were conducted ±CVX using the first 10 mls, 10–20 mls, 20–30 mls and 30–40 mls of blood. No steady increase in Ti was noted showing that TF from phlebotomy was not leading to eventual initiation. (B) Addition of antibodies against P-selectin or Gp1bα did not prolong initiation either in the absence or presence of high dose CVX, (C) Addition of antibodies against PDI or cathepsin G did not prolong initiation either in the absence or presence of high dose CVX. (D) The ribosome inhibitor puromycin; the Clk1 kinase inhibitor Tg003; (E) antibodies against TF, VII/VIIa; or (F) VIIai did not prolong initiation either in the absence or presence of high dose CVX. This shows that initiation is unaffected by either ‘bloodborne’ or platelet synthesized TF on the time scales of our experiments.
Figure 6.
Saturation of the effects of CTI.
To evaluate the possibility of leakage past CTI, experiments were conducted with no CTI, CTI addition in well plate, CTI addition during phlebotomy and large quantities of CTI during phlebotomy as well as in well plate. The inclusion of 50 µg CTI/ml whole blood (before a 5× dilution in the well plate) produced saturating effects.
Figure 7.
Effect of anti-XI and anti-XII.
Addition of 50 µg/ml of anti-XI and anti-XII will completely prevent initiation of clotting in resting blood showing that initiation is a result of leak past saturated effects of CTI. However, on CVX activated platelets initiation is still unaffected by the presence of both CTI and these antibodies. Initial thrombin production during the propagation phase is however diminished due to abolition of thrombin feedback on FXI. Insert shows initial rates of thrombin formation in the presence of these antibodies.
Figure 8.
Prevention of initiation on CVX activated platelets.
CTI-treated regular plasma or plasma deficient in factors VII, XI or XII were supplemented with washed (plasma free) platelets. These samples were left untreated or were treated with antibodies against TF; VII; XI; XII or XI, XII and VII simultaneously; and tested for thrombin generation without exogenous TF addition after activating platelets with 25 nM CVX. Simultaneous inhibition of XI, XII and VII activity was required to completely abolish thrombin generation.
Figure 9.
Titration of TF and active proteases into blood.
(A) Effect of exogenous TF on initiation time. TF was titrated from 0.5 fM to 5 pM in 5× diluted blood. The black solid line is the simulated initiation time for the Hockin-Mann model (with SIIa) and the blue dashed line is the prediction of the Platelet-Plasma model. The light green solid line is the experimental control with no added TF. (B) Addition of prothrombinase components. Xa (red) and Va (green) was added to 5× diluted blood. The black solid line is the simulated initiation time for the Hockin-Mann model (with SIIa). The red and green dashed lines are the prediction of the Platelet-Plasma model for Xa and Va, respectively. The light green solid line is the experimental control with no added proteins. (C) Addition of VIIa, IXa and XIa. Various concentrations of VIIa (green), IXa (red) and XIa (blue) were added to blood at 5× dilution. The dashed lines of the corresponding color are simulations done with the Platelet-Plasma model. The light green solid line is the experimental control.
Figure 10.
Simulating clotting times in whole blood.
The very small reaction volumes in a 384 well plate prevent us from studying coagulation reactions in whole blood (See text). To simulate the kinetics of initiation in whole blood we simulated clotting times for additions of TF, thrombin (IIa), IXa, Xa or combinations of all 3 proteases at low and high doses reported by Butenas et al. [49] in the Mann laboratory. We found good qualitative agreement between experimental clotting times (blue) and initiation times predicted by the Platelet-Plasma model (green). The Hockin-Mann (with fluorogenic thrombin substrate, SIIa) model (red) predicts finite initiation times only in the presence of high dose TF or Xa.
Figure 11.
Simulating platelet activation.
(A) Mean substrate conversion across all replicates for the donors shown in Figure 4. Substrate conversion traces without platelet activation are shown in solid lines and conversion traces upon activated platelets are shown in dashed lines. (B) Setting the initial activation state allows us to simulate platelet activation and its dynamic effect on all platelet dependent unbinding rates (see Methods). The red line indicates simulations of substrate conversion without prior platelet activation
. The blue dashed line indicate simulations of substrate conversion upon instantaneously fully activated platelets at t = 0.
.