Platelet Activation in Heparin-Induced Thrombocytopenia is Followed by Platelet Death via Complex Apoptotic and Non-Apoptotic Pathways

Heparin-induced thrombocytopenia (HIT) is an adverse drug reaction characterized by thrombocytopenia and a high risk for venous or arterial thrombosis. HIT is caused by antibodies that recognize complexes of platelet factor 4 and heparin. The pathogenic mechanisms of this condition are not fully understood. In this study, we used flow cytometry, fluorimetry, and Western blot analysis to study the direct effects of pathogenic immune complexes containing platelet factor 4 on human platelets isolated by gel-filtration. HIT-like pathogenic immune complexes initially caused pronounced activation of platelets detected by an increased expression of phosphatidylserine and P-selectin. This activation was mediated either directly through the FcγRIIA receptors or indirectly via protease-activated receptor 1 (PAR1) receptors due to thrombin generated on or near the surface of activated platelets. The immune activation was later followed by the biochemical signs of cell death, such as mitochondrial membrane depolarization, up-regulation of Bax, down-regulation of Bcl-XL, and moderate activation of procaspase 3 and increased calpain activity. The results show that platelet activation under the action of HIT-like immune complexes is accompanied by their death through complex apoptotic and calpain-dependent non-apoptotic pathways that may underlie the low platelet count in HIT.


Introduction
Heparin-induced thrombocytopenia (HIT) is an immune-mediated adverse reaction to heparin, a broadly used anticoagulant [1]. Heparin exposure leads to the formation of IgG antibodies (Abs) that recognize multimolecular complexes of platelet factor 4 (PF4) and heparin that form in the blood and on the surface of platelets and other cells [2]. These immune complexes bind to the FcγRIIA receptors of platelets, resulting in platelet activation associated with the exposure of procoagulant phosphatidylserine (PS) on the platelet membrane and on platelet-derived microparticles [3] as well as the expression of P-selectin [4]. Alternatively, platelets are transactivated by thrombin which is generated by monocytes [5]. Irrespective of the underlying mechanism, platelet activation is central to the pathogenesis of HIT. The end result is the development of venous and arterial thromboses associated with a persistent low platelet count.

Subpopulations of Live Activated and Dying Platelets
To reproduce platelet perturbations in HIT, isolated human platelets were treated in vitro for 15 and 60 min with a combination of PF4 and pathogenic (KKO) or non-pathogenic (RTO) monoclonal anti-PF4/heparin Abs. The resulting alterations in platelet functionality were evaluated using flow cytometry. Quantification showed a slightly higher level of expression of the active integrin αIIbβ3 ( Figure 1A-D) and a significantly higher exposure of phosphatidylserine (PS) ( Figure 1E-H) under the action of pathogenic KKO and PF4 or calcium ionophore used as a positive control. Treatment of platelets with KKO/PF4 and calcium ionophore also resulted in an increased formation of PS-expressing platelet-derived microvesicles ( Figure 1I-L). No significant difference was detected when comparing the expression of active αIIbβ3, PS and microvesiculation between platelets treated with PF4 mixed with RTO, or PF4, KKO or RTO alone (not shown), similar to the untreated platelets presented in Figure 1B,D (αIIbβ3); Figure 1F,H (PS); and Figure 1J,L (microvesiculation).
Platelets treated with PF4 plus KKO also displayed increased expression of P-selectin (CD62P) on the cell membrane ( Figure 1M-P). The effect of the KKO/PF4 combination was similar to the calcium ionophore A23187, known as a powerful inducer of platelet activation [14], which also led to an increase in the fraction of CD62P-positive platelets. Compared to the negative control, pure PF4, KKO or RTO alone, as well as non-pathogenic Ab RTO with PF4, had no detectable effect on the P-selectin expression in platelets (not shown) and the dot-plots were similar to the untreated platelets presented on Figure 1N,P. Platelets were incubated for 60 min without (negative control) or with a mixture of KKO (pathogenic monoclonal anti-PF4/heparin Ab) (50 μg/mL KKO) and PF4 (10 μg/mL) or 10 μM calcium ionophore (positive control). Red squares indicate the quadrants and black circles (Q-S) indicate the gates with quantified signals. Each experiment was performed with platelets isolated from at least three independent donors. *p < 0.05, **p < 0.01, ***p < 0.001, ****p < 0.0001 compared to a corresponding negative control (Kruskal-Wallis test with Dunn's post-hoc test). Red squares indicate the quadrants and black circles (Q-S) indicate the gates with quantified signals. Each experiment was performed with platelets isolated from at least three independent donors. ** p < 0.01, *** p < 0.001, **** p < 0.0001 compared to a corresponding negative control (Kruskal-Wallis test with Dunn's post-hoc test).
Of the four parameters altered in KKO/PF4-treated platelets, as determined by flow cytometry ( Figure 1A-P), the increased expression of active αIIbβ3 and P-selectin are unequivocal biochemical signs of platelet activation and degranulation, while the two others (exposure of PS and microvesiculation) may have a dual nature: they can reflect either cellular activation [15] or apoptosis [16][17][18]. Therefore, we further aimed at studying the possible death pathways of platelets under the action of the pathogenic HIT-like Abs. First, we performed flow cytometry to measure the mitochondrial transmembrane potential (∆Ψm), a general characteristic of cell viability. It was shown that after treatment for 1 h with KKO/PF4 or calcium ionophore A23187 (positive control), the fraction of platelets with reduced or lost mitochondrial transmembrane potential was significantly increased, while the fraction of platelets with normal ∆Ψm was substantially reduced ( Figure 1Q-T). This effect was in contrast to the negative control as well as isolated KKO, PF4, RTO, or the non-pathogenic RTO + PF4 (not shown), which maintained a high ∆Ψm, similar to the untreated platelets shown in Figure 1R,T.
To confirm that the described effects of the KKO/PF4 were mediated by the platelet FcγRIIA receptor, the measurements were performed in the presence of an FcγRIIA-blocking monoclonal Ab (mAb IV.3). Importantly, suppression of FcγRIIA reduced the level of P-selectin and PS expression down to the baseline, indistinguishable from the background control values in unstimulated platelets, indicating that platelet activation with KKO/PF4 was mediated by the FcγRIIA receptor (Figure 2A,B). This was consistent with preservation of ∆Ψm ( Figure 2C) and a decrease in the subpopulation of dying platelets after blockage of the FcγRIIA receptor ( Figure 2D). Platelets treated with PF4 plus KKO also displayed increased expression of P-selectin (CD62P) on the cell membrane ( Figure 1M-P). The effect of the KKO/PF4 combination was similar to the calcium ionophore A23187, known as a powerful inducer of platelet activation [14], which also led to an increase in the fraction of CD62P-positive platelets. Compared to the negative control, pure PF4, KKO or RTO alone, as well as non-pathogenic Ab RTO with PF4, had no detectable effect on the Pselectin expression in platelets (not shown) and the dot-plots were similar to the untreated platelets presented on Figure 1N,P.
Of the four parameters altered in KKO/PF4-treated platelets, as determined by flow cytometry ( Figure 1A-P), the increased expression of active αIIbβ3 and P-selectin are unequivocal biochemical signs of platelet activation and degranulation, while the two others (exposure of PS and microvesiculation) may have a dual nature: they can reflect either cellular activation [15] or apoptosis [16][17][18]. Therefore, we further aimed at studying the possible death pathways of platelets under the action of the pathogenic HIT-like Abs. First, we performed flow cytometry to measure the mitochondrial transmembrane potential (ΔΨm), a general characteristic of cell viability. It was shown that after treatment for 1 h with KKO/PF4 or calcium ionophore A23187 (positive control), the fraction of platelets with reduced or lost mitochondrial transmembrane potential was significantly increased, while the fraction of platelets with normal ΔΨm was substantially reduced ( Figure 1Q-T). This effect was in contrast to the negative control as well as isolated KKO, PF4, RTO, or the non-pathogenic RTO + PF4 (not shown), which maintained a high ΔΨm, similar to the untreated platelets shown in Figure  1R,T.
To confirm that the described effects of the KKO/PF4 were mediated by the platelet FcγRIIA receptor, the measurements were performed in the presence of an FcγRIIA-blocking monoclonal Ab (mAb IV.3). Importantly, suppression of FcγRIIA reduced the level of P-selectin and PS expression down to the baseline, indistinguishable from the background control values in unstimulated platelets, indicating that platelet activation with KKO/PF4 was mediated by the FcγRIIA receptor (Figures 2A,B). This was consistent with preservation of ΔΨm ( Figure 2C) and a decrease in the subpopulation of dying platelets after blockage of the FcγRIIA receptor ( Figure 2D).  Interestingly, the increased expression of P-selectin induced by KKO/PF4 was prevented by the direct inhibition of protease-activated receptors 1 (PAR1) receptors in a dose-dependent manner ( Figure 3), indicating that some of the activating effects of KKO/PF4 on platelets are secondary due to thrombin generation on or near the surface of platelets primarily activated by KKO/PF4.
MitoTrackerDeepRed. (D) Subpopulations of platelets double-stained with a ΔΨm-sensitive dye MitoTrackerDeepRed and FITC-Annexin V normalized by the total number of gated platelets taken as 100%. Platelets were segregated into 2 subpopulations: MitoTracker-negative/Annexin V-positive or "dead platelets" (dark boxes) and MitoTracker-positive/Annexin V-positive or "live activated platelets" (light boxes). Each experiment was performed with platelets isolated from three independent donors. *p  0.05, ****p < 0.0001 for the corresponding data without and with the addition of Ab IV.3 (A-C, Kruskal-Wallis test; D, 2-way ANOVA test with Dunn's post-hoc test).
Interestingly, the increased expression of P-selectin induced by KKO/PF4 was prevented by the direct inhibition of protease-activated receptors 1 (PAR1) receptors in a dose-dependent manner ( Figure 3), indicating that some of the activating effects of KKO/PF4 on platelets are secondary due to thrombin generation on or near the surface of platelets primarily activated by KKO/PF4.

Figure 3.
Inhibitory effect of SCH 79797, a selective antagonist of PAR1 receptors, on platelet activation induced by the combination of KKO and PF4. Platelets were pre-incubated for 30 min at 37 °C with SCH 79797 at 0, 1, 3, and 10 μM followed by treatment for 60 min at 37 °C with KKO/PF4 (10 and 50 µ g/mL, respectively), 1 U/mL thrombin (positive control), and without any treatment (untreated, negative control). Platelet activation was assessed by the expression of P-selectin, determined using flow cytometry with fluorescently labeled anti-CD62P-Abs. Each experiment was performed with platelets isolated from three independent donors. **p < 0.01, ***p < 0.001 compared to a corresponding control without addition of SCH 79797 (2-way ANOVA test with Dunn's post-hoc test).
Thus, the pathogenic Abs combined with platelet factor 4 caused not only pronounced activation of platelets, but also induced the signs of mitochondrial dysfunction and cell death. These effects were mediated by the direct interaction of the pathogenic KKO mixed with PF4 with the FcγRIIA receptors as well as via secondary activation of PAR1 receptors by thrombin formed on the activated platelets. Next, we explored if KKO/PF4-treated platelets could undergo apoptotic death pathway.

Apoptotic Markers in Platelets induced by KKO/PF4
Although platelets have no nucleus, they can undergo apoptosis via a mitochondrial pathway that involves up-or down-regulation of mitochondrial Bcl-2 family proteins that have anti-or proapoptotic activity [19]. Since flow cytometry data suggest the possibility of apoptosis induced by the PF4-containing complexes, we used Western blot analysis to test this assumption by quantifying apoptotic molecular markers. We quantified the expression of Bax and Bcl-XL, the components of the Bcl-2 gene family that are known to be expressed in platelets [20]. Bcl-XL is an anti-apoptotic protein that modulates apoptosis by controlling mitochondrial membrane permeability and regulating the release of cytochrome c [21]. The pro-apoptotic Bax protein is involved in the regulation of apoptosis by the formation of large oligomers in the outer mitochondrial membrane, which promote cell . Inhibitory effect of SCH 79797, a selective antagonist of PAR1 receptors, on platelet activation induced by the combination of KKO and PF4. Platelets were pre-incubated for 30 min at 37 • C with SCH 79797 at 0, 1, 3, and 10 µM followed by treatment for 60 min at 37 • C with KKO/PF4 (10 and 50 µg/mL, respectively), 1 U/mL thrombin (positive control), and without any treatment (untreated, negative control). Platelet activation was assessed by the expression of P-selectin, determined using flow cytometry with fluorescently labeled anti-CD62P-Abs. Each experiment was performed with platelets isolated from three independent donors. ** p < 0.01, *** p < 0.001 compared to a corresponding control without addition of SCH 79797 (2-way ANOVA test with Dunn's post-hoc test).
Thus, the pathogenic Abs combined with platelet factor 4 caused not only pronounced activation of platelets, but also induced the signs of mitochondrial dysfunction and cell death. These effects were mediated by the direct interaction of the pathogenic KKO mixed with PF4 with the FcγRIIA receptors as well as via secondary activation of PAR1 receptors by thrombin formed on the activated platelets. Next, we explored if KKO/PF4-treated platelets could undergo apoptotic death pathway.

Apoptotic Markers in Platelets induced by KKO/PF4
Although platelets have no nucleus, they can undergo apoptosis via a mitochondrial pathway that involves up-or down-regulation of mitochondrial Bcl-2 family proteins that have anti-or pro-apoptotic activity [19]. Since flow cytometry data suggest the possibility of apoptosis induced by the PF4-containing complexes, we used Western blot analysis to test this assumption by quantifying apoptotic molecular markers. We quantified the expression of Bax and Bcl-X L , the components of the Bcl-2 gene family that are known to be expressed in platelets [20]. Bcl-X L is an anti-apoptotic protein that modulates apoptosis by controlling mitochondrial membrane permeability and regulating the release of cytochrome c [21]. The pro-apoptotic Bax protein is involved in the regulation of apoptosis by the formation of large oligomers in the outer mitochondrial membrane, which promote cell apoptosis. Bax contributes to the release of cytochrome c into the cytosol, which in turn triggers the activation of caspases [22]. Accordingly, the activation/cleavage of pro-caspase-3 is another characteristic sign of platelet apoptosis and Western blot analysis with an anti-caspase-3 Ab can show whether 32-kDa pro-caspase-3 is cleaved into a 17-kDa fragment representing active caspase-3.
Western blot analysis showed that a 17-kDa fragment representing activated caspase-3 appeared in the platelets treated with KKO/PF4 and comprised 50% ± 18% (M ± SEM) of the total pro-caspase 3 + caspase 3 levels (Figure 4). Such a band was not formed under any other experimental conditions except for the calcium ionophore A23187, which induced almost complete conversion of the pro-caspase 3 to the active cleaved caspase 3 (not shown). The expression of the anti-apoptotic protein Bcl-X L was significantly reduced, while the level of the pro-apoptotic protein Bax was substantially increased in the platelets treated with the pathogenic KKO + PF4 complexes as opposed to control untreated platelets (Figure 4).
apoptosis. Bax contributes to the release of cytochrome c into the cytosol, which in turn triggers the activation of caspases [22]. Accordingly, the activation/cleavage of pro-caspase-3 is another characteristic sign of platelet apoptosis and Western blot analysis with an anti-caspase-3 Ab can show whether 32-kDa pro-caspase-3 is cleaved into a 17-kDa fragment representing active caspase-3.
Western blot analysis showed that a 17-kDa fragment representing activated caspase-3 appeared in the platelets treated with KKO/PF4 and comprised 50% ± 18% (M ± SEM) of the total pro-caspase 3 + caspase 3 levels (Figure 4). Such a band was not formed under any other experimental conditions except for the calcium ionophore A23187, which induced almost complete conversion of the procaspase 3 to the active cleaved caspase 3 (not shown). The expression of the anti-apoptotic protein Bcl-XL was significantly reduced, while the level of the pro-apoptotic protein Bax was substantially increased in the platelets treated with the pathogenic KKO + PF4 complexes as opposed to control untreated platelets (Figure 4). Taken together, the results show that the KKO/PF4 mixture induces platelet apoptosis that is characterized by mitochondrial dysfunction combined with up-regulation of the pro-apoptotic protein Bax, down-regulation of the anti-apoptotic protein Bcl-XL, and moderate activation of procaspase-3. In view of these data, the high level of PS expression on the platelet outer membrane, as well as the formation of microvesicles, may be considered not only signs of platelet activation, but also signatures of platelet death [23,24], as in nucleated cells [25]. Therefore, the apoptotic death pathway can be a pathogenic mechanism of thrombocytopenia in HIT. Taken together, the results show that the KKO/PF4 mixture induces platelet apoptosis that is characterized by mitochondrial dysfunction combined with up-regulation of the pro-apoptotic protein Bax, down-regulation of the anti-apoptotic protein Bcl-X L , and moderate activation of procaspase-3. In view of these data, the high level of PS expression on the platelet outer membrane, as well as the formation of microvesicles, may be considered not only signs of platelet activation, but also signatures of platelet death [23,24], as in nucleated cells [25]. Therefore, the apoptotic death pathway can be a pathogenic mechanism of thrombocytopenia in HIT.

KKO/PF4-induced Calpain Activation
In addition to the caspase 3 activation, we analyzed the activity of calpains, a family of calcium-dependent cysteine proteases that have been shown to play important roles in platelet functions, such as aggregation, adhesion, spreading, and platelet-driven contraction of blood clots [26]. In addition, high calpain activity has been shown to be associated with non-apoptotic platelet death induced by thrombin [9].
The fluorimetry of platelets in the presence of a fluorogenic calpain substrate revealed that the treatment of platelets with KKO/PF4 induced the activation of calpain with a maximum activity at 180 min compared to the RTO/PF4-treated platelets ( Figure 5). The results indicate that HIT-pathogenic Abs in the presence of PF4 induce a non-apoptotic death pathway associated with calpain activation.

KKO/PF4-induced Calpain Activation
In addition to the caspase 3 activation, we analyzed the activity of calpains, a family of calciumdependent cysteine proteases that have been shown to play important roles in platelet functions, such as aggregation, adhesion, spreading, and platelet-driven contraction of blood clots [26]. In addition, high calpain activity has been shown to be associated with non-apoptotic platelet death induced by thrombin [9].
The fluorimetry of platelets in the presence of a fluorogenic calpain substrate revealed that the treatment of platelets with KKO/PF4 induced the activation of calpain with a maximum activity at 180 min compared to the RTO/PF4-treated platelets ( Figure 5). The results indicate that HITpathogenic Abs in the presence of PF4 induce a non-apoptotic death pathway associated with сalpain activation.

Figure 5.
Fluorimetric analysis of calpain activity in platelets that were incubated for 180 min with a mixture of 10 μg/mL PF4 + 50 μg/mL KKO or RTO. The numbers that characterize the calpain activity were normalized by the corresponding negative control (untreated platelets) taken as 100%. Experiments were performed with platelets isolated from seven independent donors. *p < 0.05 (Mann-Whitney U test).
In addition to the previously published paper [27], the present manuscript contains new evidence for the existence of high calpain activity associated with the apoptotic death pathway of platelets treated with KKO/PF4. Therefore, we conclude that HIT-like immune complexes cause platelet death through complex apoptotic and non-apoptotic pathways.

Platelet Isolation and Incubation with Immune Complexes Containing PF4
Platelets were freshly isolated from the blood of 55 healthy donors not taking medications that affect platelet function for at least two weeks before the blood withdrawal. Informed consent from all the blood donors was obtained. All procedures were performed in accordance with the guidelines approved by the Ethical Committee of Kazan State Medical Academy (Kazan, Russian Federation). Venous blood was collected into 3.2% trisodium citrate tubes (9:1) and immediately centrifuged at room temperature at 200g for 10 min to obtain platelet-rich plasma (PRP). Platelets from PRP were isolated by gel filtration at room temperature on Sepharose 2B equilibrated with Tyrode's buffer (4 mM HEPES, 135 mM NaCl, 2.7 mM KCl, 2.4 mM MgCl2, 5.6 mM D-glucose, 3.3 mM NaH2PO4, 0.35 mg/mL bovine serum albumin, pH 7.4). The fraction of untreated platelets activated during isolation did not exceed 10%, as determined by expression of the activated integrin αβIIb3, PS or P-selectin ( Figure 1B,F,N).
Platelets in Tyrode's buffer with the addition of 2 mM CaCl2 were incubated at 37 °C for 15, 60 and 180 min with PF4 (10 μg/mL) and HIT-like pathogenic monoclonal Ab KKO [10] at 50 μg/mL or anti-PF4 non-pathogenic monoclonal Ab RTO (50 μg/mL) [11]. In a series of inhibitory experiments; Figure 5. Fluorimetric analysis of calpain activity in platelets that were incubated for 180 min with a mixture of 10 µg/mL PF4 + 50 µg/mL KKO or RTO. The numbers that characterize the calpain activity were normalized by the corresponding negative control (untreated platelets) taken as 100%. Experiments were performed with platelets isolated from seven independent donors. * p < 0.05 (Mann-Whitney U test).
In addition to the previously published paper [27], the present manuscript contains new evidence for the existence of high calpain activity associated with the apoptotic death pathway of platelets treated with KKO/PF4. Therefore, we conclude that HIT-like immune complexes cause platelet death through complex apoptotic and non-apoptotic pathways.

Platelet Isolation and Incubation with Immune Complexes Containing PF4
Platelets were freshly isolated from the blood of 55 healthy donors not taking medications that affect platelet function for at least two weeks before the blood withdrawal. Informed consent from all the blood donors was obtained. All procedures were performed in accordance with the guidelines approved by the Ethical Committee of Kazan State Medical Academy (Kazan, Russian Federation). Venous blood was collected into 3.2% trisodium citrate tubes (9:1) and immediately centrifuged at room temperature at 200g for 10 min to obtain platelet-rich plasma (PRP). Platelets from PRP were isolated by gel filtration at room temperature on Sepharose 2B equilibrated with Tyrode's buffer (4 mM HEPES, 135 mM NaCl, 2.7 mM KCl, 2.4 mM MgCl 2 , 5.6 mM D-glucose, 3.3 mM NaH 2 PO 4 , 0.35 mg/mL bovine serum albumin, pH 7.4). The fraction of untreated platelets activated during isolation did not exceed 10%, as determined by expression of the activated integrin αβIIb3, PS or P-selectin ( Figure 1B,F,N).
Platelets in Tyrode's buffer with the addition of 2 mM CaCl 2 were incubated at 37 • C for 15, 60 and 180 min with PF4 (10 µg/mL) and HIT-like pathogenic monoclonal Ab KKO [10] at 50 µg/mL or anti-PF4 non-pathogenic monoclonal Ab RTO (50 µg/mL) [11]. In a series of inhibitory experiments; platelets were pre-treated with 50 µg/mL of a monoclonal Ab against the Fcγ receptor IIA (anti-FcγRIIA; clone IV.3) [28] or with 1, 3, and 10 µM of a selective antagonist of PAR1 receptors (SCH 79797, Sigma-Aldrich, St. Louis, MO, USA) that were added to platelets for 15 min (for anti-FcγRIIA) and 60 min (for SCH 79797) [29] prior to the incubation with the mixture of KKO and PF4. Untreated platelets and platelets incubated with either PF4, KKO or RTO alone were used as negative controls and platelets incubated with 10 µM calcium ionophore A23187 were used as a positive control. KKO, RTO and recombinant PF4 were obtained as previously described [28].

Flow Cytometry
Platelets were gated in a flow cytometer by their forward scatter/fide scatter (FSC/SCC) characteristics after size-based calibration with 1-, 2-, and 4-µm polystyrene beads ( Figure 6A,B). About 97% of events in the gate were CD41-positive ( Figure 6C). Platelet-derived microvesicles were identified as the events that reflected particles <1 µm in size as positive for the platelet-specific marker CD41. For each sample analyzed, 30,000 events were collected using a FacsCalibur flow cytometer (BD Biosciences, USA) equipped with an argon laser (λ = 488 nm) and a diode red laser (λ = 635 nm). The data were analyzed using CellQuest Pro (BD Biosciences) and FlowJo software. Specific markers on platelets were identified by labeling with PE-conjugated monoclonal Abs to platelet-specific antigen CD41, FITC-conjugated (fluorescein isothiocyanate) PAC-1 Abs against activated αIIbβ3 (BD Biosciences, Franklin Lakes, NJ, USA), and Annexin V conjugated with FITC to assess the expression of PS on the platelet membrane. Evaluation of the expression of P-selectin (CD62P) on the platelet surface was performed by the addition of murine anti-CD62P Abs labeled with a fluorescent dye phycoerythrin (PE). Changes in the mitochondrial membrane potential (∆Ψm) were assessed after adding the ∆Ψm-sensitive lipophilic cationic carbocyanine-based fluorochrome (MitoTracker Deep Red FM). The fraction of platelets fluorescently labeled by a corresponding marker was determined from the total number of platelets in the gate.
FcγRIIA; clone IV.3) [28] or with 1, 3, and 10 µ M of a selective antagonist of PAR1 receptors (SCH 79797, Sigma-Aldrich, St. Louis, MO, USA) that were added to platelets for 15 min (for anti-FcγRIIA) and 60 min (for SCH 79797) [29] prior to the incubation with the mixture of KKO and PF4. Untreated platelets and platelets incubated with either PF4, KKO or RTO alone were used as negative controls and platelets incubated with 10 µ M calcium ionophore A23187 were used as a positive control. KKO, RTO and recombinant PF4 were obtained as previously described [28].

Flow Cytometry
Platelets were gated in a flow cytometer by their forward scatter/fide scatter (FSC/SCC) characteristics after size-based calibration with 1-, 2-, and 4-μm polystyrene beads ( Figure 6A,B). About 97% of events in the gate were CD41-positive ( Figure 6C). Platelet-derived microvesicles were identified as the events that reflected particles <1 µ m in size as positive for the platelet-specific marker CD41. For each sample analyzed, 30,000 events were collected using a FacsCalibur flow cytometer (BD Biosciences, USA) equipped with an argon laser (λ = 488 nm) and a diode red laser (λ = 635 nm). The data were analyzed using CellQuest Pro (BD Biosciences) and FlowJo software. Specific markers on platelets were identified by labeling with PE-conjugated monoclonal Abs to platelet-specific antigen CD41, FITC-conjugated (fluorescein isothiocyanate) PAC-1 Abs against activated αIIbβ3 (BD Biosciences, Franklin Lakes, NJ, USA), and Annexin V conjugated with FITC to assess the expression of PS on the platelet membrane. Evaluation of the expression of P-selectin (CD62P) on the platelet surface was performed by the addition of murine anti-CD62P Abs labeled with a fluorescent dye phycoerythrin (PE). Changes in the mitochondrial membrane potential (ΔΨm) were assessed after adding the ΔΨm-sensitive lipophilic cationic carbocyanine-based fluorochrome (MitoTracker Deep Red FM). The fraction of platelets fluorescently labeled by a corresponding marker was determined from the total number of platelets in the gate.

Statistical Analysis
Statistical analyses were performed using a Prism 7.0 software package (GraphPad Software, San Diego, CA, USA). Statistical differences were estimated using the nonparametric Kruskal-Wallis test with Dunn's post-hoc test, 2-way ANOVA with Dunn's post-hoc test after checking the data for normality using Shapiro-Wilk and KS normality criteria, and Mann-Whitney U test for a pairwise comparison with a 95% level of confidence. The results are presented as a median and intervals between the 25th and 75th, as well as between the 5th and 95 th , percentiles.

Conclusions
Platelet activation induced by pathogenic complexes proceeds directly through the FcγRIIA receptors or indirectly via PAR1 receptors. The effect of KKO/PF4 is accompanied by platelet dysfunction and death via complex apoptotic and calpain-dependent, non-apoptotic pathways. These platelet death pathways may comprise pathogenic mechanisms leading to the persistent low platelet count in HIT.

Conflicts of Interest:
The authors declare that they have no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.