ORIGINAL ARTICLE
ATP-binding cassette transporter 1 (ABCA1) deficiency
decreases platelet reactivity and reduces thromboxane A2
production independently of hematopoietic ABCA1
T . L H E R M U S I E R , *† S . S E V E R I N , * J . V A N R O T H E M , * † C . G A R C I A , ‡ J . B E R T R A N D - M I C H E L , § P . L E F A O U D E R ,§ B . H E C H L E R , ¶ C . B R O C C A R D O , * * P . C O U V E R T , † † G . C H I M I N I , ‡ ‡ P . S I E* ‡ and B . P A Y R A S T R E *‡
*Inserm U1048 et Universite Toulouse 3, I2MC; †Departement de Cardiologie, CHU de Toulouse; ‡Laboratoire d’Hematologie, CHU de Toulouse;§I2MC, Lipidomic Core Facility-MetaToul, Toulouse; ¶Inserm UMR-S949, Etablissement Francßais du Sang-Alsace, Strasbourg; **Inserm, U1037, CRCT, Toulouse;††Service de Biochimie Endocrinienne et Oncologique, H^opital Pitie Salp^etriere, Paris; and ‡‡CIML, Parc Scientifique de Luminy, Case 906, Marseille Luminy, France
To cite this article: Lhermusier T, Severin S, Van Rothem J, Garcia C, Bertrand-Michel J, Le Faouder P, Hechler B, Broccardo C, Couvert P, Chi-mini G, Sie P, Payrastre B. ATP-binding cassette transporter 1 (ABCA1) deficiency decreases platelet reactivity and reduces thromboxane A2 production independently of hematopoietic ABCA1. J Thromb Haemost 2016; 14: 585–95.
Essentials
• The role of ATP-binding cassette transporter 1 (ABCA1) in platelet functions is poorly characterized. • We studied the impact of ABCA1 deficiency on platelet
responses in a mouse model and two Tangier patients. • ABCA1-deficient platelets exhibit reduced positive
feed-back loop mechanisms.
• This reduced reactivity is dependent on external envir-onment and independent of hematopoietic ABCA1. Summary. Background: The ATP-binding cassette trans-porter ABCA1 is required for the conversion of apolipoprotein A-1 to high-density lipoprotein (HDL), and its defect causes Tangier disease, a rare disorder char-acterized by an absence of HDL and accumulation of cholesterol in peripheral tissues. The role of ABCA1 in platelet functions remains poorly characterized. Objec-tive:To determine the role of ABCA1 in platelet func-tions and to clarify controversies concerning its implication in processes as fundamental as platelet phos-phatidylserine exposure and control of platelet membrane lipid composition. Methods and results: We studied the impact of ABCA1 deficiency on platelet responses in a mouse model and in two Tangier patients. We show that
platelets in ABCA1-deficient mice are slightly larger in size and exhibit aggregation and secretion defects in response to low concentrations of thrombin and collagen. These platelets have normal cholesterol and major phos-pholipid composition, granule morphology, or calcium-induced phosphatidylserine exposure. Interestingly, ABCA1-deficient platelets display a reduction in positive feedback loop mechanisms, particularly in thromboxane A2 (TXA2) production. Hematopoietic chimera mice demonstrated that defective eicosanoids production, par-ticularly TXA2, was primarily dependent on external environment and not on the hematopoietic ABCA1. Decreased aggregation and production of TXA2 and eicosanoids were also observed in platelets from Tangier patients. Conclusions:Absence of ABCA1 and low HDL level induce reduction of platelet reactivity by decreasing positive feedback loops, particularly TXA2 production through a hematopoietic ABCA1-independent mechanism.
Keywords: ATP binding cassette transporter 1; cholesterol; eicosanoids; platelets; Tangier disease.
Introduction
The ATP-binding cassette transporter ABCA1 is a major actor of the reverse cholesterol transport pathway by load-ing cholesterol and phospholipids into apolipoprotein A-1 (apoA-1) to generate high-density lipoprotein (HDL) par-ticles [1–7]. ABCA1 has two transmembrane domains and two ATP-binding domains, which are crucial for this transport. It is highly expressed in intestine and liver, but it is also found in many other tissues including myeloid
Correspondence: Sonia Severin, I2MC, INSERM U1048, 1 Avenue Jean Poulhes, BP 84225, 31432 Toulouse, France.
Tel.: +33 5 31 22 41 43; fax: +33 5 61 32 56 02. E-mail: sonia.severin@inserm.fr
Received 8 July 2015
Manuscript handled by: C. Gachet
cells. Loss of function mutations of human ABCA1 are responsible for Tangier disease, a very rare autosomal recessive disorder characterized by a virtual absence of cir-culating HDL, a reduction of low-density lipoproteins (LDL), normal to elevated levels of triglycerides, and accu-mulation of cholesterol in peripheral tissues accompanied by atherosclerosis development [2–6]. Accordingly, disrup-tion of the abca1 gene in mice dramatically decreases HDL particles [1,7,8]. In mice, liver and intestinal ABCA1 are thought to control the production of most HDL particles, with monocyte/macrophage ABCA1 having minimal contribution [9–11]. The molecular mechanisms of ABCA1-mediated efflux of cellular lipids to apoA-1 are still incompletely understood. However, appropriate interaction of apoA-1 with the transporter and formation of a channel for passage of the substrate, likely through formation of ABCA1 dimers, are critical [9,12,13].
Platelets are the most important blood cells in the pri-mary prevention of bleeding after vascular injury and they are critical in arterial thrombosis. Hyporeactivity of blood platelets from patients with Tangier disease has been observed [14,15], and a bleeding tendency was reported in a patient with a nonsense mutation in ABCA1 gene and mild thrombocytopenia [16]. Moreover, a hem-orrhagic diathesis of ABCA1-deficient mice has been noted [7]. Previous studies have shown the expression of ABCA1 in human platelets and an increase of its mRNA during megakaryocytic maturation [15,17]. However, many questions remain concerning the role of ABCA1 in the regulation of platelet functions. It has been postulated that this transporter is involved in the calcium-dependent exposure of the anionic phospholipid phosphatidylserine (PS) to the plasma membrane, which occurs during cell apoptosis [8], and is also a crucial step in the development of the procoagulant activity of platelets [18]. Moreover, a missense mutation in the ABCA1 gene was reported in a patient with platelet Scott syndrome characterized by a lack of platelet procoagulant activity, leading to the sug-gestion that this transporter could be part of the scram-blase machinery required for PS exposure in platelets [19]. However, other reports argue against its direct implica-tion in lipid scrambling [15,18,20]. Ultrastructural defects of platelets from a Tangier patient with a reduced number of dense granules and the presence of some giant granules have been reported [15]. These ultrastructural abnormali-ties appear to be heterogeneous since in another study, platelets from Tangier patients display expanded Golgi complexes and reduced open canalicular system without modification of granules [7]. Finally, the impact of ABCA1 and hypo-HDLemia in the control of platelet lipid composition remains poorly known. The rarity of Tangier patients is a limiting factor for the characteriza-tion of the role of this lipid transporter in platelet func-tions in humans. Taking advantage of the ABCA1-deficient mouse model, we show that deficiency in this lipid transporter leads to a decrease in platelet reactivity
by reducing positive feedback loops particularly throm-boxane A2 (TXA2) production through a mechanism independent of the hematopoietic ABCA1. This reduction of TXA2 production was also observed in platelets from a Tangier patient. This mechanism may provide a long-lasting protective effect to counteract the elevated atherothrombotic risk.
Methods
Materials
Detailed materials are available in Data S1.
Tangier patients
Patient 1 is a 55-year-old white man who was referred to us for advice due to mild chronic thrombocytopenia (50– 60 g L1) before an ophthalmic surgery. Patient 2 is a 64-year-old white woman with established Tangier disease. The characterization of these Tangier patients is available as Data S1. The lipid profile and characteristics of the patients are shown in Table S1.
Citrated blood samples were obtained on two occasions at 5-month interval for patient 1 and one occasion for patient 2, with informed consent in accordance with the Declaration of Helsinki and approval by the local ethics review board. Washed platelets were prepared as previ-ously described [21]. Patient 1 did not take any medication with antiplatelet effect for at least 10 days before being tested, while patient 2 was receiving aspirin therapy. Con-trol subjects were healthy people with no medication.
Animals
Wild-type (WT) and abca1/ mice were of DBA1/lacJ genetic background as described previously [1,7,8]. The abca7/ mice, kindly provided by Dr M.W. Freeman (Massachusetts General Hospital, Boston, MA, USA), were of C57BL/6J genetic background as described previ-ously [22]. For bone marrow transplantation, recipient mice were lethally irradiated (9.2 Gy, c source) and then intravenously reconstituted with bone marrow cells from either WT or abca1/mice. Chimera mice were analyzed 6 weeks after injection. Anesthesia was performed via intraperitoneal injection of ketamine (100 mg kg1) and xylazine (10 mg kg1), and euthanasia via cervical dislo-cation. All mice experiments were approved by the insti-tutional review board in agreement with Directive 2010/ 63/EU of the European Parliament on the protection of animals used for scientific purposes.
Preparation and activation of mouse platelets
Blood was collected into acid-citrate-dextrose, and plate-lets were isolated as described previously [23]. Platelet
pellets were resuspended in modified HEPES-Tyrode’s buffer containing 2 mmol L1CaCl2(pH 7.38), pooled at a density of 49 108 platelets mL1in the presence of the adenosine-50-diphosphate (ADP) scavenger apyrase (ade-nosine-50-triphosphate [ATP] diphosphohydrolase), and incubated for 1 h at 37°C before platelet stimulation. Platelet aggregation experiments were monitored under continuous stirring at 1000 rpm at 37°C by a turbidimet-ric method, and ATP secretion was recorded by measur-ing the luminescence from the firefly luciferin-luciferase reaction using the Chrono-logÒ aggregometer (Stago, Lei-den, the Netherlands).
Reticulated platelets were quantified using thiazole orange (Retic-Count; Becton Dickinson, Franklin Lakes, NJ, USA) with an LSRFortessaTM
Cell analyzer flow cytometer and Diva software (Becton Dickinson).
Tail bleeding time
Mice were anesthetized via intraperitoneal injection of ketamine (25 mg kg1) and xylazine (10 mg kg1), and a 3-mm segment of the tail tip was cut off with a scalpel. Tail bleeding was monitored by gently absorbing the bead of blood every 15 s with filter paper without contacting the wound site [22,23]. When no blood was observed on the paper after 15-s intervals, bleeding was determined to have ceased.
Fibrinogen binding assays
Washed platelets (19 108platelets mL1) were incubated simultaneously with thrombin and Oregon Green 488– conjugated fibrinogen (150lg mL1 in final volume) in Tyrode’s buffer for 10 min at 37°C without shaking. Samples were fixed (1% formalin) and analyzed via the use of flow cytometry.
Lipid analysis
Total fatty acid and cholesterol quantification was per-formed on washed platelets by gas chromatography–flame ionization detector as described previously [24,25]. To simultaneously separate phospholipids by liquid chro-matography–mass spectrometry (MS)/MS, the previous neutral lipid extract [24] was analyzed on an ultra-high-per-formance liquid chromatography system (Agilent LC1290 Infinity) coupled to Agilent 6460 triple quadrupole MS (Agilent Technologies, Santa Clara, CA, USA) equipped with electro-spray ionization operating in negative and pos-itive mode. Eicosanoids were analyzed by MS as described [26]. Detailed methods are available in Data S1.
Measurement of PS exposure
The PS exposure was monitored by using the FITC-annexin V probe according to the manufacturer’s
recom-mendations (BD Biosciences Pharmingen, San Diego, CA, USA). Briefly, platelets were activated with indicated agonists for 4 min, incubated with annexin V-FITC for an additional 10 min in the dark, diluted, and analyzed by flow cytometry.
Laser-induced vessel wall injury
Thrombus formation over time after laser-induced deep injury of the mesenteric artery wall was performed as described [27].
Electron microscopy
Platelets were fixed in 2.5% glutaraldehyde in 0.1 mol L1 sodium cacodylate buffer at 4°C and allowed to adhere on poly-lysine–coated coverslips. They were then dehydrated in a graded ethanol series (scanning electron microscope) or were embedded in resin and sec-tioned (transmission electron microscopy) according to standard procedures.
Statistical analysis
Data were expressed as mean SEM of multiple experi-ments. Student’s t-tests were used to compare two groups,
and ANOVA tests were used to perform multiple
compar-isons using PRISM software (GraphPad, San Diego, CA, USA). A P value < 0.05 was considered statistically significant.
Results
ABCA1 deficiency in mice induces abnormal hemostasis and affects platelet size and aggregation
ABCA1 is expressed in mouse platelets at a low level com-pared with hepatocytes (Fig. 1A). Deletion of ABCA1 had no significant impact on the circulating platelet count (8.6 0.6 9 108/mL of blood for abca1/ mice and 9.2 0.4 9 108/mL of blood for WT mice, n= 5), and there was no alteration of the standard coagulation test prothrombin time (WT, 11.1 0.7 s; abca1/, 11.2 0.9 s, n = 5). Scanning electron microscopy studies of platelets stimulated in suspension for 20 s with thrombin indicated that the morphological modifications in abca1/ platelets (rounding and filopodia formation) were micro-scopically indistinguishable from those observed in WT platelets (Fig. S1A). Transmission electron microscopy analysis revealed that platelet size was weakly but signifi-cantly increased in abca1/animals, as shown by platelet width measurement (Fig. S1B). The mean volume of plate-let population was also significantly increased (7.76 0.18 lm3 for WT vs. 10.44 0.73 lm3 for ab-ca1/, n= 7, P < 0.01), and accompanied by a slight increase in the amount of total proteins (Fig. S1B). The
percentage of reticulated platelets measured by thiazole orange (Retic-Count) was similar in WT (17.10% 1.1%) and abca1/(18.03 1.7%) mice (n = 4).
Platelet aggregation responses induced by low doses of thrombin and collagen were consistently reduced in abca1/ platelets compared with control. The overall maximal aggregatory response to high doses of these ago-nists was, however, similar in abca1/ and WT platelets (Fig. 1B). The defect observed at low doses of agonists was not due to a decreased expression of platelet surface GPVI and aIIbb3 integrin, which were comparable in WT and abca1/ platelets (not shown). Consistent with the aggregation response, the level ofaIIbb3 activation moni-tored by Oregon-green fibrinogen binding was signifi-cantly reduced at low doses of agonists while raising the
doses of thrombin reduced the defect (Fig. 1D). The func-tionality ofaIIbb3 was further evaluated by analyzing the adhesion and spreading of platelets on immobilized fib-rinogen, and no significant difference was observed after 20 min of adhesion (mean platelet area: 14.19 0.2 lm2 for WT vs. 14.07 0.29 lm2for abca1/, n= 2).
The secretion of dense granules that contain ADP, an important positive feedback loop molecule that helps to propagate platelet activation, was quantified by monitor-ing ATP release by luminometry (Fig. 1C). A strong decrease in agonist-induced ATP secretion was observed in abca1/ platelets stimulated at low concentrations of thrombin and collagen. Again, raising the agonist concen-trations restored a nearly normal secretion response. The secretion ofa-granules monitored by P-selectin expression
ABCA1 Actin WT Hepat 250 kDa Platelets abca1–/– Collagen 2 µg mL–1 3 µg mL–1 0.05 IU mL–1 0.5 IU mL–1 5 µg mL–1 1 IU mL–1 WT WT WT WT WT WT Thrombin abca1–/– abca1–/– abca1–/– 1 min Aggregation 10% Light tr ansmittance Light tr ansmittance Light tr ansmittance abca1–/– abca1–/– abca1–/– 2 µg mL–1 3 µg mL–1 5 µg mL–1 1 IU mL–1 0.05 IU mL–1 0.5 IU mL–1 WT Collagen Thrombin 2 min ATP secretion 8 Ω abca1–/– WT abca1–/– WT abca1–/– WT abca1–/– WT WT abca1–/– abca1–/– D C A B 100 80 60 40 20 0 0 0.5 1 Thrombin (IU mL–1) P ercent of oregon g reen-conjugated fibr inogen positiv e cells WT
*
abca1–/–Fig. 1. Expression of ABCA1 in mouse platelets and effect of its deletion on platelet responses. (A) Washed mouse platelet homogenates (60lg proteins) from wild-type (WT) or ABCA1-deficient mice (abca1/) were submitted to Western blotting analysis with the anti-ABCA1 antibody. As a positive control, 60lg of proteins of a mouse hepatocyte homogenate were loaded (H). Results shown are representative of two independent experiments. (B) Washed platelets were stimulated with increasing doses of collagen or thrombin and aggregation was assessed. The profiles shown are representative of eight independent experiments. (C) ATP secretion of wild-type (WT) and abca1/washed platelets was recorded by measuring the luminescence from the firefly luciferin-luciferase reaction by lumiaggregometry using the Chrono-logÒ aggregometer. The ATP secretion profiles shown are representative of four independent experiments. (D) The binding of Oregon green 488– conjugated fibrinogen to WT and abca1/washed platelets activated or not by thrombin (0.5 and 1 IU mL1) for 5 min was measured by flow cytometry. Results are expressed as percentage of Oregon green 488–conjugated fibrinogen positive platelets and are the mean SEM of four independent experiments.*P < 0.05 according to two-tailed Student’s t test.
tended to be decreased at low doses of agonists, but this was not significant (Fig. S2C).
The aggregation defect was also observed with low doses of collagen-related peptide and TXA2 analog U46619, while ADP-induced platelet aggregation in whole blood was not significantly modified (Fig. S2A,B). Con-sistent with these defects, abca1/ mice exhibited a pro-longation of tail bleeding time (Fig. 2A). Bleeding stopped in all WT mice within 6 min (mean 3.16 0.26 min, n = 28), while in ABCA1-deficient ani-mals bleeding times were variable and generally increased (mean 8.52 1.04 min, n = 30). The difference was sig-nificant (P< 0.001), but all abca1/mice stopped bleed-ing within the observation period of 15 min.
To further assess platelet responses in vivo, we used a model of arterial thrombus formation monitored by intravital microscopy after locally injured mesenteric arteries [27]. Deep vessel lesion induced by a laser led to a rapid and robust thrombotic response, comparable in WT and abca1/mice (Fig. 2B). However, a significantly
faster decrease in thrombus size was observed (18 000 2000 lm2 vs. 10 600 1000 lm2, P < 0.01 at 360 s) in abca1/ mice, suggesting platelet detachment due to reduced intensity of their activation, particularly at the shell of the thrombus where platelets are known to be less activated than in the core.
These defects are specific of ABCA1 deficiency since absence of ABCA7, a structurally close relative of ABCA1 that does not play an essential role in the efflux of cholesterol to apoA-I [22] and is expressed in platelets [28], did not significantly affect mice tail bleeding time or platelet aggregation responses induced by thrombin or collagen (Fig. S3A–C). Further, platelets from mice with double ABCA1/ABCA7 deficiency had similar phenotype as abca1/platelets (Fig. S3D).
Collectively, these data indicate that loss of ABCA1 in mouse results in a slight increase in platelet size and a reduction of dense granule secretion and platelet aggrega-tion after stimulaaggrega-tion with low doses of agonists. These defects are no longer observed after high doses of agonist stimulation. Consistent with these results and those previ-ously reported [15], platelets from a Tangier patient whose characteristics are shown in Table S1 and Methods exhibited an aggregation defect in response to low doses of collagen and, to a lesser extent, of thrombin (Fig. S4A). No obvious morphological aberration could be observed on electron microscopy (Fig. S4B).
Loss of ABCA1 does not affect cholesterol and major phospholipid content of platelets and spares calcium-dependent PS exposure
The major function of ABCA1 is the efflux of cholesterol and phospholipids from cells to HDL particles. The plasma lipid profile of abca1/ mice shows a dramatic decrease in HDL cholesterol and apoA-I compared with WT littermates [1]. Whether this lipid transporter can impact the homeostasis of cholesterol and phospholipids in platelet membranes is still uncertain. The amount of cholesterol in the plasma membrane is an important determinant of platelet signal transduction efficiency, pos-sibly via its impact on lipid raft organization [29]. The level of platelet cholesterol and the molar ratio of choles-terol/phospholipids were comparable in control and ab-ca1/platelets (Fig. 3A). This is consistent with the fact that, in vitro, ABCA1-dependent H3-cholesterol efflux from mouse platelets to apoA-I was very weak and not significantly different from non-specific efflux to bovine serum albumin (15% and 12%, respectively).
Targeted mass spectrometry analysis indicated that the repartition of major phospholipids was not significantly modified (Fig. 3B). Absence of ABCA1 did not change sig-nificantly the global level of platelet saturated, monounsat-urated, or polyunsaturated fatty acid (Table 1). As shown in Table S2, molecular species of phospholipids were also comparable, with few noticeable modifications in the
15 10 5 0 30 000 25 000 20 000 15 000 Thromb us area (µm 2)
Bleeding time (min)
10 000 5000 0 0 120 240 360 Time (s) 480 600 WT abca1–/– WT abca1–/– A B
Fig. 2. Effect of ABCA1 deficiency on tail bleeding time and in vivothrombosis. (A) Tail bleeding time of wild-type (WT) (n= 28) and abca1/mice (n= 30). (B) Thrombus formation over time after laser-induced deep injury of mesenteric artery wall of WT and abca1/ mice was followed by intravital microscopy. Eight WT mice (n= 10 vessels) and three abca1/mice (n= 3 vessels) were analyzed. The mean thrombus surface was analyzed at 0.3-s intervals, and the shading over the curve represents the SEM at each time point. Time 0 corresponds to the time of injury.
repartition including, some increases in polyunsaturated long-chain fatty acid containing phospholipids (PE C40:6 and C40:4; PC C36:4, C36:3, C38:2, and C38:4). Moreover, binding of FITC-conjugated cholera toxin to the lipid raft marker ganglioside GM1was comparable on the surface of control and abca1/platelets either in resting conditions (WT: 28.2 6; abca1/ 34.5 6, mean fluorescence intensity values) or after ADP stimulation (WT: 34.6 4; abca1/36.8 5, mean fluorescence intensity values).
Overall, this lipidomic analysis indicates that the absence of ABCA1 and HDL in mouse does not change
significantly cholesterol and major phospholipid content of platelet membranes.
Platelets from Tangier patient 1 were also analyzed, and the levels of platelet cholesterol and total saturated, monounsaturated, and polyunsaturated fatty acids were comparable to those of control platelets (Table S3).
We then investigated the potential role of ABCA1 in the calcium-dependent exposure of the anionic phospho-lipid PS by monitoring annexin V-FITC binding to mouse platelets stimulated either by thrombin plus collagen or by the calcium ionophore A23187. No significant differ-ence in the exposure of PS could be observed between control and abca1/ platelets (Fig. 3C), indicating that this lipid transporter is not involved in the calcium-depen-dent scramblase machinery and the procoagulant activity of mouse platelets.
Loss of ABCA1 reduces platelet TXA2 production independently of hematopoietic ABCA1
To further explore the defect of abca1/ platelet responses observed at low doses of agonists, we investi-gated lipid mediators generated by platelets using an MS-based lipidomics approach. Quantification of eicosanoids indicated that thrombin stimulation induced a potent pro-duction of TXB2, the stable degradation product of TXA2, in WT platelets (Fig. 4). TXA2 production from arachidonic acid through the cyclooxygenase pathway is an important positive feedback loop mechanisms of plate-let activation. A significant reduction of its production was observed in abca1/ platelets. Other arachidonic acid derivatives involving lipo-oxygenases (12- and 15-hydroxyeicosatetraenoic acids or 12- and 15-HETE) were also significantly less produced, suggesting a decrease in arachidonic acid generation (Fig. 4). Consistent with this defect in TXA2 production, addition of low doses of the TXA2 analog U46619 (0.5lmol L1) increased platelet aggregation of abca1/ mice induced by low concentra-tions of thrombin (0.05 IU) or collagen-related peptide (0.5lg mL1) from 22 2% to 49 12% and from 43 10% to 82 4% (n = 2), respectively.
In mice, HDL particles are mainly produced via liver and intestinal ABCA1 [9–11]. Therefore, to investigate the relative impact of the platelet ABCA1 and of the low 0.2 A B C Total cholesterol (nmol per 10 6 platelets) Cholesterol/phospholipids Total phospholipids (%)
Annexin V positive cells (%)
0.6 0.5 0.4 0.3 0.2 0.1 0 0.15 0.1 0.05 0 50 40 30 20 10 0 0 – – – + + – + – – 20 40 60 80 100 PC Thrombin (0.5 IU mL–1): Collagen (2.5 µg mL–1): A23187 (10 µM): PS PE PI SM WT WT abca1–/– abca1–/– WT abca1–/– WT abca1–/–
Fig. 3. Loss of ABCA1 does not impact platelet major lipid compo-sition and PS exposure. (A) The amount of cholesterol and choles-terol/phospholipid molar ratio were quantified in wild-type and abca1/platelets (mean SEM of seven independent experiments). (B) The different phospholipids in wild-type and abca1/platelets were quantified by mass spectrometry. Data obtained in two inde-pendent experiments are shown and results are expressed as mol% of total phospholipids. (C) Exposure of the negatively charged phos-pholipid PS as determined by annexin V–FITC binding in control and abca1/platelets. Results are expressed as percentage of annexin V–FITC–positive platelets in resting conditions or after col-lagen (2.5lg mL1) plus thrombin (0.5 IU mL1) or A23187 stimu-lations during 4 min and are the mean SEM of four independent experiments.
Table 1 Repartition of total fatty acids from control and abca1/ mice platelets
WT abca1/
Saturated fatty acids 41.3 1.1 40.8 2.8 Monounsaturated fatty acids 13.1 0.2 13.2 0.3 Polyunsaturated fatty acids 45.5 1.2 46.0 2.8 Total fatty acids of platelets from wild-type and abca1/mice were quantified by gas chromatography–flame ionization detector. Results are expressed as percentage of total fatty acids and are the mean SEM of six independent experiments.
HDL level in this defect, we engrafted lethally irradiated WT or abca1/mice with bone marrow from WT or ab-ca1/mice. As expected, the plasma esterified-cholesterol level of the hematopoietic chimera mice was low when bone marrow from WT mice was grafted in abca1/ mice (Fig. 5A). The production of TXA2 by thrombin-sti-mulated platelets was again significantly reduced when ab-ca1/mice where engrafted with abca1/bone marrow. Interestingly, engraftment of WT mice with abca1/ bone marrow had no effect on TXA2 production, while engraftment of abca1/ mice with WT bone marrow recapitulated the defect (Fig. 5B). Similar results were obtained with 12- and 15-HETE production. This bone marrow transplantation approach demonstrates that hematopoietic ABCA1, and therefore platelet ABCA1, is not sufficient to induce a reduction in TXA2 production by platelets. Deficiency of ABCA1 in the total body, except the hematopoietic tissue, mimics the defect observed in abca1/ mice. These data strongly suggest
that the lipid environment (i.e. low HDL level) is more important than the platelet ABCA1 per se in the control of TXA2 production by stimulated platelets.
Interestingly, compared with control healthy donors, platelets from the Tangier patient 1 exhibited an impor-tant decrease in TXB2, 12-HETE, and 15-HETE produc-tion after thrombin stimulaproduc-tion (Fig. S4C). A similar profile of these bioactive lipid production was observed in two independent analysis performed at a 5-month inter-val. We had the opportunity to analyze the production of eicosanoids in platelets from a second Tangier patient (Table S1), who was, however, receiving aspirin therapy. As expected, TXB2 production was abolished, but throm-bin-induced production of 12- and 15-HETE was, again, strongly decreased (Fig. S4C).
Discussion
ABCA1 is highly expressed in hematopoietic stem and multipotential progenitor cells but poorly in megakary-ocyte-erythrocyte progenitors [30]. Here, we show that this transporter is weakly present in mouse platelets and its absence does not affect platelet count but increases plate-let size, suggesting a potential contribution of ABCA1 and/or HDL level in the last steps of megakaryocy-topoiesis. The abca1/mice have normal standard coagu-lation tests but a defect in primary haemostasis. Although this bleeding defect may involve different cell types expressing ABCA1 we found that abca1/platelets have a reduced reactivity in response to low doses of thrombin or collagen. These defects are specifically linked to the lack of ABCA1 because the deficiency of ABCA7, another ABCA transporter expressed in mouse platelets but playing a minor role in the efflux of cholesterol or phosphatidylcholine to apoA-I [22], has no significant effect on platelet activation and because double abca1/abca7/ platelets have a similar phenotype as abca1/platelets.
We show that aggregation defects of abca1/ platelets are related to a deficiency in positive feedback loop mech-anisms involving dense granule secretion and TXA2 pro-duction. Dense granules contain ADP, which, on secretion, activates the platelet P2Y12 receptor, an estab-lished target of antithrombotic drugs, to amplify platelet aggregation and, in turn, thrombus growth and stability [31]. Compared with WT mouse platelets, abca1/ plate-lets have a strong reduction of dense granule secretion when stimulated by low doses of thrombin or collagen. High doses of agonists overcome this defect. Moreover, after activation, platelets are known to produce proaggre-gatory lipids generated from arachidonate via cyclooxyge-nase or lypo-oxygecyclooxyge-nase activities. The cyclooxygecyclooxyge-nase product TXA2 is an important mediator of the positive platelet feedback loop mechanisms. The role of the lypo-oxygenase products (12- and 15-HETE) on platelets is less understood [32]. Using a targeted lipidomics approach, 50 40 30 20 10 0 2.0 8 6 4 2 0 OH COOH COOH COOH OH WT abca1–/– O OH HO
*
*
*
OH 1.5 1.0 0.5 0.0 Resting TXB 2 (f old increase) 12-HETE (f old increase) 15-HETE (f old increase) Thrombin 0.1 IU mL–1 Resting Thrombin 0.1 IU mL–1 Resting Thrombin 0.1 IU mL–1Fig. 4. Production of TXB2, 12-HETE, and 15-HETE in abca1/ platelets. Washed platelets from wild-type or abca1/mice were stimulated by thrombin (0.1 IU mL1) for 3 min and eicosanoids were analyzed by mass spectrometry. Results are expressed as fold increase of the resting wild-type value and are the mean SEM from six independent experiments. The maximal production was: thromboxane B2 (TXB2: 87.04 4.52 ng/4 9 108platelets); 12-hydroxyeicosatetraenoic acid (12-HETE: 195.77 22.36 ng/4 9 108 platelets); and 15-hydroxyeicosatetraenoic acid (15-HETE:
10.72 1.79 ng/4 9 108platelets).*P < 0.05 according to two-tailed Student’s t test.
we demonstrated that, compared with WT platelets, the generation of TXA2, 12-HETE, and 15-HETE are signifi-cantly decreased in the absence of ABCA1, suggesting that ABCA1 deficiency is responsible for a defect in arachidonic acid production via phospholipase A2. It is noteworthy that the aggregation response of abca1/ platelets is reminiscent to that of TXA2 receptor knock-out platelets [33].
In accordance to results obtained in abca1/ mice model, platelets from Tangier patients are less responsive to agonist stimulation and have an important reduction in TXA2, 12-HETE, and 15-HETE production after stimulation.
One potential hypothesis is that ABCA1 deficiency could have changed the lipid composition of platelets, leading to modification in their activation processes. 750 bp Bone marrow A B tail Bone marrow tail Bone marrow tail Bone marrow tail 500 bp Cholesterol-ester: (nmol L–1) 20 TXB 2 (fold increase)
12-HETE (fold increase) 15-HETE (fold increase)
15 10
*
*
*
*
5 0 3 2 1 0 8*
*
4 6 2 0 Resting Thrombin 0.1 IU mL–1Resting Thrombin 0.1 IU mL–1 Resting Thrombin 0.1 IU mL–1
WT > WT WT > WT 294±14 106±37 205±52 79±16 abca1–/– > abca1–/– abca1–/– > abca1–/– abca1–/– > WT abca1–/– > WT WT > abca1–/– WT > abca1–/–
Fig. 5. The defect in eicosanoids production in abca1/platelets is independent of hematopoietic ABCA1. Hematopoietic chimera mice were generated by engrafting normal bone marrow (WT) or bone marrow from abca1/mice (abca1/) in wild-type or abca1/mice. (A) Control showing the validity of the bone marrow engraftment by PCR (the 750-bp band corresponds to the WT ABCA1 gene amplicon, whereas the 500-bp band corresponds to the disrupted ABCA1 gene amplicon) and the level of esterified cholesterol in chimera mice plasma (mean SEM of three experiments). (B) Washed platelets from the different chimeras (WT> WT, abca1/> WT, abca1/> abca1/, WT> abca1/) were stimulated by thrombin (0.1 IU mL1) for 3 min, and eicosanoids (TXB2, 12-HETE, 15-HETE) were analyzed as in Fig. 4. Results are expressed as fold increase of the resting value and are the mean SEM from three independent experiments. *P < 0.05 according to two-way
Membrane cholesterol content, for instance, is an impor-tant determinant of platelet reactivity [29,34,35]. Impor-tantly, our lipidomics analysis clearly shows that platelets from abca1/ mice have no significant modification in membrane cholesterol content and major phospholipid composition. These data indicated that ABCA1 is not involved in cholesterol efflux from platelets, as also sug-gested by in vitro cholesterol efflux experiments. This is different from macrophages lacking ABCA1, which have elevated membrane cholesterol content leading to increased sensitivity to lipid raft–dependent signaling pro-cesses and proinflammatory response [36]. Our data also show that the dramatic drop in HDL found in abca1/ mice [1,7] does not significantly impact platelet membrane major lipid content. It is noteworthy that HDL scavenger class B type I (SR-BI) deficiency in mice increases free-to-total cholesterol ratio in plasma as well as platelet size, cholesterol content, clearance and reactivity [33,37–39]. Thus, when HDL particles and plasma cholesterol accu-mulate, such as in sr-b1/ mice, platelets, and possibly megakaryocytes can uptake free cholesterol, but when HDL particles are absent, as in abca1/ mice, platelets keep a normal cholesterol content. Similar results were obtained in platelets from a Tangier patient with primary HDL deficiency. Platelets of this patient have normal cholesterol content and fatty acid repartition. These are important observation in the context of dyslipidemia and the development of novel therapies modulating HDL par-ticles in cardiovascular patients.
Moreover, we show that ABCA1 is not required for calcium-dependent PS exposure in mouse platelets, excluding a role of this transporter in the regulation of the procoagulant activity of platelets.
Our data suggest that absence of platelet ABCA1 per se may not be responsible for the defect in platelet responses in mouse. Interestingly, using a combination of hematopoietic chimera mice, we demonstrated that the defect in TXA2, 12-HETE, and 15-HETE production is primarily dependent on external environment (i.e. low HDL) and not on the hematopoietic ABCA1. How exter-nal environment and low HDL level in plasma decrease the efficiency of positive feedback loop of platelet activa-tion, particularly TXA2 producactiva-tion, remains to be deter-mined. Recently, omega-3 fatty acids were shown to shunt metabolism of the arachidonate–to–lipo-oxygenases pathway in macrophages, indicating that hematopoietic cells can modulate eicosanoids signaling and, in turn, thrombosis and inflammation in response to cellular lipid homeostasis [40,41]. Understanding how physiological integrators of cellular and plasma lipid homeostasis can direct the production of lipid mediators modulating hemostasis and inflammation may lead to new therapeutic strategies to decrease the thrombotic risk in patients with cardiovascular disease. Several ABCA and ABCB trans-porters have been described in platelets, including ABCA3, ABCA4, ABCA6, ABCA7, ABCA9, ABCB6,
and ABCB4, and may contribute to the adaptation of platelets to their environment [42].
In conclusion, our data show that ABCA1 is not required to control the level of cholesterol and major phospholipid species in platelet membranes, neither directly nor indirectly via HDL particle deficiency. This transporter is not involved in PS exposure and the proco-agulant platelet activity, but its loss impacts platelet acti-vation by low doses of agonists through a reduction of positive feedback loop mechanisms. Interestingly, the decrease in TXA2 and other arachidonate metabolites production by activated platelets in abca1/mice is due to external environment and HDL level rather than hematopoietic ABCA1. Thus, this study provides new information on the mechanisms supporting the reduction of platelet aggregation in Tangier patients.
Addendum
T. Lhermusier and S. Severin equally participated in designing and performing the research and analyzed data. J. Van Rothem contributed to experiments on mouse plate-lets. C. Garcia performed experiments on the Tangier patient within the frame of French Reference Center on Platelet Diseases. J. Bertrand-Michel and P. Le Faouder performed lipidomic analysis. B. Hechler performed intrav-ital microscopy. C. Broccardo and G. Chimini provided the mouse models. P. Couvert performed the genetic analy-sis of the patients. P. Sie and P. Couvert evaluated the Tangier patients, and B. Payrastre controlled, analyzed data, and wrote the manuscript. All authors read, reviewed, and approved the final manuscript.
Acknowledgments
We thank Nicolas Cenac for helpful discussions and Ash-raf Ragab and Stephanie Magnenat for their help with in vitroplatelet analysis and intravital microscopy, respec-tively. B. Payrastre is a scholar of the Institut Universi-taire de France. This work was supported by grants from INSERM, Fondation pour la Recherche Medicale, and Fondation de France. T. Lhermusier was supported by a grant from CHU-Toulouse, and J. Van Rothem was sup-ported by the Fondation pour la Recherche Medicale. Disclosure of Conflict of Interests
The authors state that they have no conflict of interest.
Supporting Information
Additional Supporting Information may be found in the online version of this article:
Fig. S1. Loss of ABCA1 spares platelet shape change but increases platelet size.
Fig. S2. Platelet aggregation and a-granule secretion in abca1/platelets.
Fig. S3. abca7/have no bleeding phenotype and a nor-mal platelet aggregation response after stimulation by col-lagen or thrombin.
Fig. S4. Tangier patient platelet aggregation and produc-tion of eicosanoids.
Table S1. Lipid profile and biochemical measurements of the Tangier patients.
Table S2. Repartition of phospholipid molecular species in abca1/platelets.
Table S3. Repartition of total fatty acids and cholesterol content of platelets from Tangier patient 1 and healthy donors.
Data S1. Materials.
References
1 McNeish J, Aiello RJ, Guyot D, Turi T, Gabel C, Aldinger C, Hoppe KL, Roach ML, Royer LJ, de Wet J, Broccardo C, Chi-mini G, Francone OL. High density lipoprotein deficiency and foam cell accumulation in mice with targeted disruption of ATP-binding cassette transporter-1. Proc Natl Acad Sci USA 2000; 97: 4245–50.
2 Bodzioch M, Orso E, Klucken T, Langmann T, B€ottcher A, Die-derich W, Drobnik W, Barlage S, B€uchler C, Porsch-Ozc€ur€umez M, Kaminski WE, Hahmann HW, Oette K, Rothe G, Aslanidis C, Lackner KJ, Schmitz G. The gene encoding ATB-binding cas-sette transporter 1 is mutated in Tangier disease. Nat Genet 1999; 22: 347–51.
3 BrooksWilson A, Marcil M, Clee SM, Zhang LH, Roomp K, van Dam M, Yu L, Brewer C, Collins JA, Molhuizen HO, Loubser O, Ouelette BF, Fichter K, Ashbourne-Excoffon KJ, Sensen CW, Scherer S, Mott S, Denis M, Martindale D, Froh-lich J, et al. Mutation in ABCA1 in Tangier disease and familial high-density lipoprotein deficiency. Nat Genet 1999; 22: 336–45. 4 Rust S, Rosier M, Funke H, Real J, Amoura Z, Piette JC,
Deleuze JF, Brewer HB, Duverger N, Denefle P, Assmann G. Tangier disease is caused by mutations in the gene encoding ATP-binding cassette transporter 1. Nat Genet 1999; 22: 352–5. 5 Bortnick AE, Rothblat GH, Stoudt G, Hoppe KL, Royer LJ,
McNeish J, Francone OL. The correlation of ATP-binding cas-sette 1 mRNA levels with cholesterol efflux from various cell lines. J Biol Chem 2000; 275: 28634–40.
6 Attie AD. ABCA1: at the nexus of cholesterol, HDL and atherosclerosis. Trends Biochem Sci 2007; 32: 172–9.
7 Orso E, Broccardo C, Kaminski WE, B€ottcher A, Liebisch G, Drobnik W, G€otz A, Chambenoit O, Diederich W, Langmann T, Spruss T, Luciani MF, Rothe G, Lackner KJ, Chimini G, Schmitz G. Transport of lipids from golgi to plasma membrane is defective in tangier disease patients and Abc1-deficient mice. Nat Genet2000; 24: 192–6.
8 Hamon Y, Broccardo C, Chambenoit O, Luciani MF, Toti F, Chaslin S, Freyssinet JM, Devaux PF, McNeish J, Marguet D, Chimini G. ABCA1 promotes engulfment of apoptotic cells and transbilayer redistribution of phosphatidylserine. Nat Cell Biol 2000; 2: 399–406.
9 Brunham LR, Kruit JK, Iqbal J, Fievet C, Timmins JM, Pape TD, Coburn BA, Bissada N, Staels B, Groen AK, Hussain MM, Parks JS, Kuipers F, Hayden MR. Intestinal ABCA1 directly contributes to HDL biogenesis in vivo. J Clin Invest 2006; 116: 1052–62. 10 Timmins JM, Lee JY, Boudyguina E, Kluckman KD, Brunham
LR, Mulya A, Gebre AK, Coutinho JM, Colvin PL, Smith TL,
Hayden MR, Maeda N, Parks JS. Targeted inactivation of hepatic Abca1 causes profound hypoalphalipoproteinemia and kidney hypercatabolism of apoA-1. J Clin Invest 2005; 115: 1333–42. 11 Haghpassand M, Bourassa PA, Francone OL, Aiello RJ.
Mono-cyte/macrophage expression of ABCA1 has minimal contribution to plasma HDL levels. J Clin Invest 2001; 108: 1315–20. 12 Nagata KO, Nakada C, Kasai RS, Kusumi A, Ueda K. ABCA1
dimer-monomer interconversion during HDL generation revealed by single-molecule imaging. Proc Natl Acad Sci USA 2013; 110: 5034–9. 13 Westerterp M, Bochem AE, Yvan-Charvet L, Murphy AJ, Wang N, Tall AR. ATP-binding cassette transporters, atherosclerosis, and inflammation. Circ Res 2014; 114: 157–70.
14 Schmitz G, Schambeck M. Molecular defect in the ABCA1 path-way affect platelet function. Pathophysiol Haemost Thromb 2006; 35: 166–74.
15 Nofer J-R, Herminghaus G, Brodde M, Morgenstern E, Rust S, Engel T, Seedorf U, Assmann G, Bluethmann H, Kehrel B. Impaired platelet activation in familial high density lipoprotein deficiency (Tangier disease). J Biol Chem 2004; 279: 34032–7. 16 Hooper AJ, Robertson K, Ng L, Kattampallil JS, Latchem D,
Willsher PC, Thom J, Baker RI, Burnett JR. A novel ABCA1 non-sense mutation, R1270X, in Tangier disease associated with an unrecognised bleeding tendency. Clin Chim Acta 2009; 409: 136–9. 17 Barlage S, Boettcher D, Boettcher A, Dada A, Schmitz G. High
density lipoprotein modulates platelet function. Int Soc Anal Cytol2006; 69A: 196–9.
18 Lhermusier T, Chap H, Payrastre B. Platelet membrane phos-pholipid asymmetry: from the characterization of a scramblase activity to the identification of an essential protein mutated in Scott syndrome. J Thromb Haemost 2011; 9: 1883–91.
19 Albrecht C, McVey JH, Elliott JI, Sardini A, Kasza I, Mumford AD, Naoumova RP, Tuddenham EG, Szabo K, Higgins CF. A novel missense mutation in ABCA1 results in altered protein trafficking and reduced phosphatidylserine translocation in a patient with Scott syndrome. Blood 2005; 106: 542–9.
20 Williamson P, Halleck MS, Malowitz J, Ng S, Fan X, Krahling S, Remaley AT, Schlegel RA. Transbilayer phospholipid move-ments in ABCA1-deficient cells. PLoS One 2007; 2: e729. 21 Gratacap M-P, Martin V, Valera M-C, Allart S, Garcia C, Sie
P, Recher C, Payrastre B. The new tyrosine kinase inhibitor and anticancer drug dasatinib reversibly affects platelet activation in vitro and in vivo. Blood 2009; 114: 1884–92.
22 Kim WS, Fitzgerald ML, Kang K, Okuhira K, Bell SA, Man-ning JJ, Koehn SL, Lu N, Moore KJ, Freeman MW. Abca7 null mice retain normal macrophage phosphatidyl-choline and choles-terol efflux activity despite alterations in adipose mass and serum cholesterol levels. J Biol Chem 2005; 280: 3989–95.
23 Valet C, Chicanne G, Severac C, Chaussade C, Whitehead MA, Cabou C, Gratacap M-P, Gaits-Iacovoni F, Vanhaesebroeck B, Payrastre B, Severin S. Essential role of class II PI3K-C2a in platelet membrane morphology. Blood 2015; 126: 1128–37. 24 Bligh EG, Dyer WJ. A rapid method of total lipid extraction
and purification. Can J Biochem Physiol 1959; 37: 911–7. 25 Zadravec D, Brolinson A, Fisher RM, Carneheim C, Csikasz RI,
Bertrand-Michel J, Boren J, Guillou H, Rudling M, Jacobsson A. Ablation of the very-long-chain fatty acid elongase ELOVL3 in mice leads to constrained lipid storage and resistance to diet-induced obesity. FASEB J 2010; 24: 4366–77.
26 Le Faouder P, Baillif V, Spreadbury I, Motta JP, Rousset P, Ch^ene G, Guigne C, Terce F, Vanner S, Vergnolle N, Bertrand-Michel J, Dubourdeau M, Cenac N. LC-MS/MS method for rapid and concomitant quantification of pro-inflammatory and pro-resolving polyunsaturated fatty acid metabolitites. J Chro-matogr B Analyt Technol Biomed Life Sci2013; 932: 123–33. 27 Hechler B, Nonne C, Eckly A, Magnenat S, Rinckel J-Y, Denis
CV, Freund M, Cazenave J-P, Lanza F, Gachet C. Arterial thrombosis: relevance of a model with two levels of severity
assessed by histologic, ultrastructural and functional characteri-zation. J Thromb Haemost 2009; 8: 173–84.
28 Sasaki M, Shoji A, Kubo Y, Noda S, Yamaguchi A. Cloning of rat ABCA7 and its preferential expression in platelets. Biochem Biophys Res Commun2003; 304: 77–782.
29 Bodin S, Tronchere H, Payrastre B. Lipid rafts are critical mem-brane domains in blood platelet activation processes. Biochim Biophys Acta2003; 1610: 247–57.
30 Murphy AJ, Bijl N, Yvan-Charvet L, Welch CB, Bhagwat N, Reheman A, Wang Y, Shaw JA, Levine RL, Ni H, Tall AR, Wang N. Cholesterol efflux in megakaryocyte progenitors suppresses pla-telet production and thrombocytosis. Nat Med 2013; 19: 586–94. 31 Gachet C. P2 receptors, platelet function and pharmacological
implications. Thromb Haemost 2008; 99: 466–72.
32 O’Donnell V, Murphy RC, Watson SP. Platelet lipidomics. Mod-ern day perspective on lipid discovery and characterization in platelets. Circ Res 2014; 114: 1185–203.
33 Thomas DW, Mannon RB, Mannon PJ, Latour A, Oliver JA, Hoffman M, Smithies O, Koller BH, Coffman TM. Coagulation defects and altered hemodynamic responses in mice lacking recep-tors for thromboxane A2. J Clin Invest 1998; 102: 1994–2001. 34 Zimman A, Podrez EA. Regulation of platelet function by class
B scavenger receptors in hyperlipidemia. Arterioscler Thromb Vasc Biol2010; 30: 2350–6.
35 Shattil SJ, Anaya-Galindo R, Bennett J, Colman RW, Cooper RA. Platelet hypersensitivity induced by cholesterol incorpora-tion. J Clin Invest 1975; 55: 636–43.
36 Zhu X, Lee JY, Timmins JM, Brown JM, Boudyguina E, Mulya A, Gebre AK, Willingham MC, Hiltbold EM, Mishra N, Maeda N, Parks JS. Increased cellular free cholesterol in macrophage-specific Abca1 knock-out mice enhances pro-inflammatory response of macrophages. J Biol Chem 2008; 283: 22930–41.
37 Nofer J-R, van Eck M. HDL scavenger receptor class B type I and platelet function. Curr Opin Lipidol 2001; 22: 277–82. 38 van der Stoep M, Korporaal SJ, van Eck M. High-density
lipoprotein as a modulator of platelet and coagulation responses. Cardiovasc Res2014; 103: 362–71.
39 Dole VS, Matuskova J, Vasile E, Yesilaltay A, Bergmeier W, Bernimoulin M, Wagner DD, Krieger M. Thrombocytope-nia and platelet abnormalities in high-density lipoprotein recep-tor-deficient mice. Arterioscler Thromb Vasc Biol 2008; 28: 1111–6.
40 Harker LA, Kelly AB, Hanson SR, Krupski W, Bass A, Osterud B, FitzGerald GA, Goodnight SH, Connor WE. Interruption of vascular thrombus formation and vascular lesion formation by dietary n-3 fatty acids in fish oil in nonhuman primates. Circula-tion1993; 87: 1017–29.
41 Norris PC, Dennis EA. Omega-3 fatty acids cause dramatic changes in TLR4 and purinergic eicosanoids signalling. Proc Natl Acad Sci USA2012; 109: 8517–22.
42 Jedlitschky G, Greinacher A, Kroemer HK. Transporters in human platelets: physiologic function and impact for pharma-cotherapy. Blood 2012; 119: 3394–402.
