Farnesoid X Receptor and Liver X Receptor Ligands Initiate Formation of Coated Platelets

Abstract

Objectives: The liver X receptors (LXRs) and farnesoid X receptor (FXR) have been identified in human platelets. Ligands of these receptors have been shown to have nongenomic inhibitory effects on platelet activation by platelet agonists. This, however, seems contradictory with the platelet hyper-reactivity that is associated with several pathological conditions that are associated with increased circulating levels of molecules that are LXR and FXR ligands, such as hyperlipidemia, type 2 diabetes mellitus, and obesity.

Approach and results: We, therefore, investigated whether ligands for the LXR and FXR receptors were capable of priming platelets to the activated state without stimulation by platelet agonists. Treatment of platelets with ligands for LXR and FXR converted platelets to the procoagulant state, with increases in phosphatidylserine exposure, platelet swelling, reduced membrane integrity, depolarization of the mitochondrial membrane, and microparticle release observed. Additionally, platelets also displayed features associated with coated platelets such as P-selectin exposure, fibrinogen binding, fibrin generation that is supported by increased serine protease activity, and inhibition of integrin αIIbβ3. LXR and FXR ligand-induced formation of coated platelets was found to be dependent on both reactive oxygen species and intracellular calcium mobilization, and for FXR ligands, this process was found to be dependent on cyclophilin D.

Conclusions: We conclude that treatment with LXR and FXR ligands initiates coated platelet formation, which is thought to support coagulation but results in desensitization to platelet stimuli through inhibition of αIIbβ3 consistent with their ability to inhibit platelet function and stable thrombus formation in vivo.

Keywords: bile; blood coagulation; blood platelets; calcium; cholesterol.

Figure 1.

Liver X receptor (LXR) and farnesoid X receptor (FXR) ligands induce platelet procoagulant activity. Human washed platelets were treated for 10 min with or without increasing concentrations of GW3965 (1, 5, 10, 20, and 40 µmol/L) or GW4064 (1, 5, 10, 20, and 40 µmol/L) or vehicle control before analysis by flow cytometry for (A) annexin V binding, which is a measure of phosphatidylserine exposure, and data expressed as (i) median fluorescence intensity and (ii) percentage of annexin V–positive cells; (B) formation of microparticles, determined by gating for the microvesicle population using forward and side scatter profiles of ApogeeMix beads. Data expressed as number of phosphatidylserine (PS)–positive events per microliter. C and D, Platelet swelling, measured by analyzing (C) light transmission, with an increase in transmission associated with increased platelet swelling (i) and (ii) representative traces and (iii) representative images. D, (i) data expressed as % of light transmission; (ii) mean platelet volume. E, Calcein fluorescence, to measure membrane integrity, expressed as median fluorescence intensity. Results are mean+SEM for n≥3. *P<0.05 in comparison to vehicle controls.

Figure 2.

Liver X receptor (LXR) and farnesoid X receptor (FXR) agonists initiate coated platelet formation. Human washed platelets were treated with increasing concentrations of GW3965 (1, 5, 10, and 20 µmol/L), GW4064 (1, 5, 10, and 20 µmol/L), or vehicle control and tested for several different markers of coated platelet formation including (A) fibrinogen binding, (B) α-granule components, measured by detecting P-selectin exposure, (D) serine protease activity, measured by quantification of release of fluorescence as a result of conversion of Z-Gly-Gly-Arg-AMC fluorescent substrate for 1 h at 37°C, and (E) bound fibrin. Data expressed as median fluorescence intensity. C, Human platelets treated with GW3965 (20 µmol/L), GW4064 (20 µmol/L), or vehicle were stained with annexin V–Cy5 (in red) and antifibrinogen (Alexa 488 conjugated; in green) and analyzed by (i) fluorescence microscopy on a ×100 oil immersion lens and (ii) flow cytometry. (i) Representative images and (ii) representative scatter plots shown. Other results are mean+SEM for n≥3. *P<0.05 in comparison to vehicle controls.

Figure 3.

Liver X receptor (LXR) and farnesoid X receptor (FXR) agonists inhibit αIIbβ3. Human washed platelets were prepared, and the effect of GW3965 or GW4064 (10 and 20 µmol/L) on (A) integrin αIIbβ3 activation determined by PAC-1 antibody binding and (B) fibrinogen binding in the presence or absence of integrillin (5 µmol/L), an inhibitor of integrin αIIbβ3 fibrinogen binding. (Data expressed as median fluorescence intensity, mean+SEM for n≥3. C, PAC-1 binding and phosphatidylserine exposure in CRP-stimulated (1 µg/mL) and thrombin-stimulated (0.1 U/mL) platelets was determined. Representative scatter plots shown.

Figure 4.

Mitochondrial transmembrane potential and sustained calcium signaling are required for the induction of LXR- and FXR-mediated procoagulant activity. Human washed platelets were treated for 10 min with or without increasing concentrations of GW3965 (1, 5, 10, and 20 µmol/L) or GW4064 (1, 5, 10, and 20 µmol/L) or vehicle control before analysis by flow cytometry for (A) changes in the mitochondrial membrane potential, determined by using JC-1 dye and expressing the results using the FL2/FL1 ratio. An increase in the ratio compared with control indicates hyperpolarization of the mitochondrial membrane, whereas a decrease in the FL2/FL1 ratio indicates membrane depolarization. B, Platelets were treated for 10 min with or without increasing concentrations of GW3965 (10 and 20 µmol/L) or GW4064 (10 and 20 µmol/L) or vehicle control in the presence of cyclosporine A (5 µmol/L), a cyclophilin D inhibitor, before analysis by flow cytometry for annexin V binding. Data expressed as percentage of annexin V–positive events. Thrombin-treated (0.1 U/mL)/CRP-treated (1 µg/mL) and ionomycin-treated (5 µmol/L) samples were included as positive controls. C and D, Ca²⁺ imaging in single platelets under flow. Human washed platelets loaded with Fluo4-AM (2 µmol/L) were attached onto mouse antihuman PECAM-1 antibody–coated (WM59) glass-bottom Vena8 GCS biochips (Cellix Ltd, Dublin, Ireland) and vehicle. C, GW3965 (20 µmol/L) or (D) GW4064 (20 µmol/L) were then flowed through the chips at a slow shear rate of 400 s⁻¹ and calcium signaling monitored by observing fluorescence for 5 min. Single platelet, representative calcium signaling trace shown. E and F, Platelets were treated with GW3965 or GW4064 (0, 10, and 20 µmol/L) in the presence or absence of (E) BAPTA (10 µmol/L) or (F) EGTA (1 mmol/L) before analysis by flow cytometry for (E) annexin V binding, and fibrinogen binding and (F) fibrinogen binding and P-selectin exposure. Thr (0.1 U/mL) and Thr (0.1 U/mL)+CRP (5 µg/mL) included as positive controls for fibrinogen binding in the presence of EGTA. Data expressed as median fluorescence intensity. Results are mean+SEM for n≥3 and expressed as percentage of vehicle control. *P<0.05 in comparison to vehicle controls.

Figure 5.

Liver X receptor (LXR) and farnesoid X receptor (FXR) ligands do not initiate apoptosis. A and B, Human washed platelets were pretreated with GW3965 or GW4064 (10 and 20 mmol/L) or ABT-263 (10 and 20 mmol/L; initiator of apoptosis) for 2 h before (A) lysis in mitochondria isolation lysis buffer. The heavy mitochondrial membrane was then isolated from its releasate by centrifugation, and the presence of cytochrome C in the mitochondrial releasate determined by Western blotting. B, Activation and cleavage of caspases was also determined in GW3965-treated or GW4064-treated (10 and 20 µmol/L) platelet lysates by Western blotting for the cleaved activated form of the caspases ABT-263 (20 mmol/L) treated platelets were included as a positive control for initation of apoptosis. C, Annexin V binding was determined in (i) GW3965-treated and (ii) GW4064-treated platelets after pretreatment with a caspase inhibitor Z-FAD-FMK (10 mmol/L). Representative blots are shown, and results are mean+SEM for n=3 and expressed as fold increase compared with control. *P<0.05 in comparison to vehicle controls.

Figure 6.

Generation of reactive oxygen species (ROS) is required for induction of procoagulant platelet activity by liver X receptor (LXR) and farnesoid X receptor (FXR) ligands. Human washed platelets were treated for 10 min with or without increasing concentrations of GW3965 (10 and 20 µmol/L) or GW4064 (10 and 20 µmol/L; A) in the presence of the fluorescent dye H2DCFDA (an indicator of ROS) and fluorescence measured. Thrombin-treated, CRP-treated, and ionomycin-treated platelets were included as positive controls. B and C, In the presence or absence of 10 µmol/L DPPD (n,n′-diphenyl-p-phenylenediamine), a reactive oxygen species scavenger before analysis by flow cytometry for (B) annexin V binding and (C) fibrinogen binding, using an antifibrinogen antibody. Results are mean+SEM for n≥3 and expressed as median fluorescence intensity. *P<0.05 in comparison to vehicle controls.