Succinate independently stimulates full platelet activation via cAMP and phosphoinositide 3-kinase-β signaling

Authors


Björn Olde, BMC D12, 22184 Lund, Sweden.
Tel.: +46 46 2227729; fax: +46 46 2113417.
E-mail: bjorn.olde@med.lu.se

Abstract

Summary. Background: The citric cycle intermediate succinate has recently been identified as a ligand for the G-protein-coupled receptor (GPCR) SUCNR1. We have previously found that this receptor is one of the most highly expressed GPCRs in human platelets. Objective: The aim of this study was to investigate the role of SUCNR1 in platelet aggregation and to explore the signaling pathways of this receptor in platelets. Methods and Results: Using real-time-PCR, we demonstrated that SUCNR1 is expressed in human platelets at a level corresponding to that of the P2Y1 receptor. Light transmission aggregation experiments showed dose-dependent aggregation induced by succinate, reaching a maximum response at 0.5 mm. The effect of succinate on platelet aggregation was confirmed with flow cytometry, showing increased surface expression of activated glycoprotein IIb–IIIa and P-selectin. Intracellular SUCNR1 signaling was found to result in decreased cAMP levels, Akt phosphorylation mediated by phosphoinositide 3-kinase-β activation, and receptor desensitization. Furthermore, succinate-induced platelet aggregation was demonstrated to depend on Src, generation of thromboxane A2, and ATP release. Platelet SUCNR1 is subject to desensitization through both homologous and heterologous mechanisms. In addition, the P2Y12 receptor inhibitor ticagrelor completely prevented platelet aggregation induced by succinate. Conclusions: Our experiments show that succinate induces full aggregation of human platelets via SUCNR1. Succinate-induced platelet aggregation depends on thromboxane A2 generation, ATP release, and P2Y12 activation.

Introduction

The dicarboxyl acid succinate is known to be a physiologically important intermediate in the citric acid cycle, where it plays a crucial role in mitochondrial energy production [1]. In recent years, it has become clear that succinate also functions as a ligand for the membrane receptor SUCNR1 [2]. The transcript for the human SUCNR1 receptor encodes a 330 amino acid protein that belongs to the genetic superfamily of G-protein-coupled receptors (GPCRs) [3]. SUCNR1 has been identified and quantified in many important organs throughout the human body; particularly high levels of SUCNR1 are present in the spleen, liver, testis, and kidneys [4–7]. When we examined mRNA expression of GPCRs in platelets using microarray and real-time PCR, the succinate receptor displayed the third highest expression level of all GPCRs examined [7], indicating an important role in platelet regulation.

More than two decades ago, Huang et al. [8] discovered that succinate had a potentiating effect on platelet-activating substances such asADP, epinephrine, and serotonin. Huang et al. [8] also concluded that high concentrations (5–10 mm) of succinate could induce primary platelet aggregation. The potentiating effect of succinate on ADP has also been described by Macaulay et al. [6]. Succinate is normally present systemically in concentrations that range between 1 and 20 μm [9], but, during pathophysiologic conditions such as ischemia and metabolic stress, plasma levels of succinate have been reported to reach millimolar concentrations [10–12].

Succinate is a multifunctional molecule with a physiologic potential that has not yet been fully evaluated. The aim of this study was to characterize the platelet succinate receptor SUCNR1 signaling pathways and to determine the role of this novel receptor in the aggregation process. This involved investigating the regulation and function of SUCNR1 in signaling pathways with known importance for platelet function. Our experiments demonstrate that succinate alone is capable of mediating full platelet aggregation by decreasing cAMP levels and activating the phosphoinositide 3-kinase (PI3K)β/Akt pathway, leading to P-selectin and glycoprotein (GP)IIb–IIIa activation. Furthermore, SUCNR1 is dependent on Src kinase signaling, and is subject to homologous as well as heterologous desensitization.

Materials and methods

Chemicals

Sodium succinate dibasic hexahydrate, 2′-deoxy-N6-methyladenosine-3′,5′-bisphosphate tetrasodium salt (MRS2179), EDTA, prostaglandin E1 (PGE1), apyrase, forskolin, prostaglandin I2, rhodamine-6G and Src-inhibitor-1 were from Sigma-Aldrich (St. Louis, MO, USA). 3,7-Dihydro-1-methyl-3-(2-methylpropyl)-1H-purine-2,6-dione, and trans-4-([1R]-1-aminoethyl)-N-4-pyridinylcyclohexanecarboxamide dihydrochloride (Y27632) were from Tocris Bioscience Ltd (Bristol, UK). 5-(2,2-Difluoro-benzo[1,3]dioxol-5-ylmethylene)-thiazolidine-2,4-dione (AS-604850) and 7-methyl-2-(4-morpholinyl)-9-(1-[phenylamino]ethyl)-4H-pyrido(1,2-a)pyrimidin-4-one (TGX221) were from Larodan Fine Chemicals AB (Malmö, Sweden). [35S]GTPγS was from Perkin Elmer (Waltham, MA, USA). Ticagrelor (AZD6140) was a gift from AstraZeneca R&D Mölndal (Mölndal, Sweden).

Platelet preparation for light transmission aggregation (LTA)

Whole blood was drawn from healthy volunteers through venous puncture (Vacutainer tubes, 0.129 m sodium citrate; [Becton Dickinson Inc., Franklin Lakes, NJ, USA]). Whole blood was centrifuged (140 × Gmax, 20 min, room temperature), and the supernatant containing platelet-rich plasma (PRP) was collected. As a control, platelet-poor plasma (PPP) was used, and prepared from whole blood by centrifugation (2260 × Gmax, 10 min, room temperature). Each PRP sample contained 250 μL. The LTA measurements were performed with two serially connected Chronolog aggregometers, model 490 (Chronolog, Havertown, PA, USA). Antagonists were preincubated with PRP for 5 min prior to addition to succinate. Aggregation values from LTA experiments are presented as percentage of the mean of maximum platelet aggregation (MPA). Washed platelets were prepared according to the protocol by Cazenave et al. [13] before aggregation studies, as described for PRP.

Platelet quantitative real-time PCR

See Data S1 and reference [7].

Western blotting on human platelets for SUCNR1

See Data S1.

Confocal immunofluorescence microscopy on platelets

See Data S1.

Calcium flux assay on platelets

Calcium release, mediated through succinate stimulation, was measured as described previously [14], with one change in the protocol for the P2Y1 assay: 0.3 U mL−1 apyrase was added to the assay buffer to exclude indirect signaling from P2Y1. Fluorescence reading was performed with a Victor3 1420 Multilabel Counter (Perkin Elmer).

cAMP assay on platelets

See Data S1.

Flow cytometric analysis of platelet activation

See Data S1.

Platelet degranulation – ATP release

PRP was prepared as for the LTA measurement, and was incubated at 37 °C for 10 min (1200 rpm) before addition of the ligand. The released ATP was determined by luminometry in a GloMax TD-20/20 luminometer (Promega Biotech Inc., Madison, WI, USA), with the ATP SL kit (BioThema Inc, Handen, Sweden).

Western blot assay for protein kinase B (PKB)α/Akt1 determination

See Data S1 and reference [15].

Platelet [35S]GTPγS binding

PRP was prepared from acid–citrate–dextrose-treated whole blood through centrifugation (150 × Gmax, 20 min, room temperature) and passed carefully through a Pall Autostop leukocyte removal filter (Pall Medical Inc., Port Washington, NY, USA). The filtered PRP was centrifuged (17 000 × Gmax, 15 min, room temperature). The pellet was resuspended in a room temperature membrane buffer (20 mm HEPES, 20 mm EDTA, 150 mm NaCl, pH 7.4, room temperature) and the centrifugation was repeated once. The pellet was resuspended in 3 mL of ice-cold membrane buffer (20 mm HEPES, 5 mm EDTA, pH 7.4), homogenized for 30 s in a Polytron (setting 6), and diluted to 30 mL with ice-cold membrane buffer (20 mm HEPES, 5 mm EDTA, pH 7.4). The platelet homogenate was pelleted twice (35 000 × Gmax, 15 min, 4 °C). The final pellet was resuspended in ice-cold membrane buffer (20 mm HEPES, 0.6 mm EDTA, 5 mm MgCl2, 0.1 mm phenylmethanesulfonyl fluoride, pH 7.4) and stored at − 80 °C. Protein concentration was measured with a modified Lowry protein assay kit (Thermo Fisher Scientific, Waltham, MA, USA).

The binding assay was performed essentially as described by Vasiljev et al. [16]. Briefly, the final volume of the assay was 200 μL and contained 20 μg of platelet membranes in binding buffer (20 mm HEPES, 100 mm NaCl, 5 mm MgCl2, 10 μg mL−1 saponin, pH 7.4, 4 nm [35S]GTPγS, 20 μm GDP). The mixture was incubated (30 min, 30 °C) and then filtered through a glass fiber filter (GF/C) (Whatman International Ltd, Springfield Mill, UK). The filters were washed with 3 × 500 μL of ice-cold binding buffer, and then counted with an OptiPhase Hisafe 3 (Perkin Elmer) in an LS 6500 Multi-Purpose Scintillation Counter (Beckman Coulter Inc., Brea, CA, USA).

Statistical analysis

Calculations and statistical analysis were performed with GraphPad Prism 4.0 software (GraphPad Software Inc., La Jolla, CA, USA). One-way anova was used, followed by a Bonferroni correction between groups if there was statistical significance. Student’s t-test was used to compare means between two groups. Statistical significance was accepted at < 0.05. Values are presented as means ± standard errors of the mean; n = number of experiments.

Ethics

The project was approved by the Ethics Committee of Lund University, Sweden. All participants submitted written consent to take part in the study. The study conformed to US National Institutes of Health guidelines, and the Declaration of Helsinki.

Results

Quantification of SUCNR1 in human platelets

The succinate receptor SUCNR1 is expressed in human platelets at approximately the same level as P2Y1 (1.28 vs. 1.00, n = 6). The expression level of P2Y12 was about 12-fold higher than the expression level of SUCNR1 (n = 6), whereas the expression of A2a was approximately 10-fold lower (n = 6) (Fig. 1A). The expression of SUCNR1 on human platelets was confirmed by western blot (Fig. 1B) and confocal immunofluorescence microscopy (Fig. 1C). The platelet expression of SUCNR1 was further evaluated by flow cytometry (21.3% of platelets positive; data not shown); CHO cells, which have been previously used to characterize the SUCNR1 receptor, displayed negligible levels of SUCNR1 antibody binding (0.97% positive; data not shown). CHO cells stained with the SUCNR1 antibody did not display any fluorescence as evaluated by confocal microscopy (see Data S1).

Figure 1.

 (A) Quantitative real-time PCR analysis of SUCNR1, P2Y12 and A2a expression levels in platelets. The P2Y1 gene was used as reference gene. (B) Expression of SUCNR1 in platelets (Plt). SUCNR1 (38 kDa) was detected in human platelets by western blot. Caki-1 and CHO cells were used as positive and negative controls, respectively. Glyceraldehyde-3-phosphate dehydrogenase (GAPDH) was used as a loading control. Data are representative of two experiments. (C) Representative confocal image of SUCNR1 expression (green) in human washed platelets stained with rhodamine-6G (red). Right panel; merged images. Bar: 10 μm. MW, molecular weight marker.

Succinate stimulates platelet aggregation

Huang et al. [8] have previously demonstrated that succinate potentiates agonist-induced platelet aggregation. Using LTA, we tested whether succinate alone is sufficient to activate platelets. Our results clearly demonstrate that succinate independently activates platelets and causes platelet aggregation in a dose-dependent manner. Typical succinate aggregation displays biphasic aggregation kinetics. The effect of succinate was half-maximal at 0.3 mm and maximal at 0.5 mm. To exclude possible disturbances in the serum, we isolated washed platelets and repeated the experiment. This resulted in a displacement of the dose–response curve to the right (Fig. 2A). We also confirmed the potentiating effect of succinate on ADP-mediated platelet activation, which has been described previously (Fig. 2B). The effect of succinate on platelet activation was further characterized by demonstrating that both the platelet activation marker for GPIIb–IIIa, PAC-1, and the degranulation marker, P-selectin, were activated (% positive cells) (23.0 ± 3.3 vs. 6.9 ± 2.0, P < 0.001, n = 16; and 7.6 ± 1.5 vs. 2.7 ± 0.5, P < 0.05, n = 16) in response to 3 mm succinate (Fig. 3A,B). ADP was added as a positive marker both for PAC-1 binding, (58.5 ± 7.0 vs. 6.9 ± 2.0, P < 0.01, n = 16) and for P-selectin, (29.1 ± 5.1 vs. 2.7 ± 2.0, P < 0.001, n = 16).

Figure 2.

 (A) Upper left: dose–response curve of platelet platelet-rich plasma (PRP) aggregation mediated through elevated doses of succinate. The aggregation time was 6 min; data were collected from six experiments. Upper right: representative aggregation curves from each experiment. Lower left: dose–response curve presenting aggregation of washed platelets mediated by elevated doses of succinate. The aggregation time was delayed, as a result of the washing procedure, to 16 min: data were collected from six experiments. Lower right: representative aggregation curves from each experiment. (B) Potentiating effect of succinate on ADP-mediated platelet activation in washed platelets (anova: **P < 0.05, ***P < 0.001, n = 5). MPA, maximum platelet aggregation.

Figure 3.

 (A) Results from a flow cytometry study on glycoprotein (GP)IIb–IIIa activation in platelets. The PAC-1 antibody was used to detect platelet with activated GPIIb–IIIa. There was a marked difference between unstimulated platelets and platelets stimulated with 3 mm succinate or 5 μm ADP (anova: **P < 0.01, n = 16). (B) Flow cytometry study on P-selectin release from platelets. There was a statistically significant difference in vesicle secretion of α-granules mediated by succinate between untreated platelets and platelets treated with 3 mm succinate or 5 μm ADP (anova: *P < 0.05, ***P < 0.001, n = 16). (C) Effects of succinate on forskolin-induced and prostaglandin E1 (PGE1)-induced cAMP production in platelets; 1 × 108 washed platelets were used per sample. Significant differences in cAMP production were observed between platelets pretreated with 3 mm succinate and 10 μm forskolin vs. 10 μm forskolin (anova: ***P < 0.001, n = 4) and platelets pretreated with 3 mm succinate and 100 nm PGE1 vs. 100 nm PGE1 (anova: *P < 0.05, n = 4). (D) Flow cytometry study on platelets, measuring the ratio of phosphorylated/dephosphorylated vasodilator-stimulated phosphoprotein (VASP). The left bar graph represents the dephosphorylating effect of 3 mm succinate on VASP on PGE1-stimulated platelets (Student’s t-test: **P = 0.0025, n = 8). The right bar graph represents the dephosphorylating effect of ADP on VASP on PGE1-stimulated platelets (Student’s t-test: ***P = 0.0002, n = 8). (E) Calcium signaling in platelets. Left bar graph: data are presented as relative fluorescence units (RFU). There was no significant difference between platelets treated with succinate and control (anova: NS = P > 0.05, n = 4). There was a significant difference between platelets stimulated with thrombin and control (anova: ***P < 0.001, n = 4). A difference was also seen between ADP and control (anova: ***P < 0.001, n = 4). The right graph illustrates peak values from experiments: bsl00001, vehicle (control); ○, 3 mm succinate; •, 10 μm ADP; and □, 1 U mL−1 thrombin.

Activation of SUCNR1 inhibits cAMP production in platelets but has no effect on calcium signaling

As the succinate receptor has been reported to be coupled to Gαi activation [9], and is thus expected to result in an attenuation of platelet cAMP production, we tested the effect of 3 mm succinate on both forskolin-stimulated and PGE1-stimulated cAMP production, and also on the phosphorylation status of the cAMP/protein kinase A (PKA) marker vasodilator-stimulated phosphoprotein (VASP). We found that cAMP production in platelets prestimulated by forskolin was significantly reduced for platelets treated with succinate (2.6 μm ± 66 nm vs. 1.0 μm ± 70 nm, P < 0.001, n = 4). This reduction in cAMP production was also seen in platelets prestimulated with PGE1 (0.7 μm ± 30 nm vs. 0.4 μm ± 7 nm, P < 0.001, n = 4) (Fig. 3C). The decrease in cAMP production was confirmed by a corresponding decrease in VASP phosphorylation (% positive cells) (75.4 ± 6.3 vs. 66.7 ± 7.0, P = 0.0025, n = 8) activation of P2Y12, by ADP, resulted in a similar decrease in VASP phosphorylation (68.1 ± 7 vs. 23.1 ± 4.7, P = 0.0002, n = 8) (Fig. 3D).

Succinate (3 mm) stimulation of platelets loaded with Fluo-4AM did not result in any change in calcium mobilization as compared with untreated platelets (5284 ± 6.4 relative fluorescence units (RFU) vs. 4963 ± 6.2 RFU, (not significant, P > 0.05, n = 4) (Fig. 3E). Thrombin (1 U mL−1) elicited a positive calcium response vs. control (7563 ± 185 RFU vs. 4963 ± 6.2 RFU, P > 0.001, n = 4), thus demonstrating the functional integrity of the platelets. In order to avoid indirect ADP signaling, mediated through P2Y1 or caused by ATP release, the experiment was conducted in the presence of apyrase. However, a minor difference could be detected between platelets stimulated with 10 μm ADP and the control (5578 ± 7.7 RFU vs. 4963 ± 6.2 RFU, P > 0.001, n = 4).

Succinate-stimulated aggregation is dependent on PI3Kβ activation but not on the Gα12/13 pathway

As both the PI3K/Akt and the Gα12/13 pathway are involved in mediating platelet aggregation, we used the PI3K inhibitor wortmannin and the Rho-GEF/ROCK160-inhibitor Y27632 to explore the involvement of these pathways in succinate-stimulated aggregation. Whereas 10 μm Y27632 had no effect, there was a significant decrease in succinate-stimulated aggregation after preincubation of platelets with 10 μm wortmannin (75.7 ± 0.8 vs. 1.8 ± 0.7, P < 0.0001, n = 6), indicating involvement of the PI3K/Akt pathway (Fig. 4A,B).

Figure 4.

 (A) Platelet-rich plasma (PRP) pretreated with 10 μm of the Rho-GEF/ROCK160-inhibitor, Y27632 (blocking G12/13 pathways) and stimulated with 3 mm succinate vs. PRP stimulated with 3 mm succinate only (Student’s t-test: NS, P = 0.6, n = 4). (B) PRP pretreated with 10 μm wortmannin and stimulated with 3 mm succinate vs. PRP stimulated with 3 mm succinate only (Student’s t-test: ***P < 0.001, n = 6). The light transmission aggregation (LTA) curves accompanying the bar charts represent the characteristic aggregation patterns from the different experiments. (C) LTA on PRP preincubated with AS-604850 (PI3Kγ inhibitor) and TGX221 (PI3Kβ inhibitor). Samples were further stimulated with 3 mm succinate. Platelets preincubated with 1 μm AS-604850 showed a moderate decrease in aggregation compared to control (anova: *P < 0.05, n = 8). However, platelets preincubated with 1 μm TGX221 displayed a major reduction in succinate stimulated aggregation (anova: ***P < 0.001, n = 8). (D) Western blot (including bar chart) of protein kinase B (PKB)α/Akt1 on washed platelets treated with 1 μm of the PI3Kγ inhibitor AS-604850, or 1 μm of the PI3Kβ inhibitor TGX221. All platelet samples were stimulated with 3 mm succinate, except for unstimulated platelets, which represent the negative control. (anova: ***P < 0.001, **P < 0.01, n = 3). (E) Inhibition of Src kinase by 10 μm Src-inhibitor-1 on 3 mm succinate-stimulated PRP (Student’s t-test: ***P < 0.0001, n = 7).

In order to identify which PI3K isoform is involved in succinate-induced platelet aggregation, we employed TGX221 and AS-604850, which are specific inhibitors of PI3Kβ and PI3Kγ, respectively. Ten microliters of MRS2179 was added in all experiments to avoid potential interference of the P2Y1 receptor. Stimulation of platelets with 3 mm succinate in the presence of 1 μm AS-604850 resulted in a small, but significant, reduction in MPA (72.1 ± 1.1 vs. 77.1 ± 1.5, P < 0.05, n = 8). However, when AS-604850 was substituted for 1 μm of the PI3Kβ-specific TGX221, there was a 99% reduction in MPA as compared with the control (77.1 ± 1.5 vs. 0.6 ± 0.3, P < 0.001, n = 8) (Fig. 4C). The involvement of PI3Kβ was confirmed by western blot analysis of Akt phosphorylation. Treatment of platelets with 3 mm succinate resulted in Akt phosphorylation that was reversed when 1 μm TGX221 was added (Fig. 4D).

The ability of succinate to achieve platelet aggregation in the LTA experiment was abolished in the presence of an Src kinase inhibitor. A statistically significant difference was seen between platelets preincubated with 10 μm Src-inhibitor-1 and those stimulated with 3 mm succinate only (74.7 ± 1.4 vs. 7.0 ± 1.2, P < 0.0001, n = 7) (Fig. 4E).

Succinate-stimulated platelet aggregation is dependent on ATP release

Acetyl salicylic acid (ASA) inhibits cyclooxygenase-1, which converts arachidonic acid to thromboxane A2 (TXA2). Pretreatment with 1 mm ASA abolished the stimulating effect of 3 mm succinate in platelets (73.5 ± 5.2 vs. 6.9 ± 1.2, P < 0.05, n = 14), indicating a role for TXA2 in succinate-mediated aggregation (Fig. 5A). As the response of platelets to TXA2 depends on ADP serving as a positive-feedback mediator, which is required for sustained activation, these results prompted us to investigate the effect of succinate on ATP release. We found that 3 mm succinate elicited an approximately 1000-fold increase in ATP release (0.3 ± 0.1 nm vs. 0.4 μm ± 1.6 nm, P < 0.0001, n = 6), confirming the involvement of ATP release in succinate-mediated aggregation (Fig. 6). These results were further strengthened by the demonstration that the proactivating effect induced by 3 mm succinate was absent in platelets preincubated with 10 μm of the potent P2Y12 antagonist ticagrelor (79.7 ± 0.9 vs. 4.0 ± 0.7, P < 0.05, n = 12) (Fig. 5B). Platelets pretreated with 10 μm of the P2Y1 antagonist MRS2179, on the other hand, did not display any reduction in MPA (76.2 ± 1.0 vs. 74.4 ± 1.0 [not significant], P = 0.25, n = 14), (Fig. 5C).

Figure 5.

 (A) Succinate-stimulated platelet-rich plasma (PRP) vs. PRP pretreated with 1 mm acetyl salicylic acid (ASA) and further stimulated with succinate. There was a statistically significant difference between succinate-stimulated platelet aggregation in the presence of ASA and that in its absence (Student’s t-test: ***P < 0.0001, n = 14). (B) P2Y12 dependence of succinate-stimulated platelet aggregation. A significant difference was found between PRP stimulated with 3 mm succinate and PRP pretreated with 10 μm of the P2Y12 inhibitor ticagrelor and further stimulated with succinate (Student’s t-test: ***P < 0.0001, n = 13). (C) P2Y1 dependence of succinate-stimulated platelet aggregation. No statistical difference was seen between untreated PRP and PRP pretreated with the P2Y1 antagonist MRS2179 (Student’s t-test: NS, P = 0.25, n = 16). The light transmission aggregation curves accompanying the bar charts represent the characteristic aggregation patterns from the different experiments. MPA, maximum platelet aggregation; NS, not significant.

Figure 6.

 Succinate-mediated ATP release. Succinate is capable of stimulating ATP release in platelet-rich plasma. Student’s t-test: ***P < 0.0001, n = 6.

Succinate stimulates [35S]GTPγS binding to platelet cell membranes

As the demonstration of functional SUCNR1 receptors by standard radio-receptor binding techniques is impractical, owing to the low affinity of succinate, we investigated the ability of succinate to stimulate [35S]GTPγS binding in platelet plasma membranes. As shown in Fig. 7, increasing concentrations of succinate generated dose-dependent binding of [35S]GTPγS with an EC50 value of 100 μm (n = 4). Desensitization of the receptor upon ligand stimulation is a well-studied negative regulation mechanism of GPCRs. In order to investigate whether platelet SUCNR1 is subject to homologous or to heterologous desensitization, we prepared plasma membranes from platelets that were prechallenged with either 3 mm succinate or 10 μm 2MeSADP. The results showed that, in both cases, the ability of succinate to stimulate [35S]GTPγS binding was significantly reduced to 69.6% ± 3.4% (P < 0.01, n = 3) for homologous desensitization, and to 81.6% ± 2.9% (*P < 0.001, n = 3) for heterologous desensitization (Fig. 7).

Figure 7.

 Dose response of succinate-stimulated [35S]GTPγS binding in platelet membranes. EC50 values are for 110 μm untreated platelet membranes (bsl00001), platelet membranes pretreated with 10 μm 2MeSADP (○), and platelet membranes pretreated with 3 mm succinate (•). The maximal response, obtained for membranes pretreated with either 10 μm 2MeSADP or 3 mm succinate, was significantly reduced as compared with untreated membranes (anova: **P < 0.001, *P < 0.01, n = 3).

Discussion

In this study, we identified succinate as an independent agonist of full platelet aggregation. Succinate-stimulated platelet aggregation involves activation of the platelet activation marker GPIIb–IIIa and the degranulation marker P-selectin. Succinate stimulation of platelets resulted in decreased cAMP levels and VASP activation, but did not stimulate cytosolic calcium release. Furthermore, succinate-induced platelet aggregation depended on PI3Kβ/Akt1-induced signaling and on Src kinase, but did not involve Rho-GEF or PI3Kγ. SUCNR1 was desensitized both homologously by succinate and heterologously by ADP. Finally, our data showed that succinate induced platelet aggregation via TXA2 generation and ATP/ADP release. Interestingly, we observed complete abolition of succinate-mediated platelet aggregation in the presence of the P2Y12 receptor antagonist ticagrelor, whereas the P2Y1 antagonist MRS2179 had no such effect.

The influence of the citric acid cycle metabolite succinate on platelets was first described by Huang et al. [8], who reported that succinate augmented the actions of ADP and epinephrine, and that this effect was related to attenuation of cAMP levels. In a previous study, using gene array screening, we identified the message of the recently deorphanized succinate receptor SUCNR1 in human platelet mRNA. In an effort to confirm these results, we quantified the platelet message of SUCRN1. Interestingly, the level of SUCRN1 expression was the third highest among the GPCRs examined, and equaled the expression level of P2Y1 receptor, a receptor with a demonstrated role in platelet activation, thus predicting functionality at the same level.

The succinate receptor is reported to be a low-affinity receptor with an EC50 value for succinate that, depending on the assay method, ranges from 28 to 391 μm. These results are well in line with our findings, with an EC50 of approximately 300 μm based on LTA measurements and of 100 μm for [35S]GTPγS binding. Succinate at 3 mm activated the marker PAC-1, recognizing activated GPIIb–IIIa, and the degranulation marker P-selectin, thus supporting the picture of an independent platelet activator. Huang et al. [8] reported that succinate treatment resulted in a decrease in platelet cAMP levels. These results correspond well with our findings that inhibition of the cAMP/PKA pathway resulted in decreased cAMP levels and, as a consequence, a decrease in PKA-mediated phosphorylation of VASP. Phosphorylated VASP inhibits activation of the membrane-bound integrin GPIIb–IIIa [17], which is an important step for platelet–platelet adhesion, and hence for aggregation of platelets.

In contrast to the report by He et al. [4], we were unable to detect any effect on cytosolic calcium mobilization. In this study, SUCNR1 is described as a receptor capable of coupling not only to Gαi, mediating inhibition of adenylate cyclase, but also to Gαq, stimulating inositol trisphosphate (IP3) turnover and calcium release. However, only a part of the IP3 signal could be attributed to Gαq activation, as the other half was sensitive to pertussis toxin (PTX), and was thus probably caused by Gβγ-mediated activation of phospholipase C (PLC)β [18]. Furthermore, in a recent study by Hakak et al. [9], in which endogenous succinate receptors were investigated in TF-1 cells, the PTX-insensitive part of the IP3 signal was absent. Those authors suggest that, as the results of He et al. [4] were obtained in recombinant CHO cells overexpressing SUCNR1, it is likely that the PTX-insensitive signal is an artefact, caused by promiscuous G-protein coupling. Our results support the view of SUCNR1 as a receptor primarily coupled to Gαi, and we believe that the absence of a succinate-stimulated calcium signal in platelets can be explained by the lack of Gβγ-mediated activation of PLCβ in platelets. There are reports that support the absence of Gβγ-mediated calcium signaling in platelets: for example, it has been demonstrated that, whereas stimulation of the P2Y12 receptor results in Gαi-dependent activation of PI3K, it will not result in calcium mobilization [19]. Although it is currently not known which Gβγ combinations are expressed in platelets, it has been shown that the type of Gβ subunit determines whether PI3K or PLCβ activation is favored [20,21].

Although we could detect no effect on calcium mobilization, we obtained a dose-dependent increase in [35S]GTPγS binding upon stimulation with succinate, thus confirming the involvement of Gαi. Furthermore, the importance of Gβγ-mediated signaling was confirmed by strong PI3K activation resulting in Akt/PKB phosphorylation. The responsible isoform turned out to be PI3Kβ, which is somewhat surprising, as PI3Kγ is the isoform that is usually associated with stimulation by Gβγ. However, PI3Kβ has previously been identified as the dominant PI3K isoform responsible for Gi-mediated GPIIb–IIIa activation following ADP stimulation of the P2Y12 receptor [22]. The Src kinases have recently been reported by Nash et al. [23] to be involved in adrenergic Gi-coupled signaling in platelets. Nash et al. reported that Src kinases acting downstream of the α2a receptor mediate both the primary and secondary waves of aggregation. They also demonstrated that P2Y12-mediated aggregation is dependent on Src kinases, and suggested that this pathway is common for Gi-coupled receptors.

This finding led us to investigate whether Src kinases could be involved in SUCNR1 activation as well. Indeed, we found that platelet aggregation, induced by succinate, was nearly abolished, from approximately 75% to 7% MPA, in the presence of an Src kinase inhibitor.

The importance of positive feedback was demonstrated by the finding that succinate stimulated massive ATP/ADP release, confirmed both by the sensitivity to the P2Y12 antagonist ticagrelor and by the increase in the vesicular–inner membrane-bound protein P-selectin. Negative regulation of GPCRs is triggered by phosphorylation, leading to desensitization and internalization of the receptor. Our results demonstrate that platelet SUCNR1 is rapidly desensitized upon challenge with succinate. Previously, Robben et al. [24] showed that SUCNR1 is similarly desensitized in Madin–Darby canine kidney cells.

Interestingly, platelet SUCNR1 also seems to be subject to heterologous desensitization through activation of either P2Y12 and/or P2Y1. Hardy et al. [25] have previously reported that the desensitization mechanisms triggered by P2Y1 and by P2Y12 differ. Whereas P2Y12 is desensitized through a homologous mechanism involving G-protein-coupled receptor kinase (GRK) mediated phosphorylation and receptor internalization, P2Y1 desensitization is, to a great extent, dependent on protein kinase C, which phosphorylates ligand-bound and inactive GPCRs in a heterologous manner [26]. Thus, it seems likely that the observed cross-desensitization of SUCNR1 by 2MeSADP is caused by stimulation of P2Y1 rather than of P2Y12.

When discussing the physiologic implications of succinate-mediated platelet activation, it is important to note that the succinate levels in serum are generally too low to permit activation of SUCNR1 [27]. However, under pathologic conditions such as animal models of type 2 diabetes and during different forms of ischemia, succinate concentrations have been reported to rise to millimolar ranges [10,11]. The potential importance of succinate/SUCNR1 in cardiovascular disease was highlighted in a recent publication showing that succinate, at concentrations equal to ischemic levels, modulates apoptosis in rat ventricular cardiomyocytes [28], and that this effect is mediated by SUCNR1. Moreover, succinate has been suggested to be a myocardial marker of ischemia–reperfusion injury [29]. Thus, succinate could be important during myocardial infarction, both via apoptotic effects on the cardiomyocyte and by causing platelet aggregation and thereby epicardial vessel occlusion or microembolism. As succinate is ubiquitous in all cell types, the physiologic role could potentially be very important. However, to prove this hypothesis, selective succinate antagonists are needed for the performance of in vivo experiments or even clinical studies.

As for many platelet activators, succinate activation is dependent on the positive-feedback system of thromboxane and ADP for full platelet aggregation. Succinate-induced platelet aggregation was inhibited by aspirin and, interestingly, also by the potent P2Y12 receptor antagonist ticagrelor. Ticagrelor was recently tested in the phase III study PLATO, where it was superior to clopidogrel and reduced both myocardial infarctions and mortality [30]. It is possible that some of the beneficial effects of aspirin and ticagrelor are attributable to inhibition of succinate-induced platelet aggregation.

In conclusion, the succinate receptor is one of the most highly expressed receptors on human platelets. It can stimulate platelet aggregation by itself and increase P-selectin and GPIIb–IIIa expression. The effects are mediated via the cAMP–PKA–VASP pathway, but also depend on Src kinase activation and PI3Kβ/Akt1-induced signaling. Succinate-induced platelet activation can be prevented with aspirin or ticagrelor, and this observation may explain some of their beneficial effects. It is possible that selective succinate receptor blockers could have further effects in preventing thrombosis, and could have a place in the treatment of cardiovascular disease.

Addendum

The authors had full access to and take full responsibility for the integrity of the data. All authors have read and agreed to the manuscript as written.

Acknowledgements

The study was supported by the Swedish Heart and Lung Foundation, Swedish Scientific Research Council, ALF and Lund University Hospital funds.

Disclosure of Conflict of Interests

The authors declare that they have no conflict of interest.

Ancillary