Evidence that MRP4 is Only Partly Involved in S1P Secretion during Platelet Activation

Sphingosine-1-phosphate (S1P) is a lipid metabolite involved in several cell processes such as motility, survival, proliferation, vascular development and inflammation. S1P cell content results from its synthesis by phosphorylation of sphingosine by sphingosine kinases 1 and 2[1] and its dephosphorylation by S1P lyase. S1P is present in the blood[2] and its local concentration strongly increases upon platelet activation and secretion.[3] Yet, we have recently shown that platelets do not contribute to plasma S1P level under homeostasis but that S1P level in serum is mostly due to platelet secretion.[4] However, the way by which S1P may impact platelet activation is still under debate, as well as its secretion mechanism.

We read with great interest the article of Vogt et al in a recent issue of Thrombosis and Haemostasis.[5] Their results on a multi-drug resistance protein 4 (MRP4) knockout mouse model suggest that S1P release from agonist-activated platelets mainly depends on the transporter MRP4 (ABCC4). The sub-cellular localization of MRP4 is still under debate since it was first described in dense granule membrane[6] and then at the plasma membrane.[7] Vogt et al described MRP4 as part of granular structures. They also showed, such as others, a relocation of MRP4 from intracellular structures to the plasmatic membrane when platelets are activated.[8]

Our group has been working on the role of MRP4 in platelet functions,[9] [10] and has now explored its role on S1P secretion. One can hypothesize that the role of MRP4 in platelet S1P distribution would be separated into two discrete stages. In resting platelets, MRP4 would transport S1P from the cytosol into the granules, allowing its secretion with other granule components upon platelet activation. Then, following platelet activation and MRP4 re-localization at the plasmatic membrane, MRP4 might carry on its efflux function, leading to the export of cytosolic S1P to the extracellular medium.

To gain further insight in platelet MRP4 role, we first addressed the circulating levels of S1P in a MRP4^−/− mouse model, such as performed by Vogt et al. First, we found similar S1P plasma concentrations in wild-type (WT) and MRP4^−/− mice (240 nM [95% confidence interval [CI], 177–282] vs. 227 nM [95% CI, 209–247], respectively, p > 0.05; [Fig. 1A]). These results are in line with those of Vogt et al. Then, Vogt et al adjusted S1P level to platelet count to conclude that MRP4^−/− mice had reduced plasma S1P concentration. However, it has been shown by Gazit et al that a total absence of S1P in platelet does not impact its plasmatic level and that erythrocytes are one of the main sources of circulating S1P.[4] Therefore, we considered useless to adjust plasma S1P level to platelet count.

Fig. 1 Role of multi-drug resistance protein 4 (MRP4) in sphingosine-1-phosphate (S1P) secretion. (A) S1P concentrations as determined by liquid chromatography-tandem mass spectrometry (LC-MS/MS) in plasma and serum from wild-type (WT) and MRP4^−/− mice (n ≥ 3 for plasma; n = 7 for serum). (B) S1P assays were performed on platelet supernatants after 400 μM protease-activated receptor 4 (PAR4)-agonist-peptide activation for 10 minutes, in the presence or in the absence of MK571 (100 μM). Platelet activation was stopped by adding one-fifth volume of ice-cold 0.1 M ethylenediaminetetraacetic acid (EDTA) and immediate centrifugation at 12,000 × g for 2 minutes at 4°C (n ≥ 14 in absence of MK571; n ≥ 7 in presence of MK571).

Next, we studied the role of platelet MRP4 on S1P serum level, not evaluated by Vogt et al. S1P level was found significantly higher in the serum from WT compared with MRP4^−/− (463 nM [95% CI, 420–499] vs. 378 nM [95% CI, 325–419], respectively, p < 0.01; [Fig. 1A]). Noteworthy, a significant S1P increase was observed in the serum of both genotypes (p < 0.01 compared with plasma; [Fig. 1A]). Altogether, these results argue for a participation of MRP4 in the serum S1P level upon platelet secretion. To confirm this role of MRP4, we quantified in the supernatant the amount of S1P secreted upon strong washed-platelet activation (protease-activated receptor 4 [PAR4]-ap 400 μM). As expected and as demonstrated by Vogt et al, S1P was found significantly higher (∼37%) in WT platelet supernatant compared with MRP4^−/− (30.6 nM/10⁹ platelets [95% CI, 16.4–33.5] vs. 17.2 nM/10⁹ platelets [95% CI, 10.0–21.8], respectively, p < 0.01) ([Fig. 1B]). Moreover, we did not detect S1P in resting platelet supernatant, and total S1P in washed platelet lysates did not differ between WT and MRP4^−/− (102.5 nM/10⁹ platelets [95% CI, 97.1–108.3] vs. 105 nM/10⁹ platelets [95% CI, 94.0–116.6], respectively). Thus, such as in Vogt et al study, we found that the total amount of platelet S1P is not secreted during platelet activation. In contrast to their findings in favour of a S1P secretion process mainly dependent on MRP4 during platelet activation, we found that S1P secretion is only partly dependent on this transporter. That MRP4 would not be the only actor of S1P platelet secretion is in line with a recent publication of Vu et al, suggesting that S1P export in resting and activated platelets depends in part on Mfsd2b.[11]

We repeated this experiment in platelets pre-incubated with MK571, an inhibitor of MRP4. MK571 did not influence S1P secretion either from WT or MRP4^−/− platelets. Therefore, inhibition of MRP4 transport activity 10 minutes before platelet activation does not impact S1P release. We propose that MRP4 would be mostly responsible of S1P storage in granules. Such as the other granule components, MRP4-dependent S1P secretion would better result from the fusion of the granule membrane with the plasma membrane, being no longer dependent on MRP4. Interestingly, this model of two pools of S1P with one requiring granule release dependent on platelet activation was previously proposed by Jonnalagadda et al.[12]

Second question we addressed was if the decreased S1P secretion we observed in MRP4^−/− model could be responsible, at least in part, of the loss of platelet function in these mice.[9] Indeed, previous data suggested a positive, although weak, role of S1P on platelet activation and aggregation.[4] MRP4^−/− platelet aggregation was reduced by 65% compared with WT (63.4% [95% CI, 59.3–67.9] vs. 22.4 [95% CI, 10.6–72.1] for WT and MRP4^−/−, respectively). Interestingly, it was still reduced by 76% in the presence of S1P (61.2 [95% CI, 32.2–79.4] vs. 14.3 [95% CI, 3.9–84.0]). Therefore, the addition of exogenous S1P did not allow recovering a normal aggregation of MRP4^−/− platelets in response to PAR4-ap. In these conditions, S1P secretion independent of MRP4 in MRP4^−/− platelets is likely to be sufficient to play its role in the amplification of platelet activation.

In conclusion, we propose that (1) MRP4 is more likely involved in S1P storage in platelet granules than in secretion upon platelet activation, and (2) this MRP4-dependent S1P secretion has not a major effect on platelet functions.

• References

• 1 Chalfant CE, Spiegel S. Sphingosine 1-phosphate and ceramide 1-phosphate: expanding roles in cell signaling. J Cell Sci 2005; 118 (Pt 20): 4605-4612
• 2 Okajima F. Plasma lipoproteins behave as carriers of extracellular sphingosine 1-phosphate: is this an atherogenic mediator or an anti-atherogenic mediator?. Biochim Biophys Acta 2002; 1582 (1-3): 132-137
• 3 Yatomi Y, Ruan F, Hakomori S, Igarashi Y. Sphingosine-1-phosphate: a platelet-activating sphingolipid released from agonist-stimulated human platelets. Blood 1995; 86 (01) 193-202
• 4 Gazit SL, Mariko B, Thérond P. , et al. Platelet and erythrocyte sources of S1P are redundant for vascular development and homeostasis, but both rendered essential after plasma S1P depletion in anaphylactic shock. Circ Res 2016; 119 (08) e110-e126
• 5 Vogt K, Mahajan-Thakur S, Wolf R. , et al. Release of platelet-derived sphingosine-1-phosphate involves multidrug resistance protein 4 (MRP4/ABCC4) and is inhibited by statins. Thromb Haemost 2018; 118 (01) 132-142
• 6 Jedlitschky G, Tirschmann K, Lubenow LE. , et al. The nucleotide transporter MRP4 (ABCC4) is highly expressed in human platelets and present in dense granules, indicating a role in mediator storage. Blood 2004; 104 (12) 3603-3610
• 7 Cheepala SB, Pitre A, Fukuda Y. , et al. The ABCC4 membrane transporter modulates platelet aggregation. Blood 2015; 126 (20) 2307-2319
• 8 Mattiello T, Guerriero R, Lotti LV. , et al. Aspirin extrusion from human platelets through multidrug resistance protein-4-mediated transport: evidence of a reduced drug action in patients after coronary artery bypass grafting. J Am Coll Cardiol 2011; 58 (07) 752-761
• 9 Decouture B, Dreano E, Belleville-Rolland T. , et al. Impaired platelet activation and cAMP homeostasis in MRP4-deficient mice. Blood 2015; 126 (15) 1823-1830
• 10 Belleville-Rolland T, Sassi Y, Decouture B. , et al. MRP4 (ABCC4) as a potential pharmacologic target for cardiovascular disease. Pharmacol Res 2016; 107: 381-389
• 11 Vu TM, Ishizu AN, Foo JC. , et al. Mfsd2b is essential for the sphingosine-1-phosphate export in erythrocytes and platelets. Nature 2017; 550 (7677): 524-528
• 12 Jonnalagadda D, Sunkara M, Morris AJ. , et al. Biochimica et Biophysica Acta Granule-mediated release of sphingosine-1-phosphate by activated platelets. Mol Cell Biol Lipids 2014; 18 (41) 1581-1589