Glutamate induces synthesis of thrombogenic peptides and extracellular vesicle release from human platelets

Platelets are highly sensitive blood cells, which play central role in hemostasis and thrombosis. Platelet dense granules carry considerable amount of neurotransmitter glutamate that is exocytosed upon cell activation. As platelets also express glutamate receptors on their surface, it is pertinent to ask whether exposure to glutamate would affect their signalling within a growing thrombus. In this study we demonstrate that, glutamate per se induced synthesis of thrombogenic peptides, plasminogen activator inhibitor-1 and hypoxia-inducible factor-2α, from pre-existing mRNAs in enucleate platelets, stimulated cytosolic calcium entry, upregulated RhoA-ROCK-myosin light chain/myosin light chain phosphatase axis, and elicited extensive shedding of extracellular vesicles from platelets. Glutamate, too, incited platelet spreading and adhesion on to immobilized matrix under arterial shear, raised mitochondrial transmembrane potential associated with generation of reactive oxygen species and induced activation of AMP-activated protein kinase in platelets. Taken together, glutamate switches human platelets to pro-activation phenotype mediated mostly through AMPA receptors and thus targeting glutamate receptors may be a promising anti-platelet strategy.

www.nature.com/scientificreports www.nature.com/scientificreports/ We have recently reported pro-thrombotic attributes of amyloidogenic neurotoxic peptides like amyloid beta and prion protein 13,14 . In the present study we demonstrate that, glutamate switches human platelets to pro-activation phenotype as reflected from synthesis of thrombogenic peptides from pre-existing mRNAs, activation of RhoA-Rho kinase-myosin light chain (MLC) signalling axis, extensive shedding of extracellular vesicles (EVs), augmented spreading on immobilized matrix, and formation of large platelet microthrombi under arterial shear. Strikingly, AMPA receptor antagonist mitigates the thrombogenic effect of glutamate on platelets. Thus, targeting glutamate receptors combined with inhibition of cyclooxygenase and purinergic ADP receptors can be a potential anti-platelet therapeutic strategy.
Glutamate instigates platelet spreading and aggregate formation under flow upon immobilized matrix. We next explored the effect of glutamate on adhesion signalling in human platelets as described for thrombin [18][19][20] . Platelets seeded on to immobilized fibrinogen underwent adhesion, followed by spreading with protrusion of filopodia/microspikes (Fig. 3A, upper panel). Although glutamate pre-treatment did not affect the number of cells adhered on to matrix, it strongly augmented the extent of platelet spreading with expression of prominent lamellipodia-like structures (Fig. 3A, middle panel), which was notably attenuated by glutamate receptor inhibitor CNQX (100 µM) (Fig. 3A, lower panel). , followed by addition of 500 μM glutamate along with 10 µM glycine (indicated by arrow). Tracing 2 represents resting platelets without glutamate treatment. Ca 2+ (1 mM) was included in all samples except experiments with EGTA. Corresponding values are graphically presented in (C). (B) Dose-dependent rise in intracellular calcium from glutamate-stimulated platelets. Results in (B,C) represent average of atleast 5 independent experiments (mean ± SEM). *P < 0.01 as compared to resting platelets (RP); # P < 0.01 as compared to glutamate-stimulated platelets.
Glutamate induces synthesis of hypoxia-inducible factor-2α (HIF-2α) and plasminogen activator inhibitor-1 (PAI-1) in human platelets. We have earlier demonstrated that, platelets synthesize HIF-2α through an oxygen-independent non-canonical path when challenged with physiological agonists like thrombin 21 . We asked now whether glutamate, too, can stimulate synthesis of HIF-2α from pre-existing mRNA in enucleate platelets. Glutamate (500 µM) increased expression of HIF-2α in platelets by 70.28% ± 0.11 (Fig. 5A,B). Plasminogen activator inhibitor-1 (PAI-1), which stabilizes fibrin-rich thrombus, is a target gene of HIF-2α in renal carcinoma cells 20 . It has been demonstrated that, platelets synthesize PAI-1 when stimulated with agonists including thrombin 22 . Remarkably, we found that glutamate was able to induce translation of PAI-1 by 49.03% ± 0.63 in platelets (Fig. 5A,C) in parallel with expression of HIF-2α. Both synthesis of HIF-2α and PAI-1 were inhibited by 10 mM puromycin, pharmacological inhibitor of protein translation, by 32.76% ± 0.16 and Glutamate affects mitochondrial function. Glutamate has been reported to modulate mitochondrial transmembrane potential (ΔΨ) and basal cellular respiration in murine hippocampal HT22 cells 23 . In the present study glutamate significantly enhanced ΔΨ in MitoTracker Red-treated platelets by 30.43% ± 3.89 (Fig. 6A,B). Glutamate also enhanced ROS yield in platelets by 55.66% ± 6.57 (Fig. 6C,D), which can be attributed to mitochondrial membrane hyperpolarization. CNQX normalized glutamate-induced rise in ΔΨ by 44.33% ± 12.1 (Fig. 6A,B). As expected, protonophore 100 µM carbonyl cyanide m-chlorophenyl hydrazine (CCCP) used as control dissipated mitochondrial transmembrane potential. To evaluate the effect of glutamate on mitochondrial respiratory function, we measured mitochondria oxygen comsumption rate using Clark amperometric electrode at high resolution (sampling at 2 s intervals). Glutamate, however, had no stimulatory effect on platelet oxygen consumption ( Fig. 6F), which was consistent with hyperpolarized state of mitochondrial membrane in a coupled system.
As glutamate induced RhoA activation and MLC phosphorylation in platelets ( Fig. 4) in absence of upsurge in mitochondrial respiration (Fig. 6F), this could lead to drop in energy charge and activation of AMP-activated protein kinase (AMPK), the energy sensor in the cell, in glutamate-treated platelets. Keeping with this, glutamate was found to significantly increase phosphorylation of AMPK (by 67.50% ± 1.03) (Fig. 7A, upper panel), suggestive of enhanced activity of the enzyme in platelets, which was inhibited by CNQX (by 28.66% ± 1.43). In consistence, glutamate also significantly enhanced phosphorylation (by 65.11% ± 1.34) of acetyl CoA carboxylase (ACC), an AMPK substrate, following 10 min incubation with platelets (Fig. 7A, lower panel). As a positive control 1 mM 5-Aminoimidazole-4-carboxamide ribonucleotide (AICAR), an AMP analog, was employed to stimulate AMPK. These results were supportive of state of relative ATP depletion in glutamate-treated platelets consistent with hyperpolarization of mitochondrial membrane.

Discussion
In the present study we have demonstrated that, neurotransmitter glutamate can modulate platelet signalling per se in the absence of physiological agonists. Glutamate elicited extensive shedding of EVs from platelets, induced platelet adhesion on to immobilized collagen under arterial shear and cell spreading, provoked Ca 2+ entry from external medium and stimulated synthesis of thrombogenic peptides, PAI-1 (that stabilizes fibrin-rich thrombus) and HIF-2α, all being hallmarks of pro-activation phenotype. Although PAI-1 is a target gene of HIF-2α in renal cell carcinoma 24 , platelets lack genomic DNA and have limited capacity to synthesize peptides. Glutamate has been shown to stimulate HIF translation in TNBC cell lines 25 . We have recently demonstrated that, platelets express HIF-2α when challenged with thrombin 21 . Strikingly, present observation adds glutamate to the existing list of agonists that induce protein synthesis in enucleate platelets. www.nature.com/scientificreports www.nature.com/scientificreports/ As small GTPase RhoA is a critical regulator cytoskeletal reorganization in activated platelets, we probed RhoA-mediated signalling in cells challenged with glutamate. Pre-treatment with glutamate-stimulated phosphorylation of MLC and MYPT1, which were reversed by antagonists of ROCK (Rho effector kinase) as well as AMPAR. Consistent with this, glutamate potentiated GTP-loading of RhoA in platelets that validated activation of RhoA-ROCK-MLC/MYPT1 axis downstream of glutamate-AMPAR interaction. As glutamate was not found to stimulate mitochondrial oxygen flux, we predicted relative ATP depletion in glutamate-treated platelets. In agreement with this, activity of AMPK was found to be upregulated in platelets by glutamate.
Thrombin is known to activate RhoA-MLC axis 26 , release EVs 27 , induce synthesis of PAI-1 28 /HIF-2α 21 and activate AMPK 29 in platelets. Glutamate has earlier been shown to potentiate agonist-induced platelet activity 9,12 . Here we demonstrate that, glutamate per se at concentrations up to 500 µM brought about synthesis of  Supplementary Figs S3 and S4, respectively. (E) Densitometric analysis of RhoA-GTP normalized against total RhoA expression. Bar diagrams represent atleast 5 independent experiments (mean ± SEM). *P < 0.03 as compared to resting platelets; # P < 0.03 as compared to glutamate-stimulated platelets for p-MYPT1. *P < 0.05 as compared to resting platelets; # P < 0.05 as compared to glutamate-stimulated platelets for p-MLC and RhoA-GTP. (2019) 9:8346 | https://doi.org/10.1038/s41598-019-44734-x www.nature.com/scientificreports www.nature.com/scientificreports/ thrombogenic peptides and extensive modulation in platelet signalling, mediated mostly through AMPA receptors, thus switching cells to pro-thrombotic phenotype. Glutamate, however, was neither able to induce platelet aggregation nor binding of PAC-1 (that specifically identifies the active conformation of platelet membrane integrins α IIb β 3 ) at above concentrations (Supplementary Figs S1 and S2, respectively). Thus, we infer that glutamate did not incite inside-out signalling mediated through platelet integrins that would have led to fibrinogen binding and aggregation.
Plasma glutamate is elevated under neurological pathologies like blood-brain barrier breakdown 30 , autism 31 and ischemic stroke 32 . Glutamate is also released from platelets in excess of 400 µM (in the concentration range 400-800 µM) into plasma during thrombus formation 3,9,12 . As glutamate could establish positive feedback loops to potentiate platelet stimulation in autocrine/paracrine manner similar to ADP and TXA2, targeting glutamate signalling in combination with established anti-platelet regimens may be a plausible therapeutic option.
Methods. Platelet Preparation. Platelets were isolated from fresh human blood by differential centrifugation, as already described 13 . Briefly, blood was collected from antecubital veins of healthy donors and centrifuged at 200 × g for 10 min. Platelet-rich plasma (PRP) was collected carefully to avoid the contamination of red and white blood cells and was incubated with 1 mM acetylsalicylic acid at 37 °C for 15 min. EDTA (5 mM) was added to PRP and was centrifuged at 600 × g for 10 min. Platelets pellet was washed in buffer A (20 mM HEPES, 138 mM NaCl, 2.9 mM KCl, 1 mM MgCl 2 , 0.36 mM NaH 2 PO 4 and 1 mM EGTA, supplemented with 5 mM Glucose, pH 6.2) and centrifuged at 600 × g for 10 min. Platelet pellet was finally resuspended in buffer B (pH 7.4), which was the same as buffer A but without EGTA. Platelet count was adjusted to 2-4 × 10 8 /ml with Beckman Coulter Multisizer 4. Precautions were taken for asepsis and to maintain the cells in resting condition.
Measurement of cytosolic free Ca 2+ in platelets. PRP was incubated with fura-2/AM at 37 °C for 45 min. Fura-2 loaded platelets were washed and suspended in buffer B. Fluorescence was taken in non-stirring condition in 400 µl aliquots of platelets at 37 °C using Hitachi fluorescence spectrophotometer (model F-2500). Fura-2 was  Supplementary Fig. S5. (B,C) Densitometric analysis of immunoblots for HIF-2α and PAI-1, respectively, normalized with respect to β-actin. Bars represent atleast 5 independent experiments (mean ± SEM). *P < 0.05 as compared to resting platelets; # P < 0.05 as compared to glutamate-stimulated platelets.
www.nature.com/scientificreports www.nature.com/scientificreports/ excited at 340 and 380 nm and the emission wavelength was kept at 510 nm. Intracellular free Ca 2+ concentration, [Ca 2+ ] i changes were monitored from the fluorescence ratio (340/380) using intracellular cation measurement program in FL solutions software, as described earlier 33 . F max was determined by lysing platelets with 250 µM digitonin in presence of saturating CaCl 2 . F min was determined by adding 2 mM EGTA. Intracellular free calcium was calibrated according to derivation of Grynkiewicz et al. 34 .
Study of extracellular vesicle release from platelets. Washed human platelets were treated with glutamate (500 µM) for 15 min at 37 °C. Cells were sedimented by centrifugation. Supernatant containing EVs was separated and were fixed with equal volume of 4% paraformaldehyde (PFA). Fixed supernatant was characterized by nanoparticle tracking analysis (NTA) where a beam from solid-state laser source (635 nm) was allowed to pass through the sample. 20 X microscope was used to observe the light scattered by rapidly moving particles in suspension in Brownian motion at room temperature (RT). Stokes Einstein equation was used to unveil the hydrodynamic diameter of particles, within range of 10 nm to 1 µm and concentration between 10 7 -10 9 /ml. The average distance moved by each EVs in x and y directions were captured with CCD camera (30 frames per sec) attached to the microscope. Both capture and analysis were performed using NTA 2.1 analytical software, which provides an estimate of the particle size versus concentration in sample. www.nature.com/scientificreports www.nature.com/scientificreports/ In order to study fibrinogen-binding to integrin α IIb β 3 expressed on EVs surface, suspension of EVs (100 µl) was incubated with Alexafluor488-conjugated fibrinogen (10 µg/ml) for 30 min in dark at room temperature. EVs were next fixed with equal volume of 4% PFA, washed and resuspended in sheath fluid. Samples were analyzed on flow cytometer as described earlier 13 .
Static adhesion and spreading of platelets on immobilized fibrinogen. Glass slides were coated with 100 µl fibrinogen (100 µg/ml) for 2 h, followed by addition of 100 µl bovine serum albumin (10 mg/ml) for 1 h. Washed human platelets (10 7 cells/ml) were pre-treated with glutamate (500 µM) in presence or absence of CNQX (100 µM) and overlaid on fibrinogen-coated slides for 15 min. Cells were fixed with 100 µl PFA (4%) for 30 min, followed by three washing with PBS. Cells were permeabilized with 0.01% Triton X-100 for 30 s, followed by washing thrice with PBS. Permeabilized platelets were incubated with phalloidin-FITC (1 µM) for 15 min. Adhered cells were examined under Zeiss LSM 700 laser scanning confocal microscope with 63X objective and 1 AU pinhole size. Images were acquired and analyzed using ZEN imaging software as described earlier 13 .
Dynamic adhesion of platelets on immobilized collagen under flow. Washed human platelets were incubated with FITC-labeled calcein-AM (2 µg/ml) for 30 min at 37 °C. Cells were sedimented at 600 × g for 10 min followed by resuspension in platelet-poor plasma. Glass cover slip coated with Type I collagen was congregated in a parallel plate flow chamber (GlycoTech) and was mounted on the stage of an inverted epifluorescence video microscope (Nikon model Eclipse Ti-E) equipped with a monochrome CCD cooled camera. Syringe pump (Pump 22 infusion/withdrawal with standard syringe holder, Harvard Apparatus) was used to perfuse control and glutamate (500 µM)-treated platelets, in presence or absence of 100 µM CNQX, through the chamber at constant flow to yield wall shear rate 1500 s −1 (15 dyn/cm 2 ). Images of fluorescent platelets from at least 5 different fields from each group were captured with DS-QiMC digital camera using NIS-Elements AR imaging software (Nikon).
RhoA-GTP pulldown assay. The assay was performed using a kit (Cytoskeleton) and following manufactur-er′s instructions as described previously 13 . Washed human platelets, pre-treated with either glutamate (500 µM) or thrombin (1 U/ml) were lysed. Supernatant was incubated with 15 µl Rhotekin-Rho binding domain (Rhotekin-RBD) beads at 4 °C for 1 h. Samples were subjected to SDS-PAGE, western blotted and probed with mouse anti-human RhoA antibody followed by goat anti-mouse anti-IgG (1:20,000), as mentioned earlier 13 .
ΔΨ M measurement. For elucidating ΔΨ M , washed human platelets were treated with glutamate in presence or absence of CNQX, followed by incubation with MitoTracker Red (500 nM) for 45 min. FL2 fluorescence was measured using flow cytometer 35 .  Supplementary Fig. S6. Bars represent atleast 5 independent experiments (mean ± SEM). *P < 0.03 as compared to resting platelets; # P < 0.03 as compared to glutamate-stimulated platelets for pAMPK and *P < 0.05 as compared to resting platelets; # P < 0.05 as compared to glutamate-stimulated platelets for pACC. (2019) 9:8346 | https://doi.org/10.1038/s41598-019-44734-x www.nature.com/scientificreports www.nature.com/scientificreports/ Measurement of intracellular ROS. Intracellular ROS was determined using a redox-sensitive cell-permeable dye, H 2 DCF-DA 36 . Washed human platelets were incubated at 37 °C for 5 min without stirring in the presence of glutamate (500 µM). H 2 DCF-DA (1 µM) was added to each sample and incubated for 30 min in the dark at RT. Platelets were fixed with 2% PFA. Cells were washed twice with PBS and resuspended in sheath fluid, followed by flow cytometry as described above. Hydrogen peroxide (1%) was added to platelet suspension as positive control.
Quantification of PAC-1 binding. Platelets on stimulation changes surface integrins α IIb β 3 to an open conformation that binds to fibrinogen with high affinity and leads to platelets aggregation 37 . PAC-1 antibody recognizes open conformation of α IIb β 3. Washed human platelets were stimulated with 500 µM glutamate at 37 °C for 10 min in non-stirring condition, followed by incubation with PAC-1 antibody (1.25 µg/ml) for 30 min in dark at room temperature. Platelets were fixed with equal volume of 4% paraformaldehyde for 20 min, washed twice with PBS and was resuspended in sheath fluid. Samples were analyzed with flow cytometer as described earlier 13 .
Platelet aggregation. Washed human platelets suspended in buffer B were stirred at 1200 rpm in optical lumi-aggregometer (Chrono-log model 700-2) at 37 °C for 1 min, followed by addition of thrombin (1 U/ml) or glutamate (200 and 500 µM) and transmittance was recorded. Aggregation was measured as percentage change in light transmission where 100% transmittance refers to transmittance through blank buffer solution 13 .
High-resolution respirometry for mitochondrial respiration. Mitochondrial respiration was measured using a high-resolution respirometer (Oxygraph-2k; Oroboros Instruments) at 37 °C under stirring conditions (750 rpm) as previously described 13 . Washed human platelets (in buffer B containing 5.5 mM glucose) with or without pre-treatment were transferred into oxygraph chamber. Respiration was first allowed to stabilize at the routine state, i.e., in the physiological coupling state controlled by cellular energy demands for oxidative phosphorylation. Then, platelets were treated with either glutamate (500 µM) or thrombin (1 U/ml), and changes in oxygen flux were recorded in real time at high resolution (sampling at 2 s intervals). Calibration at air saturation was performed each day before starting experiments by letting Millipore water/buffer B stir with air in the oxygraph chamber until equilibration and a stable signal was obtained. All experiments were performed at an oxygen concentration in the range of 100-205 µM O 2 . Data were recorded and analyzed using DatLab 5.1 software (Oroboros Instruments) 13 .
Statistical methods. Standard statistical methods were utilized to present the data. Two-tailed Student's t test was employed for evaluation of significance. Tests were considered significant at P < 0.05. All data are presented as mean ± SEM of ≥3 individual experiments.