Phosphorylated vasodilator-stimulated phosphoprotein (P-VASPSer239) in platelets is increased by nitrite and partially deoxygenated erythrocytes

Nitrite is recognized as a bioactive nitric oxide (NO) metabolite. We have shown that nitrite inhibits platelet activation and increases platelet cGMP levels in the presence of partially deoxygenated erythrocytes. In this study, we investigated the effect of nitrite on phosphorylation of vasodilator-stimulated phosphoprotein on residue serine 239 (P-VASPSer239), a marker of protein kinase G (PKG) activation, in human platelets. In platelet-rich plasma (PRP), nitrite itself had no effect on levels of P-VASPSer239 while DEANONOate increased P-VASPSer239. Deoxygenation of PRP + erythrocytes (20% hematocrit) raised baseline P-VASPSer239 in platelets. At 20% hematocrit, nitrite (10 μM) increased P-VASPSer239 in platelets about 31% at 10–20 minutes of incubation while the levels of P-VASPSer157, a marker of protein kinase A (PKA) activation, were not changed. Nitrite increased P-VASPSer239 in platelets in the presence of deoxygenated erythrocytes at 20–40% hematocrit, but the effects were slightly greater at 20% hematocrit. In conclusion, our data confirm that nitrite increases P-VASPSer239 in platelets in the presence of deoxygenated erythrocytes. They also further support the idea that partially deoxygenated erythrocytes may modulate platelet activity, at least in part, via the NO/sGC/PKG pathway from NO formed by reduction of circulating nitrite ions.


Introduction
Nitrite is now recognized as a bioactive NO metabolite that can be converted to NO either by enzymatic reaction or non-enzymatic disproportionation under hypoxia and low pH [1,2] The vasodilatory effect of nitrite has been extensively studied during the last decade; however the effect of nitrite on platelets received little attention until the last several years. Publications from our group starting in 2012 showed that while nitrite itself at physiological concentrations had little effect on platelets, interaction between partially deoxygenated erythrocytes with a1111111111 a1111111111 a1111111111 a1111111111 a1111111111 nitrite inhibited platelet aggregation by the release of NO [3]. The effects were seen in human and rodent platelets with several different agonists [4,5]. The effect of nitrite on platelets has been confirmed by aggregometry, flow cytometry and thromboelastography by several independent laboratories as well as our own group [6][7][8][9][10]. Since nitrite in the presence of erythrocytes increased platelet cGMP levels, we suggested that nitrite inhibited platelet activation through NO/sGC pathway [3,6]. However, the effect of nitrite on P-VASP Ser239 , one marker of sGC/PKG activation in platelets has been questioned. A recent publication failed to detect the effect of nitrite and deoxygenated erythrocytes on P-VASP Ser239 levels in platelets [11]. These issues are important both because of the potential relevance of vascular erythrocytes to the blood clotting process and its pharmacology but also because changes in this pathway have been interpreted as the most definitive evidence for the ability of erythrocytes to generate NO from nitrite by reductive chemical processes.
Phosphorylation of VASP, an actin binding protein, inhibits activation of glycoprotein IIb/ IIIa [12,13]. Although PKG phosphorylates several signaling proteins in platelets, VASP, which is highly expressed in platelets, is considered as an important marker of sGC/cGMP/ PKG activation [14,15]. VASP phosphorylation at serine 239 is specific for PKG activation while phosphorylation at serine 157 is responsible for PKA pathway [16]. In contrast to a publication from Gambaryan et. al. [11], our recent data showed increased P-VASP Ser239 in human platelets in vitro by nitrite in deoxygenated whole blood and ex vivo after nitrite inhalation [17]. We have extended our previous study to investigate the role of deoxygenation and hematocrit on P-VASP Ser239 induced by nitrite ions in human platelets. The effects of nitrite and NO donors on platelet P-VASP Ser239 in whole blood were measured under various conditions to better understand this phenomenon.

Subjects
Blood samples were obtained from healthy donors on the IRB approved National Institutes of Health protocol 99-CC-0168. Research blood donors were randomly recruited by the Department of Transfusion Medicine since December 2016 to December 2017 by phone, email or flyers at NIH from the local community. Donors were non-pregnant adults who met healthy blood donor criteria and tested negative for transfusion-transmissible diseases. The written informed consent was obtained in accordance with the declaration of Helsinki. Blood samples were collected in acid citrate dextrose (ACD) and used within 3 hours of withdrawal.

Sample preparation
Whole blood was centrifuged at 120×g for 10 minutes at room temperature to obtain PRP. Erythrocytes were obtained from centrifugation of the lower portion at 2500×g for 10 minutes. Whole blood, PRP and PRP plus erythrocytes added to various hematocrit levels were used in further experiments.
Under oxygen controlled experiments, whole blood or PRP + erythrocytes (10, 15, 20, 30 and 40% hematocrit) were deoxygenated by gentle blowing helium above cell suspension for 10 minutes with a final partial oxygen pressure (pO 2 ) = 24.67 ± 1.45 mmHg. The pO 2 was measured by ABL-80 FLEX CO-OX blood gas analyzer (Radiometer American Inc.). Levels of cell-free hemoglobin before and after deoxygenation were measured by a human hemoglobin ELISA kit (Abcam, Cambridge, MA). Nitrite 10 μM was injected into the preparation and incubated for 5,10,15,20,30,40 and 60 minutes at 37˚C. In another set of experiment, 300 μM L-NAME or 10 μM ODQ was incubated with PRP + erythrocytes (20% hematocrit) for 30 and 10 minutes, respectively. Platelets were collected after 10-minute incubation with 10 μM nitrite in deoxygenated samples.
In a separate set of experiments, DEANONOate (1 μM) and sodium nitroprusside (10 μM) were incubated in PRP, whole blood and deoxygenated whole blood for 5 minutes at 37˚C. Since our preliminary data in deoxygenated whole blood showed that nitrite maximally increased P-VASP Ser239 at 15-minute of incubation, nitrite was added to deoxygenated whole blood for 15 minutes. Platelets were collected in lysis buffer plus proteinase inhibitors.

Statistics
All data are presented as means ± SEM. Data processing and statistical analysis were done by GraphPad Prism 1 Version 4 (GraphPad software Inc., San Diego, CA).

Nitrite ions alone had little effect on P-VASP Ser239 levels in platelets
Nitrite (10 μM) incubated in PRP for 5 to 60 minutes had small effects on P-VASP Ser239 levels in platelets while DEANONOate (1 μM) significantly increased P-VASP Ser239 at 5 minutes after incubation. The increased P-VASP Ser239 induced by DEANONOate reached a maximum at 10-minute of incubation and returned to baseline at 40-60 minutes (Fig 1).

Nitrite + erythrocytes increased P-VASP Ser239 levels in platelets under deoxygenation
Because we previously found that effect of nitrite on platelet aggregation in the presence of erythrocytes was maximal at about 20% hematocrit, PRP and erythrocytes at 20% hematocrit were used in these experiments. After deoxygenation (pO 2~2 5 mmHg), baseline P-VASP Ser239 but not P-VASP Ser157 increased in platelets (Fig 2A and 2B). A 10-minute incubation of nitrite (10 μM) increased P-VASP Ser239 in platelets by 31.15 ± 5.78% but levels did not increase without deoxygenation. Addition of nitrite raised P-VASP Ser239 at 10 to 20 minutes and values returned to baseline within 1 hour (Fig 2C and 2E). In addition, nitrite at 1 μM increased P-VASP Ser239 by 18.87 ± 5.62% at 10-minute incubation (data not shown). In contrast, addition of nitrite at this concentration, with or without erythrocytes did not alter P-VASP Ser157 in platelets (Fig 2D and 2F). L-NAME, a nitric oxide synthase (NOS) inhibitor, and ODQ, a soluble cyclase inhibitor, both decreased baseline P-VASP Ser239 associated with the deoxygenation of PRP + erythrocytes. In addition, both L-NAME and ODQ, inhibited P-VASP Ser239 induced by exogenous nitrite (S1 Fig). These results suggest that NOS either from the erythrocytes or the platelets themselves plays some in nitrite conversion to NO by erythrocytes. However, further studies are required to evaluate this mechanism.

Erythrocytes diminished P-VASP Ser239 expression induced by NO donors
Although NO donors dramatically increased P-VASP Ser239 in PRP, this effect was decreased by addition of erythrocytes. In PRP, P-VASP Ser239 induced by DEANONOate (1 μM) and sodium nitroprusside (10 μM) increased by 3.19 ± 0.90 and 3.60 ± 0.76-fold, respectively. However, the effect of DEANONOate was abolished in whole blood while sodium nitroprusside maintained some effect on P-VASP Ser239 (Fig 4A and 4B). However, the P-VASP Ser239 induced by  sodium nitroprusside in whole blood was significantly decreased compared to its effect in PRP (Fig 4C).

Nitrite was as effective as sodium nitroprusside for induction of P-VASP Ser239 in platelets
In order to compare the effects of nitrite on P-VASP Ser239 with NO donors in platelets, levels of P-VASP Ser239 induced by nitrite (10 μM), DEANONOate (1 μM) and sodium nitroprusside (10 μM) were measured in deoxygenated whole blood. Without addition of nitrite, deoxygenated whole blood (pO 2~2 5 mmHg) increased baseline P-VASP Ser239 in platelets slightly ( Fig  5A). Nitrite and sodium nitroprusside significantly increased P-VASP Ser239 by 37.03 ± 5.11 and 32.99 ± 13.56%, respectively under these conditions while DEANONOate had little effect ( Fig 5B, S3 Fig). In addition, DEANONOate at 10 μM did not increase P-VASP Ser239 in deoxygenated whole blood (data not shown). Therefore, nitrite is as effective as sodium nitroprusside in increasing P-VASP Ser239 in platelets in deoxygenated whole blood.

Discussion
Our results demonstrated increased P-VASP Ser239 in platelets by nitrite in the presence of erythrocytes under deoxygenated condition while nitrite itself at 10 μM had no effect on PRP. Nitrite increased P-VASP Ser239 in platelets at 20% hematocrit and above. Under hypoxic condition, nitrite was as effective as sodium nitroprusside for inducing P-VASP Ser239 in platelets.
Our finding of increased P-VASP Ser239 in platelets induced by nitrite + erythrocytes corroborates the notion of platelet inhibition, at least in part, through the sGC/PKG pathway. In agreement with a previous study [18], nitrite alone did not have effect on P-VASP Ser239 in platelets. However, even without addition of exogenous nitrite, deoxygenation of PRP + erythrocytes raised baseline P-VASP Ser239 . Decreasing pO 2 with deoxygenation facilitates endogenous minutes. Data are mean ± SEM (n = 7). Ã P < 0.05 compared with PRP and tested by one-way ANOVA with Tukey's multiple comparison. # P < 0.05 compared with control and tested by paired Student's t-test.
https://doi.org/10.1371/journal.pone.0193747.g002 erythrocyte nitrite conversion to NO leading to increase P-VASP Ser239 in platelets. In order to study the kinetics of P-VASP Ser239 induced by nitrite, 10 μM nitrite (a possible pharmacological concentration of nitrite) was used in this study. In contrast to the rapid effect of DEANONOate, nitrite and deoxygenated erythrocytes increased P-VASP Ser239 after 10 to 20-minute of incubation; this increase returned to baseline at 60 minutes of incubation. Since nitrite conversion to NO requires complex processes including nitrite uptake, conversion to NO and export the NO out of the erythrocytes, the kinetics of PKG activation induced by nitrite and deoxygenated erythrocytes might be expected to be slower compared to the effect of spontaneous NO release from DEANONOate. Although NO has also been shown to inhibit Rap1b, a glycoprotein IIb/ IIIa activator protein, via a VASP-independent pathway [19], phosphorylation of VASP via the sGC/PKG pathway appears to play an important role for the platelet inhibitory effect by nitrite and deoxygenated erythrocytes.
Since P-VASP Ser157 shifts the molecular weight of VASP from 46 kDa to 50 kDa [16], we tested whether nitrite increased P-VASP Ser157 . We used the same set of samples to detect both Ser157 and Ser239 phosphorylation by Western blots and found that deoxygenation and addition of nitrite did not alter P-VASP Ser157 in platelets. This indicated that the effect of nitrite detected at 50-kDa-band is specific for the phosphorylation at serine 239.
Under deoxygenation, nitrite increased P-VASP Ser239 in platelets in the presence of erythrocytes at and above 20% hematocrit. We found that P-VASP Ser239 induced by nitrite reached maximum effect at about 20% hematocrit. Deoxygenation of PRP + erythrocytes induced insignificant increases in cell-free hemoglobin levels, therefore the slightly decreased P-VASP-Ser239 induced by nitrite in the presence of erythrocytes at 30 and 40% hematocrit was not due to cell free hemoglobin from the deoxygenation process. These results are consistent with our previous platelet inhibitory effect of nitrite detected by both aggregometery and thromboelastography [3,20]. We showed that nitrite + erythrocytes at 10 and 15% hematocrit had little effect on P-VASP Ser239 in platelets. This may explain why a recent publication failed to demonstrate effect of nitrite on P-VASP Ser239 in platelets in the presence of erythrocytes at 0.9% hematocrit (10 8 RBCs/mL) [11]. The fact that nitrite was added to the erythrocytes before adding the platelets may also account for these negative results.
The decreased effect of NO donors on platelets by erythrocytes has been reported previously [21]. Our studies found that DEANONOate caused 3-fold increase in P-VASP Ser239 in PRP however; this effect was abolished by the presence of erythrocytes. At physiological pH (37˚C), NO is rapidly released (half-life = 2 minutes) from DEANONOate and taken up into erythrocytes. In erythrocytes, NO reacts with either oxyhemoglobin to form nitrate + methemoglobin or deoxyhemoglobin to form iron nitrosyl hemoglobin (HbNO) [22,23]. This suggests that NO release spontaneously by DEANONOate and taken up by erythrocytes would not be able to escape erythrocytes. This may also explain why high hematocrit levels reduce the effects of nitrite. Comparative effects of nitrite and NO donors in deoxygenated whole blood. Quantification of P-VASP Ser239 /VASP at baseline (A) and after treatment of deoxygenated whole blood with nitrite, DEANONOate and sodium nitroprusside (B) are shown. Whole blood was deoxygenated by helium for 10 minutes. Nitrite (10 μM) was incubated in deoxygenated whole blood for 15 minutes while DEANONOate (1 μM) and sodium nitroprusside (10 μM) were incubated for 5 minutes, respectively. Data are mean ± SEM (n = 5). Ã P < 0.05 compared with PRP and tested by one-way ANOVA with Tukey's multiple comparison. # P < 0.05 compared with control and tested by one-way ANOVA with Tukey's multiple comparison.
In contrast, sodium nitroprusside could increase platelets P-VASP Ser239 in the presence of erythrocytes in venous and deoxygenated whole blood. A previous study showed that sodium nitroprusside inhibited platelet aggregation in both PRP and whole blood [24]. Although sodium nitroprusside is clinically used to decreased blood pressure in hypertensive emergency, the mechanism of NO release remains incompletely understood [25]. NO release from nitroprusside requires an electron transfer from sulfhydryl-containing compounds (glutathione and cysteine) and leads to formation of disulfides and S-nitrosothiols [26]. S-nitrosothiol is detected in tissue after addition of sodium nitroprusside and is believed to be a part of storage and transport of NO in vivo [27].
Nitrite reductase activity of hemoglobin reaches a maximum rate at 50% oxygen saturation (pO 2~2 8 mmHg) [28,29], which is close to values found in tissues such as muscle and skin [30]. This suggests that nitrite can be converted to NO under hypoxic conditions in tissues, at levels where the endogenous NOS pathway becomes less active [31] leading to vasodilation and platelet inhibition. In addition to deoxyhemoglobin, nitrite can be converted to NO by NOS under hypoxic condition [32,33]. Our study found an unexpected decrease of P-VASP Ser239 caused by L-NAME after addition of nitrite and deoxygenated erythrocytes. Therefore, total levels of P-VASP Ser239 appear, at least in part, dependent on interactions with NOS and raises the possibility of the presence of NOS enzymes in the platelets and/or the erythrocytes. These processes are in addition to the direct effects of erythrocytes in reducing nitrite to NO. Erythrocytes eNOS has been reported by several groups [34] as has platelet NOS [35] although the many reports on platelet NOS has been questioned by others [36]. It is also possible that the use of NOS inhibitors such as L-NAME are not specific and have effects on other aspects of the nitrite induced production of NO itself. Conversely, not all effects of NO on platelet function may be reflected in P-VASP Ser239 levels.
Thus, the total effects of the presence of erythrocytes on these processes appear to depend on the mechanism and kinetics of formation of NO and its release from the erythrocytes and transfer to platelets, processes we are studying further. Although the exact mechanism of how NO escapes erythrocytes is still not completely clear, NO release has been detected from interaction between nitrite and deoxygenated erythrocytes by several laboratories [8,37]. In addition, NO-dependent vasodilation and platelet inhibition effects of nitrite reported by independent laboratories also confirm the production of NO from interaction of nitrite with erythrocytes [6,9,29]. Since NO produced inside erythrocytes by interaction between nitrite and deoxyhemoglobin is rapidly scavenged, nitrite has been proposed to react with a membrane metabolon such as the anion exchanger-1 (AE-1), carbonic anhydrase and deoxyhemoglobin [38]. The compartmentalization of NO might facilitate NO release. One recent study demonstrated that nitrite reacted hemoglobin such as metHb-NO and Hb-NO + binds with the erythrocytic membrane at high affinity [39]. This nitrite reacted hemoglobin may compete with deoxy-and oxyhemoglobin for binding with AE-1 leading to facilitate NO release. In addition, it has been suggested that nitrite inhibits platelets through facilitating S-nitrosation of the erythrocytic membrane [9]. Such surface nitrosation of erythrocytes may facilitate NO release by transnitrosylation.
In conclusion, our results confirm that the interaction between nitrite with partially deoxygenated erythrocytes inhibits platelets by NO/sGC pathway. Further, these findings also confirm and extend our previous results suggesting that reduction of nitrite by partially deoxygenated erythrocytes in the circulation modulate platelet activation and blood clotting within the various parts of the vascular bed. They may also open the potential use of nitrite therapy to regulate thrombotic processes.