Moderate Exercise Promotes Human RBC-NOS Activity, NO Production and Deformability through Akt Kinase Pathway

Background Nitric oxide (NO) produced by nitric oxide synthase (NOS) in human red blood cells (RBCs) was shown to depend on shear stress and to exhibit important biological functions, such as inhibition of platelet activation. In the present study we hypothesized that exercise-induced shear stress stimulates RBC-NOS activation pathways, NO signaling, and deformability of human RBCs. Methods/Findings Fifteen male subjects conducted an exercise test with venous blood sampling before and after running on a treadmill for 1 hour. Immunohistochemical staining as well as western blot analysis were used to determine phosphorylation and thus activation of Akt kinase and RBC-NOS as well as accumulation of cyclic guanylyl monophosphate (cGMP) induced by the intervention. The data revealed that activation of NO upstream located enzyme Akt kinase was significantly increased after the test. Phosphorylation of RBC-NOSSer1177 was also significantly increased after exercise, indicating activation of RBC-NOS through Akt kinase. Total detectable RBC-NOS content and phosphorylation of RBC-NOSThr495 were not affected by the intervention. NO production by RBCs, determined by DAF fluorometry, and RBC deformability, measured via laser-assisted-optical-rotational red cell analyzer, were also significantly increased after the exercise test. The content of the NO downstream signaling molecule cGMP increased after the test. Pharmacological inhibition of phosphatidylinositol 3 (PI3)-kinase/Akt kinase pathway led to a decrease in RBC-NOS activation, NO production and RBC deformability. Conclusion/Significance This human in vivo study first-time provides strong evidence that exercise-induced shear stress stimuli activate RBC-NOS via the PI3-kinase/Akt kinase pathway. Actively RBC-NOS-produced NO in human RBCs is critical to maintain RBC deformability. Our data gain insights into human RBC-NOS regulation by exercise and, therefore, will stimulate new therapeutic exercise-based approaches for patients with microvascular disorders.

In vitro investigations have shown that pharmacological stimuli positively influence RBC-NOS activation [9]. Application of recombinant human erythropoietin (rhEPO) increased RBC-NOS activity through phosphorylation of its serine 1177 (RBC-NOS-Ser 1177 )-residue that resulted in increased NO production. It was further shown that the phosphatidylinositol 3 (PI3)-kinase/Akt kinase pathway is involved in this process [9]. Additionally, NO produced by RBC-NOS has been found to regulate platelet activation and the deformability of RBC membranes [8]. Recent in vitro examinations have shown that the application of continuous shear stress also activates RBC-NOS and NO production [10]. This leads to the indication of important influences of mechanical stimulations in the vascular bed on RBC-NOS activation and subsequent NO formation to maintain RBCs deformability and, thus, peripheral oxygen supply.
Importantly, it is known that physical exercise results in increased peripheral shear stress in the vascular bed [11]. In this context, it was recently demonstrated in untrained rodents that physical exercise positively influences RBC deformability [12]. It can be speculated from these results that the improvement in RBC deformability is caused by RBC-NOS activation, because RBC-NOS represents an important source of NO production [8,13], with NO released in RBCs both leading to improved deformability [14] and potentially positively influencing vascular regulation. But these very important direct correlations have not been investigated in detail, yet.
In contrast, our group demonstrated in a human in vivo investigation [15] that high-intensive exercise induces protein catabolism and thus negatively affects the RBC-NOS activity Similar effects were found in trained rodents where high intensity exercise impaired characteristics of RBCs' plasma membrane resulting in diminished deformability [12,16].
It can be stated that in vitro, but also possibly in vivo, shear stress induced e.g. by moderate exercise does not lead to protein catabolism and therefore improves RBCs rheological characteristics. However, the underlying in vivo mechanisms of these regulations are far from being resolved. Therefore, we hypothesized in the present study that acute moderate physical exercise stimulates and improves RBC-NOS activity, RBC NO production and deformability in humans. Additionally, we proved the questions whether NOS-regulating upstream located signaling pathways, specifically Akt kinase signaling [8], are involved in these important processes in vivo.

Ethical Approval
The protocols used in this study were approved by the ethics committee of the German Sports University Cologne. These protocols align with the Declaration of Helsinki and all participants gave written informed consent to participate in this study.

Selection of Subjects
Fifteen male subjects participated in this study. Basal anthropometric parameters of the subjects were as follows (mean 6 SD): age [years]: 2763.6, height [cm]: 18264.0, weight [kg]: 7968.9. All subjects abstained from alcohol consumption for 24 hours prior to and during the training intervention and were non-smokers.

Exercise Protocol and Blood Sampling of the in vivo Investigations
During the acute moderate running test (AMRT) the participants exercised at a velocity corresponding to 70% of their individual 4 mmol lactate threshold for 1 hour. This velocity was determined for each subject in a preliminary incremental running step test (IRT) which was performed 21 days prior AMRT. This three week interruption phase between the tests was chosen to avoid post-exercise plasma expansion, intensified haemolysis or disturbances in erythropoiesis [17] to interfere subsequent testing. The test persons were also advised to abandon high intensity exercise during the period of rest. The IRT started at a velocity of 2.0 m*sec -1 and the speed increased every 5 min by 0.5 m*sec -1 until subjective exhaustion. Every interval was followed by a 30 sec break to obtain 20 ml blood from the earlobe for lactate analysis. Capillary blood lactate concentrations were measured with EBIO plus (EKF Diagnostic Sales GmbH, Magdeburg, Germany). A calculation matrix developed at the German Sport University was used to determine the individual 4 mmol lactate threshold, the 70% of this threshold and the corresponding velocity applied during AMRT. The average running speed during AMRT was 2.3960.41 m*sec -1 .
Blood gas analysis of capillary blood was carried out (AVL Omni Series Blood Gas Analyzer, Roche Diagnostics, Mannheim, Germany) in order to obtain pH, partial pressure of arterial carbon dioxide (pCO 2 ), partial pressure of oxygen that is dissolved in arterial blood (pO 2 ), and arterial haemoglobin oxygen saturation (sO 2 ) Pre AMRT and Post AMRT (Table 1).
Additional blood samples were taken from the subjects' vena mediana cubiti Pre and Post AMRT to determine Akt kinase and RBC-NOS phosphorylation, NO production in RBCs and RBC deformability. To avoid any additional shear stress which may alter NO production in the RBCs, blood was collected without a tourniquet Pre and Post using the Safety-Multifly-Set (21G, 0,8619 mm; Sarstedt, Nürnbrecht, Germany) connected to a 5 mL syringe. 4 mL blood was slowly drawn from the vein and immediately anticoagulated using either Heparin vacutainers (BD Vacutainer, Franklin Lakes, USA).

In vitro Analysis of RBC-NOS Activation through PI3 Kinase Pathway
An additional in vitro approach was carried out to investigate the influence of wortmannin on RBC-NOS serine 1177 phosphorylation (RBC-NOSSer 1177 ), RBC NO production and RBC deformability. Wortmannin has been shown to inhibit PI3 kinase, thereby reducing Akt signaling. This reduction decreases phosphorylation of the eNOSSer 1177 [18]. Wortmannin was used to determine a direct link between PI3 kinase/Akt kinase/RBC-NOS and NO signaling in human RBCs. NO production was determined Pre and Post AMRT via DAF fluorometry (see below). Post AMRT blood samples were additionally incubated with wortmannin (c = 10 mM) for 30 min at 37uC to investigate whether exerciseinduced NO production decreases upon PI3 kinase inhibition. The concentration and incubation time for wortmannin experiments were preliminary tested.
Post AMRT additional blood was taken to also investigate phosphorylation of RBC-NOSSer 1177 and RBC deformability depending on PI3 kinase inhibition. Parameters were determined before (0 min) and after (30 min) wortmannin incubation. Simultaneous sample incubation with phosphate buffered saline (Dulbeccos Phosphate-Buffered Saline (PBS) (1x), without calcium and magnesium, PAA, Paschin, Austria) adjusted to physiological pH 7.4 served as control.

Immunohistochemical Procedure
Immunostaining of RBCs as described hereinafter has been applied in various studies [8,10,15,19]. RBCs were immunostained and analyzed by semi-quantitative approach as described previously [15]. According to the protocol of Suhr et al. [15] venous blood from either the exercise intervention or from wortmannin incubation was fixed with 4% paraformaldehyde (v/ v; 1/1) immediately after blood sampling. After repeated washing procedures RBCs were dispersed on a slide and heat fixed. The RBCs were washed in 0.1 M tris-buffered saline (TBS), permeabilized for 45 min in 0.1% trypsin, placed in a solution of 2% hydrogen peroxide and 80% methanol TBS for 20 min, and  [20] for intensity analysis of immunostaining in RBCs, the gray values (television densitometry) of 100 RBCs (test area) from at least four randomly selected pictures and 50 RBCs (control area) from at least two randomly selected pictures of each slide were measured. The intensity of immunostaining was reported as the mean of measuring RBC gray value minus background gray value. The background gray value was detected at a cell-free area of the slide. Finally, the gray values of the test area (minus background) and the control area (minus background) were subtracted. For staining intensity detection, a Leica microscope coupled to a CCDcamera (DXC-1850P, Sony, Germany) was used and the analysis was conducted using the software ''Image J'' (National Institutes of Health, Bethesda, Maryland, USA). Magnification for all images was 400-fold.

Western Blot Analysis
Additional western blot analyses were performed to confirm the results obtained in the immunohistochemical procedure. For this purpose heparinized venous blood was taken Pre and Post AMRT. RBCs were separated at 8006g for 10 min at 4uC. RBC protein fractions were extracted using the ProteoExtract Subcellular Proteome Extraction Kit (CalBiochem, Darmstadt, Germany). The mild procedure yields proteins in their native state. 5610 6 RBCs were used for the extraction. RBCs were repeatedly washed (v/v; 1:20; RBCs : washing buffer) at 8006g for 10 min at 4uC using supplied washing buffer. RBCs were then incubated with supplied extraction buffer for 10 min at 4uC by gentle shaking to release cytosolic proteins. After centrifugation at 1,0006g for 10 min at 4uC supernatant with cytosolic proteins were stored on ice until further processing. Subsequently, a second extraction buffer was added to the remaining pellet, incubated for 30 min at 4uC and the sample was centrifuged at 20,0006g for 10 min and 4uC. The supernatant with the solubilized membrane proteins was stored on ice. A third extraction buffer was added to the remaining pellet, incubated for 10 min at 4uC on a shaker and the sample was then centrifuged at 20,0006g for 10 min and 4uC. This supernatant is aimed to contain nucleic proteins and was discarded because mature RBCs are anucleate cells [21]. Finally, a fourth extraction buffer was added to the pellet to solubilize components of the cytoskeleton.
Protein concentration of all four fractions was determined using the DC-Protein Assay Kit (BioRad, Munich, Germany). A preliminary test revealed that the relevant proteins were present in the cytosolic fraction that therefore was used in subsequent western blot analysis. A total of 150 mg protein (fraction 1) was loaded fivefold into the lanes of a 4-12% Bis-Tris gel (BioRad, Munich, Germany) next to the Precision Plus Protein TM Standard (BioRad, Munich, Germany). Proteins were separated according to their charge and mass in a 1 x MOPS running buffer (BioRad, Munich, Germany) and transferred to a polyvinylidene fluoride membrane (0.45 mm pore size) and probed with antibodies against total eNOS (dilution 1:800, Biomol, Hamburg, Germany), phosphorylated eNOSSer 1177 , phosphorylated eNOSThr 495 (dilution 1:800, Upstate, Lake Placid, USA), cGMP (dilution 1:1:800, Chemicon/Millipore GmbH, Schwalbach, Germany) or phosphorylated AktSer 473 (dilution 1:1:800, Cell Signaling, Danvers, USA). The membrane was then incubated with a secondary goatanti-rabbit peroxidise conjugated antibody (dilution 1:2000, Thermo Scientific, Bonn, Germany). After development of immunoreactive bands by means of an enhanced chemiluminescence kit (Thermo Scientific, Darmstadt, Germany), relative band intensities were quantified using software ''Image J''. The difference in intensity between Pre and Post exercise condition was calculated.

DAF Fluorometry
As described by Kojima et al. [22] and Itoh et al. [23] the cellpermeable 4,5-diaminofluorescein diacetate (DAF-2DA) (Alexis Biochemicals, Lörrach, Germany) indicates NO-production inside cells, where DAF-2A is N-nitrosated and thereby forms the fluorescing triazolofluorescein (DAF-2T). DAF-2A does not react directly with the NO radical, but rather with nitrous anhydride (N 2 O 3 ), which is formed by autoxidation of NO in air [22,24,25]. DAF-2DA was used to detect NO production in human RBCs before and after the AMRT intervention and after wortmannin application. Pre and Post AMRT, venous blood was equilibrated with DAF-2DA solution (c = 10 mM) for 3 min. After incubation of the blood with DAF-2DA, 50 ml of blood were pipetted and dispersed onto a little glass bottom petri dish. To investigate the question whether the fluorescent signal was NO-and NOSspecific, wortmannin (c = 10 mM) was added to another 100 ml blood sample (Post AMRT) and was incubated for 30 min. The whole blood was imaged on a confocal laser scanning microscope (LSM 510 Meta, Zeiss, Germany) with an Argon laser used to provide fluorescence excitation at 488 nm.

RBC Deformability
Immediately after sampling, venous heparinized blood was mixed (v/v; blood/medium; 1:250) with an isotonic viscous medium (0.14 mM Polyvinylpyrrolidone (PVP), osmolality 300 mOsmol*L 21 , viscosity 32 mPa*s at 37uC; Mechatronics, Netherlands) for deformability measurements. All measurements were performed at physiological pH 7.4 and 37uC. Deformability was determined at various physiological and supra-physiological fluid shear stresses by laser diffraction analysis using the laser-assisted optical rotational red cell analyzer (LORRCA, RR Mechatronics, Netherlands). The LORRCA has been described in detail elsewhere [26], Briefly, 1 mL of the blood/PVP mixture was sheared in a Couette system. A laser beam was directed through the sheared sample and the diffraction pattern produced by the deformed red cells was analyzed. On the basis of the geometry of the elliptical diffraction pattern, an elongation index (EI) was calculated by the software as EI = (L-W)*(L+W) -1 , where L and W represent length and width of the diffraction pattern, respectively. EI values were calculated for nine shear rates, in particular 0.3, 0.57, 1.08, 2.04, 3.87, 7.34, 13.92, 26.38 and 50 Pa. EI values were plotted as a function and according to Baskurt et al. [27], the Lineweaver-Burke equation was used to calculate the maximum deformability (EImax) and the shear stress required for one-half of this maximal deformation (SS K). Increased SS K values indicate decreased RBC deformability.
Additionally, Post AMRT heparinized blood samples were also incubated with either PBS or wortmannin (c = 10 mM) for 30 min at 37uC to inhibit PI3 kinase. RBC deformability was measured after 0 and 30 min using the LORRCA as described previously.

Statistical Analysis
Statistical analyses of the data were performed by using statistics software packages (STATISTICA for Windows, 7.0, Statsoft, Tulsa, USA and Origin 8.5 Pro, Northampton, USA). Data were analyzed by paired Student's t-test to determine the significance of any differences between the Pre and Post situation. For comparison of wortmannin data one-way analysis of variances (ANOVA) was applied. Descriptive statistics of the data is presented as mean 6 standard error of means (S.E.M.) unless described otherwise. Statistical differences were considered to be significant for values of P,0.05.

Immunohistochemical Staining of Akt Kinase, RBC-NOS and cGMP
Immunohistochemical straining was performed Pre and Post AMRT to investigate whether moderate exercise influences RBC-NOS activation pathway.
AktSer 473 . To determine the impact of the intervention on Akt kinase activation, phosphorylation of Akt kinase on its Ser 473 residue was examined before and after the test. The statistical analysis of phosphoAktSer 473 revealed that the value significantly increased from 1.1660.29 au Pre exercise to 1.6160.39 au Post exercise (P,0.05, Figure 1 A, B).
RBC-NOS and its activation sites. The total detectable RBC-NOS, RBC-NOSSer 1177 and RBC-NOSThr 495 signal intensity were examined and the statistical evaluation revealed no difference in total detectable RBC-NOS between the Presituation and the Post-situation (Pre cGMP. To further verify possible NO formation by the activated RBC-NOSSer 1177 we investigated downstream target proteins of NO. It is well known that cyclic guanosine monophosphate (cGMP) is a prominent target of NO in endothelial cells [28,29] and RBCs [30,31]. Therefore, we conducted evaluations of cGMP alterations after the applied exercise stimulation. The statistical analysis of the cGMP signal revealed significantly increased signal intensities from 1.0260.21 au Pre exercise to 1.8060.13 au Post exercise (P,0.05, Figures 3 A, B).
In vitro experiments. The ex-vivo results revealed that moderate exercise activates both Akt kinase and RBC-NOS. We thus performed in vitro experiments to verify the observed in vivo results. The conducted in vitro approach investigated RBC-NOS activation via the PI3 kinase/Akt kinase pathway. For this purpose phosphorylation of RBC-NOS at Ser 1177 was investigated after PI3 kinase inhibition using wortmannin [25]. The statistical analysis revealed that the analyzed signal intensity significantly decreased from 10.8061.20 au Pre incubation (0 min) to 3.3060.56 au Post incubation (30 min) (P,0.01, Figures 4 A, B).

NO Production in RBCs
Due to an increased level of RBC-NOS phosphorylation at the Ser 1177 residue after the AMRT intervention, the NO production in RBCs was investigated as a direct readout for RBC-NOSSer 1177 activation. DAF-2DA experiments were conducted before (Pre) and directly after the AMRT intervention (Post). Figure 6A shows that almost no nitric oxide production was detected in human RBCs by DAF-2DA imaging Pre AMRT. In contrast, the NO production in human RBCs increased after the physical activation, as measured post-exercise ( Figure 6B) and depicted by the green fluorescent signal.
The addition of the indirect RBC-NOS inhibitor wortmannin Post AMRT caused a down-regulation in NO production in human RBCs ( Figure 6C).

RBC Deformability Determined after AMRT
RBC deformability is crucial for the passage of RBCs through the microvasculature bed [32,33] to ensure the oxygen supply to working tissues. The results obtained in this study showed that the RBC deformability was significantly influenced by the intervention. EI max values increased from 0.5760.019 (Pre) to 0.660.013 (Post, P,0.05, Figure 7A). SS K significantly decreased from 2.1260.158 Pa (Pre) to 1.7560.14 Pa (Post, P,0.05, Figure 7B).

Discussion
NO generation and -mediated signaling are highly important for the regulation of the peripheral vascular bed [28,29] and NO produced by eNOS in endothelial cells plays a prominent role in these regulations (for review see [34,35]). However, very recently the NOS protein localized in RBCs came into focus as an additional important NO generating source [8].
Although it has been shown in vitro that defined shear stresses regulate and manipulate the activation of RBC-NOS with subsequent NO production [10] and that intracellular NO generation is involved in the improvement of RBCs rheological characteristics [8,14], it remains to be elucidated what molecular effects are promoted in vivo by moderate physical exercise on RBC regulation, specifically on RBC-NOS activation, NO generation, and subsequently on RBCs rheological characteristics.
In the present study, we therefore studied the in vivo effect of reinforced shear stress, caused by moderate exercise, on RBC-NOS protein, its activations sites, and -mediated NO signaling to outline that moderate physical exercise exerts a positive influence on the regulation of this biologically important enzyme.
The phosphorylation of RBC-NOSSer 1177 , indicating increased enzyme activity [8,19], was significantly increased after AMRT, whereas total RBC-NOS and the phosphorylation of RBC-NOS at its Thr 495 residue, indicating inhibition of NOS enzyme activity [36], were not affected by the applied exercise intervention. In line with RBC-NOS activation, the NO production was clearly increased after ARMT. These data indicate a beneficial effect of moderate exercise on RBC-NOS activity and subsequent NO production.
Next, signaling pathways that contribute to the activation of the RBC-NOS under exercise conditions were aimed to be identified. In endothelial cells eNOSSer 1177 phosphorylation and thus eNOS activation depends on the phosphorylation of Akt kinase at its Ser 473 through PI3 kinase [37,38,39]. Here, we report that in vivo phosphorylation of RBCs Akt kinase at Ser 473 significantly increased after AMRT, indicating a direct molecular interaction between phosphorylation of Akt-and RBC-NOS activation. These data complement the study of Mihov and colleagues who showed, using an in vitro assay, that pharmacological stimulations increase Akt kinase activity through PI3 kinase activation in RBCs [9].
NO, as an unstable radical, reacts rapidly within the cell [1]. Different possible reaction routes of NO might include the oxidation to nitrite and nitrate or the binding to hemoglobin or soluble guanylylcylase (sGC) with subsequent increase of cGMP content. It is also discussed that NO possesses the bioactive potential to regulate functions of the RBC itself, e.g. deformability [13,31]. In the present study we revealed that both, cGMP content and RBC deformability were significantly increased after AMRT by the described molecular routes. In the context of the present study it appears significant that RBC deformability increases upon exercise stimulation as RBCs exhibit an average diameter of about 7.5 mm [40], which is thus smaller than the diameter of the smallest capillaries having a diameter of 4.5 mm [40]. The RBCs transit through the capillaries to supply peripheral skeletal muscle and also cardiac areas with oxygen. Increased deformability thus allows the cells to respond to the changing demands by easing the passage through the vessel thereby increasing the oxygen supply during physical exercise. The interactions between membrane,  peripheral and cytoskeleton proteins are responsible for the maintenance of RBC deformability and some of these interactions are modulated by protein kinase activity. It has been shown that deformability is significantly decreased by inhibition of PKC [41], a kinase that might also interact with NOS [42]. Also, it has been shown that RBCs contain large amounts of adenosine triphosphate (ATP) [43], a substance activating endothelial P 2y purinergic receptors, thereby resulting in synthesis of endothelial NO [44,45] that leads to relaxation of vascular smooth muscles [46]. RBCs do not release ATP spontaneously, but upon mechanical deformation with the degree of deformation positively correlating with the ATP released [43]. This finding was supported by the a study of Wan and colleagues who showed that retraction of the spectrin-actin cytoskeleton network triggers the mechanosensitive ATP release [47]. Therefore, deformability of RBCs represents an important mechanism for the passage of RBCs through the microcirculation as well as for the control of vascular calibre. Patients with cystic fibrosis (CF) [48], primary pulmonary hypertension (PPH) [49] or diabetes [50] show less deformable RBCs and do not release ATP in response to mechanical deformation. These patients show also impaired endothelial NO synthesis [51,52]. From the data we represent here it can be assumed that also NO synthesis in RBCs might be impaired in these patients, which might cause reduced RBC deformability, reduced ATP release and consequently impaired NO synthesis in the endothelium.  The function of cGMP in RBCs is only speculative, but it is conceivable that RBC -cGMP also influences vascular tonus possibly by activating potassium channels that control membrane potentials in response to altered intracellular calcium [53,54]. However, further studies are needed to investigate this hypothesis.
Our data on human RBC-NOS modulation are the first to add important and molecular knowledge to the question whether exercise-induced increased shear stress promotes the activation of Akt kinase and RBC-NOS in RBCs thereby initiating beneficial alterations of RBC rheological properties.
To verify and support our obtained human in vivo results on RBC-NOS-related signaling, we conducted additional in vitro experiments by applying the PI3 kinase/Akt kinase inhibitor wortmannin [18] on human RBCs. We observed that the inhibition of PI3/Akt kinase pathway led to a decrease in RBC-NOSSer 1177 phosphorylation. In line with these observations, NO production and RBC deformability were significantly decreased indicating a causal relation between RBC-NOS activation and RBC deformability that depends on PI3/Akt kinase signaling.
The novel key findings of this study are that moderate physical exercise is a sufficient stimulator for the activation of RBC-NOS directly resulting in improved RBC rheological properties. The molecular mechanisms are highly similar to those observed in endothelial cells. Exercise-induced shear stress results in activation of the PI3 kinase/Akt kinase pathway that in turn activates RBC-NOS by phosphorylation at Ser 1177 . Activated RBC-NOS generates NO in RBCs that is mandatory to beneficially promote RBC rheological functions, e.g. deformability.
Therefore, the presented data gain important insights into the molecular pathways underlying RBC deformability. This knowledge is of great interest, because nowadays physical exercise programs are fundamental for the treatment of patients suffering from cardiac and cardiovascular diseases who may benefit from improved rheological properties. Further elucidations are required to investigate the effect of chronic exercise interventions on the discussed molecular RBC-NOS activation pathways and resulting biological functionality of RBC-NO-produced NO.