Shear-induced tyrosine phosphorylation in endothelial cells requires Rac1-dependent production of ROS

Li-Hong Yeh, Young J. Park, Riple J. Hansalia, Imraan S. Ahmed, Shailesh S. Deshpande, Pascal J. Goldschmidt-Clermont, Kaikobad Irani, B. Rita Alevriadou

Abstract

The shear-induced intracellular signal transduction pathway in vascular endothelial cells involves tyrosine phosphorylation and activation of mitogen-activated protein (MAP) kinase, which may be responsible for the sustained release of nitric oxide. MAP kinase is known to be activated by reactive oxygen species (ROS), such as H2O2, in several cell types. ROS production in ligand-stimulated nonphagocytic cells appears to require the participation of a Ras-related small GTP-binding protein, Rac1. We hypothesized that Rac1 might serve as a mediator for the effect of shear stress on MAP kinase activation. Exposure of bovine aortic endothelial cells to laminar shear stress of 20 dyn/cm2 for 5–30 min stimulated total cellular and cytosolic tyrosine phosphorylation as well as tyrosine phosphorylation of MAP kinase. Treating endothelial cells with the antioxidantsN-acetylcysteine and pyrrolidine dithiocarbamate inhibited in a dose-dependent manner the shear-stimulated increase in total cytosolic and, specifically, MAP kinase tyrosine phosphorylation. Hence, the onset of shear stress caused an enhanced generation of intracellular ROS, as evidenced by an oxidized protein detection kit, which were required for the shear-induced total cellular and MAP kinase tyrosine phosphorylation. Total cellular and MAP kinase tyrosine phosphorylation was completely blocked in sheared bovine aortic endothelial cells expressing a dominant negative Rac1 gene product (N17rac1). We concluded that the GTPase Rac1 mediates the shear-induced tyrosine phosphorylation of MAP kinase via regulation of the flow-dependent redox changes in endothelial cells in physiological and pathological circumstances.

  • endothelium
  • signal transduction
  • shear stress
  • oxidative stress
  • mitogen-activated protein kinase
  • reactive oxygen species

vascular endothelial cells (ECs) are constantly exposed to flow-induced shear stress. ECs respond to fluid shear stress by rapid release of bioactive compounds, such as the vasodilators prostacyclin and nitric oxide (NO) (21, 49), and by changes in the synthesis of proteins, such as the vasoconstrictor endothelin-1 (52). Specifically, for NO, exposure of cultured human umbilical vein ECs (HUVECs) to laminar shear stress increased NO release in a biphasic manner. The initial burst was dependent on Ca2+-calmodulin, whereas the sustained NO release was Ca2+independent (39, 40). Hence, activation of EC signal transduction by shear stress is believed to involve two pathways: a Ca2+-dependent and a Ca2+-independent pathway (6). The latter involves the Ca2+-independent tyrosine phosphorylation and activation of the 42- and 44-kDa mitogen-activated protein (MAP) kinases through small GTP-binding proteins, such as p21ras (Ras), and Ca2+-independent protein kinase C isozymes (46, 60). Tyrosine kinase inhibitors abolished the Ca2+-independent phase of shear-induced NO release, suggesting that shear stress activates the endothelial constitutive NO synthase (ecNOS) via a mechanotransduction that involves tyrosine kinases (3). It was found that the focal adhesion-associated tyrosine kinases p60src and p125FAK are upstream to the Ras-MAP kinase pathway (33, 45). Involvement of the tyrosine kinases in focal adhesion sites is expected, since focal adhesions reorganize during EC exposure to flow (15) and activated integrins elicit signals by inducing tyrosine phosphorylation of intracellular proteins (22). MAP kinase phosphorylates transcription factors, such as c-fos and AP-1, known to be activated by flow (29, 41), phospholipases, and other kinases (34). MAP kinase is a serine/threonine kinase, and ecNOS is phosphorylated on serine and threonine residues (13). Thus MAP kinase seems to be a suitable candidate for regulation of the sustained phase of shear-induced NO release (6).

It was shown that oxygen-derived free radicals, such as superoxide ( O2 ), and other reactive oxygen species (ROS), such as H2O2, induced MAP kinase tyrosine phosphorylation and activation in several cell types, e.g., O2 activated MAP kinase in vascular smooth muscle cells (VSMCs) (4). H2O2activated MAP kinase in NIH/3T3 cells (54) and also induced cellular tyrosine phosphorylation in bovine pulmonary artery ECs (9, 61). The role of ROS as intracellular second messengers was demonstrated by their requirement in VSMC signaling in response to platelet-derived growth factor (PDGF): increasing the intracellular concentration of the peroxide-scavenging enzyme catalase or treating cells with the antioxidant N-acetylcysteine (NAC) blocked the PDGF-induced cellular tyrosine phosphorylation and tyrosine phosphorylation of MAP kinase (57).

Laurindo et al. (43) discovered that increases in blood flow triggered free radical release in vivo and in isolated perfused rabbit aortas. ROS, in particular H2O2, were produced by cultured porcine aortic ECs subjected to a form of mechanical deformation, i.e., cyclic strain (27). Cyclic strain-induced ROS in HUVECs were involved in monocyte chemotactic protein (MCP)-1 gene expression (62) and plasminogen activator inhibitor (PAI)-1 release (10). More recently, intracellular O2 production in HUVECs was found to be elevated within minutes from the onset of laminar shear stress and was maintained at an elevated level as flow continued for 6 h (11, 16). Shear-induced ROS mediated the gene expression of intercellular adhesion molecule (ICAM)-1 (11) and c-fos (28). In this study we hypothesized that ROS are involved in several aspects of the signal transduction of shear stress in ECs:1) in shear-induced cellular tyrosine phosphorylation and, specifically,2) in tyrosine phosphorylation of MAP kinase, since MAP kinase is known to be activated by shear stress (60) or oxidative stress (4, 54, 57).

Potential sources of free radicals in cultured cells are the enzymes of the mitochondrial electron transport chain: xanthine oxidase (XO), cytochrome P-450, cyclooxygenase, lipoxygenase, and the superoxide-generating NADPH oxidase. On stimulation of phagocytic cells, such as neutrophils and monocytes, Rac1, a small GTP-binding protein of the Ras superfamily, enhances the activity of the enzyme NADPH oxidase, resulting in production of O2 (1). Sundaresan et al. (58) first provided evidence that the pathway by which ligand stimulation of ROS occurs in nonphagocytic cells involves the small GTP-binding proteins Ras and Rac1. Expression of activated Ras or activated Rac1 isoforms resulted in increased generation of O2 , which was subsequently dismutated to H2O2(32). Expression of a dominant negative Rac1 gene product (N17rac1) was shown to inhibit the intracellular burst of ROS in HUVECs after hypoxia-reoxygenation, suggesting that Rac1-dependent pathways may regulate the intracellular ROS levels during ischemia-reperfusion (38).

By employing chemical antioxidants, we demonstrated that the shear-induced increase in cellular tyrosine phosphorylation and, specifically, in tyrosine phosphorylation of MAP kinase is, at least partly, mediated by ROS. The effect of shear-induced ROS generation on cellular and MAP kinase tyrosine phosphorylation was completely blocked in cells infected with an adenovirus that encodes a dominant negative Rac1 (N17rac1), but not with a control virus, suggesting that the Ras-related small GTP-binding proteins may function as regulators of the intracellular redox status in sheared ECs.

MATERIALS AND METHODS

Materials.

DMEM, fetal bovine serum (FBS),l-glutamine, penicillin-streptomycin, collagenase, sodium pyruvate, nonessential amino acids, amphotericin B (Fungizone), trypsin-EDTA, and ATP-free medium 199 (M199) were purchased from GIBCO BRL (Gaithersburg, MD). The protease inhibitors 4-(2-aminoethyl)benzenesulfonyl fluoride, chymostatin, leupeptin, aprotinin, and pepstatin, the phosphatase inhibitor sodium orthovanadate (Na3VO4), and H2O2, NAC, pyrrolidine dithiocarbamate (PDTC), DMSO, and Triton X-100 were purchased from Sigma Chemical (St. Louis, MO). The OxyBlot oxidized protein detection kit was obtained from Oncor (Gaithersburg, MD).

Cell culture.

Fresh bovine aortas were obtained from a local slaughterhouse and filled with an ice-cold solution of 99% PBS-1% penicillin-streptomycin. The fat surrounding the aorta was cut off, and the intercostal and other branches were sealed. The aorta was filled with the same solution, to which 1 mg/ml collagenase was added, and incubated with the collagenase solution for 1 h at room temperature. Then it was filled with complete culture medium (DMEM supplemented with 10% FBS, 1% penicillin-streptomycin, 2% sodium pyruvate, 1% nonessential amino acids, 2 mMl-glutamine, and 25 mM HEPES) and massaged gently to detach ECs. The cell suspension was distributed into culture flasks and incubated at 37°C and 5% CO2. When confluency was reached, culture purity was verified by uptake of fluorescently labeled DiI-Ac-LDL (Biomedical Technologies, Stoughton, MA). ECs were passaged 1:4 twice and then frozen in 60% complete culture medium-10% DMSO-30% FBS in a liquid nitrogen tank. Some experiments were performed using bovine aortic ECs (BAECs) purchased from Clonetics (San Diego, CA). Primary cells were plated in tissue culture flasks at a cell density of 5,000 cells/cm2and grown in EGM-2 medium (Clonetics). Media were replaced every other day, and cells were subcultured on confluency. Cells inpassages 3–10 were used in the present experiments.

Shear stress experiments.

Cells (∼104/cm2) were seeded on glass slides (75 × 38 mm; Corning, Corning, NY) that had been dipped into 100% ethanol, air-dried, and coated with 0.2% gelatin (Sigma Chemical). Seeded slides were placed in 100-mm culture dishes with 10 ml of complete culture medium and incubated at 37°C and 5% CO2. Four days after the seeding, ECs were confluent (∼106 cells/slide). Confluent ECs were serum starved overnight in DMEM supplemented with 0.5% FBS, 1% penicillin-streptomycin, 2% sodium pyruvate, 1% nonessential amino acids, 2 mM l-glutamine, and 25 mM HEPES. Some cells were treated with 9 or 90 μM H2O2for 15 min before shear exposure. Other cells were incubated with 5 or 20 mM NAC for 4 h before shear exposure, or they were incubated with 100 μM PDTC for 4 h before shear exposure. Each slide comprised of one side of a parallel-plate flow chamber, and the chamber was connected at both ends to a reservoir forming a flow-loop system (19). A hydrostatic pressure-drop system was used to control the flow through the chamber. The wall shear stress on the cell monolayer was calculated using the following equation: τ = (6μQ)/(h 2 w), where τ is the wall shear stress (dyn/cm2), μ is the viscosity of the medium (0.01 dyn ⋅ s ⋅ cm−2), Q is the flow rate (cm3/s),h is the channel height (0.025 cm), and w is the channel width (2.5 cm) (2, 7). The flow circuit was primed with 20 ml of DMEM (with 25 mM HEPES) or ATP-free M199. In some experiments, cells were not serum starved before shear exposure and were subsequently sheared in ATP-free M199 containing 100 μM Na3VO4. Each EC monolayer was exposed to steady laminar shear stress of 20 dyn/cm2, typical of the arterial circulation, in recirculating flow systems maintained at 37°C and in the presence of 5% CO2 for 5–30 min. Static controls were maintained in the incubator for 5–30 min in a medium that was the same as the perfusion medium.

Western blot for cellular tyrosine phosphorylation.

At the end of the shearing period, the cells were scraped off the slides in ice-cold PBS, and cell suspensions were centrifuged at 5,000 rpm for 5 min. The cell pellets were resuspended in ice-cold Triton X-100 lysis buffer consisting of 145 mM NaCl, 0.1 mM MgCl2, 15 mM HEPES, 10 mM EGTA, 1% Triton X-100, 1 mM Na3VO4, and protease inhibitors: 4-(2-aminoethyl)benzenesulfonyl fluoride (25 g/l), leupeptin (50 mg/l), and chymostatin, aprotinin, and pepstatin (25 mg/l each). The lysates were sonicated, and their protein content was measured using the bicinchoninic acid protein assay (Pierce, Rockford, IL) and normalized with 2× nonreducing sample buffer (0.5 M Tris ⋅ HCl, pH 6.8, 20% glycerol, 10% SDS, and 0.1% bromphenol blue). To study tyrosine phosphorylation in cytosolic fractions, cell lysates were sonicated and centrifuged at 14,000 rpm for 10 min at 4°C. The cytosolic protein content (supernatant) was measured and normalized with sample buffer, as for the whole cell lysates. Before electrophoresis, all samples were heated in a 95°C water bath for 5 min.

Protein samples were applied to 4–20% or 10–20% Tris-glycine gels (Novex, San Diego, CA). SDS-PAGE was performed using a Laemmli discontinuous buffer system under nonreducing conditions, and the separated proteins were electrophoretically transferred onto polyvinylidene difluoride (PVDF) membranes (DuPont NEN, Boston, MA). PVDF membranes were washed in PBS for 10 min, incubated for 1 h in blocking buffer (0.2% I-Block and 0.1% Tween-20 in PBS; Tropix, Bedford, MA), and reacted for 1 h with the primary antibody 4G10, a mouse antiphosphotyrosine monoclonal antibody (UBI, Lake Placid, NY), diluted 1:1,000 in blocking buffer. The membranes were then washed in blocking buffer and incubated for 30 min with a biotin-labeled goat anti-mouse secondary antibody (Zymed, San Francisco, CA) diluted 1:5,000 in blocking buffer. After they were washed, the membranes were incubated with Avidx-AP conjugate diluted 1:10,000 in blocking buffer for 20 min, then washed with blocking buffer and assay buffer (0.1 M diethanolamine and 1 mM MgCl2; Tropix). Finally, the membranes were incubated for 5 min in chemiluminescent substrate solution (0.24 mM CSPD and 1:20 Nitro-Block in assay buffer; Tropix). Detection was performed by placing each membrane in contact with luminescence detection film (Hyperfilm-ECL, Amersham, Arlington Heights, IL).

Immunodetection of oxidatively modified intracellular proteins.

OxyBlot provides the methodology for immunodetection of carbonyl groups that are introduced into intracellular proteins by oxidative reactions (44). Protein carbonyl content in BAECs exposed to 75 μM H2O2was shown to triple within 10 min and reached a plateau by 15 min (12). According to the instructions, the carbonyl groups of the protein side chains in cell lysates were derivatized to 2,4-dinitrophenyl (DNP)-hydrazone by reaction with DNP-hydrazine. The DNP-derivatized protein samples were separated by SDS-PAGE, then subjected to Western blotting, as described above. The membranes were incubated with primary antibody specific to the DNP moiety of the proteins, incubated with a horseradish peroxidase-conjugated goat anti-rabbit IgG, and then treated with chemiluminescent reagents, as described above.

Immunoprecipitation and Western blot for MAP kinase activation.

After their protein content was measured, all cytosolic samples were normalized to the sample with the lowest protein content. Each sample was rotated end-over at 4°C for 4 h with 20 μl of an agarose-conjugated antiphosphotyrosine slurry (UBI) and centrifuged at 10,000 rpm for 30 s to separate the antiphosphotyrosine-agarose beads from the cell lysates. The supernatant was discarded. The beads were washed three times in PBS, resuspended in 50 μl of 2× reducing sample buffer (0.5 M Tris ⋅ HCl, pH 6.8, 20% glycerol, 10% SDS, 0.1% bromphenol blue, and 5% β-mercaptoethanol; Novex), and then heated in a 95°C water bath for 5 min to dissociate the proteins from the beads. Immunoprecipitated proteins were separated by SDS-PAGE, as described above. Western blot analysis was performed using a rabbit anti-MAP kinase polyclonal antibody (erk1-CT; UBI) diluted 1:1,000 in blocking buffer followed by a biotinylated goat anti-rabbit antibody (Tropix) diluted 1:4,000 in blocking buffer. All other reagents were identical to those used in the antiphosphotyrosine Western blots.

Cell infection with recombinant adenoviruses.

The replication-deficient adenovirus encoding the epitope-tagged dominant-negative Rac1 cDNA (Ad.N17rac1) was constructed by homologous recombination in 293 cells with use of the adenovirus-based plasmid JM17, as previously described (56). The replication-deficient adenovirus Ad.βgal containing the Escherichia coli Lac Z gene has also been previously described (57). All viruses were amplified and titered in 293 cells and purified on CsCl gradients (23). Infections were done overnight in 80% confluent BAECs. Western blot analysis of Rac1 expression used an antibody directed at the myc-epitope tag (9E10; Santa Cruz Biotech, Santa Cruz, CA), which identified the N17rac1 gene product, as described elsewhere (56). X-gal staining of BAECs infected with the Ad.βgal at a multiplicity of infection of 100 showed >90% transfection efficiency.

RESULTS

Shear stress effects on cellular tyrosine phosphorylation.

Exposure of serum-starved (quiescent) BAECs to a calculated shear stress of 20 dyn/cm2 without Na3VO4in the perfusion medium caused a rapid increase in the phosphotyrosine content of whole cell lysates that peaked within 15 min after the onset of shear stress and was still above baseline at 30 min (Fig.1 A). When nonquiescent BAECS were exposed to shear stress for 5, 20, or 30 min in the presence of the tyrosine phosphatase inhibitor (100 μM), there was much more pronounced cellular tyrosine phosphorylation in response to shear stress (Fig. 1 B). There was a time dependency between shear stress and tyrosine phosphorylation with saturation of the total cellular tyrosine phosphorylation at 30 min after initiation of shear (∼100% increase over the static control; Fig. 1 B). Experiments were repeated at least three times, and a representative experiment is shown. All corresponding static controls were incubated for 30 min in a medium that was the same as the perfusion medium (except for lane 2 in Fig.1 B). The addition of 100 μM Na3VO4for 30 min into the culture medium of static cells did not cause any significant changes at the cellular phosphotyrosine levels compared with static cells that were not exposed to Na3VO4(cf. lanes 2 and3 in Fig.1 B). The static control cells in Fig. 1 A were different from those in Fig. 1 B, because all cells in Fig.1 A were made quiescent before treatment and analysis.

Fig. 1.

Endothelial cell exposure to shear stress causes an increase in total cellular tyrosine phosphorylation. A: confluent quiescent bovine aortic endothelial cells (BAECs) exposed to shear stress (20 dyn/cm2) for 15 or 30 min in absence of Na3VO4. B: confluent nonquiescent BAECs exposed to shear stress for different times up to 30 min in presence of 100 μM Na3VO4. All corresponding static controls were exposed for 30 min to a medium that was the same as perfusion medium, except lane 2 in B, which was exposed to medium without Na3VO4. Whole cell lysates were size fractionated by SDS-PAGE, transferred onto polyvinylidene difluoride membranes, and immunoblotted with an antiphosphotyrosine monoclonal antibody.

H2O2 effects on cellular tyrosine phosphorylation.

Static nonquiescent BAECs were exposed to 9 or 90 μM H2O2for 15 min, and cell lysates were probed with an antiphosphotyrosine antibody. In the experiment shown in Fig.2 A, representative of two identical experiments, H2O2enhanced the total cellular tyrosine phosphorylation in a dose-dependent manner compared with static controls that were not exposed to H2O2. When static cells were exposed to 90 μM H2O2, the increase in cellular tyrosine phosphorylation was ∼200% (Fig.2 A). When nonquiescent cells were incubated with 9 μM H2O2for 15 min before 20 min of shear exposure (in the presence of 100 μM Na3VO4), the increase in total cellular tyrosine phosphorylation was even greater than in sheared cells that were not preincubated with H2O2(Fig. 2 B). Static control cells were exposed to a medium that was the same as the perfusion medium for 20 min (Fig. 2 B). Therefore, exposure to H2O2and application of fluid shear stress may have additive effects on cellular tyrosine phosphorylation.

Fig. 2.

Prior cell exposure to H2O2increases total cellular tyrosine phosphorylation in static or sheared endothelial cells. A: confluent nonquiescent BAECs treated for 15 min with 9 or 90 μM H2O2. B: confluent nonquiescent BAECs treated for 15 min with 9 μM H2O2, then exposed, or not, for 20 min to shear stress in presence of Na3VO4. Static control cells in B were incubated in a medium that was the same as perfusion medium for 20 min. Western blot analysis of whole cell lysates was performed with an antiphosphotyrosine monoclonal antibody.

Shear stress effects on the intracellular redox status.

Exposure of quiescent BAECs to 5 min of arterial shear stress (20 dyn/cm2) without Na3VO4caused a marked increase in the amount of oxidatively modified proteins in whole cell lysates, as measured by the OxyBlot oxidized protein detection kit (Fig. 3). Within 15 min of exposure to shear stress, there was a drop in the levels of protein oxidation, possibly because of counteraction of ROS by the intracellular antioxidant defense systems. When ECs were incubated with the membrane-permeant antioxidant NAC for 4 h before 15 min of shear exposure (without Na3VO4), it was shown that 20 mM NAC partly inhibited the shear-induced cellular protein oxidation (Fig. 3), whereas 5 mM NAC had no effect (not shown). NAC at 20 mM had no effect on protein oxidation of static controls not exposed to H2O2(not shown). This experiment was repeated twice with similar results.

Fig. 3.

Shear stress increases intracellular reactive oxygen species (ROS) concentration, and increase is counteracted by prior exposure of endothelial cells to antioxidantN-acetylcysteine (NAC). Confluent quiescent BAECs were exposed to shear stress for 5 or 15 min in absence of Na3VO4. Some cells were incubated with 20 mM NAC for 4 h before 15 min of shear exposure in absence of Na3VO4. For immunoblot detection of carbonyl groups introduced into cellular proteins by ROS, carbonyl groups in protein side chains were derivatized to 2,4-dinitrophenyl (DNP)-hydrazone by reaction with DNP-hydrazine, and DNP-derivatized proteins were separated by SDS-PAGE, then subjected to Western blotting with an antibody against DNP moiety.

Effects of antioxidants on shear-induced cytosolic tyrosine phosphorylation.

To investigate whether ROS are involved in the signal transduction caused by shear stress, nonquiescent BAECs were incubated with 5 or 20 mM NAC for 4 h before 30 min of shear exposure in the presence of Na3VO4. Static controls were incubated for 30 min in a medium that was the same as the perfusion medium. When cytosolic proteins were probed for antiphosphotyrosine, it was found that NAC effectively inhibited tyrosine phosphorylation in the cell cytosolic fraction in a dose-dependent manner (Fig. 4). Incubation with micromolar concentrations of a thiol antioxidant, PDTC, for 4 h before shear exposure also inhibited the shear-induced cytosolic tyrosine phosphorylation (Fig. 4). Neither of the antioxidants had any effect on the cytosolic tyrosine phosphorylation of static controls (not shown). This experiment was repeated at least twice. In the experiment shown, 30 min of shear stress caused a 50% increase in cytosolic tyrosine phosphorylation compared with static control, whereas previous exposure to NAC (20 mM) or PDTC (100 μM) followed by shear resulted in only a 10% and a 5% increase in cytosolic tyrosine phosphorylation, respectively, compared with static control. NAC also inhibited the increase in cytosolic tyrosine phosphorylation in quiescent BAECs that were sheared in the absence of Na3VO4(not shown).

Fig. 4.

ROS mediate shear-induced increase in cytosolic tyrosine phosphorylation. Confluent nonquiescent BAECs were exposed to shear stress (20 dyn/cm2) for 30 min in presence of Na3VO4. Static control cells were incubated in a medium that was the same as perfusion medium for 30 min. Some cells were incubated with 5 or 20 mM NAC or 100 μM pyrrolidine dithiocarbamate (PDTC) for 4 h before shear exposure. Cytosolic proteins were size fractionated by SDS-PAGE, transferred onto polyvinylidine difluoride membranes, and immunoblotted with an antiphosphotyrosine monoclonal antibody.

Effects of antioxidants on shear-induced MAP kinase tyrosine phosphorylation.

Because tyrosine phosphorylation is necessary for MAP kinase activation, we investigated the ability of antioxidants to inhibit the shear-induced tyrosine phosphorylation of MAP kinase. In Fig.5 A, quiescent cells were sheared in the absence of Na3VO4; in Fig. 5 B, nonquiescent cells were sheared in the presence of Na3VO4. In either case, cytosolic fractions were immunoprecipitated with an antibody to phosphotyrosine and immunoprecipitates were probed with an antibody (erk1-CT) to the phosphorylated 42- and 44-kDa isoforms of MAP kinase (NIH/3T3 cell lysates were included as a positive control). There was a time-dependent relationship between shear stress and MAP kinase tyrosine phosphorylation, with peak phosphorylation occurring at 5 min after exposure to shear in Fig.5 A and at 10 min after exposure to shear in Fig. 5 B (longer times are not shown). As was the case with cellular tyrosine phosphorylation, NAC blocked the shear-induced increase in tyrosine phosphorylation of MAP kinase in a dose-dependent manner (Fig.5 B). These experiments were repeated twice with similar results: BAEC incubation with 5 mM NAC followed by exposure to 5 min of shear stress did not cause any change in the tyrosine phosphorylation levels of MAP kinase compared with cells that were exposed to 5 min of shear without the NAC preincubation (Fig.5 B). However, BAEC incubation with 20 mM NAC followed by exposure to shear stress caused the shear-induced tyrosine phosphorylation of MAP kinase to remain slightly above the baseline levels (in the case of treatment with 20 mM NAC followed by 5 min of flow; Fig. 5 B) or at the baseline levels (in the case of treatment with 20 mM NAC followed by 10 min of flow; Fig. 5).

Fig. 5.

ROS mediate shear-stimulated increase in mitogen-activated protein (MAP) kinase tyrosine phosphorylation. Confluent BAECs were exposed to shear stress for 5 or 10 min. Some cells were incubated with 5 or 20 mM NAC for 4 h before shear exposure. A: quiescent endothelial cells sheared in absence of Na3VO4.B: nonquiescent endothelial cells exposed to 100 μM Na3VO4during shearing period. In either case, static controls were maintained in a medium that was the same as perfusion medium for 10 min. Cell lysates were immunoprecipitated with an antiphosphotyrosine monoclonal antibody and immunoblotted with an anti-MAP kinase polyclonal antibody.

Role of Rac1-mediated ROS production in cellular and MAP kinase tyrosine phosphorylation.

Because the GTPase Rac1 is an integral part of the NADPH oxidase complex and regulates ROS production in phagocytic and nonphagocytic cells (1, 38, 58), we asked whether the flow-induced cellular tyrosine phosphorylation was dependent on Rac1-mediated ROS production. Adenovirus-mediated expression of the dominant-negative Rac1, N17rac1, in BAECs markedly attenuated the flow-mediated increase in cellular tyrosine phosphorylation (Fig.6 A; nonquiescent BAECs that were sheared in the presence of Na3VO4). Static control cells infected with Ad.βgal displayed a profile of tyrosine-phosphorylated proteins that was slightly different from that of noninfected cells (cf. lane 3 of Fig. 1 B with lane 1 of Fig. 6 A). However, Ad.βgal-infected cells displayed a tyrosine phosphorylation response to flow that was similar to that of noninfected cells (cf.lane 6 of Fig.1 B with lane 2 of Fig. 6 A). When quiescent Ad.βgal- or Ad.N17rac1-infected BAECs were sheared in the absence of Na3VO4and cytosolic fractions were analyzed for MAP kinase tyrosine phosphorylation, expression of N17rac1, and not of βgal, completely blocked the shear-induced tyrosine phosphorylation of MAP kinase (Fig.6 B). Each of these experiments was repeated twice with identical results.

Fig. 6.

Rac1-dependent ROS production mediates shear-induced increase in total cellular and MAP kinase tyrosine phosphorylation. BAECs infected with Ad.βgal or Ad.N17rac1 were exposed to shear stress for up to 30 min.A: nonquiescent cells sheared in presence of Na3VO4. Western blot analysis of lysates was performed with an antiphosphotyrosine monoclonal antibody.B: quiescent cells sheared in absence of Na3VO4. Cell lysates were immunoprecipitated with an antiphosphotyrosine monoclonal antibody and immunoblotted with an anti-MAP kinase polyclonal antibody. Static controls were maintained in a medium that was the same as perfusion medium for 30 (A) or 10 min (B).

DISCUSSION

The present study suggests that ROS act as signal-transducing molecules in ECs exposed to fluid shear stress or, more appropriately, to a step change in fluid shear stress. ROS have already been implicated in the signal transduction of PDGF in VSMCs (57). ROS were also involved in cyclic strain-induced cellular responses in ECs, specifically, the MCP-1 gene expression (62) and release of PAI-1 (10). Recent studies demonstrated that shear flow to ECs can induce intracellular ROS generation, which results in increased ICAM-1 (11) and c-fos gene expression (28). Shear stress-mediated ROS production may be partly responsible for the activation by shear stress of the redox-sensitive transcription factors nuclear factor-κB and AP-1 (5, 41) and for activation of Ras, the small GTP-binding protein upstream of MAP kinase, which is a signaling target of ROS (42) and is also activated by shear stress (46). Recent findings on gene induction of Cu-Zn superoxide dismutase (SOD) and mitochondrial Mn SOD by shear stress suggest the activation of cellular defense mechanisms against oxidative stress caused by EC exposure to shear stress (31, 59).

Our findings on the effects of arterial shear stress on cellular tyrosine phosphorylation agree with those of other investigators. Specifically, Ayajiki et al. (3) showed that 30 min of arterial shear stress markedly increased the cellular tyrosine phosphorylation of 103- and 114-kDa proteins in cultured HUVECs. This range of proteins corresponds to the major band observed in cell lysates of quiescent BAECs that were sheared in the absence of the tyrosine phosphatase inhibitor (Fig. 1 A). In agreement with our findings on shear-induced MAP kinase tyrosine phosphorylation (Fig. 5 A), Tseng et al. (60) showed that arterial shear stress induced tyrosine phosphorylation and activation of MAP kinase in BAECs, with a peak activation time of 2–10 min after the onset of shear stress.

In this study we demonstrated the importance of intracellular ROS as modulators for shear-induced cellular tyrosine phosphorylation and, in particular, tyrosine phosphorylation of MAP kinase, by providing the first evidence that the shear-stimulated tyrosine phosphorylation response does require ROS. Thus the antioxidant NAC at 20 mM, which acts as a scavenger for ROS intermediates and as a precursor for glutathione (48, 53), inhibited the shear-induced cytosolic and MAP kinase tyrosine phosphorylation (Figs. 4 and 5). Reducing thiol agents, such as PDTC, affect the expression of thioredoxin, an oxidoreductase with antioxidant functions (50). PDTC at 100 μM was effective in inhibiting tyrosine phosphorylation in cytosolic fractions of sheared BAECs (Fig. 4). NAC at 20 mM was also shown to inhibit the shear-induced intracellular protein oxidation that preceded in time the cellular tyrosine phosphorylation signal but coincided with the MAP kinase tyrosine phosphorylation signal (Fig. 3). The importance of intracellular ROS in shear-induced MAP kinase activation agrees with findings by Sundaresan et al. (57), who demonstrated that NAC caused a concentration-dependent reduction of MAP kinase activation in VSMCs stimulated by PDGF.

The pathway by which ROS are generated is best characterized in phagocytic cells. Activation of phagocytic cells leads to the assembly of the NADPH oxidase enzymatic complex, which transfers electrons from NADPH to molecular O2 with the subsequent generation of O2 and appears to be regulated by the GTPase Rac1 (1, 8). Many, but not all, of the components of the NADPH oxidase complex have been shown to be expressed in a variety of nonphagocytic cell types, including ECs (35). In this report we provide the first evidence that the small GTP-binding protein Rac1 functions as a regulator of the shear-induced ROS production. This agrees with the results by Sundaresan et al. (58), who demonstrated that Rac1, as well as Ras, regulates the increase of intracellular ROS in NIH/3T3 cells on stimulation with growth factors, the receptors of which have intrinsic tyrosine kinase activity, or cytokines. Several other second-messenger responses to shear stress, such as changes in phosphatidylinositol lipid metabolism, intracellular free Ca2+ levels, and prostacyclin and NO release, are similar to responses resulting from agonist-receptor coupling, suggesting that they may share signal transduction pathways. In addition, Kim et al. (38) recently demonstrated a requirement for Rac1, but not for Ras proteins, in the generation of intracellular ROS after reoxygenation of hypoxic HUVECs.

All our shear experiments were repeated in the presence of the tyrosine phosphatase inhibitor Na3VO4(100 μM) in the perfusion medium to amplify the shear-induced tyrosine phosphorylation signal. Use of Na3VO4during the shear exposure is controversial, because1) vanadate has been demonstrated to stimulate NADPH oxidation (47), 2) vanadate is shown to increase the tyrosine kinase activity of the endogenous insulin receptor kinase in adipocytes (36),3) vanadate is known to act synergistically with compounds that activate protein kinase C to form ROS, which enhance protein tyrosine phosphorylation in macrophages (20), and 4) vanadate is an inhibitor of Na+-K+-ATPase (37). In our case, however, it seems unlikely that Na3VO4by itself increases tyrosine kinase activity, since addition of 100 μM Na3VO4into the medium of static control cells, for a maximum of 30 min, did not cause any significant changes in total cellular tyrosine phosphorylation (Fig. 1 B). Only with exposure to shear stress did Na3VO4greatly enhance the total cellular tyrosine phosphorylation in BAECs (Fig. 1). In contrast, it had no visible additive effect on the shear-induced tyrosine phosphorylation of MAP kinase (Fig. 5). The fact that the effect of ROS on cellular tyrosine phosphorylation was potentiated by added vanadate is due to the combined activation of protein tyrosine kinases and the inactivation of protein tyrosine phosphatases (25, 51). Although orthovanadate could be working on its own, it is more likely that it combines with H2O2to form pervanadate, a more potent inhibitor of protein tyrosine phosphatases (30). Similarly, Na3VO4alone, at micromolar concentrations, produced no increase in DNA binding of the transcription factor AP-1, whereas the same concentrations of Na3VO4greatly enhanced the H2O2-mediated activation of AP-1 in porcine aortic ECs (5).

Laurindo et al. (43) were the first to demonstrate free radical generation, in vivo and ex vivo, due to step increases in shear stress. Their techniques for measuring ROS, electron paramagnetic resonance spectroscopy and measurements of ascorbic radicals in plasma, allowed them to detect mainly extracellular O2 release. This was also suggested by the fact that the measured shear-induced increase in ROS was abolished by addition of SOD, a superoxide scavenger that does not cross intact EC membranes or intercellular junctions. However, different cellular enzymatic systems have been recognized as sources for ROS production in ECs exposed to reoxygenation after hypoxia or anoxia, depending on whether the technique employed measures intracellular or extracellular ROS. Specifically, with use of electron paramagnetic resonance spectroscopy, it was found that BAECs subjected to anoxia followed by reoxygenation generate oxygen free radicals, and the signal was partially inhibited by allopurinol, an inhibitor of XO (65). XO reduces molecular O2 and generates intracellular O2 , which then dismutates to form H2O2. With use of a fluorometric assay that measured the extracellular H2O2release by bovine pulmonary artery ECs on reoxygenation, it was shown that allopurinol had no effect and only the flavoprotein inhibitor diphenylene iodonium (DPI) reduced the H2O2release (64). DPI inhibits the membrane-bound enzyme NADPH oxidase, which reduces molecular O2 to O2 , and other flavoproteins, such as XO, ecNOS, and NADH dehydrogenase (18, 24, 55). With use of a fluorometric assay sensitive to intracellular H2O2and peroxynitrite, it was found that allopurinol andN G-methyl-l-arginine, an inhibitor of ecNOS, reduced intracellular ROS, suggesting that XO generates O2 , which partly dismutates to H2O2and partly reacts with NO to form peroxynitrite (63). Although the ROS sources in cultured ECs have been studied extensively for hypoxia-reoxygenation, the shear-induced ROS production was only recently studied. Intracellular O2 levels in sheared HUVECs were measured by lucigenin-amplified chemiluminescence (11), 2′,7′-dichlorofluorescin fluorescence (28), and ethidium fluorescence (16), and, in each case, steady laminar shear stress was found to increase the EC intracellular ROS. An elevated level was measured by 15 min, remained elevated as flow continued for up to 6 h, but returned to baseline at 24 h.

Because ECs are a rich source of XO, XO is a suitable candidate enzymatic system for flow-triggered intracellular ROS generation, with possible contributions by ecNOS, cyclooxygenase, phospholipase A2(PLA2), and the mitochondrial (NADH dehydrogenase and cytochrome oxidase) and microsomal (cytochromeP-450) electron transport chains. However, our study proved that Rac1 is a major component in generating ROS in sheared ECs, indicating the possible involvement of the plasma membrane-bound NADPH oxidase complex. Indeed, NADPH oxidase activity in cultured HUVECs was found to be induced by steady laminar shear stress at 1 and 5 h, a transient response that returned to baseline at 24 h (no shorter times were tested) (16). There are plenty of data available that suggest that exposure to shear stress might activate the NADPH oxidase enzyme. 1) MAP kinase is activated by shear stress, PLA2 is activated by MAP kinase (34), and products of PLA2 (largely arachidonic acid) are known to mediate activation of NADPH oxidase (14).2) Superoxide production in NIH/3T3 fibroblasts stably transfected with a constitutively active isoform of Ras was found to be inhibited by DPI (32). Because Ras is upstream of MAP kinase and is activated by shear stress (46), it is expected that NADPH oxidase and other flavoproteins might be a source for shear-induced ROS. 3) The activity of NADPH oxidase was increased in cyclically strained porcine aortic ECs (27), suggesting that NADPH oxidase might play a role in the generation of oxidative stress in the mechanically deformed vessel wall.

Because shear-induced ROS production mediates the tyrosine phosphorylation and, presumably, activation of MAP kinase, which may be one of the kinases that phosphorylate ecNOS, leading to the sustained release of NO, our results suggest that EC free radical production may exert an autocrine role in the control of vascular tone by shear stress. It is also known that shear stress-mediated NO formation inhibits apoptosis of cultured HUVECs (17, 26). If ROS, through MAP kinase, regulate the sustained release of NO, then shear-induced ROS generation may indirectly interfere with cell death signal transduction and contribute to EC integrity. Finally, flow-triggered oxidative stress may also play a role in hemodynamic force-induced pathological conditions, such as the endothelial dysfunction associated with atherosclerosis, the endothelial dysfunction after hypoxia and reperfusion, and the superoxide-dependent vasoconstriction that occurs after balloon angioplasty.

Acknowledgments

We acknowledge the expert technical assistance of S. J. Gips, C. C. Wilhide, and T. C. Huang during the different phases of the study. We thank A. Hall for the N17rac1 cDNA and R. G. Crystal for the Ad.βgal.

Footnotes

  • Address for reprint requests and other correspondence: B. R. Alevriadou, The Johns Hopkins University School of Medicine, BME Dept., Traylor Bldg., Rm. 619, 720 Rutland Ave., Baltimore, MD 21205 (E-mail:ralevria{at}bme.jhu.edu).

  • B. R. Alevriadou was supported in this work by National Heart, Lung, and Blood Institute Grant HL-54089, a Whitaker Biomedical Engineering research grant, and a grant from the Center for Alternatives to Animal Testing. K. Irani was supported by the Johns Hopkins Clinician Scientist Award, the Bernard Foundation, and an endowment from Mr. and Mrs. Abraham Weiss.

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

REFERENCES

View Abstract