INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

AS THE INNER LINING of the vessel wall, vascular endothelial cells (ECs) are poised to act as a signal transduction interface for hemodynamic forces. Depending on the flow magnitude and proximity to vessel bifurcations, fluid shear stress may exert different effects on the vascular endothelium and, hence, regulate vessel wall function and structure as well as potentiate atherogenesis (8). The concept that fluid flow provides two distinct stimuli, namely, temporal gradients in shear (TGS) and steady shear, has been previously suggested in endothelial prostacyclin release and NO and cGMP generation as well as pinocytosis (9-11, 26). Recent studies from our laboratory (3, 4) have demonstrated that TGS but not steady shear stimulates sustained endothelial expression of atherogenesis-related genes for platelet-derived growth factor (PDGF)-A and macrophage chemoattractant protein (MCP)-1 as well as the activation of cell proliferation-related signal pathways involving nuclear factor (NF)-B, egr-1, c-fos, and mitogen-activated protein kinases (MAPKs). In contrast, only steady shear continuously upregulates three potential antiatherogenic genes: manganese superoxide dismutase, cyclooxygenase-2, and endothelial nitric oxide (NO) synthase (NOS) (40). While it is apparent that TGS and steady shear are two distinct fluid shear stimuli that may contribute differently to vascular physiology and pathology and that TGS is a potent mitogenic mechanic stimulus to cultured ECs, there is little knowledge on the effect of TGS on EC proliferation as well as the underlying signaling cascade.

MAPKs are important intermediaries in signal transduction pathways, acting as a point of convergence or integration for various extracellular stimuli such as growth factors, hormones, neurotransmitters, and physical and chemical stress agents (12). By regulating the activity of transcription factors such as c-Fos (3), c-Jun, Egr-1 (6), c-Myc (38), NF-B (4), and cyclin D1, MAPKs ultimately participate in the regulation of cell growth, differentiation, and metabolism. At least four members of the MAPK family, extracellular signal-regulated kinases 1 and 2 (ERK1/2) (3, 41), c-Jun NH₂-terminal kinase (JNK) (29), and Big MAPK (44), are involved in fluid shear-induced endothelial signaling pathways. The 44- and 42-kDa ERK1/2 isoforms are ubiquitously expressed and activated by dual-specific MAPK kinases [MAP or ERK kinase 1 and 2(MEK1/MEK2)] in response to diverse stimuli (12). Fluid shear stress has been demonstrated to rapidly activate ERK1/2 in ECs in a biphasic manner, with activity peaking at 5-10 min and returning to basal level by 30 min of shear stress (23, 41). The activation of ERK1/2 by shear stress involves protein kinase C (41), pertussis toxin (PTX)-sensitive G_i2 (23), and Ras-p60src (21) pathways. Moreover, specific inhibition of ERK diminished fluid shear stress-induced c-fos expression (3) and Egr-1 promoter activity (6) in ECs. To distinguish the precise influences of TGS and steady shear on ECs, several laminar fluid flow profiles have been created in our laboratory: impulse, step, and ramp flow (see METHODS) (3, 4). Step flow, which has been widely applied in previous in vitro flow studies, contains both sharp increase and steady flow components, whereas the impulse and ramp flow contain only temporal gradient and steady shear, respectively. With the use of this unique system, we found that the activation of ERK1/2 was very sensitive to TGS (impulse flow and the onset of step flow) but not steady shear (ramp flow and steady component of step flow). Interestingly, ERK1/2 phosphorylation was actually downregulated by ramp flow (3).

The aim of the current study was to further differentiate the upstream signaling pathways that mediate the effect of TGS and steady shear on endothelial ERK1/2 activation and to investigate the role of this pathway in EC proliferation. Activation of G proteins, particularly G_q/11 and G_i3, may serve as the earliest signaling step for the shear stress-induced response in ECs (14). Fluid shear-induced NO superoxide and peroxynitrite appear to function as signaling molecules in flow-dependent activation of MAPKs (13, 19, 30). We therefore sought to examine the roles of G proteins, NO, and reactive oxygen species (ROS) in fluid shear-dependent ERK signaling pathways as well as the fluid shear-associated growth response in ECs. By adding a well-defined pulsatile flow, which contained multiple flow impulses, to the above-mentioned three laminar flow profiles, we were able to mimic the biological counterpart of TGS in vitro. We observed a more enhanced and sustained activation of ERK1/2 induced by TGS compared with steady flow. We were able to demonstrate a dual role of NO in regulating ERK1/2 activity depending on its concentration and the flow profile applied. With the use of pharmacological inhibitors as well as antisense oligodeoxynucleotides (ODN) against G protein -subunits (G_q/11 and G_i3), a TGS-G protein-NO/ROS-ERK1/2 signaling pathway in EC proliferation was demonstrated.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cell culture and experimental preparation. The isolation of primary human umbilical vein ECs (HUVECs) was performed as previously described (10). Briefly, cells were harvested from human umbilical veins by collagenase (Boehringer-Mannheim) treatment and were then seeded onto glass microscope slides. The cells grew to confluence within 3-4 days in ATP-free M199 medium supplemented with 20% fetal bovine serum (FBS; Hyclone), 2 mmol/l L-glutamine, 0.5 U/ml penicillin, and 0.05 mg/ml streptomycin (Sigma). Before shear exposure, the confluent cells were serum starved for 48 h in 0.5% FBS-supplemented M199 to establish quiescence in the monolayers. For pharmacological studies, N^G-amino-L-arginine (L-NAA; Alexis), N-acetyl-cysteine (NAC; Sigma), or PD98059 (Calbiochem) was incubated (independently or in combination) with the ECs for 1 h before the ECs were subjected to shear. For NO donor experiments, quiescent control ECs and impulse flow-treated ECs were incubated with spermine-NONOate (SPR/NO; Alexis) for 10 min, respectively. All cell cultures were maintained in a humidified 5% CO₂-95% air incubator at 37°C.

1

1

4

1

1

Flow experiments. Cell monolayers on glass slides were subjected to well-defined laminar flow for various times in a parallel-plate flow chamber, where the perfusing medium was driven by a syringe pump (pump 44, Harvard Apparatus) or a constant head flow loop as described previously (10). The computer-controlled syringe pump was programmed to generate different flow profiles. The flow rate (and hence shear stress) was changed in 1-s microsteps with volume-flow increments determined from the slope of the overall increase in shear stress. Four well-defined laminar flow profiles were generated to which ECs were exposed (4). These profiles were designed and used to separate the effects of different dynamic flow stimuli presented to ECs. The following profiles were applied: 1) step flow (instantaneous shear stress increase from 0 to 16 dyn/cm², followed by steady shear for a sustained period); 2) ramp flow (shear stress smoothly transited from 0 to 16 dyn/cm² over 2 min and then sustained for a desired period); 3) impulse flow (a 3-s flow impulse of 16 dyn/cm²); and 4) pulsatile flow (multiple flow impulses with 3-s intervals).

ODN design and transfection. On the basis of the mRNA sequences of human G_q (5), G₁₁ (NCBI GenBank AF011497/gi:2286216), and G_i3 (20), the following ODN were synthesized and used: anti-G_q/11com, 5'-CGAGCAGCAGCAGCTTGAGCTC-3'; anti-G_i3, 5'-GCCCATGGCGGCGGCGGGAGAG-3'; and scrambled nonspecific ODN, 5'-TACCGGGAATGCTGGCAACCGC-3'. The ODNs were synthesized at Genosys Biotechnologies (The Woodlands, TX).

Protein preparation and Western blotting. The cells were washed twice with PBS and lysed at 4°C with SDS sample buffer [62.5 mmol/l Tris · HCl (pH 6.8), 2% (wt/vol) SDS, 10% glycerol, 50 mmol/l dithiothreitol, and 0.1% (wt/vol) bromphenol blue]. Cellular lysate was scraped, transferred into a microtube, and sonicated for 10 s, followed by centrifugation (15 min, 4°C, 14,000 rpm). The protein concentration in the supernatant was determined by the Bio-Rad colorimetric protein assay method (Bio-Rad Laboratories). Twenty micrograms of total cellular protein were size fractionated in a 10% SDS-polyacrylamide gel and electroblotted to nitrocellulose membranes. The membranes were stained with Ponceau S (Sigma) to confirm that equal amounts of protein were loaded in each lane and that the transfer was homogeneous. After the membranes were washed with 20 mmol/l Tris · HCl and 0.5 mol/l NaCl (TBS; pH 7.4) to remove the stain, the filters were blocked with 5% nonfat milk in TBS overnight at 4°C and incubated with diluted primary antibodies [rabbit anti-human phosphospecific ERK1/2 antibody (New England Biolabs), anti-human ERK1/2 antibody (Santa Cruz), anti-G_q/11 (Chemicon), or anti-G_i3 (Chemicon)], followed by incubation with horseradish peroxidase-conjugated goat anti-rabbit (New England Biolabs) antibodies for 1 h. The filters were then washed twice with 0.05% Tween 20 in TBS (pH 7.4) for 5 min each and once with TBS. Protein-antibody conjugates were detected by chemiluminescence (Super Signal CL-HRP, Pierce Chemical).

Cell proliferation analysis. Proliferation of HUVECs was examined using an in situ monoclonal antibody kit for the detection of bromodeoxyuridine (BrdU) incorporation into DNA (Boehringer-Mannheim). Briefly, cells on the slides were grown to 50% confluency. Immediately after exposure to impulse flow, slides were quickly removed from the chamber and incubated in M199-BrdU (10 µmol/l BrdU) for 24 h in 37°C incubator. Slides were fixed in 70% ethanol (in 50 mM glycine buffer; pH 2.0) and immunofluorescent stained for BrdU incorporation. BrdU-positive cells were visualized under a fluorescence microscope (Nikon, Diaphot TMD). Proliferating cells were counted at ×40 magnification under the microscope. Counts were averaged over four to five adjacent high-power fields (×40) per slide, and four slides were examined at each time points.

Statistics. The results are presented as means ± SE. The difference between each control and experimental group was evaluated by unpaired Student's t-test. A value of P < 0.05 was considered statistically significant.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

TGS induces sustained ERK1/2 activation. To determine the effects of different fluid flow profiles on ERK1/2, serum-starved ECs were subjected to ramp, step, impulse, and pulsatile flow, respectively. As shown in Fig. 1, a single 3-s impulse flow of 16 dyn/cm² induced a marked increase in phosphorylated ERK1/2 (pERK1/2) at 10 min, as detected by Western blotting using the phosphospecific anti-ERK1/2 antibody. pERK1/2 remained elevated at 30 min and returned to baseline at 1-2 h after the shear exposure. Pulsatile flow, which contains multiple flow impulses with 3-s intervals, showed similar stimulatory effects on ERK1/2 (Fig. 1B). In contrast, step flow (which contains a temporal gradient in shear at flow onset followed by steady shear) induced only a mild increase in pERK1/2 at 10 min (Fig. 1B), which returned to the basal level at 30 min (data not shown). Interestingly, ramp flow actually downregulated the level of pERK1/2 compared with the static control (Fig. 1B). These results indicated that TGS (impulse flow and pulsatile flow) induced a significantly higher and more sustained level of ERK1/2 activation compared with step flow, whereas ramp flow caused a decrease in pERK1/2. The finding that step flow only induced relatively mild and transient ERK1/2 activation was similar to that in a previous report (41).

View larger version (41K): [in this window] [in a new window]   Fig. 1.   Effect of fluid shear stress on activation of extracellular signal-regulated protein kinases 1 and 2 (ERK1/2) in human umbilical vein endothelial cells (HUVECs). Cells were serum starved for 48 h and then subjected to various flow regimes. Total cellular proteins were prepared at various time points (10-120 min) for Western blotting assay for the total and phosphorylated 44- and 42-kDa ERK (pERK1/2). A: time course of impulse flow-induced activation of ERK1/2. Relative levels of pERK1/2 were obtained by dividing the levels of pERK1/2 in cells exposed to 3-s impulse flow by that in static control cells. Data are presented as means ± SE (n = 3). B: different effects of ramp, step, impulse, and pulsatile flow on the activation of ERK1/2. ERK1/2 represent the total amount of cellular ERK1/2 detected using the anti-ERK1/2 antibody, and pERK1/2 represent the phosphorylated/activated ERK1/2 detected using the phosphospecific anti-ERK1/2 antibody, respectively. Static, serum-starved control cells.

NO mediates both TGS-induced activation and steady shear-induced downregulation of pERK1/2. Because fluid shear stress has been shown to stimulate NO production in ECs (11, 26), we examined whether NO is involved in flow shear-dependent changes in ERK1/2 activity. Before being subjected to impulse or ramp flow, the ECs were treated with the NOS inhibitor L-NAA at various concentration for 1 h. Treatment of cells with L-NAA alone had no effect on the basal level of pERK1/2 (data not shown). However, L-NAA dose dependently decreased impulse flow-induced ERK1/2 activation, with a significant reduction of the pERK1/2 level at a concentration of 300 µmol/l (60 ± 10%, P < 0.05) (Fig. 2, A and B). In contrast, treatment of cells with L-NAA reversed the downregulatory effect of ramp flow on pERK1/2, with its level even higher than that of the static control (Fig. 3). These results show the essential and selective role that NO plays in the signaling pathways leading to the activation and downregulation of ERK1/2 by impulse flow and ramp flow, respectively.

View larger version (42K): [in this window] [in a new window]   Fig. 2.   A: effect of the nitric oxide (NO) synthase (NOS) inhibitor NG-amino-L-arginine (L-NAA; 100 and 300 µmol/l) and the NO donor spermine-NONOate (SPR/NO; 0.01 and 0.1 µmol/l) on impulse flow-induced ERK1/2 activation. B: dose-dependent effect of L-NAA on the impulse flow-induced ERK1/2 activation. *P < 0.05 vs. the pERK1/2 level in cells exposed to impulse flow but not L-NAA (control). C: dose-dependent effect of SPR/NO on the impulse flow-induced ERK1/2 activation. *P < 0.05 and **P < 0.01 vs. the pERK1/2 level in cells exposed to impulse flow but not SPR/NO (control). Data in B and C are presented as means ± SE (n = 3).

View larger version (51K): [in this window] [in a new window]   Fig. 3.   Effect of the NOS inhibitor L-NAA (300 µmol/l) on ERK1/2 phosphorylation in ramp flow-treated endothelial cells.

View larger version (30K): [in this window] [in a new window]   Fig. 4.   High concentrations of the exogenous NO donor SPR/NO (10 and 100 µmol/l) stimulate ERK1/2 activation in serum-starved endothelial cells. *P < 0.05 and **P < 0.01 vs. the pERK1/2 level in quiescent cells (control). Data are presented as means ± SE (n = 3).

ROS are involved in TGS-induced ERK1/2 activation. Fluid shear induces sustained production of ROS (13), which induces endothelial expression of intercellular adhesion molecule (ICAM-1) and c-fos genes (7, 18). There is also evidence that shear stress-induced oxidants play a role in the signaling pathways leading to flow-dependent activation of JNK (13). We therefore sought to explore the potential role of ROS in impulse flow-induced ERK1/2 activation. NAC is a well-known antioxidant, which exerts its function by scavenging ROS and maintaining intracellular reduced glutathione concentrations (2). We treated serum-starved ECs with 10 µmol/l NAC for 1 h and then subjected the cells to impulse flow. NAC itself showed no effect on the basal ERK1/2 activity. As shown in Fig. 5, treatment with NAC significantly diminished the stimulatory effect of impulse flow on ERK1/2 activation. More importantly, although NAC or L-NAA (300 µmol/l) alone did not completely block the impulse flow-stimulated ERK1/2 activation (Fig. 2A and Fig. 5), together they almost abolished such an effect (Fig. 5). Therefore, it appears that, besides their independent roles, TGS-induced NO and ROS also act together to mediate ERK1/2 activation potentially through the formation of peroxynitrite (ONOO), as suggested in a previous report (13).

View larger version (45K): [in this window] [in a new window]   Fig. 5.   Effect of the reactive oxygen species scavenger N-acetyl-cysteine (NAC; 10 µmol/l) and the NOS inhibitor L-NAA (300 µmol/l) on the ERK1/2 activation in impulse flow-stimulated endothelial cells. Control, cells only exposed to impulse flow.

Involvement of G_q/11 and G_i3 in TGS-induced ERK1/2 activation. G protein -subunits (G_q/11 and G_i3) have been implicated in mechanochemical signal transduction in HUVECs (14, 15). To test whether G_q/11 and G_i3 are involved in impulse flow-induced ERK1/2 activation, cells were transfected with antisense ODN (1.8 µmol/l) against G_q/11 or G_i3 for 1 h. Forty-eight hours after the ODN transfection, cells were either collected to evaluate the level of target protein or subjected to impulse flow for 10 min. As shown in Fig. 6A, antisense ODN treatment decreased the corresponding G subunit (G_q/11 and G_i3) expression by 83 ± 12% and 69 ± 8%, respectively. More importantly, G_q/11 and G_i3 antisense ODN treatment decreased the impulse flow-induced ERK1/2 activation by 92 ± 4% (P < 0.01) and 63 ± 13% (P < 0.05), respectively (Fig. 6B). The scrambled nonspecific ODN showed no effect on the stimulatory effect of TGS (Fig. 6B). These results clearly demonstrated that G proteins mediate the stimulatory effect of TGS on ERK1/2.

View larger version (61K): [in this window] [in a new window]   Fig. 6.   Effect of oligodeoxynucleotides (ODN) on G protein -subunits Gi3 and Gq/11 expression and fluid shear stress-induced ERK1/2 activation in cultured HUVECs. A: antisense (AS) ODN against Gi3 and Gq/11 specifically decreased the respective G protein -subunit expression. Static, protein levels in serum-starved quiescent cells detected with Western blotting using the anti-Gi3 and anti-Gq/11 antibodies. A and B are from the same filter; thus they serve as controls for each other. B: AS ODN against Gq/11 and Gi3, especially anti-Gq/11 ODN, inhibited the impulse flow-induced activation of ERK1/2 in endothelial cells. A scrambled nonspecific ODN was used for comparison with the AS ODNs. Static and control, levels of pERK1/2 in serum-starved cells and impulse flow-stimulated cells, respectively.

ERK1/2 mediate TGS-induced EC proliferation. We recently found that TGS but not steady shear (ramp flow) induced EC proliferation, as evidenced by a significant increase in the number of cells with positive BrdU incorporation (43). Because TGS but not steady shear also activates ERK1/2 in cells, we examined whether ERK1/2 mediate the mitogenic effect of TGS. Before shear stimulation, the cells were treated with 50 µmol/l PD98059 (a specific MEK antagonist) (1) for 1 h. Treatment with PD98059 blocked the impulse flow-induced ERK1/2 activation in the cells (data not shown). Four separate experiments were carried out to examine the effect of MEK inhibition on the TGS-induced proliferative response in ECs. As shown in Fig. 7, 24 h after exposure of cells to TGS (impulse flow), the cell proliferation index was significantly higher than that of the static unstimulated cells (32.3 ± 5.6 for cells subject to TGS versus 20.9 ± 4.2 BrdU-labeled nuclei per high-power field for the control cells). Pretreatment of cells with PD98059 entirely blocked the TGS-induced increase in the cell proliferation index (21.7 ± 3.1). The results indicate that ERK1/2 activation is necessary for the TGS-associated growth response in ECs.

View larger version (12K): [in this window] [in a new window]   Fig. 7.   Effect of the mitogen-activated protein kinase kinase inhibitor PD98059 on impulse flow-induced endothelial cell proliferation. The cell proliferation index was expressed as the number of bromodeoxyuridine (BrdU)-labeled nuclei per high-power field (HPF; ×40). Data are presented as means ± SE (n = 4). *P < 0.05 vs. the static quiescent cells as well as the PD98059 + impulse flow-stimulated cells.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Exposure of ECs to local fluctuating shear results in increased EC proliferation (43), which in turn is involved in the pathogenesis of several pathological vascular conditions including atherosclerosis (31), reperfusion injury (42), and anastomotic intimal hyperplasia (34). In the present study, we investigated the intracellular signal pathways linked to TGS-induced EC proliferation. Our results indicate that MAPK ERK1/2 and its upstream signaling cascade can be activated by TGS, and its activation is required for TGS-induced EC proliferation (Fig. 8).

View larger version (16K): [in this window] [in a new window]   Fig. 8.   Mechanism of temporal gradients in shear-induced ERK1/2 activation and proliferation in endothelial cells. + and , Positive and negative effects, respectively. O, superoxide; ONOO, peroxynitrite.

ERK1/2 are thought to be directly involved in transmitting signals from growth factor receptors to the nucleus to regulate gene transcription and protein synthesis, leading to proliferation, differentiation, or apoptosis (3, 6, 12, 38). In a previous study (3), we showed that the activation of ERK1/2 is sensitive to a single impulse flow, suggesting a mitogenic effect of TGS on ECs (3). We demonstrated in the current study that impulse flow, as well as pulsatile flow (multiple impulse flow), induced sustained activation of ERK1/2. We also provided evidence that TGS stimulated BrdU incorporation into DNA in ECs, and ERK1/2 activation mediates the proliferative effect of TGS.

The role of NO in regulating cell growth is controversial because both proliferative and antiproliferative effects have been reported (17, 35). In angiogenesis, NO elevation has been shown to positively correlate with neovascularization and tumor growth (22, 28) in adult rodent models. Conversely, in the chorionallantoic membrane of the chick embryo and during the developmental maturation of Drosophila, NO acts as an antiproliferative agent (27, 36). This paradoxical role of NO is also manifested in the expression of several proliferation-related genes because both stimulatory (16) and inhibitory (25) effects have been observed. Fluid shear stress induces NO production in vascular ECs (11, 26), and shear stress associated-NO production is important in regulating the functions of vascular cells and blood cells, such as smooth muscle relaxation and platelet inhibition. On the other hand, NO also acts as a signaling molecule in the mechanotransduction pathways in ECs (27, 39). The kinetics of NO production and signaling pathways involved vary among different flow profiles. For example, TGS (impulse flow) stimulates a transient high-level burst of NO release, whereas steady shear (ramp flow) induces a sustained low rate of NO production in ECs (11). The former is dependent on and the latter is independent of G protein and calcium/calmodulin-associated pathways (11, 26). In a previous study (4), we found that NO mediated both the stimulatory effect of TGS and the inhibitory effect of steady shear on atherogenesis-related genes such as MCP-1 and PDGF-A expression (4). This dual role was further demonstrated by the ability of NO to both activate and inactivate their transcription factors, NF-B and Egr-1 (4). Consistent with previous findings, we showed here that NO generated by both TGS and steady shear mediates their stimulatory and inhibitory effect on ERK1/2 activation, respectively. With the use of an NO donor, we further found that exogenous NO at a low concentration (<1.0 µmol/l) inhibited ERK1/2 activation, whereas a high concentration of NO (>10 µmol/l) stimulated ERK1/2 activation. It is apparent that depending on the level of NO production induced by different flow profiles, NO exerts different effects on ECs.

In addition to NO, fluid shear can induce sustained production of ROS, which mediate the expression of c-fos and ICAM-1 (7, 18). This raises a question about the potential significance of the simultaneous formation of NO and O in the endothelium, which in other contexts is generally viewed as being cytotoxic and apoptotic. Peroxynitrite, a product of NO and O, has been shown to activate p38 (32), JNK (13), and ERK1/2 (24) pathways in various cell types. Particularly, there is evidence that peroxynitrite may act as a signaling molecule in fluid flow-dependent activation of the JNK pathway in cultured bovine ECs (13). However, direct evidence of the involvement of ROS in the shear stress-associated ERK1/2 pathway has not yet been shown. In the present study, NAC, a nonspecific ROS and peroxynitrite scavenger (2), significantly blocked the activation of ERK1/2 induced by TGS, similar to the effect of NOS inhibitor L-NNA. When the cells were cotreated with L-NAA and NAC, interestingly, the effect of TGS on ERK1/2 activity was almost entirely blocked. These results suggest that besides the independent roles of NO and ROS, there is also a synergistic effect of NO and O, possibly through the formation of peroxynitrite, on the activation of ERK1/2 induced by TGS.

Heterotrimetric G proteins transduce signals from activated transmembrane receptors to intracellular effectors. It has been shown that fluid shear stress rapidly activates G protein -subunits G_q/11 and G_i3, but not G_i2, in HUVECs (14). Activation of G proteins has been suggested to serve as the initial signaling step for the shear stress-induced MAPK activation in ECs (23, 41). Activation of G proteins has also been shown to lead to generation of ROS, which then activate MAPKs as well as immediate-early genes in smooth muscle cells (37). With the use of antisense ODN, we demonstrated here that G_q/11 and G_i3 were both involved in TGS-induced ERK1/2 activation, suggesting that both PTX-sensitive (G_i3) and PTX-insensitive (G_q/11) G protein -subunits mediate TGS-induced ERK1/2 activation. One study (41) demonstrated that PTX-insensitive G protein -subunits mediated shear-dependent ERK activation in fetal bovine aortic ECs, whereas another study (23) showed that only G_i2 was involved in fluid shear-induced ERK1/2 activation in adult bovine aortic ECs. This discrepancy may be due to the different flow profiles to which the cells are exposed (impulse vs. step flow), different cell types used (vein vs. artery or human vs. bovine), or the subtle differences in cell culture and serum-starvation conditions. A similar discrepancy has been reported previously regarding the PTX sensitivity of G proteins mediating the production of NO/cGMP by fluid shear stress in bovine aortic ECs and HUVECs (26, 33).

In summary, this study presents the direct evidence that TGS, even a single 3-s impulse flow, is a potent mitogenic stimulus to ECs. ERK1/2 and its upstream signaling cascade (G proteins, NO, and ROS) can be activated by TGS and are involved in TGS-associated EC proliferation. In addition, NO exerts both stimulatory and inhibitory effect on ERK1/2 phosphorylation depending on the level of NO production induced by different flow profiles. The paradoxical dual role of NO may have in vivo significance with respect to the different effects of TGS-stimulated acute and steady shear-mediated basal release of NO on the function and structure of blood vessels.