Reactive Oxygen Species-mediated TRPC6 Protein Activation in Vascular Myocytes, a Mechanism for Vasoconstrictor-regulated Vascular Tone*

Both TRPC6 and reactive oxygen species (ROS) play an important role in regulating vascular function. However, their interplay has not been explored. The present study examined whether activation of TRPC6 in vascular smooth muscle cells (VSMCs) by ROS was a physiological mechanism for regulating vascular tone by vasoconstrictors. In A7r5 cells, arginine vasopressin (AVP) evoked a striking Ca2+ entry response that was significantly attenuated by either knocking down TRPC6 using siRNA or inhibition of NADPH oxidases with apocynin or diphenyleneiodonium. Inhibition of TRPC6 or ROS production also decreased AVP-stimulated membrane currents. In primary cultured aortic VSMCs, catalase and diphenyleneiodonium significantly suppressed AVP- and angiotensin II-induced whole cell currents and Ca2+ entry, respectively. In freshly isolated and endothelium-denuded thoracic aortas, hyperforin (an activator of TRPC6), but not its vehicle, induced dose- and time-dependent constriction in TRPC6 wide type (WT) mice. This response was not observed in TRPC6 knock-out (KO) mice. Consistent with the ex vivo study, hyperforin stimulated a robust Ca2+ entry in the aortic VSMCs from WT mice but not from KO mice. Phenylephrine induced a dose-dependent contraction of WT aortic segments, and this response was inhibited by catalase. Moreover, H2O2 itself evoked Ca2+ influx and inward currents in A7r5 cells, and these responses were significantly attenuated by either inhibition of TRPC6 or blocking vesicle trafficking. H2O2 also induced inward currents in primary VSMCs from WT but not from TRPC6 KO mice. Additionally, H2O2 stimulated a dose-dependent constriction of the aortas from WT mice but not from the vessels of KO mice. Furthermore, TIRFM showed that H2O2 triggered membrane trafficking of TRPC6 in A7r5 cells. These results suggest a new signaling pathway of ROS-TRPC6 in controlling vessel contraction by vasoconstrictors.

Both TRPC6 and reactive oxygen species (ROS) play an important role in regulating vascular function. However, their interplay has not been explored. The present study examined whether activation of TRPC6 in vascular smooth muscle cells (VSMCs) by ROS was a physiological mechanism for regulating vascular tone by vasoconstrictors. In A7r5 cells, arginine vasopressin (AVP) evoked a striking Ca 2؉ entry response that was significantly attenuated by either knocking down TRPC6 using siRNA or inhibition of NADPH oxidases with apocynin or diphenyleneiodonium. Inhibition of TRPC6 or ROS production also decreased AVP-stimulated membrane currents. In primary cultured aortic VSMCs, catalase and diphenyleneiodonium significantly suppressed AVP-and angiotensin IIinduced whole cell currents and Ca 2؉ entry, respectively. In freshly isolated and endothelium-denuded thoracic aortas, hyperforin (an activator of TRPC6), but not its vehicle, induced dose-and time-dependent constriction in TRPC6 wide type (WT) mice. This response was not observed in TRPC6 knock-out (KO) mice. Consistent with the ex vivo study, hyperforin stimulated a robust Ca 2؉ entry in the aortic VSMCs from WT mice but not from KO mice. Phenylephrine induced a dose-dependent contraction of WT aortic segments, and this response was inhibited by catalase. Moreover, H 2 O 2 itself evoked Ca 2؉ influx and inward currents in A7r5 cells, and these responses were significantly attenuated by either inhibition of TRPC6 or blocking vesicle trafficking. H 2 O 2 also induced inward currents in primary VSMCs from WT but not from TRPC6 KO mice. Additionally, H 2 O 2 stimulated a dose-dependent constriction of the aortas from WT mice but not from the vessels of KO mice. Furthermore, TIRFM showed that H 2 O 2 triggered membrane trafficking of TRPC6 in A7r5 cells. These results suggest a new signaling pathway of ROS-TRPC6 in controlling vessel contraction by vasoconstrictors.
Canonical transient receptor potential 6 (TRPC6) is a nonselective cation channel and participates in a diverse array of cellular functions by regulating intracellular Ca 2ϩ signaling (1). In particular, TRPC6 channels are highly expressed in vascular smooth muscle cells (VSMCs) 2 and play a key role in regulating myogenic tone in vascular tissues (2)(3)(4). Multiple mechanisms are involved in TRPC6 channel activation and regulation. These include membrane receptor activation (5), Ca 2ϩ store depletion (6), stretch (7,8), membrane lipids (9), and trafficking (10,11). The distinct activation/regulation mechanisms may be tissue/cell type-specific and thus render TRPC6 a specific function in a particular site. For instance, mechanosensitive TRPC6 residing in glomerular podocytes (7,12) and mesangial cells (13,14) may regulate hydrostatic pressure-driven ultrafiltration in response to changes in glomerular capillary pressure. Most likely, different mechanisms may exist in the same cell and work together in a synergistic way to regulate the cell function more precisely and efficiently (8). We recently demonstrated that TRPC6 also was a redox-sensitive channel that was activated by H 2 O 2 in a TRPC6-expressing cell line (11). However, the physiological relevance of activation of the channel by reactive oxygen species (ROS) is completely unknown.
ROS are produced in G protein-coupled receptor-signaling pathway (15,16), a pathway also linked to TRPC6 channel activation. ROS not only function as an intracellular signaling molecule in a variety of cells but are also associated with many diseases, such as hypertension (15,17). In blood vessels, all types of vascular cells can produce ROS that modulate vasoactive agent-induced endothelial cell and myocyte responses (18). In VSMCs, ROS play an important role in mediating vasoactive hormone-induced proliferation and hypertrophy (15). With respect to vascular contractile function, many studies have demonstrated that H 2 O 2 evoked constriction in a variety of vascular beds (19). However, the molecular mechanism for ROS-induced vasoconstriction is poorly understood.
Because both TRPC6 and ROS play an important role in regulation of vascular function, we used VSMCs and freshly isolated blood vessel segments as a model in this study to investigate the physiological significance of activation of the TRPC6 channel by ROS and the underlying mechanism. The findings from this study for the first time suggest that ROS is a physiological intermediator to mediate vasoconstrictorinduced vessel contraction by activating TRPC6 in VSMCs.

EXPERIMENTAL PROCEDURES
Animals-All procedures were approved by University of North Texas Health Science Center Institutional Animal Care and Use Committee. Trpc6 Ϫ/Ϫ mice were obtained from Dr. Shmuel Muallem (Department of Physiology, University of Texas Southwestern Medical Center). These TRPC6 knock-out (KO) mice were originally generated by Dr. Lutz Birnbaumer (20). The KO mice were bred and raised at University of Texas Southwestern Medical Center and were transferred to University of North Texas Health Science Center 2-3 days prior to experiments. Age-matched wild type (WT) C57BL/6 mice were purchased from Charles River (Wilmington, MA). The KO and WT mice were euthanized with ketamine (95 mg/kg)/xylazine (5 mg/kg), and the thoracic aortas were removed for vessel contraction assay or isolation of VSMCs.
Cell Culture and Transient Transfection-Both human embryonic kidney 293 (HEK293T) and A7r5 cells were purchased from ATCC (Manassas, VA) and were cultured as described previously (11). For HEK293 cells, all plasmids were transiently transfected using GenJet (SignaGen, Gaithersburg, MD) following the protocols provided by the manufacturer. For TRPC6 knockdown studies in A7r5 cells, we transfected the cells with a commercial siRNA against rat TRPC6 or a scrambled sequence (Dharmacon Inc., Lafayette, CO) as described in our previous publication (11). The cells were used for functional studies 24 -48 h after transfection. To identify the positively transfected A7r5 cells in patch clamp experiments, a GFP expression plasmid was co-transfected with siRNAs at a ratio of 1:9 (GFP/siRNA).
Isolation and Culture VSMCs-Smooth muscle cells were dissociated from thoracic aorta using standard enzymatic techniques. The aorta was removed and cleaned of connective tissue under ice-cold Hanks' balanced salt solution. The vessels was then placed in a tube containing 1 ml of PBS with 0.2% Sigma collagenase IV and shaken for 45 min at 37°C. Arteries were chopped into pieces, transferred to another tube containing 1 ml PBS with 0.2% collagenase IV and 4 l of elastase (Worthington Biochemical), and incubated for 60 min at 37°C with shaking. After incubation, gentle agitation with a fire-polished Pasteur pipette was used to dissociate cells from vessel tissues. The cell suspension was centrifuged, resuspended with culture medium (the same as the medium for A7r5 cells), and then transferred to a T25 mm 3 flask or cell culture dishes for growth. Cells were used for experiments after culturing for 2-3 weeks, and only the cells within passage 1-3 were used. VSMCs were verified by positive staining of ␣-smooth muscle actin (data not shown).
Patch Clamp Procedure-Conventional whole cell voltage clamp configuration was employed as described in our previous study (11). Channel currents were measured with a Warner PC-505B amplifier (Warner Instrument Corp., Hamden, CT) and pClamp 9.2 (Axon Instrument, Foster City, CA). The extracellular solution contained the following (in mM): NaCl 130, KCl 2.8, CsCl 10, MgCl 2 2, CaCl 2 0.1, HEPES 10, glucose 10, pH 7.4; and the pipette solution contained the following (in mM): CsCl 135, NaCl 4, MgCl 2 2, EGTA 5, Mg-ATP 2, GTP 1, HEPES 10, pH 7.2. In the experiments utilizing transfected cells, only GFP-labeled cells were targeted for patching. Cell capacitance and series resistance were compensated prior to recording. The whole cell currents were continuously measured at a holding potential of Ϫ60 mV. Currents were filtered at 5 kHz. To exclude the influence of fluid flow on channel activity upon delivery of chemicals, the bathing solution continuously flowed throughout the experiments. The flow rate was adjusted by gravity and controlled by a multiple channel perfusion system (ValveLink TM 8, Automate Scientific, Inc.). The whole cell currents were normalized to the cell capacitance and expressed as current density (pA/picofarads). Clampfit 9.2 software (Axon Instrument, Foster City, CA) was used to analyze channel currents.
Fluorescence Measurement of [Ca 2ϩ ] i -Intracellular Ca 2ϩ concentration ([Ca 2ϩ ] i ) was assessed by measuring fura-2 fluorescence using dual excitation wavelengths as described previously (11). A7r5 cells, grown on a coverslip (22 ϫ 22 mm), were loaded with fura-2 by incubation for ϳ50 min at room temperature in the dark in physiological saline solution containing 2 M acetoxymethyl ester of fura-2 (fura-2/AM) and 0.018 g/dl pluronic F-127 (Molecular Probes, Eugene, OR). The coverslip was then placed in a perfusion chamber (Warner, Model RC-2OH) mounted on the stage of a Nikon Diaphot inverted microscope. Fura-2 fluorescence was monitored by ratiometry (excitation at 340 and 380 nm and emission at 510 nm) using NIS Elements AR TM software (Nikon Instruments Inc., Melville, NY) at room temperature. [Ca 2ϩ ] i was calculated using the software following the manufacturer's instructions. Calibrations were performed in vivo at the end of each experiment, and conditions of high [Ca 2ϩ ] i were achieved by addition of 5 M ionomycin, whereas conditions of low [Ca 2ϩ ] i were obtained by addition of 5 mM EGTA.
Measurement of Intracellular H 2 O 2 Level-Intracellular H 2 O 2 level was estimated by measuring 2Ј,7Ј-dichlorodihydrofluorescein (DCF) fluorescence as described in our previous study (11). In brief, A7r5 cells grown in 60-mm plates with various treatments as indicated in Fig. 3C were washed three times with cold Hanks' balanced salt solution and loaded with 15 M DCF diacetate (Molecular Probes, Eugene, OR) for 10 min at 37°C in the dark. DCF fluorescence was measured by 2030 Multilabel Reader (Victor TM X3, PerkinElmer Life Sciences) at excitation and emission wavelengths of 480 and 520 nm, respectively.

TRPC6 Mediates ROS-regulated Vascular Tone
Reverse Transcription-PCR (RT-PCR)-Total RNA from thoracic aortas was extracted using the High Pure RNA tissue kit (Roche Applied Science). Reverse transcription was carried out with oligo(dT) and random hexamers (IDT, Coralville, IA) as primers, using Moloney murine leukemia virus reverse transcriptase (Promega, Madison, WI). Products were amplified using TRPC6-specific primers and ␤-actin primers as controls. The primer sequences for TRPC6 were as follows: forward, 5Ј-CAG ATC ATC TCT GAA GGT CTT TAT GC-3Ј, and reverse, 5Ј-TCT GAA TGC TTC ATT CTG TTT TGC GC-3Ј, which gave a predicted product size of ϳ200 bp. The primer sequences for ␤-actin were as follows: forward, 5Ј-TCT TTT CCA GCC TTC CTT CTT G-3Ј, and reverse, 5Ј-GCA CTG TGT TGG CAT AGA GGT C-3Ј, which gave a predicted product size of ϳ100 bp. PCRs were carried out using the following conditions: initial denaturation for 2 min at 94°C and 45 cycles of 94°C for 15 s, 50°C for 30 s, and 72°C for 30 s, followed by a final extension at 72°C for 7 min.
Western Blot-Western Blot was described in our previous publications (11,13,14,21). In brief, A7r5 cell and VSMC lysates (ϳ50 g/well) were fractionated by 10% SDS-PAGE, transferred to PVDF membranes, and probed with primary transient receptor potential channel or ␣-actin antibodies. Bound antibodies were visualized with SuperSignal West Femto or Pico Luminol/Enhancer Solution (Pierce).
Vessel Contraction Assay-The isolated thoracic aorta was placed in a vessel chamber. The two ends of the vessel were cannulated with glass micropipettes and secured. The endothelium was removed by the passage of an air bubble through the lumen. The vessels were pressurized to 100 mm Hg with a modified Krebs solution containing the following (in mM): 122 NaCl, 4.73 KCl, 15.5 NaHCO 3 , 1.19 KH 2 PO 4 , 1.19 MgCl 2 , 1.8 CaCl 2 , 11.5 glucose, and 0.026 EDTA, pH 7.2. The warmed solution (37°C) was continuously superfused following aeration with a gas mixture of 21% O 2 , 5% CO 2 , balance N 2 . The preparations were equilibrated 60 min prior to experiments, and the inner diameter was monitored using video microscopy and edge-detection software. All vessels were pretested with 80 mM KCl solution to ensure vessel vitality, and all agent-induced contractile responses were normalized to the high KCl-induced constriction. At the end of experiments, all vessels were treated with acetylcholine (100 M) to verify removal of endothelium. The vessels were considered endothelium-free if no relaxation was observed after acetylcholine.
TIRFM-TIRFM generates an evanescent field that declines exponentially with increasing distance from the interface between the coverslip and the plasma membrane, illuminating only a thin section (ϳ100 nm) of the cell in contact with the coverslip, including the plasma membrane. An Olympus IX71 fluorescence microscope (Olympus America Inc.) was used to measure single cell evanescent field fluorescence (EFF) intensity. Cells transfected with EGFP or rTRPC6-EGFP expression plasmids were excited with a 488-nm line of an argon air-cooled laser IMA101010-B0S (Melles Griot) with output power of up to 65 milliwatts. The excitation light was focused onto a single-mode optical fiber with a 5-axis fiber coupler and brought through the TIRF adapter to the rear illumination port of the Olympus IX71 fluorescence microscope. The laser light was directed through the dichroic mirror system and then passed through a high numeric aperture oil emersion objective (N.A. 1.45, ϫ60, Olympus, Melville, NY). The excitation light entered the glass coverslip 1 (ϳ120 m thick) and then was totally internally reflected at the glass-water interface on the sample side. The excitation beam did not enter the sample but formed a shallow evanescent field penetrating the sample only up to ϳ100 nm. In this way excitation was limited to the cell membrane. Fluorescence emitted from tagged proteins passed through an HQ515-30-m filter for GFP before being collected by an Avalanche photodiode detector. Fluorescence images were collected with a back-illuminated electron-multiplying charge-coupled device camera (Hamamatsu Photonics, Japan). SimplePCI software was used to control the protocol of acquisition and to perform data analysis.
Expression Plasmids and siRNA Oligonucleotides-We generated the TRPC6-EGFP construct. We also designed TRPC6 siRNA and its control scramble oligonucleotides, which were synthesized by Integrated DNA Technology. The detailed information on the TRPC6-EGFP construct and TRPC6 siRNA was described previously (11).
Chemicals and Others-Methanol was purchased from VWR (Radnor, PA). Brefeldin A was purchased from EMD Chemicals, Inc. (Gibbstown, NJ). All other chemicals and primary antibodies were purchased from Sigma.
Statistical Analysis-Data were reported as means Ϯ S.E. One-way analysis of variance and one-way repeated measures analysis of variance plus Student-Newman-Keuls test were used to analyze the differences among multiple groups and repeated measurements for the same treatment, respectively. Student t test was used to analyze the difference between two groups. p Ͻ 0.05 was considered statistically significant. Statistical analysis was performed using SigmaStat (Jandel Scientific, San Rafael, CA).

TRPC6-mediated Agonist-stimulated Ca 2ϩ Entry and Membrane Currents in A7r5
Cells-To assess the function of TRPC6 in VSMCs, we carried out Ca 2ϩ imaging and electrophysiological experiments. As shown in Fig. 1, A-C, application of 100 nmol/liter AVP dramatically increased Ca 2ϩ entry using a classic Ca 2ϩ add-back protocol (11). The Ca 2ϩ entry response was nearly abolished in the cells treated with siRNA-TRPC6 but not a scrambled control sequence (Fig. 1, B and C). However, knockdown of TRPC6 did not affect the AVP-induced initial Ca 2ϩ transient, which was primarily attributed to Ca 2ϩ release from the sarcoplasmic reticulum (Fig. 1B). Although Ca 2ϩ influx from the extracellular compartment may also participate in the initial response, these results suggest that the TRPC6 channel may not be involved in AVP-induced Ca 2ϩ release and the initial stage of Ca 2ϩ entry (see "Discussion"). Fig. 1D showed ϳ80% reduction of TRPC6 protein by siRNA-TRPC6.
Patch clamp experiments showed that AVP at 100 nmol/liter evoked robust inward currents within ϳ1 min with peaking at ϳ3 min (Fig. 1, E and F). To determine whether the currents TRPC6 Mediates ROS-regulated Vascular Tone SEPTEMBER 9, 2011 • VOLUME 286 • NUMBER 36 were carried by TRPC6, we measured the AVP response in the presence of an antibody against TRPC6, which recognizes amino acid residues 24 -38 of TRPC6. Dyachenko et al. (22) and Saleh et al. (4) have used a TRPC6 antibody bound to the same epitope to successfully block the channel (4,22). As shown in Fig. 1, E and F, dialysis of the antibody at a 1:200 dilution for ϳ10 min nearly abolished the AVP-induced currents. However, the same concentration of rabbit immunoglobin without immune reactivity only caused a slight and insignificant reduction in AVP currents (Fig. 1, E and F).
Activation of TRPC6-Induced Aortic Contraction-To assess physiological relevance of TRPC6 activation in VSMCs, we carried out contraction assays in the endothelium-denuded aortas. Hyperforin is a recently identified specific TRPC6 activator (23,24). Superfusion of hyperforin over isolated mouse aortas for 30 min resulted in a dose-dependent constriction, and a significant response occurred at 5 mol/liter with peaking at 10 mol/liter ( Fig. 2A). Fig. 2B showed the responses normalized to the constriction induced by 80 mM KCl. The hyperforin-induced response was also time coursedependent. At a dose of 10 mol/liter, which gave the maximum response (Fig. 2, A and B), a significant contraction was not observed until 20 min after treatment and a greater response occurred at 30 min (Fig. 2C). The hyperforin response was mediated by TRPC6 because the same concentration of vehicle (methanol) did not affect the vessel tone (Fig. 2, B and C), and hyperforin at 10 mol/liter with 30 min of treatment was not able to induce contraction of the vessels from TRPC6 KO mice (Fig. 2D). Consistent with the ex vivo study, pretreatment with 10 mol/liter hyperforin for 30 min, but not methanol, significantly increased Ca 2ϩ entry in the cultured VSMCs isolated from TRPC6 wild type mouse aortas. However, the same concentration of hyperforin was not able to evoke this Ca 2ϩ response in TRPC6 KO VSMCs (Fig. 2E). RT-PCR assays verified loss of TRPC6 gene in KO mice (Fig. 2F).
ROS Contributed to Agonist-evoked Ca 2ϩ Entry and Membrane Currents in VSMCs-To determine whether ROS are intermediators in a physiological pathway of TRPC6 activation in VSMCs, we evaluated the Ca 2ϩ entry and membrane currents in response to AVP in A7r5 cells with and without suppression of ROS production. As shown in Fig. 3, A and B, AVP induced a robust Ca 2ϩ influx that was significantly inhibited by an NADPH oxidase inhibitor, apocynin (1 mmol/liter), or a flavin-containing enzyme (including NADPH oxidases) inhibitor, diphenyleneiodonium (DPI, 10 mol/liter). Ca 2ϩ release was not affected by these inhibitors (Fig. 3A). NADPH oxidases are the major source of ROS in vascular myocytes (17,25). DCF florescence measurement verified production of ROS in response to AVP stimulation in A7r5 cells, and both apocynin and DPI treatments prevented the AVP response (Fig. 3C). Consistent with an inhibition on Ca 2ϩ entry response, suppression of ROS production by treating A7r5 cells with DPI (10 mol/liter) also significantly attenuated AVP-induced membrane currents (Fig. 3D). . Actin served as a loading control. Panel a, original TRPC6 bands; panel b, normalized TRPC6 expressions averaged from five independent repeats. The expression level of TRPC6 in untreated was estimated by a ratio of optical density of TRPC6 band to that of actin and was taken as 100%. The TRPC6 expression levels in scrambled and siRNA-T6 were estimated by normalization of the ratios in each group to that in untreated. *, p Ͻ 0.05, compared with both untreated and scrambled groups. E, whole cell patch clamp, showing representative inward currents in response to 100 nmol/liter AVP without (UT) or with pipette inclusion of an anti-TRPC6 antibody (T6-Ab) or a rabbit IgG (IgG) without immune reactivity at a concentration of 2.5 g/ml. Currents were normalized to cell capacitance. The holding potential was Ϫ60 mV. F, summary data from the experiments described in E. n represents cell numbers analyzed. *, p Ͻ 0.05, compared with both untreated and IgG.

TRPC6 Mediates ROS-regulated Vascular Tone
The whole cell currents in response to 100 nmol/liter AVP were measured in primary cultured mouse aortic VSMCs. As shown in Fig. 4A, an inward current developed within 1 min after application of AVP, and a peak response was observed within 3-4 min. The AVP currents were markedly attenuated by preincubation of the cells with 250 units/ml PEG-catalase (Fig. 4, A and B). This antioxidant effect was corroborated by a significant inhibition of PEG-catalase but not PEG on the AVPinduced Ca 2ϩ entry in VSMCs (Fig. 4C).
Saleh et al. (4) reported that a low dose of Ang II activates TRPC6 channel in rabbit mesenteric arterial VSMCs. It is known that ROS are downstream messengers in an Ang IIsignaling pathway (16,26). In this study, 2 nmol/liter Ang II evoked a robust inward current in primary VSMCs (Fig. 4D). The Ang II response was much greater and reached peak faster after an ϳ1-min delay as compared with AVP. Decreasing ROS production by pretreating the cells with 10 mol/liter DPI significantly reduced the Ang II currents (Fig. 4, D and E). Consistent with the inhibitory effect on the membrane currents, DPI (10 mol/liter) also significantly attenuated Ang II-evoked Ca 2ϩ entry (Fig. 4F).

ROS-mediated Vasoconstrictor-induced Vessel Contraction-
ROS involvement in the vasoconstrictor-induced response was further examined using ex vivo settings. In this series of experiments, we measured the phenylephrine (PE) response twice in freshly isolated aortas, and the second measurement was undertaken with and without PEG-catalase. As shown in Fig. 5A, PE induced a dose-dependent constriction in response to both challenges. However, after washing out PE and pretreating the vessels with 250 units/ml PEG-catalase for 30 min, PE responses were attenuated from a dose as low as 0.1 mol/liter, and a significant decrease was observed at concentrations of 10 and 100 mol/liter. The lower PE response in the presence of catalase was not due to PE desensitization because the second PE challenge in the absence of catalase was still able to evoke a comparable constriction (Fig. 5B). Comparison of the maximum responses between the first and second PE treatment (100 mol/liter) showed a significant reduction in the presence of catalase (Fig. 5C).
ROS Activated TRPC6 in A7r5 and Primary VSMCs-The in vitro and ex vivo experiments above showed that both TRPC6 and ROS contributed to agonist-induced VSMC responses and vessel contraction. We previously reported that TRPC6 is a redox-sensitive channel (11). Thus, we reasoned that ROS might be an upstream molecule from TRPC6 in the G protein-coupled receptor-signaling pathway in VSMCs. If so, then direct application of ROS should be able to activate TRPC6 and reproduce the agonist responses. This hypothesis was tested in both A7r5 cells and primary VSMCs. As expected, Ca 2ϩ entry was robust in the presence of 100 mol/liter H 2 O 2 in A7r5 cells. This Ca 2ϩ response was significantly suppressed by knockdown of TRPC6 (Fig. 6, A and  B). H 2 O 2 evoked inward currents that have a similar profile to that induced by AVP (Fig. 6C). Again, the current response was significantly suppressed in cells treated with siRNA against TRPC6, but not with scrambled, and in the cells treated with TRPC6 antibody, but not control immunoglobulin (Fig. 6, C and D). H 2 O 2 -dependent current was also detected in the primary VSMCs. Like Ang II, H 2 O 2 produced a fast developed inward current after an ϳ1-min delay in WT cells (Fig. 6E). However, the H 2 O 2 -dependent currents were significantly smaller in the TRPC6 KO VSMCs (Fig. 6, E and  F). Taken together, the Ca 2ϩ imaging and electrophysiological data strongly suggest that ROS-induced cellular response was primarily mediated by TRPC6 in VSMCs.
TRPC6-mediated ROS-induced Vessel Contraction-Because H 2 O 2 activated TRPC6 (Fig. 6) and the activation of TRPC6 triggered aortic contraction (Fig. 2), we postulated that superfusion of ROS to an endothelium-denuded vessel should induce a contractile response. Fig. 7 shows the effect of H 2 O 2 on freshly isolated aortic segments from TRPC6 WT and KO mice. In WT vessels, H 2 O 2 evoked a linear and dosedependent constriction in the range of 0.5 mol/liter, 5 mmol/liter (Fig. 7A). However, the responses were not observed in TRPC6-deficient vessels (Fig. 7A). Normalization of the H 2 O 2 responses to the constriction induced by 80 mM KCl exhibited the same profile of response (Fig. 6B). There was no difference in 80 mM KCl-evoked vessel contraction between WT and KO mice (Fig. 7C). These ex vivo studies suggest that TRPC6 in VSMCs is required for H 2 O 2induced vessel constriction.
ROS Triggered Membrane Trafficking of TRPC6 Protein-To determine whether a trafficking mechanism was involved in ROS-induced TRPC6 activation, we conducted TIRFM in A7r5 cells. TIRFM generates an evanescent field that selectively illuminates fluorophores within 100 nm of the plasma membrane-coverslip interface. In A7r5 cell transfected with TRPC6 whose C terminus was tagged with EGFP (TRPC6-EGFP), application of 100 mol/liter H 2 O 2 induced a modest increase in EFF within 1 min, and the response peaked at ϳ4 min. However, H 2 O 2 failed to increase EFF in the cell transfected with EGFP alone (Fig. 8, A and B). Results summarized from four independent experiments revealed a consistently significant increase in TRPC6-specific EFF 3 min after appli-

TRPC6 Mediates ROS-regulated Vascular Tone
cation of H 2 O 2 (Fig. 8C). Live movies showing a real time change in EFF of TRPC6 and EGFP were presented in supplemental Fig. S1.
We have shown that H 2 O 2 stimulated Ca 2ϩ entry, which was mediated by TRPC6 in A7r5 cells (Fig. 6, A and B). If recruiting TRPC6 to the cell surface is a mechanism for ROS-  induced TRPC6 activation, blocking the protein migration should inhibit the Ca 2ϩ response. Fig. 8D showed that brefeldin A (5 mol/liter), a blocker of vesicle translocation from the endoplasmic reticulum to the Golgi apparatus, significantly attenuated H 2 O 2 (100 mol/liter)-evoked Ca 2ϩ entry in A7r5 cells.

DISCUSSION
Multiple mechanisms are involved in ROS-induced vessel contraction. In this study we proposed TRPC6 being a novel target of ROS in a physiological regulation of vascular tone. This conclusion is based on several major findings from our in vitro and ex vivo assays as shown. 1) AVP-induced Ca 2ϩ   stimulating Ca 2ϩ entry and inward membrane currents, and importantly, the ROS-dependent responses were significantly inhibited by either knockdown of TRPC6 or blocking TRPC6 channel with a specific antibody (Figs. 2 and 7, A-D).
3) In primary aortic VSMCs, AVP and Ang II evoked robust inward currents and Ca 2ϩ influx, which were significantly inhibited by catalase and DPI, respectively (Fig. 4). Application of H 2 O 2 also caused a similar current response in TRPC6 WT VSMCs but not in TRPC6 KO cells (Fig. 6, E and  F). 4) Ex vivo vessel contraction assays showed that catalase significantly suppressed PE-induced aortic contraction (Fig.  5), and consistent with this, H 2 O 2 itself or selective and direct activation of TRPC6 channels by hyperforin reproduced PE responses in aortas from WT but not TRPC6 KO mice (Figs. 3 and 7).
It is generally accepted that diverse pathways activated by ROS are convergent to increase [Ca 2ϩ ] i , which triggers smooth muscle contraction (19). For instance, ROS can activate voltage-operated Ca 2ϩ channels (27), stimulate Ca 2ϩ release from the internal stores (28), and inhibit Ca 2ϩ -AT-Pase on the sarcoplasmic reticulum (29). TRPC6 is a Ca 2ϩpermeable cation channel (1). Our findings are fully compatible with the Ca 2ϩ -dependent mechanism. Suppression of TRPC6 attenuated and genetic removal of TRPC6 abolished ROS-induced intracellular Ca 2ϩ response, membrane currents, and vessel constriction, suggesting a necessity of TRPC6 for ROS effects in VSMCs. The TRPC6-mediated increase in [Ca 2ϩ ] i involves multiple mechanisms. These include the following: 1) allowing Na ϩ to enter the cells, which either depolarizes the membrane and opens voltageoperated Ca 2ϩ channels (30,31) or activates a reverse mode of Na ϩ -Ca 2ϩ exchangers (32); 2) directly conducting Ca 2ϩ influx. Which TRPC6 mechanism underlies the ROS-associated Ca 2ϩ increase observed in the current study is unknown. However, the previous findings that voltage-operated Ca 2ϩ channels contributed to H 2 O 2 -induced Ca 2ϩ response (27) suggest that membrane depolarization could be involved. It could explain why there was no difference in vessel contraction by high KCl between WT and TRPC6deficient vessels (Fig. 7C).
We further provided evidence that ROS activate TRPC6 in VSMCs by stimulating the translocation of channel vesicles to the plasma membrane. This is supported by correlation of the time course of the vesicle trafficking with the development of membrane current and [Ca 2ϩ ] i in response to H 2 O 2 . Membrane trafficking of TRPC6 occurs after activation of G q protein-coupled receptors (10). Because ROS are also generated in this signaling pathway (16), we reason that inserting the TRPC6 channel into the plasma membrane by ROS may be a general mechanism in the G q -coupled receptor-signaling pathway in TRPC6-enriched cells, such as podocytes (12). How ROS stimulate migration of TRPC6 to the cell surface remains unknown. Several possibilities exist, the first of which is that ROS could directly oxidize TRPC6 protein or membrane proteins or lipids that results in movement and binding of TRPC6 to the plasma membrane. Second, ROS could indirectly alter the phosphorylation-dephosphorylation state of the TRPC6 protein or membrane components that promote physical interactions between TRPC6 and the plasma membrane. Oxidative activations of protein-tyrosine kinase Src (33) and oxidative inactivation of protein-tyrosine phosphatases (34) have been described as a downstream mechanism in ROS-dependent cellular responses, particularly in VSMCs (16). A recent study demonstrated that TRPC6 was activated by tyrosine phosphorylation (35). However, whether stimulation of trafficking is the