Functional TRPM7 Channels Accumulate at the Plasma Membrane in Response to Fluid Flow
http://www.100md.com
《循环研究杂志》
the Howard Hughes Medical Institute, Department of Cardiology, Children’s Hospital Boston, Harvard Medical School, Mass.
Abstract
Many cells are constantly exposed to fluid mechanical forces generated by flowing blood, and wall shear stresses modulate aspects of their structure and function. However, the mechanisms for mechanotransduction of flow are not well understood. Here we report that TRPM7, which is both an ion channel and a functional kinase, is translocated within cells in response to laminar flow. After exposure of cells to physiological values of laminar fluid flow, the number of TRPM7 molecules localized at or near the plasma membrane increased up to 2-fold, in less than 100 seconds. This increase in membrane-localized GFP-TRPM7, as seen by total internal reflection fluorescence microscopy, closely correlated with increases in TRPM7 current. Both endogenous and heterologously expressed TRPM7 was found in tubulovesicular structures that were translocated to the region of the plasma membrane on induction of shear stress. In vascular smooth muscle cells, but not in several types of endothelial cells, fluid flow increased endogenous native TRPM7 current amplitude. We hypothesize that TRPM7 plays a role in pathological response to vessel wall injury.
Key Words: TRP ion channels TRPM7 shear force total internal reflection fluorescence (TIRF) microscopy vascular smooth muscle cells
Introduction
TRP ion channels are often found in sensory systems, but their wide distribution suggests they are cellular sensors in a broad sense.1 TRPM6 and TRPM7 are unique channels that possess both ion channel and protein kinase activities. The C-terminal kinase domain bears little sequence identity with other kinases but is enzymatically active and structurally homologous to protein kinase A.2,3 The physiological mechanism for activation of the widely expressed TRPM7 is unknown. TRPM7 channels are inhibited by phospholipase C (PLC)-catalyzed phosphatidylinositol 4,5-bisphosphate (PIP2)hydrolysis4 and reportedly regulated through protein kinase A by receptors coupled to adenylyl cyclase.5 TRPM7 may be involved in anoxia-induced cell death in brain6 and has been reported to be required for cell viability in TRPM7-null avian DT40 cells.7
Cells sense shear stresseCinduced mechanical stimulation and convert it into a biochemical response that impacts normal and abnormal tissue development, including growth, differentiation, migration, gene expression, protein synthesis, and apoptosis.8 The response of cells to fluid flow depends on the cell type and the magnitude and characteristics of the shear stress applied under physiological or pathological conditions. Endothelial cells are constantly subjected to blood flow that regulates physiological blood vessel responses as well as pathological arterial wall responses. In large arteries, the endothelium is exposed to shear stress values in the range 10 to 40 dyne/cm2.9 The vascular smooth muscle cells are protected from shear stress by endothelial cell lining under physiological conditions but become exposed to shear stress after endothelial injury.
Ion channels can respond to shear stress.10,11 A few ion channels sensitive to shear stress have been identified in endothelial cells and partially characterized.9 Among these are K+ channels,10 Na+ channels,12 and CleC channels.13 Most of these channels respond on a rather fast-time scale (minutes) to mechanical stimulation. However, the role of these channels in transforming a mechanical stimulus into a biochemical response is poorly understood. Here we report that TRPM7 responds within 10 of seconds to physiologically relevant fluid flow rates by increasing its presence at the plasma membrane. We show the effect of fluid shear stress on TRPM7 by 2 independent methods: total internal reflection fluorescence (TIRF) imaging and electrophysiological recordings of GFP-TRPM7 in HEK-293 cells. We also show that TRPM7 is highly expressed endogenously in vascular smooth muscle cells that respond to shear stress under pathological conditions. The endogenous TRPM7 and the expressed GFP-TRPM7 share the same highly dynamic subcellular distribution of "vesicles on tracks." Moreover, the endogenous TRPM7-like current increases significantly in response to fluid flow. Our results show that intracellular TRPM7 channels are found in tubulovesicular structures that are delivered to the plasma membrane in response to fluid flow.
Materials and Methods
cDNA Constructs
Green fluorescent protein (GFP) was amplified by PCR and inserted in the NotI site, between the FLAG tag and the N terminus of mTRPM7. FLAG-mTRPM7 in pcDNA4/TO was provided by Dr Andrew Scharenberg (University of Washington). GFP-TRPC5 was obtained by fusing GFP to the C terminus of TRPC5 into the pCI vector. GFP-TRPV5 was cloned in the eGFP/C1 vector by fusing TRPV5 to the C terminus of GFP and was provided by Dr Rene Bindels (University of Nijmegen). pEYFP-Mem, pEYFP-ER cDNA constructs were purchased from Clontech.
Cell Culture
HEK293T cells were cultured in medium containing DMEM/Ham’s F12, 10% FBS, and GHT supplement (Gibco). Cells were cultured on glass coverslips for the imaging experiments and on plastic Petri dishes for the electrophysiology experiments. In some imaging experiments, the glass coverslips were coated with Poly-L-Lysine (Sigma). A7R5 cells were cultured in DMEM supplemented with 10% FBS. Cells were tested periodically for a smooth muscle phenotype by assessing spontaneous calcium spikes mediated by L-type calcium channels14 using Fluo-4 imaging. Cells were split before becoming confluent (70% to 80% density) to a density >20% and the media changed every 2 to 3 days.
DNA Transfection
HEK293T cells were transiently transfected using Lipofectamine 2000 (Invitrogen). Cells were incubated with the transfection agent for 2 to 3 hours and then cultured on glass or plastic 24 to 48 hours before the experiment. A7R5 cells were transiently transfected with Superfect (Qiagen). Cells were plated at least 24 hours before the transfection and used at 50% to 70% density. Thirty-five-millimeter dishes were transfected with 20 e蘈 of Superfect and 2 e of DNA mixed with 100 e蘈 of OPTI-MEM and added to 500 e蘈 of medium. Cells were incubated with the transfection reagent for 2 to 4 hours and then washed with normal medium.
Imaging and Image Analysis
All live imaging experiments were performed at &25°C with cells grown on glass coverslips and placed in a custom chamber with standard extracellular buffer. Extracellular buffer contained (in mmol/L): 135 NaCl, 5 KCl, 1.5 CaCl2, 1.5 MgCl2, 20 HEPES, pH 7.4. Scanning confocal microscopy was used to image the distribution of GFP-TRPM7 in different cell types. Cells expressing GFP-TRPM7 or immunostained with anti-TRPM7 primary antibody against the C terminus of TRPM7 and Alexa 488 secondary antibody were excited by 488-nm laser illumination; the emission light was bandpass filtered from 512 to 527 nm (Olympus Fluoview 300). For TIRF imaging, we used a custom-built, objective-based evanescent field microscope.15 Cells grown on glass coverslips were placed in a custom chamber with standard extracellular buffer and imaged at &25°C. All image acquisition and some of the analysis was performed using MetaMorph (Universal Imaging, Westchester, Pa). Time series of 100 to 400 images at 1- to 4-s intervals were recorded. Fluorescence intensity in a region of interest comprising at least 50% of the visible "footprint" of the cell was measured as a function of time using MetaMorph. The data were then analyzed using Deltagraph (Red Rock Software, Salt Lake City, Utah) as described in Results.
Electrophysiological Recordings and Data Analysis
GFP-TRPM7eCexpressing HEK-293T cells were detected by GFP fluorescence. Membrane currents were digitized at 20 kHz and filtered at 1 kHz. Voltage stimuli were delivered at 5-s intervals, with voltage ramps from eC100 to +100 mV (holding potential, 0 mV). The internal pipette solution contained (in mmol/L); 120 Cs-methanesulfonate, 8 NaCl, 10 EGTA, 4.1 CaCl2, 10 HEPES; pH adjusted to 7.2 with CsOH. For perforated patch recording, 150 e/mL Amphotericin B was included in the pipette solution. Amphotericin B was dissolved in DMSO at 1.5 mg/mL and diluted 1:100 into the pipette solution. Recordings were initiated when series resistance reached values below 30 M. The extracellular solution was designed to minimize chloride currents, and consisted of (in mmol/L) 140 Na-methanesulfonate, 5 Cs-methanesulfonate, 2.8 KCl, 1 CaCl2, 10 HEPES, 10 mmol/L glucose; pH adjusted to 7.4 with NaOH.
Because initial current amplitudes and rates of current run-up in the outward direction varied from cell to cell, a slope-based method was used to quantify shear force-induced changes in ITRPM7. Average current amplitude at 100 mV was plotted as a function of time, and the slope of the data points was taken before and after application of shear force. The baseline rate of current increase (the slope before shear force was applied; S1) was determined with at least 6 data points (over a 30-s period). The current increase caused by flow was determined by taking the slope of the data points after an increase in slope (S2), until a plateau was reached. An increase in slope was only counted if observed within 30 s of flow application. The reported increase in current caused by flow was calculated as (S2eCS1)xt, where t was the length of time over which the increase in slope (S2) occurred.
Flow Setup and Calculation
A syringe pump (Harvard Apparatus, Harvard, Mass) connected through thin tubing to a 250-e ID plastic pipette provided fluid flow and flow values set between 0.123 and 1.23 mL/min. The pipette was positioned &150 e from the cell.
The magnitude of the shear stress generated by the fluid flow was calculated using the formula:
where =1025 kg/m3 is the density of water and V is the fluid velocity, calculated by V=Q/A, where A=d2/4=0.785x10eC8m2 for a pipette diameter of d=250 e; Q is the flow generated by the syringe pump measured in m3/s.
Rx, the Reynolds number, was Rx=Vxx/, where =1.139x10eC6 m2/s is the kinematic viscosity of the water, x=150 e is the distance between pipette and cell. Rx varied from 5.5 to 27.5, depending on flow rate.
To determine whether flow was laminar, we also used the Reynolds number defined as Rh=Vh/v, where h, the fluid level in the dish (3 mm), was kept constant; and V are the kinematic viscosity of the water and the fluid velocity, respectively.
For the highest value of the flow used in our experiments the Reynolds numbers were Rx=27.5 and Rh=627. Laminar flow occurs for Rx<500 000 and Rh<3900.16
Results
GFP-TRPM7 Forms a Functional Channel and Reveals a Dynamic Tubulovesicular Subcellular Distribution
To follow the intracellular distribution of TRPM7, GFP was fused to the N terminus of the channel via a (GA)2 linker (Figure 1A). We first tested the functionality of the GFP-tagged channel by transfecting GFP-TRPM7 into HEK-293 cells and measuring the TRPM7 current in response to a voltage ramp (Figure 1B). GFP-TRPM7 retained all the characteristics of native and recombinant TRPM7 current, including a gradual increase in outward rectification or "run-up"17,18 (Figure 1C).
The intracellular distribution of GFP-TRPM7 was initially studied using confocal microscopy. GFP-TRPM7 in HEK-293 cells is expressed in a reticular pattern throughout the cytoplasm (Figure 1D). Although the cells have functional TRPM7 current, the plasma membrane localized GFP-TRPM7 is relatively less abundant compared with its intracellular distribution and thus not apparent in these images.
Because of the limited resolution of the confocal images, we used evanescent field objective-based (or TIRF) microscopy to visualize GFP-TRPM7 in HEK-293 cells. The evanescent field selectively illuminates fluorophores within &250 nm of the plasma membrane-glass cover-slip interface.15,19 Evanescent light decreases exponentially in the axial direction, such that fluorescence emission is brighter as the fluorophore approaches the plane of the interface. Although TIRF axial resolution is superior to confocal microscopy, it is insufficient by itself to determine whether a protein is truly in the &4-nm plasma membrane.
TIRF imaging of HEK-293 cells expressing GFP-TRPM7 revealed the GFP-labeled channels at or near the plasma membrane (Figure 1E). The proteins appear punctate, which we assume to be vesicular structures, and in tubular structures with diameters <100 nm. Most cells contain a mix of the 2 types of structures (first cell in Figure 1E), with some cells containing predominantly tubular structures (second cell in Figure 1E) or predominantly vesicular structures (third cell in Figure 1E). This distribution pattern is markedly different from TIRF images of TRPC5-GFP15 but similar to that of GFP-TRPM6 and GFP-TRPV5 channels (data not shown).
We recorded series of 100 images at 1- or 2-s intervals, under steady-state conditions at 22°C. The time series show that both the tubular and the vesicular structures are highly mobile (Movie I in the online data supplement available at http://circres.ahajournals.org). Most strikingly, linear tracks or trajectories form in seconds, and vesicles begin to travel on them; then just as quickly, they disappear (Figure 1F). The same type of movement is not observed by stepping the axial (z) position of the objective in fixed cells (data not shown), suggesting that the observed motility is attributable to movement in the x-y plane and not to movement in and out of the z-plane.
Increased GFP-TRPM7 at the Plasma Membrane in Response to Shear Force
To determine the effects of shear stress on the distribution and functional current of GFP-TRPM7 channels, individual cells were exposed to flow as shown in Figure 2A (see Materials and Methods). When sequential TIRF images were recorded before and after the application of flow, we observed a gradual increase in the fluorescence at and near the plasma membrane (Figure 2C and supplemental Movie II). Higher resolution TIRF imaging (Figure 2D) shows that the global increase in fluorescence is attributable to both the appearance of new vesicles in the evanescent field and an increase in the background fluorescence of the cell, most likely because new tubulovesicular structures approach the plasma membrane.
The formula in Figure 2B quantified the flow-induced change in fluorescence (F) for a region of interest in each cell. The change in fluorescence in response to flow for the cell shown in Figure 2C is graphed in Figure 2E and compared with a cell recorded under the same conditions in the absence of flow. When flow was repetitively applied (Figure 2F), the fluorescence signal increased additively. Flow termination resulted in a variable decrease in the fluorescence signal, attributable either to retrieval of GFP-TRPM7 from the plasma membrane or to reversal of mechanical deformation.
The flow-induced change in fluorescence observed for GFP-TRPM7 (Figure 2G) was not observed for TRP proteins representative of 2 other subfamilies. GFP-TRPC5 is found in large (&300 nm) vesicles and is inserted in the plasma membrane in response to epidermal growth factor15 but not in response to flow (supplemental Movie IV). GFP-TRPV5 displayed a tubulovesicular distribution similar to that of GFP-TRPM7, but no change in its distribution was observed in response to flow (supplemental Movie V).
To determine whether GFP-TRPM7 fluorescence changes resulted from plasma membrane motion or to mechanical deformation of the cells under flow, experiments were repeated using a fluorescent plasma membrane marker (PMem-EYFP), as well as a fluorescent ER marker (pEYFP-ER). We did not detect a global change in fluorescence in response to flow with either of these probes (supplemental Movie VI). These findings suggest that the increase in GFP-TRPM7 fluorescence is attributable to the transport of new molecules to the region of the plasma membrane.
To test whether application of flow leads to an increase in the number of functional GFP-TRPM7 channels in the plasma membrane, we used the perforated-patch clamp technique to monitor whole-cell GFP-TRPM7 current in response to a voltage ramp (Figure 3A). Before flow application, the mean current amplitude at +100 mV increased at a rate of & 1 pA/s, as is commonly observed with TRPM7 recordings under our conditions. On application of flow (shear force of &20 dyne/cm2), the gradual increase in the maximal current was accelerated by 3- to 4-fold and reached a plateau after persistent flow (Figure 3B). In response to successive applications of flow, the GFP-TRPM7 current increased repetitively, each time with a similar rate (Figure 3C and 3D).
Was the shear force-induced run-up a new component of TRPM7 current or a separate, endogenous current The use of low chloride concentrations (&4 mmol/L) should have eliminated any significant anionic conductances arising in response to shear force.13 TRPM7 channels, also known as magnesium-inhibitable channels (MICs),20 are blocked by millimolar concentrations of Mg2+ (via entry and block from the inside of the channel). Current amplitude at +100 mV decreased to &5% of the initial value within a few seconds of adding 10 mmol/L extracellular Mg2+ (Figure 3E; inset in Figure 3F). These results show that Mg2+ similarly blocked both TRPM7 and the flow-induced current. Mg2+ blocked the total current after flow application to the same extent as TRPM7 channels in the absence of flow (95±1%; Figure 3F), suggesting that TRPM7 channels carried the flow-induced current.
GFP-TRPM7 Fluorescence and ITRPM7 Response to Increasing Rates of Flow
The change in fluorescence and ITRPM7 amplitude were measured over a range of flow rates (Figure 4 and Materials and Methods) and plotted as a function of applied shear force. Flow rates varied between 1.2 mL/min and 0.03 mL/min, generating shear stress values between 30 and 0.2 dyne/cm2. The response of the cells for the same flow rate was variable, both in amplitude (Figure 4B and 4E) and time course. The data show that increasing amounts of shear stress increase both GFP-TRPM7 fluorescence at and near the plasma membrane, and ITRPM7. More importantly, the 2 measured responses have a similar dependence on flow rate; the responses are minimal for Q <5 dyne/cm2, and reach a maximum at 15 dyne/cm2. Shear stress greater than 30 dyne/cm2 detached the cells from the surface.
These results suggest that the 2 types of response to shear stress describe the same cellular process. The observed increase in current amplitude on flow application in perforated patch (&6 to 8 pA/s) corresponds to the incorporation of &1 to 2 channels/s (TRPM7 at 100 mV &105 pS17). The flow-induced increase in current may result from an increase in the rate of TRPM7-containing vesicle fusion, increased channel gating, or both. To further characterize the mechanism mediating the response of TRPM7 to fluid flow, we tested a signaling pathway that we had previously found to regulate TRPC5.15 Both the increase in fluorescence measured by TIRF, and the increase in current amplitude, were unchanged in the presence of phosphatidylinositol 3-kinase (PI3K) inhibitors (50 nmol/L wortmannin; supplemental Figure I; or 100 eol/L LY294002). The PLC inhibitor, U73122, also had no effect on the response of TRPM7 to flow. Thus the flow enhancement of TRPM7 does not appear to require PI3K or PLC, and in this respect is different from the mechanism mediating growth factor-induced TRPC5 incorporation into the plasma membrane.15
TRPM7 contains both functional channel and kinase domains. A kinase inactive mutant of the channel (GFP-TRPM7-KD; containing the K1645A mutation) showed the same current amplitude and fluorescence response to flow as wt TRPM7 (supplemental Figure I). We conclude that the kinase domain of TRPM7 is not required in the immediate response to flow but cannot rule out the possibility that it mediates downstream effects of shear stress either by phosphorylating protein targets or by acting as a binding and/or anchoring protein for downstream effectors.
Distribution and Mobility of GFP-TRPM7 in a Vascular Smooth Muscle Cell Line
Several cells from tissues that are exposed to flow-induced shear stress values in the 10 to 20 dyne/cm2 range were examined. Among these, endothelial cells (HUVECs, HAECs, and HPAECs) had very low levels of TRPM7-like current, and showed no significant response to flow. However, TRPM7 is present in vascular smooth muscle cells,21 which are exposed to blood flow under pathological conditions. In A7R5 aortic smooth muscle cells expressing GFP-TRPM7, the distribution of the channel was markedly different from that in HEK-293 cells (Figure 5A versus Figure 1D), attributable in part to the particular morphology of vascular smooth muscle cells. A7R5 cells are very thin (maximum height=3 to 5 e in the region of the nucleus), so that the distribution pattern of GFP-TRPM7 is more readily apparent, even under the lower z-axis resolution of confocal microscopy (&0.5 to 1 e). In A7R5 cells, GFP-TRPM7 appears mostly vesicular and these vesicles are restricted to linear filamentous structures (Figure 5B). A similar distribution is observed for the endogenous TRPM7 protein when A7R5 cells were immunostained with an antibody against the C terminus of TRPM7 (Figure 5C and 5D). Endogenous TRPM7 was also detected in A7R5 cells by Western blot using the same antibody (data not shown).
Under TIRF microscopy, the same type and size of vesicular structures were seen in A7R5 cells expressing GFP-TrpM7 (Figure 5E) and in A7R5 cells immunostained for native TRPM7 (Figure 5G). When a series of 100 images of A7R5 cells expressing GFP-TRPM7 were recorded, the highly dynamic nature of the vesicles became apparent (supplemental Movie III). The vesicles moved along linear trajectories with variable speeds in the 1 e/s range (Figure 5F). Flow did not induce any significant increase in the fluorescence at or near the plasma membrane in A7R5 cells. This may be attributable to their flat morphology; the pool of vesicles above the evanescent field may be too small to compensate for photobleaching. To determine whether the paths along which the vesicles moved were related to actin or tubulin cytoskeletal elements, fixed A7R5 expressing GFP-TRPM7 were immunostained with Alexa 568 labeled phalloidin or antieC-tubulin antibody. Confocal images did not reveal any significant correlation between actin or -tubulin and the distribution of GFP. In HEK-293 cells expressing GFP-TRPM7 and RED2-ER (endoplasmic reticulum) marker, colocalization was inconclusive because both molecules appear diffusely localized throughout the cytosol. In A7R5 cells, however, the distributions were distinct, with the RED2-ER marker homogeneously distributed throughout the cell, whereas GFP-TRPM7 displayed the typical "vesicles on tracks" distribution.
Increase in Endogenous TRPM7-Like Current in Response to Fluid Flow
TRPM7-like currents were recorded in A7R5 cells, but the initial current amplitude at +100 mV varied greatly (from <50 pA to >1000 pA; Figure 6C). When shear force corresponding to 15-dyne/cm2 was applied to the cells, the maximal amplitude of the current increased significantly (Figure 6A). The amplitude of the increase in TRPM7-like current at +100 mV (Imax, calculated as described in Materials and Methods) was 148±15 pA, with a half-time to maximum current (T1/2) of 42±1.7 s (Figure 6B). The amplitude and time course of the response to flow was in the same range for all the measured cells, despite highly variable initial values of the TRPM7-like current. This observation suggests that the number of channels incorporated or activated in response to flow is not dependent on the number of functional channels already in the membrane.
The flow-induced component of the TRPM7-like current was blocked by 10 mmol/L extracellular Mg2+ (Figure 6). In the absence of flow, 10 mmol/L extracellular Mg2+ blocked 93±2% of the initial TRPM7-like current, whereas in the presence of flow the same concentration of Mg2+ blocked 90±2.5% of the TRPM7-like plus the flow-induced component of the current. We conclude that the observed response to flow in A7R5 cells, and in our recombinant expression system, involves the same channel and uses a similar mechanism.
Discussion
We have shown that exposure of cells to shear stress leads to a significant accumulation of functional TRPM7 channels at the plasma membrane in <2 minutes. Shear force increased both GFP-TRPM7 and endogenous TRPM7-like current and GFP-TRPM7 fluorescence at and near the plasma membrane. The rapid time course for this response suggests that TRPM7 is among the first molecules to respond to shear stress. Most cellular events induced by shear stress, such as cytoskeletal remodeling22,23 and changes in gene expression pattern,24,25 occur in the time scale of minutes to hours.
Under steady-state conditions, TRPM7 channels exhibit a gradual increase in outward current amplitude (run-up), both in perforated and ruptured whole-cell voltage-clamp. Among the TRP channel family, this effect is characteristic for TRPM7 and its close relative, TRPM6. This may represent gradual insertion of channels into the plasma membrane or an alteration in baseline channel gating. By either mechanism, this increase is accelerated in response to physiological levels of shear stress. The observed translocation of vesicular and tubulovesicular GFP-TRPM7 (in HEK-293, CHO-K1, COS, and MEF cells) suggests that flow leads to an increase in the fusion rate of the TRPM7-containing vesicles with the plasma membrane. We did not detect vesicle fusion followed by diffusion of the fluorescent molecules in the plasma membrane as described for post-Golgi carriers by Schmoranzer et al26 and Toomre et al.27 Insufficient temporal imaging resolution, or increased fluorescence background generated by the reticular structures overlapping with the vesicles, could have prevented such observations in our system.
Potential Mechanisms for the Response of TRPM7 to Fluid Flow
We examined several general hypotheses of how flow is translated into TRPM7 translocation. The most readily testable hypothesis involved the TRPM7 kinase domain, activation of PI3K or PLC pathways, and stretch sensing via the cytoskeleton. Unlike the mechanism mediating growth factoreCinduced TRPC5 incorporation into the plasma membrane,15 flow-induced TRPM7 enhancement is not likely to involve PI3K or PLC. Similarly, the kinase activity of TRPM7 is not involved in the immediate response to flow. The role of the cytoskeleton in translocation could not be adequately tested and remains an unexplored area for research. Thus, we do not know the steps that lead to the increase in plasma membrane TRPM7 in response to flow. Our hypothesis is that shear stress, first sensed by the extracellular matrix interaction molecules of the cell, signals to cytoskeletal proteins, which in turn lead to activation of motor proteins that transport TRPM7 toward the plasma membrane. The cytoskeletal elements and the motor proteins involved in this process remain to be determined.
Endogenous TRPM7 As an Acute Responder to Shear Stress
Unlike endothelial cells, vascular smooth muscle cells (VSMCs) derived from aortic smooth muscle had high levels of TRPM7-like current and TRPM7 protein, and a significant and consistent response to flow. VSMCs are in direct contact with the endothelium cells via myoendothelial bridges. Under normal physiological conditions, they are subject to low levels of flow but are directly exposed to shear stress at sites of vascular damage caused by the rupture of atherosclerotic lesions or invasive techniques, such as angioplasty. Such exposure of these cells to shear stress may elicit an acute response to fluid flow. Cells that are permanently exposed to higher levels of shear stress, like endothelial cells, may downregulate their responses to adapt to steady-state conditions. In contrast, cells like VSMCs, which are exposed to such levels of shear stress only under extreme conditions, may respond quickly, therefore expressing higher levels of TRPM7 and other molecules that may be involved in the immediate response. In response to TRPM7 activation, intracellular calcium increases. Intracellular calcium is known to induce smooth muscle proliferation at lower levels and cell death at high levels.
Acknowledgments
The Howard Hughes Medical Institute supported this work. We thank members of the Clapham laboratory for experimental support and insightful discussions and Drs Andrew Scharenberg, T. Lambers, and Rene Bindels for constructs.
Both authors contributed equally to this study.
References
Clapham DE. TRP channels as cellular sensors. Nature. 2003; 426: 517eC524.
Yamaguchi H, Matsushita M, Nairn AC, Kuriyan J. Crystal structure of the atypical protein kinase domain of a TRP channel with phosphotransferase activity. Mol Cell. 2001; 7: 1047eC1057.
Ryazanova LV, Dorovkov MV, Ansari A, Ryazanov AG. Characterization of the protein kinase activity of TRPM7/ChaK1, a protein kinase fused to the transient receptor potential ion channel. J Biol Chem. 2004; 279: 3708eC3716.
Runnels LW, Yue L, Clapham DE. The TRPM7 channel is inactivated by PIP(2) hydrolysis. Nat Cell Biol. 2002; 4: 329eC336.
Takezawa R, Schmitz C, Demeuse P, Scharenberg AM, Penner R, Fleig A. Receptor-mediated regulation of the TRPM7 channel through its endogenous protein kinase domain. Proc Natl Acad Sci U S A. 2004; 101: 6009eC6014.
Aarts M, Iihara K, Wei WL, Xiong ZG, Arundine M, Cerwinski W, MacDonald JF, Tymianski M. A key role for TRPM7 channels in anoxic neuronal death. Cell. 2003; 115: 863eC877.
Schmitz C, Perraud AL, Johnson CO, Inabe K, Smith MK, Penner R, Kurosaki T, Fleig A, Scharenberg AM. Regulation of vertebrate cellular Mg2+ homeostasis by TRPM7. Cell. 2003; 114: 191eC200.
Alenghat FJ, Nauli SM, Kolb R, Zhou J, Ingber DE. Global cytoskeletal control of mechanotransduction in kidney epithelial cells. Exp Cell Res. 2004; 301: 23eC30.
Resnick N, Yahav H, Shay-Salit A, Shushy M, Schubert S, Zilberman LC, Wofovitz E. Fluid shear stress and the vascular endothelium: for better and for worse. Prog Biophys Mol Biol. 2003; 81: 177eC199.
Hoger JH, Ilyin VI, Forsyth S, Hoger A. Shear stress regulates the endothelial Kir2.1 ion channel. Proc Natl Acad Sci U S A. 2002; 99: 7780eC7785.
Nauli SM, Alenghat FJ, Luo Y, Williams E, Vassilev P, Li X, Elia AE, Lu W, Brown EM, Quinn SJ, Ingber DE, Zhou J. Polycystins 1 and 2 mediate mechanosensation in the primary cilium of kidney cells. Nat Genet. 2003; 33: 129eC137.
Traub O, Ishida T, Ishida M, Tupper JC, Berk BC. Shear stress-mediated extracellular signal-regulated kinase activation is regulated by sodium in endothelial cells. Potential role for a voltage-dependent sodium channel. J Biol Chem. 1999; 274: 20144eC20150.
Barakat AI, Leaver EV, Pappone PA, Davies PF. A flow-activated chloride-selective membrane current in vascular endothelial cells. Circ Res. 1999; 85: 820eC828.
Byron KL, Taylor CW. Spontaneous Ca2+ spiking in a vascular smooth muscle cell line is independent of the release of intracellular Ca2+ stores. J Biol Chem. 1993; 268: 6945eC6952.
Bezzerides VJ, Ramsey IS, Kotecha S, Greka A, Clapham DE. Rapid vesicular translocation and insertion of TRP channels. Nat Cell Biol. 2004; 6: 709eC720.
Street RL WG, Vennard JK. Elementary Fluid Mechanics. 7th ed. New York: John Wiley and Sons; 1996.
Runnels LW, Yue L, Clapham DE. TRP-PLIK, a bifunctional protein with kinase and ion channel activities. Science. 2001; 291: 1043eC1047.
Nadler MJ, Hermosura MC, Inabe K, Perraud AL, Zhu Q, Stokes AJ, Kurosaki T, Kinet JP, Penner R, Scharenberg AM, Fleig A. LTRPC7 is a Mg.ATP-regulated divalent cation channel required for cell viability. Nature. 2001; 411: 590eC595.
Steyer JA, Almers W. A real-time view of life within 100 nm of the plasma membrane. Nat Rev Mol Cell Biol. 2001; 2: 268eC275.
Prakriya M, Lewis RS. Separation and characterization of currents through store-operated CRAC channels and Mg2+-inhibited cation (MIC) channels. J Gen Physiol. 2002; 119: 487eC507.
He Y, Yao G, Savoia C, Touyz RM. Transient receptor potential melastatin 7 ion channels regulate magnesium homeostasis in vascular smooth muscle cells: role of angiotensin II. Circ Res. 2005; 96: 207eC215.
Galbraith CG, Skalak R, Chien S. Shear stress induces spatial reorganization of the endothelial cell cytoskeleton. Cell Motil Cytoskeleton. 1998; 40: 317eC330.
Wojciak-Stothard B, Ridley AJ. Shear stress-induced endothelial cell polarization is mediated by Rho and Rac but not Cdc42 or PI 3-kinases. J Cell Biol. 2003; 161: 429eC439.
Chiu JJ, Lee PL, Chang SF, Chen LJ, Lee CI, Lin KM, Usami S, Chien S. Shear stress regulates gene expression in vascular endothelial cells in response to tumor necrosis factor-alpha: a study of the transcription profile with complementary DNA microarray. J Biomed Sci. 2005; 12: 481eC502.
Miyagi M, Miwa Y, Takahashi-Yanaga F, Morimoto S, Sasaguri T. Activator protein-1 mediates shear stress-induced prostaglandin d synthase gene expression in vascular endothelial cells. Arterioscler Thromb Vasc Biol. 2005; 25: 970eC975.
Schmoranzer J, Goulian M, Axelrod D, Simon SM. Imaging constitutive exocytosis with total internal reflection fluorescence microscopy. J Cell Biol. 2000; 149: 23eC32.
Toomre D, Steyer JA, Keller P, Almers W, Simons K. Fusion of constitutive membrane traffic with the cell surface observed by evanescent wave microscopy. J Cell Biol. 2000; 149: 33eC40.(Elena Oancea, Joshua T. W)
Abstract
Many cells are constantly exposed to fluid mechanical forces generated by flowing blood, and wall shear stresses modulate aspects of their structure and function. However, the mechanisms for mechanotransduction of flow are not well understood. Here we report that TRPM7, which is both an ion channel and a functional kinase, is translocated within cells in response to laminar flow. After exposure of cells to physiological values of laminar fluid flow, the number of TRPM7 molecules localized at or near the plasma membrane increased up to 2-fold, in less than 100 seconds. This increase in membrane-localized GFP-TRPM7, as seen by total internal reflection fluorescence microscopy, closely correlated with increases in TRPM7 current. Both endogenous and heterologously expressed TRPM7 was found in tubulovesicular structures that were translocated to the region of the plasma membrane on induction of shear stress. In vascular smooth muscle cells, but not in several types of endothelial cells, fluid flow increased endogenous native TRPM7 current amplitude. We hypothesize that TRPM7 plays a role in pathological response to vessel wall injury.
Key Words: TRP ion channels TRPM7 shear force total internal reflection fluorescence (TIRF) microscopy vascular smooth muscle cells
Introduction
TRP ion channels are often found in sensory systems, but their wide distribution suggests they are cellular sensors in a broad sense.1 TRPM6 and TRPM7 are unique channels that possess both ion channel and protein kinase activities. The C-terminal kinase domain bears little sequence identity with other kinases but is enzymatically active and structurally homologous to protein kinase A.2,3 The physiological mechanism for activation of the widely expressed TRPM7 is unknown. TRPM7 channels are inhibited by phospholipase C (PLC)-catalyzed phosphatidylinositol 4,5-bisphosphate (PIP2)hydrolysis4 and reportedly regulated through protein kinase A by receptors coupled to adenylyl cyclase.5 TRPM7 may be involved in anoxia-induced cell death in brain6 and has been reported to be required for cell viability in TRPM7-null avian DT40 cells.7
Cells sense shear stresseCinduced mechanical stimulation and convert it into a biochemical response that impacts normal and abnormal tissue development, including growth, differentiation, migration, gene expression, protein synthesis, and apoptosis.8 The response of cells to fluid flow depends on the cell type and the magnitude and characteristics of the shear stress applied under physiological or pathological conditions. Endothelial cells are constantly subjected to blood flow that regulates physiological blood vessel responses as well as pathological arterial wall responses. In large arteries, the endothelium is exposed to shear stress values in the range 10 to 40 dyne/cm2.9 The vascular smooth muscle cells are protected from shear stress by endothelial cell lining under physiological conditions but become exposed to shear stress after endothelial injury.
Ion channels can respond to shear stress.10,11 A few ion channels sensitive to shear stress have been identified in endothelial cells and partially characterized.9 Among these are K+ channels,10 Na+ channels,12 and CleC channels.13 Most of these channels respond on a rather fast-time scale (minutes) to mechanical stimulation. However, the role of these channels in transforming a mechanical stimulus into a biochemical response is poorly understood. Here we report that TRPM7 responds within 10 of seconds to physiologically relevant fluid flow rates by increasing its presence at the plasma membrane. We show the effect of fluid shear stress on TRPM7 by 2 independent methods: total internal reflection fluorescence (TIRF) imaging and electrophysiological recordings of GFP-TRPM7 in HEK-293 cells. We also show that TRPM7 is highly expressed endogenously in vascular smooth muscle cells that respond to shear stress under pathological conditions. The endogenous TRPM7 and the expressed GFP-TRPM7 share the same highly dynamic subcellular distribution of "vesicles on tracks." Moreover, the endogenous TRPM7-like current increases significantly in response to fluid flow. Our results show that intracellular TRPM7 channels are found in tubulovesicular structures that are delivered to the plasma membrane in response to fluid flow.
Materials and Methods
cDNA Constructs
Green fluorescent protein (GFP) was amplified by PCR and inserted in the NotI site, between the FLAG tag and the N terminus of mTRPM7. FLAG-mTRPM7 in pcDNA4/TO was provided by Dr Andrew Scharenberg (University of Washington). GFP-TRPC5 was obtained by fusing GFP to the C terminus of TRPC5 into the pCI vector. GFP-TRPV5 was cloned in the eGFP/C1 vector by fusing TRPV5 to the C terminus of GFP and was provided by Dr Rene Bindels (University of Nijmegen). pEYFP-Mem, pEYFP-ER cDNA constructs were purchased from Clontech.
Cell Culture
HEK293T cells were cultured in medium containing DMEM/Ham’s F12, 10% FBS, and GHT supplement (Gibco). Cells were cultured on glass coverslips for the imaging experiments and on plastic Petri dishes for the electrophysiology experiments. In some imaging experiments, the glass coverslips were coated with Poly-L-Lysine (Sigma). A7R5 cells were cultured in DMEM supplemented with 10% FBS. Cells were tested periodically for a smooth muscle phenotype by assessing spontaneous calcium spikes mediated by L-type calcium channels14 using Fluo-4 imaging. Cells were split before becoming confluent (70% to 80% density) to a density >20% and the media changed every 2 to 3 days.
DNA Transfection
HEK293T cells were transiently transfected using Lipofectamine 2000 (Invitrogen). Cells were incubated with the transfection agent for 2 to 3 hours and then cultured on glass or plastic 24 to 48 hours before the experiment. A7R5 cells were transiently transfected with Superfect (Qiagen). Cells were plated at least 24 hours before the transfection and used at 50% to 70% density. Thirty-five-millimeter dishes were transfected with 20 e蘈 of Superfect and 2 e of DNA mixed with 100 e蘈 of OPTI-MEM and added to 500 e蘈 of medium. Cells were incubated with the transfection reagent for 2 to 4 hours and then washed with normal medium.
Imaging and Image Analysis
All live imaging experiments were performed at &25°C with cells grown on glass coverslips and placed in a custom chamber with standard extracellular buffer. Extracellular buffer contained (in mmol/L): 135 NaCl, 5 KCl, 1.5 CaCl2, 1.5 MgCl2, 20 HEPES, pH 7.4. Scanning confocal microscopy was used to image the distribution of GFP-TRPM7 in different cell types. Cells expressing GFP-TRPM7 or immunostained with anti-TRPM7 primary antibody against the C terminus of TRPM7 and Alexa 488 secondary antibody were excited by 488-nm laser illumination; the emission light was bandpass filtered from 512 to 527 nm (Olympus Fluoview 300). For TIRF imaging, we used a custom-built, objective-based evanescent field microscope.15 Cells grown on glass coverslips were placed in a custom chamber with standard extracellular buffer and imaged at &25°C. All image acquisition and some of the analysis was performed using MetaMorph (Universal Imaging, Westchester, Pa). Time series of 100 to 400 images at 1- to 4-s intervals were recorded. Fluorescence intensity in a region of interest comprising at least 50% of the visible "footprint" of the cell was measured as a function of time using MetaMorph. The data were then analyzed using Deltagraph (Red Rock Software, Salt Lake City, Utah) as described in Results.
Electrophysiological Recordings and Data Analysis
GFP-TRPM7eCexpressing HEK-293T cells were detected by GFP fluorescence. Membrane currents were digitized at 20 kHz and filtered at 1 kHz. Voltage stimuli were delivered at 5-s intervals, with voltage ramps from eC100 to +100 mV (holding potential, 0 mV). The internal pipette solution contained (in mmol/L); 120 Cs-methanesulfonate, 8 NaCl, 10 EGTA, 4.1 CaCl2, 10 HEPES; pH adjusted to 7.2 with CsOH. For perforated patch recording, 150 e/mL Amphotericin B was included in the pipette solution. Amphotericin B was dissolved in DMSO at 1.5 mg/mL and diluted 1:100 into the pipette solution. Recordings were initiated when series resistance reached values below 30 M. The extracellular solution was designed to minimize chloride currents, and consisted of (in mmol/L) 140 Na-methanesulfonate, 5 Cs-methanesulfonate, 2.8 KCl, 1 CaCl2, 10 HEPES, 10 mmol/L glucose; pH adjusted to 7.4 with NaOH.
Because initial current amplitudes and rates of current run-up in the outward direction varied from cell to cell, a slope-based method was used to quantify shear force-induced changes in ITRPM7. Average current amplitude at 100 mV was plotted as a function of time, and the slope of the data points was taken before and after application of shear force. The baseline rate of current increase (the slope before shear force was applied; S1) was determined with at least 6 data points (over a 30-s period). The current increase caused by flow was determined by taking the slope of the data points after an increase in slope (S2), until a plateau was reached. An increase in slope was only counted if observed within 30 s of flow application. The reported increase in current caused by flow was calculated as (S2eCS1)xt, where t was the length of time over which the increase in slope (S2) occurred.
Flow Setup and Calculation
A syringe pump (Harvard Apparatus, Harvard, Mass) connected through thin tubing to a 250-e ID plastic pipette provided fluid flow and flow values set between 0.123 and 1.23 mL/min. The pipette was positioned &150 e from the cell.
The magnitude of the shear stress generated by the fluid flow was calculated using the formula:
where =1025 kg/m3 is the density of water and V is the fluid velocity, calculated by V=Q/A, where A=d2/4=0.785x10eC8m2 for a pipette diameter of d=250 e; Q is the flow generated by the syringe pump measured in m3/s.
Rx, the Reynolds number, was Rx=Vxx/, where =1.139x10eC6 m2/s is the kinematic viscosity of the water, x=150 e is the distance between pipette and cell. Rx varied from 5.5 to 27.5, depending on flow rate.
To determine whether flow was laminar, we also used the Reynolds number defined as Rh=Vh/v, where h, the fluid level in the dish (3 mm), was kept constant; and V are the kinematic viscosity of the water and the fluid velocity, respectively.
For the highest value of the flow used in our experiments the Reynolds numbers were Rx=27.5 and Rh=627. Laminar flow occurs for Rx<500 000 and Rh<3900.16
Results
GFP-TRPM7 Forms a Functional Channel and Reveals a Dynamic Tubulovesicular Subcellular Distribution
To follow the intracellular distribution of TRPM7, GFP was fused to the N terminus of the channel via a (GA)2 linker (Figure 1A). We first tested the functionality of the GFP-tagged channel by transfecting GFP-TRPM7 into HEK-293 cells and measuring the TRPM7 current in response to a voltage ramp (Figure 1B). GFP-TRPM7 retained all the characteristics of native and recombinant TRPM7 current, including a gradual increase in outward rectification or "run-up"17,18 (Figure 1C).
The intracellular distribution of GFP-TRPM7 was initially studied using confocal microscopy. GFP-TRPM7 in HEK-293 cells is expressed in a reticular pattern throughout the cytoplasm (Figure 1D). Although the cells have functional TRPM7 current, the plasma membrane localized GFP-TRPM7 is relatively less abundant compared with its intracellular distribution and thus not apparent in these images.
Because of the limited resolution of the confocal images, we used evanescent field objective-based (or TIRF) microscopy to visualize GFP-TRPM7 in HEK-293 cells. The evanescent field selectively illuminates fluorophores within &250 nm of the plasma membrane-glass cover-slip interface.15,19 Evanescent light decreases exponentially in the axial direction, such that fluorescence emission is brighter as the fluorophore approaches the plane of the interface. Although TIRF axial resolution is superior to confocal microscopy, it is insufficient by itself to determine whether a protein is truly in the &4-nm plasma membrane.
TIRF imaging of HEK-293 cells expressing GFP-TRPM7 revealed the GFP-labeled channels at or near the plasma membrane (Figure 1E). The proteins appear punctate, which we assume to be vesicular structures, and in tubular structures with diameters <100 nm. Most cells contain a mix of the 2 types of structures (first cell in Figure 1E), with some cells containing predominantly tubular structures (second cell in Figure 1E) or predominantly vesicular structures (third cell in Figure 1E). This distribution pattern is markedly different from TIRF images of TRPC5-GFP15 but similar to that of GFP-TRPM6 and GFP-TRPV5 channels (data not shown).
We recorded series of 100 images at 1- or 2-s intervals, under steady-state conditions at 22°C. The time series show that both the tubular and the vesicular structures are highly mobile (Movie I in the online data supplement available at http://circres.ahajournals.org). Most strikingly, linear tracks or trajectories form in seconds, and vesicles begin to travel on them; then just as quickly, they disappear (Figure 1F). The same type of movement is not observed by stepping the axial (z) position of the objective in fixed cells (data not shown), suggesting that the observed motility is attributable to movement in the x-y plane and not to movement in and out of the z-plane.
Increased GFP-TRPM7 at the Plasma Membrane in Response to Shear Force
To determine the effects of shear stress on the distribution and functional current of GFP-TRPM7 channels, individual cells were exposed to flow as shown in Figure 2A (see Materials and Methods). When sequential TIRF images were recorded before and after the application of flow, we observed a gradual increase in the fluorescence at and near the plasma membrane (Figure 2C and supplemental Movie II). Higher resolution TIRF imaging (Figure 2D) shows that the global increase in fluorescence is attributable to both the appearance of new vesicles in the evanescent field and an increase in the background fluorescence of the cell, most likely because new tubulovesicular structures approach the plasma membrane.
The formula in Figure 2B quantified the flow-induced change in fluorescence (F) for a region of interest in each cell. The change in fluorescence in response to flow for the cell shown in Figure 2C is graphed in Figure 2E and compared with a cell recorded under the same conditions in the absence of flow. When flow was repetitively applied (Figure 2F), the fluorescence signal increased additively. Flow termination resulted in a variable decrease in the fluorescence signal, attributable either to retrieval of GFP-TRPM7 from the plasma membrane or to reversal of mechanical deformation.
The flow-induced change in fluorescence observed for GFP-TRPM7 (Figure 2G) was not observed for TRP proteins representative of 2 other subfamilies. GFP-TRPC5 is found in large (&300 nm) vesicles and is inserted in the plasma membrane in response to epidermal growth factor15 but not in response to flow (supplemental Movie IV). GFP-TRPV5 displayed a tubulovesicular distribution similar to that of GFP-TRPM7, but no change in its distribution was observed in response to flow (supplemental Movie V).
To determine whether GFP-TRPM7 fluorescence changes resulted from plasma membrane motion or to mechanical deformation of the cells under flow, experiments were repeated using a fluorescent plasma membrane marker (PMem-EYFP), as well as a fluorescent ER marker (pEYFP-ER). We did not detect a global change in fluorescence in response to flow with either of these probes (supplemental Movie VI). These findings suggest that the increase in GFP-TRPM7 fluorescence is attributable to the transport of new molecules to the region of the plasma membrane.
To test whether application of flow leads to an increase in the number of functional GFP-TRPM7 channels in the plasma membrane, we used the perforated-patch clamp technique to monitor whole-cell GFP-TRPM7 current in response to a voltage ramp (Figure 3A). Before flow application, the mean current amplitude at +100 mV increased at a rate of & 1 pA/s, as is commonly observed with TRPM7 recordings under our conditions. On application of flow (shear force of &20 dyne/cm2), the gradual increase in the maximal current was accelerated by 3- to 4-fold and reached a plateau after persistent flow (Figure 3B). In response to successive applications of flow, the GFP-TRPM7 current increased repetitively, each time with a similar rate (Figure 3C and 3D).
Was the shear force-induced run-up a new component of TRPM7 current or a separate, endogenous current The use of low chloride concentrations (&4 mmol/L) should have eliminated any significant anionic conductances arising in response to shear force.13 TRPM7 channels, also known as magnesium-inhibitable channels (MICs),20 are blocked by millimolar concentrations of Mg2+ (via entry and block from the inside of the channel). Current amplitude at +100 mV decreased to &5% of the initial value within a few seconds of adding 10 mmol/L extracellular Mg2+ (Figure 3E; inset in Figure 3F). These results show that Mg2+ similarly blocked both TRPM7 and the flow-induced current. Mg2+ blocked the total current after flow application to the same extent as TRPM7 channels in the absence of flow (95±1%; Figure 3F), suggesting that TRPM7 channels carried the flow-induced current.
GFP-TRPM7 Fluorescence and ITRPM7 Response to Increasing Rates of Flow
The change in fluorescence and ITRPM7 amplitude were measured over a range of flow rates (Figure 4 and Materials and Methods) and plotted as a function of applied shear force. Flow rates varied between 1.2 mL/min and 0.03 mL/min, generating shear stress values between 30 and 0.2 dyne/cm2. The response of the cells for the same flow rate was variable, both in amplitude (Figure 4B and 4E) and time course. The data show that increasing amounts of shear stress increase both GFP-TRPM7 fluorescence at and near the plasma membrane, and ITRPM7. More importantly, the 2 measured responses have a similar dependence on flow rate; the responses are minimal for Q <5 dyne/cm2, and reach a maximum at 15 dyne/cm2. Shear stress greater than 30 dyne/cm2 detached the cells from the surface.
These results suggest that the 2 types of response to shear stress describe the same cellular process. The observed increase in current amplitude on flow application in perforated patch (&6 to 8 pA/s) corresponds to the incorporation of &1 to 2 channels/s (TRPM7 at 100 mV &105 pS17). The flow-induced increase in current may result from an increase in the rate of TRPM7-containing vesicle fusion, increased channel gating, or both. To further characterize the mechanism mediating the response of TRPM7 to fluid flow, we tested a signaling pathway that we had previously found to regulate TRPC5.15 Both the increase in fluorescence measured by TIRF, and the increase in current amplitude, were unchanged in the presence of phosphatidylinositol 3-kinase (PI3K) inhibitors (50 nmol/L wortmannin; supplemental Figure I; or 100 eol/L LY294002). The PLC inhibitor, U73122, also had no effect on the response of TRPM7 to flow. Thus the flow enhancement of TRPM7 does not appear to require PI3K or PLC, and in this respect is different from the mechanism mediating growth factor-induced TRPC5 incorporation into the plasma membrane.15
TRPM7 contains both functional channel and kinase domains. A kinase inactive mutant of the channel (GFP-TRPM7-KD; containing the K1645A mutation) showed the same current amplitude and fluorescence response to flow as wt TRPM7 (supplemental Figure I). We conclude that the kinase domain of TRPM7 is not required in the immediate response to flow but cannot rule out the possibility that it mediates downstream effects of shear stress either by phosphorylating protein targets or by acting as a binding and/or anchoring protein for downstream effectors.
Distribution and Mobility of GFP-TRPM7 in a Vascular Smooth Muscle Cell Line
Several cells from tissues that are exposed to flow-induced shear stress values in the 10 to 20 dyne/cm2 range were examined. Among these, endothelial cells (HUVECs, HAECs, and HPAECs) had very low levels of TRPM7-like current, and showed no significant response to flow. However, TRPM7 is present in vascular smooth muscle cells,21 which are exposed to blood flow under pathological conditions. In A7R5 aortic smooth muscle cells expressing GFP-TRPM7, the distribution of the channel was markedly different from that in HEK-293 cells (Figure 5A versus Figure 1D), attributable in part to the particular morphology of vascular smooth muscle cells. A7R5 cells are very thin (maximum height=3 to 5 e in the region of the nucleus), so that the distribution pattern of GFP-TRPM7 is more readily apparent, even under the lower z-axis resolution of confocal microscopy (&0.5 to 1 e). In A7R5 cells, GFP-TRPM7 appears mostly vesicular and these vesicles are restricted to linear filamentous structures (Figure 5B). A similar distribution is observed for the endogenous TRPM7 protein when A7R5 cells were immunostained with an antibody against the C terminus of TRPM7 (Figure 5C and 5D). Endogenous TRPM7 was also detected in A7R5 cells by Western blot using the same antibody (data not shown).
Under TIRF microscopy, the same type and size of vesicular structures were seen in A7R5 cells expressing GFP-TrpM7 (Figure 5E) and in A7R5 cells immunostained for native TRPM7 (Figure 5G). When a series of 100 images of A7R5 cells expressing GFP-TRPM7 were recorded, the highly dynamic nature of the vesicles became apparent (supplemental Movie III). The vesicles moved along linear trajectories with variable speeds in the 1 e/s range (Figure 5F). Flow did not induce any significant increase in the fluorescence at or near the plasma membrane in A7R5 cells. This may be attributable to their flat morphology; the pool of vesicles above the evanescent field may be too small to compensate for photobleaching. To determine whether the paths along which the vesicles moved were related to actin or tubulin cytoskeletal elements, fixed A7R5 expressing GFP-TRPM7 were immunostained with Alexa 568 labeled phalloidin or antieC-tubulin antibody. Confocal images did not reveal any significant correlation between actin or -tubulin and the distribution of GFP. In HEK-293 cells expressing GFP-TRPM7 and RED2-ER (endoplasmic reticulum) marker, colocalization was inconclusive because both molecules appear diffusely localized throughout the cytosol. In A7R5 cells, however, the distributions were distinct, with the RED2-ER marker homogeneously distributed throughout the cell, whereas GFP-TRPM7 displayed the typical "vesicles on tracks" distribution.
Increase in Endogenous TRPM7-Like Current in Response to Fluid Flow
TRPM7-like currents were recorded in A7R5 cells, but the initial current amplitude at +100 mV varied greatly (from <50 pA to >1000 pA; Figure 6C). When shear force corresponding to 15-dyne/cm2 was applied to the cells, the maximal amplitude of the current increased significantly (Figure 6A). The amplitude of the increase in TRPM7-like current at +100 mV (Imax, calculated as described in Materials and Methods) was 148±15 pA, with a half-time to maximum current (T1/2) of 42±1.7 s (Figure 6B). The amplitude and time course of the response to flow was in the same range for all the measured cells, despite highly variable initial values of the TRPM7-like current. This observation suggests that the number of channels incorporated or activated in response to flow is not dependent on the number of functional channels already in the membrane.
The flow-induced component of the TRPM7-like current was blocked by 10 mmol/L extracellular Mg2+ (Figure 6). In the absence of flow, 10 mmol/L extracellular Mg2+ blocked 93±2% of the initial TRPM7-like current, whereas in the presence of flow the same concentration of Mg2+ blocked 90±2.5% of the TRPM7-like plus the flow-induced component of the current. We conclude that the observed response to flow in A7R5 cells, and in our recombinant expression system, involves the same channel and uses a similar mechanism.
Discussion
We have shown that exposure of cells to shear stress leads to a significant accumulation of functional TRPM7 channels at the plasma membrane in <2 minutes. Shear force increased both GFP-TRPM7 and endogenous TRPM7-like current and GFP-TRPM7 fluorescence at and near the plasma membrane. The rapid time course for this response suggests that TRPM7 is among the first molecules to respond to shear stress. Most cellular events induced by shear stress, such as cytoskeletal remodeling22,23 and changes in gene expression pattern,24,25 occur in the time scale of minutes to hours.
Under steady-state conditions, TRPM7 channels exhibit a gradual increase in outward current amplitude (run-up), both in perforated and ruptured whole-cell voltage-clamp. Among the TRP channel family, this effect is characteristic for TRPM7 and its close relative, TRPM6. This may represent gradual insertion of channels into the plasma membrane or an alteration in baseline channel gating. By either mechanism, this increase is accelerated in response to physiological levels of shear stress. The observed translocation of vesicular and tubulovesicular GFP-TRPM7 (in HEK-293, CHO-K1, COS, and MEF cells) suggests that flow leads to an increase in the fusion rate of the TRPM7-containing vesicles with the plasma membrane. We did not detect vesicle fusion followed by diffusion of the fluorescent molecules in the plasma membrane as described for post-Golgi carriers by Schmoranzer et al26 and Toomre et al.27 Insufficient temporal imaging resolution, or increased fluorescence background generated by the reticular structures overlapping with the vesicles, could have prevented such observations in our system.
Potential Mechanisms for the Response of TRPM7 to Fluid Flow
We examined several general hypotheses of how flow is translated into TRPM7 translocation. The most readily testable hypothesis involved the TRPM7 kinase domain, activation of PI3K or PLC pathways, and stretch sensing via the cytoskeleton. Unlike the mechanism mediating growth factoreCinduced TRPC5 incorporation into the plasma membrane,15 flow-induced TRPM7 enhancement is not likely to involve PI3K or PLC. Similarly, the kinase activity of TRPM7 is not involved in the immediate response to flow. The role of the cytoskeleton in translocation could not be adequately tested and remains an unexplored area for research. Thus, we do not know the steps that lead to the increase in plasma membrane TRPM7 in response to flow. Our hypothesis is that shear stress, first sensed by the extracellular matrix interaction molecules of the cell, signals to cytoskeletal proteins, which in turn lead to activation of motor proteins that transport TRPM7 toward the plasma membrane. The cytoskeletal elements and the motor proteins involved in this process remain to be determined.
Endogenous TRPM7 As an Acute Responder to Shear Stress
Unlike endothelial cells, vascular smooth muscle cells (VSMCs) derived from aortic smooth muscle had high levels of TRPM7-like current and TRPM7 protein, and a significant and consistent response to flow. VSMCs are in direct contact with the endothelium cells via myoendothelial bridges. Under normal physiological conditions, they are subject to low levels of flow but are directly exposed to shear stress at sites of vascular damage caused by the rupture of atherosclerotic lesions or invasive techniques, such as angioplasty. Such exposure of these cells to shear stress may elicit an acute response to fluid flow. Cells that are permanently exposed to higher levels of shear stress, like endothelial cells, may downregulate their responses to adapt to steady-state conditions. In contrast, cells like VSMCs, which are exposed to such levels of shear stress only under extreme conditions, may respond quickly, therefore expressing higher levels of TRPM7 and other molecules that may be involved in the immediate response. In response to TRPM7 activation, intracellular calcium increases. Intracellular calcium is known to induce smooth muscle proliferation at lower levels and cell death at high levels.
Acknowledgments
The Howard Hughes Medical Institute supported this work. We thank members of the Clapham laboratory for experimental support and insightful discussions and Drs Andrew Scharenberg, T. Lambers, and Rene Bindels for constructs.
Both authors contributed equally to this study.
References
Clapham DE. TRP channels as cellular sensors. Nature. 2003; 426: 517eC524.
Yamaguchi H, Matsushita M, Nairn AC, Kuriyan J. Crystal structure of the atypical protein kinase domain of a TRP channel with phosphotransferase activity. Mol Cell. 2001; 7: 1047eC1057.
Ryazanova LV, Dorovkov MV, Ansari A, Ryazanov AG. Characterization of the protein kinase activity of TRPM7/ChaK1, a protein kinase fused to the transient receptor potential ion channel. J Biol Chem. 2004; 279: 3708eC3716.
Runnels LW, Yue L, Clapham DE. The TRPM7 channel is inactivated by PIP(2) hydrolysis. Nat Cell Biol. 2002; 4: 329eC336.
Takezawa R, Schmitz C, Demeuse P, Scharenberg AM, Penner R, Fleig A. Receptor-mediated regulation of the TRPM7 channel through its endogenous protein kinase domain. Proc Natl Acad Sci U S A. 2004; 101: 6009eC6014.
Aarts M, Iihara K, Wei WL, Xiong ZG, Arundine M, Cerwinski W, MacDonald JF, Tymianski M. A key role for TRPM7 channels in anoxic neuronal death. Cell. 2003; 115: 863eC877.
Schmitz C, Perraud AL, Johnson CO, Inabe K, Smith MK, Penner R, Kurosaki T, Fleig A, Scharenberg AM. Regulation of vertebrate cellular Mg2+ homeostasis by TRPM7. Cell. 2003; 114: 191eC200.
Alenghat FJ, Nauli SM, Kolb R, Zhou J, Ingber DE. Global cytoskeletal control of mechanotransduction in kidney epithelial cells. Exp Cell Res. 2004; 301: 23eC30.
Resnick N, Yahav H, Shay-Salit A, Shushy M, Schubert S, Zilberman LC, Wofovitz E. Fluid shear stress and the vascular endothelium: for better and for worse. Prog Biophys Mol Biol. 2003; 81: 177eC199.
Hoger JH, Ilyin VI, Forsyth S, Hoger A. Shear stress regulates the endothelial Kir2.1 ion channel. Proc Natl Acad Sci U S A. 2002; 99: 7780eC7785.
Nauli SM, Alenghat FJ, Luo Y, Williams E, Vassilev P, Li X, Elia AE, Lu W, Brown EM, Quinn SJ, Ingber DE, Zhou J. Polycystins 1 and 2 mediate mechanosensation in the primary cilium of kidney cells. Nat Genet. 2003; 33: 129eC137.
Traub O, Ishida T, Ishida M, Tupper JC, Berk BC. Shear stress-mediated extracellular signal-regulated kinase activation is regulated by sodium in endothelial cells. Potential role for a voltage-dependent sodium channel. J Biol Chem. 1999; 274: 20144eC20150.
Barakat AI, Leaver EV, Pappone PA, Davies PF. A flow-activated chloride-selective membrane current in vascular endothelial cells. Circ Res. 1999; 85: 820eC828.
Byron KL, Taylor CW. Spontaneous Ca2+ spiking in a vascular smooth muscle cell line is independent of the release of intracellular Ca2+ stores. J Biol Chem. 1993; 268: 6945eC6952.
Bezzerides VJ, Ramsey IS, Kotecha S, Greka A, Clapham DE. Rapid vesicular translocation and insertion of TRP channels. Nat Cell Biol. 2004; 6: 709eC720.
Street RL WG, Vennard JK. Elementary Fluid Mechanics. 7th ed. New York: John Wiley and Sons; 1996.
Runnels LW, Yue L, Clapham DE. TRP-PLIK, a bifunctional protein with kinase and ion channel activities. Science. 2001; 291: 1043eC1047.
Nadler MJ, Hermosura MC, Inabe K, Perraud AL, Zhu Q, Stokes AJ, Kurosaki T, Kinet JP, Penner R, Scharenberg AM, Fleig A. LTRPC7 is a Mg.ATP-regulated divalent cation channel required for cell viability. Nature. 2001; 411: 590eC595.
Steyer JA, Almers W. A real-time view of life within 100 nm of the plasma membrane. Nat Rev Mol Cell Biol. 2001; 2: 268eC275.
Prakriya M, Lewis RS. Separation and characterization of currents through store-operated CRAC channels and Mg2+-inhibited cation (MIC) channels. J Gen Physiol. 2002; 119: 487eC507.
He Y, Yao G, Savoia C, Touyz RM. Transient receptor potential melastatin 7 ion channels regulate magnesium homeostasis in vascular smooth muscle cells: role of angiotensin II. Circ Res. 2005; 96: 207eC215.
Galbraith CG, Skalak R, Chien S. Shear stress induces spatial reorganization of the endothelial cell cytoskeleton. Cell Motil Cytoskeleton. 1998; 40: 317eC330.
Wojciak-Stothard B, Ridley AJ. Shear stress-induced endothelial cell polarization is mediated by Rho and Rac but not Cdc42 or PI 3-kinases. J Cell Biol. 2003; 161: 429eC439.
Chiu JJ, Lee PL, Chang SF, Chen LJ, Lee CI, Lin KM, Usami S, Chien S. Shear stress regulates gene expression in vascular endothelial cells in response to tumor necrosis factor-alpha: a study of the transcription profile with complementary DNA microarray. J Biomed Sci. 2005; 12: 481eC502.
Miyagi M, Miwa Y, Takahashi-Yanaga F, Morimoto S, Sasaguri T. Activator protein-1 mediates shear stress-induced prostaglandin d synthase gene expression in vascular endothelial cells. Arterioscler Thromb Vasc Biol. 2005; 25: 970eC975.
Schmoranzer J, Goulian M, Axelrod D, Simon SM. Imaging constitutive exocytosis with total internal reflection fluorescence microscopy. J Cell Biol. 2000; 149: 23eC32.
Toomre D, Steyer JA, Keller P, Almers W, Simons K. Fusion of constitutive membrane traffic with the cell surface observed by evanescent wave microscopy. J Cell Biol. 2000; 149: 33eC40.(Elena Oancea, Joshua T. W)