Research ArticlePhysiology

A PLCγ1-Dependent, Force-Sensitive Signaling Network in the Myogenic Constriction of Cerebral Arteries

See allHide authors and affiliations

Science Signaling  27 May 2014:
Vol. 7, Issue 327, pp. ra49
DOI: 10.1126/scisignal.2004732

Abstract

Maintaining constant blood flow in the face of fluctuations in blood pressure is a critical autoregulatory feature of cerebral arteries. An increase in pressure within the artery lumen causes the vessel to constrict through depolarization and contraction of the encircling smooth muscle cells. This pressure-sensing mechanism involves activation of two types of transient receptor potential (TRP) channels: TRPC6 and TRPM4. We provide evidence that the activation of the γ1 isoform of phospholipase C (PLCγ1) is critical for pressure sensing in cerebral arteries. Inositol 1,4,5-trisphosphate (IP3), generated by PLCγ1 in response to pressure, sensitized IP3 receptors (IP3Rs) to Ca2+ influx mediated by the mechanosensitive TRPC6 channel, synergistically increasing IP3R-mediated Ca2+ release to activate TRPM4 currents, leading to smooth muscle depolarization and constriction of isolated cerebral arteries. Proximity ligation assays demonstrated colocalization of PLCγ1 and TRPC6 with TRPM4, suggesting the presence of a force-sensitive, local signaling network comprising PLCγ1, TRPC6, TRPM4, and IP3Rs. Src tyrosine kinase activity was necessary for stretch-induced TRPM4 activation and myogenic constriction, consistent with the ability of Src to activate PLCγ isoforms. We conclude that contraction of cerebral artery smooth muscle cells requires the integration of pressure-sensing signaling pathways and their convergence on IP3Rs, which mediate localized Ca2+-dependent depolarization through the activation of TRPM4.

INTRODUCTION

Localized increases in intraluminal pressure provoke vasoconstriction of small arteries and arterioles (1). This important part of the autoregulatory mechanism, first described by Bayliss and termed the vascular myogenic response (2), allows regional blood flow to remain essentially constant during transient changes in perfusion pressure (3, 4). Myogenic vasoconstriction results from pressure-induced depolarization of the vascular smooth muscle cell plasma membrane (5), which causes Ca2+ influx through voltage-dependent Ca2+ channels (VDCCs) (6), as well as increases in the Ca2+ sensitivity of the contractile apparatus (79). Longitudinal stretch depolarizes the plasma membrane of isolated arterial myocytes (10), demonstrating that these cells can directly detect mechanical strain within the vascular wall resulting from pressure increases. However, the underlying molecular mechanisms responsible for sensing stretch and eliciting membrane depolarization remain elusive.

In the simplest theoretical scheme to account for pressure-induced depolarization, stretch of the plasma membrane alters the activity of inherently mechanosensitive ion channels. In support of this possibility, various studies have implicated several members of the transient receptor potential (TRP) cation channel superfamily, including TRPP2 (11), TRPV2 (12), TRPC6 (13), and TRPM4 (14, 15), as well as the Ca2+-activated Cl channel TMEM16A (16, 17), in pressure-induced membrane depolarization. TRPC6 (13) and TRPM4 channels appear to be indispensable for pressure-induced cerebral artery constriction (14, 18). TRPC6 is a Ca2+-permeable, nonselective cation channel that is directly activated by the second messenger diacylglycerol (DAG) (19). Some studies have reported that TRPC6 channels are inherently mechanosensitive (20), whereas others have suggested that DAG generated by the activity of phospholipase C (PLC) is involved in mediating the response to stretch (21). Unlike TRPC6, TRPM4 is impermeant to Ca2+, selective for monovalent cations, and activated by increased concentrations of intracellular Ca2+ (22, 23). In vascular smooth muscle cells, TRPM4 activity is stimulated by Ca2+ released from the sarcoplasmic reticulum (SR) through inositol 1,4,5-trisphosphate (IP3) receptors (24, 25). Despite a report that TRPM4 channels are inherently mechanosensitive based on data obtained from heterologous expression systems (26), the mechanosensitivity of TRPM4 has not been clearly established.

Although TRPC6 and TRPM4 channels have been implicated in the myogenic response, their position in the pressure-responsive pathway and the mechanism by which their activities are integrated and regulated to achieve a graded response remain unclear. However, the activation of TRPC6 and TRPM4 by DAG and IP3, respectively, which are generated by cleavage of the membrane phospholipid phosphatidylinositol 4,5-bisphosphate [PtdIns(4,5)P2] by PLC, suggests that PLC may be centrally involved in stretch-induced activation of both TRPM4 and TRPC6. Because PLC isoforms lack transmembrane domains and thus are unlikely to respond directly to mechanical stimuli, the nature of the stress-sensing entity remains an open question. Several pathways can sense and signal stretch, including actin polymerization (9, 27, 28) and integrin-mediated (29) and Gq protein–coupled receptor–mediated signaling (21). It is posited that both PLC and TRPC6 play a role in this latter pathway, but details of this signaling cascade remain incompletely characterized and certain key features of the model rely largely on data obtained from heterologous expression systems.

Here, we sought to determine how second messengers generated by PLC and interactions between TRPM4 and TRPC6 channels are linked to bring about pressure-induced membrane depolarization. We showed that generation of IP3 by the γ1 isoform of PLC (PLCγ1) was essential for pressure-induced TRPM4 channel activity, smooth muscle membrane depolarization, and generation of myogenic tone. We also showed that PLCγ1 activity alone was insufficient to stimulate TRPM4 currents in response to stretch and that Ca2+ influx through TRPC6 was also required, implying that TRPC6 transduces a mechanical stimulus to TRPM4. Finally, we demonstrated that TRPC6, but not TRPM4, is directly mechanosensitive and that the stress-sensing mechanism upstream of PLCγ1 depended on the tyrosine kinase Src. Together, these findings describe a mechanism in which converging signaling pathways integrate the sensation of stretch in vascular smooth muscle cells to mediate pressure-induced membrane depolarization and myogenic vasoconstriction.

RESULTS

TRPM4 is not inherently mechanosensitive

Previous studies have indicated that TRPC6 and TRPM4 are critically involved in pressure-induced cerebral arterial myocyte membrane depolarization (1315, 18), suggesting that these channels can be activated by mechanical stretch. We examined the mechanosensitivity of each channel by measuring changes in single-channel activity in response to membrane stretch induced by applying negative pressure (−20 mmHg) through the recording electrode in patch-clamp experiments using human embryonic kidney (HEK) cells expressing recombinant yellow fluorescent protein (YFP)–tagged TRPC6 or green fluorescent protein (GFP)–tagged TRPM4 (30). In agreement with previous studies (20, 21), the total open probability of TRPC6 channels increased nearly threefold in response to stretch of the plasma membrane (Fig. 1A). Current-voltage relationships recorded under both pressure conditions were linear with a slope conductance consistent with the previously reported unitary conductance of TRPC6 [~35 pS (19)]. Basal channel activity recorded from HEK cells expressing TRPM4 channels did not change when negative pressure was applied (Fig. 1B). The slope conductance of these currents was ~25 pS, similar to the reported unitary conductance of TRPM4 channels (23). These data indicate that heterologously expressed TRPC6 channels, but not TRPM4 channels, are mechanosensitive, under the conditions used for this study.

Fig. 1 TRPM4 is not intrinsically mechanosensitive, and requires PLC and TRPC6 activity for cell swelling–induced activity.

(A) Left: Representative on-cell patch-clamp recording of changes in single-channel activity in a HEK 293 cell transfected with TRPC6-YFP as negative pressure (−20 mmHg) was applied to the patch pipette. Middle: Single-channel current-to-voltage (I/V) amplitude relationship plotted for currents recorded with and without negative pressure (n = 4 to 7 cells in three experiments). Right: Summary of total open probability (NPo) at the indicated intrapipette pressures at a holding potential of +70 mV (n = 5 cells in three transfection experiments; *P ≤ 0.05). (B) Left: Representative on-cell patch-clamp recording of changes in single-channel activity in a HEK 293 cell transfected with TRPM4-GFP as negative pressure (−20 mmHg) was applied to the patch pipette. Middle: I/V relationship at the indicated intrapipette pressures (n = 4 to 5 cells in three transfection experiments). Right: Summary NPo at the indicated intrapipette pressures at a holding potential of +70 mV (n = 5 cells in three transfection experiments; *P ≤ 0.05). (C) Representative recording and summary of TICC activity after 80 mosmol/liter osmotic challenge (n = 5 cells from three animals; *P ≤ 0.05 compared to control, #P ≤ 0.05 compared to hypotonic). (D) Representative recording and summary of the effects of the PLC inhibitor U73122 on swelling-induced TICC activity (n = 5 cells from three animals; *P ≤ 0.05 compared to control).

Phospholipase activity is necessary for stretch-induced TRPM4 activity in arterial myocytes

In native, freshly isolated cerebral artery smooth muscle cells, TRPM4 activity can be continuously recorded as transient inward cation currents (TICCs) (25). TICC activity increased after membrane stretch associated with cell swelling induced by an 80 mosmol/liter hyposmotic challenge (Fig. 1C). We also examined the mechanosensitivity of TRPM4 in these cells using voltage-clamp in the amphotericin B–perforated configuration, and recorded TICCs when defined amounts of suction were applied to the plasma membrane by the recording electrode. TICC activity increased in proportion to the amount of applied negative pressure (fig. S1A). Activation was maximal at an applied intrapipette pressure of −20 mmHg, and the half-maximal pressure of activation (P50) was −12.8 mmHg (fig. S1, A and B). Stretch-activated increases in TICC activity returned to baseline after release of intrapipette pressure, indicating that current activation did not result from permanent damage to the plasma membrane (fig. S1, A and C). Basal and stretch-induced TICC activity was attenuated by the selective TRPM4 inhibitor 9-phenanthrol (15, 31, 32) (fig. S1D), demonstrating that currents originated from TRPM4 channels. These findings indicate that TRPM4 is activated by stretch of the plasma membrane in native cerebral artery smooth muscle cells, consistent with the channel’s role in myogenic vasoconstriction.

TRPM4 lacks inherent mechanosensitivity, but endogenous TRPM4 channels are activated when stretch is applied to the plasma membrane of arterial myocytes, suggesting that the channel is indirectly activated by force-sensitive signaling pathways in these cells. TRPM4 activity requires high concentrations (~1 to 10 μM) of intracellular Ca2+ (22, 23), and in native smooth muscle cells, the channel is stimulated by Ca2+ released from proximal IP3 receptors (IP3Rs) (24, 25). IP3 is generated by PLC-mediated hydrolysis of PtdIns(4,5)P2, suggesting that tonic PLC activity is necessary for basal TICC activity. Consistent with this possibility, swelling-induced TICC activity was lost in the presence of U73122 (Fig. 1D). Furthermore, basal TRPM4 activity was eliminated by the PLC inhibitor U73122 and was restored by the membrane-permeant IP3 analog Bt-IP3 (fig. S1E). TICC activity was insensitive to U73343, an inactive analog of U73122 (fig. S2A). TICC activity stimulated by applying negative pressure to the plasma membrane was also attenuated after administration of U73122 (fig. S1F). Control experiments demonstrated that the frequency and amplitude of spontaneous transient outward currents (STOCs), activated by Ca2+ released from the SR through ryanodine receptors (33), were not altered by U73122 (fig. S2B), suggesting that the compound did not alter SR Ca2+ stores under these conditions. These findings indicate that tonic generation of IP3 by PLC is necessary for basal and stretch-induced TICC activity in smooth muscle cells.

We assessed the consequences of disrupted stretch-induced, PLC-dependent activation of TRPM4 activity for myogenic constriction using isolated cerebral resistance arteries. In agreement with earlier findings (34), arteries pretreated with U73122 exhibited diminished pressure-dependent vasoconstriction compared with vessels treated with vehicle controls (fig. S3A). U73122 pretreatment also decreased vasoconstriction in response to the purinergic receptor vasoconstrictor agonist uridine 5′-triphosphate (UTP) (fig. S3B), but constriction in response to increased concentrations of extracellular KCl did not differ between groups (fig. S3C). These findings indicate that general PLC activity is necessary for both pressure- and agonist-induced vasoconstriction, and its blockade does not directly influence constriction in response to membrane depolarization.

PLCγ1 regulates TRPM4 activity in arterial myocytes

Several PLC isoforms are present in vascular smooth muscle cells (35). We detected PLCγ1, PLCγ2, and PLCβ2 in native cerebral artery smooth muscle cells using reverse transcription polymerase chain reaction (RT-PCR) and immunocytochemistry (Fig. 2, A and B). PLCγ1 and PLCγ2 immunolabeling appeared at or near the plasma membrane, whereas PLCβ2 immunolabeling was detected as discrete puncta located primarily within the cytosol (Fig. 2B). TRPM4 immunolabeling was also concentrated at or near the plasma membrane (Fig. 2B). We used an in situ proximity ligation assay (36) to identify PLC isoforms that are closely associated with TRPM4 in cerebral myocytes. This technique produces positive immunolabeling only if the two targeted proteins are localized within 40 nm of each other. Dual immunolabeling for PLCγ1 and TRPM4 produced significantly higher numbers of puncta per cell near the plasma membrane than were detected with dual labeling for TRPM4 and PLCγ2 or PLCβ2 (Fig. 2C). These findings suggest that PLCγ1 and TRPM4 are within 40 nm of each other near the plasma membrane in contractile smooth muscle cells, indicating that PLCγ1 could be selectively involved in the regulation of TRPM4 activity.

Fig. 2 PLCγ1 colocalizes with TRPM4 in cerebral artery smooth muscle cells.

(A) PLCγ1, PLCγ2, and PLCβ2 detected by RT-PCR in RNA extracted from whole cerebral arteries. Representative of data obtained from at least three animals. NT, no template control. (B) Immunofluorescent images of labeling for PLCγ1, PLCγ2, or PLCβ2 (green) and TRPM4 (red) in freshly isolated cerebral artery smooth muscle cells. Scale bar, 10 μm. Representative of 15 cells from three animals. (C) Immunofluorescent images and summary of proximity ligation assay (PLA) labeling for PLCγ1:TRPM4 (top), PLCγ2:TRPM4 (middle), and PLCβ2:TRPM4 (bottom). Autofluorescence of cytosol is green, and the occurrences of protein colocalization are visualized as bright fluorescent puncta (red, insert). Representative of data obtained from at least three animals (n = 15 cells from three animals; *P ≤ 0.05 compared to PLCγ1:TRPM4). Scale bars, 10 and 1 μm (insert). (D) Representative recordings and summary of TICC activity in cells isolated from arteries treated with control (top) and PLCγ1 siRNA (bottom). Bt-IP3 was administered to cells isolated from arteries treated with PLCγ1 siRNA as indicated (n = 6 to 11 cells from five animals; *P ≤ 0.05 compared to control siRNA, #P ≤ 0.05 compared to PLCγ1 siRNA baseline).

Small interfering RNA (siRNA) was used to test the hypothesis that PLCγ1 regulates TRPM4 activity in native cerebral artery smooth muscle cells. siRNA targeting PLCγ1 or control (scrambled) siRNA was introduced into the arterial myocytes of intact cerebral arteries using a previously described reversible permeabilization technique (25) (fig. S4). TICCs were recorded from arterial myocytes, enzymatically isolated from vessels treated with PLCγ1 or control (nonsilencing) siRNA, whereas stretch was applied to the plasma membrane. Cells isolated from PLCγ1 siRNA–treated vessels exhibited an about 73% decrease in stretch-induced TICC activity compared with cells from vessels treated with control siRNA (Fig. 2D). TICC activity in cells isolated from vessels treated with PLCγ1 siRNA was partially restored by administration of Bt-IP3 (Fig. 2D), indicating that diminished IP3 production after PLCγ1 silencing contributes to the loss of TRPM4 activity. These findings demonstrate that the activity of PLCγ1 is necessary for the generation of TRPM4 currents in cerebral myocytes in response to stretch of the plasma membrane.

PLCγ1 is essential for pressure-induced smooth muscle cell depolarization and the development of myogenic tone

Acute knockdown of TRPM4 largely eliminates pressure-induced smooth muscle cell depolarization and myogenic constriction of cerebral arteries (14, 15), and the current findings indicate that stretch-induced TRPM4 activity is dependent on the generation of IP3 by PLCγ1. Therefore, we examined the role of PLCγ1 in pressure-induced smooth muscle membrane depolarization and vasoconstriction using siRNA-mediated silencing in intact cerebral arteries. Microelectrode recordings obtained from pressurized vessels demonstrated that the mean membrane potential of smooth muscle cells in control siRNA–treated arteries was −59.5 ± 6.6 mV at rest and was depolarized to −39.0 ± 3.0 mV by an increase in pressure to 80 mmHg. This response was significantly attenuated in PLCγ1 siRNA–treated arteries, demonstrating a role for PLCγ1 in pressure-induced depolarization of cerebral artery smooth muscle cells (Fig. 3, A and B). Moreover, myogenic tone was essentially abolished in PLCγ1 siRNA–treated arteries (Fig. 3, C and D), demonstrating a requirement for PLCγ1 in the myogenic response to pressure. In contrast, we found that agonist-induced vasoconstriction with UTP did not differ between PLCγ1 siRNA– and control siRNA–treated vessels (Fig. 3E), indicating that PLCγ1 activity is not involved in purinergic receptor-mediated vasoconstriction. Arterial constriction in response to increased extracellular K+ ion concentration did not differ between PLCγ1 siRNA– and control siRNA–treated vessels (Fig. 3F), indicating that absence of PLCγ1 does not directly impair membrane depolarization-dependent vasoconstrictor pathways. These findings show that PLCγ1 is specifically involved in pressure-induced smooth muscle cell depolarization and is required for the development of myogenic tone. Considering the inhibitory effects of U73122 on UTP-mediated vasoconstriction (Fig. 2B), the absence of an effect of PLCγ1 on the contractile response to UTP suggests that agonist-induced vasoconstrictor responses require the activity of a different PLC isoform (for example, PLCβ). The close proximity of PLCγ1 and TRPM4 (Fig. 2C) and our previous work showing that TRPM4 and IP3 receptors are functionally coupled (24) suggest that TRPM4, PLCγ1, and IP3 receptors are part of a force-sensitive signaling network in vascular smooth muscle cells that mediates membrane potential depolarization in response to increases in intraluminal pressure.

Fig. 3 PLCγ1 siRNA attenuates pressure-induced membrane depolarization and vasoconstriction.

(A) Representative recordings of smooth muscle membrane potential in cerebral arteries treated with either control (top) or PLCγ1 (bottom) siRNA at the indicated intraluminal pressures. (B) Summary data demonstrating the effect of PLCγ1 siRNA on cerebral arterial myocyte resting membrane potential (Em) (n = 5 arteries from four animals; *P ≤ 0.05 compared to control siRNA at 20 mmHg, #P ≤ 0.05 compared to control siRNA at 80 mmHg). (C) Recordings demonstrating the effects of increasing intraluminal pressure on luminal diameter from arteries treated with either control (top) or PLCγ1 (bottom) siRNA. (D) Summary data showing the effects of PLCγ1 siRNA treatment on myogenic tone (n = 6 to 7 arteries from four animals; *P ≤ 0.05). (E) Summary data demonstrating the effect of increasing concentrations of UTP on luminal diameter from arteries treated with either control or PLCγ1 siRNA (n = 5 arteries from four animals). (F) Summary data of the vasoconstriction in response to increased extracellular [KCl] from arteries treated with either control or PLCγ1 siRNA (n = 7 arteries from four animals).

Ca2+ influx through TRPC6 is required for pressure-induced Ca2+ waves and activation of TRPM4

Previous studies have shown that knocking down either TRPM4 or TRPC6 largely eliminates (~80 to 90%) the myogenic response in cerebral arteries, consistent with a serial mechanism (13, 14). However, we previously have shown that TRPM4 channels are not responsive to extracellular Ca2+ influx, arguing against a direct link between TRPC6-mediated Ca2+ influx and TRPM4 activation in the myogenic response. In particular, TRPM4 is activated by IP3R-mediated release of Ca2+ from intracellular stores. An alternative mechanism that is compatible with these observations and a serial linkage between TRPC6 and TRPM4 is a Ca2+-induced Ca2+-release mechanism in which Ca2+ influx through stretch-activated TRPC6 channels activates IP3Rs, promoting the release of Ca2+ from intracellular stores that, in turn, stimulates TRPM4 activity. This mechanism predicts that stretch induces Ca2+ influx and release from intracellular stores and that both events are required to activate TRPM4. We examined the role of TRPC6-mediated Ca2+ influx in pressure-induced Ca2+ waves in arteries and stretch-activated TRPM4 currents in freshly isolated myocytes. In intact pressurized arteries that were continuously exposed to the L-type Ca2+ channel blocker nimodipine and the ryanodine receptor blocker tetracaine to isolate IP3R-mediated events, the percentage of vascular smooth muscle cells exhibiting spontaneous propagating Ca2+ waves increased in a pressure-dependent manner (Fig. 4A). Preincubation with a TRPC6 antibody targeting an extracellular epitope eliminated these intraluminal pressure-induced Ca2+ waves (Fig. 4, A and B) and blocked currents in TRPC6-expressing HEK 293 cells (fig. S6). In vessels preincubated with the inhibitory antibody, administration of the purinergic receptor agonist UTP restored Ca2+ waves (Fig. 4, A and B), possibly through Gq protein–coupled receptor–dependent activation of TRPC3 (37). These results suggest that increases in intraluminal pressure and stretch of the plasma membrane elicit Ca2+ release through IP3Rs and that this response is governed by activation of plasma membrane TRPC6 channels.

Fig. 4 Ca2+ influx through TRPC6 is required for stretch-induced activation of TRPM4 in arterial myocytes.

(A) Representative images and recordings demonstrating the effects of increasing intraluminal pressure from 30 to 70 mmHg on Ca2+ waves in fluo-4 AM–loaded arteries preincubated with TRPC6 inhibitory antibody. Images are composed of compressed image stacks illustrating the number of cells (grayscale) and number of cells with Ca2+ waves (color topography). Scale bars, 5 μm. Spontaneous Ca2+ signals were detected as fractional fluorescence (F/F0) from 15 regions of interest (ROIs) marked as colored squares. Summary of the effects of the TRPC6 inhibitory antibody on pressure- or UTP-induced Ca2+ waves of intraluminal pressures (n = 3 to 4 vessels from three animals; *P ≤ 0.05 compared to control at 30 mmHg, #P ≤ 0.05 compared to control at 70 mmHg, and P ≤ 0.05 compared to anti-TRPC6 at 70 mmHg). (B) Representative recordings and summary of stretch-induced TICC activity in the presence (top) and absence (bottom) of extracellular Ca2+ (n = 7 cells from four animals; *P ≤ 0.05 compared to control at 0 mmHg). (C) Representative recordings and summary data of stretch-dependent TICC activity in the presence of the SERCA pump blocker cyclopiazonic acid (CPA) (n = 5 cells from three animals; *P ≤ 0.05 compared to control at 0 mmHg). (D) Immunofluorescent images and summary of PLA labeling for TRPM4:TRPC6 and TRPM4:BKCa (n = 15 cells from three animals; *P ≤ 0.05 compared to TRPM4:TRPC6). Autofluorescence of cytosol is green, and the areas of protein colocalization are visualized as bright fluorescent puncta (red, insert). Scale bars, 10 or 1 μm (insert). (E) Representative recordings and summary of stretch-dependent TICC activity from myocytes preincubated with GsMTx-4 (top), inhibitory TRPC6 antibody (middle), or the inhibitory TRPC6 antibody and antigenic peptide (bottom). Immunofluorescent images (right) of TRPC6 labeling in the absence (top) and presence of antigenic peptide (bottom). DAPI (4′,6-diamidino-2-phenylindole)–stained cell nuclei are blue. Scale bars, 10 μm (n = 15 cells from three animals). Summary of the effects of GsMTx-4, TRPC6 inhibitory antibody, and antigenic peptide on stretch-dependent TICC activity (n = 5 cells from three animals; *P ≤ 0.05 compared to control at 0 mmHg). (F) Representative recording and summary of swelling-induced TICC activity in myocytes preincubated with an inhibitory TRPC6 antibody (n = 4 cells from three animals).

To determine whether Ca2+ influx through TRPC6 was necessary for the pressure-dependent, IP3R-mediated activation of TRPM4, we examined the role of TRPC6 in stretch-induced TICC activity. Internal SR Ca2+ stores were maintained after acute removal of extracellular Ca2+, accomplished by immediate (<1 to 2 min) exposure to zero extracellular Ca2+ (fig. S5). Under these conditions, stimulated Ca2+ release events from internal stores are maintained (38). Removal of extracellular Ca2+ prevented stretch-induced increases in TRPM4 activity (Fig. 4B). In agreement with our previous findings (24, 25), depletion of SR Ca2+ stores with the sarco/endoplasmic reticulum Ca2+–adenosine triphosphatase (ATPase) (SERCA) inhibitor cyclopiazonic acid essentially abolished both basal and stretch-induced TICC activity (Fig. 4C). These data indicate that Ca2+ influx and intact SR Ca2+ stores are necessary for stretch-induced increases in TRPM4 activity in native smooth muscle cells.

We next sought to establish that TRPC6, which is Ca2+-permeable, activated by plasma membrane stretch (13, 20, 21) (see also Fig. 1A), and necessary for myogenic vasoconstriction of cerebral arteries (13), is the mechanosensitive ion channel involved in this response. Using in situ proximity ligation assay, we found that TRPC6 channels, but not large-conductance Ca2+-activated K+ (BKCa) channels, are within 40 nm of TRPM4 channels (Fig. 4D). These data indicate that TRPC6 channels are positioned to provide the Ca2+ influx necessary for stretch-induced activation of TRPM4. Stretch-induced increases in TRPM4 activity were attenuated by GsMTx-4, a peptide derived from the venom of the Chilean rose hair tarantula that inhibits mechanosensitive ion channels, including TRPC6 (20) (Fig. 4E). GsMTX-4 blocked recombinant TRPC6 channels expressed in HEK 293 cells, but had no direct effects on heterologously expressed TRPM4 channels (fig. S6). These results indicate a functional linkage between TRPC6 and TRPM4. Stretch-activated TRPM4 currents in cerebral artery myocytes were also attenuated by an antibody that binds to an extracellular epitope of TRPC6 (Fig. 4E, middle), providing additional support for functional communication between TRPC6 and TRPM4. The anti-TRPC6 antibody blocked currents recorded from TRPC6-expressing HEK 293 cells, but not those recorded from HEK 293 cells transfected with TRPM4 (fig. S6), and did not affect currents recorded from cerebral artery myocytes when coadministered with its antigenic peptide (Fig. 4E), indicating that these effects were specific. Consistent with these observations and the results of in situ proximity ligation assay experiments, the anti-TRPC6 antibody immunolabeled the plasma membrane of native cerebral artery smooth muscle cells but failed to immunolabel these cells when the antibody was preincubated with the antigenic peptide (Fig. 4E, right). In addition, swelling-induced TICC activity was absent in cells that were preincubated with the TRPC6 inhibitory antibody (Fig. 4F). Collectively, these data indicate that Ca2+ influx through TRPC6 is necessary for stretch-induced IP3R Ca2+-mediated activation of TRPM4 in contractile cerebral artery smooth muscle cells.

AT1 receptors and Src family tyrosine kinase activity contribute to stretch-induced TRPM4 activity in arterial myocytes

The dependence of the myogenic response on PLCγ1 and the presumed absence of PLC mechanosensitivity prompted us to address the possible upstream involvement of the Src family of nonreceptor tyrosine kinases, which can signal through PLCγ isoforms (35) and rapidly activate in response to mechanical stress (39). Consistent with a role for Src family tyrosine kinases in this pathway, we found that two structurally unrelated Src inhibitors, PP2 (Fig. 5A) or SKI-1 (fig. S7A), inhibited stretch-induced TRPM4-mediated currents. Administration of the Src kinase inhibitors PP2 (Fig. 5B) and SKI-1 (fig. S7B) nearly abolished predeveloped myogenic tone in isolated cerebral arteries. PP2 did not affect KCl-induced constriction (Fig. 5C). These data indicate that Src tyrosine kinase activity is required for the development of myogenic tone. Previous work indicated that stimulation of angiotensin II (Ang II) receptor type 1 (AT1R) initiates Src tyrosine kinase–dependent phosphorylation of PLCγ1 in smooth muscle cells (40). In agreement with these results, administration of the AT1R inhibitor losartan prevented stretch-induced increases in TRPM4 activity (Fig. 5D), suggesting a possible upstream role for this Gq/11-coupled receptor in the pressure-response pathway.

Fig. 5 Src family tyrosine kinase activity is necessary for stretch-induced TRPM4 activity in arterial myocytes.

(A) Representative recording and summary of the effects of the Src family kinase inhibitor PP2 on stretch-induced TICC activity [n = 5 cells from three animals; *P ≤ 0.05 compared to control (0 mmHg), #P ≤ 0.05 compared to control at −20 mmHg]. (B) Representative recoding of the effects of repeated administration of PP2 on the luminal diameter of a cerebral artery with predeveloped myogenic tone (80 mmHg) (n = 3 to 5 arteries from four animals; *P ≤ 0.05 compared to control). (C) Summary data indicating the effects of PP2 administration on vasoconstriction in response to increased extracellular [KCl] (n = 3 vessels from three animals). (D) Representative recording and summary of the effects of losartan on stretch-induced TICC activity (n = 5 cells from four animals; *P ≤ 0.05 compared to control at 0 mmHg, #P ≤ 0.05 compared to control at –20 mmHg).

The findings presented here support a signaling pathway (Fig. 6) in which pressure induces PLCγ1 activity through Src tyrosine kinase, likely with the involvement of the upstream Gq/11-coupled receptor AT1R, leading to the generation of IP3 and DAG. PLCγ1-derived IP3 sensitizes IP3Rs located on the SR to the influx of Ca2+ through DAG-stimulated (fig. S8) and/or pressure-activated TRPC6 channels to generate localized Ca2+ nanodomains that activate TRPM4 channels located on the plasma membrane. The resulting influx of Na+ depolarizes the membrane to initiate Ca2+ influx through VDCCs and causes myocyte contraction.

Fig. 6 Proposed mechanism of stretch-dependent activation of TRPM4.

Convergence of stretch-sensing pathways in cerebral artery myocytes includes (i) the activation of AT1R, Src tyrosine kinase, and PLCγ1 in the generation of IP3, and (ii) Ca2+ influx through TRPC6 channels onto IP3Rs to activated TRPM4 channels that are responsible for pressure-induced depolarization of the plasma membrane. PM, plasma membrane; ΔVm, change in membrane potential.

DISCUSSION

Contraction of arterial myocytes in response to increased intraluminal pressure constitutes a basic intrinsic autoregulatory blood flow control mechanism. The ability of vascular smooth muscle cells to sense and appropriately respond to pressure is central to this critical response, but the underlying mechanisms are poorly understood. The findings of the current study incorporate previous studies implicating PLC (34, 41, 42), PKC (protein kinase C) (30, 43, 44), TRPC6 (13), and TRPM4 (14, 15, 18) as major players in this process and describe a signaling pathway responsible for the sensation of stretch and initiation of myogenic constriction. According to this model, stretch of the plasma membrane has a twofold effect: (i) AT1R- and Src tyrosine kinase–dependent activation of PLCγ1 and generation of IP3, and (ii) stimulation of Ca2+ influx through TRPC6. IP3 acts primarily by enhancing the sensitivity of IP3Rs to local cytosolic Ca2+ (45); therefore, IP3Rs play a key role in integrating these two signaling pathways. IP3-mediated sensitization of IP3Rs acts synergistically with TRPC6-mediated Ca2+ influx to promote Ca2+ release from IP3Rs through a Ca2+-induced Ca2+-release mechanism, resulting in activation of proximal TRPM4 channels in response to increases in luminal pressure. Our previous observation that the TRPM4 inhibitor 9-phenanthrol relaxes vessels with tone independent of a change in pressure (15) suggests that this mechanism also operates under steady-state conditions, although it is possible that other pathways contribute.

PLC activity is necessary for the development of myogenic tone (34, 41, 42), such that pharmacological inhibition leads to hyperpolarization and disruption of sustained pressure-induced constriction of arteries (34, 41). Intracellular concentrations of endogenous IP3 in smooth muscle cells are higher in myocytes isolated from vessels exposed to higher intraluminal pressures (46), suggesting a positive relationship between PLC activity and membrane stretch. In addition, Mufti et al. reported an increase in IP3R-mediated Ca2+-release events in cerebral arteries in response to higher intraluminal pressure (47). These data suggest that pressure can stimulate PLC activity, thereby generating IP3 and the subsequent release of Ca2+ through IP3Rs. Yet, PLC is a common signaling enzyme with multiple isoforms that contribute to different signaling pathways. PLCγ1 is located proximal to TRPM4 and is a key element in stretch-induced, but not agonist-dependent, TRPM4 activation, suggesting that isoform specificity and microdomain localization can prevent crosstalk between different PLC subtypes.

PLCγ isoforms are downstream targets of Src tyrosine kinase activity (35), and previous studies have indicated that PLCγ1 can be activated by Src tyrosine kinase–dependent phosphorylation after Ang II–induced AT1R stimulation in smooth muscle cells (40). These studies have advanced the concept that Gq protein–coupled receptors, as direct mechanosensors (21, 4850), contribute to myogenic vasoconstriction by activating depolarizing cation currents (21). Zou et al. demonstrated that agonist-independent, stretch-induced hypertrophy of cardiomyocytes is blocked by inhibition of AT1Rs but remains intact in cells isolated from angiotensinogen knockout mice (48). Furthermore, Mederos y Schnitzler and co-workers have shown that losartan diminishes swelling-induced increases in intracellular Ca2+ in primary renal arterial smooth muscle cells and partially reverses the myogenic tone of isolated rat cerebral arteries (21). These authors further reported that all Gq protein–coupled receptors tested in an exogenous expression system (HEK cells) confer responsiveness to osmotic swelling when coexpressed with TRPC6, and suggested that mechanosensitivity may be a general feature of Gq protein–coupled receptors (21). Although our data also support a role for AT1R in the pressure-dependent activation of TRPC6 (and TRPM4) in cerebral arteries, we found that these effects did not extend to the P2Y receptor, another class of Gq protein–coupled receptor. Specifically, whereas general inhibition of PLC eliminated the tone induced by the P2Y receptor agonist UTP (fig. S3), knockdown of PLCγ1 in vessels did not alter UTP-induced tone (Fig. 3E). These data suggest that P2Y receptors are not involved in the activation of PLCγ1, and thus, activation of this pathway does not appear to be a general property of Gq protein–coupled receptors, in contrast to the results reported by Mederos y Schnitzler et al. (21). One possible explanation is that different PLC isoform(s) could couple to different Gq protein–coupled receptors, a question that was not specifically addressed in this previous study. It has been suggested that the purinergic Gq protein–coupled receptors P2Y4 or P2Y6, but not AT1Rs, are involved in generating myogenic tone in smaller penetrating parenchymal arterioles (51), implying that different mechanosensitive Gq protein–coupled receptors are involved in myogenic tone in different vascular beds. Further investigation is needed to clarify the details of mechanotransduction mechanisms, particularly PLC isoform dependence, and assess the involvement of other PLC-dependent mechanosensitive pathways, such as those involving integrins (52, 53) or other extracellular matrix proteins.

TRPC6 clearly contributes to the generation of myogenic tone (13), but whether these channels are directly or indirectly mechanosensitive has been a matter of controversy. TRPC6-deficient mice have been reported to exhibit enhanced membrane depolarization, myogenic tone, and increased blood pressure (54), a phenotype that has been attributed to up-regulation of constitutively active, DAG-sensitive TRPC3 channels. In addition, an acute antisense strategy directed against TRPC6 in intact cerebral arteries, which prevents functional compensation by closely related channels, results in attenuation of pressure-dependent depolarization (13). These data implicate TRPC6 as a player in the pressure-sensing mechanism but do not address whether this channel responds directly or indirectly to mechanical stretch. Various groups have characterized mechanosensitive channels in smooth muscle cells (5557), including Park et al., who have identified stretch-activated cation channels in native vascular smooth muscle cells with a unitary conductance of ~30 pS that are stimulated by a DAG analog and require PLC activity (58), properties that are reminiscent of TRPC channels. These observations are consistent with the possibility that DAG generated by pressure-induced increases in PLCγ1 activity stimulate TRPC6 in cerebral artery smooth muscle, but do not preclude direct mechanosensitivity of the channel. In agreement with this possibility, Spassova et al. have reported that stretch-induced activation of recombinant TRPC6 is independent of PLC activity, consistent with the channel exhibiting direct mechanosensitivity (20).

Collectively, our results define a force-sensitive signaling module in arterial smooth muscle cells in which converging mechanosensitive inputs (namely, AT1Rs and TRPC6) are integrated by PLCγ1- and IP3R-mediated Ca2+ release to link membrane stretch with membrane depolarization (TRPM4) to generate myogenic tone. Although our model assumes that TRPC6 is directly activated by stretch, it is capable of accommodating an indirect TRPC6 activation mechanism.

MATERIALS AND METHODS

General electrophysiological recordings

Currents were recorded using an Axopatch 200B amplifier equipped with an Axon CV 203BU headstage (Molecular Devices). Recording electrodes (2 to 3 megohms) were pulled, polished, and coated with wax to reduce capacitance. Currents were filtered at 1 kHz, digitized at 40 kHz, and stored for subsequent analysis. Clampex and Clampfit versions 10.2 (Molecular Devices) were used for data acquisition and analysis, respectively. All experiments were performed at room temperature (22°C).

HEK 293 culture and transfection

HEK 293 cells were cultured in 1× high-glucose Dulbecco’s modified Eagle’s medium (DMEM) (Gibco) supplemented with 10% fetal bovine serum (Gibco) and 0.5% penicillin-streptomycin (Gibco). Cells were incubated at 37°C with 6% CO2 and subcultured when confluent using 0.25% trypsin-EDTA (Gibco), and media were changed every 2 to 3 days. HEK 293 cells were transfected with DNA encoding TRPC6-YFP or TRPM4-GFP using a Gene Pulser Xcell (Bio-Rad, 110-V 25-ms pulse) and transferred to culture flasks. Transfected cells were replated on coverslips 4 hours before patch-clamp experiments.

Electrophysiological recordings from TRPM4- and TRPC6-expressing HEK 293 cells

TRPM4 and TRPC6 single-channel activity was recorded using the cell-attached patch-clamp configuration. The patch pipette solution contained 80 mM Cs-aspartate, 30 mM CsCl, 1.5 mM MgCl2, 10 mM tetraethylammonium (TEA), 60 mM d-mannitol, and 10 mM Hepes at pH 7.2 (CsOH). Bath solution contained 120 mM KCl, 1.5 mM MgCl2, 10 mM Hepes, 50 mM d-mannitol, and 1 mM EGTA at pH 7.4 (KOH). Mechanical stimulation of the cell membrane was accomplished by applying suction (or negative pressure of −20 mmHg) through the recording pipette with a DPM-1B Pneumatic Transducer Tester (Fluke Biomedical).

Conventional whole-cell patch-clamp studies were conducted using nontransfected HEK 293 cells or cells transiently transfected with TRPC6-YFP or TRPM4-GFP. Transfected cells were identified by fluorescence. Cells were pretreated (20 min) with the TRPC6 inhibitor GsMTx-4 (5 μM), a peptide derived from the venom of Grammostola spatulata (Chilean rose hair tarantula), or an anti-TRPC6 inhibitory antibody [1:250, Alomone (ALC-120)]. Macroscopic whole-cell currents from HEK 293 cells transfected with TRPM4-GFP were recorded in external bathing solution containing 146 mM NaCl, 5 mM CaCl2, 10 mM Hepes, and 10 mM glucose (pH 7.4). The bathing solution was supplemented with the potassium channel blocker TEA (10 mM). The pipette solution contained 146 mM CsCl, 1 mM MgCl2, 10 mM Hepes, and 0.1 mM CaCl2. Macroscopic whole-cell currents from HEK 293 cells transfected with TRPC6-YFP were recorded in external bathing solution containing 150 mM NaCl, 0.08 mM KCl, 0.8 mM MgCl2, 5.4 mM CaCl2, 10 mM Hepes, and 10 mM glucose (pH 7.4). The pipette solution contained 130 mM CsOH, 110 mM aspartic acid, 1 mM MgCl2, 10 mM EGTA, and 10 mM Hepes. After baseline currents were recorded, carbachol (10 μM) was administered to induce receptor-mediated TRPC6 currents. For all experiments, cells were initially voltage-clamped at a membrane potential of 0 mV. Currents were measured during voltage ramps between −100 and +100 mV (1.6 s) repeated every 4 s.

Animals

Male Sprague-Dawley rats (250 to 350 g; Harlan) were used for these studies in accordance with protocols approved by the Institutional Animal Care and Use Committees of the Colorado State University. Animals were deeply anesthetized with pentobarbital sodium (50 mg, intraperitoneally) and euthanized by exsanguination. Cerebral and cerebellar arteries were dissected from brains and cleaned of connective tissue in cold Mops-buffered saline composed of 3 mM Mops (pH 7.4), 145 mM NaCl, 5 mM KCl, 1 mM MgSO4, 2.5 mM CaCl2, 1 mM KH2PO4, 0.02 mM EDTA, 2 mM pyruvate, 5 mM glucose, and 1% bovine serum albumin. Arteries were stored in Mops-buffered saline on ice before further manipulation.

Cerebral artery smooth muscle cell isolation

Vessels were washed in magnesium-based physiological saline solution (Mg-PSS) containing 5 mM KCl, 140 mM NaCl, 2 mM MgCl2, 10 mM Hepes, and 10 mM glucose. Arteries were initially digested in papain (0.6 mg/ml) (Worthington) and dithioerythritol (1 mg/ml) in Mg-PSS at 37°C for 15 min, followed by a 15-min incubation at 37°C in type II collagenase (1.0 mg/ml) (Worthington) in Mg-PSS. The digested arteries were washed three times in ice-cold Mg-PSS solution and incubated on ice for 30 min. After this incubation period, vessels were triturated to liberate smooth muscle cells and stored in ice-cold Mg-PSS before use. Smooth muscle cells were studied within 6 hours after isolation.

Electrophysiological recordings from isolated smooth muscle cells

Isolated smooth muscle cells were placed in a recording chamber (Warner Instruments) and allowed to adhere to glass coverslips for 20 min at room temperature. Gigohm seals were obtained in Mg-PSS. Amphotericin B (40 μM) was included in the pipette solution to perforate the membrane. Perforation was deemed acceptable if series resistance was less than 50 megohms. STOC and TICC activities were recorded in normal external bathing solution containing 134 mM NaCl, 6 mM KCl, 1 mM MgCl2, 2 mM CaCl2, 10 mM Hepes, and 10 mM glucose at pH 7.4 (NaOH). The pipette solution contained 110 mM K-aspartate, 1 mM MgCl2, 30 mM KCl, 10 mM NaCl, 10 mM Hepes, and 5 μM EGTA at pH 7.2 (NaOH). Additional external solutions include Ca2+-free external solution containing 134 mM NaCl, 6 mM KCl, 1 mM MgCl2, 0.73 mM CaCl2, 1 mM EGTA, 10 mM Hepes, and 10 mM glucose at pH 7.4 (NaOH). Where specified, a hyperosmotic solution was reached using 100 mM mannitol. A hypotonic challenge of 80 mosmol/liter was achieved by incubating and recording baseline activity in the above hypertonic solution (376 mosmol/liter) followed by exposure to the original recording solution (without mannitol, 296 mosmol/liter). In our recording solutions, the calculated reversal potentials for total monovalent cations were −1.8 and −30.6 mV for monovalent anions (Cl). STOCs were defined as transient events greater than 10 pA (more than one BKCa channel), and the frequency was calculated by dividing the number of events by the time between the first and the last event. Defined amounts of stretch of the plasma membrane were achieved by applying negative pressure through the recording electrode using a DPM-1B Pneumatic Transducer Tester (Fluke Biomedical). Stretch-induced TICC activity was recorded after the administration of 9-phenanthrol (30 μM), U73122 (10 μM), Bt-IP3 (10 μM), Ang II (100 nM), losartan (1 μM), PP2 (300 nM), GsMTx-4 (5 μM), U73343 (10 μM), or Src family kinase inhibitor-1 (SKI-1, 300 nM). TICC activity at −70 mV was quantified as the calculated sum of the open channel probability of multiple open states of 1.75 pA. This value was based on the reported unitary conductance of TRPM4 (25 pS). Channel open probability was calculated using the following equation:NPo=j=1N(tjjT)where tj is the time spent in seconds, with j = 1, 2, … N channels open; N is the max number of channels observed; and T is the duration of measurement.

Isolated vessel experiments

Arteries were harvested and transferred to a vessel chamber (Living Systems Inc.). The proximal end of the vessel was cannulated with a glass micropipette and secured with monofilament thread. Blood was gently rinsed from the lumen, and the distal end of the vessel was cannulated and secured. Arteries were pressurized to 10 mmHg with PSS composed of 119 mM NaCl, 4.7 mM KCl, 1.8 mM CaCl2, 1.2 mM MgSO4, 24 mM NaHCO3, 0.2 mM KH2PO4, 10.6 mM glucose, and 1.1 mM EDTA, and superfused (5 ml/min) with warmed (37°C) PSS aerated with a normoxic gas mixture (21% O2, 6% CO2, balance N2). After a 15-min equilibration period, intraluminal pressure was slowly increased to 80 mmHg, vessels were stretched to remove bends, and pressure was reduced to 20 mmHg for an additional 15-min equilibration period. Inner diameter was continuously monitored using video microscopy and edge detection software (IonOptix). Arteries pressurized to 20 mmHg were exposed to isotonic PSS containing 60 mM KCl to assess viability of the preparation. Arteries were pressurized to 80 mmHg, and stable myogenic tone was allowed to develop before the administration of U73122 (10 μM). To determine maximum (or passive) diameter, vessels were later superfused with Ca2+-free PSS: 119 mM NaCl, 4.7 mM KCl, 1.2 mM MgSO4, 24 mM NaHCO3, 0.2 mM KH2PO4, 10.6 mM glucose, 1.1 mM EDTA, 3 mM EGTA, and 0.01 mM diltiazem.

Ca2+ imaging experiments of isolated vessels

Arteries were harvested and incubated in the dark at room temperature in Mops-buffered saline containing 10 μM fluo-4 AM (Invitrogen) and pluronic acid (2.5 μg/ml) for 1.5 hours. Vessels were cannulated on a glass micropipette, pressurized, and superfused with warmed (37°C) modified PSS (125 mM NaCl, 3 mM KCl, 26 mM NaHCO3, 1.25 mM NaH2PO4, 1 mM MgCl2, 4 mM d-glucose, and 2 mM CaCl2,) and aerated with a gas mixture consisting of 20% O2, 5% CO2, 75% N2 to maintain pH (pH 7.4). IP3R-mediated Ca2+ waves were isolated by continuously exposing vessels to the L-type Ca2+ channel blocker nimodipine (300 nM) and the ryanodine receptor blocker tetracaine (100 μM). Images were acquired at 20 frames per second on a 138 × 138–μm field of view (512 × 512 pixels) using a Solamere confocal scanning unit (QLC 100), an electron-multiplying CCD (charge-coupled device) camera, and a 40× water immersion objective. The fluorophore was excited by illumination with a 488-nm krypton/argon laser, and fluorescence emission was collected above 495 nm. Ca2+ images were acquired using Andor Revolution TL software (Andor Technology) and analyzed with custom software (SparkAn) created by A. D. Bonev (University of Vermont, Burlington, VT).

Reverse transcription polymerase chain reaction

After enzymatic dispersal of rat cerebral arteries, total RNA was extracted (RNeasy Protect Mini Kit, Qiagen Inc.) and first-strand complementary DNA (cDNA) was synthesized using an Omniscript Reverse Transcription Kit (Qiagen Inc.). RT-PCR was performed with the following PLC-specific QuantiTect Primers (Qiagen Inc.) spanning intron/exon boundaries for the following: PLCγ1, QT00184100 (Rn_Plcg1_1_SG); PLCγ2, QT01080338 (Rn_Plcg2_1_SG); and PLCβ2, QT00188062 (Rn_Plcb2_1_SG). PCR products were resolved on 2% agarose gels. These primers yielded PCR products of 84 base pairs (bp) (PLCγ1), 108 bp (PLCγ2), and 102 bp (PLCβ2). All PCRs included a template-free control. RT-PCR experiments were repeated using RNA extracted from at least three different animals.

Membrane staining and Mag-fluo-4 Ca2+ imaging

Cells were enzymatically dissociated from cerebral vessels as described above, and allowed to adhere to glass slides for 20 min at 4°C. Membrane-specific fluorescent staining was performed with ER-Tracker Red (5 μg/ml) (Invitrogen), and SR Ca2+ stores were assessed using the low-affinity (Kd = 22 μM) Ca2+ indicator Mag-fluo-4 AM (5 μM, Invitrogen). Cells were incubated in ER-Tracker Red and Mag-fluo-4 AM in the patch-clamp recording solution for 30 min at 37°C. Fluorescent images were obtained using a spinning disk confocal microscope (Andor). ER-Tracker Red and Mag-fluo-4 fluorescence were excited by illumination with 543- and 488-nm line, respectively. All images were acquired at 1024 × 1024 pixels and were analyzed using ImageJ [National Institutes of Health (NIH)].

Immunocytochemistry

Cells were enzymatically dissociated from freshly isolated and siRNA-treated vessels as described above. Cells were allowed to adhere to glass slides for 20 min at 4°C, then were fixed with 4% formaldehyde for 10 min, permeabilized with cold methanol (−80°C), and blocked with 2% bovine serum albumin. Cells were incubated in 2% bovine serum albumin blocking solution containing polyclonal primary antibodies overnight at 4°C. Cells were washed and incubated in appropriate fluorescent secondary antibodies for 2 hours at room temperature. Fluorescent images were obtained using a spinning disk confocal microscope (Andor) and a 100× oil immersion objective. Texas Red and fluorescein isothiocyanate (FITC) were excited by illumination with 543- and 488-nm line, respectively. All images were acquired at 1024 × 1024 pixels and were analyzed using a Volocity imaging software package (version 6.0, PerkinElmer Inc.) and ImageJ version 1.42q (NIH). Total fluorescence was determined using the mean fluorescence of the ROI encompassing the whole cell.

Proximity ligation assay

Colocalization of proteins was studied in freshly isolated cerebral artery myocytes using an in situ proximity ligation assay detection kit (Duolink, Olink Biosciences Inc.). Cells were enzymatically dissociated as described. Glass slides were pretreated with a blocking solution containing 1% bovine serum albumin and 1% fish gelatin in Mg-PSS for 10 min at room temperature. Cells were allowed to adhere to glass slides for 20 min at 4°C and fixed with 4% formaldehyde for 10 min at room temperature and 2 hours at 4°C. After subsequent washes in Mg-PSS, cells were permeabilized with cold methanol (−80°C) and incubated overnight in blocking solution containing paired primary antibodies. After incubation in primary antibodies, cells were washed in blocking solution followed by three 10-min washes in 10 ml of Duolink In Situ Wash Buffer A containing 10 mM tris, 150 mM NaCl, and 0.05% Tween 20 (pH 7.4). Cells were incubated in a humidified chamber at 37°C for 60 min with secondary anti-rabbit PLUS and anti-goat MINUS proximity ligation assay probes and later washed three times for 5 min with 10 ml of Wash Buffer A at room temperature. Cells were incubated in ligation ligase solution for 30 min at 37°C in a humidified chamber and later washed three times for 2 min with 10 ml of Wash Buffer A at room temperature. Last, cells were incubated in amplification polymerase solution for 100 min at 37°C in a humidified chamber and later washed twice for 2 min with 10 ml of Duolink In situ Wash Buffer B containing 200 mM tris and 100 mM NaCl (pH 7.5). Cells were further washed with a 1% Wash Buffer B for 1 min and mounted using Duolink In situ Mounting Medium containing DAPI nuclear stain. Fluorescent images were obtained using a spinning disk confocal microscope (Andor) and a 100× oil immersion objective. Generation of positive signal (bright puncta) occurs only when the two proximity ligation assay probes are in close proximity (<40 nm). Excitation of fluorescent puncta was achieved with a 543-nm laser, and autofluorescence of the cytosol was illuminated using a 488-nm laser. Images were analyzed in the Volocity imaging software package (version 6.0, PerkinElmer Inc.). To demonstrate the specificity of the primary antibodies, we performed proximity ligation assay experiments using two antibodies (goat and rabbit) targeting different epitopes of the same protein. Under these conditions, positive proximity ligation assay puncta were observed as the two primary antibodies were bound in close proximity. The density of positive puncta per cell was determined using an automated object-finding protocol in the Volocity imaging software package.

Antibodies

The following primary antibodies were used: anti-PLCγ1 (Santa Cruz, sc-426) for immunocytochemistry and proximity ligation assay experiments; anti-PLCγ2 (Santa Cruz, sc-31751) for immunocytochemistry and proximity ligation assay experiments; anti-PLCβ2 (Santa Cruz, sc-206) for immunocytochemistry and proximity ligation assay experiments; anti-TRPC6 (Alomone, ACC-120) for immunocytochemistry, electrophysiology, and Ca2+ imaging experiments; anti-TRPC6 (Santa Cruz, sc-19197) for proximity ligation assay experiments; anti-TRPM4 (Abcam, ab63080) for immunocytochemistry and proximity ligation assay experiments; anti-TRPM4 (Santa Cruz, sc-67125) for immunocytochemistry and proximity ligation assay experiments; and anti-BKCa (Alomone, APC-021) for proximity ligation assay experiments. Secondary antibodies used were the following: sheep anti-rabbit Texas Red (Abcam, ab6793-1), bovine anti-rabbit FITC (Santa Cruz, sc-2365), donkey anti-goat Texas Red (Santa Cruz, sc-2783), and donkey anti-goat FITC (Abcam, ab6881).

RNA interference and reverse permeabilization

siRNAs directed against PLCγ1 were used to decrease protein abundance in isolated cerebral arteries. siRNAs purchased from Qiagen [1027280 (AllStars Negative Control) and S101961939 (Rn_Plcg1_1)] were dissolved at a concentration of 20 μM in siRNA suspension buffer. Control or PLCγ1 siRNAs were introduced into intact cerebral arteries using a reversible permeabilization procedure. Arteries were permeabilized by first incubating segments for 20 min at 4°C in the following solution: 120 mM KCl, 2 mM MgCl2, 10 mM EGTA, 5 mM Na2 ATP, and 20 mM TES (pH 6.8). Arteries were then placed in a similar solution containing siRNA (40 nM) for 3 hours at 4°C and then transferred to a third siRNA-containing solution with increased MgCl2 (10 mM) for 30 min at 4°C. Permeabilization was reversed by placing arteries in a Mops-buffered physiological siRNA-containing solution consisting of 140 mM NaCl, 5 mM KCl, 10 mM MgCl2, 5 mM glucose, and 2 mM Mops (pH 7.1, 22°C) for 30 min at room temperature. Ca2+ was gradually increased in the latter solution from nominally Ca2+-free to 0.01, 0.1, and 1.8 mM over a 45-min period. After the reversible permeabilization procedures, arteries were cultured for 2 to 3 days in DMEM/F-12 culture medium supplemented with l-glutamine (2 mM) (Gibco) and 0.5% penicillin-streptomycin (Gibco).

Real-time RT-PCR

Arteries treated with control or PLCγ1 siRNA were enzymatically dissociated, and RNA was immediately isolated, purified, and synthesized into cDNA as described above. PLCγ1 mRNA abundance was measured using a real-time SYBR Green detection assay (Bio-Rad), QuantiTect (Qiagen Inc.) primers for PLCγ1 [QT00184100 (Rn_Plcg1_1_SG)], and iQ5 Multicolor Real-Time PCR Detection System (Bio-Rad). Samples were normalized to β-actin [QT00193473 (NM_031144)], and cycling parameters were selected on the basis of the protocol for QuantiTect primer assays (Qiagen Inc.). The decrease in PLCγ1 mRNA abundance was calculated according to the Pfaffl method (59).

Smooth muscle cell membrane potential

To measure smooth muscle cell membrane potential, cerebral arteries were isolated and pressurized to either 20 or 80 mmHg, and smooth muscle cells were impaled through the adventitia with glass intracellular microelectrodes (tip resistance 100 to 200 megohm). A WPI Intra 767 amplifier was used to record membrane potential (Em). Analog output from the amplifier was recorded using IonWizard software (sample frequency 20 Hz). Criteria for acceptance of Em recordings were (i) an abrupt negative deflection of potential as the microelectrode was advanced into a cell, (ii) stable membrane potential for at least 1 min, and (iii) an abrupt change in potential to about 0 mV after the electrode was retracted from the cell.

Calculations and statistics

All data are means ± SE. Values of n refer to the number of arteries for the isolated vessel experiments or the number of cells for immunocytochemistry, proximity ligation assay, and patch-clamp experiments. Unpaired Student’s t tests were performed to detect differences in total PLCγ1-dependent fluorescence from cells isolated from control compared to PLCγ1 siRNA–treated vessels and to demonstrate differences in TRPM4:TRPC6 compared to TRPM4:BKCa proximity ligation assay labeling. In addition, unpaired Student’s t tests were performed to detect differences in TICC total open probability from cells isolated from control compared to PLCγ1 siRNA–treated vessels and control compared to PLCγ1 siRNA–treated vessels treated with Bt-IP3. Paired Student’s t tests were performed to determine differences in single-channel and macroscopic TRPC6-YFP currents in response to intrapipette pressure, GsMTx-4, and anti-TRPC6 inhibitory peptide. In addition, paired Student’s t tests were performed to determine differences in TICC total open probability in response to Bt-IP3 in cells isolated from PLCγ1 siRNA–treated vessels. One-way analysis of variance (ANOVA) was performed to demonstrate differences in proximity ligation assay labeling for TRPM4:PLCγ1, TRPM4:PLCγ2, and TRPM4:PLCβ2, and the percentage of cells with Ca2+ waves in Ca2+ imaging experiments in pressurized vessels. One-way repeated-measures ANOVAs were performed to demonstrate differences in TICC total open probability in response to intrapipette pressures and after administration of U73122, Bt-IP3, and CPA. Two-way repeated-measures ANOVAs were performed to demonstrate differences between TICC total open probability at different intrapipette pressures and in the presence or absence of 9-phenanthrol or extracellular Ca2+; myogenic tone at different intraluminal pressures in the presence or absence of U73122; myogenic tone and membrane potential at different intraluminal pressures from control versus PLCγ1 siRNA–treated vessels; and arterial tone at different concentrations of UTP from vessels treated with U73122. Individual groups were compared using a Student-Newman-Keuls post hoc test. P ≤ 0.05 was accepted as statistically significant for all experiments.

SUPPLEMENTARY MATERIALS

www.sciencesignaling.org/cgi/content/full/7/327/ra49/DC1

Fig. S1. PLC activity is necessary for stretch-induced TRPM4 activity in arterial myocytes.

Fig. S2. Control experiments for U73122.

Fig. S3. PLC activity is necessary for myogenic and agonist-induced constriction.

Fig. S4. Control experiments for siRNA-mediated PLCγ1 down-regulation.

Fig. S5. Rate of Ca2+ depletion within the SR in freshly isolated smooth muscle cells.

Fig. S6. GsMTx-4 and a TRPC6 inhibitory antibody do not block TRPM4 channels.

Fig. S7. Src kinase inhibitor-1 blocks stretch-dependent TICC activity and myogenic tone.

Fig. S8. Activation of TRPC6 with OAG (1-oleoyl-2-acetyl-sn-glycerol).

REFERENCES AND NOTES

Acknowledgments: We thank M. M. Tamkun and the members of his laboratory for generating the TRPC6-YFP expression plasmid used for these studies; Z. Garcia and A. Simpson for technical assistance; and A. Howard for statistical assistance. Funding: This work was supported by a Monfort Excellence Award from the Monfort Family Foundation (to S.E.), Fondation Leducq for the Transatlantic Network of Excellence on the Pathogenesis of Small Vessel Disease of the Brain (to M.T.N.), Totman Medical Research Trust (to M.T.N.), and the NIH [grants R01HL091905 (to S.E.), HL044455 (to M.T.N.), R37 DK053832 (to M.T.N.), 1PO1HL095488 (to M.T.N.), T32 HL007594 (to A.L.G.), and F31HL094145 (to A.L.G.)]. Author contributions: A.L.G., Y.Y., M.N.S., L.S., F.D., and S.E. designed and performed the experiments; A.L.G., Y.Y., M.N.S., F.D., and S.E. analyzed the data; A.L.G., F.D., and S.E. constructed the figures; A.L.G., D.C.H.-E., M.T.N., and S.E. conceptualized and wrote the paper. Competing interests: The authors declare that they have no competing interests.
View Abstract

Navigate This Article