INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

ACTIVE FORCE DEVELOPMENT BY vascular smooth muscle (VSM) in response to elevation of luminal pressure, or stretch, is termed the myogenic response. This response is independent of neural, metabolic, hormonal, and endothelial factors (5). The myogenic response is important for local regulation of blood flow and capillary pressure and for generation of basal vascular tone. However, the underlying cellular mechanisms are incompletely understood.

Myogenic contractions are associated with a sustained depolarization of smooth muscle (14, 15), bringing the membrane potential to threshold for opening voltage-gated L-type Ca²⁺ channels (23). Ca²⁺ influx through these channels is thought to initiate and/or sustain contraction. Inhibition of the L-type channel blocks the myogenic response in nearly all vascular preparations (5). Although L-type Ca²⁺ channels can be directly activated by stretch if single cells are pressurized through a patch pipette (20), the resulting current is too small to account for stretch-induced depolarization of VSM (21); moreover, stretch-induced depolarization persists in the presence of L-type channel blockers (19, 26).

For these reasons, alternative mechanisms have been proposed to account for stretch-induced depolarization, including 1) inhibition of a resting K⁺ conductance (15), 2) activation of a Cl conductance (22), and 3) activation of a nonselective cation conductance (18). Although there is evidence to support each of these mechanisms, a cation channel is consistently activated by whole cell stretch over a range of length changes that would be experienced by VSM cells in vivo (4, 26, 31). The characteristics of this channel, as studied in visceral (31) and VSM cells (4, 26), appear to be appropriate to produce depolarization and initiate Ca²⁺ influx (6).

However, the physiological importance of the cation channel remains to be fully characterized. This is, in part, due to the lack of a selective blocker (12) that could be used to test its function in an intact vascular preparation. The interaction of this current with other mechanosensitive currents described in VSM (5) also remains to be determined. Other currents could potentially augment or counteract the effect of cation channel activation. Therefore, the purpose of this study was to determine the interaction of whole cell currents activated by longitudinal stretch of coronary smooth muscle cells.

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Cell isolation technique. Pigs (6-10 wk old, of either sex) were sedated with telazol (4.4 mg/kg im) and xylazine (Rompun, 2.2 mg/kg im), anesthetized with pentobarbital sodium (20 mg/kg iv), and administered heparin (1,000 U/kg iv). Pigs were then intubated and ventilated with room air. A left thoracotomy was performed, and the heart was electrically fibrillated, excised, and immediately placed in 4°C saline solution.

Patch-clamp techniques. Standard patch-clamp techniques as devised by Hamill et al. (11) were used. Micropipettes for whole cell recordings were pulled from 1.5-mm-outer diameter glass tubing (Corning 8161; Warner Instruments; Hamden, CT) on a programmable puller and fire polished. Pipette resistances ranged from 1 to 3 M. For conventional whole cell recordings, pipettes were back filled with high-K⁺ pipette solution (high K⁺) containing (in mM) 6.0 NaCl, 115 KCl, 10 HEPES, 1.15 NaH₂PO₄, 1.18 MgCl₂, 2.0 CaCl₂, 11 glucose, and 10 EGTA [pH adjusted to 7.2 with KOH; free Ca²⁺ concentration ([Ca²⁺]) = 17 nM]. For perforated whole cell recording, the pipettes also contained 240 µg/ml amphotericin B. An EPC-7 amplifier (HEKA, Darmstadt-Eberstadt, Germany) was used to record current, and Narishige MO-series hydraulic manipulators (Tokyo, Japan) were used for fine control of the micropipettes. Analog-to-digital conversions were made using a TL-1 interface (Axon Instruments; Foster City, CA) and stored on a Pentium computer for subsequent analysis. Data were sampled at 5-10 kHz and filtered at 1-2 kHz with the use of an 8-pole Bessel filter (Frequency Devices, Haverhill, MA). Series resistances varied from 2 to 9 M. Series resistance compensation was used in whole cell recordings to improve voltage control. Current records were analyzed by use of pCLAMP (version 6.0.3, Axon Instruments). All experiments were performed at 22°C.

Cell stretch technique. Before they firmly attached to the chamber bottom, individual cells were stretched with the use of two patch pipettes. The first pipette (Fig. 1, pipette 1), also used for intracellular recording, gently pressed one end of the cell to the bottom of the experimental chamber. After establishment of a gigaseal, a sharp pulse of suction was applied to the rear of the pipette to enter the whole cell recording mode. A second "stretch" pipette (Fig. 1, pipette 2) was sealed to the distal end of the cell in the cell-attached mode. The cell was gently lifted off the chamber bottom by use of the stretch pipette and slowly extended to its slack length as measured on the video monitor. The inner tip diameter of this pipette (Corning 7052; Garner Glass; Claremont, CA) was 3-5 µm after pulling and 2-3 µm after heat polishing. The stretch pipette was filled with the same solution used for superfusion. The position of the stretch pipette was controlled from a computer-driven amplifier that powered a piezoelectric translator attached to the pipette micromanipulator. Length steps, typically in the shape of a trapezoid, were delivered by the computer. The reference length (L) was designated as the minimum length of the cell between the two pipettes when the slack in the cell was removed. Maximum length changes to L = 130% corresponded to the magnitude of the transient distention previously observed in isolated arterioles when pressure was stepped over the entire myogenic range (7). Length changes beyond this level were usually associated with irreversible cell damage. During stretch protocols, the controlling voltage to the piezoelectric manipulator was recorded as well as the signal of the responding current of the cell (sampled at 20-400 Hz) with the use of LabView (National Instruments; Austin, TX). At the same time, the holding potential (HP), seal test, ramp or step voltage, and current changes were recorded using pCLAMP.

View larger version (44K): [in this window] [in a new window]   Fig. 1.   Cell stretch method. Pipette 1 was used for recording and to gently hold one end of the cell. Pipette 2 was sealed to the cell near its distal end, and its position was controlled using a piezoelectric translator. The cell was never firmly attached to the chamber bottom. Pipette 3 was used to apply solutions and/or inhibitors. Pipette 1 contained high K+; pipettes 2 and 3 contained physiological saline solution (PSS). L, reference length; L, change in length; Vm, membrane voltage. Bar in B is 15 µm.

Fura 2 microfluorometry. Intracellular Ca²⁺ concentration ([Ca²⁺]_i) was measured using a fura 2 microfluorometry system (6). Cells were preloaded with the acetoxymethyl ester of fura 2 (fura 2-AM; 10 µM) for 10 min and then superfused for at least 30 min with PSS before fluorescence measurements were started. The ratio of 340/380 fluorescence was converted to [Ca²⁺]_i using a standard method (10). Minimum ratio (R_min) was determined in 10 mM EGTA, and maximum ratio (R_max) was determined in 10 µM Ca²⁺.

Solution changes. Solution changes and drug application to individual cells were made with the use of a picospritzer pipette before and during stretch (Fig. 1B, pipette 3). Cells at the outflow end of the chamber were studied first and cells at the inflow end last. Grammostolla spatulata spider venom was obtained from Spider Pharm (Feasterville, PA). Iberiotoxin and apamin were obtained from Alomone Labs (Jerusalem, Israel). Tetraethylammonium (TEA), hexamethyleneamiloride (HMA), and all other chemicals were obtained from Sigma.

Data analysis. We used data only from those experiments in which the distal cell end did not pull away from the stretch pipette and in which a stable gigaseal was maintained throughout the stretch protocol. The high-resolution traces from pCLAMP were used to determine current amplitude. In most analyses, raw current values were normalized to cell capacitance (an index of cell size) and expressed as current density (in pA/pF). Whole cell capacitance ranged from 6 to 22 pF. Statistical comparisons were performed with repeated-measures analysis of variance (ANOVA) followed by post hoc tests or with an independent two-tail t-test, as appropriate. Summary values are presented as means ± SE. Values of P < 0.05 were considered to be statistically significant. When normalized current was plotted versus cell stretch, the relationship was fit to the Boltzmann equation (I = {1 + exp[(L  L_0.5)/k]}¹), where I is current, L_0.5 is the length producing 50% current activation, and k is the slope factor.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

An inward current is activated by whole cell stretch. Longitudinal stretch typically induced an inward cation current, termed I_CAT, at negative holding potentials (n = 285 cells from 112 animals). The magnitude of I_CAT varied from 0.8 to 6.9 pA/pF (HP = 60 mV) for length increases between 105 and 135% of L. I_CAT was reversible, and its amplitude increased monotonically with stretch. I_CAT was noisier than the resting membrane current, suggesting that it arose from a channel. The delay in I_CAT activation ranged from 20 to 1,000 ms after stretch. In most cells, I_CAT could be recorded for at least 90 s without substantial inactivation.

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Longitudinal stretch consistently induced an inward cation current (ICAT). A: a single vascular smooth muscle (VSM) cell was held at 60 mV while being stretched to 120% of the reference length (L) for 15 s. Inward ICAT is indicated by a downward deflection in the current (I) density trace. B: ICAT was also induced by a 1-Hz sine-wave stimulus. Trace is representative of 6 cells. C: gigaseal breakdown during stretch. Inward ICAT density (ICAT, 2 to 3.4 pA/pF) was recorded after 2 successive stretches to 115 and 120% of L. However, near the end of the second stretch (arrow), the seal broke down, resulting in a large and noisy current (greater than 15 pA/pF). The first inward ICAT recording reversed when stretch was released, but the second current recording did not. Bath solution, PSS; pipette solution, high K+. Holding potential (HP) = 60 mV. Acquisition rate = 80-120 Hz.

I_CAT amplitude as a function of the magnitude of stretch. Graded increases in the magnitude of stretch were used to investigate whether the amplitude of I_CAT was proportional to the magnitude of stretch. With cells stretched to 115, 120, 125, 130, or 135% of L, there was a corresponding increase in the amplitude of I_CAT. An example is shown in Fig. 3A, in which length changes to 120, 125, and 130% of L were imposed at HP = 60 mV. The duration of stretch was 6 s in each case. The amplitude of the inward I_CAT increased progressively with increasing stretch, and the entire data set for this cell is plotted in Fig. 3B. A fit of the Boltzmann equation to the data in Fig. 3B shows that half-maximum inward I_CAT was evoked by a stretch to 123.3% of L; k was 1.8% above L. It was difficult to consistently analyze this relationship for individual cells because most cells could not successfully be stretched through the entire range of lengths. The composite data set for 25 different cells is shown in Table 1. The threshold for consistent, measurable I_CAT was a stretch to 115% of L, and 135% of L was usually the maximum length change that could be imposed without damage. This magnitude of stretch is consistent with that observed in isolated arterioles when pressure is stepped through the entire range over which the vessels respond myogenically (7).

View larger version (12K): [in this window] [in a new window]   Fig. 3.   Amplitude of ICAT was graded with increases in the magnitude of stretch. A: length changes to 120, 125, and 130% of L elicited graded increases in inward ICAT density. B: the magnitude of ICAT plotted as a function of L for the cell in A. Data were fit by a Boltzmann equation with half-maximal stretch equal to 123.3% of L and slope factor (k) = 1.8% above L. Bath solution, PSS; pipette solution, high K+. HP = 60 mV. Acquisition rate = 300 Hz.

Current-voltage relationship for I_CAT. To determine the whole cell current-voltage (I-V) relationship for I_CAT, voltage ramps from 100 to +60 mV (duration = 200 ms, HP = 60 mV, average of 3-5 ramps) were applied before and during longitudinal cell stretch. We used only data from I_CAT recordings that were stable during stretch (duration = 5-10 s). Figure 4A summarizes the I-V relationship of membrane currents before and during 115% stretch (n = 5). Stretch produced a 248% increase in inward I_CAT at 70 mV and 20% enhancement in outward current at +60 mV. The difference current recorded before and after stretch is shown in Fig. 4B. The I-V relationship of the difference current was approximately linear between 70 and 10 mV, rectified inwardly at voltages negative to 70 mV, and rectified outwardly at voltages positive to 10 mV. The reversal potential of the stretch-induced current in this group of cells was 18.2 ± 3.2 mV (stretch to 115% of L). The reversal potentials at other levels of stretch, 105, 110, 115, and 120% of L, were not significantly different from this (Table 2). Membrane conductance increased from 0.3 ± 0.04 to 1.0 ± 0.07 nS at 100 mV and from 1.5 ± 0.05 to 1.8 ± 0.04 nS at +50 mV during stretch to 115% of L.

View larger version (13K): [in this window] [in a new window]   Fig. 4.   Current-voltage (I-V) relationship of stretch-induced current. A: I-V plot of the average membrane current evoked by voltage ramp from 100 to +60 mV (200-ms duration), with HP = 60 mV. Current density was recorded before () and during () stretch to 115% of L (n = 5). Each line is the average of 5 responses. B: I-V curve of the difference current, i.e., obtained before and during stretch. The reversal potential of the difference current was 18.2 ± 3.2 mV for stretch to 115% of L. Bath solution, PSS; pipette solution, high K+.

[Ca²+]_i changes during I_CAT recording. To test whether longitudinal cell stretch induced an increase in [Ca²⁺]_i concomitant with I_CAT, global [Ca²⁺]_i was measured using fura 2 microfluorometry while current recordings were performed in the perforated-patch, whole cell mode. With cells bathed in PSS and high-K⁺ solution in the recording pipette, [Ca²⁺]_i levels consistently increased in response to 120% stretch, as shown in Fig. 5. The onset of the changes was normally delayed 0.2-2 s after the change in length, and the peak [Ca²⁺]_i increase was delayed significantly with respect to the peak I_CAT. In the example shown in Fig. 5, an inward I_CAT of approximately 3 pA/pF was evoked by stretch. This current activated over the same time course in which [Ca²⁺]_i increased from 100 to a peak of 500 nM, followed by a plateau to 300 nM. Data for five cells showed an average increase in [Ca²⁺]_i to 278 ± 35.8% of control levels in response to a 120% stretch.

View larger version (17K): [in this window] [in a new window]   Fig. 5.   Time course of intracellular Ca2+ concentration ([Ca2+]i) change relative to the change in stretch-evoked current density. Trapezoidal stretch to 120% of L activated an inward ICAT at HP = 60 mV. A concomitant rise in [Ca2+]i occurred. In this cell, the inward ICAT and [Ca2+]i declined slightly with time, while stretch was maintained. When the cell was returned to its original L, the current inactivated, but the decline in [Ca2+]i was delayed (complete recovery is not shown). Test pulses from 60 to +50 mV (duration = 400 ms) were delivered before, during, and after stretch to estimate magnitude of outward K+ current (IK) density. Outward IK could be recorded at +50 mV in the absence of stretch, but the amplitude of the current was increased during stretch. Perforated patch-clamp mode with cell preloaded with fura 2-AM. Bath solution, PSS; pipette solution, high K+ with amphotericin B. R340/380, ratio of 340- to 380-nm fluorescence.

An outward K+ current is enhanced secondary to stretch. An outward whole cell current was also enhanced during stretch, concomitant with the increase in [Ca²⁺]_i and I_CAT. This is noted in the cell shown in Fig. 5 by the current response to the three brief voltage pulses (60 to +50 mV, duration = 400 ms) delivered before, during, and after stretch. Although little outward current was evident at the holding potential (HP = 60 mV), 14.5 pA/pF of outward current were evoked by the first depolarizing pulse. The amplitude of this current increased to 20.2 pA/pF during the application of a 120% stretch and then returned toward control (to 16.2 pA/pF) after stretch was released. This observation suggested that an outward Ca²⁺-dependent current might be enhanced secondary to I_CAT.

View larger version (17K): [in this window] [in a new window]   Fig. 6.   Time course of the outward IK activated during stretch. A: stretch to 120% and then to 125% of L, each stretch lasting 10 s. An inward ICAT was activated by each stretch when HP was 60 mV, as noted by a sustained downward deflection in the current density trace coinciding with each increase in stretch. Transient voltage steps to +50 mV (duration = 400 ms) elicited larger outward IK superimposed on the inward ICAT. a-d: outward IK evoked by 4 successive voltage pulses, with a and c recorded before stretch and b and d recorded during stretch. B: a higher-resolution trace (from pCLAMP) of the outward IK elicited during voltage pulses c and d. d-c, Difference current between c and d. An outward difference current suggests that enhancement of IK occurred during stretch. Bath solution, PSS; pipette solution, high K+.

View larger version (15K): [in this window] [in a new window]   Fig. 7.   Characteristics of the outward IK enhanced by stretch. A: amplitude of the outward IK was not increased during stretch if the cell was bathed in Ca2+-free ([Ca2+]o free, where [Ca2+]o is extracellular Ca2+ concentration) PSS. ICAT was still recorded in response to stretch, as noted by a sustained downward deflection in the current density trace coinciding with the L increase. Voltage steps from 60 to +50 mV during this time elicited an outward IK, the amplitude of which was not noticeably enhanced during stretch. B: 2 stretches to 130% of L were used to test the effect of adding 1 mM tetraethylammonium (TEA) to the bath solution (PSS). Voltage steps from 60 to +50 mV revealed an increase in outward IK during stretch (trace b vs. a). TEA caused a substantial decrease in outward IK at +50 mV before stretch (trace d vs. c), and, in the presence of TEA, outward IK did not significantly increase during stretch (trace e vs. trace d). There were no substantial differences in the amplitude of outward IK with time in the absence of TEA and stretch (trace a vs. trace c vs. trace f). Also, the amplitude of ICAT was similar between the 2 stretches in the presence of TEA. C: summary of effects of TEA (1 mM) and iberiotoxin (IbTX; 100 nM) on IK before and after stretch. For TEA protocols, stretch was to 130% of L. For IbTX protocols, stretch was to 125% of L. HP = 60 mV; test pulse = +50 mV; n = no. of cells. Bath solution, Ca2+-free PSS (A) and PSS (B and C); pipette solution, high K+. * P < 0.05 vs. control.

View larger version (15K): [in this window] [in a new window]   Fig. 8.   Stretch-induced depolarization was enhanced by Ca2+-activated K+ channel blockade. A: with the use of the current-clamp configuration, longitudinal stretch to 130% of L induced a 17-mV membrane depolarization from a resting potential near 60 mV. B: depolarization induced by stretch to 130% of L in another cell was enhanced by addition of IbTX (100 nM) to the bath. In addition, spontaneous fluctuations in the membrane potential recording were reduced in the presence of IbTX (see portion of trace above the dashed line). C: summary of effects of IbTX (100 nM) or apamin (200 nM) on membrane potential at rest and during stretch to 130% of L. Nos. in parentheses indicate no. of cells. Bath solution, PSS; pipette solution, high K+. * P < 0.05 vs. control group (no stretch). # P < 0.05 vs. stretch group.

Effects of putative mechanosensitive channel blockers. Although no highly selective blockers of mechanosensitive channels have been described to date, HMA is reported to block stretch-activated single-channel cation currents in Xenopus oocytes (13), and G. spatulata venom is reported to block mechanosensitive whole cell cation currents in GH3 cells, Xenopus oocytes, and chick heart cells (3, 16). We tested the effects of these compounds on the inward and outward components of stretch-activated current in coronary smooth muscle cells.

View larger version (18K): [in this window] [in a new window]   Fig. 9.   Effect of putative mechanosensitive channel blockers on ICAT and IK . A: 2 identical stretches to 120% of L evoked an inward ICAT at 60 mV and enhanced the outward IK at +50 mV. Three voltage ramps (from 100 to +60 mV, 200-ms duration) were delivered at periodic intervals to measure IK, and the changes in IK (in C) are the average of the 3 currents. ICAT was prevented by 50 µM hexamethyleneamiloride (HMA) as indicated by the absence of inward ICAT at 60 mV during HMA perfusion (for stretch 2). Stretch-induced increases in IK did not occur in the presence of HMA (trace e vs. trace d). HMA did not block control IK in the absence of stretch (trace d vs. trace c or trace a). There were no substantial differences in the amplitude of IK with time (trace a vs. trace c vs. trace f). B: same protocol repeated in the presence of G. spatulata spider venom (1:100,000). ICAT was blocked by the spider venom as indicated by the absence of an inward ICAT evoked by stretch (for stretch 2). Stretch-induced increases in IK also did not occur in the presence of spider venom (e vs. d), but the venom significantly blocked IK even in the absence of stretch (d vs. c). C: summary of effects of HMA (50 µM) and G. spatulata venom (1:100,000) on ICAT and IK. For IK, test pulse = +50 mV. Nos. in parentheses indicate no. of cells. Stretch to 120% of L in all protocols. Bath solution, PSS; pipette solution, high K+; HP = 60 mV. * P < 0.05 vs. control.

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

To investigate the mechanism underlying myogenic depolarization of VSM, whole cell currents were recorded from single coronary myocytes during longitudinal stretch. Stretch to 115-135% of control length caused activation of an inward cation current, induced membrane depolarization, and produced an increase in global [Ca²⁺]_i. The magnitude of I_CAT ranged from 0.8 to 6.9 pA/pF (HP = 60 mV), was graded with stretch, inactivated on release of the stimulus, and could be repeatedly induced. An outward I_K was also enhanced by stretch at depolarized potentials. In Ca²⁺-free bath, I_CAT could still be recorded, but the stretch-induced enhancement in I_K was absent. The reputed mechanosensitive channel blockers HMA and G. spatulata venom blocked stretch-induced increases in both I_CAT and I_K, but G. spatulata venom also blocked basal I_K, suggesting it was not selective for mechanosensitive current. K⁺ channel blockers, including TEA and IbTX, blocked the stretch-induced increase in I_K and increased the magnitude of the stretch-induced depolarization but did not alter I_CAT. We concluded that longitudinal stretch of coronary myocytes directly activates an inward I_CAT at the resting membrane potential and secondarily activates a K_Ca channel. The enhancement in I_K may serve to counteract stretch-induced depolarization and thereby limit the magnitude and/or duration of a myogenic contraction.

Evidence for a depolarization-linked mechanism underlying myogenic tone derives from several lines of experimentation. 1) Graded VSM depolarization occurs as pressure in isolated small arteries and arterioles is increased (17). 2) Smooth muscle strips (2) and single myocytes depolarize in response to longitudinal stretch (4, 26, 31), such that depolarization by 10-15 mV from a resting potential of about 55 mV would be sufficient to increase the open probability of voltage-gated Ca²⁺ channels by severalfold (23). 3) The myogenic response is blunted when the normal ion gradients across the plasma membrane are altered so as to minimize the extent of stretch-induced depolarization (5). 4) Ca²⁺ channel blockers attenuate myogenic tone in almost all types of smooth muscle (5). Depolarization of smooth muscle would activate Ca²⁺ entry and thereby enhance contraction (14, 17). This is not the only mechanism involved, however, because there is considerable evidence that Ca²⁺-independent pathways, e.g., protein kinase C activation, are also important (5), perhaps for long-term maintenance of contraction at high pressure.

The mechanism by which stretch is coupled to depolarization is controversial. Both the speed of the response and the change in membrane potential point to a role for ion channels. One explanation is that depolarization is mediated by activation of a cation channel promoting Na⁺/Ca²⁺ influx. Such channels have been recorded in myocytes isolated from a number of smooth muscles (4, 18, 25, 32). These channels typically exhibit a relative permeability of K⁺ Na⁺ > Ca²⁺ and single-channel conductance between 30 and 40 pS (for monovalent ions) (4, 18, 32). In the presence of Ca²⁺, they often rectify (18, 32) and show reduced monovalent conductance (18, 26), suggesting that a Ca²⁺-dependent inactivation mechanism exists. In terms of physiological relevance, it is important to note that stretch-activated whole cell currents and/or depolarizations have now been recorded in pig coronary myocytes (4), urinary bladder (30, 31), and rat mesenteric arterioles (25, 26). The reversal potentials for whole cell currents and their dependence on extracellular Na⁺ concentration are consistent with activation of a nonselective cation channel rather than a pure Ca²⁺ conductance (4, 25, 26, 31).

Voltage-gated Ca²⁺ channels might also be modulated directly by stretch. L-type Ca²⁺ currents in rat cerebral arterial myocytes, as recorded by use of the conventional whole cell mode, are enhanced by inflating cells through a patch pipette (20, 21) and, in perforated-patch recordings, by hypoosmotic cell swelling (20). Yet current flow through these channels is probably too small to account for stretch-induced depolarization (21), and stretch-induced depolarization persists in the presence of Ca²⁺ channel blockade (19, 26).

Stretch-induced depolarization might also be mediated by activation of Cl efflux if a mechanosensitive Cl channel in smooth muscle responded to longitudinal stretch and if the Cl equilibrium potential of the smooth muscle cell were more positive than the resting membrane potential (22). These requirements have not been thoroughly tested, but the reversal potential of stretch-induced whole cell current appears to be independent of changes in extracellular Cl concentration (4, 26). Although Cl channel blockers inhibit myogenic responsiveness (22), their actions may be explained by nonspecific effects on voltage-gated Ca²⁺ channels (9). Thus the evidence supporting a key role for a Cl channel is not compelling at this time.

Inhibition of K⁺ efflux might also account for stretch-induced depolarization if one or more of the various K⁺ channels identified in smooth muscle were inhibited by membrane stretch (5). Mechanosensitive K⁺ channels have been identified in smooth muscle by use of single-channel recording techniques, but the channels are activated, rather than inactivated, by stretch (8). To mediate stretch-induced depolarization, a putative mechanosensitive I_K must exhibit basal activity when a vessel has resting vascular tone. Although some isolated vessel data are consistent with this requirement (19), electrophysiological support is largely lacking. Voltage-dependent K⁺ (K_V) channel blockers depolarize VSM cells in pressurized arterioles and enhance myogenic tone (19), but the more likely candidate to mediate a stretch-sensitive I_K is the large-conductance K_Ca (BK) channel (8). Wesselman et al. (33) concluded that BK channels were necessary for pressure-induced contraction of isolated mesenteric arteries because charybdotoxin, a relatively specific inhibitor, caused a decrease in myogenic index. However, those experiments were performed in vessels precontracted with norepinephrine, which might have led to basal BK channel activation. On the basis of the effect of specific BK inhibitors on basal tone, it appears that not all blood vessels have basal BK channel activity that could potentially be inhibited by stretch (5).

A more likely role for K⁺ channels in myogenic responsiveness is to counteract, rather than initiate, myogenic tone. Indeed, because the conductance of BK channels is so large and activation of only a few would be necessary to substantially change smooth muscle membrane potential, it has been proposed that stretch-induced depolarization could not be maintained unless an endogenous inhibitor of the channels is produced (15). There is evidence that 20-hydroxyeicosatetraenoic acid (20-HETE), a cytochrome P-450 metabolite of arachidonic acid, may be one such substance (15). However, the use of (reputedly) selective inhibitors has confirmed a role for 20-HETE only in some (33), not all (1), vessels.

Data in the present study support the idea that a K_Ca current is activated secondary to stretch-induced Ca²⁺ entry, whereas a cation channel underlies the depolarization controlling Ca²⁺ entry. The enhanced I_K had the following characteristics under the conditions of our experiments: 1) it activated over a voltage range appropriate for K_V or K_Ca channels (Fig. 4), 2) its activation coincided with the time course of stretch-induced [Ca²⁺]_i increases (Fig. 5), 3) its enhancement was prevented in Ca²⁺-free bath solution (Fig. 7), and 4) it was blocked by inhibitors of K_Ca channels (Figs. 7 and 8). These observations are all consistent with this channel being a K_Ca channel. Importantly, its activation must have occurred secondary to that of the cation channel, because I_K was blocked when I_CAT was blocked, but not vice versa (Fig. 9). The idea that I_K counteracts stretch-induced depolarization is supported by our observation that stretch-induced depolarization was potentiated by the specific K_Ca channel inhibitor IbTX (Fig. 8). Importantly, IbTX had no effect on resting membrane potential, suggesting it blocked only a component of current that was activated after cell stretch (Fig. 8C). Our conclusions are consistent with the findings of Wellner and Isenberg (31), who reported that TEA caused a shift in reversal potential of I_CAT in bladder myocytes toward more positive values. In that study, stretch-enhanced outward I_K was also blocked by intracellular Ca²⁺ chelation (31). Collectively, these characteristics match those of a K_Ca current.

Our data support the conclusion that a cation channel is activated by longitudinal stretch in coronary arteries and arterioles, leading to membrane depolarization. The characteristics of the whole cell current associated with this channel were as follows: 1) it was reversibly and repeatedly activated by stretch (Fig. 2); 2) it showed graded increases in amplitude with increasing stretch, up to 135% of control cell length (Fig. 3); 3) it activated over the same time course as stretch-induced depolarization (Figs. 2 and 8); and 4) it had a reversal potential in PSS between 15 and 20 mV (Fig. 4 and Table 2). These characteristics are nearly identical to those for stretch-activated whole cell currents recorded in other smooth muscle cells (26, 31). Furthermore, I_CAT persisted in Ca²⁺-free bath solution (Fig. 7) and in the presence of K⁺ channel blockers (Figs. 7 and 8) and was blocked by inhibitors of mechanosensitive channels such as Gd³⁺ (28), HMA, and G. spatulata venom (Fig. 9). Because the stretch-activated current can be recorded at negative potentials after extracellular Cl substitution (4), it is likely that current is carried predominantly by Na⁺ at negative potentials. At positive potentials, much of the current must be carried by K⁺, because it can still be recorded as the equilibrium potential for Na⁺ is approached. The channel is likely to be Ca²⁺ permeable as well, as is typical of other non-voltage-gated, mechanosensitive channels (4, 18, 32). The observation that this current sometimes decays markedly after stretch (16, 32) suggests a possible Ca²⁺-dependent inactivation mechanism that will make accurate determination of Na⁺-to-Ca²⁺ permeability ratio difficult.

Is activation of a nonselective cation channel necessary for initiation of the vascular myogenic response? This question was not answered by the present experiments and cannot be answered at this time. There are several reputed inhibitors of mechanosensitive channels used to block mechanotransduction in muscle (12). Yet all of these compounds, including Gd³⁺, streptomycin antibiotics, and G. spatulata venom, also inhibit voltage-gated Ca²⁺ channels. For example, Gd³⁺ inhibits I_CAT in mesenteric artery and bladder myocytes (26, 31), inhibits stretch-induced depolarization of single VSM cells, inhibits stretch-induced Ca²⁺ influx (6), and blocks myogenic tone of arterioles (27, 29). However, Gd³⁺ inhibits L-type channels in smooth muscle at 1,000-fold lower concentrations than required to inhibit mechanosensitive cation channels (28), so its physiological effect on arterioles could be explained simply by a downstream action on the L-type channel. Thus circumstantial evidence linking stretch-induced depolarization to activation of a nonselective cation channel is growing, but the lack of a selective, high-affinity antagonist of that channel prevents definitive testing of the hypothesis that its activation is required for myogenic responsiveness.