Recently, we have shown that Rho and Rho-activated kinase (ROCK) may become activated by high-millimolar KCl, which had previously been widely assumed to act solely through opening of voltage-dependent Ca2+ channels. In this study, we explored in more detail the relationship between membrane depolarization, Ca2+ currents, and activation of Rho/ROCK in bovine tracheal smooth muscle. Ca2+ currents began to activate at membrane voltages more positive than −40 mV and were maximally activated above 0 mV; at the same time, these underwent time- and voltage-dependent inactivation. Depolarizing intact tissues by KCl challenge evoked contractions that were blocked equally, and in a nonadditive fashion, by nifedipine or by the ROCK inhibitor Y-27632. Other agents that elevate intracellular calcium concentration ([Ca2+]i) by pathways independent of G protein-coupled receptors, namely the SERCA-pump inhibitor cyclopiazonic acid and the Ca2+ ionophore A-23187, evoked contractions that were also largely reduced by Y-27632. KCl directly increased Rho and ROCK activities in a concentration-dependent fashion that paralleled closely the effect of KCl on tone and [Ca2+]i, as well as the voltage-dependent Ca2+ currents that were measured over the voltage ranges that are evoked by 0–120 mM KCl. Through the use of various pharmacological inhibitors, we ruled out roles for Ca2+/calmodulin-dependent CaM kinase II, protein kinase C, and protein kinase A in mediating the KCl-stimulated changes in tone and Rho/ROCK activities. In conclusion, Rho is activated by elevation of [Ca2+]i (although the signal transduction pathway underlying this Ca2+ dependence is still unclear) and possibly also by membrane depolarization per se.
- myosin light chain phosphatase
- Rho-activated kinase
- voltage-dependent Ca2+ channels
constriction of smooth muscle in general is a complex event, involving a wide variety of disparate signaling mechanisms. One major pathway involves elevation of intracellular calcium concentration ([Ca2+]i), in turn leading to activation of myosin light chain kinase, a serine/threonine kinase that phosphorylates serine-19 of the 20-kDa myosin light chain, thereby causing a 10-fold increase in actomyosin ATPase activity. The elevation of [Ca2+]i is a product of both influx of external Ca2+ via voltage-dependent Ca2+ channels and release of internal Ca2+ sequestered primarily within the sarcoplasmic reticulum. To further complicate the matter, there may be interactions between these two pathways: Ca2+ influx can trigger Ca2+ release (4), and Ca2+ release can be modulated directly by changes in membrane potential (9, 15).
Contraction in smooth muscle may involve not only an elevation of [Ca2+]i but also an enhancement of the Ca2+ sensitivity of the contractile apparatus (31). Myosin is dephosphorylated by myosin light chain phosphatase (MLCP), which is tonically active but can also be upregulated by relaxants (20). Constrictors, on the other hand, suppress MLCP activity (31), resulting in greater contraction for any given change in [Ca2+]i (increased Ca2+ sensitivity). In some airway preparations, MLCP activity is downregulated by protein kinase C (PKC) (1), whereas in others it involves the monomeric G protein Rho and its downstream effector Rho-activated kinase (ROCK) (16, 34).
In airway smooth muscle (ASM), voltage-dependent constrictor responses have been particularly puzzling. Although substantial voltage-dependent Ca2+ currents can be measured by the patch-clamp electrophysiological technique (10, 11, 17, 24, 26), the voltages required to see appreciable Ca2+ channel activation are rarely seen, even during aggressive stimulation with bronchoconstrictors, and those voltages required for maximal activation are essentially never achieved by physiologically relevant bronchoconstrictors (18). High-millimolar potassium chloride (KCl), on the other hand, is a much more effective agent for depolarizing the membrane and is routinely used to evoke sustained constrictor responses (which, it should be pointed out, are generally only a fraction of the magnitude of those contractions evoked by cholinergic agonists). Many ASM physiologists operate from the assumption that KCl acts solely through triggering of voltage-dependent Ca2+ influx. However, we have recently shown that Rho and ROCK are activated by 60 mM KCl, leading to increased Ca2+ sensitivity of the contractile apparatus (21). The mechanisms underlying this phenomenon are poorly resolved in ASM. Studies done in non-ASM preparations (2, 12, 28–30, 32) may not be helpful, as there appear to be important tissue- and species-related differences. For example, one group found this phenomenon to be present in rat ureter but not that of the guinea pig (30); another group found it in rat and rabbit aorta but not chicken gizzard (28). In some preparations, it is associated with a change in Ca2+ homeostatic mechanisms (12, 30), whereas in others these are unaffected (28, 32). One group found the underlying mechanism included activation of a Ca2+ entry pathway distinct from voltage- or store-operated channels (12), whereas another emphasized the importance of a Ca2+-dependent colocalization of ROCK with caveolae on the membrane (32), and a third focused upon reorganization of the actin cytoskeleton (28).
In this study, we studied the potential mechanisms by which voltage-dependent Ca2+ influx stimulated the activity of the Rho/ROCK pathway in ASM. In particular, we examined the effects of KCl at concentrations ranging from 0 to 120 mM (which produce increasing degrees of membrane depolarization) on tone, [Ca2+]i, and Rho/ROCK signaling and compared these effects with the voltage dependence of activation of voltage-dependent Ca2+ channels.
Preparation of isolated tissues and single cells.
All experimental procedures were approved by the McMaster University Animal Care Committee and the McMaster University Biosafety Committee and conform to the guidelines set out by the Canadian Council on Animal Care.
Lobes of lung and tracheae were obtained from cows (135–455 kg) killed at a local abattoir and immediately put in ice-cold physiological solution for transport to the laboratory. Tracheal smooth muscle (TSM) was isolated by removing connective tissue, vasculature, and epithelium and then cut into strips parallel to the muscle fibers (∼1 mm wide).
Muscle bath technique.
Tracheal strips were immersed in 2.5-ml muscle baths (one end tied to a glass rod serving as an anchor, the other end tied to a force transducer by silk thread). Tissues were bathed in Krebs-Ringer buffer (see below for composition) containing indomethacin (10 μM), Nω-nitro-l-arginine(l-NNA, 10−4 M), and atropine [1 μM, to block the possible effects of KCl-induced release of acetylcholine from cholinergic nerve endings (25)], bubbled with 95% O2/5% CO2, and maintained at 37°C; tissues were passively stretched to impose a preload tension of ∼1 g. Isometric changes in tension were amplified, digitized (2 samples/s), and recorded on-line (DigiMed System Integrator; MicroMed, Louisville, KY) for plotting on the computer. Tissues were equilibrated for 1 h before the experiments commenced, during which time the tissues were challenged with 60 mM KCl three times to assess the functional state of each tissue.
Assay for Rho activity.
Tissues that had been flash frozen were homogenized in ice-cold buffer [50 mM Tris·HCl, pH 7.5, 0.1 mM EDTA, 0.1 mM EGTA, 750 mM NaCl, 5% Igepal CA-630, 50 mM MgCl2, 10% glycerol, 10 μg/ml aprotinin, 10 μg/ml leupeptin, 1 mM phenylmethylsulfonyl fluoride, 1 mM 4-(2-aminoethyl)-benzenesulfonyl fluoride, and 2 mM sodium orthovanadate]; total protein content was determined (Bradford method) and adjusted (by addition of media) to make uniform. Tissue homogenates were incubated (60 min, at 4°C) with rhotekin-coated cellulose beads (rhotekin specifically binds activated Rho and not inactive Rho). The sample was then lightly centrifuged (14,000 g for 10 s, at 4°C) to “pull down” the beads, and the supernate (unbound material) was discarded, after which Rho was dissociated from the beads by incubation with Laemmli sample buffer [62.5 mM Tris·HCl, pH 6.8, 2% SDS, 10% glycerol, 50 mM dithiothreitol (DTT), 0.1% 2-mercaptoethanol, and 0.01% bromphenol blue]. Samples were boiled for 5 min, subjected to electrophoresis, and then transferred to nitrocellulose membrane (blocked with 3% BSA in 0.05% Tris-buffered saline Tween). Rho was visualized with a rabbit anti-Rho polyclonal antibody preparation (Upstate Biotechnology, Waltham, MA).
Tissues were homogenized, and protein content was adjusted as outlined above. Tissue homogenates were incubated (10 min, at 30°C) with 0.5 ng MYPT (the myosin-targeting subunit of myosin light chain phosphatase), after which the reaction was terminated by addition of Laemmli sample buffer. Samples were then subjected to Western blot analysis, as outlined above. Phospho-MYPT was visualized with a rabbit anti-phospho-MYPT1 polyclonal antibody preparation (Upstate Biotechnology).
Protein kinase A assay.
Smooth muscle strips were homogenized in ice-cold homogenization buffer [50 mM Tris·HCl, pH 7.5, 0.1 mM EDTA, 0.1 mM EGTA, 0.1% 2-mercaptoethanol, 25 μg/ml aprotinin, 25 μg/ml leupeptin, and 1 mM 4-(2-aminoethyl)-benzenesulfonyl fluoride] and centrifuged at 13,000 g for 10 min at 4°C, and the supernatants were collected for assay of protein kinase A (PKA) using a commercially available assay kit (Upstate). In brief, the protein concentrations of the samples were determined by the Bradford method and adjusted to 3 mg/ml. Protein (60 μg) was then incubated with 32P-labeled ATP (2 μCi) and kemptide (20 μM), a substrate for which PKA is highly selective, for 10 min at 30°C. Inhibitor cocktail, provided within the kit, was added to prevent dephosphorylation of the substrate. A 25-μl aliquot of the reaction mixture was blotted on P81 cellulose and washed four times with phosphoric acid (0.75%) and once with acetone (99.5%). Incorporated radioactivity was then counted with a scintillation counter.
TSM was minced and placed in dissociation buffer containing collagenase (Sigma blend type-F, 66.5 mg/ml) and elastase (type IV, 25 mg/ml) and incubated for 30 min at 37°C. Papain (2 g/ml) and l-DTT [(−)-1,4-dithio-l-threitol, 50 mg/ml] were subsequently added and incubated an additional 20–30 min. Cells were gently triturated and then centrifuged for 1 min at 200 rpm in a Hermle Z 233 M centrifuge (Mandel Scientific). Cells were resuspended in standard Ringer solution and stored up to 24 h at 4°C. Several drops of cell suspension were added to the bottom of a recording chamber (1.5-ml volume). Cells were allowed to settle and adhere and then superfused with standard Ringer solution at room temperature. Electrophysiological responses were tested in cells that where phase dense, appeared relaxed, and responded to a 2-s application of caffeine (5 mM). Whole cell currents were recorded using the nystatin perforated-patch configuration of the standard patch-clamp technique. Pipettes with tip resistances of 3–5 MΩ when filled with sterile filtered (0.2 μm) electrode solution were fashioned from borosilicate glass with a P-87 Flaming/Brown micropipette puller (Sutter Instrument, Novato, CA). Electrophysiological recordings commenced once series resistance compensation dropped <30 MΩ. Membrane currents were measured (filtered at 1 kHz, sampled at 2 kHz) using an Axopatch-1D patch-clamp amplifier, digitized using a DigiData 1200 analog-to-digital converter, recorded on a local hard drive, and analyzed with pCLAMP6 software (Axon Instruments, Foster City, CA). Pharmacological agonists were delivered via micropipettes driven by a pressure ejection system (Picospritzer II, General Valve, Fairfield, NJ).
Bovine tracheal strips were digested and dissociated as described above (see Patch clamping), the tissue digest was centrifuged (200 rpm for 1 min at room temperature), the pellet was resuspended in Ringer buffer, and cells were incubated with fluo-4 AM (2 μM, dissolved in DMSO containing 0.1% Pluronic F-127, 37°C, 30 min). Cells were placed in the recording chamber of an Olympus microscope equipped with a charge-coupled device camera and imaging software (ImageMaster for Windows 2.0; Photon Technology Int., Lawrenceville, NJ). Cells studied were chosen on the basis of morphology (spindle-shaped, relaxed, absence of blebs, phase-bright) and responsiveness to 10 mM caffeine (applied by micropipette brought into vicinity of the cell with a hydraulic micromanipulator). Dye was excited by light from a Hg arc lamp (75 W; 494-nm excitation), and emitted light images (516 nm) were acquired at 1 Hz. Fluorescence intensities from regions of interest (central, nonnuclear regions of cell) were saved and plotted against time.
Solutions and chemicals.
Unless indicated otherwise, tissues were studied using standard Krebs-Ringer buffer containing (in mM): 116 NaCl, 4.2 KCl, 2.5 CaCl2, 1.6 NaH2PO4, 1.2 MgSO4, 22 NaHCO3, and 11 d-glucose, bubbled to maintain pH at 7.4. l-NNA (10−4 M) and indomethacin (10−5 M) were also added to prevent generation of nitric oxide and of cyclooxygenase metabolites of arachidonic acid, respectively. Isotonic challenge with KCl was accomplished by replacing NaCl with KCl (either 60 or 120 mM). Freshly dissociated cells were studied using standard Ringer solution containing (in mM): 130 NaCl, 5 KCl, 1 CaCl2, 1 MgCl2, 20 HEPES, and 10 d-glucose, pH 7.4.
All chemicals were obtained from Sigma Chemical. Pharmacological tools were prepared as 10 mM stock solutions, either in distilled water (trifluoroperazine), absolute EtOH (nifedipine, Y-27632, H89), or DMSO (W7, KN93, bis-indolylmaleimide). Aliquots were then added to the muscle baths; the final bath concentration of EtOH did not exceed 0.1%, which we have found elsewhere to have little or no effect on mechanical activity.
Constrictor responses were generally expressed as a percentage of the response to 60 mM KCl delivered during the equilibration period (i.e., before blockers were added), unless noted otherwise. All responses are reported as means ± SE; n refers to the number of animals. Statistical comparisons were made by Student's t-test (for single pairwise comparisons), one-way ANOVA (for multiple comparisons of mean values), or linear regression analysis, as indicated in the text; P < 0.05 was considered statistically significant.
Electrophysiological characteristics of voltage-dependent Ca2+ current in bovine TSM.
Many have previously described the voltage-dependent Ca2+ currents in TSM, finding these to be dihydropyridine sensitive or “L-type” in nature (11, 13, 14, 17, 24). Figure 1 summarizes the voltage dependence of activation and inactivation of these currents in our bovine tracheal myocytes. This current became progressively more activated at voltages more positive than −40 mV, becoming maximally activated at voltages at or exceeding 0 mV. At the same time, this current exhibited significant time-dependent inactivation (Fig. 1A), such that its magnitude was markedly decreased within a few seconds of membrane depolarization. Also, the inactivation was voltage dependent, being half-maximal and maximal at approximately −40 and 0 mV, respectively, producing a “window current” with a peak at approximately −20 mV (Fig. 1B).
Pharmacological characteristics of KCl-evoked contractions.
If one assumes the membrane to be permeable primarily to K+ and cytosolic [K+] to be 120 mM, the Nernst equation predicts that 30, 60, and 120 mM KCl will depolarize the membrane from a resting value of approximately −60 mV to approximately −40, −20, and −10 mV, respectively; actual measurements made decades ago corroborate these predictions (7). In view of the electrophysiological properties of the Ca2+ channels described above, it is clear that in an intact tissue that is partially depolarized by 60 mM KCl for many minutes or even hours, the resultant Ca2+ current would measure only a few pA in amplitude and would decrease further in magnitude as the tissue was more fully depolarized to 0 mV with 120 mM KCl. In other words, the concentration-response relationship for KCl is expected to be U-shaped rather than sigmoidal. It is further predicted that the responses to both 60 and 120 mM KCl should be completely abolished by nifedipine (a dihydropyridine blocker of L-type Ca2+ channels). We tested these predictions in intact tissues using the standard muscle bath technique.
Tissues were challenged with 0, 30, 60, 90, and 120 mM KCl in cumulative fashion. Neither adding KCl to the bathing medium directly nor replacing external NaCl with KCl is an ideal approach, since the former leads to osmotic changes, whereas the latter compromises Na+-dependent events such as Na+/Ca2+ exchange (leading to elevation of [Ca2+]i): we therefore compared both approaches. Figure 2 shows that the KCl concentration-response relationship obtained by hypertonic addition of KCl is sigmoidal, rather than U-shaped. Blocking voltage-dependent Ca2+ influx using nifedipine (10−6 M) markedly suppressed the responses but did not totally abolish them, particularly those at higher [KCl]: for example, approximately one-third of the mean response to 120 mM KCl remained in the presence of nifedipine (Fig. 2). This nifedipine-resistant component was apparently entirely due to osmotic effects, as it was mimicked by challenge with hyperosmolar sucrose (Fig. 2). Coapplication of the ROCK-selective inhibitor Y-27632 (5) or of cyclopiazonic acid (CPA) along with nifedipine had no additional inhibitory effect (Fig. 2): in fact, CPA significantly enhanced the response to 30 mM KCl, as we have described previously and found to be due to abolition of the superficial buffer barrier (19).
To test whether the contractions evoked by high [KCl] were caused in part by the change in Cl− equilibrium potential brought on by KCl addition, we examined the effect of the Cl− channel blocker niflumic acid (10−4 M) on KCl-evoked contractions. Niflumic acid had no effect on contractions evoked by 0–90 mM KCl but significantly reduced those evoked by 120 mM KCl (Fig. 2).
To further examine the extent to which osmotic changes contribute to the contractions evoked by KCl, we modified our experimental protocol such that the bathing medium was replaced with isotonic medium containing first 60 mM KCl and then 120 mM KCl (NaCl replaced with KCl). A representative tracing is given in Fig. 3A, and the mean responses are given in Fig. 3B. Using this approach, we did indeed find that tone actually decreased upon raising [KCl] from 60 to 120 mM, consistent with the predictions arising from our patch-clamp data above. The high [KCl]-evoked contractions under these conditions were abolished by either nifedipine (10−6 M) or Y-27632 (10−5 M) (Fig. 3).
Collectively, these data suggest that KCl stimulates ROCK activity at concentrations that are expected to depolarize the membrane to potentials that are insufficient for appreciable activation of voltage-dependent Ca2+ currents.
To test that the ROCK-dependent contractions were due to elevation of [Ca2+]i per se, we examined the Y-27632 sensitivity of contractions evoked by other agonists that act through elevation of [Ca2+]i alone, without involvement of G proteins. CPA (10−5 M) suppresses Ca2+ uptake into the internal store and thereby elevates [Ca2+]i (19, 22); A-23187 (3 × 10−5 M) is a Ca2+-specific ionophore that allows passive influx of external Ca2+. Both of these agents evoked contractile responses (peak values of 7.2 ± 1.5 g for CPA and 1.9 ± 0.6 g for A-23187, n = 9 for both), which were essentially abolished by 10−5 M Y-27632 (Fig. 4).
KCl directly stimulates Rho and ROCK activities.
The findings summarized above suggest that a rise in [Ca2+]i may lead in some way to activation of Rho and/or ROCK activity. We next examined this hypothesis directly using Western blot techniques. Tissues were incubated for 20 min with 0, 15, 30, 60, 90, or 120 mM KCl, flash frozen, and subjected to analysis for Rho and ROCK activation (see methods). We found the mean magnitudes of both activities were increased in a dose-dependent fashion by KCl in parallel with the mean changes in tone (Fig. 5). Figure 5 also shows a plot of the changes in Rho (Fig. 5B) or ROCK (Fig. 5C) activities vs. the magnitude of contractile responses for each individual tissue. Linear regression analysis of the relationships between Rho or ROCK activities and tone yielded R values of 0.55 and 0.72, respectively. Fitting of the data was greatly improved if sigmoidal (R = 0.88 and 0.89, respectively) or exponential models (R = 0.76 and 0.86, respectively) were used rather than a linear one, but we have no a priori reason to justify those types of comparisons. On the other hand, the linear fits were also greatly improved if we omitted those datum points in which Rho or ROCK activities were >200% above baseline (R = 0.69 and 0.82, respectively). The implications of this are discussed below.
KCl-induced changes in [Ca2+]i.
In addition to stimulating voltage-dependent Ca2+ influx, membrane depolarization may also enhance Ca2+-induced Ca2+ release from the sarcoplasmic reticulum (9), also resulting in elevation of [Ca2+]i and contraction. We examined whether this mechanism might account for the effect of KCl on [Ca2+]i and tone by comparing caffeine-triggered Ca2+ transients in cells incubated in Ringer medium comprising various [KCl].
In normal Ringer containing 5 mM KCl, caffeine evoked a large, spike-like elevation of [Ca2+]i that reached a peak within seconds and then fell back to a plateau level well above baseline until caffeine was washed off, as we have previously described in more detail (19, 22). When [KCl] was raised to 60 mM, there was also a statistically significant and sustained elevation in [Ca2+]i (Fig. 6A), which in general was smaller than that evoked by caffeine (Fig. 6, B and C); more importantly, subsequent application of caffeine again elicited a spike-like elevation that was not significantly different in magnitude compared with the control response evoked in 5 mM KCl. When [KCl] was further elevated to 120 mM, baseline [Ca2+]i tended to rise further, although this change was not statistically significant, nor was the caffeine-evoked transient evoked under these conditions statistically different in magnitude from the control response. Upon restoration of [KCl] to the normal value of 5 mM, baseline [Ca2+]i and the peak caffeine-evoked responses returned to control levels (Fig. 6B); there was no statistically significant difference between the first and final stimulation in 5 mM KCl, indicating that photobleaching was negligible.
Potential mechanisms underlying KCl-induced changes in Rho/ROCK activities.
In vascular smooth muscle, Ca2+-dependent stimulation of Rho/ROCK signaling is mediated by Ca2+/calmodulin-dependent CaM kinase II (29); alternatively, elevation of [Ca2+]i might stimulate PKC, which is also Ca2+ dependent. To test whether either of these pathways might be involved in the KCl-evoked responses we describe here, we examined the effects of the CaM kinase II inhibitor KN93 (3 × 10−5 M) or the PKC inhibitor bis-indolylmaleimide (5 × 10−6 M) on KCl-evoked contractions: tissues were incubated for 20 min with these blockers before the KCl concentration-response relationship was examined. Neither blocker had a statistically significant effect on this relationship (Fig. 7A). We also incubated tissues for 30 min in the presence of the CaM kinase II inhibitors trifluoroperazine (10−5 M) or W-7 (3 × 10−4 M) and then recorded the contractile response to 60 mM KCl for 20 min before flash-freezing and assaying Rho and ROCK activities. Neither inhibitor had a significant effect on KCl-induced Rho or ROCK activities nor on the magnitude of the contractions (Fig. 7B).
Alternatively, it may be that elevation of [Ca2+]i downregulates cAMP-dependent PKA activity, which removes an inhibitory influence on ROCK (6). We therefore assayed PKA activity in bovine tracheal tissues challenged with KCl (0–60 mM); some tissues challenged with 60 mM KCl were also pretreated with the PKA inhibitor H89 (10−5 M) or with the nonselective phosphodiesterase inhibitor isobutylmethylxanthine (IBMX, 10−4 M). As expected, PKA activity was markedly suppressed by H89 (10−5 M) and augmented by IBMX; more importantly, though, PKA activity was not affected whatsoever by KCl (Fig. 7C).
We have recently demonstrated that Rho/ROCK signaling in ASM is enhanced by 60 mM KCl (21) but did not examine in detail the relationships between membrane depolarization (using a wider range of [KCl]), elevation of [Ca2+]i, and Rho/ROCK activation. In this follow-up study, we examined constrictor responses evoked by 0–120 mM KCl to correlate the biochemical changes with the electrophysiological events. Not only were KCl-evoked contractions abolished by the dihydropyridine nifedipine (as expected) and by the ROCK-selective inhibitor Y-27632 (previously unexpected), but coadministration of Y-27632 and nifedipine had no greater effect than either agent alone, suggesting these two blockers act upon a common target or pathway. We also found Rho and ROCK activities per se were directly enhanced by KCl (Fig. 5) and that the concentration dependence of KCl-stimulated Rho/ROCK activities closely paralleled that of KCl-induced contraction and KCl-induced elevation of [Ca2+]i, as illustrated in Fig. 8. As noted above, linear regression analysis revealed a very close relationship between Rho or ROCK activities and tone (R values of 0.55 and 0.72, respectively), and this fitting was greatly improved (R = 0.69 and 0.82, respectively) if we omitted those tissues in which Rho/ROCK activities were increased >200% above baseline. This would suggest that only relatively modest changes in Rho/ROCK activities (less than one doubling) are sufficient for the effect on mechanical activity, and increases beyond this have very little additive effect.
Superimposed upon the data in Fig. 8, we have plotted the mean peak Ca2+ currents (from Fig. 1) measured at voltages ranging from −60 to 0 mV, which is the same range of membrane voltages produced by 0–120 mM KCl (7) (it is relevant to point out here that ASM is notable for not exhibiting action potentials under physiological conditions, and to depolarize to potentials much less than 0 mV following stimulation with bronchoconstrictor agents). Plotting the data in this way highlights the relationship between voltage-dependent Ca2+ currents and the KCl-evoked elevation of [Ca2+]i. Many have described the electrophysiological characteristics of the voltage-dependent Ca2+ current in ASM, finding it to have L-type properties (11, 17, 24), including marked inactivation that is time, voltage, and Ca2+ dependent. As such, although it is possible to demonstrate substantial voltage-dependent Ca2+ currents for a couple hundred milliseconds in a very artificial setting (a cell held under voltage clamp conditions and immediately depolarized to positive membrane potentials), voltage-dependent Ca2+ influx is actually quite small in the more physiologically relevant situation in which a cell is depolarized to subthreshold potentials for minutes or hours. Despite the small size of this persistent inward current, it is clearly sufficient to elevate [Ca2+]i (Fig. 6) and evoke a contractile response (Fig. 8). A steady inward Ca2+ current of only 3 pA (equivalent to 3 × 10−17 mol of Ca2+ per second) in a typical ASM cell with a volume of 3 pl (Drs. C. Seow and A. Herrera, University of British Columbia, unpublished communication) would result in a change in [Ca2+]i of 10−5 M if the Ca2+ were uniformly distributed and not subject to Ca2+-homeostatic mechanisms such as extrusion out of the cell by the plasmalemmal pump and Na+/Ca2+ exchange, sequestration into the sarcoplasmic reticulum and mitochondria, and buffering by cytosolic proteins.
Other sources of Ca2+ appear to also be sufficient for activation of Rho, since we found that constrictor responses to agents that elevate [Ca2+]i by pathways independent of G protein-coupled receptors, namely the SERCA-pump inhibitor CPA and the Ca2+-ionophore A-23187, were also largely abolished by Y-27632. Altogether, the data suggest that a rise in [Ca2+]i alone leads in some way to activation of Rho and ROCK activities. This phenomenon has been described recently in vascular smooth muscle and attributed to stimulation of Ca2+/calmodulin-dependent CaM kinase II (27, 29, 33). We explored several possible mechanisms by which Rho activity in ASM might be coupled to [Ca2+]i, including stimulation of CaM kinase II or PKC, as well as removal of the inhibitory influence of cAMP-dependent protein kinase. Our data rule out all three of these possibilities as accounting for the KCl-induced change in Rho activity. Thus the coupling mechanism underlying KCl-induced Rho-activation in ASM is as yet unclear.
Another possibility that remains to be explored is that Rho activation per se is in part voltage dependent. For example, depolarization may in some way facilitate the translocation of Rho to the membrane and/or the interactions between the different components of this signaling cascade. Alternatively, it may be that there is a direct interaction between voltage-dependent ion channels and Rho. That is, the conformation of many ion channels is known to undergo rearrangement in response to changes in membrane potential: it might be that rearrangement of some cytosolic portion of the channel might stimulate Rho (or some other signaling molecule upstream from it). Inactivation of those ion channels may not interfere with this interaction, as inactivation typically involves conformational changes distinct from those producing activation. Data from rat ureter foretell a direct interaction between ROCK and the voltage-dependent Ca2+ channels (30). We feel it is now necessary, therefore, to carry out experiments using techniques that can directly probe the interactions between various ion channels and Rho/ROCK.
We have argued previously that ASM is distinct from vascular and gastrointestinal smooth muscles in that electromechanical coupling contributes very little to the responses to physiologically relevant autacoids, such as cholinergic agonists (18). This contention is based in part on the observation that membrane potentials in ASM rarely reach levels necessary for substantial Ca2+ current activation, even during stimulation with half-maximally effective concentrations of agonist. However, these agonists also stimulate G12/G13, which are known to directly activate Rho. It may be that G protein-mediated and Ca2+ channel-mediated stimulation of Rho activity are synergistic. This could shed light on the extremely puzzling observation that voltage-dependent Ca2+ channel inhibitors partially suppress the contractile responses to low concentrations of cholinergic agonist (3, 7, 8, 23) despite the fact that such modest cholinergic stimulation results in membrane depolarization of only a few millivolts above baseline (which in turn is far more negative than the voltages required for even marginal Ca2+ influx).
In conclusion, we have compared the effects of membrane depolarization on Ca2+ currents, [Ca2+]i, Rho/ROCK activities, and contractile tone in bovine ASM. Rho and ROCK activities are elevated during membrane depolarization, apparently in response to the elevation of [Ca2+]i. The coupling between membrane depolarization and Rho activation is still unclear, though we have ruled out involvement of CaM kinase II, as well as PKA and PKC. Instead, it may involve some form of interaction between Rho and the activated conformation of the voltage-dependent Ca2+ channels.
These studies were supported by a Career Award from the Canadian Institutes of Health Research, as well as operating grants from the Canadian Institutes of Health Research, the Ontario Thoracic Society, and Astra-Zeneca Canada.
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
- Copyright © 2005 the American Physiological Society