Mechanosensitive nonselective cation channel facilitation by endothelin-1 is regulated by protein kinase C in arterial myocytes

Abstract

Objective The mechanosensitive nonselective cation channel (NSC_MS) and endothelin-1 (ET-1) play critical roles in the regulation of vascular tone. This study was undertaken to investigate the effect of ET-1 on NSC_MS and on the myogenic response of arteries.

Methods Cell-attached patch-clamp techniques were applied to rabbit pulmonary and cerebral arterial smooth muscle cells using a 140 mM CsCl pipette and bath solutions (Ca²⁺-free, 1 mM EGTA). Myogenic responses were determined by video analysis of pressurized arteries.

Results The application of negative pressures through the pipette activated NSC_MS, and this was augmented by bath application of ET-1 (1 pM–30 nM). ET-1 lowered the lowest pressure required for NSC_MS activation. NSC_MS facilitation by ET-1 was prevented by BQ-123 (1 μM, an ET_A antagonist) but not by BQ-788 (1 μM, an ET_B antagonist). Phorbol 12-myristate 13-acetate (PMA, 100 nM), a protein kinase C activator, also increased the activity of NSC_MS. ET-1- or PMA-induced facilitation of NSC_MS was abolished by GF109203X (10 μM), a protein kinase C inhibitor. Video analysis of pressurized cerebral artery showed inhibition of the myogenic response by the NSC_MS channel blockers GsMTx-4 (5 μM) and DIDS (3–100 μM). Treatment with ET-1 (10 pM) augmented the myogenic response and this was inhibited by DIDS (30 μM).

Conclusion Stimulation of ET-1 receptor (ET_A) facilitates NSC_MS via a protein kinase C-dependent signaling pathway in rabbit arterial myocytes. Our findings suggest that NSC_MS play a role in the myogenic response and its augmentation by ET-1.

This article is referred to in the Editorial by Coleman and Parkington (pages 197–198) in this issue.

1 Introduction

In many organs, blood flow is held at a relatively constant level when perfusion pressure changes substantially [1]. This autoregulation of blood flow is achieved in several ways including the myogenic contraction triggered by mechanical stretch of vascular smooth muscle [2]. Since vascular smooth muscle cells are depolarized as luminal pressure increases, the existence of mechanosensitive nonselective cation channels (NSC_MS) provides a plausible explanation to the myogenic response. Actually, NSC_MS have been found in a wide variety of cells including vascular myocytes [3–6]. Our previous studies showed that NSC_MS are more abundant in pulmonary arterial smooth muscle cells (PASMCs) than systemic arterial myocytes of rabbits [7].

Endothelin-1 (ET-1) is a peptide hormone released from endothelium and vascular smooth muscle cells, and it is a highly potent vasoconstrictor [8,9]. ET-1 also induces mitogenic and vascular remodeling effects, which are critically required for several pathophysiological mechanisms [9–14]. The release and expression of ET-1 are triggered by various physicochemical stimuli, such as mechanical stress and hypoxia [15–19]. Moreover, it has been argued that ET-1 dependent signaling is critical for hypoxia-induced pulmonary vasoconstriction and chronic hypoxia-induced pulmonary hypertension [12,18,19]. In vascular myocytes, stimulation of the two known ET-1 receptors, ET_A and ET_B[20], increases cytosolic Ca²⁺ concentration ([Ca²⁺]_c) via various ways which include phospholipase C (PLC)/InsP₃-dependent release of Ca²⁺ from intracellular stores and the activation of Ca²⁺ permeable nonselective cation channels [20–26].

In addition to the direct vasoconstrictive effects, ET-1 also enhances myogenic vascular tone [10,13,27–29]. Because NSC_MS may play a critical role in myogenic autoregulation, it might be expected that NSC_MS are activated by ET-1. However, no previous study has investigated the effect of ET-1 on NSC_MS. Moreover, it is unclear whether the augmented myogenic contraction by ET-1 is mediated by activation of NSC_MS. Therefore, in the present study, we examined whether ET-1 modulates NSC_MS in rabbit pulmonary and cerebral arterial myocytes. Also, we sought to identify the relevant intracellular signaling pathway and the relation between these actions of ET-1 on NSC_MS and the myogenic response of arteries.

2 Materials and methods

2.1 Preparation of smooth muscle cells

New Zealand White rabbits (1.5–2.0 kg, n=75) of either sex were used for this study, and the investigation conforms with the Guide for the Care and Use of Laboratory Animals published by the US National Institutes of Health (NIH Publication No. 85-23, revised 1996). Rabbits were fully anaesthetized with sodium pentobarbital (50 mg/kg) and simultaneously injected with heparin (100 U/kg). Pulmonary arteries (outer diameters between 0.5 and 1 mm), the main trunk of basilar arteries, and the second and third branches of middle cerebral arteries were used as sources of vascular smooth muscles. After arteries had been dissected and cut into small pieces, they were incubated in 1 mL of first digestion medium (Ca²⁺-free Tyrode's solution containing papain 1 mg/mL) for 10–15 min and in second digestion medium (Ca²⁺-free Tyrode's solution containing collagenase 2.8 mg/mL) for 10–20 min; both of the two digestion mediums contained serum albumin (1.5 mg/mL) and dithiothreitol (1.5 mg/mL). Cells were isolated from tissue strips by gentle agitation using a fire-polished glass pipette in high K⁺- and low Cl⁻ storage medium. Isolated cells were maintained in the storage medium (see below) at 4 °C until required.

2.2 Experimental solutions

Ca²⁺-free normal Tyrode's solution contained (in mM): NaCl 143, KCl 5.4, MgCl₂ 1, glucose 5.5, and HEPES 5 adjusted to pH 7.4 with NaOH. High K⁺/low Cl⁻ storage medium contained (in mM): KCl 50, l-glutamate 50, KH₂PO₄ 20, taurine 20, MgCl₂ 3, glucose 20, HEPES 10, and EGTA 0.5, adjusted to pH 7.3 with KOH. The bath and pipette solutions for single-channel recording contained (in mM): CsCl 140, HEPES 10, EGTA 2 and MgCl₂ 1, adjusted to pH 7.4 with CsOH. Cells were immersed in CsCl bath solution for at least 30 min before performing electrophysiological recording. The bath used for patch recording was superfused by hydrostatic pressure (50 cm H₂O) when a drug application was required. The bath solution for the pressurized artery system contained (in mM): NaCl 140, KCl 5.4, NaH₂PO₄ 0.33, glucose 10, HEPES 10, CaCl₂ 1.8, and MgCl₂ 1 adjusted to pH 7.4 with NaOH.

2.3 Electrophysiological recording

Recordings were obtained in the cell-attach configurations using a patch-clamp amplifier (Axopatch 200A, Axon Instruments, Foster City, CA). Gigaseals were formed using Sylgard coated thin-walled borosilicate capillaries (Clark Electromedical Instruments, Pangbourne, UK). The negative pressure required to obtain gigaseals was applied minimally only during the initial phase of patch-clamp recording. Throughout the experiment, pipette pressure was equilibrated versus ambient pressure (i.e. 0 mm Hg). To activate NSC_MS, negative pressure was applied through a patch pipette using a syringe connected to a water-filled meter [6,7]. Pressures (in cm H₂O) were calibrated using a manometer and later converted into mm Hg. Electrical tip resistance of patch pipettes filled with experimental solution ranged between 2.3 and 2.5 MΩ. Care was taken to maintain a constant tip resistance since the size of the pipette tip affected cell membrane area and the degree of stretch induced by negative pressure. If not specifically mentioned, all NSC_MS recordings were performed using a standard cell-attached patch-clamp setup (+50 mV, −15 mm Hg) at room temperature (22–24 °C). Polarities of clamp voltage refer to values on the cytoplasmic side. Recorded signals were filtered at 1 kHz and transferred to a computer using the Digidata 1200 interface (Axon Instruments, Foster City, CA) at a sampling rate of 5 kHz. Continuous single-channel currents were analyzed using the pCLAMP program (version 9, Axon Instruments). Channel activities are presented as NP_o values, which was determined from continuous recordings of during ca. 1 min. Amplitudes of the single-channel currents at each membrane potential were obtained from all-point histograms.

2.4 Video analysis of pressurized arteries

Small arteries (250–350 μm, i.d., passive) was identified using a Nikon SMZ645 (Tokyo, Japan) stereo-zoom microscope. The dissected segments of small arteries were placed in a glass-bottomed vessel chamber (Model CH/1/SH, Living System Instrument, Burlington, VT, USA). The vessel chamber contained a HEPES-buffered physiological salt solution (pH 7.40±0.03; 140 mmol/L NaCl, 5.4 mmol/L KCl, 1.8 mmol/L CaCl₂, 1 mmol/L MgCl₂, 0.33 mmol/L NaH₂PO₄, 10 mmol/L glucose and 10 mmol/L HEPES), and its temperature was set at 37 °C by using a temperature controller (Model TC-01, Living Systems Instrument). Two pipettes were utilized to cannulate the vessel that was secured with 12-0 suture. To evaluate pressure-induced myogenic response, the luminal flow was maintained at zero. After confirming vessel integrity, intraluminal pressure was set to 40 mm Hg for at least an additional 60 min; spontaneous tone usually developed within this time frame. Vessel dimensions were measured using a video dimension analyzer (Model V94, Living Systems Instrument) and a data acquisition system (Digidata 1200 and Axoscope, Axon Instruments).

To assess myogenic behavior, internal diameters were measured during successive 20 mm Hg increments in luminal pressure over the range 20–100 mm Hg. After each step change in pressure, arteriolar diameter usually attained a steady state within 2 to 3 min. After a series of step changes, luminal pressure was returned to 40 mm Hg and the vessel concerned was allowed to re-equilibrate for a minimum of 5 min. At the end of each experiment, passive diameter measurements were recorded at each of the above pressures in 0 Ca²⁺ PSS containing 10 μM sodium nitroprusside. For pulmonary arteries, outer diameter was measured instead of inner diameter because the edge detector of our video analysis did not stably discriminate the margin of luminal wall of pulmonary arteries.

2.5 Drugs and chemicals

All drugs and chemicals except GsMTx-4 peptides were purchased from Sigma Chemical (St. Louis, MO). GsMTx-4 and its negative analogue were synthesized as reported previously [30].

2.6 Statistical analysis

Results are expressed as mean±S.E.M. Statistical significance was evaluated using the paired or unpaired Student t-test (P-value<0.05). N numbers in the Results section and in the figure legends indicate numbers of cells or vessels tested. The presence of NSC_MS and their regulation by tested chemicals were consistently observed in this study, and were not confined to a particular animal. The numbers of rabbits sacrificed for each statistical summary of patch-clamp study were ten (Fig. 1), five (Fig. 2), seven (Fig. 3C,D), six (Fig. 3E,F), nine (Fig. 4C,D), seven (Fig. 4E,F), four (Fig. 7B, basilar artery), and three (Fig. 7B, cerebral artery), respectively.

3 Results

3.1 Recording of NSC_MS in arterial myocytes

Fig. 1 demonstrates representative current traces of NSC_MS and response to ET-1 recorded in PASMCs. At a clamp voltage of 50 mV, negative pressure-induced single-channel currents with amplitude of about 1.5–1.8 pA, which corresponds to the known unitary conductance of NSC_MS (30–33 pS) [16,17]. In the present study, the probability of recording NSC_MS was about 50% in the PASMCs derived from the 3rd to 4th branch vessels (ϕ, 0.5–1.0 mm). In all patches confirmed to have NSC_MS, no spontaneous activity was observed under control condition, i.e. without applying negative pressure. The effect of stretch was immediate and reversible: channel opening was turned on and off by applying and releasing pressure without a noticeable delay. As the magnitude of negative pressure increased, channel opening became more frequent and the number of open channels increased (Fig. 1A, see also Fig. 2).

NSC_MS were inhibited by GsMTx-4, the venom of the tarantula Grammostola spatulata which possesses specific extracellular inhibitory effects on stretch-activated channels [31]. In c–a patches, GsMTx-4 was included in the pipette solution and stretch was applied as soon as the gigaseal has been formed. As demonstrated in Fig. 1B, when 3.5 μM GsMTX-4 was present in pipette solution, the initially observed stretch-activated channels disappeared permanently within 20–30 s after a gigaseal was formed. Similar results were obtained in three cases of c–a patch.

3.2 Effects of ET-1 on NSC_MS

ET-1 (30 nM) increased NSC_MS activity within 1–2 min of bath application, and this facilitation was maintained stably, and almost reversed by washout (Fig. 1C,E). In the absence of negative pressure, ET-1 did not activate any channels in the c–a condition even though NSC_MS were present in the same membrane patch (Fig. 1D, n=33). Increases in NP_o by 30 nM ET-1 at −15 mm Hg are summarized in Fig. 1E (n=14). NSC_MS facilitation in the presence of negative pressure was observed from ET-1 concentrations of 1–100 pM (n=3–7 at each concentration, Fig. 1F). Although not shown, NSC_MS facilitation by ET-1 was observed at both positive and negative membrane voltages and the slope conductance of NSC_MS (33 pS) was not changed by ET-1.

The mechanosensitivities of NSC_MS were compared in the presence and absence of ET-1 (Fig. 2A,B). Without ET-1, NSC_MS activity was started to be observed from −7.5 or −15 mm Hg, but the application of 30 nM ET-1 sensitized NSC_MS to lower negative pressures (−1.5 or −4.5 mm Hg). Interestingly, ET-1-induced NSC_MS activity was less prominent at higher negative pressures (−22.5 and −30 mm Hg), which suggests that the effects of ET-1 are not simply additive. To elucidate whether sensitization of NSC_MS sensitization to membrane stretching was also induced by lower concentrations, the apparent threshold pressure for activation was compared in the presence and absence of 1 pM ET-1. Interestingly, the lowest pressure required for NSC_MS activation was found to decrease from 8.4±2.45 to 3.9±1.12 mm Hg (n=5) on adding 1 pM ET-1 (Fig. 2B,C).

Two types of ET-1 receptors, ET_A and ET_B, are specifically blocked by BQ-123 and BQ-788, respectively [9]. After confirming NSC_MS facilitation by ET-1, we added 1 μM BQ-123 to bath solution, and found that NSC_MS activity was reduced to the control level (Fig. 3A,C). Also, in the presence of BQ-123, ET-1 addition had no effect on NSC_MS (Fig. 3B,D). In contrast, BQ-788 had no effect on ET-1 induced NSC_MS facilitation (Fig. 3E,F).

It is known that ET_A stimulation activates phospholipase C and protein kinase C (PKC) signaling cascades [9]. To test whether the effects of ET-1 are mediated via a PKC pathway, GF 109203X (10 μM, an inhibitor of PKC), was applied before or after the ET-1 treatment. As was observed for BQ-123, the application of GF 109203X (10 μM) in the presence of ET-1 returned NSC_MS activity to the control level (Fig. 4A,C, n=4). Also, pretreatment with GF 109203X prevented NSC_MS facilitation by ET-1 (Fig. 4B,D; n=4). Further supporting the involvement of PKC, the application of 100 nM phorbol 12-myristate 13-acetate (PMA, an activator of PKC) also facilitated NSC_MS activity, and this activation by PMA was blocked by 10 μM GF 109203X (Supplementary Fig. 1). Responses to PMA and GF 109203X are summarized in Fig. 4E and F as bar graphs.

Because PLC activation via ET-1 receptors could also lead to stored Ca²⁺ release, we wondered whether the increased [Ca²⁺]_c, in addition to PKC, might modulate NSC_MS, or whether it might independently activate Ca²⁺-dependent nonselective cation channels. However, experimental conditions in the present study were designed to minimize the [Ca²⁺]_c changes; myocytes were immersed in Ca²⁺-free CsCl solution (1 mM EGTA) for more than 20 min. Actually, when the [Ca²⁺]_c of PASMCs under the above condition was measured by Fura-2 fluorometry, it was found that ET-1 (30 nM) did not increase [Ca²⁺]_c (n=6, Fig. 5). In each case of this experiment, the large increase in [Ca²⁺]_c by ionomycin in the presence of 10 mM CaCl₂ was confirmed as a positive control.

3.3 Myogenic response and NSC_MS

To investigate the role of NSC_MS and their modulation by ET-1 in terms of myogenic response in blood vessels, the change of inner diameter was measured in pressurized cerebral arteries. Arterial diameters were either unchanged or reduced in response to luminal pressure increases, indicating the presence of myogenic constriction. The application of GsMTx4 (5 μM) attenuated, but did not completely abolish this response in cerebral arteries (n=4, Fig. 6A,C). Our previous study [6] and the result shown below (Fig. 7A) show that DIDS blocks NSC_MS. Consistent with the suspected role of NSC_MS, the myogenic response was also inhibited by DIDS (Fig. 6B,D). Interestingly, the inhibitions by GsMTx-4 and DIDS were more prominent at the relatively low pressures, and the slow development of myogenic response was less affected by these blockers. When various concentrations were tested, DIDS showed more potent inhibitory effects on myogenic responses at lower luminal pressure range (Fig. 6E). Due to the high cost of experimentation with GsMTx-4, DIDS was used as a NSC_MS blocker in the following experiments.

The effect of ET-1 on NSC_MS was also examined in basilar and middle cerebral arterial myocytes (Fig. 7A,B). The blocking effects of DIDS on NSC_MS in cerebral arterial myocytes were also confirmed (n=3, Fig. 7A). To compare responses with those of pulmonary artery, PASMCs were isolated on the same date and the effect of ET-1 was examined. As reported previously [7], control NSC_MS activity was higher in PASMCs than in basilar and cerebral arteries (Fig. 7B). However, ET-1 (30 nM) also increased the activity of NSC_MS in basilar artery (n=4) and middle cerebral arteries (n=3), and this effect was reduced by PKC inhibitor (data not shown here, n=2).

The application of 10 pM ET-1 did not significantly reduce the diameter of middle cerebral artery at 20 mm Hg but augmented myogenic constriction at higher luminal pressures (Fig. 7C), and the subsequent addition of DIDS reversed this augmented myogenic response (Fig. 7C,D). In contrast with its effects on myogenic responses, treatment with 30 μM DIDS did not inhibit vasoconstriction induced by high KCl (60 mM, Fig. 7E,F). The lack of an effect of DIDS on KCl-induced constriction indicates that it does not directly inhibit L-type Ca²⁺ channels or contractile molecules in vascular smooth muscle.

Since the distribution of NSC_MS and their mechanosensitivities are higher in PASMCs than in systemic arteries, we also investigated pulmonary artery myogenic response. Unfortunately, unlike that observed for cerebral arteries, typical myogenic responses were not observed at luminal pressures between 5 and 25 mm Hg (physiological pulmonary arterial pressures). In fact, a slow decrease in diameter in response to raised luminal pressure was observed in only two of 25 trials (Supplementary Fig. 2). We also tested whether repetitive increases in luminal pressure from 5 to 15 mm Hg induced myogenic responses. In four of six pulmonary arteries, the application of ET-1 (100 pM) decreased inner diameters at 5 and 15 mm Hg, and this was subsequently reversed by DIDS (30 μM, Fig. 8). The other two arteries showed only passive dilation either in the absence or presence of ET-1 or DIDS (data not shown).

4 Discussion

The present study demonstrates that the mechanosensitivity of NSC_MS in rabbit arterial myocytes is increased by ET-1 via ET_A receptors in a manner that involves the PKC-dependent signaling pathway. The pressure–activity relation of NSC_MS was shifted to the left by ET-1 (Fig. 2A). Also, cerebral artery myogenic tone was augmented by ET-1 and NSC_MS blockers attenuated myogenic tone under control conditions and that augmented by ET-1. Previous studies have shown that ET-1 modulates various kinds of ion channels, e.g. it inhibits K_ATP and K_v channels [32–34], and stimulates L-type Ca²⁺ channels, Cl⁻ channels, and nonselective cation channels [23,35–39]. These changes commonly imply that ET-1 has a depolarizing effect on vascular smooth muscle cells. Moreover, the facilitation of NSC_MS by ET-1 would also be expected to depolarize arterial myocytes especially in vivo where cells are exposed to continuous mechanical stimuli.

Because stimulation with ET-1 is known to activate nonselective cation channels in vascular myocytes via signaling cascades like phospholipase C [37–39], we were initially concerned as to whether increased channel activity reflects NSC_MS facilitation or an additive activation of receptor-operated cation channels. However, the present study demonstrates that the application of ET-1 alone does not activate single-channel currents (Fig. 1D). In contrast, once the presence of NSC_MS has been confirmed in test membrane patches, channel activity was always increased by ET-1 (n=33). Therefore, we believe that the present study demonstrates genuine facilitation of NSC_MS by ET-1. Nonetheless, considering previous report on ET-1 activated cationic currents in whole-cell recordings [37–39], the absence of channel activation by ET-1 alone in the present study is difficult to explain. The Ca²⁺-free condition might be supposed as a factor suppressing the activation of other cation channels by ET-1.

The effects of BQ-123 and of PKC agonist and antagonist on the modulation of NSC_MS by ET-1 indicate that ET_A receptors mediate NSC_MS facilitation via a PKC-dependent signaling pathway. Similarly, ATP-sensitive K⁺ channels are also inhibited by ET_A via PKC-dependent signaling in rabbit PASMCs and coronary arterial myocytes [33] whereas PLC-InsP₃ dependent Ca²⁺ release is mainly mediated by ET_B. The expression of both ET_A and ET_B in the rabbit PASMCs has been confirmed by an immunoblot assay [26]. Therefore, ET-1 receptor subtypes appear to regulate different targets in vascular myocytes.

Although the present study demonstrates PKC involvement in NSC_MS facilitation by ET-1, it does not determine how PKC-dependent signaling, i.e. protein phosphorylation, actually changes NSC_MS mechanosensitivity. Since the molecular nature of NSC_MS in vascular myocytes remains controversial, we were unable to target a specific site for PKC phosphorylation. TRPM4 was recently proposed to be a mechanosensitive cation channel protein responsible for the myogenic response, and is activated by PKC [40]. TRPM4 was originally considered as a molecular candidate of Ca²⁺-activated nonselective cation channel, and the PKC-dependent phosphorylation of TRPM4 was reported to sensitize it to intracellular Ca²⁺[41]. However, in the present study, NSC_MS were well-demonstrated under Ca²⁺-free conditions, which is inconsistent with the key property of TRPM4 as a Ca²⁺-activated channel [40,41]. TRPC6 has also been proposed as a NSC_MS in vascular myocytes, and is known to be positively modulated by diacylglycerol [42], similar with the property of NSC_MS reported by us [6]. However, the activation of PKC does not stimulate but rather desensitizes TRPC6 [43]. Therefore, none of the currently proposed candidates well-match the properties of NSC_MS, and the precise mechanism of mechanosensitive activation and its facilitation by PKC remain to be determined.

NSC_MS are regarded to be probably involved in the mechanism underlying arterial myogenic tone. In this respect, NSC_MS facilitation might have implications in terms of increased myogenic response by ET-1 [10,13,28]. Since the inner diameter of the patch pipette was around 1 μm, when stretched by −15 mm Hg of pressure, a hemispheric plasma membrane in the pipette tip would be under the tension of 20 dyn/cm or higher depending on the geometry of a patch membrane. The reported wall tension of arteries for the study of myogenic response is around several hundreds of dyn/cm [44], and this seems to be sufficient to activate NSC_MS. However, in the intact vessel, the wall tension is more likely to be transmitted by extracellular matrix and intracellular cytoskeletons rather than by direct stretch of plasma membranes. Therefore, one should be cautious to correlate the results of isolated myocytes with intact vessel study.

The present study demonstrates that myogenic responses shown by rabbit cerebral arteries and their augmentation by ET-1 are suppressed by NSC_MS blockers, although the drug effects were incomplete. Interestingly, the myogenic responses were found to be generally more sensitive to NSC_MS blockers at relatively low pressure ranges (Fig. 6). These results suggest that mechanism(s) other than the NSC_MS could be also involved in the myogenic responses, especially at higher luminal pressure. Further investigation is required to determine the NSC_MS-independent component of myogenic responses.

In contrast with the clear demonstration of myogenic responses in cerebral arteries, changes in luminal pressure with physiological range rarely induced myogenic response-like diameter decrease in pulmonary arteries. In fact, the presence of myogenic response in pulmonary arteries from normal adult animal is controversial [45–47]. In this respect, it is not likely that the NSC_MS are active under the low luminal pressures of pulmonary circulation. However, the inhibition of ET-1 induced contraction by DIDS suggests that the facilitated NSC_MS might participate, at least to a partial extent, in the agonist-induced vascular tone (Fig. 8).

Although typical myogenic responses are difficult to show, the majority of patch-clamp experiments have been performed in PASMCs because they have higher NSC_MS densities and mechanosensitivities than systemic arterial myocytes [7]. In addition to this practical advantage, ET-1 is known to play a crucial role in the pathophysiology of the pulmonary circulation [11,12,48]. In view of ET-1 release under hypoxic conditions, our study suggests that NSC_MS might participate in the hypoxia-induced changes of pulmonary arteries. The pulmonary arteries from pulmonary hypertensive animals or under chronic hypoxia started to show myogenic responses and altered ion channel expression [12,48,49]. In this respect, further investigations using specific pharmacological tools and the elucidation of the molecular identity of NSC_MS are required to determine the roles of NSC_MS in the pulmonary circulation and in classical myogenic responses of systemic arteries.

Appendix A. Supplementary data

Supplementary data associated with this article can be found, in the online version, at doi:10.1016/j.cardiores.2007.06.021.

Acknowledgement

This work was supported by the Korea Research Foundation funded by the Korean Government (MOEHRD, Basic Research Promotion Fund, KRF-2006-E00021).

References

Author notes