Smooth muscle membrane potential modulates endothelium-dependent relaxation of rat basilar artery via myo-endothelial gap junctions

Jou

rnal

of P

hysi

olog

y

Endothelium-dependent relaxation of vascular smoothmuscle has been attributed to release of nitric oxide (NO),prostaglandin I2 (prostacyclin), and a third mechanisminvolving hyperpolarization by as yet ill-defined factor(s)and/or gap junctional communication from the endo-thelium to smooth muscle (Fltou & Vanhoutte, 1999).The release of endothelium-derived NO by agonists, suchas acetylcholine (ACh), is dependent on a rise inintracellular calcium levels ([Ca2+]i) within endothelialcells and activation of NO synthase (NOS). The entry ofCa2+ from the extracellular space is required to maintainproduction of NO (Lckhoff et al. 1988; Kruse et al. 1994)and several lines of evidence indicate that endothelial cellmembrane potential plays a critical role in the regulationof Ca2+ entry (Nilius & Droogmans, 2001). Ca2+ entry is not

dependent on voltage-gated Ca2+ channel activity, rather itoccurs via a nonselective cation pathway and is dependenton membrane hyperpolarization to provide an inwardlydirected electrochemical driving force for Ca2+ movement(Busse et al. 1988; Lckhoff & Busse, 1990a; Kamouchi etal. 1999; Nilius & Droogmans, 2001). Significantly, severalstudies show that membrane depolarization inhibits bothagonist-induced increases in [Ca2+]i and NO release fromcultured endothelial cells (Adams et al. 1989; Schilling,1989; Lckhoff & Busse, 1990b). However, to date, thereare no data available concerning the influence of vascularsmooth muscle membrane potential on endothelial cellfunction within intact arteries.

Following its release from the endothelium, NO is thoughtto cause vasorelaxation, at least in part, by affecting the

Smooth muscle membrane potential modulatesendothelium-dependent relaxation of rat basilar artery viamyo-endothelial gap junctionsTracy Allen, Mircea Iftinca, William C. Cole and Frances PlaneThe Smooth Muscle Research Group, Canadian Institutes of Health Research Group in Regulation of Vascular Contractility, University of Calgary,Alberta, Canada

The release of endothelium-derived relaxing factors, such as nitric oxide (NO), is dependent on anincrease in intracellular calcium levels ([Ca2+]i) within endothelial cells. Endothelial cell membranepotential plays a critical role in the regulation of [Ca2+]i in that calcium influx from the extracellularspace is dependent on membrane hyperpolarization. In this study, the effect of inhibition ofvascular smooth muscle delayed rectifier K+ (KDR) channels by 4-aminopyridine (4-AP) onendothelium-dependent relaxation of rat basilar artery to acetylcholine (ACh) was assessed. ACh-evoked endothelium-dependent relaxations were inhibited by N-(V)-nitro-L-arginine (L-NNA) or1H-[1,2,4]oxadiazolo[4,3-a]quinoxalin-1-one (ODQ), confirming a role for NO and guanylylcyclase. 4-AP (300 mM) also suppressed ACh-induced relaxation, with the maximal responsereduced from ~92 to ~33 % (n = 11; P < 0.01). However, relaxations in response to exogenous NO,applied in the form of authentic NO, sodium nitroprusside or diethylamineNONOate(DEANONOate), were not affected by 4-AP treatment (n = 311). These data are not consistentwith the view that 4-AP-sensitive KDR channels are mediators of vascular hyperpolarization andrelaxation in response to endothelium-derived NO. Inhibition of ACh-evoked relaxation by 4-APwas reversed by pinacidil (0.51 mM; n = 5) or 18b-glycyrrhetinic acid (18bGA; 5 mM; n = 5),indicating that depolarization and electrical coupling of the smooth muscle to the endotheliumwere involved. 4-AP caused depolarization of both endothelial and vascular smooth muscle cells ofisolated segments of basilar artery (mean change 11 1 and 9 2 mV, respectively; n = 15).Significantly, 18bGA almost completely prevented the depolarization of endothelial cells (n = 6),but not smooth muscle cells (n = 6) by 4-AP. ACh-induced hyperpolarization of endothelium andsmooth muscle cells was also reduced by 4-AP, but this inhibition was not observed in the combinedpresence of 4-AP and 18bGA. These data indicate that 4-AP can induce an indirect inhibition ofendothelium-dependent relaxation in the rat basilar artery by electrical coupling of smooth musclemembrane depolarization to the endothelium via myo-endothelial gap junctions.

(Resubmitted 30 August 2002; accepted 26 September 2002; first published online 8 November 2002)Corresponding author W. C. Cole: The Smooth Muscle Research Group, Faculty of Medicine, University of Calgary,3330 Hospital Drive, N.W., Calgary, Alberta, Canada T2N 4N1. Email: [email protected]

Journal of Physiology (2002), 545.3, pp. 975986 DOI: 10.1113/jphysiol.2002.031823 The Physiological Society 2002 www.jphysiol.org

Jou

rnal

of P

hysi

olog

y

activity of K+ channels within vascular smooth muscle cellsto elicit hyperpolarization and thereby a decline in L-typeCa2+ channel activity, [Ca2+]i and tone development(reviewed by Fltou & Vanhoutte, 1999). There isconsiderable evidence that NO stimulates large conduct-ance Ca2+-activated K+ (BKCa) channel activity of vascularsmooth muscle cells (Robertson et al. 1993) due to: (1)direct effects of NO on channel gating (Lang et al. 2000);(2) phosphorylation by cGMP-dependent protein kinase(PKG; Robertson et al. 1993); and/or (3) modulation of SRCa2+ release leading to an increased frequency of Ca2+

sparks (Porter et al. 1998). However, other K+ channeltypes, including ATP-sensitive K+ channels (Murphy &Brayden, 1995) and delayed rectifier K+ (KDR) channels(Zhao et al. 1997; Sobey & Faraci, 1999; Lovren & Triggle,2000), have also been suggested to play a role in mediatingendothelium-dependent vasorelaxation due to NO. Forexample, in the rat basilar artery in vivo, endothelium-dependent relaxations in response to ACh were inhibitedby 4-AP (200 mM), and thus a role for KDR channels inmediating the effects of endothelium-derived NO wasproposed (Sobey & Faraci, 1999). The stimulation of KDRchannel activity was postulated to be due to a signallingcascade involving activation of guanylyl cyclase, cGMP,PKG and phosphorylation of the channels or associatedmodulatory subunits (Sobey & Faraci, 1999). Previousstudies provide compelling evidence for the modulation ofvascular KDR channels by serine/threonine kinase-mediatedphosphorylation, for example by cAMP-dependent (PKA)and Ca2+/phospholipid-dependent (PKC) protein kinase(Aiello et al. 1995, 1998; Clment-Chomienne et al. 1996).However, direct evidence for their modulation by apathway involving cGMP and PKG is lacking.

In this study, the hypothesis that endothelium-dependentNO release produces relaxation of the rat basilar artery viaa modulation of KDR channel activity was tested using wiremyography and intracellular microelectrode recordings.Although we found that 4-AP depressed ACh-inducedendothelium-dependent relaxation of rat basilar artery,the data were not consistent with the view that KDR channelactivity was modulated by NO released from the endo-thelium. Rather, our findings suggest a novel, alternativeexplanation for the depression of endothelium-dependentrelaxation by 4-AP. Specifically, our findings suggest thatelectrotonic communication of the depolarization due to4-AP treatment from the smooth muscle cells to theendothelial layer via myo-endothelial gap junctions resultsin a depression of endothelium-dependent relaxation dueto a suppression of NO synthesis and/or release.

METHODS Tissue preparationMale Sprague-Dawley rats (300350 g) were maintained andkilled by halothane inhalation and exsanguination according to

the standards of the Canadian Council on Animal Care and aprotocol reviewed by the Animal Care Committee of the Facultyof Medicine, The University of Calgary. Basilar arteries weregently removed, placed in ice-cold Krebs buffer and carefullycleaned of adherent connective tissue. The Krebs solutioncontained (mM): NaCl, 120; KCl, 4.8; NaHCO3, 25; NaH2PO4, 1.2;MgSO4, 1.2; glucose, 11; and CaCl2, 1.8.

Wire myographyFor wire myograph studies, basilar arteries were cut into segments(2 mm in length) and mounted between two tungsten wires (40 mmin diameter) in a Mulvany-Halpern myograph (J. P. Trading,Denmark) for recording of isometric tension. Tissues weremaintained at 37 C in 5 ml of oxygenated (95 % O25 % CO2)Krebs buffer and resting tension was set to approximately 2 mN.Cumulative concentrationresponse relations for ACh, authenticNO (applied as bolus doses), sodium nitroprusside (SNP) anddiethylamineNONOate (DEANONOate) were determined usingarterial segments preconstricted with 5-hydroxytryptamine (5-HT; 0.11 mM) at a concentration that produced 75 % of maximalcontraction. Peak relaxations in response to ACh were determinedfor each concentration and normalized to developed tension toobtain the percentage relaxation. To determine the concentrationof 5-HT that would produce 75 % level of peak contraction,complete doseresponse curves were determined for 26 tissueswith an average peak developed force of 6.9 0.3 mN observedwith 3 mM 5-HT; the 75 % value, 5.3 mN, was obtained on averagewith 0.10.3 mM 5-HT. In some instances, it was necessary toadjust the concentration of 5-HT to obtain a similar level ofprecontraction in the presence of inhibitors as observed undercontrol conditions. Contractility data were recorded to hard diskand analysed via a Powerlab A/D convertor and software(ADInstruments, USA).

Membrane potential recordingsFor measurement of smooth muscle and endothelial cellmembrane potential, basilar arteries were cut open longitudinallyand pinned to the bottom of a Sylgard chamber, endothelialsurface uppermost. Tissues were maintained at 37 C andconstantly perfused with warmed Krebs buffer at a rate of5 ml min_1. Measurements of membrane potential were madewith sharp glass microelectrodes, back-filled with 3 M KCl andwith resistances of 60100 MV. All data were recorded through aPowerlab (ADInstruments) and stored on disk. Drugs were addedeither as bolus doses or to the perfusate as indicated.

Solutions and chemicalsAll drugs were obtained from Sigma Chemical Co. (St Louis, MO,USA) except for DEANONOate (Cayman Chemical Company,USA). All drugs were dissolved in Krebs solution except forDEANONOate, which was dissolved in degassed distilled waterand, pinacidil and 1H-(1,2,4)oxadiazolo(4,3-a)quinoxalin-1-one(ODQ), which were dissolved in DMSO and then diluted in Krebssolution to the desired concentration (DMSO at the maximal finaldilution of 0.1 % had no effect on 5-HT-induced contraction(control 6.9 1.0 mN versus DMSO 7.2 1.1 mN; n = 5; P > 0.05)or ACh-induced relaxation (control 95.4 2.3 % versus DMSO90.8 3.5 % at 1 mM ACh; n = 3; P > 0.05)). All Krebs solutionscontaining 4-AP were checked for appropriate pH of 7.4.Solutions of authentic NO were prepared by injecting researchgrade NO gas (Praxair, Canada) into deoxygenated H2O.Authentic NO solutions were injected into the myograph close tothe arterial segments and in volumes of less than 250 ml.

T. Allen, M. Iftinca, W. C. Cole and F. Plane976 J. Physiol. 545.3

Jou

rnal

of P

hysi

olog

y

StatisticsFor wire myograph studies, all data were expressed as means S.E.M.of n rats with only one tissue per rat employed for a givenexperimental group. Changes in membrane potential (in mV)were expressed as means S.E.M. of the number of rats. Thestatistical significance of all differences between mean values wascalculated using Students paired t test or repeated measuresANOVA followed by Bonferonnis post hoc test. A level of P < 0.05was considered to be statistically significant.

RESULTSEffect of 4-AP on endothelium-dependentrelaxations in response to AChAs shown in Fig. 1, ACh (1 nM to 1 mM) evokedconcentration-dependent relaxation of segments of basilarartery precontracted with 5-HT. Removal of the endo-thelium prevented relaxation in response to ACh (data notshown) and the relaxations were significantly inhibited inthe presence of L-NNA (100 mM) or ODQ (10 mM; Fig. 1),but not indomethacin (data not shown). Mean cumulativeconcentrationrelaxation curves for ACh in the absenceand presence of L-NNA and ODQ are shown in Fig. 1.Maximal relaxation was reduced from ~90 and ~88 to ~18(n = 14; P < 0.01) and ~26 % (n = 13; P < 0.01) in L-NNAand ODQ, respectively. Several tissues showed completesuppression with L-NNA alone (e.g. representative data ofFig. 1), but on average, relaxation in response to ACh was

still observed. Complete suppression of endothelium-dependent relaxation was obtained, however, with thecombination of L-NNA and N G-nitro-L-arginine methylester (L-NAME; 100 mM; Fig. 1).Application of 4-AP (300 mM) induced an increase in basaltone in 6 of 11 basilar artery segments studied (meanincrease 2.6 0.4 mN; n = 11 arteries) and regularoscillations in tone development following 4-AP exposurewere observed in some tissues (Fig. 2). In the presence of4-AP, relaxations in response to ACh were significantlyinhibited, with the maximal response reduced from ~92 to~33 % (n = 11; P < 0.01; Fig. 2). Pretreatment withtetrodotoxin (10 mM; n = 5) or selective inhibition ofneuronal Kv1.1 subunit-containing KDR channels withk-dendrotoxin (Akhtar et al. 2002; 10 nM; n = 3) werewithout effect on basal tone or ACh-induced relaxations inthe absence or presence of 4-AP (data not shown). We alsoemployed clofilium for comparison with 4-AP; this druginhibits recombinant Kv1.2 and Kv1.5 channels (Malayevet al. 1995; Yamagishi et al. 1995) and these channel

Myo-endothelial coupling in rat basilar arteryJ. Physiol. 545.3 977

Figure 1. Effect of L-NNA (100 mM) and ODQ (10 mM) onACh-evoked relaxation of segments of rat basilar arteryprecontracted with 5-HTA, representative recordings showing ACh (1 nM to 3 mM)-evokedrelaxations of 5-HT (0.1 mM)-induced tone in the absence andpresence of L-NNA. B and C, cumulative concentrationrelaxationcurves for ACh in the absence and presence of L-NNA andL-NNA + L-NAME (100 mM), and ODQ, respectively. Data are themeans of 14, 4 and 13 experiments, respectively, S.E.M. (in somecases the error bars in this and subsequent figures are within thesize of the data symbols).

Figure 2. Effect of 4-AP (300 mM) on ACh-evokedrelaxation of segments of rat basilar arteryprecontracted with 5-HTA, representative recording showing oscillatory increase in basaltension induced by 4-AP in some tissues. B, representativerecording showing ACh (1 nM to 3 mM)-evoked relaxations of5-HT (0.1 mM)-induced tone in the absence and presence of 4-AP.C, cumulative concentrationrelaxation curves for ACh in theabsence and presence of 4-AP and clofilium (3 mM). Data are theaverage of 11 and 5 experiments S.E.M. with 4-AP and clofilium,respectively,

Jou

rnal

of P

hysi

olog

y

subunits are expressed by vascular myocytes (Thorneloe etal. 2001). Clofilium at 3 mM (n = 5) inhibited ACh-induced relaxation, in a similar way to the suppressionproduced with 4-AP (Fig. 2), but it had no effect onrelaxations produced by the putative BKCa channelactivator, NS1619 (30 mM; n = 4) up to the maximalconcentration tested of 10 mM (data not shown).Effect of 4-AP on relaxations in response toexogenous NOTo test whether NO released from the endotheliummodulated vascular KDR channel activity, exogenous NOdonors were applied to vessel segments in the absence andpresence of 4-AP (300 mM). If KDR channels contributed tothe relaxation induced by endothelium-derived NO, wereasoned that relaxations in response to authentic NO,or a NO donor, would be similarly depressed by 4-AP.However, in contrast to ACh, relaxations of basilar arterysegments in response to authentic NO (1 nM to 1 mM;n = 11), SNP (1 nM to 3 mM; n = 5) or DEANONOate (1 nMto 0.3 mM; n = 3) were not inhibited by 4-AP (300 mM).Figure 3 shows a representative recording for authenticNO and mean cumulative concentrationrelaxationcurves for authentic NO, SNP and DEANONOate in theabsence and presence of 4-AP. Similarly, clofilium (3 mM)did not affect relaxations in response to authentic NO(n = 4; data not shown).

Prevention of 4-AP-induced inhibition of relaxationin response to ACh by pinacidilTo assess the role of membrane potential depolarization asa cause of the inhibition of ACh-induced endothelium-dependent relaxation induced by 4-AP, the K+ channelopener, pinacidil, was employed to activate ATP-sensitiveK+ channels and evoke hyperpolarization within thevessel segments. Pinacidil was applied in increasingconcentrations until reversal of 4-AP-induced (300 mM)increase in basal tension was observed (relaxation wasnoted at 1 mM pinacidil in most tissues). Significantly, inthe combined presence of pinacidil and 4-AP, the

inhibitory effect of KDR channel inhibition on ACh-evokedrelaxations of segments of basilar artery was not apparent.A representative trace illustrating the lack of effect of 4-APon ACh-induced relaxations in the presence of pinacidiland mean concentrationrelaxation curves for ACh in thepresence and absence of 4-AP and pinacidil (1 mM; n = 5)are shown in Fig. 4.

Prevention of 4-AP-induced inhibition of ACh-induced relaxation by 18bGAEndothelial cells are not thought to express voltage-gatedK+ channels (Himmel et al. 1993; Nilius & Droogmans,2001) and for this reason we felt that the depolarizationwas not due to a direct effect of 4-AP on endothelial cellmembrane potential. However, we reasoned that ACh-induced relaxation might be inhibited indirectly if thedepolarization evoked by 4-AP inhibition of smooth muscleKDR channels was communicated to the endothelium viaelectrical coupling by myo-endothelial gap junctions. Totest this hypothesis, ACh was applied in the combinedpresence of 4-AP and the gap junction uncoupling agent,18bGA (Davidson & Baumgarten, 1988; Guan et al. 1996;Yamamoto et al. 1998). In preliminary experiments,18bGA was found to depress 5-HT-induced contractionsin a concentration-dependent fashion between 10 and50 mM, probably as a result of nonspecific inhibition oftone development, as previously reported (Chaytor et al.2000; Tare et al. 2002). However, at 5 mM, no effect on5-HT-induced tone development (control 8.4 0.7 versus18bGA 8.3 0.6 mN; n = 4; P > 0.05) or ACh-inducedrelaxation (control 78.4 2.3 versus 18bGA 82.7 9.8 %with 1 mM ACh; n = 3; P > 0.05) was observed, so thisconcentration was employed to determine the effect onendothelium-dependent relaxation in the presence of 4-AP.Treatment with 18bGA was found to prevent the inhibitoryeffect of 4-AP (300 mM) on ACh-induced relaxation.A representative recording and mean concentrationrelaxation curves for ACh in the presence of 4-AP and18bGA are shown in Fig. 5 (n = 5).


Figure 3. Effect of 4-AP (300 mM) onrelaxations in response to NO in rat basilararteryA, representative recordings showing authenticNO (1 nM to 1 mM)-induced relaxations of 5-HT(0.03 mM)-induced tone in the absence andpresence of 4-AP. B, C and D, mean cumulativeconcentrationrelaxation curves for authenticNO, SNP and DEANONOate (NONOate) in theabsence and presence of 4-AP. Each point is themean of 11, 5 and 3 experiments, respectively, S.E.M.

Jou

rnal

of P

hysi

olog

y

Effect of 4-AP on membrane potential of endothelialand vascular smooth muscle cells of rat basilar arteryTo directly assess the effect of 4-AP on the membranepotential of vascular smooth muscle and endothelial cellsof rat basilar artery and to determine if the gap junctionuncoupling drug, 18bGA, would prevent depolarizationof the endothelium by 4-AP, segments of rat basilar arterywere opened and impaled with sharp microelectrodesfrom the luminal surface. In most cases, the endothelialcells were the first cells to be impaled using this approachand recordings of smooth muscle membrane potentialwere obtained by further advancing the pipette into thevessel wall. To confirm the specific cell type impaledduring each recording, the response of membranepotential to BayK8644 (1 mM), an activator of L-typevoltage-gated Ca2+ channels, was determined in theabsence and presence 18bGA (5 mM). The rationale for theuse of BayK8644 was as follows: L-type Ca2+ channels areexpressed by vascular smooth muscle but not byendothelial cells (Nilius & Droogmans, 2001), so BayK8644would be not expected to depolarize endothelial cells in theabsence of electrical communication with the underlyingsmooth muscle cells. The representative recordings ofFig. 6 show that BayK8644 (1 mM) depolarized bothendothelial and smooth muscle cells, but in the presence ofthe gap junction uncoupler, 18bGA (5 mM), only the

membrane potential of smooth muscle cells was affectedby the Ca2+ channel activator. The resting membranepotential of endothelial cells was depolarized compared tothat of smooth muscle cells under control conditions(Table 1). BayK8644 caused depolarization of endothelialand smooth muscle cells by +10.4 1.2 and +12.9 1.5 mV, respectively (Table 1). Application of 18bGA(5 mM) induced a small, but significant depolarization of+4.1 1.1 mV in endothelial cells (P < 0.01), but it had noeffect on smooth muscle cell membrane potential (Table 1).In the presence of 18bGA, however, the membranepotential of endothelial cells was unaffected by BayK8644,but the membrane potential of vascular smooth musclecells exhibited a depolarization of +14.2 1.4 mV (P < 0.01;Table 1). In contrast to BayK8644, application of 5-HT(10 mM) had only a minimal effect on membrane potentialof smooth muscle cells (+2.1 0.9 mV; n = 15).

The representative recordings of Fig. 7 show thatapplication of 4-AP (200 mM) evoked a similar level ofdepolarization in endothelial and vascular smooth musclecells of rat basilar artery by +11.0 1.0 (P < 0.05) and+10.0 1.0 mV (P < 0.05), respectively. Significantly,however, the depolarization of endothelial cells induced by4-AP was reversed by 18bGA (5 mM), whereas uncouplinggap junctions had no effect on the depolarization of thesmooth muscle cells. Similarly, Fig. 7 shows that if 18bGA


Figure 4. Effect of pinacidil (1 mM) on 4-AP-inducedinhibition of relaxation in response to ACh in rat basilararteryA, representative recordings showing the prevention of 4-AP(300 mM)-induced inhibition of ACh (1 nM to 10 mM)-evokedrelaxation of 5-HT (0.3 mM)-induced tone by pinacidil (1 mM).Note that relaxation of 4-AP-induced increase in basal tone wasunaffected by 0.1, but it was reversed by 1 mM pinacidil. B, meancumulative concentrationrelaxation curves for ACh in theabsence and presence of 4-AP and pinacidil (1 mM). Each point isthe mean of 5 experiments S.E.M.

Figure 5. Effect of 18bGA (5 mM) on 4-AP-inducedinhibition of relaxation in response to ACh in rat basilararteryA, representative recordings showing the prevention of 4-AP(300 mM)-induced inhibition of ACh (1 nM to 10 mM)-evokedrelaxations of 5-HT (0.3 mM)-induced tone by 18bGA (5 mM).B, mean cumulative concentrationrelaxation curves for ACh inthe absence and presence of 4-AP and 18bGA. Each point is themean of 5 experiments S.E.M.

Jou

rnal

of P

hysi

olog

y

was added before the addition of 4-AP, then themagnitude of the depolarization of endothelial cells wasreduced. Under control conditions, 4-AP depolarizedendothelial cells by +12.4 3.9 mV (n = 6), but aftertreatment with 18bGA, 4-AP treatment had no effect onmembrane potential (Table 2). In contrast, application of4-AP depolarized smooth muscle cells by +9.2 1.1 mVbefore, and by a similar magnitude of +10.5 1.1 mVafter, treatment with 18bGA (Table 2).Effect of 4-AP on ACh-induced hyperpolarization ofendothelium and vascular smooth muscle cells of ratbasilar arteryAs shown in Fig. 8, bolus doses of ACh (10 mmol) causedtransient hyperpolarizations of the membrane potential ofendothelial and smooth muscle cells that were of a similarpeak amplitude of approximately _10 mV (Table 3). Thehyperpolarization of endotheliaI cells was unaffected bytreatment with L-NNA, but hyperpolarization of smooth

muscle cells was completely blocked (Table 3). These dataindicate that the hyperpolarization of the smooth musclecells in response to ACh was due to NO released from theendothelium. In the presence of 4-AP (200 mM), the


Table 1. Effect of BayK8644 on endothelial and vascular smooth muscle cell membranepotential (mV) in the absence and presence of 18bGA

Endothelium Vascular smooth muscleControl _37 0.9 * (n = 38) _49.1 1.1 (n = 38)

Before After D Before After DBayK8644 (n = 14, 14) _37.4 1.5 _27.0 2.0 +10.4 1.2 _49.9 2.2 _37.0 2.6 +12.9 1.5

18bGA (n = 38, 38) _37.8 0.9 _32.7 0.7 +4.1 1.1 _49.1 1.1 _48.8 1.0 +0.3 1.018bGA + BayK8644 (n = 38, 38) _32.3 1.0 _29.0 1.2 +3.3 0.9 _49.0 1.7 _34.8 2.0 +14.2 1.4

* Significantly different from smooth muscle cell membrane potential (P < 0.01). Significantly differentfrom value before BayK8644 (P < 0.01). Significantly different from value before 18bGA (endothelium) or18bGA + BayK8644 (vascular smooth muscle; P < 0.01). Significantly different from value in BayK8644but absence of 18bGA (P < 0.01).

Figure 6. Effect of BayK8644 (1 mM) on membranepotential of endothelial and smooth muscle cells of ratbasilar artery in the absence and presence of 18bGA(5 mM)Representative microelectrode recordings of BayK8644-induceddepolarization of endothelial (Endo) and smooth muscle cells(VSM) in the absence (left) and presence (right) of 18bGA. Notethat the depolarization of the endothelial cell, but not the VSM cell,due to BayK8644 was prevented by 18bGA.

Figure 7. Effect of 4-AP (200 mM) on membrane potentialof endothelial and smooth muscle cells of rat basilarartery in the absence and presence of 18bGA (5 mM) A, representative microelectrode recordings showing 4-AP-induced depolarization of an endothelial cell (left) and the abilityof 18bGA to reverse the depolarization caused by 4-AP.B, representative microelectrode recordings showing 4-AP-induced depolarization of a vascular smooth muscle cell (left) andthe inability of 18bGA to reverse the depolarization caused by4-AP. C, representative microelectrode recordings showing 4-AP-induced depolarization of an endothelial cell (left) and the lack ofdepolarization in response to 4-AP in the presence of 18bGA.D, representative microelectrode recordings showing 4-AP-induced depolarization of a vascular smooth muscle cell in theabsence (left) and presence of 18bGA (right).

Jou

rnal

of P

hysi

olog

y

membrane potential of both cell types was significantlydepolarized and the hyperpolarizations induced by AChdid not reach the level of membrane potential attained inthe absence of 4-AP. Additionally, 4-AP did not alter themagnitude of the ACh-induced hyperpolarization ofendothelium, but the magnitude of smooth musclehyperpolarization was significantly reduced (Fig. 8 andTable 3). To quantify the change in ACh-inducedhyperpolarization in the presence of 4-AP, the amplitudeand duration of the hyperpolarization were simultaneouslyconsidered by calculating the area under the curve forthe endothelial and smooth muscle hyperpolarizations.Endothelial and smooth muscle cell hyperpolarizationinduced by ACh were significantly reduced from2164 141 to 1156 80 mV s (n = 9; P < 0.01) and

from 4427 367 to 1982 210 mV s (n = 9; P < 0.01),respectively by 4-AP. Thus, 4-AP reduced the extent ofACh-induced hyperpolarization of the endothelium andsmooth muscle cells, as well as shifting the range overwhich the hyperpolarizations occurred to more positivepotentials. Application of 18bGA did not alter hyper-polarization of endothelial or smooth muscle cells inresponse to ACh, indicating that myo-endothelial gapjunctional coupling is not required for endothelium-dependent relaxation in rat basilar artery (n = 6).However, in the combined presence of 4-AP and 18bGA,ACh-induced hyperpolarization of smooth muscle cellmembrane potential (_12.3 1.6 mV) was significantlygreater than in 4-AP alone (_6.5 1.1 mV; Fig. 8 andTable 3). Additionally, the area under the curves for ACh-


Table 2. Effect of 4-AP on endothelial and vascular smooth muscle cell membrane potential(mV) in the absence and presence of 18bGA

Endothelium Vascular smooth muscle

Before After D Before After D4-AP (n = 15, 11) _38.9 1.4 _27.9 0.7 +11.0 1.0 _46.3 1.8 _36.4 1.8 +10.0 1.0

4-AP (before 18bGA) (n = 6, 6) _40.2 1.9 _27.8 2.0 +12.4 3.9 _43.4 2.5 _34.2 2.4 +9.2 1.14-AP (after 18bGA) * (n = 6, 6) _34.7 1.7 _33.0 1.4 +1.7 1.0 _47.4 2.7 _36.9 3.3 +10.5 1.1

* Paired impalements maintained during 4-AP application in the absence and presence of 18bGA. Significantly different from value of membrane potential in the absence of 4-AP (P < 0.01). Significantlydifferent from change in membrane potential due to 4-AP in the absence of 18bGA (P < 0.01).

Figure 8. Effect of L-NNA (100 mM) and 4-AP (200 mM) on ACh-induced hyperpolarization ofthe membrane potential of endothelial and smooth muscle cells of rat basilar arteryA, representative microelectrode recordings showing an ACh (bolus concentration 100 mmol)-inducedhyperpolarization of a smooth muscle cell (VSM) in the absence (left) and lack of change in membranepotential following ACh application in the presence of L-NNA (right). B, representative microelectroderecordings of ACh-induced hyperpolarizations of an endothelial cell (Endo) in the absence (left) and in thepresence of 4-AP alone (middle) and 4-AP + 18bGA (right). C, representative microelectrode recordings ofACh-induced hyperpolarization of a smooth muscle cell (VSM) in the absence (left) and in the presence of4-AP (midde) and 4-AP + 18bGA (right).

Jou

rnal

of P

hysi

olog

y

induced hyperpolarization of endothelial and smoothmuscle cell membrane potential were 2297 96 and3645 244 mV s (n = 4) in the presence of 4-AP and18bGA, and significantly greater than the values obtainedin 4-AP alone (P > 0.05), but not different from thosedetermined for control conditions.

DISCUSSIONThis study makes the novel observation that depolariz-ation of vascular smooth muscle cells of the rat basilarartery due to inhibition of KDR channels with 4-AP, or theactivation of L-type Ca2+ channels with BayK8644, leads toa corresponding depolarization of endothelial cellmembrane potential due to electrical communicationbetween the two cell types via myo-endothelial gapjunctions. Significantly, depolarization of the endotheliumin the presence of 4-AP was associated with a depression ofACh-induced endothelium-dependent relaxation mediatedby NO. Based on our findings, it is unlikely that vascularsmooth muscle KDR channels mediate NO-inducedhyperpolarization and relaxation of the rat basilar artery.Rather, the results indicate that 4-AP-induced smoothmuscle depolarization may indirectly inhibit ACh-induced relaxation via electrotonic depolarization of theendothelium and suppression of NO synthesis and/orrelease. The ability of endothelial cell hyperpolarization toaffect smooth muscle membrane potential via myo-endothelial gap junctional communication is wellrecognized as a potential mechanism to account for non-NO, non-prostanoid endothelium-dependent relaxation,i.e. relaxation due to an endothelium-dependent hyper-polarizing factor (reviewed by Bny, 1999; Fltou &Vanhoutte, 1999). However, only limited evidence existsto indicate that electrical communication can occur in thereverse direction, from vascular smooth muscle cells to theendothelium, and the functional consequences of thiscoupling have not been assessed previously. The results of

this study provide the first evidence that selectivedepolarization of vascular smooth muscle cells caninfluence the level of endothelial cell membrane potentialand thereby endothelium function by reducing NO-dependent control of vascular tone.

A substantial body of evidence indicates that endothelium-dependent relaxation of arterial smooth muscle ismediated by at least three different factors/mechanisms,including NO, prostacyclin and an as-yet-unidentifiedendothelium-derived hyperpolarizing factor, which maybe a P450-derived epoxide, K+ or myo-endothelial gapjunctional electrical coupling (Edwards et al. 1998;Coleman et al. 2001; Fleming, 2001). Endothelium-dependent relaxation of the rat basilar artery evoked byACh was shown to be completely suppressed by inhibitionof NOS by the combination of L-NNA and L-NAME, andby the inhibitor of soluble guanylyl cyclase, ODQ, but notby indomethacin or 18bGA. This indicates that in thisvessel, endothelium-dependent relaxation is exclusivelydue to the release of NO. There is compelling evidence thatBKCa channels are important effectors of the action of NOin mediating vascular smooth muscle hyperpolarizationand relaxation in many vessels (Homer & Wanstall, 2000).This role for BKCa was established in part by experimentsshowing that the specific BKCa inhibitor iberiotoxinsuppressed endothelium-dependent relaxation inducedby NO (Khan et al. 1993; Bialecki & Stinson-Fisher, 1995;Kitazono et al. 1997; Li et al. 1997; Satake et al. 1997). Asimilar approach has been employed to define a role forother K+ channels as effectors of endothelium-derivedNO. For example, inhibition of ACh-evoked relaxation by4-AP was interpreted to indicate that KDR channelsmediate NO-induced relaxation of rat basilar arteries invivo (Sobey & Faraci, 1999). In the light of the present data,however, this mechanism would not appear to beinvolved. This view is based on the lack of effect of 4-AP onrelaxations evoked by exogenous sources of NO, including


Table 3. Effect of 4-AP on ACh-induced hyperpolarization of endothelial and vascularsmooth muscle cell membrane potential (mV) in the absence and presence of 18bGA

Endothelium Vascular smooth muscle

Before After D Before After DACh (n = 3, 3) _35.7 1.8 _47.3 1.0 _11.7 1.1 _43.3 3.0 _55.2 2.0 _11.8 1.6

ACh + L-NNA * (n = 3, 3) _35.3 2.4 _44.3 2.4 _12.0 1.8 _40.7 2.4 _41.5 3.1 _0.8 1.0

ACh (n = 16, 16) _39.8 1.2 _48.5 1.6 _9.5 1.1 _46.1 1.4 _56.4 0.7 _10.3 1.1

ACh + 4-AP (n = 9, 9) _27.1 1.1 _37.8 1.1 _7.9 0.9 _33.8 2.1 _40.5 2.1 _6.5 1.1

ACh + 4-AP and 18bGA (n = 4, 4) _39.0 2.1 _52.0 1.8 _12.3 1.0 _35.3 2.0 _47.5 1.9 _12.3 1.6 * Impalements maintained during 4-AP application in the absence and presence of L-NNA. Significantlydifferent from value of membrane potential in the absence of ACh (P < 0.01). Significantly different fromchange in membrane potential due to ACh in the absence of L-NNA (P < 0.01). Significantly different fromvalue in the absence of 4-AP (P < 0.01). Significantly different from value in the presence of 4-AP(P < 0.01).

Jou

rnal

of P

hysi

olog

y

authentic NO, sodium nitroprusside and DEANONOate.Similar results with 1 mM 4-AP and DEANONOate werepreviously reported by Hemplemann et al. (2001). In theabsence of a role for KDR channels in mediating the effect ofNO on rat basilar arterial smooth muscle cells, analternative explanation for the suppression of ACh-evokedendothelium-dependent relaxation by 4-AP is required.

This study indicates that membrane potential depolariz-ation underlies the suppression of endothelium-dependentrelaxation by 4-AP. Pinacidil was employed to activateATP-sensitive K+ channels and hyperpolarize vascularsmooth muscle cells (Nelson & Quayle, 1995) and wasfound to prevent the suppression of ACh-evokedrelaxation by 4-AP. This indicates that depolarization ofthe vascular smooth muscle cells and/or the endotheliumby 4-AP was responsible for the suppression ofendothelium-dependent relaxation by NO. This isconsistent with previous observations showing thatdepolarization of pulmonary arteries with elevatedextracellular potassium (25 mM) blocked endothelium-dependent relaxation (Seiden et al. 2000). Depolarizationof the basilar arterial smooth muscle cells by 4-AP per sedid not block the actions of endothelium-derived NO onthe myocytes because 4-AP-induced-depolarization hadno effect on relaxations in response to exogenous NO.Alternatively, depolarization of the endothelium is knownto reduce NO synthesis and/or release. Previous studiesshow that NO synthesis and/or release is dependent on arise in intracellular [Ca2+]i in endothelial cells, that Ca

2+

influx is required for sustained endothelium-dependentrelaxation, and that this Ca2+ influx is decreased byendothelial cell depolarization (Adams et al. 1989;Lckhoff & Busse, 1990b; Wang & Van Breeman, 1999).Endothelial Ca2+ influx is not mediated by voltage-dependent Ca2+ channels and is not activated by depolariz-ation. Rather, hyperpolarization appears to be required, inpart to maintain an appropriate electrochemical drivingforce for ACh-induced Ca2+ influx through a nonselectivecation conductance (Schilling, 1989; Kamouchi et al. 1999;Wang & Van Breeman, 1999), but it may also prevent Ca2+

entry by channel inactivation and/or reduction in unitaryconductance. The latter is suggested by the fact that thedecline in Ca2+ influx with depolarization positive to~_35 mV exceeds that which can be explained by areduction in driving force (Oike et al. 1994; Wang & VanBreeman, 1999). According to these observations, areduction in basal and agonist-evoked NO release wouldaccompany endothelial depolarization from _39 to_27 mV under resting conditions and reduction in ACh-induced peak hyperpolarization from _49 to _38 mV inthe presence of 4-AP. A limitation in this study is the lackof direct measurements to confirm the suppression ofACh-induced NO release from the basilar arterialendothelium by 4-AP treatment. However, the presentresults provide indirect evidence that NO release was

reduced. First, the ACh-induced hyperpolarization ofsmooth muscle cells was blocked by L-NNA, indicatingthat it was induced by NO. Second, both the amplitudeand the area under curve for the ACh-induced smoothmuscle hyperpolarization were significantly reduced in thepresence of 4-AP, suggesting that the magnitude of NOrelease was reduced.

It is unlikely that a direct effect of 4-AP on endothelial K+

channels was responsible for the depolarization observedin this study. This is based on: (1) the lack of effect of 4-APon the membrane potential of endothelial cells in thepresence of 18bGA; and (2) the lack of evidence of 4-AP-sensitive KDR channels in endothelial cells (Nilius &Droogmans, 2001) despite immunocytochemical evidenceof voltage-gated K+ channel subunit expression in theendothelium of some resistance arteries and capillaries(Fan & Walsh, 1999; Cheong et al. 2001) and reports oftheir presence in cultured endothelial cells (Takeda et al.1987; Hogg et al. 1999). Nonspecific effects would alsoseem to be an unlikely explanation for the suppression ofACh-induced relaxation by 4-AP. Although 4-AP is widelyemployed as an inhibitor of voltage-gated K+ channels insmooth muscle and other cell types (Nelson & Quayle,1995), actions unrelated to channel block were previouslyreported and its selectivity and suitability for use in intacttissue and/or cell preparations has been questioned (e.g.Wood & Gillespie, 1998). However, the concentration of4-AP (200 or 300 mM) used in this study is consistent withthe IC50 for inhibition of KDR currents of vascular myocytes(Kerr et al. 2001), and significantly lower than thatassociated with reduced Ca2+ release, intracellular pHand/or BKCa channel activity (i.e. 520 mM; Petkova-Kirova et al. 2000; Quignard et al. 2000; Ghisdal & Morel,2001). Additionally, clofilium (3 mM) was also employedto selectively inhibit KDR channels and found to mimic theactions of 4-AP treatment. For this reason, it is unlikelythat 4-AP affected endothelial or vascular smooth musclecell function via a mechanism(s) unrelated to aninhibition of smooth muscle KDR channels.

The electrophysiological data obtained in this studyare consistent with the view that the inhibition ofendothelium-dependent relaxation by 4-AP was the resultof an indirect effect on endothelial cell membranepotential mediated by electrical coupling through myo-endothelial gap junctions. We employed 18bGA touncouple gap junctions and to prevent myo-endothelialgap junctional communication based on the previousfindings showing a phosphatase-mediated dephosphoryl-ation of connexins and loss of cell-to-cell coupling viadisassembly of gap junction plaques with this agent (Guanet al. 1996). 18bGA was found to prevent and/or reverseendothelial cell depolarization induced by 4-AP orBayK8644, but it did not affect smooth muscle celldepolarization by these treatments. These results indicate


Jou

rnal

of P

hysi

olog

y

that uncoupling myo-endothelial gap junctions canprevent endothelial cell depolarization as a result of theselective depolarization of smooth muscle cells by twodistinct mechanisms, inhibition of KDR channels andactivation of L-type Ca2+ channels. Moreover, in thepresence of 18bGA, we found that the inhibition of ACh-induced hyperpolarization and relaxation by 4-AP wassignificantly reduced. This indicates that uncoupling myo-endothelial gap junctions permitted a recovery ofendothelium function despite the presence of maintainedsmooth muscle depolarization due to 4-AP. We do notattribute the effect of 18bGA to the nonspecific effects onmembrane potential (Tare et al. 2002) and tonedevelopment (Chaytor et al. 2000) that were noted whenthis drug was employed at 50100 mM. We approached theuse of 18bGA with caution and used a concentration of5 mM that did not: (1) cause relaxation of arterial segmentsprecontracted with 5-HT; (2) alter the extent of ACh-induced relaxation or sensitivity to ACh; or (3) changesmooth muscle membrane potential. 18bGA did cause aslight depolarization of basilar arterial endothelial cells,but at approximately +4 mV, the change in membranepotential was much smaller than the +30 mV depolariz-ation observed with higher concentrations (Coleman et al.2001). We attribute this +4 mV depolarization to theuncoupling of the endothelium from the hyperpolarizinginfluence of the underlying smooth muscle cells, aspreviously reported (Popp et al. 1996; Yamamoto et al.1998).

The role of the endothelium in control of vascular smoothmuscle contractility in health is well established anddysfunctional regulation of arterial tone development dueto abnormal endothelial function and reduced NO releaseis recognized to be a important cause of abnormalcontractility associated with vascular disease (Vanhoutte,1997; Schiffrin, 2001; Baron, 2002). This study shows forthe first time that vascular smooth muscle cell membranedepolarization can alter endothelial function. The presenceof myo-endothelial gap junctions between endothelial andsmooth muscle cells of several arteries has been identified(e.g. Sandow & Hill, 2000), and these structures arepresent in the rat basilar artery (S. L. Sandow & C. E. Hill,unpublished observations). Moreover, previous studieshave shown that changes in arterial smooth musclemembrane potential can be communicated to endothelialcells in several vessels (von der Weid & Bny, 1993; Bny &Pacicca, 1994; Marchenko & Sage, 1994; Murai et al. 1999;White & Hiley, 2000). For example, noradrenaline-induceddepolarization (rat aorta; Marchenko & Sage, 1994) andlevcromakalim-induced hyperpolarization (rabbit and ratmesenteric arteries; Murai et al. 1999; White & Hiley,2000) were reported to be transmitted from smoothmuscle to endothelial cells. However, the physiologicaland pathophysiological significance of smooth musclecontrol of endothelial function via myo-endothelial

gap junction communication has not been addressedpreviously. In the present study, 4-AP was found to causean impairment of agonist-induced, endothelium-dependentregulation of tone development in the rat basilar artery.Based on this novel finding, it would seem worthwhile toconsider the possibility that abnormal smooth muscledepolarization in disease may contribute to a loss ofendothelium-dependent control of arterial tonedevelopment, for example, as a result of depressed smoothmuscle K+ channel, or enhanced Ca2+ channel, activityand/or expression (Smirnov et al. 1994; Platoshyn et al.2001). NO is also known to have additional roles withinblood vessels, including control of proliferation andphenotype, and abnormal NO synthesis and/or release dueto smooth muscle depolarization and electrotonicdepolarization of the endothelium could also contribute todysfunctional regulation of vasculogenesis, athero-scelerosis and retinopathy. To address these possibilities,further studies will be required to determine: (1) theelectrical properties, such as the ratio of electricalimpedance between the endothelium and smooth musclelayers, as well as the level of myo-endothelial gapjunctional conductance, that are required for control ofendothelial membrane potential by vascular smoothmuscle; and (2) whether the properties of the endo-thelium, smooth muscle cells and myo-endothelial gapjunctions of other arteries are appropriate for smoothmuscle to endothelium communication in health and/ordisease states.

REFERENCESADAMS, D. J., BARAKEH, J., LASKEY, R. & VAN BREEMAN, C. (1989). Ion

channels and regulation of intracellular calcium in vascularendothelial cells. FASEB Journal 3, 23892400.

AIELLO, E. A., MALCOLM, A. T., WALSH, M. P. & COLE, W. C. (1998).Beta-adrenoceptor activation and PKA regulate delayed rectifierK+ channels of vascular smooth muscle cells. American Journal ofPhysiology 275, H448459.

AIELLO, E. A., WALSH, M. P. & COLE, W. C. (1995). Phosphorylationby protein kinase A enhances delayed rectifier K+ current invascular smooth muscle cells. American Journal of Physiology 268,H924934.

AKHTAR, S., SHAMOTIENKO, O., PAPAKOSTA, M., ALI, F. & DOLLY, J. O.(2002). Characteristics of brain Kv1 channels tailored to mimicnative counterparts by tandem linkage of alpha subunits.Implications for K+ channelopathies. Journal of BiologicalChemistry 277, 1637616383.

BARON, A. D. (2002). Insulin resistance and vascular function.Journal of Diabetes Complications 16, 92102.

BNY, J. L. (1999). Information networks in the arterial wall. News inPhysiological Sciences 11, 6873.

BNY, J. L. & PACICCA, C. (1994). Bidirectional communicationbetween smooth muscle and endothelial cells in the pig coronaryartery. American Journal of Physiology 266, H14651472.

BIALECKI, R. A. & STINSON-FISHER, C. (1995). KCa channelantagonists reduce NO donor-mediated relaxation of vascular andtracheal smooth muscle. American Journal of Physiology 268,L152159.


Jou

rnal

of P

hysi

olog

y

BUSSE, R., FICHTNER, H., LCKHOFF, A. & KOHLHARDT, M. (1988).Hyperpolarization and increased free calcium in acetylcholine-stimulated endothelial cells. American Journal of Physiology 255,H965969.

CHAYTOR, A. T., MARSH, W. L. HUTCHESON, I. R. & GRIFFITH, T. M.(2000). Comparison of glycyrrhetinic acid isoforms andcarbenoxolone as inhibitors of EDHF-type relaxations mediatedvia gap junctions. Endothelium 7, 265278.

CHEONG, A., DEDMAN, A. M., XU, S. Z. & BEECH, D. J. (2001). Kva1channels in murine arterioles: differential cellular expression andregulation of diameter. American Journal of Physiology 281,H10571065.

CLMENT-CHOMIENNE, O., WALSH, M. P. & COLE, W. C. (1996).Angiotensin II activation of protein kinase C decreases delayedrectifier K+ current in rabbit vascular myocytes. Journal ofPhysiology 495, 689700.

COLEMAN, H. A., TARE, M. & PARKINGTON, H. C. (2001). K+ currentsunderlying the action of endothelium-derived hyperpolarizingfactor in guinea-pig, rat and human blood vessels. Journal ofPhysiology 531, 359373.

DAVIDSON, J. S. & BAUMGARTEN, I. M. (1988). Glycyrrhetinic acidderivatives: a novel class of inhibitors of gap junctionalintracellular communication. Structure-activity relationships.Journal of Pharmacology and Experimental Therapeutics 246,11041107.

EDWARDS, G. A., DORA, K. A., GARDENER, M. J., GARLAND, C. J. &WESTON, A. H. (1998). K+ is an endothelium-derivedhyperpolarizing factor in rat arteries. Nature 396, 269272.

FAN, J. & WALSH, K. B. (1999). Mechanical stimulation regulatedvoltage-gated potassium currents in cardiac microvascularendothelial cells. Circulation Research 84, 451457.

FLTOU, M. & VANHOUTTE, P. M. (1999). The third pathway:endothelium-dependent hyperpolarization. Journal of Physiologyand Pharmacology 50, 525534.

FLEMING, I. (2001). Cytochrome P450 enzymes in vascularhomeostasis. Circulation Research 89, 753762.

GHISDAL, P. & MOREL, N. (2001). Cellular target of voltage andcalcium-dependent K+ channel blockers involved in EDHF-mediated responses in rat superior mesenteric artery. BritishJournal of Pharmacology 134, 10211028.

GUAN, X., WILSON, S., SCHLENDER, K. K. & RUCH, R. J. (1996). Gap-junction disassembly and connexin 43 dephosphorylation inducedby 18 beta-glycyrrhetinic acid. Molecular Carcinogenesis 16,157164.

HEMPELMANN, R. G., SEEBECK, J., KRUSE, M. L., ZIEGLER, A. &MEHDORN, H. M. (2001). Role of potassium channels in therelaxation induced by the nitric oxide (NO) donor DEA/NO in theisolated rat basilar artery. Neuroscience Letters 313, 2124.

HIMMEL, H. M., WHARTON, A. R. & STRAUSS, H. C. (1993).Intracellular calcium currents and stimulus response coupling inendothelial cells. Hypertension 21, 112127.

HOGG, D. S., ALBARWANI, S., DAVIS, A. R. & KOZLOWSKI, R. S. (1999).Endothelial cells freshly isolated from resistance-sized pulmonaryarteries possess a unique K+ current profile. Biochemical andBiophysical Research Communications 24, 405409.

HOMER, K. L. & WANSTALL, J. C. (2000). Cyclic GMP-independentrelaxation of rat pulmonary artery by spermine NONOate, adiazeniumdiolate nitric oxide donor. British Journal ofPharmacology 131, 673682.

KAMOUCHI, M., DROOGMANS, G. & NILIUS, B. (1999). Membranepotential as a modulator of the free intracellular Ca2+

concentration in agonist-activated endothelial cells. GeneralPhysiology and Biophysics 18, 199208.

KERR, P. M., CLMENT-CHOMIENNE, O., THORNELOE, K. S. CHEN, T. T.,ISHII, K., SONTAG, D. P. WALSH, M. P. & COLE, W. C. (2001).Heteromultimeric Kv1.2Kv1.5 channels underlie4-aminopyridine-sensitive delayed rectifier K+ current of rabbitvascular myocytes. Circulation Research 89, 10381044.

KHAN, S. A., MATHEWS, W. R. & MEISHERI, K. D. (1993). Role ofcalcium-activated K+ channels in vasodilation induced bynitroglycerine, acetylcholine and nitric oxide. Journal ofPharmacology and Experimental Therapeutics 267, 13271335.

KITAZONO, T., IBAYASHI, S., NAGAO, T., FUJII, K. & FUJISHIMA, M.(1997). Role of Ca2+-activated K+ channels in acetylcholine-induced dilatation of the basilar artery in vivo. British Journal ofPharmacology 120, 102106.

KRUSE, H. J., GRUNBERG, B., SIESS, W. & WEBER, P. C. (1994).Formation of biologically active autacoids is regulated by calciuminflux in endothelial cells. Arteriosclerosis Thrombosis 14,18211828.

LANG, R. J., HARVEY, J. R., MCPHEE, G. J. & KLEMM, M. F. (2000).Nitric oxide and thiol reagent modulation of Ca2+ activated (BKCa)channels myocytes of the guinea-pig taenia caeci. Journal ofPhysiology 525, 363376.

LI, P. L., ZOU, A. P. & CAMPBELL, W. B. (1997). Regulation ofpotassium channels in coronary arterial smooth muscle byendothelium-derived vasodilators. Hypertension 29, 262267.

LOVREN, F. & TRIGGLE, C. (2000). Nitric oxide and sodiumnitroprusside-induced relaxation of the human umbilical artery.British Journal of Pharmacology 131, 521529.

LCKHOFF, A. & BUSSE, R. (1990a). Activators of potassium channelsenhance calcium influx into endothelial cells as a consequence ofpotassium currents. Naunyn-Schmiedebergs Archives ofPharmacology 342, 9499.

LCKHOFF, A. & BUSSE, R. (1990b). Calcium influx into endothelialcells and formation of endothelium-derived relaxing factor iscontrolled by the membrane potential. Pflgers Archiv 416,305311.

LCKHOFF, A., POHL, U., MLSCH, A. & BUSSE, R. (1988). Differentialrole of extra- and intracellular calcium in the release of EDRF andprostacyclin from cultured endothelial cells. British Journal ofPharmacology 95, 189196.

MALAYEV, A. A., NELSON, D. J. & PHILIPSON, L. H. (1995). Mechanismof clofilium block of the human delayed rectifier potassiumchannel. Molecular Pharmacology 47, 198205.

MARCHENKO, S. M. & SAGE, S. O. (1994). Smooth muscle cells affectendothelial membrane potential in rat aorta. American Journal ofPhysiology 267, H804811.

MURAI, T., MURAKI, K., IMAIZUMI, Y. & WATANABE, M. (1999).Levcromakalim causes indirect endothelial hyperpolarization via amyo-endothelial pathway. British Journal of Pharmacology 128,14911496.

MURPHY, M. E. & BRAYDEN, J. E. (1995). Nitric oxide hyperpolarisesrabbit mesenteric arteries via ATP-sensitive potassium channels.Journal of Physiology 486, 4758.

NELSON, M. T. & QUAYLE, J. M. (1995). Physiological roles andproperties of potassium channels in arterial smooth muscle.American Journal of Physiology 268, C799822.

NILIUS, B. & DROOGMANS, G. (2001). Ion channels and theirfunctional role in vascular endothelium. Physiological Reviews 81,14151459.

OIKE, M., GERICKE, M., DROOGMANS, G. & NILIUS, B. (1994). Calciumentry activated by store depletion in human umbilical veinendothelial cells. Cell Calcium 16, 367376.


Jou

rnal

of P

hysi

olog

y

PETKOVA-KIROVA, P., GAGOV, H., KRIEN, U., DURIDANOVA, D., NOACK,T. & SCHUBERT, R. (2000). 4-Aminopyridine affects rat arterialsmooth muscle BK(Ca) currents by changing intracellular pH.British Journal of Pharmacology 131, 16431650.

PLATOSHYN, O., YU, Y., GOLOVINA, V. A., MCDANIEL, S. S., KRICK, S.,LI, L., WANG, J. Y., RUBIN, L. J. & YUAN, X.-J. (2001). Chronichypoxia decreases KV channel expression and function inpulmonary artery myocytes. American Journal of Physiology 280,L801812.

POPP, R., BAUERSACHS, J., HECKER, M., FLEMING, I. & BUSSE, J. (1996).A transferable, beta-naphthoflavone-inducible, hyperpolarizingfactor is synthesized by native and cultured porcine coronaryendothelial cells. Journal of Physiology 497, 699709.

PORTER, V. A., BONEV, A. D., KNOT, H. J., HEPPNER, T. J., STEVENSON,A. S., KLEPPISCH, T., LEDERER, W. J. & NELSON, M. T. (1998).Frequency modulation of Ca2+ sparks is involved in regulation ofarterial diameter by cyclic nucleotides. American Journal ofPhysiology 274, C13461355.

QUIGNARD, J. F., FLTOU, M., THOLLON, C., VILAINE, J. P., DUHAULT,J. & VANHOUTTE, P. M. (2000). Role of endothelial cellhyperpolarization in EDHF-mediated responses in the guinea-pigcarotid artery. British Journal of Pharmacology 129, 11031112.

ROBERTSON, B. E., SCHUBERT, R., HESCHELER, J. & NELSON, M. T.(1993). CGMP-dependent protein kinase activates Ca-activatedK channels in cerebral artery smooth muscle cells. AmericanJournal of Physiology 265, C299303.

SANDOW, S. L. & HILL, C. E. (2000). Incidence of myoendothelial gapjunctions in the proximal and distal mesenteric arteries of the rat issuggestive of a role in endothelium-derived hyperpolarizingfactor-mediated responses. Circulation 86, 341346.

SATAKE, N., SHIBATA, M. & SHIBATA, S. (1997). The involvement ofKCa, KATP and KV channels in vasorelaxing responses toacetylcholine in rat aortic rings. General Pharmacology 28,453457.

SCHIFFRIN, E. L. (2001). A critical review of the role of endothelialfactors in the pathogenesis of hypertension. Journal ofCardiovascular Pharmacology 38, (suppl. 2), S36.

SCHILLING, W. P. (1989). Effect of membrane potential on cytosoliccalcium of bovine aortic endothelial cells. American Journal ofPhysiology 257, H778784.

SEIDEN, J. E., PLATOSHYN, O., BAKST, A. E., MCDANIEL, S. S. & YUAN,J. X.-J. (2000). High K+-induced membrane depolarizationattenuates endothelium-dependent pulmonary vasodilation.American Journal of Physiology 278, L261267.

SMIRNOV, S. V., ROBERTSON, T. P., WARD, J. P. T. & AARONSON, P. I.(1994). Chronic hypoxia is associated with reduced delayedrectifier K+ current in rat pulmonary artery muscle cells. AmericanJournal of Physiology 266, H365370.

SOBEY, C. G. & FARACI, F. M. (1999). Inhibitory effect of4-aminopyridine on responses of the basilar artery to nitric oxide.British Journal of Pharmacology 126, 14371443.

TAKEDA, K., SCHINI, V. & STOECKEL, H. (1987). Voltage-activatedpotassium, but not calcium currents in cultured bovine aorticendothelial cells. Pflgers Archiv 410, 385393.

TARE, M., COLEMAN, H. A. & PARKINGTON, H. C. (2002).Glycyrrhetinic acid derivatives inhibit hyperpolarization inendothelial cells of guinea pig and rat arteries. American Journal ofPhysiology 282, H335341.

THORNELOE, K. S., CHEN, T. T., KERR, P. M., GRIER, E. F., HOROWITZ,B., COLE, W. C. & WALSH, M. P. (2001). Molecular composition of4-aminopyridine-sensitive voltage-gated K+ channels of vascularsmooth muscle. Circulation Research 89, 10301037.

VANHOUTTE, P. M. (1997). Endothelial dysfunction andatherosclerosis. European Heart Journal 18 (suppl. E), E1929.

VON DER WEID, P. Y. & BNY, J. L. (1993). Simultaneous oscillationsin the membrane potential of pig coronary artery endothelial andsmooth muscle cells. Journal of Physiology 471, 1324.

WANG, X. & VAN BREEMAN, C. (1999). Depolarization-mediatedinhibition of Ca2+ entry in endothelial cells. American Journal ofPhysiology 277, H14981504.

WHITE, R. & HILEY, C. R. (2000). Hyperpolarisation of rat mesentericendothelial cells by ATP-sensitive K+ channel openers. EuropeanJournal of Pharmacology 397, 297290.

WOOD, P. G. & GILLESPIE, J. J. (1998). In permeabilised endothelialcells IP3-induced Ca

2+ release is dependent on the cytoplasmicconcentration of monovalent cations. Cardiovascular Research 37,263270.

YAMAGISHI, T., ISHII, K. & TAIRA, N. (1995). Antiarrhythmic andbradycardic drugs inhibit currents of cloned K+ channels, KV1.2and KV1.4. European Journal of Pharmacology 281, 151159.

YAMAMOTO, Y., FUKATA, H., NAKAHIRA, Y. & SUZUKI, H. (1998).Blockade by 18b-glycyrrhetinic acid of intercellular electricalcoupling in guinea-pig arterioles. Journal of Physiology 511,501508.

ZHAO, Y. J., WANG, J., RUBIN, L. J. & YUAN, X. J. (1997). Inhibition ofKv and KCa channels antagonizes NO-induced relaxation inpulmonary artery. American Journal of Physiology 272, H904912.

Acknowledgements This work was supported by the Alberta Heritage Foundation forMedical Research (AHFMR), Canadian Institutes of HealthResearch (MT-13505) and The Wellcome Trust (UK).


Documents

Smooth muscle membrane potential modulates endothelium-dependent relaxation of rat basilar artery via myo-endothelial gap junctions