relaxation by bradykinin in porcine ciliary artery -...

Relaxation by Bradykinin in Porcine Ciliary ArteryRole of Nitric Oxide and K+-Channels

Peili Zhu* Jean-Louis Beny,^ Josef Flammer* Thomas F. Luscher,%and Ivan 0. Haefliger*

Purpose. To assess the effects of K+-channel blockers on bradykinin-induced relaxations inporcine ciliary artery.

Methods. Vascular isometric forces were measured with a myograph system. Ciliary vascularrings were precontracted with thromboxane A2 analog (U 46619, 10~7 M) to assess dose-dependent (10~'°-3 X 10~6 M) bradykinin-induced relaxation after addition of one of thefollowing: the nitric oxide (NO) synthase inhibitor Mu-nitro-L-arginine methyl ester (L-NAME,lO"4 M) or inactive enantiomer (D-NAME, 10~4 M); the nonspecific K+-channel blocker tetra-ethylammonium (TEA, 10~2 M); or the ATP-sensitive K+-channel blocker glibenclamide (10~5

M). The effect of TEA on relaxations to the NO donor, sodium nitroprusside (SNP, 10~'°-lO"4 M) was investigated. The membrane potential of vascular smooth muscle cells (VSMC)was recorded after exposure to bradykinin (2.5 X 10~7 M).

Results. Endothelium-dependent relaxations to bradykinin (maximal [max], 99% ± 3%) werestrongly inhibited by L-NAME (max, 39% ± 4%, P < 0.01) and partially by TEA (max, 62%± 2>%,P< 0.01) or glibenclamide (max, 77% ± 4%,P< 0.01). Administration of glibenclam-ide plus L-NAME further suppressed bradykinin-induced relaxation (max, 23% ± 6%; P <0.01), whereas TEA and L-NAME (max, 6% ± 2%; P < 0.01) abolished the relaxation. SNPrelaxations were unaffected by TEA. Bradykinin had no effect on the membrane potential ofVSMC.

Conclusions. In porcine ciliary artery, the endothelium-dependent relaxations to bradykininare primarily mediated by NO and involve K+-channels. As only relaxations to bradykinin,but not those mediated by SNP, were inhibited by TEA, this implies that K+-channel blockersmost likely affect the bradykinin-evoked NO production or release by the endothelium. InvestOphthalmol Vis Sci. 1997;38; 1761-1767.

JL he endothelium plays an important role in the localregulation of vascular tone by releasing endothelium-derived relaxing factors (EDRFs),1'2 both under basalconditions and when activated by different vasodilat-ing agents, such as bradykinin.3'4 It is well acceptedthat one of the EDRFs is nitric oxide (NO) or a closelyrelated molecule derived from the metabolism of L-

From the * Laboratory of Ocular Pharmacology and Physiology, University EyeHospital Basel, the f Department of Zoology and Animal Biology, University ofGeneva, and the % Laboratory of Cardiovascular Research, University HospitalBern, Sxuitzerland.Supported by the Swiss National Research Foundation (grant 32-42564.94), theSchwickerl Foundation, and the Velux Foundation.Submitted for publication August 22, 1996; revised January 22, 1997; acceptedMarch 17, 1997.Proprietary interest category: N.Reprint requests: Ivan 0. Haefliger, Laboratory of Ocular Pharmacology andPhysiology, University Eye Hospital Basel, Mittlere Strasse 91, Basel CH-4012,Switzerland.

arginine by constitutive endothelial nitric oxide syn-thase (NOS).5 However, a number of pharmacologicaland physiological studies have also shown the pres-ence of additional EDRFs6"8 distinct from NO, suchas prostacyclin4 or the putative endothelium-derivedhyperpolarizing factor (EDHF).9'10

On vascular smooth muscle cells (VSMC), bothNO"'12 and EDHF10 have been shown to activate K+-channels, and in some vessels, bradykinin-induced en-dothelium-dependent relaxation could be inhibitedby K+-channel blockers.1314

In porcine ciliary artery, it is known that bradyki-nin can induce endothelium-dependent relax-ations15'16; however, it remains unclear which EDRFscontribute to this response and whether K+-channelsare involved in these relaxations. Therefore, this studyinvestigates the potential role of NO, EDHF, and K+-

InvestigaLive Ophthalmology & Visual Science, August 1997, Vol. 38, No. 9Copyright © Association for Research in Vision and Ophthalmology 1761

Downloaded From: http://iovs.arvojournals.org/pdfaccess.ashx?url=/data/journals/iovs/933425/ on 05/30/2018

1762 Investigative Ophthalmology & Visual Science, August 1997, Vol. 38, No. 9

channels on endothelium-dependent relaxations tobradykinin in porcine ciliary artery.

MATERIALS AND METHODS

Preparation of Vessels and Experimental Setup

In adherence with the ARVO Statement for the Useof Animals in Ophthalmic and Vision Research, eyesfrom farm pigs (100 to 125 kg, 5 to 6 months old)were obtained from a slaughterhouse 5 minutes afterdeath and were transported in cold (4°C) modifiedKrebs-Ringer bicarbonate solution: NaCl, 118.6 mM;KC1,4.7 mM; CaCl2, 2.5 mM; MgSO4,1.2 mM; KH2PO4,1.2 mM; NaHCO3, 25.7 mM; EDTA, 0.026 mM; andglucose, 11.1 mM. Under a microscope (Wild M38,Heerbrugg, Switzerland) the ciliary arteries were dis-sected free and cut into small rings (~2 mm). Twotungsten wires (30 //m and 80 //m) were passedthrough the vascular lumen (vascular diameter 200 to400 (im). One of the wires was connected to a forcetransducer (Showa Sokki LB-5, Rikadenki GmbH,Freiburg, Germany) to measure isometric forces, andthe other was fixed to a micromanipulator (Narishige,Tokyo, Japan) to adjust the muscle length.15 Duringthis procedure, tissues were kept in cold (4°C) modi-fied Krebs-Ringer bicarbonate solution.

Assessment of Endothelial Function

Mounted rings were immersed into organ chambersfilled with Krebs-Ringer bicarbonate solution (37°C;95% O2; 5% CO2; indomethacin 10"5 M to inhibitprostaglandin synthesis). Vessels were stretched in100-mg stepwise increments until their optimal passivetension was reached. The optimal passive tension wasdefined as the level of vascular wall tension at whichcontractions to 100 mM potassium chloride (KC1) be-came maximal. The level of optimal passive tension forporcine ciliary arteries was determined in preliminaryexperiments and used in all subsequent experimentalprotocols (1025 ± 48 mg, n = 4, data not shown).The integrity of the endothelial function was assessedby adding bradykinin (10~7 M) on vessels precon-tracted with thromboxane A2 analog, U 46619 (10~7

M). The endothelial function was considered to bepreserved if under these conditions bradykinin evokedno less than 80% relaxation. In a few vessels the integ-rity of the endothelium was sufficiently compromisedby rubbing the luminal surface of the vessels with ahuman hair as determined by the absence of a bradyki-nin response.17

Isometric Tension Measurements

The effect of bradykinin was assessed by adding cumu-lative concentrations (10"10-3 X 10~6 M) of the drugson vessels precontracted with U 46619 (10~7 M). Con-

centration-response curves to bradykinin were con-structed in the presence of the following: Krebs solu-tion (control, n = 6); the inhibitor of NOS, M*;-nitro-L-arginine methyl ester (L-NAME; 10~4 M; n = 6);the biologically inactive D-enantiomer of L-NAME, Nu-nitro-D-arginine methyl ester (D-NAME; 10~4 M; n =6); the nonspecific K+-channel blocker, tetraethylam-monium (TEA; 10'2 M; n =5); the ATP-sensitive K+-channel blocker, glibenclamide (10~5 M; n = 5) alone;the coadministration of 10"4 M L-NAME and 10~2 MTEA (n = 6); or 10"4 M L-NAME and 10"5 M gliben-clamide (n = 5). These drugs were used with a 30-minute prei'ncubation time. Because contractionsevoked by U 46619 tended to be increased by L-NAMEor TEA and decreased by glibenclamide, the concen-tration of U 46619 used for precontraction was ad-justed (3 X 10~8-3 X 10~7 M) to achieve, in all experi-ments, an equivalent level of precontraction (~70%of 100 mM KC1 contractions). The effect of the NOdonor, sodium nitroprusside (SNP), was also assessedby adding cumulative concentrations (10~10-10~4 M)to vessels precontracted with U 46619 (10~7 M). Con-centration-response curves to SNP were constructedin the presence or absence of 10~2 M TEA.

Electrophysiology of Smooth Muscle Cells

Transmembrane potentials of VSMC were recorded,as described elsewhere.18 The strip was incubated ina 200-^1 Perspex bath continuously perfused with oxy-genated Krebs solution (3.7 ml/min) at 36°C. Drugswere delivered to the preparations by diluting themdirectly in the perfusion solution. One extremity ofthe strip was pinned on a silicon rubber surface withthe intimal surface facing up. The membrane poten-tial was measured with a conventional glass microelec-trode (60 to 80 Mfi) filled with 3 M KC1. Cells in whichmembrane potentials were recorded were locatednear the fixed points of the tissue to reduce problemsrelated to muscle movements. The criteria for ac-cepting a recording were a stable membrane potentialfor more than 1 minute and a sharp rise to 0 mV whenthe electrode was withdrawn from the recorded cell.

Lucifer Yellow Injection

To verify that the experimentally selected cell was actu-ally a VSMC, lucifer yellow was used.19'20 Briefly, thecell membrane potential was measured with a glassmicroelectrode with its tip filled with a lucifer yellowsolution (5% in water) and back-filled with 150 mMLiCl. The microiontophoretic injection of the fluo-rescent dye was achieved by passing a direct currentof 0.35 nA through the electrode for 0.5 to 5 min. Toidentify the injected cells, the tissue was fixed with 4%paraformaldehyde in phosphate-buffered saline (pH7.4). The luminal face of the strip was examined witha fluorescence microscope (Nikon diaphot; excitation


Nitric Oxide and K+-Channels in Porcine Ciliary Artery 1763

wavelength, 450 to 490 nm) or a confocal microscope(Zeiss LSM) and then processed for electron micros-copy.

Electron Microscopy for Cytologic Detection ofthe Dye

After fluorescence examination, the fixed tissues wereembedded in polyethylene glycol (MWt, 1000) usingthe methods already described.19 The embedded tis-sues were cut with a "tissue chopper" in thick slices.The polyethylene glycol was removed from the slices.Only the slice (s) containing fluorescent material wasthen embedded in Epon without osmification for 1-fim serial semithin sectioning. After having reachedthe first semithin sections containing fluorescent ma-terial, thin sections of about 60 nm were cut. Luciferyellow-immunoreactive sites were localized on thesethin sections by the protein A-gold-citrate techniqueusing an anti-lucifer yellow antiserum. The thin sec-tions were examined with a Zeiss EM 10 electron mi-croscope.

Membrane Potential Measurement

The VSMC membrane potential was first recorded un-der basal conditions. Then, the tension and VSMCmembrane potential were measured after exposure to1.5 X 10"7 M of U 46619 alone, and after addition of2.5 X 10~7 M bradykinin. If the membrane potentialof recorded cells was unaffected after exposure to U46619 and bradykinin, exposure to phenylephrine (5X 10"5 M) was used as a control.

Drugs and Statistical Analysis

Bradykinin, Mi>nitro-L-arginine methyl ester, Nu>-ni-tro-D-arginine methyl ester, sodium nitroprusside, in-domethacin, U 46619, glibenclamide, and tetraethyl-ammonium were purchased from Sigma (St. Louis,MO). Glibenclamide was dissolved in dimethylsulfox-ide (0.06% DMSO), indomethacin in Na2CO3 (10~5

M), and all other drugs in distilled water. Concentra-tions are expressed as final molar organ chamber con-centrations. Relaxations were expressed as percentageof contractions to U 46619 (10~7 M). For each concen-tration-response curve, the maximal response (max),the area under the concentration-response curve(AUC), and the concentrations (expressed as negativelog M concentration) of bradykinin evoking 25%(pD25) or 50% (pD50) relaxation were calculated. Re-sults are given as mean and standard error of the mean(mean ± SEM), and n equals the number of animalsstudied (one eye per animal). Unpaired Student's t-tests or one factor ANOVA Scheffe F test were usedfor statistical comparison. Two-tailed P =s 0.05 wasconsidered statistically significant.

C MO

rcen

i

©

a:

OS

4661

©

c©

ei

oU

01

20

40

60

80

100

10

Bradykinin (-log M)

FIGURE l. Effect of the inhibitor of nitric oxide formation,L-NAME, on the endothelium-dependent relaxation to bra-dykinin. In porcine ciliary artery, bradykinin evoked a con-centration-dependent relaxation that was almost completelyabolished (P < 0.001) after removal of the endothelium.The relaxations to bradykinin were significantly blunted bypretreatment with L-NAME (10~4 M, P < 0.001), but not byD-NAME (10~4 M, NS). O = control (n = 6); • = withoutendothelium (n = 6); A = with L-NAME (n = 6). A = withD-NAME (n = 6).

RESULTS

Endothelium and Relaxations to Bradykinin

In porcine ciliary arteries precontracted with thethromboxane A2 analog, U 46619 (10~7 M), bradyki-nin (10~"10-3 X 10~6M) evoked concentration-depen-dent relaxations. After removal of the endothelium,the relaxing effect of bradykinin was completely abol-ished, confirming that relaxations to bradykinin areendothelium-dependent (max, 99 ± 3% versus 5.4 ±2%; AUC, 188 ± 23 versus 447 ± 9; P < 0.001;Fig. 1).

Nitric Oxide and Relaxations to Bradykinin

In the presence of the NOS inhibitor, L-NAME (10~4

M), relaxations to bradykinin were significantly inhib-ited (max, 99 ± 3% versus 39 ± 4 %; pD25, 8.6 ± 0.2versus 6.3 ± 0.8; AUC, 188 ± 23 versus 377 ± 9; P <0.001; Fig. 1). In contrast, in the presence of D-NAME,which is the biologically inactive D-enantiomer of L-NAME, relaxations to bradykinin were not significantdifferent from that of controls (max, 94 ± 4% versus99 ± 3%, NS; Fig. 1).

K+-Channels and Relaxations to BradykininTo address whether in porcine ciliary arteries relax-ations to bradykinin involved K+-channels, endothe-lium-dependent relaxations to bradykinin were as-sessed in the presence of K+-channel blockers. In ves-



sels precontracted with U 46619 (10 7 M), relaxationsto bradykinin (10"10-3 X 10"6 M) were inhibited bythe nonspecific K+-channel blocker, TEA (10~2 M).In the presence of TEA, the concentration-responsecurve to bradykinin was shifted to the right, and themaximal response was reduced (max, 99 ± 3% versus62 ± 3%; pD50, 8.1 ± 0.2 versus 6.4 ± 0.3; AUC, 188± 23 versus 334 ± 14; P < 0.001; Fig. 2, upper panel).Furthermore, in the presence of both TEA (10~2 M)and the inhibitor of NOS, L-NAME (10~4 M), relax-ations to bradykinin were completely inhibited (max,62 ± 3% versus 6 ± 2%; AUC, 334 + 14 versus 444± 10; P < 0.001; Fig. 2, upper panel).

Similar results were observed in vessels precon-tracted with U 46619 (10~7 M) and pretreated with theATP-sensitive K+-channel blocker, glibenclamide (10~5

M). In the presence of glibenclamide, the concentration-relaxation curve to bradykinin (10"10-3 X 10~6 M) wasalso significantly reduced and shifted to the right (max,99 ± 3% versus 77 ± 4%; P < 0.01; pD50, 8.1 ± 0.2versus 7.3 ± 0.3, P < 0.05; AUC, 188 ± 23 versus 300±22, P < 0.01; Fig. 2, lower panel). In the presence ofboth glibenclamide (10~5 M) and L-NAME (10~4 M),relaxations to bradykinin were almost abolished (max,77 ± 4% versus 23 ± 6%; AUC, 300 ± 22 versus 420 ±15; P < 0.001; Fig. 2, lower panel).

K+-Channels and Relaxations to Nitric Oxide

To address whether in porcine ciliary arteries relax-ations to NO involve opening of K+-channels, the ef-fect of the NO donor, SNP, was assessed in the pres-ence or absence of the nonspecific K+-channelblocker, TEA (10~2 M). In porcine ciliary arteries pre-contracted with U 46619 (10"7 M), SNP (10-10-10-4

M) evoked concentration-dependent relaxations.These relaxations were not affected by the presenceor absence of TEA (max, 101 ± 3% versus 7 ± 4%;pD50, 5.9 ± 0.2 versus 5.7 ± 0.2; AUC, 387 ± 12 versus396 ± 18; NS; Fig. 3).

Membrane Potential After Bradykinin Exposure

Electrophysiologic experiments showed that, in por-cine ciliary arteries, the VSMC membrane potentialwas -66 ± 4 mV (n = 5) under basal conditions. Thethromboxane A2 analog, U 46619 (1.5 X 10"7 M),contracted the strips and did not significantly changethe VSMC membrane potential (-68 ± 5 mV, n = 5,NS). In the strips incubated with U 46619 (1.5 X 10~7

M), bradykinin (2.5 X 10~7 M) relaxed the strips buthad no significant effect on the membrane potentialof VSMC. The change of VSMC membrane potentialinduced by bradykinin was 0.2 ±1.7 mV (n = 5, NS,Fig. 4). As a control, phenylephrine (5 X 10~5 M) wasused to show that this electrically "silent cell" wasactually able to depolarize despite the lack of mem-brane potential changes after bradykinin exposure.

£ 5 20i

I 5u&

~ O

40

60

h so

100

10

4)

§ sou soId ^

S S.2 e

« isou

0 1

20

40

60

80

100

10

Bradykinin (-log M)

FIGURE 2. (upper panel) Effects of the nonspecific K+-channelblocker, tetraethylammonium, and of the inhibitor of NOformation, L-NAME, on endothelium-dependent relaxationsto bradykinin in porcine ciliary arteries. The relaxation tobradykinin was significantly blunted (P< 0.001) by tetraeth-ylammonium alone (10~2 M) and almost completely abol-ished (P < 0.001) by the coadministration of tetraethylam-monium (10~2 M) and L-NAME (10~4 M). O = control (n= 6); • = tetraethylammonium (n = 5). D = tetraethylam-monium and L-NAME (n = 6). (loiuer panel) Effects of theATP-sensitive K+-channel blocker, glibenclamide, and L-NAME on endothelium-dependent relaxations to bradyki-nin. The relaxation to bradykinin was partially inhibited (P< 0.01) in the presence of both glibenclamide (10~5 M)and L-NAME (10~4 M). O = control (n = 6); A = gliben-clamide (n = 5); A = glibenclamide and L-NAME (n = 5).

Because of the lack of effect of U 46619 and bradyki-nin on the membrane potential of recorded cells, itwas important to demonstrate that the electrodes wereinserted into smooth muscle cells. This was done byidentifying the recorded cells by dye injection andfurther examination with fluorescence and sometimeselectron microscopy, as shown in Figure 4.


Nitric Oxide and K+-Channels in Porcine Ciliary Artery

0

£ *r 20

•tat OS

33 vo

a is

M

40

80-

100-

10 8 7 6

SNP(-logM)

FIGURE 3. Effect of tetraethylammonium on endothelium-inde-pendent relaxations to the NO donor, sodium nitroprusside(SNP). The relaxation to SNP was not affected by tetraethylam-monium (10~a M; NS). • = with tetraethylammonium (n =7); O = without tetraethylammonium (n = 7).

DISCUSSION

The major findings from the current study are that,in porcine ciliary artery, bradykinin-evoked endothe-lium-dependent relaxations mediated are essentiallyby NO and do not involve changes in the membranepotential of VSMC. The endothelium-dependent re-laxations to bradykinin were blunted by K+-channelblockers, such as TEA or glibenclamide. Furthermore,only relaxations to bradykinin, but not those to SNP,were inhibited by the nonspecific K+-channel blocker,TEA, suggesting that K+-channel blockers inhibit bra-dykinin-evoked relaxations most likely by affecting theproduction or the release of NO by the endothelium.

Relaxations to bradykinin were almost completelyabolished after damaging the endothelium with a hu-man hair, or after exposing the vessels to the NOSinhibitor, L-NAME (but not by its biological enanti-oiner, D-NAME), demonstrating that NO is a majorEDRF involved in endothelium-dependent relaxationsto bradykinin (Fig. 1), The current observation is inagreement with previous reports showing that relax-ations to bradykinin were significantly inhibited, al-though only partially, after damaging the endothe-lium chemically with saponin or after inhibiting NOformation with A^-monomethyl-L-arginine (L-NMMA).15'21 The more pronounced inhibition of bra-dykinin-induced relaxations observed in the currentstudy reflects the fact that L-NAME is a more efficientinhibitor of NOS than is L-NMMA,22"24 and the factthat the endothelial damage after rubbing the lumenof the vessel with a human hair is probably more pro-

1765

nounced than after perfusing the lumen of the vesselwith saponin.

It is known that NO can relax VSMC by openingK+-channels, either directly, ">2t> or through the activa-tion of a cGMP-dependent protein kinase pathway.12'21'To assess whether in porcine ciliary artery, the endo-thelium-dependent relaxations to bradykinin, primar-ily mediated by NO, involve the opening of K+-chan-nels, the relaxing properties of bradykinin were as-sessed in the presence of different K+-channelblockers. The maximal relaxations evoked by bradyki-nin were partially inhibited by the nonspecific K+-channel blocker, TEA,27 and further abolished by

1 min

-50

-60

-70

FIGURE 4. (lower panel) Effect of 2.5 X 10 7 M bradykinin(BK) and 5 X 10~5 M phenylephrine (Ph) on the membranepotential of a smooth muscle cell in a strip of a porcineciliary artery continuously exposed to 1.5 X 10~7M U 46619.Phenylephrine was used to show that this "silent cell" couldbe depolarized despite the lack of effect to bradykinin. Toconfirm that the recorded cell was a smooth muscle cell,fluorescent dye lucifer yellow was microiontophoretically in-jected through the recording electrode into the recordedcell. After fixation, the strip was examined with a confocalmicroscope (left upper panel) the visualized slice was parallelto the endothelium) and then processed for electron mi-croscopy. In that case, lucifer yellow was revealed using ananti-lucifer yellow antibody that was visualized by protein Agold particles (small black spots in right upper panel). L =lumen of the artery; E = endothelium; IEL = internal elasticlamina. The anoxv in the middle upper panel shows therecorded cell as demonstrated in the right upper panel.



coadministration of TEA and L-NAME (Fig. 2). Similarresults could be found when glibenclamide, an ATP-sensitive K+-channel blocker,28 was used, demonstra-ting that opening of K+-channels is involved in therelaxation of porcine ciliary artery to bradykinin. Inthe literature, a similar observation has also been re-ported.29 For example, in the rabbit cerebral arteryglibenclamide could inhibit endothelium-dependentrelaxations to bradykinin.29

Because the relaxations of porcine ciliary arteryto bradykinin, which are mainly mediated by NO, canbe partially inhibited by K+-channel blockers, wewanted to determine whether the relaxations to theNO donor, SNP, could also be modulated by the non-specific K+-channel blocker, TEA. The presence ofTEA had no significant effect on SNP relaxations,demonstrating that the relaxing effect of NO doesnot involve opening of K+-channels in porcine ciliaryartery (Fig. 3). This observation was also corroboratedby electrophysiological studies, in which no VSMCmembrane potential hyperpolarization could be de-tected after exposure of porcine ciliary artery to brady-kinin (Fig. 4). Indeed, as the opening of K+-channelsshould induce a hyperpolarization,712 the absence ofa change in VSMC membrane potential after exposureto bradykinin further confirms that the relaxation ofporcine ciliary artery to bradykinin does not involvethe opening of K+-channels on VSMC. In addition, thefact that bradykinin-induced endothelium-dependentrelaxation in porcine ciliary artery was not accompa-nied by a hyperpolarization in VSMC membrane po-tential also indicates that, in contrast to the porcinecoronary artery,9'30 relaxations to bradykinin are notmediated by a EDHF.

The observation that endothelium-dependent re-laxations to bradykinin are mainly mediated by NOand that only relaxations to bradykinin (but not thoseto the NO donor, SNP) are inhibited by the ^-chan-nel blocker, TEA, suggests that, in porcine ciliary ar-tery, the inhibitory effects of K+-channel blockers onthe bradykinin-induced relaxations take place at thelevel of the endothelium. It has been reported that byopening K+-channels, bradykinin can induce a hyper-polarization of endothelial cells,6'3132 which in turncan lead to a transient increase of cytosolic free cal-cium.33'34 As constitutive NO synthase is calcium-cal-modulin-dependent, intracellular increase in Ca2+ en-hances the synthesis of NO.35 This mechanism wouldexplain why, by interfering with the endothelial syn-thesis of NO,36 K+-channel blockers specifically inhibitrelaxations to bradykinin but not those to SNP, eventhough NO is the main mediator of bradykinin-in-duced relaxations. This hypothesis was actually sup-ported by electrophysiological experiments per-formed in porcine ciliary artery, in which bradykinincould induce a hyperpolarization of endothelial cells.

In summary, this study shows that in porcine cili-ary artery endothelium-dependent relaxations to bra-dykinin are primarily mediated by NO and can bepartially inhibited by K+-channel blockers. Relaxationsto bradykinin, but not those to SNP, were inhibitedby the nonspecific K+-channel blocker, TEA, sug-gesting that K+-channel blockers interfere with thebradykinin-induced synthesis or release of NO by theendothelium. Since K+-channel blockers can modu-late relaxations to bradykinin in porcine ciliary artery,K+-channel openers, such as nicorandil, pinacidil, orcromakalim,38 might be considered as useful tools forthe management of some functional ocular blood flowdisturbances,39'40 an approach that has already beenproposed for the treatment of cerebral vasospasm.41

Key Words

endothelium-derived hyperpolarizing factor (EDHF), endo-thelium-derived relaxing factors (EDRFs), K+-channel, ocu-lar blood flow, vasoactive agents

References

1. Furchgott RF, Zawadzki JV. The obligatory role of en-dothelial cells in the relaxation of arterial smoothmuscle by acetylcholine. Nature. 1980;288:373-376.

2. Furchgott RF. The discovery of endothelium-depen-dent relaxation. Circulation. 1993;87(Suppl. V):3-8.

3. Haefliger IO, Meyer P, Flammer J, Luscher TF. Thevascular endothelium as a regulator of the ocular cir-culation: a new concept in ophthalmology? Surv Oph-thalmol 1994; 39:123-132.

4. Luscher TF, Tanner FC. Endothelial regulation of vas-cular tone and growth. Am J Hypertens. 1993;6:283S-293S.

5. Palmer RMJ, Ferrige AG, Moncada S. Nitric oxide re-lease accounts for the biological activity of endothe-lium-derived relaxing factor. Nature. 1987; 327:524-526.

6. BenyJL. Endothelial and smooth muscle cells hyper-polarized by bradykinin are not dye coupled. Am JPhysiol. 1990;258:H836-H841.

7. Angus JA, Cocks TM. Endothelium-derived relaxingfactor. Pharmacol Ther. 1989;41:303-352.

8. Rubanyi G, Vanhoutte PM. Nature of endothelium-derived relaxing factor: are there two relaxing media-tors? CircRes. 1987;61(Suppl. II):61-67.

9. BenyJL, Brunet P, Huggel H. Interaction of bradyki-nin and des-Arg-bradykinin with isolated pig coronaryarteries: mechanical and electrophysiological events.RegulPept. 1987; 17:181-190.

10. Suzuki H, Chen G, Yamamoto Y. Endothelium-derivedhyperpolarizing factor (EDHF). Jpn Circ J. 1992;56:170-174.

11. Bolotina VM, Najibi S, Palacino JJ, Pagano PJ, CohenRA. Nitric oxide directly activates calcium-dependentpotassium channels in vascular smooth muscle. Na-ture. 1994; 368:850-853.

12. Tare M, Parkington HC, Coleman HA, Neild TO,Du'sting GJ. Hyperpolarization and relaxation of arte-


Nitric Oxide and K+-Channels in Porcine Ciliary Artery 1767

rial smooth muscle caused by nitric oxide derivedfrom endothelium. Nature. 1990;346:69-71.

13. Hecker M, Bara AT, Bauersachs J, Busse R. Character-ization of endothelium-derived hyperpolarizing factoras a cytochrome P450-derived arachidonic acid metab-olite in mammals. JPhysol. 1994;481:407-414.

14. Holzmann S, Kukovetz WR, Windischofer W, PaschkeE, Graier WF. Pharmacologic differentiation betweenendothelium-dependent relaxations sensitive and re-sistant to nitro-L-arginine in coronary arteries. JCardio-vasc Pharmacol. 1994;23:747-756.

15. Yao K, Tschudi M, Flammer J, Luscher TF. Endothe-lium-dependent regulation of vascular tone of the por-cine ophthalmic artery. Invest Ophthalmol Vis Sci.1991;32:1791-1798.

16. Haefliger IO, Flammer J, Luscher TF. Nitric oxideand endothelin-1 are important regulators of humanophthalmic artery. Invest Ophthalmol Vis Sci. 1992;33:2340-2343.

17. Meyer P, Flammer J, Luscher TF. Local action of therenin angiotensin system in the porcine ophthalmiccirculation: effects of ACE-inhibitors and angiotensinreceptor antagonists. Invest Ophthalmol Vis Sci 1995;36:555-562.

18. Beny JL, Brunet PC, Huggel H. Effect of mechanicalstimulation, substance P and vasoactive intestinal poly-peptide on the electrical and mechanical activities ofcircular smooth muscles from pig coronary arteriescontracted with acetylcholine: role of endothelium.Pharmacology. 1986; 33:61-68.

19. BenyJL, Gribi F. Dye and electrical coupling of endo-thelial cells in situ. Tissue Cell. 1989; 21:797-802.

20. BenyJL, ConnatJL. An electron-microscopic study ofsmooth muscle cell dye coupling in the pig coronaryarteries. Circ Res. 1991;70:49-55.

21. Haefliger IO, Flammer J, Luscher TF. Heterogeneityof endothelium-dependent regulation in ophthalmicand ciliary arteries. Invest Ophthalmol Vis Sci. 1993;34:1722-1730.

22. Rees DD, Palmer RMJ, Schulz R, Hodson HF, Mon-cada S. Characterization of three inhibitors of endo-thelial nitric oxide synthase in vitro and in vivo. BrJPharmacol. 1990; 101:746-752.

23. Nakaike R, Shimokawa H, Yasutake H, et al. Effectsof L-arginine analogues on vasomotion of isolated por-cine coronary arteries. AmJPhysiol. 1995;268:H1966-H1972.

24. GureviciusJ, Salem MR, Metwally AA, Silver JM, Crys-tal GJ. Contribution of nitric oxide to coronary vasodi-lation during hypercapnic acidosis. Am J Physiol.1995;268:H39-H47.

25. Bialecki RA, Stinson-Fisher C. Kca-channel antagonistsreduce NO donor-mediated relaxation of vascular andtracheal smooth muscle. Am J Physiol. 1995;268:L152-L159.

26. Robertson BE, Schubert R, Hescheler J, Nelson MT.cGMP-dependent protein kinase activates Ca-activated

K channels in cerebral artery smooth muscle cells. AmJ Physiol. 1993;265:C299-C303.

27. Cook NS. The pharmacology of potassium channelsand their therapeutic potential. Trends Pharmacol Sci.1988;9:21-28.

28. Klieber HG, Daut J. A glibenclamide-sensitive potas-sium conductance in terminal arterioles isolated fromguinea pig heart. Cardiovasc Res. 1994;28:823-830.

29. Brayden JE. Membrane hyperpolarization is a mecha-nism of endothelium-dependent cerebral vasodila-tion. AmJPhysiol. 1990;259:H668-H673.

30. Tschudi M, Richard V, Biihler FR, Luscher TF. Impor-tance of endothelium-derived nitric oxide in porcinecoronary resistance arteries. Am J Physiol. 1991;260:H13-H20.

31. Brunet PC, BenyJL. Substance P and bradykinin hy-perpolarize pig coronary artery endothelial cells inprimary culture. Blood Vessels. 1989;26:228-234.

32. Luckhoff A, Busse R. Calcium influx into endothelialcells and formation of endothelium-derived relaxingfactor is controlled by the membrane potential. Pfliig-ersArch. 1990; 416:305-311.

33. Schilling WP. Effect of membrane potential on cyto-solic calcium of bovine aortic endothelial cells. AmJPhysiol. 1989;257:H778-H784.

34. Colden-Stanfield M, Schilling WP, Ritchie AK, EskinSG, Navarro LP, Kunze DL. Bradykinin-induced in-creases in cytosolic calcium and ionic currents in cul-tured bovine aortic endothelial cells. Circ Res.1987;61:632-640.

35. Luscher TF, Vanhoutte PM. The Endothelium—Modula-tor of Cardiovascular Function. Boca Raton, CRC Press.1990:1-215.

36. Cooke JP, Rossitch JE, Andon NA, Loscalzo J, DzauV. Flow activates an endothelial potassium channel torelease an endogenous nitrovasodilator. / Clin Invest.1991;88:1663-1671.

37. Beny J-L, Zhu P, Haefliger IO. Lack of bradykinin-induced smooth muscle hyperpolarization despiteheterocellular dye coupling and endothelial cell hy-perpolarization in porcine ciliary artery. / Vase Res.1997(in press).

38. Edwards G, Weston AH. Potassium-channel openersand vascular smooth muscle relaxation. PharmacolTher. 1990; 48:237-258.

39. Haefliger IO, Anderson DR. Blood flow regulation inoptic nerve head. In: Ritch R, Shields MB, Krupin T,eds. The Glaucomas, 2nd ed. St Louis: Mosby-YearBook; 1996:189-197.

40. Flammer J. To what extent are vascular factors in-volved in the pathogenesis of glaucoma? In: KaiserHJ, Flammer J, Hendrickson PH, eds. Ocular BloodFlow. Basel: Karger; 1996:12-39.

41. Zhang H, Cook D. Cerebral vascular smooth musclepotassium channels and their possible role in the man-agement of vasospasm. Pharmacol Toxicol. 1994;75:327-336.


relaxation by bradykinin in porcine ciliary artery -...

Documents