679

Mechanism of Action and Systemic and RegionalHemodynamics of the Potassium ChannelActivator BRL34915 and Its Enantiomers

R.P. Hof, U. Quast, N.S. Cook, and S. Blarer

BRL34915 (BRL) is a vasodilator with a novel structure. Its mechanism of action, its effects ondepolarization-induced and receptor-mediated blood vessel contraction, and its hemodynamic effectswere investigated. In the rat portal vein, BRL inhibited spontaneous mechanical activity [IC,,0.013 ±0.001 M.M (mean±SEM) for (-)-BRL], the initial effect being a reduced frequency ofcontraction. At higher concentrations, the spontaneous contractions were abolished and *Rb+effluxwas increased. These results suggest that BRL preferentially acts on the pacemaker cells, the K+

channels in other cells being activated only at higher BRL concentrations in this vessel. In experimentson the rabbit aorta, ( —)-BRL shifted the KC1 concentration-response curve to the right andnoncompetitively inhibited responses to angiotensin II. A concentration of 3 |xM ( — )-BRL reducedmaximal angiotensin II contractions by around 50%, higher concentrations having little further effect.This inhibition of angiotensin II contractions is notably greater than that seen with Ca" antagonistsin this vessel. In anesthetized rabbits, (— )-BRL was a peripheral vasodilator at doses of 3-30 jig/kg,but it had no relevant effects on heart rate and myocardial contractile force. This suggests tissueselectivity of this compound or this mechanism of action. BRL preferentially dilated the coronary,gastrointestinal, and cerebral vessels but not those of the kidneys or skeletal muscle as measured withtracer microspheres. This profile of activity is different from that of calcium antagonists or nonspecificvasodilators like dihydralazine. All effects were stereoselective, the (— )-enantiomer being 100 to 200times more active than the (+ )-enantiomer. (Circulation Research 1988;62:679-686)

BRL34915 (BRL, Figure 1) is a peripheralvasodilator.1 Evidence from experiments invitro suggests that this compound relaxes

smooth muscle via the opening of membrane K+

channels.2^1 This novel mechanism is at present notfully understood, and its therapeutic potential andpossible side effects are only now emerging. It has alsorecently been shown that other vasodilator drugs,notably nicorandil,4-5 pinacidil,67 and the sulfatedmetabolite of minoxidil,8 may also act, at least in part,via this mechanism.

BRL is a racemic mixture of two enantiomers (Figure1). Previously published pharmacological (and clini-cal) investigations have been obtained with the race-mate. However, evidence has recently been presentedthat the vasorelaxant and K+-channel activation prop-erties of BRL are stereoselective.9 Recent studies withagents having affinities to both Ca2+ channels and Na+

channels have clearly shown that different stereoiso-mers can have differing or even opposite pharmacody-namic effects.10"12 We have thus prepared the separateenantiomers of BRL by stereochemical synthesis fordetailed evaluation of their biological effects.

The experiments presented here provide a detailedstudy of the effects of BRL and its enant iomers on bloodvessels in vitro and their hemodynamic effects in vivo.The results obtained validate the suggestion that BRLis capable of activating K+ channels and that its effects

From the Cardiovascular Unit of Preclinical Research, SandozLtd., Basel, Switzerland.

Address for correspondence: R.P. Hof, Cardiovascular Unit ofPreclinical Research, Sandoz Ltd., CH-4002 Basel, Switzerland.

Received March 26, 1987; accepted November 10, 1987.

are stereoselective, but they do not exclude the pos-sibility of other actions contributing to its vasodilatoractivity.

Materials and MethodsPreparation of (-)-(3S,4R) and( + )-(3R,4S)-BRL34915

3,4-Dihydro-3-hydroxy-2,2-dimethyl-4-(2-oxo-l-pyrrolidinyl)-2H-l-benzopyran-6-carbonitrile was syn-thesized as follows: 3-bromo-3,4-dihydro-4-hy-droxy-2,2-dimethyl-2H-l-benzopyran-6-carbonitrile13

was acetylated with (S)-camphanic acid chloride. Thetwo resulting diastereoisomers were separated bycolumn chromatography (Silicagel) and cleaved(NaOH, dioxane/water, 97%) to the correspondingepoxides. Epoxide ring opening13 yields the two opti-cally active forms of BRL34915: ( - )-(35,47?)[a]D20 =-52 .1° (c = 0.5,CHC13); and ( + )-(3/?,45): M D 2 0 =+ 51.9 (c = 0.5,CHC13).14 An 'H-NMR shift experi-ment [CDC13, (#)-(-)-l-(anthryl)-2,2,2-trifiuorethanol] snowed the isomeric purity to be better than 97%.

Contraction of Rabbit Aortic RingsMongrel rabbits of either sex weighing 1.8-3.0 kg

were kil led by a sharp blow to the base of the skul 1. Thedescending thoracic aorta was excised immediately,cleaned of connective tissue, and aortic rings 2-3 mmwide were prepared and suspended in 10-ml organ bathscontaining a modified Krebs-Henseleit solution con-taining (mM) NaCl 118.0, KC14.7, MgSO,1.2, CaCl2

1.2, KH2PO4 1.2, NaHCO3 25.0, EDTA 0.03, andglucose 10.1. The solution was maintained at 37° C andcontinually gassed with 5% CO2 in oxygen. The tension

by guest on May 10, 2018

http://circres.ahajournals.org/D

ownloaded from

680 Circulation Research Vol 62, No 4, April 1988

HO CNNC

FIGURE 1. Structure of the enantiomers of BRL34915 (BRL).Left: (- )-(3S,4R)-BRL; right: (+ )-(3R,4S)BRL The molecu-lar weight of the compounds is 286.3.

of the rings was recorded isometrically with electro-mechanical transducers (model UC3, Gould-Statham,Oxnard, California) coupled to an OKI if-800 Model30 microcomputer (Oki Denki, Japan) capable ofsimultaneous recording from eight separate organbaths. At the beginning of the experiments, the ringswere stretched to an initial tension of 1 g and allowedto equilibrate for approximately 1.5 hours, until astable baseline of 100-200 mg tension was obtained.During the equilibration periods, the bathing mediumwas changed at 15-minute intervals to prevent theaccumulation of metabolites.

Cumulative concentration-response curves were ob-tained to either potassium chloride or angiotensin II.15

The maximum angiotensin II concentration was re-stricted to 10~7 M since it was found that higherconcentrations failed to produce a greater steady-statecontraction of the tissue and restricted the subsequentresponsiveness of the tissue to angiotensin II. Ringswere then washed and allowed to relax to baseline, thesubstance under investigation was added to the bath,and 30 minutes later the agonist concentration-responsecurves were repeated. The effects in the presence ofsubstances were expressed as percentages of the pre-ceding maximum response to the agonist ( = 100%) foreach aortic ring. Stock solutions of the enantiomers ofBRL (10~2 M in ethanokporyethylene grycol 400, 1:1)were freshly prepared each day and diluted further withKrebs-Henseleit solution as necessary. One of everyfour rings served as a control, receiving only the vehiclesolution.

S6Rb+ Efflux and Myogenic Activity (Portal Veins)Portal veins of freshly killed rats were incubated for

30 minutes in a HEPES-buffered physiological saltsolution (PSS) gassed with 95% O2-5% CO2 at 37° C.The PSS contained (mM) NaCl 120, KC1 5, NaHCO3

15, NaH2PO4 1.2, MgCl2 2.5, CaCl2 2.5, glucose 11,and HEPES 20, pH 7.4 at 37° C. For loading with^Rb"1", the vein was incubated for an additional 90minutes in PSS to which 5 (xCi/ml 86Rb+ had beenadded. The vein was then mounted in a thermostaticperfusion chamber similar to that described by Boltonand Clark.16 A preload of 500 mg was applied, and thechamber was perfused with PSS at 37° C at a rate of2.5 ml/min. The upper cotton thread was attached toan isometric force transducer (Gould cell, Statham)connected to a home-built amplifier from which thesignal was given to a recorder and an integrator forquantitation of myogenic activity.

For measurement of^Rb* efflux, the perfusate wascollected at a sampling rate of 2 minutes and countedfor radioactivity in the Cerenkov mode at 50% effi-

ciency. The radioactivity remaining in the portal veinat the end of the assay was determined by dissolvingthe vessel in 500 ^1 Lumasolve (Lumac, Schaesberg,The Netherlands) at 50° C overnight. The sample wasthen supplemented with 500 ^1 IN HC1 and 10 mloptifiuor (Packard, Zurich, Switzerland) and countedin the 32P-channel at 100% efficiency.17

The rate constant, k, of ^Rb* efflux was calculatedas described by Quast.17 Drug effects on the efflux rateconstant were calculated as the peak value of k obtainedduring drug application divided by the basal value ofk averaged over 6-10 minutes before drug application.Concentration-effect curves were fitted to a saturationhyperbola or to the Hill equation by nonlinear least-squares analysis. Errors in the parameters were esti-mated in the univariate approximation.18

Hemodynamic Experiments: AnimalsLarge mongrel rabbits (body weight 3.5-4.5 kg)

were anesthetized by injection into an ear vein of 25mg/kg pentobarbital followed by 50 mg/kg phenobar-bital 10-15 minutes later as described in detailelsewhere." The animals were tracheotomized andventilated with a Loosco MK2 infant ventilator usingroom air. The ventilation was adjusted to keep theend-expiratory CO2 between 4.0 and 4.5 volumepercent (measured continuously with a Gould-Godartcapnograph). A positive end-expiratory pressure wasapplied as soon as the thorax was opened. Catheterswere placed in the lower abdominal aorta, the inferiorvena cava, and the jugular vein. The anesthesia wasthen deepened by a further 50 mg/kg phenobarbital.Through a thoracotomy in the left third intercostalspace, the left atrium was cannulated for the injectionof microspheres. The aortic root was cleaned ofconnective tissue, and a flow probe (Narco RT 500,Narco Bio-Systems, Houston, Texas; inner diameter4.5-5.5 mm) was fitted on it. The electromagnetic flowprobe was calibrated in vivo by the reference flowmethod at the time of the last microsphere injection.20

A second thoracotomy in the fifth right intercostal spacewas used to sew a Walton-Brodie strain-gauge onto theright ventricle parallel to the superficial muscle fibers.

MicrospheresWe have described the use of the microsphere

method previously in detail.20"22 In brief, for eachmeasurement of regional blood flow about 1.5 xlO5

microspheres were used, selected from various batcheswith one of the following labels: 12SI, 14lCe, 51Cr, MSr,or ^Sc, as best suited. The spheres were injected intothe left atrium with 1 ml of 0.9% saline. The referencesample was withdrawn from the lower abdominal aortathrough the catheter in the femoral artery at a rate ofapproximately 6 ml/min.

At the end of the experiment, the animals were killedwith an overdose of pentobarbital. The radioactivity ofthe samples was determined in a Packard gammacounter (model 5921) and the spectra processed on anOKI if-800 Model 30 microcomputer according to themethod of Rudolph and Heymann23 with the modifi-cations of the calculations described by Schosser et al.24

by guest on May 10, 2018

http://circres.ahajournals.org/D

ownloaded from

Hofet al Mechanism of BRL34915 and Its Enantiomers 681

Experimental ProtocolThe racemic form and two enantiomers of BRL were

dissolved in a vehicle containing ethanol 0.02 ml,polyethylene glycol 400, and glucose 5% 2 ml/mg ofsubstance (total amount infused per kilogram bodyweight).

The same amount of vehicle alone was infused intoa fourth group of animals (placebo group).

The effects of the two enantiomers were assessedusing the Kruskal-Wallis test, and when it indicated asignificant difference, this was followed by the Dunn-Bonferroni test to determine to which group the effectcould be attributed.

ResultsEffect of BRL34915 and Its Enantiomers on the RatPortal Vein

Figure 2 shows the effect of the two enantiomers ofBRL34915 on spontaneous activity and 86Rb+ efflux inthe rat portal vein in representative experiments. In theleft panel, superfusion of the vessel with 0.03 LIM ofthe active enantiomer, ( - )-BRL, for 20 minutes is seento drastically diminish spontaneous activity by de-creasing the frequency of the spontaneous contractionswhile leaving their amplitude almost unaltered. Therate constant, k, of MRb+ efflux was not affected. Theeffect of (— )-BRL was readily washed out, and appli-cation of 0.3 \xM ( - )-BRL led to complete mechanicalquiescence accompanied by an increase in the flux rateconstant of 24%. After a washout phase of 30 minutes,a concentration of 30 \iM ( - )-BRL was applied. Afterhaving reached a peak value of 165%, the efflux rate

o2.0

E 1.5

*". 1.0 .

100

-.••/ "v.

O0320mm.

30

• \

12PJJM(- )BRL

FIGURE 2. Effects of the (-)- and (+ )-enantiomers of BRL

34915 on spontaneous activity and KRb* efflux in the rat portalvein. Left panel: Continuous simultaneous measurement ofspontaneous activity (upper trace) andofRb* efflux, expressedas the efflux rate constant, k (lower trace). (-)-BRL wassuperfused for 20 minutes (0.03 and 0.3 \iM) or 10 minutes (30JJLA/J. (-)-BRL (0.03 \LM) inhibited spontaneous activity(control level, 0.25 mNlmin) by 83% due to a decrease in thefrequency of the contractions. (-)-BRL (0.3 and 30 \iM)inhibited spontaneous activity completely and increased k by24% and 165%, respectively Right panel: The control level ofspontaneous activity was 0.08 mNlmin. ( + )-BRL (3 \xM),applied for 10 minutes, inhibited spontaneous activity by 55%,again by decreasing the frequency of the contractions withoutany detectable effect on "Rb* efflux. ( + )-BRL, 100 \iM,completely suppressed myogenic activity and increased flux by22%.

250

200

2 150

100'

" 9 8 7 6 5 9 8 7 6 5 4

-log [(-/•) BRL,M] -log [(-/•)BRL, M ]

FIGURE 3. Concentration dependence of the effects of BRL andits enantiomers on the rat portal vein. (*)(-)BRL, (o)( + )BRL,(t>) racemic BRL. The data were obtained as in Figure 2. Shownare mean±SEM, n = 4 per point. Left panel: Inhibition ofspontaneous activity. The curves present a nonlinear least-squares fit to the Hill equation, giving Hill coefficients of1.6±0.2for(-)-BRL, 1.4±0.3 for ( + J-BRL, and 1.4±0.4for (±)-BRL The respective IC^ were 0.013±0.001 pjW,2.2±0.4 \iM, and0.024±0.005 iiM. Right panel: StimulationofRb* efflux. The data for (-)- and ( + )-BRL excluding thepoints at 100 \iM were fit to a saturation hyperbola (see text forvalues of fit).

constant decreased again despite the continued pres-ence of ( — )-BRL. This decrease probably reflectsinactivation of the response (channel? See Quast17).

The effects with ( + )-BRL are shown in the rightpanel of Figure 2. Concentrations at least 100 timeshigher were needed to produce effects similar to thoseseen with ( - )-BRL. At 3 ^M, (+ )-BRL depressedspontaneous activity by 55%, again reducing thefrequency but not the height of the contractile spikes.The rate constant of MRb+ flux was minimally affected.At 100 |xM, ( + )-BRL inhibited spontaneous activitycompletely and increased the flux rate constant by 22%.In this case, washout of the substance was followed byan overshoot of mechanical activity. The racemicmixture gave similar results to those obtained with theactive enantiomer (not shown).

The concentration dependence of the effects of thetwo enantiomers and the racemate on 86Rb+ efflux ispresented in Figure 3. Inhibition of spontaneousactivity (left panel) gave IC^ values of 0.013 ± 0.001^M for ( - ) -BRL and 2.2 ±0 .4 M-M for ( + )-BRL,separated thus by a factor of 170 (mean±SEM). TheIQo value for the racemate (0.024 ± 0.005 LIM) is aboutthe theoretical factor of two higher than that of theactive enantiomer. The concentration-effect curveswere steep, as indicated by the Hill coefficients ofapproximately 1.5.

The concentration dependencies of the ( - ) -en-antiomer and of the racemate were bell-shaped as thepeak response decreased again at 100 (xM, the highestconcentration used. This decrease probably reflects thedesensitization process mentioned above, which mayprevent full development of the effect. Leaving out thepoint at 100 M-M allows a fit of the remaining data toa saturation hyperbola, which gives midpoints of1.7 ±0 .3 v.M for ( - ) -BRL and 2.5 ±0 .5 |xM for(±)-BRL with theoretical saturation values of

by guest on May 10, 2018

http://circres.ahajournals.org/D

ownloaded from

682 Circulation Research Vol 62, No 4, April 1988

131 ±6% and 116 ±5%, respectively. The apparentsaturation behavior of the flux response may well bedue to the occurrence of desensitization and need notreflect saturation of agonist binding.

It should be noted that stimulation of MRb+ efflux isobserved only at relatively high concentrations; 10%stimulation of efflux occurs at concentrations about 10times those necessary to inhibit spontaneous activity,and the midpoints of the flux curves occur at concen-trations 100 times higher than the IQ,, values forspontaneous activity (see Figure 3). As with theinhibition of spontaneous activity, the stimulation of^Rb* efflux by BRL was also stereoselective, theactivity residing in the ( - )-enantiomer.

Inhibition of Contraction of Rabbit AortaIn rabbit aortic rings contracted by potassium

chloride depolarization, concentrations of 0.01 and 1.0\LM (-)-BRL caused a small rightward shift ofthe potassium chloride concentration-response curve(Figure 4, upper panel). Contractions due to potassiumchloride concentrations up to 32 mM were diminishedin the presence of (-)-BRL, but contractions due to

1OOn

5O-

T /

yJ.D 16 19 23 27[KCI

wr

32 39 46 55.mM]

(-)BRL

O1O'8M10 "6M1O"4M

34915

( • )(A)(D)

1OCH

1CT' M

3x10 -

1O"4

U)

(•)(•)

9 83 8 7.5 7-log[Anglot»nslnTj.M]

FIGURE 4. Effects of (-)-BRL upon contractile responses ofrabbit aortic rings elicited either by angiotensin II (lower) ordepolarization with potassium chloride (upper). Thirty minutesprior to the start of the agonist concentration-response curves,(-)-BRL was added to the bath at the concentrations shown inthe diagrams. (-)-BRL did not modify the basal tone of theaortic rings (indicated in the diagrams as drug D) at concen-trations up to 10~4 M. The responses of the rings to angiotensinII (lower diagram) were noncompetitivefy inhibited in thepresence of ( — )-BRL, half maximal-inhibition being seen withapproximately 3 \iAf (-)-BRL. The potassium chlorideconcentration-response curve was shifted slightly to the right by(-)-BRL Contractions are expressed as a percentage of tltepreceding tissue maximum response of each ring to the agonistin the absence of any drug. Symbols represent means of 3 to 7experiments, bars show the SEM.

1OOn

5O-

(*)BRL 34915

O (•)10-6 M (D)1O"5M (o)1O"4M <•)

0 T-iD 16 19 23 27 32 39 46 55

[KCI.mM]

1OOn O (•)

1O"7 M (A)

3X10-6M ( • )

1O"4M (•)

as 8 75 7

- log [Anglotentln TJ. M l

FIGURE 5. Effects of (+)-BRL upon contractile responses ofrabbit aortic rings elicited either by angiotensin II (lower) ordepolarization with potassium chloride (upper). Thirty minutesprior to the start of the agonist concentration-response curves,( + )-BRL was added to the bath at the concentrations shown inthe diagrams. ( + )-BRL did not modify the basal tone of theaortic rings (indicated in the diagrams as drug D) at concen-trations up to 10'' M. ( + )-BRL showed no relevant effects.atconcentrations below 10~4 M against angiotensin 11 or potas-sium chloride responses. At 10~' M, ( + )-BRL inhibited angio-tensin 11 contractions in a noncompetitive manner. Contractionsare expressed as a percentage of the preceding tissue maximumresponse of each ring to the agonist in the absence of any drug.Symbols represent means of 3 to 7 experiment, bars show theSEM.

higher potassium chloride concentrations were littleaffected. At the higher concentration of 100 n,M( - )-BRL, there was also evidence of a reduction in themaximum tissue response to potassium chloride. Withthe ( + )-enantiomer of BRL (Figure 5, upper), con-centrations of 1 and 10 JJLM had no effect on theresponses to potassium chloride and only at 100 \x.Mwas there some evidence of a rightward shift of thepotassium chloride concentration-response curve, witha small reduction in the maximum tissue response topotassium chloride. This diminution of the maximalresponse of aortic rings to depolarization-inducedcontraction by 100 JJLM concentrations of both enan-tiomers of BRL thus appears to lack the stereoselec-tivity that characterizes the other actions of this drug,possibly indicating the existence of nonspecific effectsat these high concentrations.

Aortic rings contracted by receptor stimulation withangiotensin II revealed a quite different antagonisticprofile of BRL. The (-)-enantiomer potently antag-onized angiotensin II responses, resulting in a non-competitive inhibition of the angiotensin IIconcentration-response curve (Figure 4, lower). At aconcentration of 3 \LM, ( —)-BRL inhibited contrac-

by guest on May 10, 2018

http://circres.ahajournals.org/D

ownloaded from

Hofet al Mechanism of BRL34915 and Its Enantiomers 683

tions due to 10~7 M angiotensin II by approximately50%, the higher concentration of 100 \iM ( - )-BRLcausing little further inhibition. Only the highestconcentration (100 p-M) of the ( + )-enantiomer (Figure5, lower) inhibited the angiotensin II responses to anyappreciable degree, 0.1 and 3 u,M appearing inactive.

Experiments in Whole Animals: SystemicHemodynamic Effects

The baseline values of the four groups before drugadministration (Table 1) were comparable (Kruskal-Wallis test). BRL and its enantiomers induced a fallin blood pressure as shown in Figure 6. The race-mic mixture was about half as active as the pure( - )-enantiomer, and the (+ )-enantiomer was morethan 100 times less active. Heart rate showed a weaktendency to increase with the racemic mixture and theless active enantiomer, and it increased significantlywith the (— )-enantiomer (9% at the highest dose).Myocardial contractile force, measured directly with astrain gauge, remained unchanged as compared withplacebo. Cardiac output increased dose-dependentry.The effect reached statistical significance at the highestdoses of both enantiomers. The increases in totalperipheral conductance were more pronounced andreached statistical significance with the highest dosesof each compound, as compared with placebo-treatedanimals.

Regional CirculationThe large increase in total peripheral conductance

induced by BRL indicates that this agent is a peripheralvasodilator. This dilatation was, however, not gener-alized, but showed instead selectivities for somevascular beds. Figures 7A and 7B show the regional

TABLE 1. Baseline Values for Hemodynamic Effects

distribution of the peripheral vascular effects expressedas changes in regional conductance. BRL in all itsforms caused coronary vasodilatation (Figure 7A),which tended to be more pronounced in the outer layerof the left ventricular free wall than in the subendo-cardial layer.

Figure 7B shows further organs where BRL and itsenantiomers elicited vasodiJatation. In the brain, thiseffect was weak, reaching significance only at thehighest dose of the racemic mixture and the( - )-enantiomer. Vasodilatation was pronounced in thestomach, where the effect reached significance at thetwo higher doses of each compound and in the case ofthe (— )-enantiomer at all three doses used. A some-what weaker vasodilatation was seen in the smallintestine, where the effects were again dose-related butsignificant only at the highest dose of the three formsof the compound. Finally, no relevant vasodilatationwas found in the kidney and in skeletal muscle.

DiscussionBRL and its enantiomers are peripheral vasodilators.

Their systemic hemodynamics, but not their regionaleffects, were qualitatively similar to those of vasose-lective calcium antagonists."•25-26 The mechanisms ofaction of calcium antagonists and potassium channelactivators resemble each other in that both are assumedto reduce calcium influx into cells, calcium antagonistsby directly interfering with the opening of slow L-typecalcium channels and K+-channel activators by hyper-polarizing the cell membrane2-4 and thus opposing theopening of such voltage-dependent Ca2+ channels. Thepresent study shows that striking differences exist in theprofiles of action of these two classes of vasodilatorsboth in vitro and in vivo. First, we wish to consider the

Baseline values

Systemic variables

MAP (mm Hg)

HR (beats/min)

CVP (mm Hg)

CF(g)CO (ml/min/kg)

TPC (ml/min/mm Hg/kg)

Regional conductance(ml/min/mm Hg/100 g)

Heart total

Epi

Mid

Endo

Brain

Kidneys

Muscle

Stomach

Small intestine

MAP, mean arterial pressure; HR,TPC, total peripheral conductance.

Control

80 + 3.37

2% ±8.29

2.97±0.52

26±1.04

92 ±10.23

1.16±0.17

2.02 + 0.123

2.53±0.273

3.02±0.251

2.96 + 0.181

0.509 ±0.062

3.54±0.770

0.075 ±0.027

0.796 ±0.247

0.624 + 0.090

heart rate; CVP,

(±)

68 ±3.29

285 + 21.0

2.97±0.45

27 ±4.05

90 + 6.24

1.32±O.O3

2.33±0.109

3.11 ±0.208

2.79±0.183

2.89±0.097

0.552±0.049

4.18 ±0.445

0.068±0.018

0.826 + 0.216

0.627 ±0.132

central venous pressure;

( + )

74 ±1.95

258±7.56

3.47 + 0.74

29 ±1.69

94±5.23

1.27 + 0.06

2.26±0.099

3.18±0.162

2.94±0.146

3.09±0.198

0.597 ±0.023

4.20±0.355

0.045 ±0.003

0.754±0.110

0.449 ±0.030

CF, contractile force;

( - )

71 ±4.77

263 ±10.6

3.76±0.35

32 ±2.39

85±7.51

1.21±0.11

2.08±0.187

2.87 ±0.324

2.49 + 0.196

2.74±0.213

0.452 ±0.032

3.52±0.250

0.049 ±0.006

0.704±0.113

0.432 ±0.041

CO, cardiac output;

by guest on May 10, 2018

http://circres.ahajournals.org/D

ownloaded from

684 Circulation Research Vol 62, No 4, April 1988

c (±

FIGURE 6. Systemic hemodynamic effects of the infusion ofBRL. Cumulative doses of 3, 10, and 30 \ig/kg of the racemate(±), 0.3, 1, and 3 mglkgofthe (+)-enantiomer and 3, 10, and30 \iglkg of the (-)-enantiomer were infused. C indicatesspontaneous changes observed in a control group infused withthe vehicle of the active agents. Effects are shown as percentchanges from the baseline values listed in Table 1. Bars showthe SEM, n = 5. *Significant differences from the vehicle-treatedanimals (Kruskal-Wallis and Dunn-Bonferroni tests, p<0.05).HR, heart rate (beats/min); MAP, mean arterial pressure (mmHg); CF, contractile force (gram); CVP, central venouspressure (mm Hg); CO, cardiac output (ml/min/kg); TPC, totalperipheral conductance (ml/min/mm Hg/kg).

experimental evidence that BRL acts indeed as aK+-channel activator.

Effects of BRL on Rat Portal Vein: Evidence forIncreased K+ Permeability

At concentrations above 0.1 \xM, BRL stimulatedMRb+ efflux from the portal vein (Figure 2), whichreflects an increase in the K+ permeability of thesmooth muscle membrane. This activity of BRLresided primarily in the (— )-enantiomer, with the(+ )-enantiomer increasing flux only at concentrations100 times higher. Additional support for the notion thatBRL acts by increasing the cell membrane K+ perme-ability stems from the particular pattern by whichspontaneous activity of the portal vein was inhibited.As shown in Figure 2, BRL first decreased thefrequency of the contractions without reducing theiramplitude. This is typical of hyperpolarization of apacemaker, by activation of K+ channels after a burstof activity.27 It is known that spontaneous activity in theportal vein originates in some electrically unstableregions (i.e., pacemaker regions from where theexcitation is propagated electrically along the vessel).(For a review, see Ljung.28)

A preferential action of BRL on the pacemakerregions of the portal vein (perhaps due to the particularsliding membrane potential there) would also explainwhy stimulation of ^Rb* efflux is observed only atconcentrations higher than those required to decreasethe frequency of spontaneous activity. If low concen-trations of BRL primarily increase the K+ permeabilityof the pacemaker cells, this will inhibit myogenicactivity by decreasing the frequency without producinga measurable increase in overall ^Rb* efflux, since

pacemaker cells are scarce. The bulk of the smoothmuscle cells presumably responds only at higher BRLconcentrations, leading to the observed increase in^Rb* flux. Moreover, spiking activity itself causesefflux of MRb+, and the decreased frequency after BRLprobably masks the increases elicited by BRL. If spikegeneration in the portal vein is prevented by additionof a Ca2+ antagonist, then BRL-stimulated 86Rb+ effluxcan be detected from concentrations as low as 60 nM,which is more in line with the concentrations thatreduce spontaneous activity in this vessel.17 Two othervasodilators now known to act like BRL via K+-channel activation, nicorandil and pinacidil,46-7 alsodecrease spontaneous activity of the portal vein atconcentrations 10 times lower than those needed tocause 86Rb+ efflux. This, together with the stereose-lectivity of BRL's effects, further implicates theactivation of membrane K+ channels in the vasodilatoraction of BRL.

Profile of Activity on Rabbit Aortic RingsIn rabbit aortic rings contracted by potassium

chloride depolarization, calcium channel antagonistsare known to noncompetitively inhibit the potassiumchloride concentration-response curves but have rela-tively little effect on receptor-mediated contractions.29

C (±)

160

2120

Z 80S 40X

0-

IJ

(+; (-! c* 160-|

j

120

18040

ft-u

(±) (+!

*

I 1#

T1

o160

120

8 0

40

0

j*J

i

[I

j120

80

40

0

c

Bra

i

80

40

0'-20J

C

^x

(i) (

*

ftI

ifl Jll

O» (A 10-

S S1 o

O 2 -10* - 2 0

40n

240

160

80

0-

C

"V

-

* *

i

j

B -40J

m120

80

40

0

j

T

I

iFIGURE 7. Changes of regional conductance induced by theinfusion of BRL and its enantiomers. Same experiments as inFigure 6 where details are given and the symbols explained. Alleffects are shown as percent changes from the baseline valuesgiven in Table 1. Epi, Mid, Endo: subepicardial, middle, andsubendocardial layer of the left ventricular free wall.

by guest on May 10, 2018

http://circres.ahajournals.org/D

ownloaded from

Hof el al Mechanism of BRL34915 and Its Enantiomers 685

However, BRLnoncompetitively inhibited angiotensinII contractions of the aorta but caused only a smallrightwards shift of the concentration-response curve topotassium chloride. In both cases, the activity appearedto reside primarily in the ( - )-enantiomer, with the(+ )-enantiomer being approximately 100 times lessactive. This apparent rightward shift of the potassiumchloride curve by BRL, which has been found previ-ously in the rat aorta and portal vein,2-4 is what onemight expect as a consequence of K+-channel activa-tion causing membrane hyperpolarization, which func-tionally antagonizes the depolarization elicited bypotassium chloride. The inability of BRL to inhibitcontractions by high concentrations of potassiumchloride is indeed predictable from the Nernst equa-tion. These high potassium chloride concentrationsshift the potassium equilibrium potential towards morepositive membrane potentials. Hence, when the bathpotassium chloride concentration exceeds about 32mM (dependent on the intracellular K+ concentrationand the resting membrane potential of the aorta), themembrane potential will always be above the thresholdpotential for opening of voltage-sensitive Ca2+ chan-nels (around - 40 mV), with the consequence that K+-channel activating drugs should no longer be able tomodulate the opening of the Ca2+ channels (see alsoHamilton et al2). Only at 100 \\M was any effect of BRL(or its enantiomers) elicited against these contractionsby high potassium chloride concentrations, the lack ofstereoselectivity at this concentration suggesting it tobe a nonspecific effect.

A 50% inhibition of angiotensin II contractions wasseen with 3 JJLM (— )-BRL. Higher concentrations werefound to cause little further reduction of the angiotensinII responses (Figure 4, lower panel). The underlyingmechanism is not known. Extracellular Ca2+ plays animportant role in the contraction of this blood vessel toangiotensin II.30 However, organic Ca2+ antagonistsinhibit only a small component of the angiotensin IIresponse in rabbit aorta.15-31 Therefore, indirect inhi-bition of Ca2+ entry through classical (L-type) voltage-sensitive Ca2+ channels cannot explain the inhibitoryaction of the K+-channel activators against angiotensinII contraction of the rabbit aorta.31

It is not known to what extent alterations of the cellmembrane potential might influence the release ofintracellularry stored Ca2+. Whether the part of theangiotensin II response antagonized by BRL is duesolely to inhibition of Ca2+ influx has still to beestablished.

Hemodynamic Effects in Whole AnimalsThe systemic vascular effects of BRL and its

enantiomers are qualitatively similar to those of othervasodilators such as hydralazine32 or, perhaps moreappropriately, those of vasoselective calciumantagonists.23 They clearly differed from those ofcalcium antagonists with strong cardiac effects such asverapamil or diltiazem.25 This is an indication that themechanism of action of BRL shows a certain degree ofselectivity for K+ channels in vascular smooth muscle

over those in the heart as has been shown in apreliminary report in vitro.33 Major differences betweenBRL and its enantiomers and these other vasodilatorswere found, however, with respect to the peripheralvascular regions studied. A relatively vasoselectivecalcium antagonist, darodipine, has been examined inthe same experimental model so that a direct compar-ison is possible.18 Darodipine increased coronary andespecially cerebral blood flow much more than BRL atdoses causing comparable falls in blood pressure.Calcium antagonists, especially dihydropyridines,23

but not BRL, strongly dilate skeletal muscle vessels.By contrast, BRL dilated the gastrointestinal vascularbed, which is affected little by calcium antagonists.34-33

Only in the rabbit, but not in other species (discussedin H o f ) , dihydropyridines cause vasoconstriction inthe kidney, whereas BRL was inactive in this vascularbed. In cats, BRL has been found to increase renalflow,1 whereas calcium antagonists cause little or nochange. This looks similar to, but probably is not,autoregulation since these agents have been shown tointerfere with renal autoregulation.36"38 It is perhapsmore appropriate to say that the vasodilator effects ofcalcium antagonists are of sufficient magnitude tocompensate for the simultaneously occurring fall inblood pressure.

It is interesting to note that in cats, the effects of BRLon other vascular beds (femoral and mesenteric bloodflow measured with electromagnetic flowmeters) weresimilar to those we observed in the rabbit experiments.Interestingly, dihydralazine dilated the same vascularbeds as calcium antagonists,32 thus presenting a hemo-dynamic profile of activity different from that of BRL.

In conclusion, all effects of BRL examined in ourexperiments were stereoselective. Our results do notsuggest a separate biological activity for the ( + )-enantiomer and are compatible with a very small(0.5-1.0%) admixture of the (— )-enantiomer. Theresults with the rat portal vein confirm the ability ofBRL to evoke "'Rb* efflux from this tissue but alsodemonstrate effects of BRL (and its enantiomers) onmechanical activity at concentrations below those atwhich ^Rb* efflux could be detected. Additionally, inthe rabbit aorta, the ability of these compounds toattenuate angiotensin II contractile responses could notbe explained merely in terms of an indirect inhibitionof Ca2+ entry through dihydropyridine-sensitive Ca2+

channels. The in vivo results also demonstrate majordifferences in the vasodilator profiles of BRL and Ca2+

antagonists, suggesting once again that this class ofdrugs should not be regarded merely as indirect-actingCa2+ antagonists. It will be interesting to see how theseresults extrapolate to the clinical situation where suchcompounds with a new specific mechanism of actioncould offer exciting new insights into the humanpathophysiology of the cardiovascular system.

References1. Buckingham RE, Clapham JC, Hamilton TC, Longman SD,

Norton J, Poyser RH: BRL34915, a novel antihypertensiveagent: Comparison of effects on blood pressure and other

by guest on May 10, 2018

http://circres.ahajournals.org/D

ownloaded from

686 Circulation Research Vol 62, No 4, April 1988

haemodynamic parameters with those of nifedipine in animalmodels. J Cardiovasc Pharmacol 1986;8:798-804

2. Hamilton TC, Weir SW, Weston AH: Comparison of the effectsof BRL34915 and verapamil on electrical and mechanicalactivity in rat portal vein. Br J Pharmacol 1986;88:103-lll

3. Weir SW, Weston AH: Effect of apamin on responses toBRL34915, nicorandil and other relaxants in the guinea-pigtaenia caeci. Br J Pharmacol 1986;88:113-120

4. Weir SW, Weston AH: The effects of BRL34915 and nicorandilon electrical and mechanical activity and on **Rb efflux in ratblood vessels. Br J Pharmacol 1986;88:121-128

5. Furukawa K, Itoh T, Kajiwara M, Kitamura K, Suzuki H, HoY, Kuriyama H: Vasodilating actions of 2-nicotinamido-ethylnitrate on porcine and guinea-pig coronary arteries. J Phar-macol Exp Ther 1981;218:248-259

6. Bray KM, Newgreen DT, Small RC, Southerton JS, Taylor SG,Weir SW, Weston AH: Evidence that the mechanism of theinhibitory action of pinacidil in rat and guinea-pig smoothmuscle differs from that of gryceryl trinitrate. BrJ Pharmacol1987;91:421-429

7. Cook NS, Quast U, Hof RP, Baumlin Y, Pally C: Similaritiesin the mechanism of action of the two new vasodilator drugs,Pinacidil and BRL34915. J Cardiovasc Pharmacol 1987 (inpress)

8. Meisheri KD, Lipkus LA: Minoxidil sulfate acts as a K*channel agonist to produce vasodilatation. Fed Proc 1987;46:553

9. Buckingham RE, Clapham JC, Coldwell MC, Hamilton TC,Howlett DR: Stereospecific mechanism of action of the novelantihypertensive agent BRL34915 (abstract). BrJ Pharmacol1986;87: 78P

10. Hof RP, Hof A, Ruegg UT, Vogel A: Stereoselectivity at thecalcium channel: Opposite action of the enantiomers of a1,4-dihydropyridine. J Cardiovasc Pharmacol 1985;7: 689-693

11. Hof RP, Hof A, Ruegg UT, Cook NS, Vogel A: Stereoselec-tivity at the calcium channel: Different profiles of haemody-namic activity of the enantiomers of the dihydropyridijiederivative PN200-110. J Cardiovasc Pharmacol 1986;8: 221-226

12. Romey G, Quast U, Pauron D, Frelin C, Renaud JF, LazdunskiM: Na+ channels as sites of action of the cardioactive agentDP1201-106 with agonist and antagonist enantiomers. ProcNatl Acad Sci USA 1987;84:896-900

13. Evans JM, Fake CS, Hamilton TC, Poyser RH, Watts EA:Synthesis and antihypertensive activity of substituted trans-4-Amino-3,4-dihydro-2,2-dimethyl-2H-l-benzopyran3-ols. JMedChem 1983;26:1582-1589

14. Ashwood VA, Buckingham RE, Cassidy F, Evans JM, FarukEA, Hamilton TC, Nash DJ, Stemp G, Willcocks K: Synthesisand antihypertensive activity of 4-(cyclic amido)-2H-l-benzopyrans. J Med Chan 1986;29:2194-2201

15. Hof RP, Vuorela HJ, Neumann P: PY108-O68, a new, potent,and selective inhibitor of calcium-induced contraction of rabbitaortic rings. J Cardiovasc Pharmacol 1982;4:344-351

16. Bolton TB, Clark JP: Actions of various muscarinic agonistson membrane potential, potassium efflux, and contraction oflongitudinal muscle of guinea-pig intestine. Br J Pharmacol1981;72:319-334

17. Quast U: Effect of the K+ efflux stimulating vasodilatorBRL34915 on 86Rb+ efflux and spontaneous activity in guinea-pig portal vein. BrJ Pharmacol 1987;91:569-578

18. Draper NR, Smith H: Applied Regression Analysis, ed 2. NewYork, John Wiley & Sons, Inc, 1981, pp 85-96, 458-517

19. Hof RP: Modification of vasopressin- and angiotensin II-

induced changes by calcium antagonists in the peripheralcirculation of anaesthetized rabbits. Br J Pharmacol1985;85:75-87

20. Hof RP, Hof A: Very small microspheres are useful for thedetermination of cardiac output but not organ blood flow inconscious rabbits. Experientia 1981^37:438-^39

21. Hof RP, Hof A, Salzmann R, Wyler F: Trapping and intra-myocardial distribution of microspheres with different diam-eters in cat and rabbit hearts in vitro. Basic Res Cardiol 1981;76:630-638

22. Hof RP, Wyler F, Stalder G: Validation studies for the use ofthe microsphere method in cats and young minipigs. Basic ResCardiol 1980;75:747-756

23. Rudolph AM, Heymann MA: The circulation of the fetus inutero: Methods for studying distribution of blood flow, cardiacoutput and organ blood flow. Circ Res 1967;21:163-184

24. Schosser R, Arfors KE, Messmer K: Mic-II-A program for thedetermination of cardiac output, arterio-venous shunt andregional blood flow using the radioactive microsphere method.Comput Programs Biomed 1979;9:19-38

25. Hof RP: Calcium antagonists and the peripheral circulation:Differences and similarities between PY108-068, nicardipine,verapamil and diltiazem. Br J Pharmacol 1983;78:375-394

26. Hof RP: Selective effects of different calcium antagonists on theperipheral circulation. Trends Pharmacol Sci 1984;5:100-102

27. Hille B: Ionic Channels of Excitable Membranes. Sunderland,Mass, Sinauer Associates, Inc, 1984

28. Ljung B: Nervous and myogenic mechanisms in the control ofa vascular neuroeffector system. Ada Physiol Scand 1970;349(suppl):33-68

29. Cauvin C, Loutzenhiser R, VanBreemen C: Mechanisms ofcalcium antagonist-induced vasod ilation. Annu Rev PharmacolToxicol 1983;23:373-396

30. Deth R, VanBreemen C: Relative contributions of Ca2+ influxand cellular Ca2+ release during drug induced activation of therabbit aorta. Pflugers Arch 1974;348:13-22

31. Cook NS: The pharmacology of potassium channels and then-therapeutic potential. Trends Pharmacol Sci 1988;9:21-28

32. Hof RP, Hof A: Effects of bromocriptine on the heart and theperipheral circulation of anesthetized cats. / CardiovascPharmacol 1984;6:68-75

33. Grosset A, Hicks PE: Evidence for blood vessel selectivity ofBRL34915. BrJ Pharmacol 1986;89:500P

34. Hof RP, Hof A, Kalkmann HO: Modification of the vasocon-strictor effects of noradrenaline and serotonin by the selectivecalcium antagonist PY108-068 in anaesthetized cats. J Car-diovasc Pharmacol 1985;7(suppl 6):S53-S60

35. Hof RP, Hof A, Neumann P: Effects of PY108-068, a newcalcium antagonist, on general hemodynamics and regionalblood flow in anesthetized cats: A comparison with nifedipine.J Cardiovasc Pharmacol 1982;4:352-362

36. Hashimoto K, Taira N, Chiba S, Hashimoto K, Endoh M,Kokubun M, Kokubun H, Iijima T, Kimura T, Kubota K, OguroK: Cardiohemodynamic effects of BAY a 1040 in the dog.Arzneimittelforschung 1972;22:15-21

37. Loutzenhiser R, Epstein M: Effects of calcium antagonists onrenal hemodynamics. Am J Physiol 1985;249:F619-F629

38. Navar LG, Champion WJ, Thomas CE: Effects of calciumchannel blockade on renal vascular resistance responses tochanges in perfusion pressure and angiotensin-convertingenzyme inhibition in dogs. Circ Res 1986;58:874-881

KEY WORDS • BRL34915 • enantiomers • potassium channels• "Rb* efflux • hemodynamic effects • microspheres

by guest on May 10, 2018

http://circres.ahajournals.org/D

ownloaded from

R P Hof, U Quast, N S Cook and S Blareractivator BRL34915 and its enantiomers.

Mechanism of action and systemic and regional hemodynamics of the potassium channel

Print ISSN: 0009-7330. Online ISSN: 1524-4571 Copyright © 1988 American Heart Association, Inc. All rights reserved.is published by the American Heart Association, 7272 Greenville Avenue, Dallas, TX 75231Circulation Research

doi: 10.1161/01.RES.62.4.6791988;62:679-686Circ Res. 

http://circres.ahajournals.org/content/62/4/679World Wide Web at:

The online version of this article, along with updated information and services, is located on the

  http://circres.ahajournals.org//subscriptions/

is online at: Circulation Research Information about subscribing to Subscriptions: 

http://www.lww.com/reprints Information about reprints can be found online at: Reprints:

  document. Permissions and Rights Question and Answer about this process is available in the

located, click Request Permissions in the middle column of the Web page under Services. Further informationEditorial Office. Once the online version of the published article for which permission is being requested is

can be obtained via RightsLink, a service of the Copyright Clearance Center, not theCirculation Research Requests for permissions to reproduce figures, tables, or portions of articles originally published inPermissions:

by guest on May 10, 2018

http://circres.ahajournals.org/D

ownloaded from