European Journal of Pharmacology, 222 (1992) 143-151 0 1992 Elsevier Science Publishers B.V. All rights reserved 0014-2999/92/$05.00

143

EJP 52722

Effect of Rb+ on cromakalim-induced relaxation and ion fluxes in guinea pig trachea Keith A. Foster, Jonathan SmitltKiitre

Beecham

R.S. Arch, Penny N. Newson, Pl~armace~tticals,

Great

Burgh,

Deborah

Yew Tree Bottom

Road,

Shaw and Stephen Epsom,

Surrey

KT18

SXQ,

G. Taylor UK

Received 19 March 1992, accepted 4 August 1992

The effects of cromakalim, verapamil and.salbutamol have been examined in guinea pig trachealis smooth muscle in both Krebs physiological salt solution and Krebs solution where K+ has been replaced by Rb+. Cromakalim-induced relaxation in the presence of Rb+ was reduced in extent and became transient, whilst the relaxation response to verapamil was enhanced and that to salbutamol unaffected. The transient relaxation occurring in Rb+ was blocked by quinidine and glibenclamide. The presence of extracellular Rb+ also prevented cromakalim-stimulated efflux of both “Rb+ and 42/43K+. There was, however, no effect on cromakalim-stimulated “Rb+ uptake. It is proposed that cromakalim is opening two populations of potassium channel in guinea pig tracheal smooth muscle, one of which is susceptible to blockade by Rb+ and one of which is not. The latter channel appears to play the dominant role in cromakalim-stimulated uptake, and is responsible for the transient relaxation response in the presence of rubidium, whilst the former is responsible for the maintained relaxation. Cromakalim; Ion flux; K+ channels; Rb+; Smooth muscle relaxation

1. Introduction

Cromakalim (BRL34915) is a novel anti-hypertensive and bronchodilator agent that has potential for the treatment of asthma (Buckingham et al., 1986; Arch et al., 1988a,b; Williams et al., 1990; Bowring et al., 1991). It is believed that cromakalim causes smooth muscle relaxation by opening plasmalemmal potassium channels (Hamilton et al., 1986). The consequent hyperpolarization of the smooth muscle cell membrane opposes the opening of voltage-dependent Ca*+ channels and may also affect other potential-dependent aspects of Ca*+ regulation in a way which opposes contraction (Cook et al., 1988). Opening of potassium channels is routinely monitored by measurement of “Rb+ efflux. This provides a convenient marker for potassium because of its relatively long half-life (18.6 days) compared with 42K+ (12.4 h) or 43K+ (22.4 h), but it may not be an appropriate substitute in all cases (Smith et al., 1986). In the majority of smooth muscle preparations cromakalim has been reported to stimulate *‘Rb+ efflux, indicating that it opens Rb+-permeable potassium channels (Al-

Correspondence to: K.A. Foster, SmithKline Beecham Pharmaceuticals, Great Burgh, Yew Tree Bottom Road, Epsom, Surrey KTlS

5XQ, UK. Tel. 44 131 364222,fax 44 737 364250.

len et al., 1986; Hamilton et al., 1986; Weir and Weston, 1986a,b; Coldwell and Howlett, 1987; Quast, 1987). The cromakalim-stimulated efflux of 86Rb+ from guinea pig trachealis is, however, difficult to detect (Allen et al., 19861, cromakalim-induced relaxations of rat uterus are associated with little or no “Rb+ efflux (Hollingsworth et al., 1987) and Brading has reported that cromakalim opens K+ channels that are impermeable to Rb+ ions in the guinea pig bladder detrusor muscle (Foster and Brading, 1987; Foster et al., 1989). These findings raise the possibility that cromakalim activates more than one type of potassium channel in smooth muscle and that these can be differentiated by their permeabilities to Rb+ ions. In the study reported herein the effect of replacing K+ in the Krebs solution by Rb+ has been investigated with regard -to tension and K+ and Rb+ fluxes in guinea pig trachealis smooth muscle in response to cromakalim. The mechanism of action of cromakalim involves K” channel opening, membrane hyperpolarization and consequent inhibition of voltage-operated Ca2+ channels (see Arch et at., 1988a). Comparison has been made, therefore, with the Ca*+ channel blocker verapamil. Cromakalim has also been compared with another smooth muscle relaxant, the p2adrenoceptor agonist salbutamol. &Adrenoceptor agonists in addition to increasing intracellular CAMP levels also cause hyperpolarization of the smooth mus-

144

cle plasma membrane and enhance *‘Rb+ efflux (Allen et al., 1985; Cook and Small, 1992).

2. Materials

and methods

2.1. Solutions The Krebs solution had the following composition (mM): NaCl (118), NaHCO, (25), KC1 (4.81, CaCI, (2.51, KH,PO, (1.21, MgSO, (1.181, glucose (11). The Rb-Krebs had the same composition except that KH,PO, and KC1 were replaced, at equivalent concentrations, by NaH,PO, and RbCl respectively. Krebs solutions were bubbled with 5% CO, in 0, to give pH 7.4. Exogenous RbCl and KC1 were added from 2 M aqueous stock solutions. Cromakalim was prepared as a 10 mM stock solution in 70% ethanol, and diluted with buffer as required. Salbutamol, +/verapamil, histamine HCl and quinidine were prepared as stock solutions in 10% DMSO or distilled water. Indomethacin was prepared as a stock solution in ethanol. Glibenclamide was freshly prepared each day in 0.1 N NaOH. Appropriate vehicle control experiments were performed. 2.2. Eflu..x studies Sections of guinea pig trachealis, from male Dunkin Hartley guinea pigs, 300-600 g, were prepared as described by Allen et al. (19861, and following a 30-min equilibration in Krebs solution at 37°C were loaded with either *‘Rb+ (74 MBq/l) or 42/43K3+ (37-74 MBq/l) or both at 37°C in 2.5 ml Krebs solution. Loading with 86Rb+ (alone or in combination with 42/43K+) was performed for 90 min, whilst loading with 42/43K+ alone was performed for 60 min. Efflux of isotope was followed by transferring tissues through a sequence of 17 tubes each containing 2.5 ml of either normal or Rb-Krebs solution at 37°C with a resident time in each tube of 3 min. When present, cromakalim (10 PM) was added to tubes 10-13. After the final efflux period, tissues were blotted dry, and, for experiments involving 86Rb+, were digested overnight in Optisolve. Radioactivity in both tissues and wash tubes was measured by liquid scintillation counting (experiments involving *‘Rb+) or by y-counting (42/43K+ single label experiments). For dual label experiments samples were stored for 10 days following counting to allow decay of 42/43K+, and recounted. Correction for 86Rb+ decay then allowed estimation of the relative *‘Rb+ and 42/43K+ content of the original count. Efflux was calculated as a rate coefficient (fractional loss of 42K+ or *‘Rb+ from the tissue over a l?min period), and was then expressed as a percentage of the basal efflux coefficient, where the basal efflux was the average

efflux coefficient occurring in the three efflux tubes immediately prior to drug or vehicle addition. 2.3. Uptake studies Segments of guinea pig trachealis, prepared as for the efflux studies, were incubated in normal Krebs solution at 37°C for 30 min, and then transferred to either normal or .Rb-Krebs solution containing 86Rb+ (74 MBq/l) with or without cromakalim (10 FM). After 10 min, uptake was ended by transferring the tissues to ice-cold Krebs solution for 5 min. Tissues were then blotted dry, weighed and, following digestion in Optisolve, their 86Rb+ content measured by liquid scintillation counting. 2.4. Tension studies Tracheal spiral strips (two per animal) were prepared and suspended isometrically under 2 g tension. In some experiments tone was allowed to develop spontaneously, tension being maintained at 2 g by frequent adjustment. In other experiments tone was induced using histamine at an EC,, concentration (5 PM), determined from preliminary concentration-effect experiments. Indomethacin (2.8 PM) was present in the histamine experiments to prevent endogenous production of prostaglandins. Concentration-response curves were constructed in a cumulative fashion. Inhibitory effects of the compounds against the maintained tension were calculated as a percentage of the maximal relaxation induced by isoprenaline (1 mM) or as a percentage of the contraction induced by histamine, and intrinsic activities were expressed relative to the isoprenaline (1 mM) effect. To test the effect of exogenous Rb+, tone was induced using histamine (5 PM) and when the tone had reached a maximum various concentrations of RbCl(1 mM, 3 mM or 5 mM) were added to the baths. The RbCl was in contact with the tissues for 20 min before cumulative dose-response curves to cromakalim were performed in the continued presence of RbCl. Control experiments were performed with equivalent concentrations of KCl. The relaxations were calculated as a percentage of the histamine-induced relaxation. For the potassium channel blocker experiments, tissues were challenged with quinidine (10 PM) or glibenclamide (1 FM) after a maximal histamine (5 PM)-induced contraction had developed. Alter a further 20 min cromakalim (10 PM) was added to the bath and the response studied for 20 min in the continued’ presence of the potassium channel blockers. A maximum relaxation was then induced using isoprenaline (1 mM). Only one potassium channel blocker was tested on a single tissue.

145

2.5. Statistics

IC,, values were calculated from individual concentration-effect experiments, and the results are expressed as arithmetic means f S.E.M. or as geometric means with 95% confidence intervals. Significance was assessed relative to time-matched control tissues using a two-tailed unpaired t-test. In dual label efflux studies significance between the efflux responses of the two isotopes was assessed by a paired t-test. A difference between means was assumed to be significant when P < 0.05.

% Basal Efflux A 200

2.6. Materials

Cromakalim was synthesized by SmithKline Beecham Pharmaceuticals. 86Rb+ was supplied by Amersham International and the mixed 42K+, 43.KK+isotope by the M.R.C. Cyclotron Unit, Hammersmith Hospital, London, UK. Optisolve and Optiphase liquid scintillant were from LKB. All other reagents were from B.D.H. or Sigma, and were A.R. grade or better.

B

3. Results 3.1. K + and Rb+ efj7u.x

Cromakalim (10 PM) caused a significant stimulation of both “Rb+ and 42/43Kf efflux from guinea pig trachealis (fig. 1). In single isotope experiments the stimulation of 42/43K+ efflux by cromakalim was significantly greater than that of 86Rb+ efflux (fig. lA), whilst in dual isotope experiments the stimulation of 42/43K+ efflux was reduced to that of the 86Rb+ efflux (fig. 1B). There was no significant difference between the basal efflux of 42/43K+ in single label (1.092 f O.O42%/min, mean + S.E.M., n = 59) and dual label (1.262 + O.l03%/min; mean f S.E.M., n = 24) experiments. The stimulation of “Rb+ efflux was similar in the single- and dual-labelled experiments (fig. 1). There was no difference in the basal efflux of 86Rb+ between single label (1.192 + O.l12%/min; mean f S.E.M., n = 27) and dual label (1.167 + O.O93%/min; mean f S.E.M., n = 27) experiments. The influence of rubidium on potassium efflux was studied further by performing single-isotope experiments in a Krebs solution in which K+ ions were replaced by Rb+ ions (Rb-Krebs solution). In Rb-Krebs solution, cromakalim caused no measurable stimulation of 42/43K+ efflux from guinea pig trachealis, which was in marked contrast to the significant stimulation observed in normal Krebs solution (fig. 2A). Surprisingly, cromakalim also failed to stimulate 86Rb+ efflux in RbKrebs solution (fig. 2B). Basal efflux rates of 42/43K+ (1.105 f O.O35%/min; mean f S.E.M., n = 59)

50

L

Cromakalim I

I

I

24 27 30

I

I

33 36

I

1

39 42

I

,

I

45 48 51

1

54

Time (min) Fig. 1. 42/43K+ (0, 0) and “Rb’ (A, A> efflux from guinea pig trachealis in response to cromakalim (10 PM) (closed symbols) or vehicle controls (open symbols). (A) Measured separately in single isotope experiments; (B) measured in the same tissue in dual isotope experiments. Points show means with S.E.M. indicated by vertical bars, n 2 6. * Significantly different from time-matched vehicle controls; + significantly different from corresponding *‘Rb+ efflw.

and 86Rb+ (1.218 f O.l27%/min, mean f S.E.M., n = 25) in Rb-Krebs solution were not significantly different from those in normal Krebs solution. 3.2. Rb + uptake

Basal uptake of “Rb+ by guinea pig trachealis was 235 f 24 cpm/mg per min (n = 12). This was significantly stimulated (P < 0.02) by 10 PM cromakalim to 317 + 19 cpm/mg per min (n = 12). The concentration of Rb+ in the Krebs solution in these experiments was between 4 and 10 PM depending upon the specific activity of the *‘Rb+ used. Basal and stimulated uptake were always measured in parallel using 86Rb+ of

146

the same specific activity. Basal uptake of “Rb+ by guinea pig trachealis in Rb-Krebs solution was not significantly different from that in normal Krebs solution (221 f 15 cpm/mg per min, n = 12). In contrast to the efflux experiments in Rb-Krebs solution, cromakalim (10 PM) again stimulated uptake significantly (286 it 19 cpm/mg per min, n = 121, and the extent of stimulation of uptake by cromakalim was not significantly different in Rb- or normal Krebs solution.

% Basal Efflux A 200 -

150 -

3.3. Tissue tension

B

50

1,

Cromakalim I I I I I I 24 27 30 33 36 39 42 Time (min)

I I 45 46

I I 51 54

Fig. 2. Cromakalim (10 PM)-stimulated (closed symbols) or basal (open symbols) 42/43K+ (A) or &Rb+ (B) efflux from guinea pig trachealis measured in single-isotope experiments in normal (0, O) or Rb (A, A) Krebs solution. Points show means with S.E.M. indicated by vertical bars for 12 measurements in (A) and 5 measurements in (B). * Significantly different from time-matched vehicle control experiment performed in the appropriate Krebs solution; + significantly different from the corresponding efflux in Rb-Krebs solution.

Following a l-h pre-equilibration of the tissues in the appropriate Krebs solution, cromakalim was less potent and less effective at relaxing spontaneous or histamine-induced tone in Rb-Krebs solution compared to normal Krebs solution (table 1). This is in contrast to the efflux experiments where effects of cromakalim were not detectable in Rb-Krebs solution. The relaxations in response to salbutamol were less affected by Rb-Krebs solution and verapamil caused more marked relaxations (table 1). It was apparent when performing the cromakalim experiments that the relaxations obtained in Rb-Krebs solution were transient, whereas those in normal Krebs solution were well maintained. Time course experiments (fig. 3A) performed using a single concentation of cromakalim showed that in normal Krebs solution the onset of action of cromakalim against spontaneous tone was at approximately 0.5 min, with a maximal, well maintained, relaxation being achieved within 6 min. In Rb-Krebs solution, the time of onset of cromakalim’s action and time to maximal effect were unaltered, but the relaxation was markedly reduced compared to that in normal Krebs solution and was transient, declining to control levels by 20 min. Salbutamol caused a maximal relaxation in normal Krebs solution within 4 min (fig. 3A), and this response was well maintained. Unlike cromakalim, the response to salbutamol in RbKrebs solution, although smaller, was well maintained. There was no alteration in salbutamol’s onset of action

TABLE 1 Relaxation of guinea pig tracheal spirals in Rb-Krebs solution compared with normal Krebs solution. ’ IC,, (PM), mean (95% confidence interval); intrinsic activity at 10 PM relative to isoprenaline 1 mM (mean f S.E.M.). Nq: not quantifiable. n = 4 in all cases. Dw Cromakalim Salbutamol Verapamil

Krebs solution

Spontaneous tone ’

Histamine (5 PM)-induced

Normal Krebs Rb-Krebs Normal Krebs Rb-Krebs Normal Krebs Rb-Krebs

0.6 (0.5-0.7); 0.90 f 0.03 2.2 (0.9-5.0); 0.84 f 0.07 0.02 (0.01-0.06); 0.99f 0.01 0.04 (0.02-0.07); 0.99f 0.01 Nq; 0.20 f 0.02 1.5 (0.6-4.0); 0.79* 0.05

3.0 (1.9-4.9); 0.77 f 0.03 6.2 (5.2-7.5); 0.23 f 0.10 0.05 (0.03-0.08); 0.99 kO.01 0.25 (0.16-0.45); 0.87 f 0.04 5.4 (3.1-9.7); 0.54 f 0.05 1.0 (0.8-1.2); 0.85 f 0.07

tone ’

147 %

Relaxation

to

isoprenaline

B

A

.’I. a. a. ..

.. . 20 -: . .. . 40 - :. .I. I. 60:

\i ‘+II .I .* *. .* .* . * .*

II I.

O-K

I)y.t’~ l

iT:

y

i

60-

A..

,. I. .

~ .

.

--

~*,@=’

. -.“*-a

T.*V--. loo-

I

7°F

0

5

I

I

1

IO 15 20 25 Time (min)

I

o

1

5

’

’

’

10 15 20 Time (min)

1

25

Fig. 3. Time course of cromakalim (10 PM, 01, salbutamol-(100 nM, n ) and verapamil (10 PM, A)-induced relaxations of (A) spontaneous tone and (B) histamine (5 FM&induced tone in guinea pig tracheal spirals measured in normal (dashed line) and Rb (solid line) Krebs solution. Each point is the mean with S.E.M. indicated by the vertical lines for z four measurements.

or the time taken to reach a maximum. The relaxation produced by cromakalim against histamine-induced tone in Rb-Krebs solution was also transient (fig. 3B), whilst the responses to both salbutamol and verapamil, as was found against spontaneous tone, were well maintained. The mechanism of the transient relaxation induced by cromakalim in Rb-Krebs solution was investigated using the K+ channel blockers quinidine and glibenclamide. In normal Krebs solution neither potassium channel blocker, at the concentrations used, caused additional contractions or relaxations in histamine pre-contracted tracheal spirals. Both quinidine (10 PM) and glibenclamide (1 PM) inhibited the cromakalim concentration-response curve compared to control tissues (cromakalim IC,,: control, 1.0 PM (0.6-1.7); quinidine, 4.0 PM (3.3-5.6); glibenclamide, 9.9 PM (3.0-32.8); mean (95% confidence intervals), n = 4). In Rb-Krebs solution, quinidine totally abolished and glibenclamide inhibited the cromakalim (10 PM)-induced. transient relaxation of the histamine-induced tone (fig. 4). Quinidine was without effect on salbutamol (0.1 PM)-induced relaxation of histamine tone in Rb-Krebs solution (data not shown). 3.4. Effects of additional Rb ’ on eflux and tension

Cromakalim-stimulated efflux of 42/43K+ in Krebs solution containing a normal concentration of potassium was completely blocked by the additional presence of 5 mM RbCl throughout the efflux period, and

only extremely small stimulations were observed in the presence of 3 mM and 1 mM RbCI. Cromakalim did enhance 42/43K+ efflux markedly in the presence of 0.3 mM RbCI, although this was reduced compared to that observed in the absence of RbCl (table 2). Control experiments established that addition of 0.3-5 mM KC1 was without effect on cromakalim-stimulated efflux. Addition of RbCl (1 mM, 3 mM and 5 mM) to normal Krebs solution induced concentration-dependent contractions of tracheal spirals additional to that produced by 5 PM histamine (3 + ll%, 12 f 12% and 19 + 13% respectively). These were not significantly different from those produced by adding equivalent concentrations of KCl. With all concentrations of RbCl tested there was a concentration-dependent, rightward shift of the concentration response curve to cromakalim performed 20 min after the addition of the %

Relaxation

to

isoprenaline

-10

0

10

20

30

40

50

60

70

I 0

I 5

I 10

Time

I 15

I 20

(min)

Fig. 4. Effect of quinidine (10 PM) or glibenclamide (1 PM) on cromakalim (10 PM&induced relaxations of guinea pig tracheal spirals in Rb-Krebs solution. 0, 10 PM cromakalim alone (quinidine control, 0.1% (v/v) DMSO) (n = 4); n , 10 pM cromakalim plus 10 @M quinidine (n = 3); +, 10 @M cromakalim alone (glibenclamide control, 0.1 mM NaOH) (n = 4); A, 10 FM cromakalim plus 1 PM glibenclamide. Points show means with S.E.M. indicated by vertical bars.

148 TABLE

2

TABLE

Effect of various concentrations of RbCl on the peak 42/43K+ efflux response to cromakalim (10 FM). Where added, RbCl was present in the wash tubes throughout the efflux experiment. Values are mean f S.E.M. (number of measurements). ’ Significantly different from time-matched vehicle control measurements in the presence of the same concentration of RbCI. Concentration of RbCl fmM) 0 0.3 1 3 5

Peak efflux response to 10 PM cromakalim (% basal) 163.4* 13.0 (8) ’ 138.8* 16.8 (3) a 118.3 f 4.6 (6) ’ 114.5 f 9.2 (6) ’ 100.5 f 4.0 (6)

3

Effect of the time of addition of Rb-Krebs solution on the peak 42’43K+ efflux response to cromakalim (10 PM). Efflux was performed as described in Methods into tubes: containing normal Krebs solution throughout (control), containing Rb-Krebs solution throughout (Rb-Krebs (33-min pre-incubation)) or containing normal Krebs solution in the tubes prior to the addition of cromakalim and Rb-Krebs solution in the tubes containing cromakalim (Rb-Krebs (no pre-incubation)). Values are mean f S.E.M. (number of measurements). D Significantly different from time-matched vehicle controls performed under equivalent conditions. Condition

RbCl (fig. 5). As was found for the efflux experiments, addition of 5 mM RbCl to normal Krebs solution was equieffective with Rb-Krebs solution at inhibiting the activity of cromakalim. Addition of exogenous KC1 at equivalent concentrations to RbCl did not influence subsequent cromakalim-induced relaxations (data not shown). The effect of Rb” on cromakalim-stimulated 42/43K+ efflux was not influenced by the time of pre-incubation with Rb+ (table 3). Following a 6.5-h pre-incubation of the trachealis in normal Krebs solution, cromakalim (10 PM) caused a small but significant stimulation of 42/43K+ efflux, which was completely inhibited when the pre-incubation was in Rb-Krebs solution (table 4). Cromakalim-induced relaxation of histamine (5 PM)-contracted guinea pig tracheal spirals was reduced by prolonged (300 min) incubation in

Relaxation (% of histamine contraction) 0

50

Control Rb-Krebs (33-min pre-incubation) Rb-Krebs (no pre-incubation)

Peak efflux response to 10 PM cromakalim (o/c basal) 167.7* 12.1 (4) ’ 108.4& 8.3 (7) 98.2f 7.6 (7)

either normal or Rb-Krebs solution (fig. 6). The ability of Rb-Krebs solution to reduce cromakalim-induced relaxation relative to appropriate time-matched control experiments in normal Krebs solution was not different at 30 or 300 min, however.

4. Discussion This study was undertaken to investigate the Rb+ permeability of the potassium channels opened in guinea pig trachealis by cromakalim, and to compare the effects of extracellular Rb+ on the relaxant responses induced by cromakalim, verapamil and salbutamol. Although Allen et al. (1986) have previously reported a enhancement of 86Rb efflux from guinea pig trachealis by cromakalim, the effect was small compared to that reported for many other smooth muscles. A small, but significant, stimulation of 86Rb+ efflux from guinea pig trachealis was also found in the present work, whereas there was a marked stimulation of 42/43K+ efflux from this tissue. When both isotopes were present in the same tissue, in a dual isotope experiment, however, the extent of the stimulation of 42/43K+ efflux was reduced to that of the 86Rb+. This

.

TABLE 4

100

I

10-E

lo-’

10-6

10-s

10-4

Cromakalim [M] Fig. 5. Effect of various concentrations of RbCl on cromakalim dose-response cmves against histamine (5 PMhinduced tone in guinea pig tracheal spirals. l , control; n , 1 mM RbCI; A 3 mM RbCI; +, 5 mM RbCI. Points show means with S.E.M. indicated by vertical bars for four determinations.

Cromakalim (10 PM)-stimulated 42/43K+ efflux from guinea pig trachealis following a 6.5-h incubation in either normal or Rb-Krebs solution. The values quoted are for the efflux occurring in the second of four sequential efflux tubes containing cromakalim or vehicle as indicated, and are the mean& S.E.M. (number of measurements). ’ Significantly different from corresponding vehicle control. Krebs solution Normal Krebs solution Rb-Krebs solution

42/43K+ efflux response (%I basal Vehicle control 10 PM cromakalim 102.4k5.1 (12) 126.2+ 9.1 (12) ’ 105.3 rt5.6 (12) 109.4* 10.7 (12)

149

Relaxation (% of histamine contraction)

I

I

I

I

1U8

1o-5

IO-6

lo-’

I

1o-4

Cromakalim [Ml Fig. 6. Effect of the incubation time in Rb- or normal Krebs solution on cromakalim-induced relaxation of histamine (5 KM)-induced tone in guinea pig tracheal spirals. Closed symbols: normal Krebs solution; open symbols: Rb-Krebs solution. Solid lines: 30-min pre-incubation; dashed lines: 300-min pre-incubation. Points show means with S.E.M. indicated by vertical bars for four determinations.

+ -+ 0 l ------* o------o

0

ICsa values PM (mean (range))

Intrinsic activity at 10 PM relative to isoprenaline 1 mM (mean f S.E.M.)

1.90 (1.80-2.10)

0.98 f 0.03

5.30 (3.80-7.62) 5.50 (5.13-5.72) 7.13 (5.12-10.53)

0.64 f 0.04 0.68 f 0.02 0.19*0.05

suggested that Rb+ was impairing the efflux of 42/43K+ through the channel opened by cromakalim. In agreement with these results, Edwards and Weston (1989) have reported that 1 PM, although not 10 PM, cromakalim-stimulated 42K” efflux in rat portal vein is reduced in dual isotope experiments. Foster and Brading (1987) and Foster et al. (1989) have reported that the activity of cromakalim in guinea pig detrusor muscle can be blocked by using a Krebs solution in which K+ is replaced by Rb+. Such a Rb-Krebs solution was found in the present study to completely prevent the efflux of 42/43K+ from guinea pig trachealis in response to cromakalim (10 FM). This suggested that even the Rb+-permeable component of the efflux response, as determined by “Rb+ efflux, was being prevented by incubation in Rb-Krebs solution. This conclusion was confirmed by showing that cromakalim failed to stimulate 86Rb+ efflw in Rb-

Krebs solution. The paradoxical ability of extracellular rubidium to block “Rb+ efflux from the cytosol could not be explained in terms of the potassium channel being permeable to intracellular rubidium but blocked by extracellular rubidium, because cromakalim was capable of stimulating 86Rb+ uptake. Thus cromakalim opens a bidirectional channel permeable to “Rb+ in both directions. The cromakalim-stimulated Rb+ uptake, by contrast with efflux, was not inhibited by extracellular Rb+ in Rb-Krebs solution, suggesting that it is occurring largely through a channel which is not blocked by Rb+. The blockade of efflux is probably related to the concentration of Rb+ involved: in the “Rb+ efflux studies, at the specific activity used, and assuming equilibrium between the intra- and extracellular pools, the intracellular concentration was likely to have been of the order of 0.1-0.2 mM, whilst in RbKrebs solution the extracellular concentration was 4.8 mM. When RbCl was added to normal Krebs solution a concentration of 0.3 mM was found to reduce, but not abolish, cromakalim-stimulated 42/43K+ efflux, a result in keeping with the reduction of cromakalimstimulated 42/43K+ efflux by rubidium in the dual isotope experiments. Thus in guinea pig trachea, cromakalim opens a potassium channel that is poorly permeable to Rb+ at low concentrations of the ion and becomes blocked as the concentration is increased, and also opens a channel, which is not blocked by Rb’, and appears to be the same channel by which uptake 06 curs. The blockade of efflux in Rb-Krebs solution was associated with an impaired relaxation response to cromakalim against both spontaneous and histamineinduced tone. Such an effect was not observed with either salbutamol- or verapamil-induced relaxations. The response to verapamil was, if anything, enhanced in Rb-Krebs solution, suggesting an increased dependency on extracellular Ca*” ‘entry through voltage-operated calcium channels relative to other calcium sources. This might be expected in Rb-Krebs solution because a proportion of K+ channels would be blocked by Rb+, leading to a depolarization of the smooth muscle cell plasma membrane and enhanced opening of voltage-operated Ca*+ channels. An increased “Rb+ efflux, indicating membrane depolarization, in the presence of Rb+ has been reported in rat aorta (Bray and Quast, 19911, although no effect of Rb-Krebs solution on basal efflux was found in the present study. An influx of extracellular Ca*+ is supported by the finding that addition of an extra 5 mM KC1 to the Krebs solution caused an equivalent contraction to that caused by 5 mM RbCI, .as it is known that KCI-induced contractions in guinea pig trachea are due to a membrane depolarization resulting in an influx of extracellular Ca*+ (Foster et al., 1983). An enhanced Ca*+ influx could be suggested to explain the inhibition of

150

the cromakalim-induced effects by Rb+, as the enhanced Ca2” entry would oppose cromakalim’s action on voltage-operated Ca 2f channels. This cannot be the mechanism underlying the inhibition, however, as the 5 mM KC1 would then also be expected to inhibit cromakalim’s relaxant effects, and there was no evidence of this. Hence Rb+ is doing more than simply depolarizing the tissue in the way K+ does. In guinea pig detrusor muscle it has been observed that the extent of the reduction of the cromakalim response increased with the incubation time in RbKrebs solution, becoming maximal at about G h (Foster and Brading, 1987; Foster et al., 1989) This was suggested to be a consequence of the replacement of intracellular K+ with Rb+. The effects of Rb+ on the responses of guinea pig trachealis to cromakalim reported here were, by contrast, mostly observed after approximately 30-min pre-incubation, and in the case of 42/43K+ efflux could be observed within 3 min (table 3). Prolonged incubation caused no enhancement of the inhibitory action of Rb-Krebs solution toward cromakalim, either in terms of relaxation of histamine-induced tone (see fig. 6) or 42/43K+ efflux (see table 3). Thus in guinea pig trachealis the inhibitory effects of Rb-Krebs solution toward cromakalim do not appear to involve replacement of intracellular K” with Rb+. Indeed the ability to detect blockade of efflux in the very first efflux tube, when contact with Rb had only been for a maximum of 3 min, suggests that the Rb effect is occurring extracellularly. The effect of RbKrebs solution on cromakalim’s activity could be reproduced by adding 5 mM RbCl to normal Krebs solution, suggesting that it was the presence of Rb+, rather than the absence of K+, which was producing the effect. Addition of RbCl (0.3-5 mM) to normal Krebs solution produced a concentration-dependent impairment of the effects of cromakalim on both 42/43K” efflux and tracheal tone. The response to Rb+ in the tension experiments indicated a concentration-dependent blockade by Rb+. Thus, unlike detrusor muscle, the effects of Rb+ in guinea pig trachealis appear to be due to blockade of the channel opened by cromakalim rather than the replacement of intracellular K+. Although tracheal relaxation by cromakalim was markedly inhibited by Rb+ blockade, a transient, repeatable response still occurred. The time of onset of the relaxant response and time to maximal effect were unaltered between Rb- and normal Krebs solution. This would appear to rule out the suggestion that Rb+ is simply passing through the channel(s) more slowly than K+ as one might then expect a delay in these parameters. Two explanations for the transient relaxation, with no detectable 42/43K efflux, observed in the presence of Rb+ have been considered. Firstly, cromakalim has a second mechanism of action independent of opening plasmalemmal K+ channels. However,

the blockade of the transient relaxations by both quinidine and glibenclamide would suggest an involvement of K+ channel opening, as it is unlikely that two such diverse potassium channel blockers are both affecting a second, non-potassium channel-dependent effect of cromakalim. The second, more propable, explanation is that cromakalim opens two types of K+ channel as characterized by their Rb+ permeability. One is blocked by Rb+, whilst the other is not blocked by Rb+ and opens only transiently. If the transient relaxation observed in the presence of Rb+ is due to the opening of a K+ channel, it might be expected that an efflux response should be observed in the presence of Rb+. The efflux measurement is, however, a relatively insensitive one and a small, transient response may well be missed. The decrease in the level of stimulated efflux that is observed in the continued presence of cromakalim (fig. 1 and also: Coldwell and Howlett, 1987; Quast, 1987; Quast and Baumlin, 1988) might represent the closing of such a transient channel. A transiently opened potassium channel which was not affected by extracellular Rb+ would also explain the capacity to stimulate 86Rb uptake by cromakalim even in Rb-Krebs solution, a channel freely permeable to Rb being the more likely candidate for 86Rb uptake to occur through. Alternatively, the transient relaxation may be due to the opening of an intracellular potassium channel of the type described by Fink and Stephenson (1987) in skeletal muscle, which would inhibit Ca2+ release from the sarcoplasmic reticulum. Such an intracellular potassium channel would not produce an efflux of 42/43K+ or “Rb+ from sections of whole tissue, but could be sensitive to the effects of potassium channel blockers such as quinidine (Shah and Pant, 1988). The possibility that cromakalim has an effect on potassium channels associated with an intracellular site such as the sarcoplasmic reticulum has been suggested previously (Bray et al., 1989; Chopra et al., 1990). An effect of cromakalim on the refilling of intracellular Ca2+ stores has also been shown (Chopra et al., 1990). Intracellular sites of action for cromakalim do not, however, explain the ability to stimulate uptake of “Rb+ in Rb-Krebs solution. The suggestion that cromakalim is opening more than one potassium channel in a given smooth muscle is not a novel one, and has been proposed by a number of workers (Hamilton et al., 1986; Nakao et al., 1988). Differences in the permeability to 86Rb+ and 42K+ in response to cromakalim have also been reported in a number of vascular tissues (Quast and Baumlin, 1988; Bray and Weston, 1989; Bray and Quast, 19911, and it was proposed that this might represent the opening of two cromakalim-sensitive potassium channels differing in their Rb+ permeability. Fuller characterization and identification of the potassium channels opened in guinea pig trachealis in

151

response to cromakalim will require electrophysiological studies. Measurement of 86Rb efflux does not appear to offer an ideal method for investigating potassium channel opening by compounds such as cromakalim. In conjunction with potassium efflux studies, however, it is a valuable tool for the characterization of these channels.

Acknowledgements The authors are grateful to Mr J.E.J. Morris for his technical assistance in performing some of the experiments described, and also to Miss A.N. Russell for typing this manuscript.

References Allen, S.L., D.J. Beech, R.W. Foster, G.P. Morgan and R.C. Small, 1985, Electrophysiological and other aspects of the relaxant action of isoprenaline in guinea pig isolated trachealis, Br. J. Pharmacol. 86, 843. Allen, S.L., J.P. Boyle, J. Cortijo, R.W. Foster, G.P. Morgan and R.C. Small, 1986, Electrical and mechanical effects of BRL34915 in guinea pig isolated trachealis, Br. J. Pharmacol. 89, 395. Arch, J.R.S., D.R. Buckle, J. Bumstead, G.D. Clarke, J.F. Taylor and S.G. Taylor, 19SSa, Evaluation of the potassium channel activator cromakalim (BRL34915) as a bronchodilator in the guinea pig: comparison with nifedipine, Br. J. Pharmacol. 95, 763. Arch, J.R.S., D.R. Buckle, J. Bumstead and J.F. Taylor, 19SSb, Comparison of the effects of cromakalim (BRL34915) and pinacidil in guinea pig models of bronchoconstriction, Br. J. Pharmacol. Proc. Suppl. 95, 794P. Bowring, N.E., D.R. Buckle, G.D. Clarke, J.F. Taylor and J.R.S. Arch, 1991, Evaluation of the potassium channel activator BRL 38227 as a inhaled bronchodilator in the guinea pig: contrast with nifedipine and salbutamol, Pul. Pharmacol. 4, 99. Bray, K.M. and U. Quast, 1991, Inhibition of cromakalim-induced effects in rat isolated aorta by Ba2+ and Rb+, Fundam. Clin. Pharmacol. 5, 407. Bray, K.M. and A.H. Weston, 1989, Differential concentration-dependent effects of K channel openers on 42K and S6Rb efflux in rabbit isolated aorta, Br. J. Pharmacol. Proc. Suppl. 98, SS4P. Bray, K.M., S. Duty and A.H. Weston, 1989, Analysis of the spasmogenie effect of cromakalim in rabbit isolated aorta, J. Physiol. (London) 417, 67P. Buckingham, R.E., J.C. Clapham, T.C. Hamilton, S.D. Longman, J. Norton and R.H. Poyser, 19S6, BRL34915, a novel anti-hypertensive agent; comparison of effects on blood pressure and other haemodynamic parameters with those of nifedipine in animal models, J. Cardiovasc. Pharmacol. 8, 798. Chopra, L.C., C.H.C. Twort and J.P.T. Ward, 1990, Effects of BRL 38227 on calcium uptake by intracellular stores in cultured rabbit airway smooth muscle cells, Br. J. Pharmacol. Proc. Suppl. 100, 36SP.

Coldwell, M.C. and D.R. Howlett, 1987, Specificity of action of the novel anti-hypertensive agent, BRL34915, as a potassium channel activator. Comparison with nicorandil, Biochem. Pharmacol., 36, 3663. Cook, N.S., S.W. Weir and M. Danzeisen, 1988, Anti-vasoconstrictor effects of the K+-channel opener cromakalim on the rabbit aorta - comparison with the calcium antagonist isradipine, Br. J. Pharmacol. 95, 741. Cook, S.J. and R.C. Small, 1992, Role of Kc-channel opening in salmeterol-induced relaxation of trachealis muscle, Br. J. Pharmacol. Proc. Suppl. 106, 12P. Edwards, G. and A.H. Weston, 1989, Effects of cromakalim on potassium and rubidium efflux rate following dual isotope labelling, Br. J. Pharmacol. Proc. Suppl. 98, 926P. Fink, R.H.A. and D.G. Stephenson, 1987, Ca2+-movements in muscle modulated by the state of Kf-channels in the sarcoplasmic reticulum membranes, Pflilgers Arch. 409, 374. Foster, CD. and A.F. Brading, 1987, The effect of potassium channel antagonists on the BRL34915 activated potassium channel in guinea pig bladder, Br. J. Pharmacol. Proc. Suppl. 92, 751P. Foster, CD., K. Fujii, J. Kingdon and A.F. Brading, 1989, The effect of cromakalim on the smooth muscle of the guinea pig urinary bladder, Br. J. Pharmacol. 97, 281. Foster, R.W., R.C. Small and A.H. Weston, 1983, The spasmogenic action of potassium chloride in guinea pig trachealis, Br. J. Pharmacol. SO, 553. Hamilton, T.C., S.W. Weir and A.H. Weston, 1986, Comparison of the effects of BRL34915 and verapamil on electrical and mechanical activity in rat portal vein, Br. J. Pharmacol., 88, 103. Hollingsworth, M., T. Amedee, D. Edwards, J. Mironneau, J.P. Savineau, R.C. Small and A.H. Weston, 1987, The relaxant action of BRL 34915 in rat uterus, Br. J. Pharmacol. 91, 803. Nakao, K., K. Okabe, H. Kitamura and A.H. Weston, 1988, Characteristics of cromakalim-induced relaxations in the smooth muscle cells of guinea pig mesenteric artery and vein, Br. J. Pharmacol. 95, 795. Quast, U., 1987, Effect of the K+ efflux stimulating vasodilator BRL34915 on *‘Rb+ efflux and spontaneous activity in the guinea pig portal vein, Br. J. Pharmacol. 91, 569. Quast, U. and Y. Baumlin, 1988, Comparison of the efflwes of 42Kc and *‘Rb+ elicited by cromakalim (BRL34915) in tonic and phasic vascular tissue, Naunyn-Schmiedeb. Arch. Pharm’acol. 338, 319. Shah, J. and H.C. Pant, 1988, Potassium-channel blockers inhibit inositol trisphosphate-induced calcium release in the microsomal fractions isolated from the rat brain, Biochem. J. 250, 617. Smith, J.M., A.A. Sanchez and A.W. Jones, 1986, Comparison of rubidium-S6 and potassium-42 fluxes in rat aorta, Blood Vessels 23, 297. Weir, S.W. and A.H. Weston, 1986a, Effect of apamin on responses to BRL34915, nicorandil and other relaxants in the guinea pig taenia caeci, Br. J. Pharmacol. 88, 113. Weir, S.W. and A.H. Weston, 1986b, The effects of BRL34915 and nicorandil on electrical and mechanical activity and on *‘Rb+ efflux in rat blood vessels, Br. J. Pharmacol., 88, 121. Williams, A.J., T.H. Lee, G.M. Cochrane, A. Hopkirk, T. Vyse, F. Chiew, E. Lavender, D.H. Richards, S. Owen, P. Stone, M. Church and A.A. Woodcock, 1990, A potassium channel activator (cromakalim) attenuates nocturnal asthma, Lancet 336, 334.