Acta Ph?jsiol Scand 1992, 145, 423-428

Effects of cromakalim ( B R L 3491 5) on resting and evoked activity in frog semicircular canals G. ZUCCA, L. BOTTA, A. BARBIERI", E. G R A N A " and P. V A L L I T

* Institute of General Physiology

and Institute of Pharmacology, University of Pavia ;

j-Department of Animal Biology, University of Torino, Italy ZUCCA,G., BOTTA, L., BARBIERI, A., GRANA, E. & VALLI,P. 1992. Effects of cromakalim (BRL 34915) on resting and evoked activity in frog semicircular canals. Actu Physiol Scand 145, 423428. Received 29 July 1991, accepted 16 February 1992. ISSN 00016772. Institute of General Physiology and Institute of Pharmacology, University of Pavia and Department of Animal Biology, University of Torino, Italy. The effects of endolymphatic administration of cromakalim and glibenclamide were tested in isolated semicircular canals of the frog. The actions of the drugs were evaluated by recording : (1) the transepithelial potential between the endolymphatic and the perilymphatic sides of the crista ampullaris; (2) the slow nerve potential from the ampullar nerve; (3) the action potential discharge in afferent ampullar nerve fibres; and (4) the perilympathatic potassium concentration in the fluid bathing the outer surface of the crista ampullaris. The above mentioned parameters were recorded both at rest and during mechanical stimulation of the sensory organ. The results demonstrated that the endolymphatic administration of cromakalim ( W 4M) produced an increase in both ampullar receptor resting activity and in perilymphatic K' resting levels. By contrast all the parameters related to the mechanically evoked responses were practically unaffected. Glibenclamide (lo-' M) proved able to cancel or to prevent cromakalim effects. These data suggest that the membrane of the hair cells is endowed with K+ channels regulated by internal ATP whose activation is mainly involved in the processes sustaining ampullar receptors' resting firing rate.

Key words ; cromakalim, glibenclamide, hair cells, K+-channels, frog, semicircular canals.

T h e apical pole of hair cells in the sensory organs of the acoustico-lateralis system is surrounded by a fluid that has a high potassium concentration, whereas the basal pole contacts normal extracellular fluid (Sterkers et al. 1988). This concentration gradient of K+ across the sensory cells gives rise to a flow of K' ions throughout hair cell bodies, the receptor current, which sustains the whole conversion process in inner ear end-organs. Shearing displacements of the hair cell cilia produces changes in the impedance of their hair-bearing membrane. This change modulates the K+-current flowing across Correspondence: Dr G. Zucca, Institute of General Physiology, Via Forlanini, 6, I27100 Pavia, Italy.

the sensory cells which in turn controls transmitter release and the onset of EPSPs in the afferent nerve terminals which contact the hair cells (Rossi et al. 1977). These and other observations (Zucca et al. 1982, Valli et al. 1984, 1990, Hudspeth & Lewis 1988, Sterkers et al. 1988, Ashmore 1991) indicate that the hairbearing membrane of labyrinthine sensory cells is endowed with Kf-channels, but very few data are available on the characteristics of these channels. T h e present study was devised in order to gain some insight into the nature of the K+channels which are present in hair cell apical membrane and are involved in the conversion process taking place in vestibular organs.

423

Drugs und niunipulutions o f t h e endolymphatic @id. The drugs used were: Cromakalim (BRL 34915, Beecham, Retchworth, UK) and Glibenclamide (Sigma Chemical Co, St Louis, MO, USA). In different experiments the endolymphatic fluid \\as replaced by one of the following media: (1) cromakalim solutions: 0.5, 5 or Sop1 of a stock solution of cromakalim ( M) dissolved in 70");) v/v ethanol: double distilled water was added to the endolymphatic fluid (5 ml) in order to reach the final A 1 AT I.:R I .A 1, S -4%KI \IE T H 0 D S concentrations of lo-", lo-;, 10 ' 51 respectively-; ( 2 ) Experiments nere carried out on \-ertical posterior ethanol solutions: the effects of 0 . 5 , 5, S O pl of 70:;) semicircular canals isolated from frogs (Rona esrrrleniu ethanol solution, i.e. of the ethanol present in the I-.) prebiously anaesthetized by immersion in 0.1 different cromakalim solutions, were also tested; (3) >lS--222 solution. Isolated preparations were placed Glibenclamide solutions: 0.5, 5 or 50 pl of' a stock in a two-compartment chamber (Fig. 1) which made solution of glibenclamide (IOV' M), dissolved in it possible to maintain the endol!-mphatic and the dimethj-lsulphoxide (DMSO), was added to the peril!mphatic canal sides in contact with two ionic endol!mphatic fluid to reach final concentrations of and electricall! separated fluids. The canal's peri10P, 10-' or 10F M; and (4) DMSO solutions: as lymphatic nall and the ampullar nerve w-ere in fact with the ethanol solutions, control experiments were bathed in the fluid of one compartment (peril! mphatic performed to test whether DMSO, at the doses compartment) while the canal interior communicated emplo!ed (0.01, 0.1, 1""), might influence receptor onl?. with the other (endolymphatic compartment). activit!-. The chambers (capacity 5 ml each) were filled l\iirh In all the different experimental conditions, the artificial endolymph (XaCI 19.j m\r; KCI 100 m11; peril!-mph was aha!-s normal Ringer's solution. NaHCO, 1.2 m~rI;NaH,Po,O.li m\r;CaCI, 1.8 msl; .2lerhunizul stimulation of the sensory organ. Hair cell glucose 5.5 mbr; pH 7.3) and perilymph (NaCI stimulation was performed by producing sinusoidal 117 n n i ; KCI 2.5 mxl; NaHCO, 1.2 mxi; NaH,Po, fluid flow (0.05 Hz) inside the canal by- means of a 0.17 mM; CaCI, 1.8 mM; glucose 5.5 mxi; p H 7.3), microsyringe whose plunger (diameter 0.5 mm; disi.e. nith solution of ionic composition similar to those placements + 5 pm) was operated by a servothat hair cells face ' I n i i i o ' (Valli rt ul. 1990). controlled stepper motor (Valli & Zucca 1076). Rrrordrng of runal urtiuty. Slow nerve potentials (Ndc), due to electrotonic spreading of EPSPs in afferent fibres (Rossi et al. 1977), and spike discharge Ad c Ndc Nfr P 7 V (Nfr) were recorded from the whole nerve by means of _1? r-a fluid electrode. Spike frequency was measured using F/v a n indon- discriminator and a frequency-to-voltage con\-erter. .%mpullar potentials (Adc), which may be considered an index of the hair cell receptor current (Valli & Zucca 1976, Valli et a / . 1990), were recorded by means of Ag/AgCI electrodes placed in the endolymphatic and perilymphatic compartments respectively. In nature, ampullar potentials become negative during excitatory cupula deflections and positive during inhibitory ones. For ease of comparison between tracings, Adcs are reproduced with inverted polarities. Canal actiiity was evaluated both at rest, the values (R-rldc, R-Ndc, R-Nfr) being determined in the 30 s preceding stimulation, and during mechanical stimub c lation of the sensor). organ (E-Adc, E-Ndc, E-Nfr). Fig. 1. Schematic representation of the experimenral .+leasurrment of K* conrentrution in the perilymphatic set-up. Adc, [IC-],Ndc-Nfr, amplifiers and associated electrodes used to record ampullar potential (.ldc), K- jhrid. K- concentrations (Fig. l), evaluated only in the activit! ([K-1) and nerve potentials (Ndc, slow nerve perilymphatic fluid, were measured by means of K"potentials ; Nfr, nerve firing rate) respectively. F/Y, sensitive electrodes (Valli e t al. 1990) positioned close to the external surface of the ampulla. the frequent!-to-voltage converter ; SP, the stimuAs already reported (Valli et al. 1990), perilymphatic lating pipette used to produce sinusoidal fluid flow K--fluctuations, both at rest and during mechanical inside the canal.

'To this end, the effects of cromakalim, a new opener of the I(--channels regulated by internal .ATP (I(,T,,), and glibenclamide, a blocker of E;,,.,,, often used as a cromakalim antagonist (Cook 1988, Quast & Cook 1989, Cook & Quast 1990, Richer rt al. 1990) were tested in isolated semicircular canals of the frog.

'Iu

, D , p,

Cromakalim and frog semicircular canals

a Adc

Ndc Nfr

@+I

b

425

C

-

Stir r

M glibenclamide (c) on canal bioelectrical M cromakalim (b) and Fig. 2. The effects of responses (Adc, Ndc, Nfr) and perilymphatic K+ concentrations ([KC]). (a) Controls (b), (c) 10 min. after the administration of the drugs. Stim, stimulus. Adc, Ndc, Nfr, [K+], as in Fig. 1.

stimulations, were evident only if the K+-sensitive electrode was positioned close (30-50pm) to the region where the ampullar nerve enters the ampulla and decreased rapidly when the K+-sensitive electrode was pulled back. (At about 150-200pm from the ampulla no changes were detected.) Each K+-sensitive electrode was calibrated before and at the end of each experiment by measuring its voltage output in physiological solutions with known K' contents (0.1, 1, 10, 100 mM). If calibration values differed by more than 10% the experiment was discarded. A conventional microelectrode, filled with perilymphatic solution and positioned close to the K+sensitive electrode, was used as reference electrode (Fig. 1). To obtain signals related only to K' activity, the potentials from the reference electrode and from the K+-sensitive electrode were differentially amplitied, to eliminate common-mode field potentials. Perilymphatic K+ concentrations, as for canal bioelectrical activity, were measured both at rest (R-[K']) and during mechanical stimulation of the sensory organ [E-[K']). Recordings und calculations. Signals were sampled digitally using an analogue-digital converter coupled to a personal computer, analysed and plotted on paper.

RESULTS Preliminary experiments were devoted to determining the endolymphatic concentration of cromakalim which produced significant changes in ampullar receptor activity. T h e results demonM cromakalim was completely strated that devoid of effects (n = 3); lo-' M cromakalim (n = 5 ) produced small, but not significant,

changes (data not shown) whereas distinct changes were observed only when the endolymphatic concentration of the drug was as high as lo-* M ( n = 6). Therefore, in the present M study, only the results obtained with cromakalim will be presented in detail. A typical example of cromakalim action on both bioelectrical canal activity and perilymphatic K+-levels is shown in Figure 2. It will be noted that the drug produces different effects on resting and on evoked activity. I n fact, as shown in Figure 2(b) (beginning of the tracings), all the parameters related to the resting conditions (resting bioelectrical activity, resting perilymphatic K+-levels) are clearly affected by the drug. I n effect: (1) the resting transepithelial potential decreased and settled, in about 5 min, at a new level lower than resting level by about 0.5 m V (Adc tracings are reproduced with inverted polarity); (2) a distinct spike discharge, associated with nerve depolarization, took place in afferent fibres; and (3) a clear increase in K+ concentration occurred in the perilymphatic medium, i.e. in the fluid bathing the outer surface of the crista ampullaris sensory organ. I n contrast, the mechanically evoked responses, as demonstrated by the constancy of the peak-to-peak amplitude of the evoked responses, were nearly unchanged. Cromakalim effects remained stable for long periods (more than 30 min) and were almost completely irreversible even after protracted washing of the preparation with normal endolymphatic solutions. Glibenclamide ( M)

R-Adc

R-Nfr

R-Ndc

R-[K']

120 80

8o

-0

0

10 20 30

cr

glib

Time (min)

E-Adc

glib

100 80

i

oc

0

r

'

.

i 0 ' i O ' j 0 Time

-

cram (min) glib

E-Ndc

glib

2.3

of,. 0

,

..

,

, ,

Time crom (min) glib

10 20 30

-

E-Nfr

glib

of,

0

,

.,.

,

, ,

10 20 30 Time

c

z (min) glib

E-[K']

. , glib

Fig. 3. Time course of the etfects of cromabalim (10 ' \I) and glibenclamide (10 M) on both resting (K--\dc, R-Ndc, R-Sfr, K-LIi-1) and peak-to-peak amplitude of the evoked responses (E.ldc. E-Ndc, E-Sfr, E-[ K-J).Solid bars indicate the drug administration periods. Values, means from sis different experiments, are expressed as percentages of controls. The results are given as means & SE. Statisticall!- significant differences were evaluated with Student's t-test. Vertical linea show SE significant differences from the basal values are shown: P < 0.0.5 (*); P < 0.001

proi.ed able to cancel cromakalirn effects ( I O P , lo-" \I glibenclamide were ineft'ectii-e). I n fact, as s h o a n in Figure 2(c), it will be observed that about i--I0 min after endolymphatic administration of 10 11 glibenclamide the effects of cromakalim on both canal activity and perilymphatic K'-le\-els were completelJ-suppressed. T h e results of these experiments are summarized in the graphs in Figure 3 where the effects of cromakalim and glibenclamide on both resting (R-.ldc, R-Xdc, R-Nfr, R-[K-1) and e\-oked responses (E-;\dc, E-Ndc, E-Nfr, E[K-1) arc quantitatiwly reported as a function of treatment time. It nil1 be ohscri-ed from the graphs that onl!the parameters related to resting activity were significantl! modified by cromakalim. At the ma.iimum of its effect, reached in 3-9 min, the drug y ~ a sable to induce changes ofabout 20°,, in R-.\dc, R-Kdc and R-Nfr and of about 1Oo0 in the perilymphatic K- content (R-(K-]). T h e graphs in Figure 3 also show that glibenclarnide, in about 10 min, was able to cancel complete]!cromakalim induced modifications.

'

T h e ability of glibenclamide to prevent cromakalim effects a a s also demonstrated by the fact that a 20-min pretreatment of the preparation M), which, by itself, with glibenclamide ( was almost ineffective ( n = 4), was able to prevent cromakalirn effects (data not shown). Control experiments were also made to verify a hether ethanol and DMSO, i.e., the solvents in which cromakalim and glibenclamide, respectively, -ere dissolved, might influence canal activity. These experiments demonstrated that DMSO, at the doses employed, has no effect ( n = 1), while ethanol, but only at the highest doses tested (0.7(J,,), produced a small but significant decrease ( l.5°0 1.7) in resting nerve firing rate (n = 4). Ethanol and cromakalim, therefore, have opposite effects.

D I s c cs s I 0 N T h e present experiments h a \ e clearlj shown that the administration of cromakalim in the endollmph, i.e. in the fluid bathing the hairbearing membrane of sensory cells, produces a

Cromakalim and frog semicircular canals

427

distinct increase in the resting firing rate of cancel or prevent cromakalim effects. The ampullar receptors. This increase is accompanied concentration at which glibenclamide became by changes in canal activity which indicate that active as a cromakalim antagonist was, therefore, it is mainly due to an increased intensity in the as for cromakalim, 100 times higher than that K+ resting receptor current which flows across active on other preparations (Cook & Quast 1990). On the basis of the present experiments, sensory cell bodies. I n fact the resting transepithelial potentential (R-Adc) was reduced it cannot be established whether these differences (indicating a decrease in sensory epithelium in potency have mainly pharmacodynamic or impedance and therefore an increase in the pharmacokinetic causes. I t must be remembered intensity of the receptor current) ; the resting that the hair-bearing membrane of sensory cells slow nerve potential (R-Ndc) showed a positive is not in free diffusional exchange with the shift, indicating an increased transmitter release endolymphatic fluid but is immersed in the from the synaptic pole of sensory cells and, cupula, a mucopolysaccharidic structure, which hence, spiking in afferents (Rossi et al. 1977); the might constitute a barrier to drug diffusion. At any rate, the observation that cromakalim K + concentration in the perilymphatic fluid was significantly raised, indicating an increased and glibenclamide act in the crista ampullaris output of K+ from the sensory cells. Insofar as sensory organ at concentrations similar to that confidence may be placed in the thesis that observed in /3 pancreatic cells suggests that the cromakalim, at the doses employed, acts as a K+ channels present on the hair-bearing memspecific K+ channel opener, all the above brane of sensory cells may be, to some extent, mentioned modifications can be explained. In similar to the K+-channels regulated by internal fact the K+ concentration in this medium is ATP described in the pancreas. T h e experiments also revealed that only mainly determined by three factors (Valli et al. 1990) i s . , (1) the K+ that emanates from the ampullar receptor resting activity, but not that basal pole of sensory cells; (2) the K+ which evoked mechanically, was sensitive to crodiffuses towards the perilymphatic space ; and makalim action. T o interpret this observation at (3) the K+ that is pumped again towards the least two hypotheses may be put forward. The endolymphatic fluid. T h e fact that the endo- first hypothesis is that cromakalim action is lymphatic administration of cromakalim is able evident only when K+-flows across sensory cell to produce a distinct increase in the peri- bodies are not very high (resting conditions) but lymphatic K + content can be explained only by not when almost all K+-channels are opened (under stimulation). A similar hypothesis was admitting that the drug is able to increase the K’ permeability of the hair-bearing membrane of put forward by Cook & Quast (1990) in /3 sensory cells and, in consequence, the outflow of pancreatic cells. The second is that two different mechanisms are present in nature. T h e first K+ from their basal pole. One factor which cannot be ignored in this mechanism, sensitive to cromakalim action, discussion is the high concentration of cro- would support receptor resting discharge, wheremakalim necessary for observing the above- as the second, not affected by the drug, would be mentioned effects. This concentration is at least involved in ampullar receptors’ mechano-sen100 times higher than that sufficient to produce sitivity (Valli et al. 1974, Guth et al. 1991). hyperpolarization in different kinds of smooth muscle (Cook & Quast 1990, Grana et al. This work was supported by grants from the Minister0 1991a, b) but similar to that active on both /l dell’universita e della Ricerca Scientifica e Tecnopancreatic cells (Quast & Cook 1989) and cardiac logica (600,:). miocytes (Richer et al. 1990). T h e view that cromakalim action may be specific in nature is strongly supported by the experiments with R E F E R E N C E S glibenclamide. This compound, which by itself ASHMORE, J.F. 1991. The electrophysiology of hair ( 10-6-10-4 M), is without appreciable activity cells. Ann Rev of Physiol 53, 465-476. (contrasting descriptions of glibenclamide effects BUCKINGHAM, R.E., HAMILTON, T.C., HOWLETT, have been reported (Quast & Cook 1989, D.R., MOOTOO, S. & WILSON,C. 1989. Inhibition by glibenclamide of the vasorelaxant action of Ruckingham et ul. 1989, Longmore et al. 1990, cromakalim in the rat. Br 3 Phurmacol 97, 57-64. M, to Grana et al. 1991a, b), was able, at

COOK,N.S. 1988. The pharmacology of potassium & CA\ERO, I . 1990. Cardiovascular and biological eEects of K* openers, a class of drugs with channels and their therapeutic potential. 7 ' r d r Phnrrnacol Sci 9, 21-28. , vasorelaxant and cardioprotective properties. Z@ Sri 47, 1693-1705. COOK,N.S. & QUAST, L. 1990. Potassium channels C. 1977. Postpharmacolog! . In X.S. Cook (ed.) Potassrnni Ross[, X L L . , VALLI,P. & CASELLA, s p a p t i c potentials recorded from aRerent Chunnrls, pp. 181 225. Ellis Ilorwood. nerve tihres of the posterior semicircular canal GRAKA, E., B-IRBIERI, A . & %OUT%,F. 1991 a. Effects in the frog. Bruin Rrs 135, 67-7.5. of cromakalim (BRI, 34915) on niechamical responses of rat \-as deferens to noradrenaline and STEKKERS, o., FERR.ARV, E. & b l I E L , C. 1988. Production of inner ear fluids. Ph,ysto/ Re?, 68, naphazoline. European .j' Plinrinutd 192, 79-84, 1083-1 127. GRAY.\,F.., R A K B I ~ R IA. , & % o \ T ~ ,F. 1991 h. V. & ROW, M.L. 1971. Effects Inhihition hy glibenclamide of the effect of cro- \-.AI.I.I, P., T.AGI.IETTI, of D-tubocurarine on rhe ampullar receptors of the makalim on responses of rat \ a s deferens to frog. . k t n Oto-ltrryngol (Stockh) 78, 51-58, naphazoline. Europran 3 Phnrmiirol 202, 93-96. \-..\LLI, P. & ZCCCA,G. 1976. T h e origin of slow GLTEI. P.s., ACRERT. A , , RICCI, A.J. i+ NORRIS, CXI. potentials in semicircular canals of the frog. Actu 1901. Differential modulation of spontaneous and Ot/J-/ifr)v?p/(Stockh) 81, 395-405. evoked neurotransmitter release from hair cells : V 4 1 . 1 ~ ,P., ZVCCA, G., BOTTA,I,. & C~SELLA, C. 1984. some novel hypotheses. Heriring Rrs 56, 69-78. Specificit! of the K-channels in labyrinthine HCDSPETH, -1.J. & IXIVIS, R.S. 1988. Kinetic analysis organs: effects of Kb- and Cs- on the conversion of voltage and ion-dependent conductances in process in frog semicircular canals. .J' Comp Ph,ysiol saccular hair cells ofthe bull frog, Rana Catesbeians. (.4)155. 739-'743. 3 Ph,ysiol (London) 400, 237-274. ~ ~ 4 1 . 1 . 1P., , %cCC.d, G . & BOTT.4, I.. 1990. Perilymphatic LONGMORE, J,, NEWGREES, D.T. 8- U.ESTOA,A.H. potassium changes and potassium homeostasis in 1990. I-Rects of cromakalim, RP 49356, diazosidc, isolated semicircular canals of thc frog. 3 Ph.ysrol glihenclamide and galanin in rat portal vein. (Idondon)430, 585--.i94. Eurnpan -3 Phririnitc~of190. 75-84. Z~cc.4, G., V.ALLI, P. & CAStLLA, C. 1982. Ionic QCAS~, U. &. Cook, N.S. 1989. \lo\ing together: K mechanisms sustaining activity in ampullar recepchannel openers and .lTP-sensitive K- channels. tors of the frog. .-frta Ofo-luryngal (Stockh) 93, T r m h PI~izr~nacol Scr 10, 43143.5. 355-362. RICHER,C;.. PRATZ, J., ~ I L L D ~P..R G, I L D I C E LJ.F. L.~,