The role of dihydropydine-sensitive Ca2 + channels in stimulus-evoked catecholamine release from chemoreceptor cells of the carotid body

Neuroscience Vol. 47, No. 2, pp. 463412, 1992 Printed in Great Britain 0306-4522/92 $5.00 + 0.00 Pergamon Press plc 0 1992 IBRO THE ROLE OF DIHYDROPYRIDINE-SENSXTIVE Ca*+ CHANNELS IN STIMULUS-EVOKED CATECHOLAMINE RELEASE FROM CH~MORECEPTOR CELLS OF THE CAROTID BODY A. OBEW,* A. ROCHER,* S. FIDONE* and C. CONZALEZQ *Department of Physiology, University of Utah School of Medicine, 410 Chipeta Way, Research Park, Salt Lake City, UT 84108, U.S.A. tDepartamento de Bioquimica y Biologia Molecular y Fisiologia, Facultad de Medicina, Universidad de Valladolid, C/. Rambn y Cajal s/n, 47005Valladolid, Spain Ahatraet-The present study utilized an in vitro preparation of the rabbit carotid body, with tissue catecholamine stores labeled by incubation with ‘H-tyrosine. The goal was to characterize pharmacologically the vol~g~~n&nt Ca*+ channels present in the type I (glomus) cells of this arterial chemoreceptor organ, and to elucidate their role as pathways for Ca2+ entry. We found that release of ‘H-dopamine induced by high external potassium was over 95% dependent on external cakium concentration and that this release was 9&100% inhibited by the dihydropy~~ne antagonists, nisoldipine and nitrendipine, and was potentiated by the dihydropyridine agonist, BayK 8444. Therefore, any stimulus-induced, cakiumdependent release of 3H-dopamine that was inhibited by nisoldipine and potentiated by BayK 8644, was considered to be supported by Ca2+ entry into the cells via voltage-dependent Ca2+ channels. Significant differences were observed in the release of ‘H-dopamine induced by 75 vs 25mM K+. On prolonged stimulation, release induced by 75 mM K+ was large and transient, whilst that induced by 25 mM K+, although more moderate, was sustained. The release elicited by 75 mM K+ was inhibited approximately 90% by 1.5 mM Co2+ or 625 nM nisoldipine, while release by 25 mM K+ was completely blocked by 0.6 mM Co*+ or 125 nM nisoldipine. Low PO,-induced release of 3H-dopamine was 95% dependent on Ca*+, and was inhibited by nisoldipine (625 nM) in a manner inversely proportional to the intensity of hypoxic stimulation, i.e. 79% inhibition at a PO, of 49 Torr, and 20% inhibition at PO2 of 0 Torr. BayK 8644 potentiatcd the release induced by moderate hypoxic stimuli. Release elicited by high PCOJlow pH, or by Na+-propionate or dinitrophenol~n~ining solutions, was approximately 80% Ca’+-dependent, and the ~hyd~y~din~ failed to modify this release. It is concluded that type I mlls possess vol~~de~nd~t Ca ‘+ channels sensitive to the dihydropy~dines, which in agreement with previous el~trophysiolo~~l data should be defined as L-type Ca*+ channels. Calcium entry which supports the release of 3H-dopamine elicited by moderate hypoxia should occur mainly through these channels while the release induced by strong hypoxic stimuli will be SetNed by Ca2+ entry which occurs in part via voltage-dependent Ca2+ channels, and in part through an additional pathway, probably a Na+/Ca2+ exchanger. The insensitivity to dihydropyridines of the release of )H-dopamine induced by high 1DC02/low pH, Na+-propionate and dinitrophenol may indicate a complete loss of efficacy of the drugs to modulate Ca 2+ channels under these conditions or more likely, that other mechanisms are activated, probably the Na+-Ca’+ exchanger. Carotid body (CB) chemoreceptors are thought to be composite receptors in which the type I (glomus) cells detect changes in blood PO,, PCO, and pH and respond with the release of neurotransmitt~ to activate the closely apposed chemosensory nerve terminals.~** One such neurotransmitter that has received considerable attention in recent years and is known to be released by the type I cells is dopamine (DA). This biogenic amine has been shown to be released in proportion to both the intensity of stimulation and the resultant sensory discharge recorded from the carotid sinus nerve $To whom correspondence should be addressed. Abbr~~~~~~ CB, carotid body; CSN, carotid sinus nerve; DA, dopamine; DHMA, dihydrox~~delic acid, DOPAC, dihydroxyphenyl acetic acid; NE, norepinephrine. (CSN). This relationship between stimulus intensity, CSN sensory discharge, and DA release has been shown to be Ca*+-dependent,@ and is known to occur with both physiolo~cal (low PO*, low p~)2’,4t.42 and pharmacological (cyanide, dinitrophenol, 2-deoxyglucose)3*339 stimuli. In a previous study,3 it was postulated that type I cells might possess voltage-dependent Ca2+ channels. This suggestion was proposed to explain the Ca2+-dependent release of DA elicited by high external K+.17 It was also shown in a subsequent preliminary study” that nitrendipine, a Ca’+ channel blocker of the dihydropyridine group, markedly inhibited the high K+ and low PO,-induced release of DA. Quite recently, several different laboratories working with isolated type I cells have provided direct electrophysiological evidence for the presence of voltagedependent Ca’+ channels in these cells.‘6~“~32~33~” 463 NSC 4112-l The present study was undertaken to characterize pharmacologically the type(s) of voltage-dependent Ca’+ channels present in the chemoreceptor type I cells of the CB, and to define the role of these Ca” channeIs in the response to different forms of chemoreceptor stimulation. Since the involvement of voltage-dependent Ca*+ channels in a cellular response couples a voltage dependency to that response,” the present study should reveal which chemoreceptor stimuli produce significant changes in the membrane potential of type I cells. Calcium entry into these cells was assessed by measuring the Ca’ ’ _ dependent release of DA from the tissue. We chose this indirect method, instead of the measurement of 45Ca2t fluxes, because it has been found that, for a variety of tissues, there exists a very close relationship between Ca2+ influx and the intensity of the tissue’s Ca*+ -dependent response.6.‘3.29.3* Furthe~ore, the small size of the CB and the cellular heterogeneity of the tissue (type I and II cells, endothelial cells, neurons, etc.) make the measurement of 4rCa’ ’ fluxes both technically difficult and of questionable signihcance. The release of DA, on the other hand, is a parameter specific to the type I ceils.‘” EXPERIMENTAL PROCEDURES The present study utilized an in cirro CB preparation fram adult (1.5-2.5 kg) New Zealand rabbits. The animals were anesthetized with Na+-pentobarbitone (30-40 mg/kg, i.v.), tracheotomized, and both carotid bifurcations were exposed and removed for further dissection in a lucite chamber filled with ice-cold 100% O,-equilibrated Tyrode’s solution (in mmol/l: NaCI, 140; KCl, 5; CaCI,, 2; MgCl,, 1.1; HEPES, IO mM; glucose, 5.5 mM; pH adiusted with I N NaOH to 7.42). Carotid bodies were ileaned of surrounding tissue and incubated for 2 h with 3 H-tyrosine (20 fl M, 30 Ci/mmol), as previously described.lg Under these conditions, the organs synthesized approximately 20-22 pmol of 3H-catechoiamine/mg tissue (i.e. about IOpmol, or more than 5 x IO5 d.p.m./CB). Following incubation, the CBS were transferred to new vials containing 4 ml of tyrosine-free media and incubated for an additional 2 h to wash out the rapidly releasable pool of labeled catecholamine;’ during this period the medium (4 ml) was replaced by prewarmed fresh solutions every 30 min and the samples were discarded. Thereafter, the collection of incubation media for analysis was grouped in stimulation cycles; one stimulation cycle consisted of a period of incubation with the stimulus (test) solution, preceded by a similar period and followed by several poststimulus periods of incubation with standard solution to determine basal responses and allow the preparation to recover, respectively. The duration of each of these periods in any given experiment is provided in the Results section. The number of stimulation cycles per experiment varied from one to four depending on the nature of the test incubation media. The stimulus-evoked release declined in successive presentation of the stimulus in parallel to the basal release and therefore the ratios of evoked to basal release remained constant throughout the experiments.*’ The incubation media were continuously bubbled with the appropriate gas mixture, and the POZ was monitored with an 0, electrode in selected experiments. Under these experimental conditions, the threshold for DA release by low O2 was approximately 75 Torr Or (i.e. normal saline equilibrated with IO% 0,/90% N,). All coflected samples were added to a carrier mixture containing 0.3 M acetic acid, I mM ascorbic acid, and looI I F=l r-l OHMA B 55mM 5 %O, Fig. 1. Profile of ‘H-catechols (‘H-CA) released from the carotid body following incubation with ‘H-tyrosine. ‘H-Catechols present in the alumina eluates were taken as 100%. Note that the relative proportions of ‘H-DA and ‘H-DOPAC are reversed under basal (B) vs stimulus conditions (55 mM K+ and 5% 0,/95% N;equilibrated media; IO mm), owing to saturation of degradation mechanisms during stimulation. Nonetheless, DA + DOPAC still represent over 85% of 3H-catechols under all conditions. 100 GM unlabeled DA. Radiolabeled cdtechols in the incubation media were analysed by adsorption to alumina at pH 8.6, and elution with I N HCI.” Since the CB receives a rich sympathetic innervation,” the ‘Hsatechols released may have two origins, the type I cells and the sympathetic nerve endings. 3H-DA is synthetized and stored exclusively in type I cells while ‘H-norepinephrine (NE) is synthetized and stored, both in type I cells and in the sympathetic endings;” therefore to assess the origin of the 3H-catechols released it was necessary to identify its chemical nature. To do so, the alumina eluates were vacuum-concentrated to dryness, resusnended in 40 u1 of mobile obaseZ convening 5 UR of u&beled DA, N’E, dihydroxyphenyl acetic acid <kPAC) and dihydroxymandelic acid (DHMA), and separated using thin-layer chromatography according to the method of Fleming and Clark.22 The labeled compounds were eluted and counted as previously described.s5 The profiles of released ‘H-catechols under basal conditions and in response to high K+ and low PO, stimulation are shown in Fig. I. Obse-me that regardless of the expected variations18,2i.” m the ‘H-DA to “H-DOPAC ratios between basal and stimulus samples, ‘H-DA + 3H-DOPAC (the main catabolite of ‘H-DA in the CB)r4 represent over 85% of the total 3H-catechols released, indicating that type I cehs are their main source. Because the same analytical profile was found in pilot experiments for all the stimuli tested in this study, in most of the replicate experiments the alumina eluates were directly counted and “H-catechol release was taken as a measure of ‘H-DA release from type I cells. Results are expressed as c.p.m. of 3H-DA present in each collected sample, as the ratio of stimulus-evoke release to basal release (evoked/basal); or as per cent of tissue content. Materials. 3,S-3H-Tyrosine, 4@-@ Ci/mmol was purchased from New England Nuclear and aluminum oxide (alumna) from Serva (Heidel~rg). Nisol~pine and nitrendipine were a generous gift from Miles Laboratories, and BayK 8644 was provided by Prof. A. Garcia (U.A.M. Spain). RESULTS The Ca2+ dependency of 3H-DA release observed in this study is summarized in Table I. The results are similar to those previously reported for the cat CB.3,38-40 Ahbough not shown in the table, most of Carotid body Ca 2+ channels and catecholamine release Table 1. Ca2+ dependency of 3H-dopamine release induced by different stimuli Percentage reduction Stimulus Normal Ca2+ 0 M Ca2+ in 0 M Ca2+ High K+ (40mM) 105 + 8.2 4.80 + 0.38* 95.60 Hypoxia (2% 0,) 28.6 f 3.3 0.69 + 0.05* 97.50 High PCO,/pH 6.6 1.60*0.12 0.31 f 0.09* 81.00 Na+-Propionate (15 mM) 1.02*0.11 0.21 f 0.06* 80.00 Dinitropyridine (2.5 x 10d4 M) 47.65 + 3.82 9.00 + 1.3; 79.00 Cyanide (1 Oe4 M) 4.00 f 0.66 0.93 f 0.13* 76.75 Veratridine (5 x lo-’ M) 16.9 + 1.4 1.6 + 0.09; 90.50 Ouabain ( 10e5 M) 11.2& 1.5 0.80 + 0.21* 93.00 Data are expressed as the ratios of evoked to basal release. Except for veratridine and ouabain, the protocol in these experiments was as follows: a stimulation cycle with Caz+-containing media was applied first, then two stimulation cycles with nominally free Ca2+ media, followed by a stimulation cycle with Ca2+containing media; the incubation with the test solutions was 10 min in all stimulation cycles. In the experiments with veratridine, pairs of CBS were incubated in Ca2+-containing and in nominally free Ca2+ media 30 min prior to and during the application of veratridine (20min); media were changed and collected for analysis every 10 min. The experiments with ouabain were identical except for the duration of the exposure to the test solution (30min). In all the cases the data are means f S.E. of the mean of six individual values. Differences of the mean evoked release in Ca2+-containing vs Ca 2+-free media are statistically significant (*P < 0.001; paired Student’s r-test). 465 the stimuli were tested at varying intensities, and the Ca*+ dependency of release was present at all stimulus intensities. Time course of high K+-induced release of ‘Hdopamine The time course of ‘H-DA release induced by 75 and by 25 mM K+ are compared in Fig. 2A; in both cases the media were changed and collected for g30. 2 320. -0 % I L 10. $ ,I 0. A. n lcontrol q Stimulus -3min , , + , 75mM K’ 25mM K’ B. a’““\ 75mM K’ “1 25mM K’ 32 e = 50 s I\ I lrn - ._ a E oJ\L-s O_~ 3 9 15 3 9 15 min Fig. 2. Time course of K+-induced release of ‘H-DA. (A) Two single stimulus cycles for 75 (left) and 25 (right) mM K+. (B) Mean fractional release (* S.E.) of ‘H-DA in response to 75 and 25 mM K+. Total ‘H-DA released during the 15&n incubation with high K+ was taken as lOO%, and the fraction collected in each 3-mitt period is expressed as a per cent of this total. Data are from six experiments. analysis every 3 min for a total of 15 min. As shown in Fig. 2B, most of the ‘H-DA release evoked by 75 mM K+ was collected in the first 3-min fraction, while with 25 mM K+, release remained nearly constant for the duration of the stimulus. These results would suggest that Ca*+ entry inactivates rapidly during exposure to 75 mM K+ but that entry is sustained during exposure to 25 mM K +. The unsustained release of 3 H-DA during exposure to 75 mM K+ is, in fact, due to inactivation of Ca*+, as confirmed by Fig. 3. The first panel of Fig. 3 illustrates a typical profile of declining release in 75 mM K + . In a second CB (middle panel) removal of Ca*+ from the incubation medium not only abolished the immediate release of 3H-DA, but also prevented its subsequent release following re-introduction of Ca*+ into the medium. This suggests that prior depolarization of the cells in a Ca*+-free I 0 W2l"td u3min lh+__li oca++ .lonon 75mM 71mM 71mM Fig. 3. Effects of 0 M Ca2+ and ionomycin (30pM) on ‘H-DA release. induced by 75 mM K+. The experiment shows that neither depletion of releasable ‘H-DA stores, nor adaptation of exocytotic activity are responsible for accommodation of release induced by 75 mM K+. 466 . Control D 2 5mM K+ -3mm .OCaet. Nis Fig. 4. Effects of OM Caz4 and nisoldipine (125 nM) on IH-DA release induced bv 25 mM K+. The exneriment demonstrates the involve&W of dih~dropy~dine~sensitive Ca*+ channels in the sustained release of ‘H-DA induced by 25 mM K+. medium prevented subsequent entry of CaZt apparently by inactivating the pathways for Ca*+ entry. Data obtained using a third CB {right panel), appeared to support this notion: introduction of ionomycin, a Ca2+ ionophore, during the declining phase of release reactivated the release process. This precluded the possibility that exhaustion of the releasable ‘H-DA pool, or loss of sensitivity to Ca’ + by the exocytotic machinery, was responsible for the unsustained release normally observed during exposure to 75 mM K+ . Very different results were seen with 25 mM K+. The left panel of Fig. 4 shows the normal time course of )H-DA release during exposure to 25 mM K+; in the middle panel, removal of Ca2+ during the initial exposure to 25 mM K+ abolished the increased release, but did not prevent reactivation of increased release upon ~introdu~tion of Ca*+. Nisoldipine reduced release when introduced late in the incubation period (Fig. 4, right panel), indicating that secretion depends on Ca2+ entry into type I cells via a dihydropyridine-sensitive pathway (see also below}. These findings also indicate that, unlike the case for 75 M I(+, predepolarization with 25mM K+ does not inactivate the pathway for Ca*+ entry. 10, i 75mM I<+ K+ Pharmacology of high K’ -induced rvkust. I)/ ‘Hdopamine The effects of different concentrations of Ca? ! in the incubation media on the basal (5 mM K + ) and high K+ (25 and 75 mM)~voked release of 3 H-DA are shown in Fig. 5. In these experiments, 25 mM and 75 mM K + were applied for 10 min, and the media were sampled for analysis every 10 min. In the left of the figure, the plot of 3H-DA release vs Ca’ f concentration shows that basal release did not change appreciably at Low Ca’+, but tended to decrease as Cat + was increased. Evoked release in 25 mM K * was abolished (over 95%) in 0 mM Ca”, was maximum at 0.5 mM but then declined at higher Ca”. The release evoked by 75 mM K’ was likewise abolished in 0 mM Cat+, and increased dramatically with increasing Ca2+. To the right in Fig. 5 a plot of the ratio of evoked reiease to basal release as a function of Ca*+ illustrates the reduction of the evoked response to 25 mM K+ at 10 mM Ca2+. and the progressive increase in the evoked release with 75 mM K+ with increasing Ca’ ’ These changes in CB 3 H-DA secretion with increasing Ca2 i are similar to those observed with other tissues.” They presumably reflect the increase in the threshold for Ca: ’ channel activation with increasing Cal’ that was evident with small depolarizations (25 mM K+), but which disappeared upon stronger depolarizations (75 mM K+).27,4’ The inset in the figure shows the protocol of sample collection for analysis. Cd2+ and Co’+ , well-known Ca2+ channel blockers, produced a dose-dependent inhibition of the ‘H-DA release induced by 25 and 75 mM K+. as shown in Fig. 6. Both responses to elevated K ’ were completely blocked by 100 pm Cd’+. Variation in the effect of different Cd2+ concentrations on basal release precluded an accurate calculation of an tc,, for release evoked by 2.5 mK K A. but did not interfere significantly in calculating a value for the much larger release evoked by 75 mM K + . For this we found an tO mM Ca+’ 0 5 10 mM Ca+* Fig. 5. Effect of Ca2+ concentration on 3H-DA release under resting conditions and in response to high K+. Left: r&ease expressed as per cent of tissue content. Right: ratio of evoked to basal release. Inset: sequence of medium renewal and collection for analysis. Each data point represents the mean + S.E. of at least four samples. Carotid body Car+ channels and catecholamine release 461 ICd”lpM ICd’lmM Fig. 6. Effects of Cd*+ (left) and Co2+ (tight) on ‘H-DA release under resting conditions (5 mM K+) and during high K+ (25 and 75 mM). Experimental protocol as in Fig. 5. Cd2+ or Co*+ was introduced 10 min before and during the incubation with high K+. Release is expressed as per cent of tissue content. Data points are mean + SE. of four to six experiments. GQ of lOpM, which is lower than the 56 PM observed for PC12 cells by Freedman et d2’ We found that Co2+ at any concentration did not appreciably modify the basal release, but at 0.6 mM completely abolished the release evoked by 25 mM K+, and at 1.5 mM inhibited 93% of the release induced by 75 mM K+. The q,, for the Co2+ effect on the 75 mM Kc-induced release was 0.3 mM, with maximum inhibition at 1.5 mM. Both values are comparable to those reported for PC12 cell~.~~ Neither nisoldipine (Fig. 7A) nor nitrendipine (Fig. 7B) modified the basal release 6f ‘H-DA. However, both exhibited a dose-dependent inhibitory effect on the release induced by 25 and 75 mM K+. At lob9 M, nisoldipine inhibited 50% and 10% of the release induced by 25 and 75 mM K+, respectively. As shown in the inset, this difference was maintained throughout the range of nisoldipine concentrations (Fig. 7A). 0 lo-’ lo-’ lo-’ Nisoldipine (Mb Nitrendipine, on the other hand, required higher concentrations to inhibit release, and exhibited less marked differences in its effects on 25 and 75 mM K+-induced release (inset Fig. 7B). These data demonstrate that the voltage-dependent Ca2+ channels in the CB type I cells are dihydropyridine sensitive. Figure 8 illustrates further proof of the dihydropyridine sensitivity of the Ca2+ channels in type I cells; BayK 8644, a Ca2+ channel agonist belonging to the dihydropyridine family,49 markedly potentiated the release of ‘H-DA by 30 mM K+. Effects of dihydropyridines on low PO,-induced release of ‘H-dopamine The participation of Ca2+ channels in the response to low PO, was assessed using stimuli of moderate intensity (10 min incubation with 7% O2 equilibrated media; PO2 = 49 Torr). The results of a single experiment with two stimulation cycles are shown in Fig. 9A (left). Introduction of nisoldipine (625 nM) into the media 10 min prior to presentation of the second stimulus inhibited the response. The inhibition of the evoked release from four such experiments had a mean of 79 f 7% (Fig. 9A, right). In Fig. 9B, BayK 8644 was used in a similar protocol and produced a five-fold increase in the release of 3 H-DA in response to 7% 02. These data implicate dihydropyridinesensitive Ca2 + channels in the mediation of Ca2+ entry during low PO,-induced release of ‘H-DA. The effects of 625 nM nisoldipine on the low POrinduced release of ‘H-DA are summarized in Fig. 10. The inhibitory effect of this dihydropyridine decreased as the intensity of the low PO, stimulus was increased. Even at high stimulus intensities (e.g. 2% 02/98% N2 media) over 95% of ‘H-DA release was Ca2+-dependent (see Table l), suggesting that as the intensity of hypoxic stimulation increases, the role 4 -75mM K+ -25” ” .--_ 5 ” - 2 Nitrendipine hl) Fig. 7. Effects of nisoldipine (left) and nitrendipine (right) on the release of 3H-DA under resting conditions (5 mM K+) and during high K+ (25 and 75 mM) stimulation. Experimental protocol as in previous figure, nisoldipine and nitrendipine were present in the incubation media 10 min prior and during high K+ stimulation. Release is expressed as per cent of tissue content. Insets: dose-response inhibition curves. All data points are means + SE. of four to six values. C 3 30mM K+ c-3 BAY-K Fig. 8. Effat of the dihydropyridine Ca* + channel agonist BayK 8644 (1~ M) on 30 mM K+ induced release of 3 H-DA. Note that BayK 8644 did not affect basal release. of Ca2+ channels in mediating Ca2+ entry becomes less important. To test the possibility that the loss of efficacy of nisoldipine with increasing strength of the hypoxic stimuli is related to the proportionally increasing acidification that is produced during low PO, stimulation,15.& we utilized a strong hypoxic stimulus of shorter duration (2% 0,; 3 min). The rationale for these experiments was that acidic stimuli evoke Ca’+- dependent release that was insensitive to dihydropyridines (see Table 1 and below) and that intracellular acidification produced under low 0, increased dramatically between 3 and 10 min (see Ref. 15). In these experiments (Fig. 1 l), pairs of CBS were subjected to two cycles of low PO* stimulation; one CB served as rn7%0, 2 _-Iom,n ; P x : 40 E2’ 0 ,,h,\ d u 20 d z . 0 id BAVK - -7%O, _BAYK Fig. 9. Dihydropyridine and low P&-induced release of ‘H-DA. (A) Left: effect of nisoldipine (625 nM) on release induced by 7% O,-equilibrated media (10 min; black bars). (A) Right: evoked to basal release ratios from four experiments (x + S.E.). (B) Left: effect of 1 pM BayK 8644 was tested. (B) Right: mean evoked to basal release ratios. Differences of the mean evoked release in control vs test conditions were statistically significant, *P < 0.01; paired Student’s r-test. 100. 0 7x02 WO, 2%4 N, Fig. 10. Decreasing potency of nisoldipine (625 nM) in inhibiting the release of ‘H-DA induced by increasingly strong hypoxic stimuli. Experimental protocol as in Fig. 9. Duration of low PO, incubations was 10 min. Means + SE. of the mean of six or more samples. a control while the second was used to assess the effect of nisoldipine during the second stimulus cycle. The ratios of evoked release in the first and second cycles from four such experiments are shown in the right hand panel of Fig. 1 I. It is evident from a comparison with the results of Fig. 10, that shortening the duration of the stimulus increased the efficacy of nisoldipine inhibition of the low PO>-induced release (29% vs 73%). Efsects of dihydropyridines on the release qf ‘H-dopa - mine induced by d@erent stimuli In addition to high K+ (a depolarizing agent useful in revealing the participation of voltage-dependent Ca2+ channels in depolarization-mediated transmitter release) and low PO2 (the principal natural stimulus for the carotid chemoreceptors), we also examined the effects of the dihydropyridines on the secretory response evoked by other natural and pharmacological stimuli of the carotid chemoreceptors (Table 2). Stimuli that evoke only a moderate release of ‘H-DA (20% C02/pH 6.6 and Nat-propionate) were employed to test the effects of the Cal+ channel agonist, Fig. 11. Effect of nisoldipine on 3 H-DA release induced by severe hypoxia. Duration of stimulus (2% Oz iri N,), 3 min (black bars). Left: comparison of control (top) and nisoldipine (625 nM)-treated (bottom) carotid bodies. Right: ratios of evoked release (second/first stimulus presentation). Means + S.E. of such four paired experiments. Carotid body Ca *+ channels and catecholamine release 469 Table 2. Dihydropyridine sensitivity of 3Hdopamine release evoked by different stimuli Percentage Stimulus Control Dihvdrouvridine effect 20% CO,/pH 6.6 NaPr (15mM) Dinitropyridine (2.5 x 10e4 M) Cyanide (2.5 x IO-” M) 1.6kO.17 1.5 f 0.21t - l.OkO.16 1.1 +0.17 48.0 f 7.0 47.4 f 8.5$ 17.8 f 2.1 11.4 f l.o**$ -36% Vkratridine (5 x 10m5 M) 20 min 22.8 + 4.0 16.4 + 1.9.1 -28% Veratridine (5 x 10e5 M) 3 min 9.5 * 1.6 5.0 + 0.9**i -47% \ - tBayK 8644, 1 PM; Snisoldipine, 625 nM. Experimental protocols as in Fig. 9; duration of exposure to test solutions 10 min in all the cases except when indicated otherwise. Data are means f SE. of the means of four or more individual values. Differences of the mean evoked release in control vs test conditions were statistically significant for cyanide and veratridine (*P < 0.05; **P < 0.02, paired r-test). BayK 8644 (1 p M); stronger pharmacological stimuli were used to test the effects of the antagonist nisoldipine (625 nM). The protocol for these experiments was identical to that described above for the experiments of Fig. 9. Our findings from these experiments, summarized in Table 2, are as follows: (i) it appears that high PCO,/low pH, Nat-propionate, and the protonophore dinitrophenol, do not trigger the activation of voltage-dependent Ca2+ channels because release is unaffected by the dihydropyridines. It is noteworthy that in other preparations these three stimuli produce intracellular acidification and fail to appreciably modify the membrane potential.2SsM This suggests that the coupling between the increase in intracellular H+ and the Ca2+ dependent release of DA from type I cells follow different pathways than those used by low PO,; (ii) cyanide is a powerful stimulant classically used with carotid chemoreceptars,’ but its mechanism of action, although probably related to histotoxic hypoxia, is not understood.9~‘0~39 Whatever its precise mechanism of action, it is likely that it involves depolarization of the type I cells, because cyanide-induced release of 3H-DA is partially sensitive to blockers of voltage-dependent Ca2+ channels; (iii) it has recently been shown that rabbit type I cells are excitable cells that possess tetrodotoxin-sensitive Nat channels.‘7*32~33~53 We further showed that the ‘H-DA release induced by veratridine, a Nat channel activator, is Nat and Ca2+ dependent, and tetrodotoxin sensitive.45 It was therefore of interest to explore the effects of nisoldipine on the veratridine-induced release. Table 2 shows that this release is partially sensitive to the dihyropyridine, and that the sensitivity increases as veratridine exposure decreases. DISCUSSION The results show that high K+-induced release of ‘H-DA from type I cells is over 95% dependent on Ca2+, is inhibited in a dose-dependent manner by Cd2+ and Co’+ and is modulated by dihydropyridines. We also found that low PO,-induced release is likewise Ca2+ dependent, modulated by dihydropyridines, and that this modulation is dependent on the intensity of hypoxic stimulation. The release of 3HDA induced by acidifying stimuli also exhibited a marked Ca2+ dependence, but was insensitive to modulation by dihydropyridines. In a previous publication from our laboratory, before it was known that type I cells were electrically excitable, Almaraz et ale3 postulated the existence of voltage-dependent Ca*+ channels in these cells to explain the release of ‘H-DA induced by high K+. Quite recently, the presence of Ca2+ channels in type I cells has been directly confirmed by several groups in electrophysiological studies of freshly dissociated cells.L6~17*32~33*53 Analysis of the Ca2+ tail currents, recorded at instantaneous repolarizations, revealed a large component with a rapid decay and a small component which deactivated slowly, equivalent to about 95% and 5% of the total Ca2+ current amplitude, respectively. 53 However, because of a slow time course of inactivation, insensitivity of the closing kinetics to long conditioning pulses and the lability of Ca2+ current following intracellular dialysis during whole-cell patch-clamp recordings, two Ca2+ currents could not be resolved. It was therefore concluded that most, if not all, of the Ca2+ current was mediated by rapidly deactivating (L-type) Ca2 + channels.53 Our present data are consistent with these electrophysiological findings in showing that Ca2+ entry into type I cells activated by high K+ are dihydropyridine sensitive, a characteristic of L-channels;35*52 this conclusion also agrees with the reported potentiation of Ca2+ currents recorded in dissociated chemoreceptor cells by BayK 8641.26 However, Akaike et uZ.‘*~ have described Ca2+ channels that behave kinetically as T-type channels but nonetheless are sensitive to dihydropyridines. In order to explain differences in the time course of the secretory response induced by 25 mM and 75 mM K+, Almaraz et al.’ postulated the existence of two types of Ca2+ channels in type I cells. However, the available data more likely favor the existence of a single type of voltage-dependent Ca2+ channel in type I cell~,‘~ despite the finding in the present study of different sensitivities to nisoldipine and Co2+ in the release response induced by 25 mM and 75 mM K +, and the two components found in the analysis of the 470 A. OHESO r/ (I/. Ca’ ‘- tail currents 53 both of which could be viewed as compatible with ;he existence of two types of Ca’? channels. With the knowledge that type I cells generate action potentials, ‘7.3’,” it is possible now to explain differences in the time course of release in response to 25 and 75 mM K+ with a single set of Cal+ channels of the L-type. The large, transient release in response to 75 mM K i would result from the maximal, or near maximal, activation of Ca’+ channels that within a few seconds completely inactivate.53 The magnitude of the sustained depolarization with 75 mM K+ ensures continued inactivation, with the consequence that the secretory response subsides. At the same time, the high level of depolarization should favor the entry of Ca* + via the Na +/Ca’ + exchanger,” and thereby account for both the small response to 75 mM K+ late in the stimulus period, as well as the small difference between the Ca2+ dependency and the sensitivity to channel blockers. The moderate, sustained release in response to 25 mM K’ would result from the depolarization of type I cells to the activation threshold for Na+ and Can’t channels, and the consequent generation of action potentials.” Repolarization of the action potential resulting from the activation of voltage-gated K+ channels, both Ca2+ dependent and independent”,” will limit the entry of Ca 2+ to the duration of the action potentials. At the same time repolarization will allow the removal of inactivation from Ca2+ channels, thereby permitting their recruitment by successive action potentials. The slow kinetics of Ca” channel activation and the interspike intervals will limit the entry of Ca*+ and therefore the release response. Dihydropyridine inhibition and potentiation of low PO,-induced release of 3 H-DA is evidence for the participation of voltage-gated Ca2+- channels in the physiological response of the carotid chemoreoceptors. At the same time the data indicate the existence of additional pathways for Ca2 + entry during hypoxic stimulation: low PO,-induced release of “H-DA is over 95% Ca 2+ dependent at all intensities of the stimulus and the blockade of the release by dihydropyridines ranges from 79% with mild hypoxic stimulus to about 20% under anoxia. The recent finding that chemoreceptor cells exhibit a K + current that is reversibly inhibited by low oxygen pressure’6.3?.33 suggests a possible mechanism for the depolarization required to activate the Ca2+ channels. The additional pathways for Ca2+ entry activated during hypoxic stimulation could be related to the concomitant acidification produced under low O2 that is parallel to the intensity and duration of the hypoxia.‘4~15~5’ The data presented in Figs 10 and I 1 support this contention; as the intensity and duration of the hypoxic stimuli increase the sensitivity to dihydropyridines decreases and tends towards the insensitivity of the release response observed with pure acidic stimuli (see Table 2). It would appear then that under in vitro conditions the participation of dihydropyridine-sensitive voltage-gated Ca’ ’ channels during strong hypoxia is limited to the initial moments of the stimulus and thereafter the pathways for Ca” entry activated by acidic stimuli are those that support the Ca*’ -dependent release of ‘H-DA. In apparent contradiction with this interpretation is the reported inhibition by low pH of the $-sensitive K + current in dissociated chemoreceptor cells;” more recently, however, it has been also shown that Na’ and Ca2+ currents in these cells are similarly inhibited by low pH32 (see also Ref. 36). It must be pointed out also that in uivo, physiological hypoxia IS accompanied by systemic alkalosis. Situations comparable to the in vitro conditions are seen iri r:ico at extremely low POZ, and in certain lung pathologies and circulatory shock. Regarding the nature of the pathways used for influx of Ca’+ into chemoreceptor cells during acidic stimulation, our data are compatible with two possibilities: (i) Ca’+ enters the cells via the voltage-gated Ca*+ channels, which became insensitive to dihydropyridines under the specific conditions set by the stimuli, and (ii) Ca 2+ influx into chemoreceptor cells occurs through a dihydropyridine-insensitive pathway. Although it is not possible to distinguish between these alternatives without direct recordings of Ca2+ currents under the same conditions, there are other data that favor the second alternative. For example, on theoretical grounds it should not be expected that dinitrophenol produces any significant depolarization at the plasma membrane level. In fact it has been shown that dinitrophenol produces an intense acidification in lymphocytes without any appreciable change in their membrane potential.” It has also been shown that the other two acidifying stimuli used in the present study do not produce any significant change in the membrane potential of snail neurons.” Therefore, it appears inconceivable that these stimuli activate voltage-gated Ca’ ’ channels. In addition, a recent publication from our group suggested that the Ca’+-dependent release of ‘H-DA elicited by acidifying stimuli would be mediated by influx of Ca’ + via the Na + Ca’ + exchangerc3” it was suggested also that the accumulation of Na + accompanying proton extrusion during an acidic load, would be responsible for the reversal of the Na’ / Ca2+ exchanger.39 Confirmation of these suggestions has been obtained.” The release of DA induced by acidic stimuli was found to be Na+-dependent and inhibited by blockers of the Na + -dependent protonextruding mechanisms indicating that entry of Na+ is the driving force for Ca ‘+ entry during acidic stimulation. This information, coupled to the complete insensitivity of the acidic stimuli release response to BayK 8644 and to nisoldipine, supports the conclusion that transduction of the acidic stimuli does not involve depolarization of type I cells nor participation of voltage-gated Ca’ ’ channels. The time-dependence of nisoldipine on veratridineinduced 3H-DA release, i.e. the inhibition produced by the Ca2+ -channel block decreased with the pro- Carotid body Ca *+ channels and catecholamine release 471 longed exposure to veratridine (see Table 2), may be explained as follows. It is conceivable that, after an initial burst of Ca’+ inthtx via Ca*+ channels, the sustained depolarization produced by the alkaloid permanently inactivates the voltage-dependent Ca2+ channels in the chemoreceptor cell~.~~ With time, the depolarization itself, and the intracellular accumulation of Na+ produced by veratridine,4s12 will trigger Na+/Ca2+ exchange that will provide the influx of Ca2+ to support the veratridine-induced release.” A Na+/Ca2+ antiporter may also be an important mechanism in the secretory response induced by cyanide.” Acknowledgements-This work was supported by grants PB89/0358, FISS 89/0340 and Junta de Castilla y Leon 1101/89 (Spain), and by USPHS Grants NS12636 and NS07938. REFERENCES 1. Akaike N., Kanaide H., Kuga T., Nakamura M., Sadoshima J. and Tomoike H. (1989) Low-voltage-activated calcium current in rat aorta smooth muscle cells in primary culture. 1. Physiof. 416, 141-160. 2. Akaike N., Kostyuk P. G. and Osipchuk Y. V. (1989) Dihydropyridine sensitive low threshold calcium channels in isolated rat hypothalamic neurones. J. Physiol. 412, 181-196. 3. Ahnaraz L., Obeso A. and Gonzalez C. (1986) Effects of high potassium on the release of ‘H-dopamine from the cat carotid body in vitro. J. Physiol. 379, 293-307. 4. Amy C. and Kirshner N. (1982) =Na+ uptake and catecholamine secretion by primary cultures of adrenal medulla cells. J. Neurochem. 39, 132-142. 5. Anichkov S. V. and Belen’kii M. L. (1963) Pharmacology of the Carotid Body Chemoreceptors. Macmillan, New York. 6. Baker P. F. and Knight D. E. (1984) Calcium control of exocytosis in bovine adrenal medullary cells. Trends Neurosci. 7, 120-126. 7. Beaty G. N., Cota G., Siri L., Sanchez J. A. and Stefani E. (1987) Skeletal muscle Ca++ channels. In Structure and Phvsioloev of the Slow Inward Culcium Channel (eds Venter J. C. and Triaele D.). DD. 123-140. Alan R. Liss. New York. 8. BelmoncC:and Gonzalez C. (1983) Mechanisms of chemoreception in tK carotidbody: possible models. In Physiology of the Peripheral Arterial Chemoreceptors (eds Acker H. and G’Reagan R. G.), pp. 197-220. Elsevier, Amsterdam. 9. Biscoe T. J. and Duchen M. R. (1989) Electrophysiological responses of dissociated type I cells of the rabbit carotid body to cyanide. J. Physiol. 413, 447468. 10. Biscoe T. J., Duchen M. R., Eisner D. A., G’Neill S. C. and Valdeolmillos M. (1989) Measurements of intracellular Ca+ + in dissociated type I cells of the rabbit carotid body. J. Physiol. 416, 421-434. 11. Blaustein M. P. (1988) calcium transport and buffering in neurons. Trends Neurosci. 11, 438-443. 12. Catterall W. A. and Nirenberg M. (1973) Sodium uptake associated with activation of action potential ionophores of cultured neuroblastoma and muscle cells. Proc. natn. Acad. Sci. U.S.A. 70, 3759-3763. 13. Collazos J. M. and Sanchez A. (1987) CAMP reduces the affinity of Ca*+-triggered secretion in platelets. Fe& Eur. biochem. Sots Lett. 215, 183-186. 14. Delpiano M. A. (1987) Glycolysis as a link for chemoreception? In Chemoreceptors in Respirutory Control (eds Ribeiro J. A. and Pallot D. J.), pp. 5968. Croom Helm, London. 15. Delpiano M. A. and Acker H. (1985) Extracellular pH changes in the superfused cat carotid body during hypoxia and hypercapnia. Brain Res. 342, 273-280. 16. Delpiano M. A. and Hefscheler H. (1989) Evidence for a PO,-sensitive K+ channel in the type I cells of the rabbit carotid body. Fedn Eur. biochem. Sots Lett. 249, 195-198. 17. Duchen M. R., Caddy K. W. T., Kirby G. C., Patterson D. L., Ponte J. and Biscoe T. J. (1988) Biophysical studies of the cellular elements of the rabbit carotid body. Neuroscience 26, 291-3 11. 18. Famebo L. 0. and Malmfors T. (1971) ‘H-Noradrenaline release and mechanical response in the field stimulated mouse vas deferens. Acta physiol. stand, Suppl. 371, l-18. 19. Fidone S. J. and Gonzalez C. (1982) Catecholamine synthesis in rabbit carotid body in vitro. J. Physiol. 333, 69-79. 20. Fidone S. J. and Gonzalez C. (1986) Initiation and control of chemoreceptor activity in the carotid body. In Handbook of Physiology. The Respiratory System II (ed. Fishman A. P.), pp. 243-312. Am. Physiol. Sot., Bethesda, MD. 21. Fidone S. J., Gonzalez C. and Yoshizaki K. (1982) Effects of low oxygen on the release of dopamine from the rabbit carotid body in vitro. J. Physiol. 333, 93-110. 22. Fleming R. M. and Clark W. G. (1972) Quantitative thin-layer chromatographic estimation of labelled dopamine and norepinephrine, their precursors and metabolites. J. Chromat. 52, 305-312. 23. Freedman S. B., Dawson G., Villereal M. L. and Miller R. J. (1984) Identification and characterization of voltage-sensitive calcium channels in neuronal clonal cell lines. J. Neurosci. 4, 1453-1467. 24. Gonzalez E., Rigual R., Fidone S. J. and Gonzalez C. (1987) Mechanisms for termination of the action of dopamine in carotid body chemoreceptors. J. auton. nerv. Syst. 18, 249-259. 25. Grinstein S. and Cohen S. (1987) Cytoplasmic [Ca*+] and intracellular pH in lymphocytes. Role of membrane potential and volume-activated Na+/H+ exchange. J. gen. Physiol. 89, 185-213. 26. Hescheler J. and Delpiano M. A. (1990) Ionic currents in carotid body type I cells and the effects of hypoxia and NaCN. In Arterial Chemoreception (eds Eyzaguirre C., Fidone S. J., Fitzgerald R. S., Lahiri S. and MacDonald D. M.), pp. 58862. Springer, New York. 27. Hille B. (1984) Ionic Channels of Excitable Membranes. Sinauer, Sunderland, MA. 28. Kidokoro Y. and Ritchie A. (1980) Chromaffin cell action potentials and their possible role in adrenaline secretion from rat adrenal medulla. J. Phystol. %7, 199-216. 29. Kileatrick D. L.. Slenetis R. J.. Corcoran J. J. and Kirshner N. (1982) Calcium uotake and catecholamine secretion by cultured bovine adrenal medulla cells. J. Neurochem. 38, 427231. ’ 30. Llinas R., Sugimori M. and Simon S. M. (1982) Transmission by presynaptic spike-like depolarization in the squid giant synapse. Proc. natn. Acad. Sci. U.S.A. 79, 2415-2419.