Cell Physiology

Retardation of cation channel deactivation by mitochondrial dysfunction in adrenal medullary cells

M. Inoue, N. Fujishiro, I. Imanaga


The mechanism for cyanide (CN) activation of a nonselective cation (NS) channel coupled with a muscarinic receptor in a guinea pig chromaffin cell was studied with the perforated-patch method. Bath application of a protein kinase inhibitor resulted in a dose-dependent inhibition of muscarine-induced current (I M) but had no apparent effect on the CN-induced current (I CN). On the other hand, production of I CN occluded muscarine activation of NS channels in an amplitude-dependent manner. Deactivation of I M after washout was retarded whileI CN was also active, and the extent of the retardation increased with an increase in the relative production ofI CN on muscarinic stimulation. Restoration of Na+ pump activity from CN suppression was conspicuously retarded below 19–20°C, and the apparent diminution ofI M and I CN after washout was retarded in parallel with a decrease in temperature. The results suggest that CN activation of NS channels is due to suppression of deactivation of the channel.

  • muscarinic receptor
  • mitochondria
  • temperature
  • cyanide

secretion of catecholamines in response to hypoxia is vital for the body to deal with the life-threatening event. Secreted catecholamines, especially adrenaline, increase cardiac output and enhance gluconeogenesis and glycogenolysis in the liver with the consequent increase in blood glucose (6). Although carotid body type I cells are a sensing site for O2 tension, recent studies demonstrated that adrenal medullary cells are also capable of promptly detecting a decrease in O2 tension with a subsequent secretion of catecholamines (14, 24). Accumulating evidence indicates that hypoxia produces a depolarization with the consequent activation of voltage-dependent Ca2+ channels and that the subsequent increase in intracellular Ca2+ concentration ([Ca2+]i) is responsible for catecholamine secretion in type I and chromaffin cells (2, 3, 14, 26). How hypoxia is detected is controversial, and two main hypotheses have been proposed. One is the membrane ion channel hypothesis that a K+ channel itself or its closely associated regulator in the membrane senses O2 levels (11, 22). Although K+ channel activity in isolated patch membranes was documented to be suppressed by hypoxia (11), it remains to be determined whether the hypoxia-sensitive K+ channel is active at resting membrane potentials and whether suppression of its activity is responsible for depolarization in response to hypoxia (2). The other is that mitochondria play a primary role for the detection (8, 9). This proposal is principally based on findings that the effects of hypoxia can be mimicked by various types of mitochondrial inhibitors.

Thompson and Nurse (25) reported that anoxia suppressed two distinct voltage-evoked K+ currents, Ca2+-dependent and delayed rectifier type, in adrenal medullary cells obtained from newborn rats. However, a depolarizing response or receptor potential to anoxia was not suppressed by bath addition of 10 mM tetraethylammonium, which completely suppressed the anoxia-sensitive K+currents. On the other hand, Mojet et al. (24) reported that the depolarization of mitochondrial membrane potential preceded an increase in [Ca2+]i in response to hypoxia in the rat chromaffin cell, and they proposed that mitochondria can serve as a site for detection of a decrease in O2. Consistent with this hypothesis, cyanide (CN) and anoxia induced activation of a nonselective cation (NS) channel and inhibition of the Na+ pump in guinea pig adrenal chromaffin cells (15). This cation channel may be the same as that activated by muscarinic receptor stimulation (19), because muscarine failed to induce a further inward current during the full production of a current in response to CN (14). If this notion is tenable, then anoxia and CN might activate the channel through phosphorylation because the muscarinic activation of channels and deactivation may be mediated by a protein kinase and a Mg2+-dependent phosphatase, respectively (16, 20). In the present experiment, we examined the mechanism for CN activation of NS channels. Our findings are consistent with the notion that exposure to CN diminishes the deactivation process with the consequent dominance of the activation.


Whole cell recordings.

Experiments on dissociated adrenal medullary cells were done, as described elsewhere (17). Briefly, female guinea pigs weighing 250–300 g were killed by a blow to the neck, and the adrenal glands were eliminated and immediately put into ice-cold Ca2+-free solution in which 1.8 mM Ca2+ was simply removed from a standard saline containing (mM): 137 NaCl, 5.4 KCl, 1.8 CaCl2, 0.5 MgCl2, 0.53 NaH2PO4, 5 d-glucose, 5 HEPES, and 4 NaOH. Adrenal medullas were cut into three to six pieces and incubated for 30 min with 0.25% collagenase dissolved in the Ca2+-free solution. After the incubation, the tissues were washed three times in the Ca2+-free solution and then kept in it at room temperature (23–25°C). A few pieces of the tissues were put in the bath apparatus, which was placed on an inverted microscope, and adrenal chromaffin cells were dissociated mechanically with fine needles. Then, dissociated cells were allowed to adhere to the bottom for a few minutes before the bath apparatus was perfused with saline at a rate of 1 ml/min. The whole cell current was recorded with the perforated-patch method (13). The current was recorded with an Axopatch 200A amplifier (Axon Instruments) and then fed into a brush recorder after low-pass filtering at 3 or 5 Hz and into a video tape after being digitized with an analog-to-digital converter. The pipette solution contained (mM): 120 potassium isethionate, 20 KCl, 10 NaCl, 10 HEPES, and 2.6 KOH. On the day of the experiment, nystatin dissolved in dimethyl sulfoxide (10 mg in 50 μl) was added to the pipette solution at a final concentration of 100 μg/ml. Glucose and NaCl in the standard saline were equimolarly replaced with sucrose and NaCN in a CN solution. The pH of the pipette solution and external solutions was adjusted to 7.2 and 7.4 with KOH and NaOH, respectively. All chemicals were bath applied, and a CN-induced current (I CN) or muscarine-induced current (I M) was evoked by perfusion with the CN solution or 3 μM muscarine-containing solution, unless otherwise noted. The membrane potential was corrected for a liquid junction potential of −12 mV between the pipette solution and the standard solution. Experiments except for those at low temperatures were carried out at 23–25°C. When bath temperature was lowered from room temperature, the perfusate was cooled with a Peltier device. Thus the temperature around the cell examined was expected to be lower than that at the outflow, where temperature was measured. In a separate experiment, the temperature at the center of bath was lower by ∼3°C than the 19.7°C at the outflow, by 1.6°C than the 20.8°C, and by 1.4°C than the 21.8°C. Data are expressed as means ± SD, and statistical significance was determined with Student's t-test.

Fluorescence recordings.

To label the cell surface with a fluorescence dye, we incubated dissociated cells for 30 min in a standard solution to which 5 μM di-8-ANEPPS and 0.05% Pluronic F-127 were added. The dish in which the cells settled was placed on a Zeiss Axiovert microscope attached to a Zeiss LSM 410 laser confocal scanning unit (Carl Zeiss). The objective lens was an oil-immersion lens with a magnification of ×63 and a numerical aperture of 1.25. Illumination with 488 nm was provided by an argon laser, and emission was monitored above 570 nm because the peak emission wavelength was reported to be 570–580 nm (4). The theoretical spatial resolution given by the equation d= λ/(2 × NA), where d represents the smallest resolvable distance, λ is the wavelength of emission light (∼580 nm), and NA is the numerical aperture (1.25), is ∼0.2 μm. Thus images and line-scan images were obtained in all experiments with a pixel size of <0.2 μm and with a full width at half-maximal intensity of ∼0.7 μm. The actual value of d may be twice or three times larger than the theoretical one because of diffraction (23). To study effects of CN, one-half of the 2-ml solution in the dish was replaced with the CN solution and the administration was completed within 8 s.


Nystatin, (±)-muscarine chloride, and Pluronic F-127 were obtained from Sigma; di-8-ANEPPS was from Molecular Probes;N-(2-guanidinoethyl)-5-isoquinolinesulfonamide hydrochloride (HA-1004) was from Seikagaku (Tokyo, Japan); collagenase was from Yakult (Tokyo, Japan); NaCN was from Hayashi Pure Chemical (Tokyo, Japan).


Retardation of IM deactivation by CN.

One of the findings for involvement of phosphorylation in the activation of NS channels is a reversible suppression of the muscarinic production of a nonselective cation current (I NS) by protein kinase inhibitors (16, 18). Consistent with the previous result, addition of 300 μM HA-1004, a general protein kinase inhibitor (12), to a 3-μM muscarine-containing perfusate resulted in a gradual decline of I M in a reversible manner, whereas the inhibitor failed to suppress I CN (Fig.1 A). In the previous experiment (15), 57 and 43% of I CN were estimated to be due to activation of NS channels and inhibition of the Na+ pump current (I pump), respectively.I CN in this particular cell, however, might have comprised only inhibition of the I pump, but this possibility may not be feasible because the subsequent application of 3 and 10 μM muscarine did not induce a further inward current, as was noted previously. In a total of eight cells, HA-1004 suppressedI M in a concentration-dependent manner but did not affect I CN (Fig. 1 B). This finding raises the possibility that exposure to CN activates NS channels independent of phosphorylation. One of such possibilities is that a change in membrane tension is responsible for the NS channel activation, because hypoxia and metabolic suppression were reported to increase volume in various types of cells (10, 21). Thus we investigated whether or not a short exposure to CN would induce any alteration in cell size. To this end, the cell surface was labeled with di-8-ANEPPS and fluorescence was observed in line-scan or z-axis images. Figure2 shows results of line-scan analysis. The line marked in the image (Fig. 2 A) was scanned every 0.1 s, and these scans were used to construct the line-scan image (Fig.2 B). Addition of 1 ml CN solution to 1 ml of dish solution (final CN concentration, 2.5 mM) did not induce any change in cell diameter for 200 s in nine of ten cells tested, as noted in one cell (Fig. 2, A and B, left). However, in another cell (Fig. 2, A and B, right), the diameter began to increase 26 s after the onset of CN exposure and the increase continued. Figure 2 C summarizes relative diameters of 10 cells during 2-min exposure to CN. It is evident that short exposure to CN did not alter cell size. Similarly, z-axis images were not altered during the CN exposure (not shown). The results indicate that short application of CN does not induce an increase in diameter at least in the order of ∼0.5 μm.

Fig. 1.

Failure of protein kinase inhibitor to suppress cyanide (CN)-induced current. Values are means ± SD of 2–5 observations. A: chart record of a 3 μM muscarine (M)-induced current and a 5 mM CN-induced current (I M and I CN, respectively). Whole cell current was recorded at a holding potential of −67 mV with perforated-patch method. Here and in subsequent figures, horizontal arrows are zero-current level. M, CN, and HA-1004 (HA: protein kinase inhibitor) were bath applied during the period indicated by the bar. Nos. are μM concentrations used. B: summary of relative amplitudes of I M andI CN in presence of 100, 150, and 300 μM HA. Amplitudes are expressed as fractions of currents in absence of HA in same cell.

Fig. 2.

No change in cell volume during short exposure to CN. Values are means ± SD of 10 cells. A: fluorescence image of chromaffin cells obtained in confocal mode. Image was obtained approximately in middle of cells. Straight line is a site for scanning.B: line-scan image. Line scan was performed every 100 ms. Intensity of fluorescence was indicated by heights. Arrows, onset of addition of CN solution. C: summary of relative diameters of cells during CN exposure. Relative diameters are expressed as fractions of those before exposure.

If phosphorylation is indeed involved, then the activation of channels might be attributed to a decrease in phosphatase activity with the consequent dominance of kinase activity. If this is the case, thenI M should diminish slowly after washout of muscarine while I CN is also active. This inference was examined by applying muscarine at a maximum of the current evoked by 0.5 mM CN; thus I M was expected to be reversible under such conditions because the amplitude of the fully developedI CN was about one-half of I Mand activation of the channel may not have been saturated (14). Figure3 A shows that deactivation ofI M after washout was considerably retarded inI CN production, an effect that rapidly disappeared after washout of CN. To elucidate the relation between retardation ofI M deactivation and I CNproduction, half-decay times of I M in the production of I CN were expressed as a fraction of those of control I M in the same cells and the ratio was plotted against the relative amplitude of I CN, which had developed on muscarinic stimulation. The relative amplitude of I CN was expressed as a fraction ofI M evoked in the absence of CN. Figure 3 Bshows that as the relative amplitude of I CNincreased, deactivation of I M slowed. Furthermore, the amplitude of I M evoked in the presence of CN diminished with an increase in the relative amplitude ofI CN (not shown).

Fig. 3.

Retardation of I M deactivation inI CN production. A: chart records of whole cell currents. Current was recorded at −62 mV with perforated-patch method. CN and muscarine were bath applied during indicated period (double line for CN and bars for M). B: relative amplitudes of I CN are plotted against ratios of half-decay times for I M in presence of CN to those for I M in absence of CN. Amplitudes ofI CN, which were evoked in 2 cells by 0.1 mM CN, in 6 by 0.25, in 2 by 0.5, and in 2 by 5, were measured on application of muscarine, and values were expressed as fractions ofI M elicited in absence of CN.

Effects of low temperature.

The foregoing observations suggest that suppression of mitochondrial function retards the deactivation of I M. We then asked if low temperatures have a similar retarding action onI M, because mitochondrial F1-ATPase activity diminished markedly below ∼20°C (1, 7). First, to determine how low temperature affects mitochondrial function, restoration of the Na+ pump activity from CN inhibition was examined. The I CN elicited in the presence of 6 mM Ba2+ was in a major part attributed to inhibition ofI pump, and suppression of the pump activity by CN was thought to be due to a decrease in intracellular ATP contents (15). Thus, if ATP production in mitochondria is indeed impaired by a decrease in temperature, then restoration of the pump activity should be slow. Figure 4 shows that this is the case. As the temperature in the outflowing perfusate increased from 19.6°C (actual temperature around the cell, ∼16.6°C), restoration of the pump activity became rapid (Fig. 4 A). This rapid event could not be ascribed to a general effect of temperature on kinetics. The time required for half-suppression ofI pump after CN administration increased only slightly with decreasing temperature, whereas the time for half-restoration after washout increased steeply (Fig.4 B). Furthermore, the temperature dependence of half-times for production and diminution of I CN in the absence of Ba2+ did not apparently differ from those for I CN in Ba2+ (Fig. 4 B). In Fig. 4 C, amplitudes of I CN in the presence and the absence of 6 mM Ba2+ were plotted against temperature. The apparent x-intercepts forI CN with and without Ba2+ were 18.7 and 17.8°C, respectively. Because restoration of the pump activity is likely to reflect that of mitochondrial functions, a decrease in temperature below 21°C (actual temperature, ∼19.5°C) may impair the restoration.

Fig. 4.

Retardation of restoration of Na+ pump activity from CN inhibition at low temperatures. A: chart records ofI CN in presence (Ba) and absence of 6 mM Ba2+. Whole cell current was recorded at −65 mV with perforated-patch method. Nos. beside traces are temperature in outflowing perfusate. CN was applied during indicated period (bar). Sequence of application was from top to bottom. B: half-times for production (• and ▴) and diminution (□ and ▵) of I CN are plotted against temperatures. Half-times are times required for half-production or-diminution ofI CN. Triangles and squares meanI CN in presence and absence of 6 mM Ba2+, respectively. I CN in presence of Ba2+ was mainly due to suppression of Na+ pump current. C: amplitudes of I CN in presence (▵) and absence (□) of Ba2+ are plotted against temperatures.

Whether or not deactivation of I M is retarded at low temperatures was then examined. I M was successively evoked as the temperature in the outflow decreased from 28.5 to 15.7°C and then increased up to 29°C (Fig. 5, BD). It is evident that the half-decay time markedly increased with decreasing temperature below 21°C (Fig. 5 C), whereas the half-rise time only slightly increased. A similar difference in temperature dependence of the half-rise time and the half-decay time (Fig 5, B and C) of I CN was noted below 21°C. The half-rise times of the first two I CNs elicited at 28.5 and 22.7°C were apparently smaller; these results, however, may not indicate the marked temperature dependence of the half-rise time, because results of I CN elicited at 22.2 and 26.4°C in the increasing phase of temperature did not differ from those of I CN below 21°C. The time course of the developing process of I CN sometimes slowed with time of the recording and then became stable. A comparison of temperature dependence of the decay time for I CNwith that for I M reveals that the decay process ofI M and I CN had the same temperature dependence. In Fig. 5 D, amplitudes ofI M and I CN are plotted against temperature. The regression line for I CN had a steeper slope than that for I M, and the apparentx-intercepts for I M andI CN were 9 and 13°C, respectively.

Fig. 5.

Parallel retardation of diminution of I M andI CN at low temperatures. A: chart records of I M and I CN at various temperatures. Whole cell currents were recorded at −62 mV with perforated-patch method. M and CN were bath applied during indicated period (double line for CN and bar for M). No. beside each trace is temperature in outflow. B and C: half-rise times and half-decay times of I M (○) andI CN (▵) after washout are plotted against temperatures, respectively. D: amplitudes ofI M (○) and I CN (▵) are plotted against temperatures. Open and closed symbols correspond to decreasing and increasing phases of temperatures, respectively. Note that a temperature decrease from 22.7 to 15.7°C reversibly induced an inward shift of holding current by 0.9 pA.

To facilitate analysis of temperature effects in different cells, slopes and x-intercepts for I CN in the presence of 6 mM Ba2+ and for I M were expressed as fractions of those for I CN in its absence in the same cells. The relative x-intercept ofI CN in the presence of Ba2+ and that ofI M thus obtained significantly differed from one, and similarly, the relative slope of I M was noticeably different from one (Fig. 6,A and B). These results indicate thatI pump inhibition and I NS, which constitute I CN, have a different temperature dependence. On the other hand, ratios of half-decay times at 20°C (actual temperature, ∼17°C) to those at 24°C forI CN in the presence of Ba2+ and forI M did not differ appreciably, compared with those for I CN in its absence, thereby suggesting the involvement of a common mechanism for the retardation of both restoration of pump activity and diminution of I NS.

Fig. 6.

Temperature dependence of I CN in absence and presence of Ba2+ and I M. Slopes of temperature-amplitude relations were calculated with a least-squares method. Relative slopes and x-intercepts ofI CN in presence of Ba2+ (Ba-CN) and ofI M (M) are expressed as fractions ofI CN in its absence in same cells. Half-decay times of I CN in presence and absence of Ba2+and I M after washout were measured at 24 and 20°C (actual temperature, ∼18°C). Values at 20°C are expressed as fractions of those at 24°C, and then ratios forI CN in presence of Ba2+ and forI M are calculated as fractions of those forI CN in its absence. Asterisks are statistical significance (P < 0.05). Values are means ± SD of 3 and 5 cells for Ba-CN and M, respectively.


Our previous study (15) suggested that 57 and 43% of 5 mM CN-induced inward currents can be attributed to activation of NS channels and inhibition of the I pump, respectively. Because amplitude of I NS diminished progressively in the absence of Na+ pump activity and this decrease was enhanced by replacement of sucrose with glucose, i.e., by glycolytic production of ATP, activation of the channel was presumed to be due to an ATP decrease that resulted from consumption by energy-dependent processes, such as Na+-K+-ATPase. Thus, to obtain a maximum stimulation of NS channels, CN activation of the channel was observed as I CN without isolation of theI NS. In the present experiment, the notion thatI CN comprises production of I NSand inhibition of the Na+ pump was further supported by a difference in temperature dependence of I M andI pump inhibition. The relative values ofx-intercept for I M andI pump inhibition significantly differ from one. Similarly, I M production andI pump inhibition had relative slopes of 0.82 and 1.05, respectively, and the former appreciably differed from one. Thus temperature dependence of I CN is thought to reflect a combination of those related to I M production andI pump inhibition.

In carotid body type I cells, a decrease in O2 tension was proposed to be sensed by an O2-sensitive K+channel or its closely associated regulator (11, 22) or by mitochondria (8, 9). Similarly, these two mechanisms were suggested to be involved in O2 sensing in adrenal medullary cells (14, 24, 25). The findings that CN-induced activation of NS channels and inhibition of the Na+ pump were reproduced by exposure to anoxia are consistent with the mitochondria hypothesis. The rapid secretion of catecholamines by hypoxia would be accounted for readily by the membrane ion channel hypothesis, whereas how the dysfunction of mitochondria promptly induces a change in membrane excitability would be a challenging issue for the mitochondrial hypothesis. The sequence of our experiments revealed that dysfunction of mitochondria indeed induces a rapid depolarization through inhibition of the Na+ pump and activation of NS channels. In particular, anoxia or CN activation of NS channels is estimated to correspond to ∼60% of anoxia- or CN-induced currents and exposure to anoxia or CN even in the absence of Na+ pump activity results in activation of the channel. Thus mechanisms for stimulation of NS channels by anoxia or chemical hypoxia seem to develop specially for transduction of O2 signal to catecholamine secretion. The present results suggest that CN activation of NS channels is due to suppression of the deactivation process for the channel. First, the diminution of I M after washout was retarded inI CN generation, and the degree of retardation depended on the relative production of I CN on muscarinic stimulation. As the relative amplitude ofI CN compared with that of I Mincreased, I M diminution was even more retarded. This close correlation between relative production ofI CN and retardation of I M decay indicates that the deactivation process for the NS channel diminishes in generation of I CN. Because the biophysical properties of CN or anoxia-sensitive channels resemble those of the muscarinic one and the muscarinic stimulation of the channel was occluded in an amplitude-dependent manner by generation ofI CN (Figs. 1 and 3; Ref. 13), there would be no doubt that exposure to CN or hypoxia activates the same NS channel as that regulated by the muscarinic receptor. Thus it is likely that diminished deactivation process for NS channels is responsible for CN activation of the channel. The decrease in rate constant for deactivation is expected to shift the equilibrium between activation and deactivation toward the former with the consequent production ofI NS. Secondly, the failure of HA-1004 to suppressI CN is consistent with our hypothesis. The muscarinic activation of NS channels was reversibly suppressed by various isoquinoline sulfonamide derivatives with different potencies (16, 18), and this inhibition was assumed to be due to inhibition of the protein kinase involved and the consequent shift of equilibrium between phosphorylation and dephosphorylation toward the latter (20). The fact that HA-1004 failed to suppress I CNsuggests that the rate constant for deactivation decreased almost to null during exposure to 5 mM CN. Thirdly, deactivation ofI M was retarded at low temperatures. This low-temperature effect is probably due to the dysfunction of mitochondria, because restoration of the Na+ pump activity from CN inhibition was retarded below 21°C (actual temperature, ∼19.5°C). Because the CN inhibition of the pump activity is probably attributed to a decrease in ATP contents, mitochondrial production of ATP in chromaffin cells may have a steep temperature dependence below 19–20°C. This temperature dependence of ATP production might be ascribed to that of ATP synthase in the terminal step of oxidative phosphorylation, because temperature dependence of F1-ATPase activity obtained from heart mitochondria was biphasic with a steep decrease below 20°C (1, 7). Because it took a few minutes to change the bath temperature in 1°C and thus studying quantitative effects of low temperatures on the whole cell current was difficult, we did not consistently investigate whether or not a decrease in temperature induces an inward current. In one cell shown in Fig. 4, decreasing the bath temperature to 15.7°C reversibly induced an inward current of 0.9 pA. The fact that metabolic inhibition with CN and low temperatures induce a similar retardation ofI M deactivation indicates that the deactivation process of NS channels is closely associated with the mitochondrial function. Based on three lines of evidence, it would be rational to assume that exposure to CN or anoxia activates NS channels through suppression of the deactivation process. Our previous studies (16, 20) suggest that a Mg2+-dependent phosphatase is responsible for deactivation of the NS channel. Thus our notion means that exposure to CN results in a decrease in an apparent activity of the Mg2+-dependent phosphatase. This thesis was not examined directly in the present experiment, because the Mg2+-dependent phosphatase has not been identified molecularly and there is no specific inhibitor for protein phosphatase IIC (5), a candidate for the phosphatase involved. The best means to inhibit the phosphatase activity is removal of Mg2+. In preliminary experiments, we found that the CN potency to induce an inward current was almost abolished by decreasing concentrations of free Mg2+ inside the cell.


  • Address for reprint requests and other correspondence: M. Inoue, Department of Physiology, School of Medicine, Fukuoka University, Fukuoka 814–0180, Japan (E-mail:minoue{at}fukuoka-u.ac.jp).

  • The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. §1734 solely to indicate this fact.


View Abstract