O2 Modulates Large-Conductance Ca2+-Dependent K+ Channels of Rat Chemoreceptor Cells by a Membrane-Restricted and CO-Sensitive Mechanism
Hypoxic inhibition of large-conductance Ca2+-dependent K+ channels (maxiK) of rat carotid body type I cells is a well-established fact. However, the molecular mechanisms of such inhibition and the role of these channels in the process of hypoxic transduction remain unclear. We have examined the mechanisms of interaction of O2 with maxiK channels exploring the effect of hypoxia on maxiK currents recorded with the whole-cell and the inside-out configuration of the patch-clamp technique. Hypoxia inhibits channel activity both in whole-cell and in excised membrane patches. This effect is strongly voltage- and Ca2+-dependent, being maximal at low [Ca2+] and low membrane potential. The analysis of single-channel kinetics reveals a gating scheme comprising three open and five closed states. Hypoxia inhibits channel activity increasing the time the channel spends in the longest closed states, an effect that could be explained by a decrease in the Ca2+ sensitivity of those closed states. Reducing maxiK channels with dithiothreitol (DTT) increases channel open probability, whereas oxidizing the channels with 2,2′-dithiopyridine (DTDP) has the opposite effect. These results suggest that hypoxic inhibition is not related with a reduction of channel thiol groups. However, CO, a competitive inhibitor of O2 binding to hemoproteins, fully reverts hypoxic inhibition, both at the whole-cell and the single-channel level. We conclude that O2 interaction with maxiK channels does not require cytoplasmic mediators. Such interaction could be mediated by a membrane hemoprotein that, as an O2 sensor, would modulate channel activity.
Type I cells of the rat carotid body (CB) are chemoreceptors that sense changes in blood pH, CO2, and O2. These cells respond to stimulation with a membrane depolarization that induces Ca2+ entry through voltage-dependent Ca2+ channels and the subsequent release of a neurotransmitter that activates the afferent fibers of the carotid sinus nerve.1 In the case of hypoxia, the cause of membrane depolarization seems to be the inhibition of K+ channels. Three different types of O2-sensitive K+ channels have been postulated to elicit such depolarization: a voltage-gated channel in rabbit type I cells2 and a large-conductance Ca2+-dependent K+ channel (maxiK)3 and a TASK-like channel4 in rat type I cells.
Although the role of K+ channels in the hypoxic transduction in type I cells is generally accepted, mechanisms involved in the process of O2 sensing remains unclear.5 The prevalent theory to explain hypoxic modulation of K+ channels links O2-sensing to the process of O2 reduction with some electron donor such as NADPH and postulates that channel activity is modulated by changes in GSH/GSSG or some still undefined reactive oxygen species (ROS).6 In fact, oxidizing and reducing agents have been shown to alter channel activity,7–11 even in the O2-sensing channels in type I cells.5 In the case of maxiK channels, reducing agents have been reported to increase,9,10 decrease,8 or not having effect12 on channel activity. However, although hypoxia and reducing agents could have similar effects on channel activity, they cannot be considered equivalent.11
An alternative hypothesis to explain hypoxic modulation of ion channels proposes the existence of a hemoproteic sensor located in the plasma membrane that modulates channel activity in the absence of soluble factors. Evidence supporting this model has been obtained mainly in rabbit type I cells, where hypoxic modulation of a K+ current remains in isolated membrane patches2 and can be reverted by CO,13 a well-known ligand of hemoproteins. However, there are reports in neonatal rat chemoreceptor cells suggesting that the effect of hypoxia could require some cytosolic factor,3,4 although the hypothesis of the hemoproteic sensor has not been directly tested.
In the present study, we investigate the modulation of rat CB maxiK channels by O2 at the single-channel level. We demonstrate that channel activity is markedly affected by changes in O2 concentration, both in whole-cell and inside-out membrane-excised patches, suggesting that O2 modulation is a membrane-restricted event. The voltage and Ca2+ dependence of O2 modulation of maxiK channels is compatible with a physiological role of hypoxic inhibition, because the effect was maximal at threshold levels for channel activation. Finally, we show that two redox agents, dithiothreitol (DTT) and 2,2′-dithiopyridine (DTDP), modulate maxiK activity. However, the effect of hypoxia is mimicked by the oxidant, which is what makes the redox hypothesis very unlikely. On the contrary, CO completely reverts the effect of low O2, as predicted by the hemoproteic sensor hypothesis.
Materials and Methods
Cell Isolation and Culture
Experiments were performed on cultured rat CB chemoreceptor cells. Adult Wistar rats (3 to 4 months old) were obtained from the animal care unit of the University of Valladolid. Rats were anesthetized with pentobarbital sodium (100 mg · kg−1 IP). After tracheostomy, the carotid artery bifurcations were removed, and the animals were killed by an intracardiac bolus injection of sodium pentobarbital. All measures were taken to ensure the animals did not suffer distress at any time. Protocols were approved by the Institutional Animal Care and Use Committee of the University of Valladolid. The CBs were cleaned of surrounding connective tissue and enzymatically dispersed as described elsewhere.14 Dispersed cells were plated onto small poly-l-lysine–coated coverslips and maintained in culture for up to 48 hours.
Ionic currents were recorded at room temperature (20°C to 25°C) using the whole-cell and inside-out modes of the patch-clamp techniques. Whole-cell currents were recorded and acquired as previously described.14 Borosilicate glass patch pipettes used for single-channel recordings (2 mm, World Precision Instruments) were double-pulled (Narishige PP-83) and heat-polished (Narishige MF-83) to resistances between 10 and 15 MΩ when filled with the internal solution. Recordings were made with an Axopatch-200A patch-clamp amplifier and a Digidata 1200 A/D interface, driven by CLAMPEX version 8.0 software (Axon Instruments). Single-channel recordings were filtered at 1 to 5 kHz and digitized at 20 kHz.
Analysis of the data were performed with pCLAMP software. Single-channel amplitudes and open probabilities were measured from amplitude histograms generated with FETCHAN. Open probabilities were expressed as NPo, where N represents the number of single channels present in the patch, and Po is the open probability of a single channel. NPo was calculated using the following expression:equation
where A0 is the area under the curve of the amplitude histogram corresponding to the closed state, and A1–An represent the histogram areas reflecting the different open-state current levels for 1 to n channels present in the patch. Histogram parameters were obtained from multiple least-squares gaussian fits of the data using ORIGIN 4.0 software (Microcal Inc).
Single-channel kinetic analysis was performed using the QuB Suite package (New York State University, Buffalo, NY). Single-channel traces were idealized with SKM, and model fitting to data were carried out with MIL.15,16
Pooled data are expressed as mean±SEM. Statistical comparisons were performed with the two-tailed t test for paired or unpaired data as appropriate. Values of P<0.05 were considered statistically different.
The composition (in mmol/L) of the bathing solution used for gigaseal forming and whole-cell recordings was NaCl 141, KCl 4.7, MgCl2 1.2, CaCl2 1.8, glucose 10, and HEPES 10 (adjusted to pH 7.4 with NaOH). Whole-cell experiments were performed with a pipette solution containing (in mmol/L) KCl 125, MgCl2 4, EGTA 10, MgATP 5, NaGTP 5, and HEPES 10 (adjusted to pH 7.2 with KOH). Inside-out recordings were carried out with a pipette solution containing (in mmol/L) KCl 125, EGTA 5, and HEPES 10 (adjusted to pH 7.2 with KOH). After patch excision, the bath solution was switched to a solution of identical composition to that of the pipette but containing different amounts of CaCl2 to get free [Ca2+] of 1, 5.5, or 10 μmol/L. To obtain 100 μmol/L free Ca2+, HEDTA instead of EGTA was used as Ca2+ buffer. Total Ca2+ was calculated using the program CHELATOR.17 Titrated stock solutions of CaCl2 and EGTA (HEDTA) were used to minimize errors due to impurities of EGTA (HEDTA) and hydration of CaCl2.
Bathing solutions were perfused by gravity into the recording chamber at 1 to 2 mL/min. Control conditions are defined as the O2 level present in solutions equilibrated with air (150 mm Hg). Hypoxia was achieved by bubbling the bath solution with 100% N2, giving a Po2 in the recording chamber of 9.5±1.2 mm Hg (n=5), as determined with an O2 microelectrode (Diamond General Corp). O2 electrode was calibrated using solutions equilibrated with air (150 mm Hg) or containing 1 mmol/L sodium dithionite (Na2S2O4) as an O2 scavenger (0 mm Hg). When CO was used, the bathing solution was bubbled with a mixture of 80% N2 and 20% CO, that gives a Po2 of 8.7±2.3 mm Hg (n=5). In selected experiments, control solution was also bubbled with air to avoid possible artifacts introduced in the record by bubbling. No differences were found between these two control groups.
O2 Sensitivity of MaxiK Channels in Isolated Patches
MaxiK channel activity was recorded from inside-out patches. Patches with several (typically 2 to 5) large conductance channels (186±12 pS, in symmetrical K+, n=21) were routinely obtained. In most cases, the activity of these channels was completely abolished on removal of Ca2+ from the bath solution. Patches with channels of smaller conductance and/or Ca2+-independent were not considered for analysis. A typical experiment is shown in Figure 1. Figure 1A represents a diary plot of channel activity in the patch. This plot was built by computing the NPo in 5-second intervals and representing the obtained value against recording time. The activity of three channels was evident in control conditions, as shown in the sample current trace (Figure 1A, 1). When hypoxia was applied, channel activity decreased, and only two channels opened simultaneously (Figure 1A, 2). The channel activity recovered after returning to control solution (Figure 1A, 3). Removing Ca2+ from the bath solution completely abolished channel activity in the patch. The all-point histograms corresponding to periods of 2 minutes in control (1), hypoxia (2), and recovery (3) conditions are shown in Figure 1B. Average NPo calculated from these histograms were 0.73, 0.32, and 0.75, respectively. Similar results have been obtained in 49 patches of 65 tested.
Voltage and Ca2+ Dependence of the Effect of Hypoxia
Figure 2 shows the effect of membrane potential on the open probability of a single maxiK channel when [Ca2+]i was kept at 10 μmol/L. The increase of open probability on depolarization is well described by a Boltzmann equation (Figure 2B, filled squares). Hypoxia (empty circles) decreased the open probability of the channel by shifting the voltage curve 24.6 mV to the right. As expected from the Boltzmann functions, when expressed as percentage of inhibition of open probability (open triangles), the effect of hypoxia is maximal at lower potentials (around the activation threshold) and decreases sharply with depolarization. Similar results were obtained in two additional patches (average shift: 26.8±5.1 mV).
The effect of hypoxia on maxiK channels was also investigated at different [Ca2+] (Figure 3). Figure 3A shows the all-point histograms obtained in a single-channel patch with a holding potential of +10 mV at three different [Ca2+], both in control and hypoxic conditions. Insets show sample current traces obtained in each situation. Channel activity increased by raising [Ca2+] (top to bottom), but this condition concurrently decreased the effect of hypoxia. At 1 μmol/L Ca2+ hypoxia inhibited the channel activity by ≈70%, but this effect was ≈46% at 5.5 μmol/L and only 3% at 100 μmol/L. The same three [Ca2+] were also tested in this channel at −20 and +30 mV (Figure 3B). Patch depolarization decreased the low O2-induced inhibition, as expected from the above-described voltage dependence of this effect. The effect of hypoxia was tested at different voltages and [Ca2+] in 9 additional patches, although not all voltages and [Ca2+] were checked in every experiment. Average data, expressed as percentage of inhibition of NPo, are represented against membrane potential at three different [Ca2+] in Figure 3C. Hypoxic inhibition was maximal at lower potentials and lower Ca2+ concentrations.
Effects of Low Po2 on Channel Kinetics
The effect of hypoxia on channel kinetics was analyzed in detail in 4 single-channel patches. The Po measured in one of those patches is shown in Figure 4A. The kinetic behavior of this channel was investigated computing the stability plots depicted in Figure 4B. These plots represent the mean duration of groups of 50 sequential open and closed intervals plotted against the sequential interval number. Inspection of these plots reveals that the kinetic behavior of the channel remains constant over time, and that lowering the Po2 increases the duration of the closed intervals. This effect becomes more evident when the dwell-time histograms corresponding to control and hypoxic conditions are represented superimposed (Figure 5). Whereas the open-time histograms are almost identical, the closed-time histograms markedly differ in the number of counts corresponding to the longer closed durations.
Open-time histograms were properly described in all single-channel patches by the sum of 3 exponential functions, whereas closed-time histograms required 4 or 5. The number of states suggested by this analysis is compatible with the gating scheme proposed by McManus and Magleby18 to describe the kinetics of maxiK channels from rat skeletal muscle.equation
To test if this model also fits the data obtained in the present work, we used the program MIL, which computes the likelihood of the experimental sequence of open and closed times given a set of trial rate constants and changes the rate constants to maximize the likelihood. Proper constraints were introduced during the fitting procedures to ensure the microscopic reversibility of the loops in the model. The program also allows to fit simultaneously sets of data obtained under different experimental conditions. Therefore, data under control and hypoxic conditions were fitted simultaneously, making O2-sensitive different rate constants of the scheme. Because the effect of hypoxia was in all cases an increase in the duration of the longest closed states without changing the open states, only the forward transitions between the closed states were considered O2-dependent during the fit. In the experiment shown in Figure 4, the maximum likelihood was obtained when the forward rate constants of the transitions between C8–C7 and C7–C6 were made O2-dependent. The best fit was obtained when the effect of hypoxia on these rate constants was to decrease their value by a factor (h) of 2.8. The dwell-time histograms and the probability-density functions derived from the fitted parameters are shown in Figure 5. Results obtained in another two channels were also well fitted by this model, with h averaging 3.5±0.85 (n=3).
Effect of Redox Agents on Channel Activity
One possible mechanism to explain the effect of hypoxia could be the redox modulation of the cysteine sulfhydryl residues of the channel. In that case, the reducing environment created by hypoxia should be mimicked by a reducing agent such as DTT and blocked by an oxidant such as DTDP. The effect of DTT on channel activity was explored in five different inside-out patches containing 2 to 4 maxiK channels. Figure 6A shows representative current traces corresponding to one of those patches. DTT increased the activity of the channels in the patch; this effect was reversible on returning to control solution. The effect of DTDP has been also explored in 5 additional patches. Average results obtained at different holding potentials and expressed as percentage of NPo in control conditions are also depicted in Figure 6B.
The interaction between redox and O2 modulation has been further investigated using the whole-cell configuration of the patch-clamp technique. Cells were voltage-clamped at −80 mV, and depolarizing pulses to +20 mV were applied at 0.05 Hz. Figure 7A shows the plot of peak current versus time of pulse application obtained in a typical experiment. Peak current slowly runs down along the experiment, because of the washout of Ca2+ channels. The contribution of maxiK channels to the whole K+ current can be estimated by inhibiting Ca2+ entry during the pulse with Cd2+ at the end of the experiment. Both hypoxia and DTDP produced an inhibition of K+ current, as expected from the single-channel data shown in the previous figures. Interestingly, the effect of hypoxia was not affected by the presence of DTDP, suggesting different mechanisms for both stimuli. The effect of DTDP was almost irreversible (data not shown), but fully reverted with the application of DTT. Average results obtained in 5 different cells are shown in Figure 7B. DTT data correspond to 4 cells where this agent was applied without previous application of DTDP.
Reversion of Hypoxic Inhibition by CO
Alternatively to the redox mechanism, O2 could interact with a sensor in the plasma membrane just as a ligand, so that its degree of saturation would modulate channel activity. In this hypothesis, the effect of hypoxia should revert after saturating the sensor with an alternative ligand. If the sensor were a hemoprotein, a putative ligand would be CO. Therefore, we have investigated the effect of CO on the inhibition of channel activity by hypoxia. First, we have studied the effect of CO using a protocol similar to that in Figure 7A. Figure 7C shows a typical experiment. When hypoxia was applied to the bath solution, peak current decreased by ≈30%, and this inhibition was almost fully reverted adding 20% CO to the hypoxic solution. When the effect of hypoxia is measured on the Cd2+-sensitive current, inhibition increased to ≈50%. Average results obtained in 15 different cells are shown in Figure 7D.
The effect of CO on the hypoxic response was also tested in isolated patches in the inside-out configuration. Figure 8A shows the NPo diary plot of a patch containing at least 3 maxiK channels. In this experiment, hypoxia decreased NPo by ≈50%, and adding CO to the hypoxic solution not only reverted the effect of hypoxia but increased NPo to ≈150%. Current traces corresponding to control conditions, hypoxia, and hypoxia plus CO are depicted below the diary plot. Similar experiments were performed in 4 additional patches at the same [Ca2+] and at 3 different potentials. In all cases, CO reverted the effect of hypoxia. Average results are shown in Figure 8B.
The present study demonstrates a membrane-delimited mechanism for hypoxic modulation of maxiK channels of rat CB type I cells. The [Ca2+] and voltage dependence of this effect are consistent with a physiological role for maxiK channels in the process of hypoxic transduction. Furthermore, our results provide some insights into the mechanisms of O2 modulation of maxiK channels, suggesting the existence of a hemoproteic sensor located in the plasma membrane.
O2 modulation has been described for many different ion channels, and not only in classical chemoreceptor cells.5,19 However, very little is known about the molecular mechanisms of such modulation. Data presented in this study clearly show that interaction of O2 and maxiK channels can occur through a membrane-delimited mechanism, because hypoxic modulation occurs in excised membrane patches, in the absence of cytoplasmic factors. Although we cannot exclude the involvement of soluble factors in the modulation of the response to hypoxia, certainly such factors are not a requirement sine qua non for the response. Membrane-delimited mechanisms for O2 modulation of different K+ channels have been proposed both in native cells2,20–22 and in heterologous expression systems.11,22 Cloned human brain maxiK channels stably expressed in HEK293 have also been reported to be modulated by hypoxia through some direct mechanism.23 However, the requirement of cytosolic factors has been postulated in the case of maxiK channels of neocortical neurons from young mice24 or neonatal rat type I cells.3 Although we do not have an explanation for the discrepancy between the latter findings and ours, the fact that the studies of Wyatt and Peers3 were carried out in neonatal cell preparations makes conceivable the hypothesis that the mechanism of O2 detection can change during postnatal development.
The effect of [Ca2+] and voltage on maxiK kinetics has been extensively studied in different preparations,18,25 and a kinetic scheme comprising 3 open and 5 closed states has been proved to explain many properties of maxiK channels, although more complicated extensions of the model are needed to fully characterize channel properties.26 In spite of some discrepancies, Ca2+ and voltage seem to activate maxiK channels increasing mainly the forward rate constants between closed (C-C) and open (O-O) states. Although we have not performed a detailed study of the Ca2+ and voltage dependence of type I cell maxiK channels, our results are consistent with such effects (data not shown). Hypoxia increases longer closed durations without changing significantly the shape of the open-dwell time histograms. Taking into account the pronounced influence of [Ca2+] on the hypoxic modulation of the channel, O2 could affect the Ca2+ sensitivity of maxiK channels. If we assume that the forward rate constants between closed (C-C) and open (O-O) states are Ca2+-dependent, as reported for other preparations, hypoxia would be modifying the rate constants of forward transitions between the longest closed states, C8, C7, and C6. This assumption provides a reasonable fit of the data when activity of the channel in control and hypoxic conditions is fitted simultaneously. The model also predicts a larger effect of hypoxia at lower [Ca2+] and lower potentials, when the channel spends more time in the closed states. To date, apart of this study, the only attempt of investigating the effects of lowering O2 in the gating properties of a K+ channel has been the work of Ganfornina and López-Barneo2 on rabbit chemoreceptor cells. In spite of differences in the nature of the K+ channels in both preparations (maxiK versus voltage-dependent), it is remarkable that in both cases hypoxia modulates open probability, stabilizing the closed sates of the channels.
O2 could interact directly with the channel molecule or, alternatively, with a putative O2 sensor that would modulate channel activity. K+ channels have cysteine and methionine residues susceptible of redox modulation,27 and that property has been related in some preparations with their hypoxic sensitivity. In this hypothesis, O2 sensing is linked to the process of O2 reduction with some electron donor. Changes in ROS production would modulate channel activity, directly or through changes in GSH/GSSG. Our results are clearly not compatible with such a mechanism. The effect of O2 in excised inside-out patches excludes both the requirement of an electron donor and the possible role of a soluble molecule. Furthermore, the reversibility of the effect under these cell-free conditions rules out the reduction by hypoxia of methionine residues of the channel as a possible mechanism.27 In addition, application of the reducing agent DTT produces an increase of maxiK channel activity, an effect exactly opposite to that of hypoxia. Furthermore, application of the oxidant agent DTDP decreases channel activity, although the effect of hypoxia is not affected by the presence of the oxidant. Moreover, whereas the redox sate of the channels in whole-cell and excised patches varies,10 as evidenced by the effects of DTT and DTDP, hypoxic inhibition remains essentially unchanged in the two environments. Taken together, the data indicate that reduction of channel cysteine or methionine residues cannot account for the effect of hypoxia. By contrast, data provided in this study suggest the presence in the plasma membrane of a hemoproteic sensor, capable of binding both O2 and CO. This sensor would inhibit channel activity when desaturated. As predicted, CO reverts the hypoxic inhibition of maxiK channel activity. In fact, this activity is slightly larger when solutions are equilibrated with 20% CO than with air (20% O2), suggesting that the sensor could have a higher affinity for CO. A hemoproteic sensor has already been postulated to explain hypoxic transduction in rabbit chemoreceptor cells,13 pulmonary artery vasoconstriction,28 or the effect of hypoxia on cloned channels expressed in HEK293.11
The role of maxiK channels in type I cells of rat CBs in response to hypoxia is not clear in the literature.3,29,30 In order to participate in the depolarization triggered by hypoxia, maxiK channels should contribute to the resting membrane potential; if that were not the case, these channels would just act as negative-feedback regulators of membrane potential and [Ca2+]i changes elicited by activation of chemoreceptor cells, and O2 modulation of maxiK channels would contribute to control such regulation. In the present study, we do not attempt to solve this controversy, although data regarding the voltage and Ca2+ dependence of hypoxic modulation of maxiK channels can provide some insights in both possible scenarios. Hypoxia produces a shift to the right of the voltage dependence curve of maxiK channels, leading to an inhibition of channel activity at lower potentials. If some maxiK channels are open at resting potentials, hypoxia could lead to membrane depolarization. If they are not open, hypoxic membrane depolarization should be just a consequence of TASK channel inhibition.4 In this case, inhibition of maxiK channels at the initial states of such depolarization should prevent a prompt repolarization induced by the activation of maxiK by Ca2+ and depolarization. Hypoxic inhibition of maxiK channels would be released at higher potentials or higher [Ca2+]i, allowing full maxiK channel activation. As a consequence, inhibition of maxiK channels would amplify the cell response to hypoxia without losing the negative-feedback role of the channels.
This work was supported by Spanish Dirección General de Investigación Científica y Técnica Grant PB97/0400. We thank Maria de los Llanos Bravo for technical help.
Original received April 23, 2001; revision received June 28, 2001; accepted July 6, 2001.
Gonzalez C, Almaraz L, Obeso A, Rigual R. Carotid body chemoreceptors: from natural stimuli to sensory discharges. Physiol Rev. 1994; 74: 829–898.
Ganfornina MD, López-Barneo J. Gating of O2-sensitive K+ channels of arterial chemoreceptor cells and kinetic modifications induced by low Po2. J Gen Physiol. 1992; 100: 427–455.
Wyatt CN, Peers C. Ca2+-activated K+ channels in isolated type I cells of the neonatal rat carotid body. J Physiol. 1995; 483: 559–565.
Buckler KJ, Williams BA, Honore E. An oxygen-, acid- and anaesthetic-sensitive TASK-like background potassium channel in rat arterial chemoreceptor cells. J Physiol. 2000; 525: 135–142.
López-Barneo J, Pardal R, Ortega-Saenz P. Cellular mechanism of oxygen sensing. Annu Rev Physiol. 2001; 63: 259–287.
Acker H. Cellular oxygen sensors. Ann NY Acad Sci. 1994; 718: 3–10.
Ruppersberg JP, Stocker M, Pongs O, Heinemann SH, Frank R, Koenen M. Regulation of fast inactivation of cloned mammalian IKA channels by cysteine oxidation. Nature. 1991; 352: 711–714.
Park MK, Lee SH, Lee SJ, Ho WK, Earm YE. Different modulation of Ca-activated K channels by the intracellular redox potential in pulmonary and ear arterial smooth muscle cells of the rabbit. Pflugers Arch. 1995; 430: 308–314.
Wang ZW, Nara M, Wang YX, Kotlikoff MI. Redox regulation of large conductance Ca2+-activated K+ channels in smooth muscle cells. J Gen Physiol. 1997; 110: 35–44.
DiChiara TJ, Reinhart PH. Redox modulation of hslo Ca2+-activated K+ channels. J Neurosci. 1997; 17: 4942–4955.
Perez-Garcia MT, López-López JR, Gonzalez C. Kvβ1.2 subunit coexpression in HEK293 cells confers O2 sensitivity to kv4.2 but not to Shaker channels. J Gen Physiol. 1999; 113: 897–907.
Thuringer D, Findlay I. Contrasting effects of intracellular redox couples on the regulation of maxi-K channels in isolated myocytes from rabbit pulmonary artery. J Physiol. 1997; 500: 583–592.
López-López JR, Gonzalez C. Time course of K+ current inhibition by low oxygen in chemoreceptor cells of adult rabbit carotid body: effects of carbon monoxide. FEBS Lett. 1992; 299: 251–254.
López-López JR, Gonzalez C, Pérez-García MT. Properties of ionic currents from isolated adult rat carotid body chemoreceptor cells: effect of hypoxia. J Physiol. 1997; 499: 429–441.
Qin F, Auerbach A, Sachs F. Estimating single-channel kinetic parameters from idealized patch-clamp data containing missed events. Biophys J. 1996; 70: 264–280.
Qin F, Auerbach A, Sachs F. Maximum likelihood estimation of aggregated Markov processes. Proc R Soc Lond B Biol Sci. 1997; 264: 375–383.
Schoenmakers JM, Visser GJ, Flik G, Theuvenet APR. CHELATOR: an improved method for computing metal ion concentrations in physiological solutions. Biotechniques. 1992; 12: 870–879.
McManus OB, Magleby KL. Accounting for the Ca2+-dependent kinetics of single large-conductance Ca2+-activated K+ channels in rat skeletal muscle. J Physiol. 1991; 443: 739–777.
Peers C. Oxygen-sensitive ion channels. Trends Pharmacol Sci. 1997; 18: 405–408.
Jiang C, Haddad GG. A direct mechanism for sensing low oxygen levels by central neurons. Proc Natl Acad Sci USA. 1994; 91: 7198–7201.
Conforti L, Millhorn DE. Selective inhibition of a slow-inactivating voltage-dependent K+ channel in rat PC12 cells by hypoxia. J Physiol. 1997; 502: 293–305.
Osipenko ON, Tate RJ, Gurney AM. Potential role for Kv3.1b channels as oxygen sensors. Circ Res. 2000; 86: 534–540.
McKenna F, Ashford MLJ, Peers C. Hypoxia reversibly inhibits the activity of cloned human brain BKCa channels stably expressed in HEK 293 cells. J Physiol. 1998; 509: 188P.
Liu H, Moczydlowski E, Haddad GG. O2 deprivation inhibits Ca2+-activated K+ channels via cytosolic factors in mice neocortical neurons. J Clin Invest. 1999; 104: 577–588.
DiChiara TJ, Reinhart PH. Distinct effects of Ca2+ and voltage on the activation and deactivation of cloned Ca2+-activated K+ channels. J Physiol. 1995; 489: 403–418.
Rothberg BS, Magleby KL. Voltage and Ca2+ activation of single large-conductance Ca2+-activated K+ channels described by a two-tiered allosteric gating mechanism. J Gen Physiol. 2000; 116: 75–99.
Hoshi T, Heinemann SH. Regulation of cell function by methionine oxidation and reduction. J Physiol. 2001; 531: 1–11.
Tamayo L, López-López JR, Castaneda J, Gonzalez C. Carbon monoxide inhibits hypoxic pulmonary vasoconstriction in rats by a cGMP-independent mechanism. Pflugers Arch. 1997; 434: 698–704.
Jackson A, Nurse C. Dopaminergic properties of cultured rat carotid body chemoreceptors grown in normoxic and hypoxic environments. J Neurochem. 1997; 69: 645–654.
Buckler KJ. A novel oxygen-sensitive potassium current in rat carotid body type I cells. J Physiol. 1997; 498: 649–662.