| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
RECEPTORS AND SIGNAL TRANSDUCTION
1Departamento de Fisiología, Facultad de Medicina, Universidad Nacional Autónoma de México, México D. F., México; and 2Department of Biochemistry and Molecular Biology, University of Maryland School of Medicine, Baltimore, Maryland
Submitted 20 October 2006 ; accepted in final form 12 February 2007
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
S produces a nonreversible 
34% decrease in Qmax, a 10 mV shift in Vh, and a 63% increase in k with respect to the control. Norepinephrine induces a 7 mV shift in Vh and 40% increase in k. Overexpression of G protein 
14 subunits produces a 13% decrease in Qmax, a 9 mV shift in Vh, and a 28% increase in k. We correlate charge movement modulation with the modulated behavior of voltage-gated channels. Concurrently, G protein activation by transmitters and GTPS also inhibit both Na+ and N-type Ca2+ channels. These results reveal an inhibition of gating charge movement by G protein activation that parallels the inhibition of both Na+ and N-type Ca2+ currents. We propose that gating charge movement decrement may precede or accompany some forms of GPCR-mediated channel current inhibition or downregulation. This may be a common step in the GPCR-mediated inhibition of distinct populations of voltage-gated ion channels. ion channel modulation; G protein-coupled receptors; charge movement
The activity of almost all types of voltage-gated ion channels is modulated by at least one kind of extra- or intracellular signaling process. A broad array of neuromodulators—monoamines, peptides, glutamate, and acetylcholine—have been shown to alter the function of voltage-gated channels in neurons. The modulation of voltage-gated ion channels by signaling pathways activated by G protein-coupled receptors (GPCRs) constitutes a widespread mechanism for regulating neuronal excitability, neurotransmitter release, and plasticity (15, 18). The mechanisms that underlie the ion channel modulation mediated by GPCR-dependent signaling pathways alter channel function by modifying their opening probability, kinetics, and voltage dependence without altering, however, ionic selectivity and unitary conductance (15, 17, 23, 32, 37, 39). In almost all cases, this has been established by solely examining the properties of ionic current flowing through ion channels. Only in a few cases has voltage-gated ion channel modulation been examined at the level of gating charge movement (7, 36). This study was undertaken to investigate whether G protein-dependent signaling pathways modulate gating charge movement in a native neuronal system. We analyzed total gating charge movement in rat superior cervical ganglion (SCG) neurons. Our results reveal that G protein activation induces a mixed pattern of inhibitory modulation on gating charge movement. Part of the inhibitory pattern is voltage dependent, whereas the other is voltage independent. We establish a correlation between gating charge movement modulation and the modulated behavior of two populations of voltage-gated ion channels present in these neurons. We report that TTX-sensitive Na+, similar to N-type Ca2+-channel currents, is inhibited by GPCR-dependent signaling pathways. Our results show an inhibition of gating charge movement by G protein activation in a manner that closely parallels the inhibition of Na+ and N-type Ca2+ channels. We propose that gating charge movement trapping or its decrement may precede or accompany some forms of ion channel inhibition and/or ion channel downregulation. This process might be a common step in the GPCR-mediated modulation of distinct voltage-gated ion channels.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Intranuclear microinjection.
For some experiments, after a 4-h wait for attachment to the substrate, the neurons were intranuclearly microinjected using an Eppendorf 5242 pressure microinjector and a 5171 micromanipulator system (Eppendorf, Madison, WI) as described previously (25). The injection solution contained cDNA construct coding for a green fluorescent protein (GFP) mutant fused to the G protein 1-subunit (G1-GFP, 100 ng/µl) and the G protein 4-subunit (G4, 100 ng/µl) expression plasmids mixed with 1 mg/ml of 10,000 kDa dextran-fluorescein used as an injection marker. The employment of the tagged subunits facilitated the identification of cells that positively express G subunits. Injection pressures of 10–20 kPa for 0.5–0.8 s resulted in no obvious increase in cell volume. After 12–16 h, successfully injected neurons were identified by their characteristic greenish-blue GFP fluorescence using an inverted microscope equipped with epifluorescence and fluorescent optics.
Solutions.
Neurons were constantly and locally perfused during recording (1–2 ml/min) with appropriate solutions designed to isolate total nonlinear capacitive currents, Ba2+ currents (flowing through N-type Ca2+ channels), Na+ currents, and K+ currents. For current measurements through Ca2+ channels, the bath solution contained (in mM) 162.5 tetraethylammonium (TEA) chloride (TEACl), 2–5 BaCl2, 10 HEPES, 8 glucose, 1 MgCl2, 0.0001 TTX, and 0.005 nifedipine, pH adjusted to 7.4 with TEAOH. The Ca2+ current of rat SCG neurons is carried 85–90% in N-type channels and the remainder in L-type channels, with no detectable P- or Q-type channels component (26, 46, 49). Therefore, the N-type Ca2+ current was defined as the component of the current sensitive to 100 µM CdCl2 in the presence of 5 µM nifedipine. For Na+ current measurements shown in 
Figs. 2 and 6, the bath solution contained (in mM) 103 TEACl, 60 NaCl, 10 HEPES, 8 glucose, 2.9 MgCl2, and 0.1 CdCl2, pH adjusted to 7.4 with TEAOH. For Na+ currents measurements shown in Fig. 3, the bath solution contained (in mM) 153 NaCl, 10 TEACl, 10 HEPES, 8 glucose, 2.9 MgCl2, and 0.1 CdCl2, pH adjusted to 7.4 with NaOH. For K+ currents measurements, the bath solution contained (in mM) 163 NaCl, 10 HEPES, 8 glucose, 5 KCl, 2.9 MgCl2, 0.0002 TTX, and 0.1 CdCl2, pH adjusted to 7.4 with NaOH. The external solution for nonlinear capacitive currents was designed to eliminate voltage-gated ionic currents and consisted of (in mM) 161.5 TEACl, 10 HEPES, 8 glucose, 2.8 MgCl2, 1 CsCl, 0.0001 TTX, 0.1 CdCl2, 0.1 LaCl3, pH adjusted to 7.4 with TEAOH.
|
|
|
|
|
|
4 mV shift) in surface potential. Where noted, 100 µM norepinephrine (NE) was added to the external solution. Nifedipine-containing external solution was prepared daily from 25 mM absolute ethanol stock solution and protected from light whenever possible. For most of the experiments of this study, we evaluated short-term actions of transmitters on membrane currents. Here we define short-term effects as the alterations in the electrophysiological properties of membrane currents that occur within seconds to minutes (<10 min) after a change in an experimental condition.
The standard pipette solution (internal solution) contained (in mM) 140 CsCl, 10 HEPES, 11 EGTA, 1 CaCl2, 5 MgCl2, 5 Na2ATP, 0.3 Na2GTP, and 0.1 leupeptin, pH adjusted to 7.4 with CsOH. The choice of such a solution was the result of a large series of trial and error tests in preliminary experiments. When compared with large nonpermeable inorganic (N-methyl-D-glucamine+ and TEA+) or organic (aspartate and glutamate) cations, we obtained with a Cs+-based solution a large improvement in sealing, durability, blocking of ionic currents, and comparability with experiments at the level of ionic current. Where noted, CsCl was replaced with KCl and 0.3 Na2GTP was replaced with 2 mM Li3GDPS or with 0.3 mM Li4GTPS. For some Na+ current measurements, final intracellular [Na+] was set at 30 mM. All drugs were obtained from Sigma (St. Louis, MO). The used internal Na+ concentration was within the range of values measured experimentally at 25°C in mammalian sympathetic neurons by different methods of analysis (8, 14).
Electrophysiology and data analysis.
Membrane current measurements were performed in the whole cell configuration of the patch-clamp technique (27) using a EPC-7 or a EPC-10 (Heka Instruments) amplifiers and Sylgard 184 (Dow Corning)-coated pipettes of borosilicate glass with resistance of 0.8–1.2 M
when filled with the internal solution. Once that whole cell configuration was established, cells were held at –60 mV and the capacitance compensation was adjusted to cancel linear capacity transients and series resistance [1.5 M (SD 0.44)] was compensated to >70% and periodically monitored. Care was taken to select spherical neurons of small diameter with no evident process to improve clamp. Neurons with signs of qualitative clamp error were rejected from the analysis. Measurements were started >1–3 min after membrane rupture and access to whole cell patch-clamp configuration.
Voltage protocols were generated and current responses were digitized and stored for subsequent analysis using either Basic-Fastlab (Indec Systems, Capitola, CA) for the EPC-7 or Patchmaster and Fitmaster (Heka Instruments). Command pulses were delivered at every 10-s intervals to the levels and duration indicated in each figure from a holding potential of –100 mV, unless otherwise indicated. Currents were typically low-pass filtered at 3 to 10 kHz (3-pole Bessel filter) and digitized using a 12-bit analog-to-digital Tecmar Lab Master board (maximum digitization rate, 125 kHz). Steady-state currents were sampled at 10 kHz and tail currents at 50 kHz using a "split-clock" protocol. Traces from experiments measuring Ba2+, Na+, and K+ currents were not corrected for linear leakage and capacitive currents. Periods of 200–400 µs at the onset and 100–300 µs at the end of the command step were removed from almost all ionic current records to eliminate spurious points due to incomplete capacitive compensation.
For calculation of gating charge movement, nonlinear capacitive currents were sampled at 10 kHz. Linear capacitive and leak currents were routinely subtracted by a P/–5 protocol from a subholding potential of –120 mV (see voltage protocol in Fig. 1, top). Fig. 1B shows the zero sum response to opposite polarity voltages jumps, indicating the linearity of membrane properties over the range of voltages used to obtain leak templates, usually in the range of –120 to –140 mV. It seems reasonable to assume that leak templates in this range are a little contaminated with nonlinear capacitive currents themselves, since there was usually little or no nonlinearity in the range of –140 to –100 mV.
Gating charge that was moved during test depolarization (Qon) was quantified by calculating the area under the curve of each trace of nonlinear capacitive currents using the baseline of each trace as a steady-state level. Total charge moved during repolarization (Qoff) was calculated similarly for experiments shown in Fig. 1.
Nonstationary noise analysis.
Ensembles of currents were generated by a series of identical voltage pulses from –90 to –10 mV delivered every 4 s. Data were sampled at 20 µs after filtering at 3 kHz (3-pole Bessel). The variance, 
2, in trial-to-trial fluctuations of the ensemble current evoked by identical depolarizing steps was calculated from two to six sets of 80–200 consecutive sweeps. Background variance at the holding potential was subtracted from the variance during the test pulse. Variance-mean plots were fitted with a least-square algorithm to the following equation (3, 55):
 | (1) |
Because the magnitude of the Ba2+, Na+, K+, and nonlinear capacitive currents were dependent on cell size, aggregate current data and the amount of gating charge movement are presented as current or charge density normalized to cell capacitance. To avoid one source of systematic bias, experimental and control measurements were alternated whenever possible, and concurrent controls were always performed. All recordings were obtained at room temperature (19–22°C). Where appropriate, data are expressed as means ± SE, with n representing the number of observations. Statistical significance was determined by using an unpaired Student's t-test. Data were considered significant when P < 0.05.
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
 | (2) |
There is a general agreement between Qon and Qoff for all voltages, except that the amplitude of Qoff was less than Qon for depolarizations above –20 mV (Fig. 1D). The ratio of Qon and Qoff at +80 mV was 1.19 ± 0.07 (n = 14). In experiments where the duration of the voltage steps was diminished to 1 ms, the values for Qon and Qoff were nearly equal (Qon/Qoff = 1.05 ± 0.02, n = 5), whereas pulses of 10 ms enhanced the inequality of charges (data not shown). Therefore, in the experiments shown in Figs. 2, 4, and 7, only Qon was analyzed.
|
What channels are actually contributing to total gating charge movement? There are no pharmacological tools to isolate and/or identify gating charge movements arising from a specific channel population. To obtain an estimation of the channels that are actually contributing to the total charge movement, we followed two approaches: 1) the use of charge movement immobilization and voltage-gated channel inactivation protocols and 2) nonstationary fluctuation analysis of Na+ and Ca2+ currents.
Charge movement immobilization and voltage-gated channel inactivation.
If gating charge of Na+, fast-transient type-A K+, and Ca2+ channels at least contribute to the total charge movement that we recorded, distinct fast and slow components should be observed. However, as can be appreciated in Fig. 1A, nonlinear capacitive currents did not exhibit distinct fast and slow components. One possibility could be that series resistances are not completely compensated; therefore, nonlinear capacitive currents might be filtered. This limits the possibility to establish a kinetic separation of distinct components. In an alternative attempt to separate such distinct components, we explored charge movement immobilization and voltage-gated channel inactivation. Inequality of moved charges (Fig. 1, C and D) might be due to a time- and voltage-dependent gating charge immobilization associated with fast voltage-dependent inactivation (5, 12). To test this possibility, we first evaluated charge movement immobilization and then voltage-gated channel inactivation by using double-pulse protocols. Figure 2A shows nonlinear capacitive currents traces elicited after a 500-ms conditioning prepulse (PP) to voltages from –100 to 0 mV. Increasing the magnitude of PP resulted in a reduction of Qon charge movement, elicited by a 4-ms pulse to +20 mV. In four neurons, the reduction of Qon averaged 64% ± 3% when nonlinear capacitive currents were preceded by a PP to –40 mV (Fig. 2A; trace c). The immobilization curve is shown as Qon versus PP voltage in Fig. 2E (solid circles). Qon began to decrease at PP potentials positive to –90 mV and was partially immobilized by a 500-ms PP to 0 mV. This raises the possibility that the component(s) of gating charge movement in SCG neurons, in which amplitude decreases after a PP (Fig. 2A), in fact arises from rapidly inactivating Na+ channels and transient type-A K+ channels since these neurons have large Na+ (INa) and transient type-A K+ (IA) currents (51, 52). We tested this possibility by comparing the voltage dependence of charge immobilization with that of both INa and IA steady-state inactivation. Figure 2, B and C, shows experiments in which voltage dependence of steady-state inactivation of both Na+ and transient type-A K+ channels was determined by voltage protocols illustrated at the top of each panel. INa and IA evocated by depolarizing steps to –20 mV after a 500-ms PP are shown in Fig. 2, B and C, respectively. Increasing the amplitude of PP resulted in the inactivation of both INa and IA (Fig. 2, B and C). The INa steady-state inactivation curve is shown as peak current versus PP voltage (Fig. 2E; open squares). INa began to inactivate at potentials positive to –90 mV and was almost fully inactivated by a 500-ms PP to –40 mV. At –20 mV, IA component was a little contaminated by other K+ currents, obviating the need for digital subtraction. As the PP was depolarized, the IA amplitude decreased until it was completely inactivated at potentials more depolarized than –60 mV. The IA steady-state inactivation curve is shown as peak current versus PP voltage (Fig. 2E; solid triangles). The voltage dependence of INa and IA inactivation and charge immobilization could be fit well by a modified Boltzmann function:
 | (3) |
40% reduction of peak Ca2+ current recorded after depolarized PP in the range of –80 to –40 mV (Fig. 2E; open diamonds), whereas an almost full inactivation of both INa and IA was observed after depolarized PP in the range of –90 to –40 mV (Fig. 2E). Note the 60–80% reduction of peak Ca2+ current recorded after depolarized PP in the range of –40 to 10 mV (Fig. 2E). These results suggest that also a portion of total gating charge that is not reduced by depolarized PP in the range of –80 to –40 mV (
1/3 of the total) might arise from Ca2+ channels. The steady-state separation of two components of total gating charge movement by digital subtraction is illustrated in Fig. 2A. The current traces a–c show the PP-sensitive component, presumably the gating charge arising from both Na+ and transient type-A K+ channels, whereas the current trace c shows the PP-insensitive component of the total charge movement, presumably the gating charge arising from Ca2+ channels. Our immobilization/inactivation experiments suggest that, from a host of voltage-gated ion channels present in the cell body of a SCG neuron, primarily Na+ and Ca2+ channels might contribute to the total gating charge movement. Nonstationary fluctuation analysis of Na+ and Ca2+ currents. We use ensemble fluctuation analysis to obtain an estimation of the number of Na+ and Ca2+ channels present in the cell body of the rat SCG neurons. Nonstationary fluctuation analysis typically involves measuring an ensemble of responses to a repeated activating stimulus. The theory is based is three general assumptions about voltage-gated channels: 1) channels are composed of a homogeneous population, 2) this channel population is composed of a fixed number of independently gated channels, and 3) channels are assumed to exist in either a conducting or nonconducting state. When the probability of being in the conducting state is maximal, the theory predicts a parabolic relation between the current variance and mean that can be used to estimate the number of channels, N, and their unitary current amplitude, i (3, 55).
Figure 3, A and C, shows representative macroscopic recordings of both INa and ICa (gray traces), as well as the corresponding time courses of the mean current (bold traces). Figure 3, A,b and C,b, shows the time of course of the variance (2) of both INa and ICa, respectively, estimated at the voltages indicated in each panel from three different neurons. Figure 3, B and D, shows plots of variance versus mean current corresponding to data shown in Fig. 3, A and C, respectively. Continuous lines are fit to the data by using Eq. 1, allowing the estimation of the single channel current and the number of the channels, indicated in each panel. These results show the number of Na+ (7.8 x 104 ± 0.39 x 104, n = 7) and Ca2+ (3.8 x 104 ± 0.27 x 104, n = 5) channels in the cell body of 1-day cultured SCG neurons.
G protein activation by GTPS inhibits gating charge movement.
A general strategy of implicating G proteins in the modulation of voltage-gated ion channels is to introduce nonhydrolysable guanine nucleotide analogs into the cell via the patch pipette (32). Figure 4A shows that averaged total nonlinear capacitive current records obtained in SCG neurons dialyzed with GTPS (300 µM) are reduced in amplitude compared with the control (Fig. 4A; traces labeled as GTPS, n = 9). The magnitude of the inhibition induced by GTPS depends on the magnitude of the test depolarization. Maximal inhibition (60%) was present around –30 mV, whereas only 37% is seen at extreme test depolarizations (Fig. 4B; solid inverted triangles).
The voltage-dependence of gating charge movement inhibition arises from a positive shift in its activation curve. Figure 4B summarizes experimental data for Q-V curves of control and GTPS-dialyzed neurons. Three main significant effects are observed upon GTPS dialysis: 1) a 34% decrease in Qmax, 2) a 10 mV positive shift in Vh; and 3) a 63% increase in k (n = 9; P < 0.05, see Table 1). This suggests that nonselective G protein activation by GTPS modifies both the voltage dependence and the number of available channels of one or more populations of voltage-gated ion channels present in the cell body of SCG neurons.
|
14-subunits (G14). Figure 4A shows that total nonlinear capacitive currents recorded at moderated test depolarizations are much more inhibited by NE than those recorded at more positive potentials (traces labeled as NE). Gating charge movement is inhibited by NE 25% at –20 mV compared with just 3% at 40 mV (Fig. 4B; solid triangles). G14 overexpression mimics part of the effects of NE. Gating charge movement is inhibited by G14 overexpression 41% at –20 mV in contrast to 14% at 40 mV (Fig. 4B; solid diamonds). A slight prolonged decaying phase of the "on" nonlinear capacitive currents is also evident at moderate test potentials, especially with NE, but we did not analyze this effect in more detail. In neurons dialyzed with GDPS, a nonhydrolysable analog of GDP that minimizes G protein activation, NE was without effect (data not shown). Therefore, NE-induced effects are G protein dependent.
In rat SCG neurons, GDPS has a fairly large effect on N-type Ca2+ current (see Table 2). Intracellular dialysis with solutions containing 2 mM of GDPS removes tonic G protein-mediated Ca2+ current inhibition, prevents the actions of GPCR activated by neurotransmitters, greatly reduced facilitation, enhanced inactivation, and shifted tail Ca2+-current activation curves to more hyperpolarized potentials (26, 33). However, on gating charge movement, dialysis with GDPS induces small and nonsignificant effects (see Table 1).
|
14 (n = 5) is a significant 6- and 9-mV positive shift in Vh, respectively (P < 0.05, see Table 1). In NE-exposed neurons, a significant decrease was found in the steepness (k) of the Q-V curve and Qmax remained essentially unchanged (see Table 1). It is important to note that, in G14-overexpressed neurons, Qmax was reduced 13% compared with control neurons, which may indicate that overexpression of G14 induces both voltage-dependent and -independent inhibition. The following reasons suggest that a possible candidate to underlie gating charge movement inhibition by G protein activation is the N-type Ca2+ channel population. First, in rat SCG neurons, most of the whole cell Ca2+-channel current flows through N-type Ca2+ channels (26, 46, 49). Second, N-type Ca2+ channels are a typical target of inhibitory modulation by several signaling pathways activated by different GPCRs (32, 37). Third, N-type Ca2+ channel inhibition by distinct GPCR-dependent signaling pathways alters channel-current properties in a voltage-dependent or -independent manner (23, 32, 37, 39). Fourth, at least part of the voltage-dependent inhibition of N-type Ca2+ current arises from an inhibition of their gating charge movement (36).
G protein inhibition of N-type Ca2+-channel currents.
In agreement with previous observations (22), Fig. 5A show that nonselective G protein activation by GTPS strongly inhibits Ba2+ currents (IBa) flowing through N-type Ca2+ channels. As in the case of gating charge movement, the inhibition of IBa induced by GTPS is partially voltage dependent. Maximal inhibition of Ba2+ tail currents (IBa tail) is 63% after test depolarizations to –10 mV and 36% after the most extreme depolarizing test pulses (Fig. 5B; solid inverted triangles).
All neurons studied under all conditions tested showed asymmetrical IBa tail activation curves around the point of half-maximal activation. These curves are well described by a two Boltzmann function (26, 33), a feature interpreted as to reflect the existence of two populations of N-type Ca2+ channels (G protein modulated and unmodulated; see Refs. 10 and 22). However, we still used a single Boltzmann approximation in the fit to IBa tail-V data to make a comparison with the data obtained from Q-V curves:
 | (4) |
S-dialyzed neurons (solid inverted triangles). Three main significant effects are observed upon GTPS dialysis: 1) a 33% decrease in Imax, 2) a 10 mV positive shift in Vh, and 3) a 74% increase in k (n = 9; P < 0.05, see Table 2).
The selective participation of different intracellular inhibitory pathways in the modulation of N-type Ca2+ channels is demonstrated by the use of a specific transmitter and G14. Figure 5A shows that a saturating dose of NE (100 µM; see Ref. 60) partially inhibits IBa in SCG neurons. In agreement with previous reports (11), much of the NE-induced inhibition of IBa is absent at extreme positive potentials (45% at –10 mV vs. 6% at extreme depolarizations), and it is therefore mostly voltage dependent (10, 11, 23). Consequently, all NE-induced effects can be ascribed to a significant +17-mV shift in Vh and a 9-mV increase in k of the activation curve, with little if any effect on Imax (n = 16; P < 0.05, see Table 2).
Previous reports strongly support the idea that voltage-dependent inhibition of N-type Ca2+ channels results from a direct interaction between activated G protein -subunits and the 
1-subunit of these channels (25, 28, 34). Figure 5, A and B, shows that a modulation of IBa similar to the inhibition caused by NE is produced after overexpression of G14 by intranuclear injection of cDNA in SCG neurons (n = 5; P < 0.05, see Table 2). As has been shown before (25, 28), we find that a microinjection of DNA for G subunits, with and without DNA for G subunits, into the nucleus of SCG neurons inhibited Ca2+ currents mostly in a voltage-dependent manner, increased the facilitation ratio, and partially occluded the actions of NE (data not shown). However, it is important to note that, in G14-overexpressed neurons, Imax was significantly reduced 17% compared with control neurons (see Table 2). This effect may indicate that overexpression of G14 induces both voltage-dependent and -independent inhibition.
From these results, and in agreement with previous interpretations, we conclude that nonselective G protein activation by GTPS induces two independent inhibitory patterns on N-type Ca2+ channels. One inhibitory pattern is voltage dependent, whereas the other is voltage independent. It can be inferred that different G proteins might induce different combinations of voltage-dependent and -independent effects.
As it can be appreciated in Figs. 4 and 5, there is a qualitative correlation between the actions of GTPS, NE, and G14 on both N-type Ca2+-channel currents and gating charge movement. On the one hand, GTPS and NE induce a predominant voltage-dependent inhibition of both IBa and gating charge movement, arising from a positive shift on both IBa tail-V and Q-V relationships (see Figs. 4B and 5B). Also, Fig. 7A shows a complete recovery of total gating charge movement at +40 mV, but a strong inhibition of IBa at this same voltage. One possibility, to explain the fact that NE is still inhibiting N-type Ca2+ current at +40 mV, could be that at such voltage, G proteins acts on transitions (near to the open state) (12) where little charge moves; however, such transitions may be strongly affected, and the excitation-conduction coupling could be less efficient.
On the other hand, note that both GTPS and G14 also inhibit IBa tail and gating charge movement by a constant proportion in addition to altering their voltage dependence (Figs. 4, 5, and 7A). Therefore, based on the above comparisons, we suggest that part of the effects on gating charge movement induced by G protein activation may arise from the inhibition of gating charge of N-type Ca2+ channels. In a previous report, Jones et al. (36) suggest that G proteins act to inhibit both voltage-sensor movement and the transduction of voltage-sensor activation into channels opening. Our results with native channels are consistent with this hypothesis.
Inhibition of Na+ currents by GTPS and angiotensin II.
Several observations suggest that not all modulation of gating charge movement can be ascribed to the regulation of N-type Ca2+ channels and that one or more different populations of voltage-gated ion channels can be involved in the observed effects. First, maximal asymptotic effects of GTPS are very similar in IBa but quite different on gating charge movement. Second, if all GTPS-observed effects were due solely to the modulation of N-type Ca2+ channels, it would imply that these channels are responsible for at least 50% of the total gating charge movement recorded in these cells, a very unlikely condition (see DISCUSSION).
Two observations led us to suspect that the modulation of a fast-inactivating voltage-gated ion channel population can contribute to the observed effects of G proteins on gating charge movement. First, as noted above, the inequality of Qon-to-Qoff ratio (Fig. 1C) suggests that rapidly inactivating channels contribute to total gating charge movement. Second, at least two-thirds of the gating charge movement is immobilized by PP to –40 mV (Fig. 2, A and E). The same protocol produces a complete inactivation of both macroscopic INa and IA (Fig. 2, B and C).
Several reports have shown that voltage-gated Na+ channels of central and sensory neurons are targets of inhibitory GPCR-dependent signaling pathways (15). The above observations and findings prompted us to look for electrophysiological evidence of INa modulation by GPCR-dependent signaling pathways in rat sympathetic neurons.
To evaluate a possible rundown of INa, we assessed changes in INa amplitude during repetitive depolarizing pulses (up to 200 sweeps) from a holding potential of –100 mV to a test potential of –20 mV, applied 10 s apart. The amplitude of INa, recorded in this manner, was minimally reduced (<5%) over the 10-min course of a typical experiment (Fig. 6A). Step depolarizations beyond –40 mV activated a fast transient-inward INa (Fig. 6E, solid circles). Figure 6B (trace labeled as control) shows a INa elicited by a 10-ms depolarizing pulse to –20 mV from a holding potential of –100 mV. Superfusion of the cell with 100 nM TTX completely blocked the current at all tested voltages (Fig. 6E, solid squares). The current versus voltage (I-V) relationship shown in Fig. 6E (solid circles) shows that the currents activated at potentials positive to –50 mV reached maximum amplitude near –20 mV and reversed near +20 mV. An activation curve plotted as normalized conductance versus test potential for control neurons is shown in Fig. 6F (solid circles). Conductance was calculated from peak INa values from the I-V relationship (Fig. 6E; solid circles) using the equation GNa = INa/(V – VNa), where V is the membrane potential and VNa the INa are reversal potential. The solid line through the activation curve in Fig. 6F (solid circles) was a least squares fit of the data to the modified Boltzmann equation:
 | (5) |
Consistent with a previous report (56), we found that the TTX-sensitive INa of SCG neurons is insensitive to both NE and acetylcholine (data not shown). However, we found that GTPS (n = 7) inhibits INa. Figure 6, A and B, shows an experiment with a significant inhibition (38% ± 3%, n = 7; P < 0.05 compared with control) of INa in GTPS-dialyzed neurons (solid inverted triangles). Figure 6F shows that inhibition affects the maximal relative conductance but does not alter its voltage dependence. Therefore, judging by the lack of effect on the current kinetics, and the voltage dependence, we can classify GTPS-mediated inhibition as voltage independent.
Bath application of 500 nM angiotensin II (ANG II) reduced INa. Figure 6C shows a representative experiment with a 16% inhibition of INa by ANG II; since it was typical, the inhibition reversed only slightly after washout of the agonist. ANG II did not induce a marked change in the activation kinetics and voltage dependence of INa (Fig. 6, D and F, open diamonds). The speed of the action of ANG II on INa was variable. In the experiment shown in Fig. 6C, 100 s elapsed between the start of the change in current (a rough estimation of when ANG II had reached its receptors) and when the inhibition had reached half its maximal level. The speed of action of ANG II on INa was qualitatively similar to that observed on both M-type K+ and Ca2+ currents (54) but slower than that induced by NE on Ca2+ currents in SCG neurons (11, 60). To test whether modulation of INa by ANG II was initiated by G proteins, we compared neurons dialyzed with either a control GTP-containing internal solution (0.3 mM GTP) or a GTP-free solution containing GDPS (2 mM), an antagonist of G protein activation. Neurons were dialyzed for >5 min using pipettes of 1.2 M before ANG II application. ANG II had a negligible effect in neurons dialyzed with GDPS. ANG II inhibited INa by 16 ± 6% in control neurons (n = 12) but only by 4 ± 2% in GDPS-dialyzed neurons (n = 8) (Fig. 6G). Therefore, we conclude that ANG II-induced effects involve the activation of G proteins.
Interestingly, the inhibition of INa by G protein activation parallels the inhibition of gating charge movement. As it can be appreciated in Fig. 7B, there is a correlation between the actions of GTPS on both INa and gating charge movement. On both the Q-V and GNa-V curves, GTPS induced a net decrease in Qmax and maximal relative sodium conductance (GNa), respectively (Fig. 7B). Based on the above comparisons, we suggest that part of the effects on gating charge movement induced by G protein activation may arise from the inhibition of gating charge of Na+ channels.
Taking all the above observations together, we conclude that G protein-induced effects on gating charge movement might underlie the mixture of G protein-induced inhibitory patterns that target, at least, Na+ and N-type Ca2+ channels.
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
The most striking result of the present study is that gating charge movement is modulated by G protein-dependent biochemical routes. This conclusion is supported by both voltage-dependent and -independent inhibition of gating charge movement induced by GTPS, NE, and G14.
Most of the charge movement comes from Na+ and Ca2+ channels. A host of voltage-gated ion channels contributes to the total gating charge movement recorded in our experiments. We used immobilization and inactivation pulse protocols and nonstationary noise measurements to approximate the relative contribution of particular channel populations to the total gating charge movement. Our results shown in Figs. 2 and 3 suggest that most of the charge movement arises from Na+ and Ca2+ channels, with perhaps a smaller contribution from transient type-A K+ channels. Delayed-rectifier K+ channels and Ca2+-dependent K+ channels seem unlikely to contribute significantly to our measured gating charge movement. These channels both activate and deactivate very slowly with time constants on the order of tens of milliseconds (51). If the gating charge arising from these channels were to move with a similar time course, it would be lost in the baseline unless it was enormous (9).
Voltage-dependent and -independent inhibition of N-type Ca2+ channels.
In our experimental conditions, only N-type Ca2+ channels are inhibited by NE. In rat SCG neurons, neither Na+ channels nor type-A K+ channels are sensitive to NE (56). Therefore, it is tempting to speculate that all NE-induced effects on the gating charge movement are the result of the voltage-dependent inhibition of N-type Ca2+ channels. This fits well with a previous study that demonstrated GTPS-mediated voltage-dependent inhibition on gating currents of N-type Ca2+ channels expressed in HEK293 cells (36). In the present study, the demonstration of the same observation in a native neuronal system with a putative transmitter and G subunits is an important advance.
The magnitude of the inhibition of both total gating charge movement and N-type Ca2+ channels by NE correlates well with the estimated contribution of these channels to the total gating charge movement. With an assumption of a contribution of one-third of these channels to the total gating charge movement, the 13% inhibition by NE at 0 mV correlates with the observed 40% inhibition of the N-type Ca2+-channel current at that same potential. The value of Qmax in unmodulated (GDPS-dialyzed) neurons is about 6.2 nC/µF (Table 1). With an assumption of a specific membrane capacitance of 1 µF/cm2, the saturating gating charge movement of surface membrane is 6.2 nC/cm2 or, translated into elementary charges (eo), 390 eo/µm2. With a consideration of an equivalent gating charge of 12 eo per each voltage-gated ion channel (12, 29, 47, 53), the predicted density of voltage-gated ion channels in rat SCG neurons would be 32 channels/µm2. A SCG neuron with a capacitance of 40 pF and a total area of 4,000 µm2 would have 12.8 x 104 channels. Nonstationary noise analysis of N-type Ca2+ currents suggests 3.8 x 104 channels in a rat SCG neuron of average size. This value is in agreement with the number of Ca2+ channels reported by McDonough et al. (44). Taking together, the above arguments suggest that gating charge arising from Ca2+ channels comprises approximately one-third of total gating charge movement.
It is well established that a direct protein-protein interaction of the G complex with the 1-subunit of N-type Ca2+ channel mediates voltage-dependent inhibition of these channels (23, 25, 28, 34). Our results strengthen this conclusion at the level of gating charge movement. Overexpression of G subunits mimics much of the effects induced by NE on both N-type Ca2+-channel current and gating charge movement. In particular, both NE and G subunits decrease the steepness of activation curves for both N-type Ca2+ channels and gating charge movement. This suggests a reduced sensitivity of modulated N-type Ca2+ channels to the membrane potential. However, several observations do not fit well with solely voltage-dependent effects of G subunits on N-type Ca2+ channels. Overexpression of G subunits was used to induce a tonic and voltage-dependent inhibition of N-type channels. Unexpectedly, this procedure appears to induce both voltage-dependent and -independent inhibitions on Qmax (see table 1).
Voltage-independent inhibition of N-type Ca2+ channels may contribute to the observed voltage-independent inhibition of total gating charge movement. Our results show that activation of G proteins by GTPS induces a reduction of Qmax even by extreme depolarization. There are two possible explanations for this observation. First, in channels modulated by a voltage-independent pathway, channel opening is blocked at a very early step where much of the gating charge moves (12). This mechanism may agree with the observations made by Carabelli et al. (16) and Lee and Elmslie (42), who reported an increase of "null sweeps" in records of single N-type Ca2+ channels exposed to neurotransmitters. Second, Tombler et al. (58) recently reported a novel mechanism of GPCR-mediated modulation of N-type Ca2+ channels involving destabilization and subsequent removal of Ca2+ channels from the plasma membrane. Concurrently with the above, our results support the initial hypothesis of Dunlap and Fischbach (21), who first suggested a reduction in the number of functionally available Ca2+ channels by neurotransmitter-mediated inhibition.
Voltage-independent inhibition of N-type Ca2+ channels is quite complex and has been much less studied than the voltage-dependent pathway (23). At least two forms of voltage-independent inhibition, fast and slow, can be distinguished in sympathetic neurons by the speed of inhibition and by the involvement of a final effector(s) (23). Clues as to which final effector might be involved in one slow form came from experiments showing that G protein-mediated hydrolysis of phosphatidylinositol 4,5-bisphosphate (PIP2) causes voltage-independent N-type Ca2+ channel inhibition (24, 59). The final effector of fast voltage-independent inhibition of N-type Ca2+ channels is still unknown. Nevertheless, our data in Fig. 4 provide an additional piece of evidence in favor of the G subunit as this final effector. Further experiments are required to establish whether PIP2 or G binds to and modulates N-type Ca2+ channels and gating currents in a voltage-independent manner.
Modulation of Na+ channels by G protein-dependent pathways.
Several pieces of evidence suggest that N-type Ca2+ channels are not the only target of voltage-independent modulation by activated G proteins. For example, with the assumption that N-type Ca2+ channels contribute one-third of the total gating charge movement, the inhibition of nonlinear capacitive currents should be at most 67% smaller than that of N-type Ca2+ current. However, GTPS inhibits 50% of both total gating charge movement and N-type Ca2+ current (see Figs. 4 and 5). Thus the only possible conclusion is that activated G proteins do not specifically inhibit N-type Ca2+-channel gating charge movement.
Early studies in rat sympathetic neurons led to the idea that Na+ channels are not modulated by GPCR-dependent signaling pathways (56). This study is the foremost report in showing that TTX-sensitive Na+ channels of rat sympathetic neurons are modulated by G protein activation. There are two observations in support of this: 1) GTPS (a nonselective and irreversible G protein activator) and ANG II (a GPCR agonist) induce an inhibition of Na+ channel-current amplitude; and 2) ANG II had no effect in neurons dialyzed with GDPS, which is expected to minimize the activation of G proteins. The observed effects of GTPS and ANG II on INa show a voltage-independent inhibition at all potentials tested and poor reversibility. Our results might echo the reported voltage-independent inhibition of Na+ channels in central neurons by GPCR linked to PKA and/or PKC signaling pathways (Refs. 43 and 57; for review, see Ref. 15). Nevertheless, further work is required to explore details about the final effector(s) of voltage-independent inhibition of Na+ channels observed in SCG neurons.
Finally, we cannot conclude that all the observed effects on gating charge movement inhibition arise exclusively from the regulation of Na+ and N-type Ca2+ channels. In particular, there is evidence that transient type-A K+ channels are modulated by GPCR through an indirect mechanism in central neurons (31, 40). Whether this modulation affects their gating charge movement is unknown. It may in turn account for some of the effects reported here. Further work is needed to elucidate this question.
Overall, the present study reveals a strong and differential inhibition of gating charge movement by G protein activation in a manner that parallels the novel inhibition of Na+ channels and the well-documented inhibition of N-type Ca2+ channels. We propose that gating charge movement trapping or decrement may precede or accompany some forms of ion channel inhibition or downregulation. This process may be a common step in the GPCR-mediated modulation of distinct populations of voltage-gated ion channels.
|
GRANTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
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. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONGRANTSREFERENCES |
|---|
Deceased 11 April 2005. 