| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1Department of Neurobiology and Physiology, Northwestern University, Evanston 60208; and 2Department of Physiology, Northwestern University Medical School, Chicago, Illinois 60611
Submitted 17 March 2004; accepted in final form 19 June 2004
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
; Hodgkin and Huxley 1952; Kuo and Bean 1994). The Na channels of cerebellar Purkinje neurons, however, tend to recover from inactivation through open states, giving rise to a "resurgent" current that flows upon repolarization from positive potentials (Raman and Bean 1997). These resurgent kinetics appear to result from channels with at least two mechanisms of rapid inactivation: the conventional or "fast" inactivation gate, corresponding to the cytoplasmic linker between domains III and IV of the Na channel 
subunit (Stühmer et al. 1989; Vassilev et al. 1989), as well as an endogenous, voltage-dependent, open-channel blocker (Raman and Bean 2001). Upon depolarization, the open-channel blocker binds more rapidly than the inactivation gate, limiting the extent of fast inactivation, particularly at positive potentials. Upon repolarization, the blocker is expelled, allowing resurgent current to flow and restoring the availability of Na channels. Because the recovery from inactivation associated with resurgent current shortens the refractory period between action potentials, high-frequency firing appears to be facilitated by Na channels with resurgent kinetics (Khaliq et al. 2003; Raman and Bean 2001).
Resurgent Na current was first characterized in cerebellar Purkinje neurons (Raman and Bean 1997), but recently it has also been described in neurons of the subthalamic nuclei (Do and Bean 2003). Several additional observations are also suggestive of the expression of resurgent currents by other classes of neurons. First, in a preliminary report (Mossadeghi and Slater 1998), resurgent current was documented in unipolar brush cells, which are neurons of the vestibulocerebellum that form a system of mossy fibers within the cerebellar cortex (Mugnaini and Floris 1994; Mugnaini et al. 1997). Second, in a study of spontaneous firing in cerebellar nuclear neurons, a subset of cells showed a small resurgent component to their Na currents (Raman et al. 2000). Third, in a modeling study, simulations of action potential firing predicted that cerebellar granule cells might express resurgent current (D'Angelo et al. 2002). These preliminary observations raise the possibility that resurgent Na currents may be more widespread than previously thought.
In this study, we compared the TTX-sensitive, voltage-gated Na currents from three classes of cerebellar neurons—cerebellar nuclear cells, unipolar brush cells, and granule cells—to those of Purkinje neurons, with a focus on testing for the presence of resurgent Na current in each cell type. The results indicate that the non–Purkinje cerebellar neurons indeed express Na currents with resurgent kinetics, although the resurgent component can be quite small in the low concentrations of external Na and high concentrations of divalent calcium-channel blocking ions that are often used for biophysical studies. When recordings are repeated in more physiological ionic species and concentrations, however, the resurgent component of sodium current is increased to a greater extent than the transient component, such that all four cerebellar cell types produce robust resurgent currents. These results suggest that the channels with resurgent kinetics may regulate firing rates in several types of neurons.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Purkinje, granule, and cerebellar nuclear neurons were acutely dissociated from 13- to 17-day-old C57BL6 mice, and unipolar brush cells were acutely isolated from 17- to 18-day-old Sprague-Dawley rats (Charles River Laboratories, Wilmington, MA). In accordance with institutional guidelines, animals were anesthetized with halothane (mice) or isoflurane (rats) and decapitated. All procedures were consistent with approved protocols.
Purkinje and granule neurons were isolated as previously described (Raman et al. 1997; Regan 1991). The superficial layers of the cerebellum were removed and minced in ice-cold, oxygenated dissociation solution containing (in mM) 82 Na2SO4, 30 K2SO4, 5 MgCl2, 10 HEPES, 10 glucose, and 0.001% phenol red (buffered to a pH of 7.4 with NaOH). The tissue was incubated for 7 min at 31°C in 10 ml of dissociation solution containing 3 mg/ml of protease XXIII (pH readjusted) with oxygen blown over the surface of the fluid. After enzymatic treatment, the tissue was washed in dissociation solution containing 1 mg/ml trypsin inhibitor and 1 mg/ml bovine serum albumin (pH readjusted), in which the meninges were removed. The tissue pieces were then triturated with a series of fire-polished Pasteur pipettes to release Purkinje and granule neurons.
Cerebellar nuclei and unipolar brush cells were dissociated with slight modifications to previous methods (Raman and Trussell 1992; Raman et al. 2000). Cerebella were removed in oxygenated Tyrode's solution containing (in mM) 150 NaCl, 4 KCl, 2 CaCl2, 2 MgCl2, 10 HEPES, and 10 glucose, pH adjusted to 7.4 with NaOH (final Na concentration of 155 mM). Parasagittal cerebellar slices (350–400 µm thick) were cut either on a tissue chopper (for cerebellar nuclear cells; McIlwain, UK) or on a Vibratome Series 1000 (for unipolar brush cells, in Tyrode's solution at 33°C; Vibratome, St. Louis, MO). Slices were incubated for 20 min at 31°C in minimal essential medium (MEM; Life Technologies, Gaithersburg, MD) with (in mM) 10 HEPES, 1 cysteine, and 0.5 EDTA, pH 7.2, to which 40 U/mL papain (Worthington, Lakewood, NJ) was added. For cerebellar nuclei, 1 mg/ml chondroitinase ABC and 5 mM Na-acetate were also included in the dissociation solution. After enzymatic treatment, the slices were washed in MEM-HEPES containing 1 mg/ml trypsin inhibitor and 1 mg/ml bovine serum albumin, pH adjusted with NaOH to 7.4. For cerebellar nuclei, the cell body–rich regions in the core of the cerebellar slices were dissected out with tungsten needles. For unipolar brush cells, the nodulus was isolated from the slices. The tissue pieces were triturated with a series of fire-polished Pasteur pipettes to release individual neurons.
Cells were allowed to settle in the recording chambers in Tyrode's solution for 
1 h before recordings. Recordings were made at room temperature within 6 h after trituration.
For slice experiments, slices were prepared using standard techniques (Telgkamp and Raman 2002). Briefly, mice (P13–P16) were anesthetized with halothane, perfused with ice-cold artificial cerebrospinal fluid (ACSF, 4°C, in mM: 123.75 NaCl, 3.5 KCl, 26 NaHCO3, 1.25 NaH2PO4, 1 MgCl2, 2 CaCl2, and 10 glucose), and rapidly decapitated. The cerebellum was bathed in ice-cold ACSF bubbled with 95% O2-5% CO2, and parasagittal cerebellar slices (300 µM) were cut on a Leica VT100S vibratome. Slices were allowed to recover in oxygenated ACSF at 34°C for 1 h before recording.
Electrophysiological recordings: isolated cells
Borosilicate pipettes (1–3 M
for Purkinje and cerebellar nuclear cells; 3–4 M for unipolar brush cells, 7–10 M for granule cells) were wrapped with parafilm or coated with Sylgard (Dow Corning, Midland, MI) and fire polished. For Purkinje, granule, and cerebellar nuclear cells, the pipette solution consisted of (in mM) 108 CsCH3SO3, 9 NaCl, 1.8 MgCl2, 9 HEPES, 0.9 EGTA, 47.7 sucrose, 14 Tris-CreatinePO4, 4 MgATP, and 0.3 TrisGTP (buffered to a pH of 7.40 with CsOH). For unipolar brush cells, 117 CsCl replaced the CsCH3SO3 and NaCl, and 4 Na2ATP replaced the MgATP. Voltage-clamp recordings were made with an Axopatch 200A or 200B (Axon Instruments, Union City, CA). The bath was grounded with a 3 M KCl agar bridge to minimize the junction potential between the pipette and bath solution (see following text in METHODS). Data were recorded either with an InstruTech ITC-18 interface (Great Neck, NY) and Pulse software (Heka Electronik, Lambrecht, Germany) or with the Digidata/pCLAMP system (Axon Instruments). After a whole cell recording was established, each cell was positioned in front of a series of gravity-driven flow pipes containing different extracellular solutions. Whole cell series resistances of 5–7 M were routinely compensated by >70%.
Several control extracellular solutions were used for recordings, and each is referred to by Na concentration and primary divalent cation. The "high Na/Ca" solution consisted of Tyrode's solution with 5 mM TEACl and 30 µM CdCl2, except in Fig. 8. The "high Na/Ca" solution of Fig. 8 and the "high Na/Co" solution consisted of (in mM) 155 NaCl, 5 TEACl, 10 HEPES, and either 2 CaCl2 + 0.3 CdCl2 or 2 CoCl2, buffered to pH 7.4 with TrisOH for a final Na concentration of 155 mM. The "low Na/Co," "low Na/Ba," and "low Na/Ca" solutions consisted of (in mM) 50 NaCl, 110 TEACl, 10 HEPES, and 2 CoCl2, 2 BaCl2 + 0.3 CdCl2, or 2 CaCl2 + 0.3 CdCl2, buffered to pH 7.4 with TrisOH for a final Na concentration of 50 mM. After recordings were made in any control solution, protocols were repeated in a solution identical to the control but which included 300 nM or 1 µM TTX. TTX-sensitive Na currents in either low Na or high Na were obtained by subtraction. All drugs were from Sigma Aldrich, except for TTX (Alomone).
|
Neurons were visualized with infra-red differential interference contrast microscopy on a Zeiss FS1 microscope. Slices were bathed in oxygenated ACSF. Borosilicate pipettes (1–4 M) contained (mM) 108 KCH3O3S, 9 NaCl, 1.8 MgCl2, 9 EGTA, 9 HEPES, 5 TEACl, 14 Tris-creatine phosphate, 4 MgATP, and 0.3 Tris GTP, buffered to pH 7.3 with KOH. Purkinje cells were voltage-clamped nominally at –90 mV, and step depolarizations and repolarizations were given to elicit transient and resurgent currents. Recordings were made first in ACSF with 5 mM TEA and 0.3 mM CdCl2 and repeated in the same solutions but including 900 nM TTX.
Analysis
Data were analyzed with IGOR 4.02 software (WaveMetrics, Lake Oswego, OR). In isolated cells, for calculation of Na conductance, the reversal potential of the Na current was extrapolated from the peak current-voltage relation for each cell, and conductance was calculated by dividing peak current amplitude by driving force at each potential. Conductance-voltage data were fit with a Boltzmann equation of the form G(V) = 1/{1 + exp[–(V – V1/2)/k]}, where G is conductance, V is voltage, V1/2 is the voltage for half-maximal activation, and k is the slope factor. With the solutions used, and at 25°C, ENa was predicted to be +44 mV in low Na and +73 mV in high Na (Purkinje, granule, and cerebellar nuclear cell experiments) or +47 mV in low Na and +76 mV in high Na (unipolar brush cell experiments). Because the reversal potentials extrapolated from the transient current-voltage relations were near the predicted ENa values, the predicted values were used for the analyses of resurgent conductance.
Steady-state availability of Na channels was plotted as peak current evoked at 0 mV following a conditioning step versus the conditioning voltage. Data were fit with a modified Boltzmann equation of the form I/Imax = 1/{1 + exp[(V – V1/2)/k]}, where I/Imax is relative current, V1/2 is the half-maximal voltage for inactivation, and k is the slope factor.
To estimate the predicted change in current amplitudes in the low Na and high Na solutions at different potentials, predicted currents (as a function of permeability) were calculated with the Goldman-Hodgkin-Katz (GHK) current equation: INa = PNaz2(EF2/RT)([Na]i – [Na]oexp[–EF/RT])/(1 – exp(–EF/RT), where INa is Na current, PNa is Na permeability, z is valence, E is voltage, F is Faraday's constant, R is the gas constant, T is temperature, [Na]o is the extracellular concentration of Na+, and [Na]i is the intracellular concentration of Na+. The ratio of currents in high Na relative to those in low Na was calculated for the relevant potentials, assuming PNa was constant. Chord conductances were also calculated by dividing the predicted current by driving force, and the ratio of conductance at –30 mV to conductance at +30 mV (G–30 mV/G+30 mV) was calculated for high Na and low Na (junction potentials accounted for; see following text in METHODS). The GHK-predicted conductance ratio in high Na was 119% of that in low Na, and this was used as a reference for comparison of the resurgent-to-transient conductance ratio in the different solutions in Fig. 4.
|
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
The four classes of cerebellar neurons studied in this series of experiments were identified by the region of the cerebellum from which they originated (see METHODS), as well as by their size and morphology. Purkinje neurons were large (>20 µm diam), with a pear-shaped soma and an apical dendritic stump, and were therefore easily distinguishable from the small (5–10 µm diam), round granule cells released by the same dissociation procedure (Fig. 1A) (Regan 1991). Unipolar brush cells were also small (5–10 µm diam) but had a single, thick dendritic stem that terminated in a brush-like formation of branchlets (Fig. 1B) (Mugnaini and Floris 1994; Rossi et al. 1995). "Large" cerebellar nuclear cells had somata that were 10–20 µm in diameter and generally had two to four dendrites (Fig. 1C) (Raman et al. 2000).
|
To compare the properties of Na currents in the various cerebellar neurons, we recorded TTX-sensitive currents evoked by depolarization ("transient currents") as well as by repolarization ("resurgent currents"). To decrease the amplitude of Na currents and to improve the quality of voltage clamp, these recordings were made in solutions in which the external Na+ concentration had been reduced from physiological levels to 50 mM. Also, to reduce currents through Ca and K channels, 2 mM Ca2+ was replaced either by 2 mM Co2+ or by 2 mM Ba2+ with 0.3 mM Cd2+. Mg2+ was omitted from the solution, and osmolarity was compensated with TEACl. Although transient currents in Purkinje cells were on average two to three times larger than those of cerebellar nuclear and unipolar brush cells, the transient currents had a similar voltage-dependence of activation in all three cell types, consistent with previously reported values (Raman et al. 1997, 2000) (Fig. 2, A and B; Table 1). In granule cells, transient currents were too small to be reliably recorded in low Na, and recordings in higher Na are discussed later in the RESULTS.
|
|
). Figure 2C (top) shows the resurgent current elicited by a repolarizing step to –30 mV. A similar voltage protocol applied to cerebellar nuclear and unipolar brush cells also elicited currents that appeared to have a small resurgent component (Fig. 2C, middle and bottom; n = 7/7 cerebellar nuclear cells, n = 8/8 unipolar brush cells). The resurgent currents were occasionally obscured by the noise, however, particularly in cells with small transient current amplitudes (quantified below in the RESULTS). Na currents measured in quasi-physiological ionic conditions
To improve our resolution of resurgent currents, as well as to allow us to compare resurgent currents across cell types, we repeated the recordings in a quasi-physiological, modified Tyrode's solution ("high Na/Ca"). This solution included physiological concentrations of Na+ (155 mM), K+ (4 mM), Ca2+(2 mM), and Mg2+ (2 mM), but included 30 µM Cd2+ and 5 mM TEA+ to reduce Ca and K currents.
As expected, transient currents were larger in the high Na/Ca solution (quantified below). More strikingly, resurgent current increased greatly in all three cell types (Fig. 3, A and B), such that all cerebellar nuclear neurons (n = 9) and all unipolar brush cells (n = 5), like Purkinje cells, unequivocally produced robust resurgent currents. An enhancement of all components of Na current at potentials negative to the reversal is expected because of the increased extracellular Na+ concentration and the increased driving force. Based on the Na+ concentration change alone, the GHK current equation predicts that, in the solutions used, the ratio of the current in 155 mM Na+ to that in 50 mM Na+ (I155 Na/I50 Na) will be 4.6 at nominally +30 mV, dropping to 3.1 at –30 mV, as the asymptotic value (155/50) is approached at negative voltages that are distant from the reversal potential (junction potentials considered; see METHODS). Measuring transient currents at +30 mV gave an I155 Na/I50 Na of 4.7 ± 0.5 for Purkinje cells (n = 15 paired observations) and 5.2 ± 0.94 for cerebellar nuclear cells (n = 7 paired observations). The observed ratios were quite close to the GHK prediction (P = 0.8 for Purkinje and P = 0.6 for cerebellar nuclear cells). As shown in Fig. 4A, deviations from the predicted ratio (dotted line) were most noticeable in Purkinje cells with the largest transient currents (>2 nA in low Na), which were most likely to be subject to some clamp error in high Na. In the same set of cells, however, the amplitude ratio of resurgent currents at –30 mV increased rather than decreased. I155 Na/I50 Na became 6.9 ± 0.7 for Purkinje cells and 6.4 ± 0.9 for cerebellar nuclear cells, considerably greater than the predicted value of 3.1 (Fig. 4B; P < 0.05 for both cell types). The population of unipolar brush cells showed the same trend; the mean I155 Na/I50 Na for transient current at +30 mV was 3.2 and for resurgent current at –30 mV was 5.8 (n = 8 in 50 Na/Ba and n = 5 in 155 Na/Ca).
|
Before considering factors that might be responsible for the high sensitivity of resurgent current to external ionic conditions, we recorded Na currents in granule cells, with a goal of testing the prediction of D'Angelo et al. (2002) that the Na currents of cerebellar granule cells might have a resurgent component. Since the total Na currents are small in these tiny cells, we made recordings only in high Na/Ca. Even under these recording conditions, Na currents, particularly those evoked by repolarizing steps, were often difficult to resolve. In each cell, therefore, we applied a single voltage protocol repeatedly to evoke transient currents at 0 and +30 mV and resurgent currents at –30 mV, and averaged the resulting traces (Fig. 5A). All cells that were tested showed robust resurgent currents (n = 10; Fig. 5, A and B), giving a mean ratio of resurgent to transient conductance of 6.8 ± 0.7%, similar to that in Purkinje neurons (Fig. 4B).
|
; Raman and Bean 2001). It seemed possible, therefore, that manipulations that increase resurgent current might also increase channel availability at moderately negative potentials.
We therefore tested the effects of changing extracellular solutions on steady-state inactivation of Na currents by comparing the steady-state availability curves measured in low Na/Co and high Na/Ca. Channels were conditioned with 200-ms steps to potentials between –90 and 0 mV, a protocol that is likely to favor binding of the fast inactivation gate rather than block by the endogenous blocker (Raman and Bean 2001). After each conditioning step, the availability of channels was assessed with a test step to 0 mV. In Purkinje cells, switching from low Na/Co to high Na/Ca significantly shifted the steady-state inactivation curves by an average of +9.5 mV and also increased their steepness (Fig. 6, A–C; Table 1). Not surprisingly, transient currents evoked during the 200-ms conditioning pulses sometimes showed signs of voltage escape in high Na/Ca, particularly at moderately negative potentials. Inadequate voltage clamp at the beginning of the conditioning step in high Na, however, is likely to lead to underestimation of the depolarizing shift in the inactivation curve, since any loss of voltage control will tend to promote rather than to relieve inactivation.
|
|
) and entry into slow inactivated states is reduced (Townsend and Horn 1997). Additionally, in preparations in which the fast inactivation gate of Na channels has been removed or mutated, high concentrations of external Na+ can relieve channel block by exogenous intracellular blockers, such as strychnine, tetra-alkylammonium ions, and peptides that mimic the inactivation gate, by expulsion, or "knock-off," of the blocker bound in the pore (O'Leary et al. 1994; Shapiro 1977; Tang et al. 1996). Regarding the effects of divalent cations, changes in divalent ion species and concentrations can change surface charge, resulting in shifts of measured voltage-dependence. While these shifts are likely to influence the measurements of steady-state inactivation, they are not large enough to account for the experimental results; for Ba2+, the shift is also of the wrong sign (see DISCUSSION). In the context of the present results, a more significant effect of divalent species may result from a direct blocking action on the Na currents themselves. Specifically, substitution of 2 mM Co2+ for 2 mM Ca2+ can reduce peak Na currents in isolated Purkinje cells by 40–50% (Swensen and Bean 2003), and this Co2+ block is most effective at negative potentials (AM Swensen and BP Bean, unpublished observations). A relief of block by external divalent cations might therefore contribute both to the larger resurgent currents and to the positively shifted availability curves measured in Purkinje and cerebellar nuclear cells in more physiological solutions.
To distinguish the effects of changing divalent ions from the effects of changing Na concentration, we repeated the experiments in Purkinje cells with a new series of solutions. Recordings were made first in low Na/Co and then were repeated in a high Na/Co solution that contained 155 mM Na+ with 2 mM Co2+ (see METHODS), so that the two solutions differed only in the Na+ and TEA+ concentration. Even with divalents held constant, the V1/2 of the inactivation curve shifted by +6 mV when the concentration of Na+ was increased (from –63.5 ± 1.3 to –57.5 ± 1.3 mV, n = 10 paired observations, P < 0.001), whereas the slope factor did not change (from k = 6.7 ± 0.1 to 6.5 ± 0.1, P = 0.11; data not shown). Additionally, the ratio of resurgent conductance at –30 mV to transient conductance at +30 mV measured in high Na/Co was 180 ± 20% of that in low Na/Co, considerably larger than the GHK prediction of 119% (P < 0.05). These results are suggestive of a significant effect on resurgent current and channel availability that is independent of divalent species in the extracellular solution.
Five of the 10 Purkinje cells were also exposed to the high Na/Ca solution used for previous experiments. Consistent with a relief of voltage-dependent block by external Co2+ ions (Swensen and Bean 2003), the resurgent current amplitude at –30 mV in high Na/Ca was 2.1 ± 0.2-fold larger than that in high Na/Co, whereas the transient current at +30 mV increased only to 1.13 ± 0.16 times the amplitude in high Na/Co. Moreover, the V1/2 of inactivation shifted further positive by 2.5 ± 0.6 mV (P < 0.01) and the slope became steeper (from k = 6.4 ± 0.2 to 5.9 ± 0.2 mV, P < 0.0001). Thus the changes in resurgent current and availability curves originally seen with switches from low Na/Co to high Na/Ca may be in part attributed to a block of Na current by Co2+.
Nevertheless, as mentioned above, even with divalents held constant, the increase in resurgent-to-transient conductance ratio with increasing Na+ concentration deviated significantly from the GHK prediction. To test the effect of increasing external Na+ with Ca2+ as the primary divalent, we repeated the experiments in cerebellar nuclear cells, with recordings made in high Na/Ca and low Na/Ca. Both solutions contained 0.3 mM Cd2+ to reduce Ca currents; this concentration of Cd2+ has only a minor effect on Na current amplitude (Swensen and Bean 2003). Also, to minimize entry into slow inactivated states, the steady-state inactivation curve was assessed with 100-ms, rather than 200-ms, conditioning steps. The results obtained for cerebellar nuclear cells were consistent with the observations in Purkinje cells. Specifically, increasing the external Na concentration shifted the V1/2 of inactivation by about +8 mV (Fig. 8A, P < 0.001, from –58.6 ± 1.8 mV in low Na/Ca to –50.7 ± 1.6 mV in high Na/Ca) with no significant effect on k (P = 0.11, from 6.5 ± 2 mV in low Na/Ca to 5.9 ± 2 mV in high Na/Ca, n = 4). Additionally, as shown in Fig. 8B, with the shift to high Na solutions, the transient current amplitude at +30 mV increased 5.7 ± 0.9-fold, not significantly different from the GHK prediction of 4.6-fold (P = 0.3, n = 4). In contrast, the resurgent current amplitude at –30 mV in high Na/Ca was 7.4 ± 3.0 times that in low Na/Ca, much larger than the GHK prediction of a 3.1-fold increase (Fig. 8, B and C, P < 0.01, n = 3). Converting these current amplitudes to conductances gave a resurgent-to-transient conductance ratio in high Na/Ca to low Na/Ca of 241 ± 47% (n = 3). When the fourth cell was exposed to high Na/Ca, the steady-state current at positive potentials also increased, resulting in a tail current at –30 mV that obscured the rising phase of any resurgent current. Consistent with the other cells, however, the amplitude of the current elicited upon repolarization measured after 4 ms (the mean time of the resurgent current peak; Grieco et al. 2002) was 8.3 times the resurgent peak in low Na/Ca. The putative resurgent-to-transient conductance ratio in this cell was therefore 289%, similar to the other cells.
The disproportionate increase in resurgent current, along with the significant positive shift in the inactivation curve with high external Na+, is suggestive of a direct influence of external Na+ ions on the Na current. Since resurgent current apparently flows when an endogenous open channel blocker, which binds on depolarization, is expelled from the pore on repolarization, the preferential enhancement of resurgent current in physiological concentrations of Na+ is consistent with knock-off models of open-channel blockers by permeating ions (see DISCUSSION).
The result that many cerebellar cells have large resurgent Na currents in physiological ionic conditions is suggestive of a role for this component of current in cerebellar function. Observations made in isolated cells, however, cannot necessarily be directly extrapolated to the intact brain. A common concern raised about recordings from isolated cells is that channel properties might change during cell preparation, particularly as a consequence of enzymatic treatment. Although little evidence exists for the idea that Na current kinetics can change during enzymatic dissociation, certain channel types do appear especially susceptible to enzymatic cleavage, thereby reducing current densities (Akaike et al. 1988; Budde et al. 1994; Raman and Trussell 1992). In the worst case, the resurgent currents that we have measured may be a nonphysiological phenomenon induced by the process of cell isolation.
Several observations make this possibility unlikely. First, across isolated neurons, the presence or absence of resurgent current varies with cell class, not with the type of enzyme used for dissociation and not with duration of enzyme exposure (Do and Bean 2003; Pan and Beam 1999; Raman and Bean 1997; the present data). Second, within isolated Purkinje cells, the presence or absence of resurgent current varies with expression of specific Na channel genes (Grieco and Raman 2004; Raman et al. 1997). Third, in any given membrane patch from an isolated Purkinje cell, the presence or absence of resurgent current varies with phosphorylation state of an intracellular, membrane-associated factor (Grieco and Raman 2004; Grieco et al. 2002).
Nevertheless, the concern remains. To address this issue directly, we made whole cell recordings from neurons in slices. Although the voltage control of voltage-gated currents is undoubtedly compromised in the slice preparation, we reasoned that resurgent Na current, if it existed, would be evident on repolarization, and that the modest amplitude and slow kinetics of this component of current would make it reasonably amenable to voltage clamp. Moreover, any resurgent-like current should be blocked by TTX. To minimize K and Ca currents, as well as synaptic currents, 5 mM TEA and 0.3 mM Cd were included in the bath. A K-based pipette solution was used to keep EK near –90 mV to avoid inward K tail currents in the range of potentials where resurgent current would be most likely to flow. The pipette solution included 5 mM TEA to decrease K current amplitudes.
Despite these procedures, the attempt to voltage-clamp Na currents in the slice was unsuccessful in many cells, particularly at temperatures above 30°C; notably, the voltage protocol (Fig. 9, top) often elicited escaping action potentials on the repolarization phase, superimposed on large inward currents. Since the predicted physiological consequence of Na channels with resurgent kinetics is the facilitation of high-frequency firing, however, this behavior was rather suggestive than not of the presence of resurgent current. At temperatures below 30°C, we were able to obtain recordings in seven Purkinje neurons and two granule cells in which the clamp appeared adequate for our measurements.
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
External ions and channel block
Divalent cations have long been known to affect Na channels (Armstrong and Cota 1991; Hille et al. 1975; Nilius 1988), both by affecting surface potential and by direct blocking interactions. Our experiments are consistent with work showing that transient as well as resurgent currents at moderately negative potentials are highly sensitive to the species and concentration of divalent cations (Swensen and Bean 2003). The effects of divalent cations cannot, however, entirely account for the increase in resurgent current and shift in availability curves in quasi-physiological solutions. First, relative to the 2 mM Ca2+ and 2 mM Mg2+ in the modified Tyrode's solution, the 2 mM Co2+ in the low Na solution will shift the surface potential by about +3 mV, whereas 2 mM Ba2+ will shift it by about –3 mV (Hille et al. 1975). Based on data from Hille et al. (1975), to obtain large enough changes in surface potential to account for the observed shifts in the availability curves, the low Na solutions would have to contain 
1 mM Ba2+ or 300 µM Co2+. Second, unipolar brush cells were studied in low Na with 0.3 mM Cd2+ as the Ca channel blocker rather than 2 mM Co2+; 0.3 mM Cd2+ is not expected to block Na currents significantly (Swensen and Bean 2003). Nevertheless, unipolar brush cells showed qualitatively the same changes as Purkinje and cerebellar nuclear cells, suggesting that factors other than divalent ions affected the Na currents. Finally, even with divalents held constant, switching Purkinje cells and cerebellar nuclear cells from low to high Na indicated that resurgent current increased more than predicted by the GHK equation and also produced a positive shift of the availability curve.
This preferential increase in resurgent current by raising external Na is consistent with several previous studies of open- channel blockers. In Shaker K channels, the N-terminus "ball" (Hoshi et al. 1990; Zagotta et al. 1990) acts as an open-channel blocker, inactivating the channel at positive potentials. During recovery at negative potentials, K channels reopen as the ball is expelled from the pore, allowing a brief passage of current before the channel deactivates (Demo and Yellen 1991), much like Na channels that carry resurgent current. Increasing the concentration of external K accelerates recovery from N-type (ball-and-chain) inactivation, probably as a result of permeant ions facilitating unbinding of the pore blocker (Baukrowitz and Yellen 1995; Demo and Yellen 1991; Gómez-Lagunas and Armstrong 1994). In contrast to Shaker K channels, in Na channels, recovery from fast inactivation does not appear to proceed through open states (Kuo and Bean 1994), and the fast inactivation gate does not behave like an open channel blocker (Tang et al. 1996). Nevertheless, in Na channels from squid axon and from cardiac cells, raising the concentration of external Na facilitates the expulsion of exogenously applied intracellular blockers, such as strychnine, tetra-alkylammonium ions, and the KIFMK inactivation peptide (Eaholtz et al. 1994; O'Leary et al. 1996; Shapiro 1977; Tang et al. 1994). Our observation that raising external Na increases resurgent current to a greater extent than predicted by the changes in permeant ion concentration, driving force, and divalent cations is therefore consistent with the idea that permeating Na ions may displace an endogenous, voltage-dependent, open-channel blocker from a binding site within the pore.
Fast and slow inactivation
Increases in external Na+ concentration may affect steady-state availability curves by limiting fast and/or slow inactivation. Evidence for an interaction between external Na and the fast inactivation gate has been obtained in recordings from CA1 pyramidal cells, in which high external Na accelerated recovery from fast inactivation (Kuo and Liao 2000). In those experiments, raising external Na from 50 to 150 mM produced a 4-mV positive shift in the steady-state inactivation curve, similar to our data from Purkinje cells. Not all preparations show a dependence of inactivation on permeating ions; however, in squid axons and cardiac cells, for instance, fast inactivation is relatively insensitive to changes in the concentration of external Na (Armstrong and Bezanilla 1974; Townsend and Horn 1997).
An additional property that may be affected by external Na+ is slow inactivation. In many Na and K channels, prolonged or repeated depolarizations can lead to slow (C-type) inactivation, which is independent of fast inactivation (Cannon 1996) and which appears to result from a constriction of the external mouth of the channel (Yellen et al. 1994). In Shaker K channels, entry into slow inactivated states can be reduced by manipulations that promote binding of K ions in the external mouth of the pore, such that a constant outward K flux is maintained (Baukrowitz and Yellen 1995; Gómez-Lagunas and Armstrong 1994). Similarly, in cardiac Na channels, high concentrations of external Na (or other alkali cations) slow the onset of slow inactivation as well as accelerate recovery from it (Townsend and Horn 1997).
The shifts in the availability curves that we have measured are consistent with an ability of high external Na+ to destabilize binding of the fast inactivation gate and/or to prevent slow inactivation, thereby increasing channel availability at moderately negative potentials. Notably, however, in cerebellar nuclear cells, the shift in inactivation was seen even with relatively short (100-ms) conditioning steps. Possibly, cerebellar nuclear cells enter "slow" inactivated states quite rapidly, and it is entry into these states that is minimized by high Na+; alternatively, the observed shift in the steady-state inactivation curve may result largely from destabilization of fast inactivation. In either case, this increase in availability may itself contribute to the increase in resurgent current amplitude. Relief of slow inactivation may provide more channels that could open, block, and unblock with repolarization, and relief of fast inactivation may make it less likely for stable binding of the fast inactivation gate to curtail the current that flows as channels become unblocked.
Maximization of availability and high-frequency firing
The net result of destabilization of open-channel block, fast inactivation, and/or slow inactivation at negative potentials is that physiological ionic concentrations limit the accumulation of channels in nonconducting states, thereby increasing channel availability. In this context, it is interesting that binding of the fast inactivation gate is also limited at positive potentials by the endogenous open-channel block (Grieco and Raman 2004; Raman and Bean 2001). The effect of the rapid unblock on repolarization, apparently promoted by physiological ion concentrations, is twofold: First, the resurgent current that flows following an action potential may provide some depolarizing drive for a subsequent spike (Do and Bean 2003). Second, the unblocked channels deactivate into "available" states, rather than remaining refractory (Raman and Bean 1997). The cycle of opening, blocking, and unblocking thereby minimizes the refractory period, a feature that may be suitable for enabling high-frequency firing (Khaliq et al. 2003; Raman and Bean 2001).
In fact, Purkinje cells, cerebellar nuclear cells, unipolar brush cells, and granule cells all produce action potentials at high rates. Purkinje neurons are not only spontaneously active at frequencies near 50 Hz, but fire at frequencies >100 Hz during cerebellar behaviors (e.g., Gilbert and Thach 1977; Kitazawa et al. 1998; Thach 1968). Cerebellar nuclear neurons tend to fire at lower rates than Purkinje cells, but they are also spontaneously active and can maintain firing frequencies of >50 Hz for prolonged periods (Aizenman and Linden 1999; Jahnsen 1986; Ledoux et al. 1998; Thach 1968). Although the electrical activity of unipolar brush cells has been studied less extensively, it is known that their excitatory postsynaptic currents have an unusually long component that results from entrapment of glutamate in the synaptic cleft (Rossi et al. 1995). The resulting excitatory postsynaptic potentials can be large and long, evoking high-frequency action potentials superimposed on the synaptic depolarization (Rossi et al. 1995). Finally, although granule cells do not appear to be spontaneously active, they fire at frequencies exceeding 100 Hz in response to excitatory stimuli (Chadderton et al. 2004; D'Angelo et al. 1998; Hartmann and Bower 2001; Mitchell and Silver 2003).
The extent to which rapid firing depends on the expression of Na channels with resurgent kinetics may vary across these cerebellar neurons. Conversely, the efficacy of resurgent current in promoting firing will depend strongly on the other ion channels expressed in the particular cells. Nevertheless, it is striking that the four classes of cerebellar neurons that we have tested share the property of producing resurgent current, particularly since this property initially seemed to be an unusual, if not unique, specialization of Purkinje cells. The idea that Purkinje cells were exceptional seemed plausible in part because resurgent current is absent from several neurons that express the same Na channel subunits as Purkinje cells (Maurice et al. 2001; Pan and Beam 1999; Raman and Bean 1997), and in part because voltage-gated Na channels in expression systems lack a resurgent component (Smith et al. 1998). Recently, however, resurgent current was identified in neurons of the subthalamic nuclei (Do and Bean 2003), and the present results reveal that the current is a typical, rather than an unusual, feature of cerebellar neurons. The list of neurons with resurgent kinetics appears likely to grow. The widespread occurrence of this once-novel component of Na current suggests that the expression of a blocking factor that confers resurgent kinetics to Na channels may be a commonly utilized mechanism by neurons whose signaling involves sustained periods of high-frequency firing.
note added in proof
A recent abstract has reported resurgent Na current in granule cells in cerebellar slices (Magistretti et al. FENS Abstracts, 2004).
|
GRANTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
|
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.
Address for reprint requests and other correspondence: I. M. Raman Dept. of Neurobiology and Physiology, 2205 Tech Dr., Northwestern Univ., Evanston, IL 60208 (E-mail: i-raman{at}northwestern.edu).
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONGRANTSACKNOWLEDGMENTSREFERENCES |
|---|
Akaike N, Kaneda M, Hori N, and Krishtal OA. Blockade of N-methyl-D-aspartate response in enzyme-treated rat hippocampal neurons. Neurosci Lett 87: 75–79, 1988.[CrossRef][ISI][Medline]
Armstrong CM and Bezanilla F. Charge movement associated with the opening and closing of the activation gates of the Na channels. J Gen Physiol 63: 533–552, 1974.
Armstrong CM and Cota G. Calcium ion as a cofactor in Na channel gating. Proc Natl Acad Sci USA 88: 6528–6531, 1991.
Baukrowitz T and Yellen G. Modulation of K+ current by frequency and external [K+]: a tale of two inactivation mechanisms. Neuron 15: 951–960, 1995.[CrossRef][ISI][Medline]
Bezanilla F and Armstrong CM. Inactivation of the sodium channel. I. Sodium current experiments. J Gen Physiol 70: 549–566, 1977.
Budde T, White JA, and Kay AR. Hyperpolarization-activated Na(+)-K+ current (Ih) in neocortical neurons is blocked by external proteolysis and internal TEA. J Neurophysiol 72: 2737–2742, 1994.
Cannon SC. Sodium channel defects in myotonia and periodic paralysis. Annu Rev Neurosci 19: 141–164, 1996.[CrossRef][ISI][Medline]
Chadderton P, Margrie TW, and Hausser M. Integration of quanta in cerebellar granule cells during sensory processing. Nature 428: 856–860, 2004.[CrossRef][Medline]
D'Angelo E, De Filippi G, Rossi P, and Taglietti V. Ionic mechanism of electroresponsiveness in cerebellar granule cells implicates the action of a persistent sodium current. J Neurophysiol 80: 493–503, 1998.
D'Angelo E, Nieus T, Maffei A, Armano S, Rossi P, Taglietti V, Fontana A, and Naldi G. Theta-frequency bursting and resonance in cerebellar granule cells: experimental evidence and modeling of a slow K+-dependent mechanism. J Neurosci 21: 759–770, 2002.
Demo SD and Yellen G. The inactivation gate of the Shaker K+ channel behaves like an open-channel blocker. Neuron 7: 743–753, 1991.[CrossRef][ISI][Medline]
Do MT and Bean BP. Subthreshold sodium currents and pacemaking of subthalamic neurons: modulation by slow inactivation. Neuron 39: 109–120, 2003.[CrossRef][ISI][Medline]
Eaholtz G, Scheuer T, and Catterall WA. Restoration of inactivation and block of open sodium channels by an inactivation gate peptide. Neuron 12: 1041–1048, 1994.[CrossRef][ISI][Medline]
Gilbert PF and Thach WT. Purkinje cell activity during motor learning. Brain Res 128: 309–328, 1977.[CrossRef][ISI][Medline]
Gómez-Lagunas F and Armstrong CM. The relation between ion permeation and recovery from inactivation of ShakerB K+ channels. Biophys J 67: 1806–1815, 1994.
Grieco TM, Afshari FS, and Raman IM. A role for phosphorylation in the maintenance of resurgent sodium current in cerebellar Purkinje neurons. J Neurosci 22: 3100–3107, 2002.
Grieco TM and Raman IM. Production of resurgent current in NaV1.6-null Purkinje neurons by slowing sodium channel inactivation with 
-pompilidotoxin. J Neurosci 24: 35–42, 2004.
Hartmann MJ and Bower JM. Tactile responses in the granule cell layer of cerebellar folium crus IIa of freely behaving rats. J Neurosci 21: 3549–3563, 2001.
Hille B, Woodhull AM, and Shapiro BI. Negative surface charge near sodium channels of nerve: divalent ions, monovalent ions, and pH. Philos Trans R Soc Lond B Biol Sci 270: 301–318, 1975.[ISI][Medline]
Hodgkin AL and Huxley AF. A quantitative description of membrane current and its application to conduction and excitation in nerve. J Physiol 117: 500–544, 1952.
Hoshi T, Zagotta WN, and Aldrich RW. Biophysical and molecular mechanisms of Shaker potassium channel inactivation. Science 250: 533–538, 1990.
Deceased 29 August 2001.