Kammermeier, Paul J. and Stephen W. Jones. High-voltage-activated calcium currents in neurons acutely isolated from the ventrobasal nucleus of the rat thalamus. J. Neurophysiol. 77: 465–475, 1997. We studied the high-voltage-activated (HVA) calcium currents in cells isolated from the ventrobasal nucleus of the rat thalamus with the use of the whole cell patch-clamp technique. Low-voltage-activated current was inactivated by the use of long voltage steps or 100-ms prepulses to −20 mV. We used channel blocking agents to characterize the currents that make up the HVA current. The dihydropyridine (DHP) antagonist nimodipine (5 μM) reversibly blocked 33 ± 1% (mean ± SE), and ω-conotoxin GVIA (1 μM) irreversibly blocked 25 ± 5%. The current resistant to DHPs and ω-conotoxin GVIA was inhibited almost completely by ω-conotoxin MVIIC (90 ± 5% at 3–5 μM) and was partially inhibited by ω-agatoxin IVA (54 ± 4% block at 1 μM). We conclude that there are at least four main HVA currents in thalamic neurons: N current, L current, and two ω-conotoxin MVIIC-sensitive currents that differ in their sensitivity to ω-agatoxin IVA. We also examined modulation of HVA currents by strong depolarization and by G protein activation. Long (∼1 s), strong depolarizations elicited large, slowly deactivating tail currents, which were sensitive to DHP antagonists. With guanosine 5′-O-(3-thiotriphosphate) (GTP-γ-S) in the intracellular solution, brief (∼20 ms), strong depolarization produced a voltage-dependent facilitation of the current (44 ± 5%), compared with cells with GTP (22 ± 7%) or guanosine 5′-O-(2-thiodiphosphate) (7 ± 4%). However, the HVA current was inhibited only weakly by 100 μM acetylcholine (8 ± 4%). Effects of the γ-aminobutyric acid-B agonist baclofen were variable (3–39% inhibition, n = 12, at 10–50 μM).
Thalamic relay neurons possess both low-voltage-activated (LVA) and high-voltage-activated (HVA) calcium channels (Hernández-Cruz and Pape 1989). The LVA (T-type; I T) channels have been well characterized (Coulter et al. 1989; Crunelli et al. 1989; Huguenard and Prince 1992; Suzuki and Rogawski 1989). That conductance underlies the low-threshold calcium spike. When the cells are driven to negative voltages, I T recovers from inactivation and induces the cells to burst (McCormick and Pape 1990). This bursting mode has been associated with the sleep state and with seizures (Steriade and Llinas 1988).
When thalamic relay neurons rest at less negative voltages, I T inactivates, abolishing the low-threshold calcium spike. The cells then enter a tonic firing mode in which the output of the cells is shaped by input from sensory systems (Steriade and Llinas 1988). However, much less is known about the calcium conductances that are active in this state, the HVA currents, although they are particularly important for their effects on cell excitability in addition to their roles in such processes as transmitter release and signaling.
Several procedures can be used to separate HVA from LVA currents, including selective inactivation of LVA current by weak depolarizations. One characteristic feature of LVA currents is slow channel closing (deactivation) on hyperpolarization, producing tail currents ≥10-fold slower than HVA tails (Huguenard and Prince 1992; Matteson and Armstrong 1986). However, in some cell types, the L-type HVA current can be potentiated by strong depolarization or by dihydropyridine (DHP) agonists, effects that also produce slowly deactivating tail currents (Fleig and Penner 1995; Forti and Pietrobon 1993; Kavalali and Plummer 1996; Pietrobon and Hess 1990; Slesinger and Lansman 1991; Thibault et al. 1993).
Neurons generally have several types of HVA calcium currents (Dunlap et al. 1995; Olivera et al. 1994). At the level of whole cell currents, they can be distinguished most clearly by their pharmacology, because kinetic criteria (e.g., inactivation rates) can be misleading (Plummer et al. 1989). Studies on peripheral neurons distinguished L current, sensitive to DHP agonists and antagonists, from N current, blocked potently by ω-conotoxin GVIA. CNS neurons have N and L currents, plus other HVA currents. P current, the predominant HVA current of cerebellar Purkinje neurons, is blocked potently by the spider toxin ω-agatoxin (ω-Aga) IVA (Mintz et al. 1992). Other proposed HVA currents include Q (blocked potently by ω-conotoxin MVIIC and weakly by ω-Aga IVA) (Randall and Tsien 1995; Zhang et al. 1993) and R (insensitive to DHPs and toxins) (Ellinor et al. 1993). Several recent studies have used these pharmacological tools to separate HVA currents in a variety of neurons (e.g., Eliot and Johnston 1994; Randall and Tsien 1995; Rothe and Grantyn 1994), but it remains to be established whether the P, Q, and R classification is generally applicable (Dunlap et al. 1995; Olivera et al. 1994).
In many neurons, HVA calcium currents can be inhibited by neurotransmitters acting through G proteins (Dunlap and Fischbach 1981; Hille 1994). Probably the most thoroughly studied mechanism for modulation of neuronal calcium channels is the G protein-mediated inhibition seen in, e.g., sympathetic neurons (Hille 1994; Jones and Elmslie 1992). One useful test for this mechanism is partial reversal of the inhibition by strong depolarization (Bean 1989; Elmslie et al. 1990; Grassi and Lux 1989; Kasai and Aosaki 1989). This type of modulation is widespread (Brown and Birnbaumer 1990) and can target different HVA channels, including N (Plummer et al. 1989), P (Mintz and Bean 1993), and Q (Mackie et al. 1995), but usually not L (Kasai and Aosaki 1989; Plummer et al. 1989). Voltage-dependent modulation of calcium channels can be mediated by different receptors and even different G proteins (Ehrlich and Elmslie 1995; Hille 1994). Thus it is possible that calcium channel modulation may play a role in thalamic relay neurons.
In this study the HVA current and its components were separated with the use of pharmacological methods. We have found that at least four different currents make up the HVA. In addition, we examined the ability of guanosine 5′-O-(3-thiotriphosphate) (GTP-γ-S), a non-hydrolyzable GTP analogue, as well as acetylcholine (ACh) and baclofen, to modulate the HVA current. We also found that long, strong depolarizations can potentiate the L current, associated with dramatic slowing of tail currents.
Cells were isolated by a method based on that of Swartz and Bean (1992). Neonatal rats (Sprague-Dawley, aged 7–16 days) were anesthetized with ether and decapitated. The brains were quickly removed and placed in a N-2-hydroxyethylpiperazine-N′-2-ethanesulfonic acid (HEPES)-buffered solution at 0°C. After removal of the cerebellum, a midsagittal cut was made and each half of the brain was mounted into a Vibratome. Coronal slices (400 μm) were then made through the thalamus. Slices containing the ventrobasal thalamus were placed into the same HEPES-buffered solution at 35°C with type XXIII protease (Sigma, 3 mg/ml) for 6 min. Next, the slices were removed from the enzyme-containing solution and placed into a HEPES-buffered solution (room temperature) containing bovine serum albumin and trypsin inhibitor (Sigma, 1 mg/ml each). When needed, each slice was removed, and the ventrobasal thalamus was dissected out and placed into a Ba2+-free transfer solution to avoid precipitation of BaSO4. The tissue was gently triturated and placed in a 35-mm dish, and the extracellular recording solution was then washed in. Cells appeared triangular in shape and possessed short processes. All solutions were bubbled with 100% oxygen until placed in the dish for recording.
The HEPES-buffered saline used in the dissociation procedure contained (in mM) 82 Na2SO4, 30 K2SO4, 5 MgCl2, 10 HEPES, and 10 glucose. The transfer solution contained (in mM) 145 tetraethylammonium (TEA) Cl, 2 CaCl2, and 10 HEPES. The intracellular recording solution for all experiments contained (in mM) 117 TEA Cl, 4.5 MgCl2, 9 HEPES, 9 ethylene glycol-bis(β-aminoethyl ether)-N,N,N′,N′-tetraacetic acid, 14 creatine phosphate, 4 tris(hydroxymethyl)aminomethane ATP, and 0.3 Li GTP. Experiments in which GTP-γ-S and guanosine 5′-O-(2-thiodiphosphate) (GDP-β-S) were used included these compounds at 0.1 and 1.0 mM, respectively, replacing GTP. The extracellular recording solution for most experiments contained (in mM) 120 TEA Cl, 25 BaCl2, and 10 HEPES. Where noted, a low-barium extracellular solution was used, which contained (in mM) 145 TEA Cl, 2 BaCl2, and 10 HEPES. All solutions were adjusted to pH 7.4. Solutions containing ω-Aga IVA also contained 1 mg/ml cytochrome c to reduce binding of peptides to plastic tubing and dishes (Mintz et al. 1992). Experiments in which ω-conotoxin MVIIC and GVIA were used were performed in 2 mM Ba2+, unless otherwise indicated. To prevent blockade of muscarinic ACh receptors (Caulfield 1991), extracellular TEA was replaced by N-methyl-d-glucamine in experiments on modulation of calcium currents by ACh or baclofen.
All recordings were made in the whole cell patch-clamp configuration. Pipettes were made from Garner EN-1 glass (0.86 mm ID, 1.5 mm OD). Pipette resistances were 2–5 MΩ, yielding series resistances of 5–15 MΩ before compensation, which was 80% in all recordings. Data acquisition and analysis were performed with the use of pClamp software (Axon Instruments, Foster City, CA). Data were recorded with the use of an Axopatch 200 patch-clamp amplifier from Axon Instruments and digitized with an Axon TL-1 A-D converter. The holding potential was −80 mV. Currents were sampled at 1–33 kHz after filtering at 1/5 the sampling rate. Data were leak subtracted during analysis (P/4). Values are given as means ± SE. Where statistical significance is noted, the two-tailed Student's t-test was used, with P < 0.05 considered to be significant.
Nearly all ventrobasal thalamus neurons project to the cortex, with <1% γ-aminobutyric acid (GABA)-positive local circuit neurons (Harris and Hendrickson 1987). Correspondingly, nearly all cells tested had similar calcium currents. Immediately after the whole cell configuration was established, the LVA and HVA currents were of comparable magnitude (<200 pA). Over the course of the next 3–10 min the HVA currents grew larger and eventually stabilized (at ∼700–2,000 pA), whereas the LVA currents ran up only slightly or not at all. After runup, calcium currents often showed a variable rundown that proceeded much more slowly than the initial runup.
ω-Aga IVA was obtained from Pfizer (Groton, CT). ω-Conotoxin GVIA, ω-conotoxin MVIIC, and nimodipine wereobtained from Research Biochemicals (Natick, MA). Additional ω-conotoxin GVIA was obtained from Peninsula Laboratories (Belmont, CA); ω-conotoxin GVIA from the two suppliers had similar effects. Protease XXIII, trypsin inhibitor, bovine serum albumin, ACh chloride, baclofen, and cytochrome c were obtained from Sigma Chemical (St. Louis, MO). (+)202-791 was obtained from Sandoz (Basel, Switzerland). GTP-γ-S and GDP-β-S were obtained from Boehringer Mannheim (Indianapolis, IN). Nimodipine and (+)202-791 were stored as stock solutions in ethanol. The final ethanol concentration in the extracellular solution was ≤0.04%, and that concentration of ethanol had no effect on calcium currents (n = 5). Other drugs were dissolved in water.
Isolation of HVA calcium current
From a holding potential of −80 mV, small depolarizing steps elicited a rapidly inactivating current (I T). At more positive voltages a noninactivating current was predominant (Fig. 1 A). The current-voltage relation revealed a large inward current, peaking near 0 mV, with a smaller “shoulder” at more negative voltages. When the current was measured at the end of long voltage steps (100 ms), after most of the I T was inactivated, the shoulder was selectively reduced(Fig. 1 B).
I T deactivates slowly compared with HVA calcium currents (Huguenard and Prince 1992; Matteson and Armstrong 1985). Indeed, tail currents measured at −50 mV after 10-ms depolarizations exhibited a biexponential time course (Fig. 2 A). When the step was preceded by a 100-ms prepulse to −20 mV, where much I T but little HVA current is activated, the slow component of the tail current was greatly reduced (Fig. 2 B). Therefore we could isolate the HVA calcium current either by measuring at the end of long voltage steps (Fig. 1) or with the use of prepulses to −20 mV (Fig. 2).
The activation curve for the fast tail had the usual sigmoid shape, with half-maximal activation near +10 mV, and was unaffected by the prepulses to −20 mV (Fig. 2 C). The activation curve for the slow tail current was more complex (Fig. 2 D). One component activated at negative voltages, roughly 30 mV negative to the fast HVA tails, as expected for I T. However, depolarization beyond 0 mV activated additional current, with a shallow voltage dependence. The HVA component of the slow tail was resistant to inactivation (Fig. 2 D, □), and is considered in more detail below.
Pharmacology of HVA current
CURRENT SENSITIVE TO DHPS.
To test for the presence of L current, we examined the effects of DHPs. The DHP agonist (+)202-791 enhanced the steady-state current at 0 mV by 124 ± 24% (mean ± SE) at 300 nM (n = 4), and dramatically slowed deactivation (Fig. 3 A). The effect of (+)202-791 was concentration dependent (Fig. 3, A and B), with an apparent dissociation constant of 160 nM (Fig. 3 C). As in other cells (Hess et al. 1984), DHP agonists also shifted the peak of the current-voltage curve to more negative voltages (Fig. 3 D).
To determine the concentration of DHP antagonist required to inhibit L current, we attempted to reverse the effect of (+)202-791 by application of nimodipine (Fig. 4). The slow tail currents induced by 300 nM (+)202-791 were blocked incompletely by 1 μM nimodipine (52 ± 16%,n = 3), with significantly more block at 5 μM (88 ± 3%,n = 4). Thus high nimodipine concentrations are required to fully block L current in these cells.
Addition of 5 μM nimodipine to the control current induced a 33.2 ± 1.1% (n = 12) reduction of the steady-state current recorded at 0 mV (Fig. 5). This result suggests that 1/3 of the HVA current is DHP-sensitive L current.
CURRENT SENSITIVE TO ω-CONOTOXIN GVIA.
We also examined the effects of the specific N current blocker ω-conotoxin GVIA. At 1 μM, the toxin irreversibly inhibited 25 ± 5% (n = 3) of peak HVA current (Fig. 6), recorded in 2 mM Ba2+ to avoid inhibition of toxin binding by high concentrations of divalent cations (Boland et al. 1994). However, the amount of inhibition was similar in 25 mM Ba2+, because peak current was inhibited by 24 ± 2% (n = 4, data not shown). Full block of N current by 1 μM ω-conotoxin GVIA in 25 mM Ba2+ is consistent with the estimated dissociation constant of 0.2–0.4 μM in 112 mM Ba2+ (Boland et al. 1994).
EFFECTS OF ω-AGA IVA AND ω-CONOTOXIN MVIIC.
At low concentrations, the funnel web spider toxin ω-Aga IVA is a potent and selective P current blocker, with an apparent dissociation constant of 2 nM (Mintz et al. 1992). In thalamic neurons, application of 100 nM ω-Aga IVA for 5–8 min had little or no effect on the total HVA current (8 ± 3% inhibition, n = 5) (Fig. 7 A). In contrast, the HVA current of cerebellar Purkinje cells was inhibited by 91 ± 2% (n = 3) under identical conditions (Fig. 7 B), confirming that the ω-Aga IVA preparation used here is active against P current.
A large amount of current remained in the presence of nimodipine and ω-conotoxin GVIA (51 ± 4%, n = 8). Nearly all of that current was inhibited by 3–5 μM ω-conotoxin MVIIC (90 ± 5%, n = 4; Fig. 8 A) when currents were recorded in the presence of 2 mM Ba2+ to speed toxin block (McDonough et al. 1996).
Because ω-conotoxin MVIIC has been reported to block N, P, and Q currents, but not L or R currents (Hillyard et al. 1992; Randall and Tsien 1995), the current resistant to nimodipine and ω-conotoxin GVIA could be either P or Q. It has been proposed that P and Q currents can be distinguished by their affinity for ω-Aga IVA (Randall and Tsien 1995). To test that hypothesis, we applied ω-Aga IVA to cells in the presence of nimodipine and ω-conotoxin GVIA. Under these conditions, 100 nM and 1 μM ω-Aga IVA blocked 26 ± 7% and 54 ± 4% (n = 5) of the current, respectively (Fig. 8 B).
Block with 100 nM ω-Aga IVA is slow, and the effect did not seem to reach steady state during the applications used, which lasted 2–4.5 min (Fig. 8 B). Two additional cells were tested with 100 nM ω-Aga IVA for 9–10 min, followed by brief application of 1 μM ω-Aga IVA. A train of strong depolarizations (50-ms pulses to +110 mV for 4 s at 10 Hz) was given during the wash, which speeded reversal of the block (Mintz et al. 1992; Randall and Tsien 1995). Assuming that incomplete reversal reflected rundown during this long protocol, inhibition was 54 and 39% at 100 nM and 72 and 47% at 1 μM in the two cells, respectively.
Modulation by G proteins and neurotransmitters
MODULATION BY INTRACELLULAR GTP-γ-S.
In many cells, activation of G proteins causes a voltage-dependent inhibition of calcium channels (Hille 1994). The inhibition can be partially relieved by strong depolarization. This can be seen as an increase in current following brief, strong depolarizing prepulses. That facilitation is a convenient measure of the modulation, especially when it is induced by irreversible means such as dialysis of GTP-γ-S.
To test for this kind of modulation, we examined facilitation in the presence of intracellular GTP-γ-S, GTP, or GDP-β-S (Fig. 9 A). Cells were held at −80 mV and stepped to 0 mV for 50 ms, with or without a 20-ms prepulse to +80 mV. With 100 μM GTP-γ-S, the prepulse to +80 mV increased the current by 44 ± 5% (n = 5), measured after 5 min to allow time for the pipette solution to dialyze into the cell. In contrast, facilitation was 7 ± 4% (n = 5) when cells were dialyzed with 1 mM GDP-β-S, and 22 ± 7% (n = 5) with 300 μM GTP. Facilitation with GTP-γ-S was significantly different from the values with GDP-β-S or GTP; there was no significant difference between facilitation with GDP-β-S and GTP.
MODULATION BY ACh AND BACLOFEN.
The effect of GTP-γ-S suggested that the calcium current might also be inhibited by activation of G protein-coupled receptors. We tested muscarinic and GABAB agonists, because they are known to inhibit calcium channels in other neurons (Dunlap and Fischbach 1981; Wanke et al. 1987) and ACh and GABA are neurotransmitters known to act on thalamic relay neurons (Sherman and Koch 1990). At 100 μM, ACh induced small, reversible inhibition of the HVA current (8.7 ± 3.7%, n = 6). Effects of 10 μM baclofen ranged from 3 to 34% inhibition (16 ± 5%, n = 7). At 50 μM, baclofen inhibited 28 ± 6% (n = 5) of the current, but in two cells tested at both concentrations, 10 and 50 μM baclofen had comparable effects (31–39% inhibition). The inhibition was partially reversed by strong depolarization (Fig. 9 B).
Potentiation of L current by strong depolarization
After inactivation of I T, tail currents measured at −50 mV following steps to negative voltages were well fit by a single, fast exponential time course (Fig. 2 B). However, following steps to more positive voltages, a slower component to the tail current emerged (Fig. 2 D). This was not I T, because it was activated only by strong depolarizations and was present even when I T was inactivated by a 100-ms prepulse to −20 mV. Furthermore, the slow tail current, which is relatively small following short voltage steps (Fig. 2 D), is dramatically increased following long depolarizations (Fig. 10 A), which would also inactivate I T. Because long steps to +70 mV were not always well tolerated by the cells, we were not able to fully define the time course, but it is clear that the effect developed very slowly, requiring >1 s to reach half maximal potentiation at +70 mV (Fig. 10 B).
We next performed pharmacological tests to determine the identity of the slow tail currents. At 1 mM, Co2+, a nonselective calcium channel blocker, strongly reduced both the fast and slow components of the tail current, suggesting that the slow tails are indeed through calcium channels (Fig. 11 A). That experiment does not fully rule out a current activated by Ba2+ influx, e.g., Ca2+-dependent Cl− current, but there should be little Ba2+ influx at +70 mV.
Strong depolarization can potentiate L current, producing slow tail currents (Pietrobon and Hess 1990). The slow tails in thalamic neurons were strongly reduced by 5 μM nimodipine (88 ± 5% inhibition, n = 5), suggesting that they reflect activity of L channels (Fig. 11 B). ω-Conotoxin GVIA had no obvious effect on the slow tails (data not shown).
This study focuses primarily on determining which types of calcium currents compose the HVA currents in thalamic relay neurons. In addition, we have further investigated some properties of these currents in an attempt to begin to understand their physiological roles.
Our results are consistent with the hypothesis that DHPs, ω-conotoxin GVIA, and ω-Aga IVA target different calcium channel types (Mintz et al. 1992). The effects of 5 μM nimodipine and 1 μM ω-conotoxin GVIA were close to additive (33 ± 1% and 25 ± 5% inhibition separately, vs. 49 ± 4% when applied together). Brief application of 100 nM ω-Aga IVA inhibited 26 ± 7% of the current remaining in nimodipine plus ω-conotoxin GVIA, which would correspond to 13% of the total current, in good agreement with the measured value of 8 ± 3%. These results imply that nonspecific effects of DHPs are not significant here, although high concentrations can even block Na+ and K+ channels (Jones and Jacobs 1990; Nerbonne and Gurney 1987; Yatani and Brown 1985), and overlap between DHP-sensitive and ω-Aga-IVA-sensitive currents has been reported in some other neurons (Brown et al. 1994; Pearson et al. 1995). High concentrations of DHP antagonists appear to be necessary, because the effect of DHP agonists was not fully reversed by 1 μM nimodipine (Fig. 4). If the interaction of DHP agonists and antagonists is competitive, that test would underestimate the potency of nimodipine by about threefold, because the concentration of (+)202-791 was twice the apparent dissociation constant. The relatively negative holding potential (−80 mV) would contribute to the low potency (Bean 1984), but Marchetti et al. (1995) found stronger block by nimodipine (apparent dissociation constant = 50 nM) in cerebellar granule cells, even from −80 mV.
The effect of ω-conotoxin GVIA was generally comparable with previous reports on thalamic neurons (Table 1). One previous study reported no effect of 100 nM ω-Aga IVA (Guyon and Leresche 1995), compared with our value of 8 ± 3%. Reported effects of DHP antagonists are more variable (Table 1), possibly resulting from incomplete block at low concentrations and/or nonspecific actions at high concentration. Indeed, Guyon and Leresche (1995) concluded that 10 μM nifedipine partially blocked N current. In our experiments, 5 μM nimodipine produced nearly complete block of DHP-agonist-induced tail currents, and the effects of nimodipine and ω-conotoxin GVIA were nearly additive, suggesting that L current is blocked effectively and selectively by 5 μM nimodipine.
The current resistant to DHPs and ω-conotoxin GVIA was blocked almost completely (90 ± 5%) by ω-conotoxin MVIIC. In contrast, ω-Aga IVA produced only 54 ± 4% block at 1 μM. If the DHP- and ω-conotoxin-GVIA-resistant current is a single homogeneous channel population blocked by 54% at 1 μM ω-Aga IVA, only 11% block would be expected at 100 nM, which is significantly different from the observed value (26 ± 7%, for brief applications). If block at 100 nM did not reach steady state, the discrepancy would be even larger. (Correction for the 10% of current that is resistant to ω-conotoxin MVIIC does not greatly affect the calculation.) The simplest explanation is existence of (at least) two channel populations that are sensitive to ω-conotoxin MVIIC, and resistant to DHPs and ω-conotoxin GVIA, but differ in their affinity for ω-Aga IVA (Randall and Tsien 1995). It is possible that one of the channels is completely resistant to ω-Aga IVA (Mackie et al. 1995; McDonough et al. 1996). Because steady-state block may not have been reached at low concentrations of ω-Aga IVA, it is difficult to compare the potency of block with previously reported P and Q currents. It is clear, however, that block by ω-Aga IVA is slower for thalamic neurons than for cerebellar Purkinje cells (Figs. 7 and 8).
We conclude that the HVA current of thalamic relay neurons is composed of at least four main components: nimodipine-sensitive L current (33%), ω-conotoxin GVIA-sensitive N current (25%), and two ω-conotoxin MVIIC-sensitive currents differing in sensitivity to ω-Aga IVA (46%). The small amount of current remaining in the presence of nimodipine, ω-conotoxin GVIA, and ω-conotoxin MVIIC (5% of the total current) may be an R current.
Voltage-dependent modulation of calcium currents via G proteins is a common and well-studied mechanism for influencing the activity of calcium channels (Hille 1994). GTP-γ-S, which should irreversibly activate all G proteins, produced substantial facilitation (44 ± 5%). There was some facilitation with GTP alone (22 ± 7%), presumably reflecting basal G protein activation (Ikeda 1991) because it was not observed with GDP-β-S. Note that any voltage-independent modulation induced by activation of G proteins would not be detected in these experiments.
It is not clear why the effect of ACh was small and the effect of baclofen was variable. Formenti et al. (1995a,b) found ∼50% inhibition by enkephalin and ACh in most ventrobasal thalamus neurons. In dorsal lateral geniculate, Guyon and Leresche (1995) found 17% inhibition by 10 μM baclofen and 40% inhibition by 50 μM baclofen. That is similar to our results, but in our experiments the difference between effects at the two concentrations could reflect variability among batches of cells. These studies have used rather different intra- and extracellular recording solutions, which could affect second-messenger systems. Another factor is the relatively large amount of basal facilitation (with intracellular GTP alone) in our experiments, which would partially occlude the effect of receptor-mediated G protein activation.
L current potentiation
In many studies, strong depolarization potentiates activity of L-type calcium channels. (We use the term “potentiation” here to avoid confusion with the G protein-mediated “facilitation” discussed above.) It appears that there are multiple mechanisms for potentiation. One pathway involves protein phosphorylation, and increases the peak inward current but does not necessarily produce slow tail currents (Artalejo et al. 1990; Bourinet et al. 1994; Sculptoreanu et al. 1993). Slow tail currents, or the corresponding single channel activity following strong depolarizations, have been reported in cardiac cells (Pietrobon and Hess 1990), in skeletal muscle (Flieg and Penner 1995), and in several neurons (Forti and Pietrobon 1993; Kavalali and Plummer 1996; Slesinger and Lansman 1991; Thibault et al. 1993). However, potentiation in thalamic relay neurons is considerably slower (Fig. 10) than in previous studies, e.g., a time constant of 300–400 ms at +70 mV in cardiac cells (Pietrobon and Hess 1990).
The physiological role of L current potentiation is unclear. Obviously, Ca2+ entry that is maintained for several milliseconds after repolarization could have powerful effects on the electrophysiology and biochemistry of a neuron, but large tail currents were produced only by highly unphysiological stimuli (depolarization to +70 mV for 1–2 s). One possibility is that even small slow tail currents, produced by more modest stimulation, could produce significant Ca2+ entry. L channel openings at negative voltages have been reported in response to trains of action-potential-like depolarizations (Slesinger and Lansman 1991). It is also conceivable that the large slow tails reflect a mode of calcium channel activity that can also be triggered by other, as yet unidentified, physiological means.
We are grateful for Pfizer for supplying ω-Aga IVA. We thank Dr. M. A. Werz for participation in some early experiments in this project, and Dr. Subba Shankar for assistance with developing the cell isolation procedure.
This work was supported in part by National Institute of Neurological Disorders and Stroke Grant NS-24471 to S. W. Jones, who was an Established Investigator of the American Heart Association.
Address for reprint requests: S. W. Jones, Dept. of Physiology and Biophysics, Case Western Reserve University, Cleveland, OH 44106.