The Journal of Neurophysiology Vol. 80 No. 4 October 1998, pp. 1981-1988
Copyright ©1998 by the American Physiological Society

Group I Metabotropic Glutamate Receptors Mediate Slow Inhibition of Calcium Current in Neocortical Neurons

Rod J. Sayer

Department of Physiology, University of Otago, New Zealand

	ABSTRACT

Abstract Introduction Methods Results Discussion References

Sayer, Rod J. Group I metabotropic glutamate receptors mediate slow inhibition of calcium current in neocortical neurons J. Neurophysiol. 80: 1981-1988, 1998. Metabotropic glutamate receptor (mGluR)-mediated inhibition of high-voltage-activated Ca²⁺ currents was investigated in pyramidal neurons acutely isolated from rat dorsal frontoparietal neocortex. Whole cell recordings were made at 30-32°C, with Ca²⁺ as the charge carrier. Selective agonists were used to classify the subgroup of mGluRs mediating the response. Ca²⁺ currents were inhibited by (1S,3R)-1-aminocyclopentane-1,3-dicarboxylic acid (1S,3R-ACPD) and by the group I agonist (RS)-3,5-dihydroxyphenylglycine (DHPG) but not by the group II agonist (2S,2'R,3'R)-2-(2',3'-dicarboxycyclopropyl)glycine (DCG-IV) or the group III agonist L(+)-2-amino-4-phosphonobutryic acid (L-AP4). (2S,1'S,2'S)-2-(carboxycyclopropyl)glycine (L-CCG-I) was effective at 10 and 100 µM but not at 1 µM, consistent with involvement of group I mGluRs. Variable results were obtained with the putative mGluR5-selective agonist (RS)-2-chloro-5-hydroxyphenylglycine (CHPG) and the putative mGluR1-selective antagonist (S)-4-carboxyphenylglycine [(S)-4CPG], indicating that the group I mGluR subtypes may vary between cells or that these compounds were activating other receptors. The actions of (+)--methyl-4-carboxyphenylglycine [(+)-MCPG] were consistent with it being a low-potency antagonist. Several features of the Ca²⁺ current inhibition evoked by DHPG distinguished it from the rapid modulation typical of a direct action of G proteins on Ca²⁺ channels; the inhibition was slow to reach maximum (tens of seconds), current activation was not slowed or shifted in the positive voltage direction, and the inhibition was not relieved by positive prepulses. Nimodipine and -conotoxin GVIA blocked fractions of the current and also reduced the magnitude of the responses to DHPG, indicating that both L- and N-type Ca²⁺ channels were regulated. These results further differentiate the slow modulatory pathway observed in neocortical neurons when Ca²⁺ is used as the charge carrier from the rapid voltage-dependent mechanism reported to inhibit Ba²⁺ currents under Ca²⁺-free conditions.

	INTRODUCTION

Abstract Introduction Methods Results Discussion References

Metabotropic glutamate receptors (mGluRs) mediate inhibition of high-voltage-activated Ca²⁺ current in a number of neuronal types, including neocortical and hippocampal pyramidal neurons (Choi and Lovinger 1996; Lester and Jahr 1990; Sahara and Westbrook 1993; Sayer et al. 1992; Swartz and Bean 1992). Regulation of Ca²⁺ channels at somata and dendrites may influence neuronal excitability or synaptic plasticity, and at presynaptic terminals it may play a role in autoinhibition of transmitter release (reviewed by Stefani et al. 1996b). mGluR subtypes are encoded by eight known genes and are classified into three groups based on amino acid sequence similarities and pharmacology: group I, mGluR1 and mGluR5; group II, mGluR2 and mGluR3; and group III, mGluR4, mGluR6, mGluR7, and mGluR8. Splice variants have been found for several mGluR subtypes (for a review of mGluR classification and pharmacology see Conn and Pin 1997). Recent studies used selective agonists to identify which groups of mGluRs are involved in Ca²⁺ channel regulation. In neocortical pyramidal neurons, receptors from all three groups have been implicated; Ca²⁺ channel inhibition was observed in response to the group I agonist (RS)-3,5-dihydroxyphenylglycine (DHPG) and the group II agonist (2S,2'R,3'R)-2-(2',3'-dicarboxycyclopropyl)glycine (DCG-IV) (Choi and Lovinger 1996) and also to L(+)-2-amino-4-phosphonobutyric acid (L-AP4), a group III agonist (Stefani et al. 1996a 1998). The Ca²⁺ channel subtypes targeted for modulation by mGluRs in neocortical neurons were N-type and possibly P-type (Choi and Lovinger 1996) or predominantly L-type (Sayer et al. 1992), depending on experimental conditions.

Studies of Ca²⁺ channel regulation in sympathetic neurons have revealed a variety of G protein-mediated mechanisms (Hille 1994). A major pathway that inhibits N-type channels is fast, membrane delimited, does not require the presence of Ca²⁺, and involves a shift in the voltage dependence of channel gating. Another pathway is slower, probably involves a diffusible second messenger, requires Ca²⁺, is not associated with a gating shift, and targets L- and N-type channels. In neocortical pyramidal neurons, the mGluR-mediated inhibition of Ba²⁺ current through N-type (and possibly P-type) Ca²⁺ channels described by Choi and Lovinger (1996) showed the hallmarks of the fast pathway with a gating shift. These included a rapid agonist action (near-maximal responses in 2-4 s), a slowing of Ca²⁺ current activation, voltage-dependent inhibition (less inhibition at more positive potentials), and temporary relief of the inhibition by strongly positive voltage steps. This contrasted with an earlier study (Sayer et al. 1992) in which Ca²⁺ was used as the charge carrier, where a slow (1-2 min to maximum) inhibition of L-type Ca²⁺ channels was observed. The aim of this investigation was to characterize further the slow modulation of Ca²⁺ current by mGluRs in neocortical pyramidal neurons, by 1) using selective agonists to classify the group of mGluRs mediating the response and by 2) obtaining evidence as to whether the targeted Ca²⁺ channels undergo a gating shift.

	METHODS

Abstract Introduction Methods Results Discussion References

Cell preparation and recording

Rats aged 10-21 days were anesthetized with ketamine (150 mg/kg ip) and xylazine (10 mg/kg ip) and killed by carotid section. Transverse slices 500 µm thick were cut from dorsal frontoparietal (sensorimotor) cortex in bicarbonate-buffered saline (Sayer et al. 1992) saturated with 95% O₂-5% CO₂ at ~4°C. The slices were chopped into pieces 1-2 mm wide and incubated for 45-60 min with 19 U/ml papain in a saline containing (in mM) 5 N-2-hydroxyethylpiperazine-N'-2-ethanesulfonic acid (HEPES), 5 Na-HEPES, 140 NaCl, 3 KCl, 1.25 NaH₂PO₄, 10 glucose, 2 MgCl₂, and 2 CaCl₂, pH 7.4, temperature 28°C, saturated with O₂. The tissue pieces were maintained in this solution (without papain) at room temperature until immediately before use. Cells were isolated by trituration through Pasteur pipettes of decreasing aperture in a similar saline but with 5 mM ethylene glycol-bis(-aminoethyl ether)-N, N,N',N'-tetraacetic acid (EGTA), no added CaCl₂, and 10 mM MgCl₂.

Recordings were made in whole cell voltage clamp (Axopatch 200A, pClamp; Axon Instruments, Foster City, CA) from neurons of pyramidal morphology with apical dendrites <100 µm. The recording micropipettes contained (in mM) 30 tris(hydroxymethyl)aminomethane (Tris)-base, 70 diTris-PO₄, 4 MgCl₂, 5 NaCl, 0.286 CaCl₂, 1 1,2-bis-(2-aminophenoxy)ethane-N,N,N',N'-tetraacetic acid (BAPTA), 0.1 leupeptin, 0.3 GTP (Tris salt), 4 ATP (Tris salt), 20 phosphocreatine, and 50 U/ml creatine phosphokinase, pH 7.2. Under these conditions a calcium association constant of 4.5-6.5·10⁶ M¹ would be expected for BAPTA (Harrison and Bers 1987), giving an estimated free intracellular Ca²⁺ concentration of 60-90 nM under resting conditions. The aim of using a relatively low concentration of BAPTA (1 mM rather than 10 mM BAPTA or EGTA) was to provide conditions more permissive to Ca²⁺-dependent mechanisms.

After gaining whole cell access, the cells were superfused with (in mM) 140 tetraethylammonium (TEA)-Cl, 2 CaCl₂, 2 MgCl₂, 5 NaCl, 3 CsCl, 0.75 MnCl₂, 10 HEPES, 10 glucose, and 0.0005 tetrodotoxin, pH 7.4. The neurons were warmed from room temperature to 30-32°C during the 10 min after patch rupture; no data were collected until after this time, to allow equilibration of solutions and stabilization of temperature. Although the effects of temperature were not studied systematically, the aim of warming the neurons was to approach conditions that are more natural than room temperature. In preliminary experiments, still higher temperatures (e.g., 35-39°C) produced unacceptably high rates of Ca²⁺ current rundown. Mn²⁺ was included in the external solution to partially block the Ca²⁺ currents (by~50%). This reduced series resistance errors and appeared to further reduce rundown. This magnitude of block was found to approximately counteract the increased amplitude of the Ca²⁺ currents caused by warming. The access resistance after patch rupture was typically 20-30 M, and this was routinely compensated with 80% correction and 95% prediction. Except where stated, leak currents were subtracted by the P/4 or P/6 protocol. The liquid junction potential associated with the micropipette solution was approximately 3 mV, and all voltages were adjusted by this value. The records were low-pass filtered at 1 kHz and digitized at 5 kHz, except that when measuring tail currents the filter was set at 5 kHz and the digitization was set at 40 kHz. The tail currents were averaged over a 125-µs window (isochronally for each cell) 425 µs after returning to 60 mV from a 5-ms duration depolarization to various potentials and were normalized to the tail current associated with the +30-mV test potential. The standard duration of agonist application was 120 s, and the percentage inhibitions were measured at the time of maximum effect between 80 and 120 s after switching flow to the test drug. The measurements were made from a line joining the baseline and recovery amplitudes so that errors caused by rundown were minimized. Group data are expressed as the means ± SE. Exponentials were fitted to the time courses of responses with Origin software (Microcal Software, Northampton, MA).

Drugs

The following drugs were obtained from Tocris Cookson (Bristol, UK): (1S,3R)-1-aminocyclopentane-1,3-dicarboxylic acid (1S,3R-ACPD), L-AP4, (2S,1'S,2'S)-2-(carboxycyclopropyl)glycine (L-CCG-I), (S)-4-carboxyphenylglycine [(S)-4CPG], (RS)-2chloro-5-hydroxyphenylglycine (CHPG), DCG-IV, DHPG, and (+)--methyl-4-carboxyphenylglycine [(+)-MCPG]. In the experiments with DCG-IV or L-CCG-I, all solutions contained 100 µM DL-2-amino-5-phosphonopentanoic acid (DL-AP5, Research Biochemicals International, Natick, MA). Nimodipine (a gift of Bayer NZ) was made as a 5-mM stock solution in polyethylene glycol 400 and diluted to 5 µM in the superfusion solution; equivalent polyethylene glycol 400 was added to all other solutions applied in the same experiments. The pH of test solutions containing drugs was adjusted to the same as the control solution with HCl or TEAOH as necessary. Drug and control solutions were applied at a constant flow rate (except where specified, ~0.4 ml/min) into a ~0.3-ml recording chamber. An exception was -conotoxin GVIA (-CgTX, Research Biochemicals International), which was added as a 0.11-ml bolus close to the recorded neuron at a concentration of 3 µM in superfusion solution with 0.1 mg/ml cytochrome C added.

	RESULTS

Abstract Introduction Methods Results Discussion References

Data were obtained from 75 neocortical pyramidal neurons. The standard protocol for the pharmacological experiments was to hold the cells at 60 mV and evoke Ca²⁺ currents by stepping to 10 mV for 50 ms at 20-s intervals. At a holding potential of 60 mV, the low-voltage-activated (T-type) Ca²⁺ current is inactivated in these neurons (Sayer et al. 1990), and the test potential of 10 mV was close to the negative peak of the current-voltage relationship, as determined for each cell at the beginning of recording (not shown).

View larger version (39K): [in this window] [in a new window]   FIG. 1. High-voltage-activated Ca2+ currents were inhibited by (1S,3R)-1-aminocyclopentane-1,3-dicarboxylic acid (1S,3R-ACPD, 200 µM) and (RS)-3,5-dihydroxyphenylglycine (DHPG, 200 µM) but not by (2S,2'R,3'R)-2-(2',3'-dicarboxycyclopropyl)glycine (DCG-IV, 10 µM). A: peak Ca2+ current amplitudes vs. time plotted for 1 neuron. Currents were evoked by a step depolarization from 60 to 10 mV (50-ms duration). Drugs were applied at the times indicated by the horizontal bars. B: mean Ca2+ current inhibition (n = 14) in response to the agonists (error bars represent SE). C: individual records from the neuron in A taken at times indicated by the lowercase letters. The scale bars represent 100 pA and 10 ms.

Group-selective agonists

DHPG is a selective agonist at group I mGluRs (Baker et al. 1995; Gereau and Conn 1995), whereas DCG-IV is selective for group II mGluRs (Hayashi et al. 1993; Ishida et al. 1993). 1S,3R-ACPD is a more "broad-spectrum" mGluR agonist, activating all receptors of groups I and II, although it is of low potency, where determined, at mGluRs of group III (see Conn and Pin 1997). All three agonists were applied, in varying order, to 14 neurons. 1S,3R-ACPD (200 µM) and DHPG (200 µM) were similarly effective in inhibiting the Ca²⁺ current, producing reductions in amplitude of 31 ± 3.4 and 30 ± 3.6% respectively. DCG-IV (10 µM), in comparison, was ineffective (1 ± 0.4% inhibition; Fig. 1, A and B). Individual current traces from one neuron are shown in Fig. 1C; note that when the current was reduced by 1S,3R-ACPD or DHPG there was no significant slowing of activation.

L-CCG-I may be useful for distinguishing group II from group I mGluRs, because the concentrations causing 50% effectiveness (EC₅₀s) for responses mediated by mGluR2 and mGluR3 in expression systems were found to be 300 nM and 1 µM respectively, compared with 50 µM for mGluR1 and mGluR5 (Hayashi et al. 1992; Pin et al. 1994). In this study, 11 neurons were tested with 200 µM 1S,3R-ACPD and 1 µM L-CCG-I, and 6 of these were additionally tested with 10 µM and 100 µM L-CCG-I. L-CCG-I was ineffective at 1 µM, a concentration that would be expected to activate group II receptors, but it clearly inhibited the Ca²⁺ current at 10 and 100 µM (Fig. 2). L-CCG-I was ineffective at 1 µM regardless of whether it was applied before (n = 6) or after (n = 5) 1S,3R-ACPD. Overall inhibition was 2 ± 0.6% for 1 µM L-CCG-I, 21 ± 2.5% for 10 µM L-CCG-I, 30 ± 4.9% for 100 µM L-CCG-I, and 23 ± 3.4% for 200 µM 1S,3R-ACPD (Fig. 2B).

View larger version (38K): [in this window] [in a new window]   FIG. 2. (2S,1'S,2'S)-2-(carboxycyclopropyl)glycine (L-CCG-I) inhibited Ca2+ currents at concentrations of 10 and 100 µM but not at 1 µM. A: data from an individual neuron. B: group data representing the mean responses to the agonists. Numbers of neurons tested are shown in parentheses. Concentration of 1S,3R-ACPD was 200 µM.

The group III-selective agonist L-AP4, at a concentration of 1 mM, did not inhibit the Ca²⁺ current consistently in these experiments. Instead, the most common finding for L-AP4 was a small increase in peak Ca²⁺ current (9/13 neurons, e.g., Fig. 3A). In the remainder there was no response, or in one case there was a 5% reduction. Overall L-AP4 enhanced the current by 6 ± 1.6%, and in the same cells 200 µM 1S,3R-ACPD induced a 26 ± 3.8% inhibition (Fig. 3B).

View larger version (25K): [in this window] [in a new window]   FIG. 3. L(+)-2-amino-4-phosphonobutyric acid (L-AP4, 1 mM) did not mimic the response to 1S,3R-ACPD (200 µM). A: example from a single neuron. B: group data (n = 13).

Compounds with differential action on group I mGluR subtypes

Two compounds putatively useful in distinguishing mGluR1 from mGluR5 were tested; CHPG, which was reported to activate mGluR5a but not mGluR1a (Doherty et al. 1997), and (S)-4CPG, which has competitive antagonist selectivity for mGluR1a over mGluR5a (Brabet et al. 1995).

Application of CHPG (1 mM) resulted in inhibition of the Ca²⁺ current in 7 of 10 neurons tested. The reduction associated with CHPG was 17 ± 3.0%, whereas in the same cells the response to 200 µM DHPG was 35 ± 3.8% inhibition. In five of these seven neurons, the response to CHPG was biphasic; the inhibitions followed small (7 ± 0.9%) and transient (peak 40 s) increases in the current seen during wash-in of the drug (Fig. 4A). The responses to DHPG in the same neurons were purely inhibitory. In the remaining 3 of the 10 neurons tested, application of CHPG was associated with monophasic increases in the current (15 ± 4.3%), whereas in the same cells DHPG produced 23 ± 6.6% inhibition (Fig. 4B).

View larger version (19K): [in this window] [in a new window]   FIG. 4. Effects of compounds differentially active at mGluR1 and mGluR5. A: typical response to 1 mM (RS)-2-chloro-5-hydroxyphenylglycine (CHPG), in which a biphasic enhancement/reduction of the Ca2+ current was seen, with the reduction being of smaller amplitude than that in 200 µM DHPG. B: in this cell the only effect of CHPG was a small enhancement of the current. C: (S)-4-carboxyphenylglycine [(S)-4CPG; 500 µM] did not to antagonize the response to 50 µM 1S,3R-ACPD in this neuron. D: example from a cell in which (S)-4CPG application was associated with a small increase in the current and loss of the response to 1S,3R-ACPD.

Antagonist activity of (S)-4CPG (500 µM) was assessed by application of 50 µM 1S,3R-ACPD in the absence and presence of the compound. In three of six neurons tested, (S)-4CPG had no apparent effect; the mean inhibition of the Ca²⁺ current in response to 1S,3R-ACPD alone was 16 ± 2.7% compared with 19 ± 3.3% in the presence of (S)-4CPG (Fig. 4C). However, for the remaining three cells in which 1S,3R-ACPD alone produced a mean inhibition of 15 ± 1.4%, the application of (S)-4CPG was associated with an increase in the Ca²⁺ current (mean 10 ± 1.9%). The subsequent application of 1S,3R-ACPD in the presence of (S)-4CPG produced no discernible effect in these neurons (Fig. 4D).

Antagonism by (+)-MCPG

(+)-MCPG is a competitive antagonist at some cloned mGluRs of groups I and II but not group III (Hayashi et al. 1994), and it has been shown to inhibit mGluR-mediated responses in a number of systems (reviewed by Roberts 1995). In six neurons, the response to 50 µM 1S,3R-ACPD was tested with and without 1 mM (+)-MCPG in the bath. In these cells, the Ca²⁺ current inhibition was 24 ± 6.4% for 1S,3R-ACPD alone and 9 ± 5.2% for 1S,3R-ACPD in the presence of 1 mM (+)-MCPG (Fig. 5). This difference was significant (P = 0.008, paired t-test). However, 1 mM (+)-MCPG had no effect on Ca²⁺ current inhibition evoked by 200 µM 1S,3R-ACPD in two other cells (not shown).

View larger version (27K): [in this window] [in a new window]   FIG. 5. (+)--Methyl-4-carboxyphenylglycine [(+)-MCPG; 1 mM] reduced the response to 1S,3R-ACPD (50 µM). A: plot Ca2+ current amplitude vs. time for 1 neuron in which an antagonist action of (+)-MCPG was particularly evident. B: mean inhibition of Ca2+ current evoked by 1S,3R-ACPD alone and in the presence of (+)-MCPG (n = 6).

Time course of Ca²⁺ current inhibition

To determine the time course of the mGluR-mediated inhibition, Ca²⁺ currents were evoked every 2 s, and the turnover time of the recording chamber was minimized by reducing the fluid level and increasing the flow rate. The time course of drug delivery was estimated for each neuron by switching to a superfusion saline with 1.5 mM Mn²⁺ (instead of the usual 0.75 mM Mn²⁺), which produced further partial block of the Ca²⁺ current. DHPG (200 µM) applied to four of these neurons inhibited the Ca²⁺ current with a much slower time course than was caused by Mn²⁺ blockade (Fig. 6). When fitted with a single exponential, the time constants obtained for inhibition by DHPG in the four neurons were (in seconds) 21.9, 21.4, 27.3, and 9.7 (mean 20.1). In the same cells the time constants for Mn²⁺ blockade were almost an order of magnitude shorter, at 2.4, 3.4, 2.7, and 1.9 s (mean 2.6), respectively, suggesting that the time courses of DHPG delivery had little influence on the measured responses. The purpose of fitting the responses with single exponentials was to quantify their time courses rather than to make any inferences about the underlying mechanisms. Besides the slow time course, in three of the four cells there was a delay of 4-6 s in the onset of the inhibition compared with the response to Mn²⁺. In the fourth neuron, which happened to be the one with the shortest time constant for the response to DHPG, the delay was not resolvable at the 2-s interval between evoked currents.

View larger version (15K): [in this window] [in a new window]   FIG. 6. Time course of the inhibition of Ca2+ current by DHPG (200 µM; ) superimposed on the time course of the block caused by raising Mn2+ from 0.75 to 1.5 mM () in a single neuron. Time = 0 represents the time of switching the solution flow to the recording chamber. , monoexponential fits to the data, which in this example gave time constants of 27.3 s for DHPG and 2.7 s for the change to 1.5 mM Mn2+.

View larger version (22K): [in this window] [in a new window]   FIG. 7. Inhibition of Ca2+ current by DHPG was not associated with a shift in the activation curve and was not relieved by positive prepulses. A: averaged activation curves (n = 5) plotted from tail currents before (baseline), during, and after (wash) application of 100 µM DHPG. Some of the error bars (SE) are obscured by the symbols. B: protocol for determining the effects of a positive prepulse. The Ca2+ current amplitude evoked by depolarization to 10 mV was measured before and after a step to 80 mV (duration 50 ms). An example is shown of currents before and during (*) DHPG application. C, top: normalized amplitudes of the currents evoked by the 1st depolarization to 10 mV (a), averaged across cells (n = 5, same neurons as in A) were reduced during application of DHPG. C, bottom: ratio of the current amplitude evoked by the 2nd depolarization to 10 mV over the 1st (b/a) did not increase during application of DHPG, indicating that the step to 80 mV did not relieve the inhibition. Roman numerals indicate the times when the protocol was stopped so that tail currents could be measured for the activation curves in A.

View larger version (52K): [in this window] [in a new window]   FIG. 8. Effects of Ca2+ channel blockers on the response to DHPG. A: example from 1 neuron of current amplitude vs. time showing the response to 200 µM DHPG before and after application of -conotoxin GVIA (-CgTX, 3 µM) and nimodipine (5 µM). B: percentages of the total (baseline) current blocked by -CgTX and nimodipine (n = 7). C: effects of Ca2+ channel blockers on the magnitude of Ca2+ current inhibited by DHPG, expressed as the percentage of the initial (baseline) DHPG-sensitive current that was blocked after application of -CgTX or nimodipine (n = 7).

Voltage dependence of Ca²⁺ channel inhibition

A shift in the voltage dependence of channel gating is a common mechanism of Ca²⁺ channel inhibition, and it was reported to occur for mGluR responses in neocortical neurons (Choi and Lovinger 1996). To investigate whether this was happening under the conditions of the current experiments, activation curves were constructed from tail currents evoked before, during, and after application of 100 µM DHPG. DHPG did not alter the voltage dependence of Ca²⁺ channel activation (n = 5, Fig. 7A). In the same cells, the inhibition of peak Ca²⁺ current induced by DHPG was not relieved by a 50-ms prepulse to 80 mV (Fig. 7, B and C). Facilitation (an increase in Ca²⁺ current after a depolarizing prepulse) was not seen in either the presence or absence of DHPG (Fig. 7C).

Inhibition of Ca²⁺ channel subtypes

To determine which Ca²⁺ channel subtypes were inhibited under the conditions of the current experiments, DHPG (200 µM) was applied before and after successive fractions of the Ca²⁺ current were blocked by -CgTX (an N-type channel blocker) and nimodipine (an L-type channel blocker). An example of the effects of the drugs on peak Ca²⁺ current amplitude is shown in Fig. 8A. In these cells (n = 7), -CgTX (3 µM) blocked 23 ± 2.3%, and nimodipine (5 µM) blocked 33 ± 2.7% of the total (baseline) current (Fig. 8B), leaving 43 ± 1.8% resistant to both. Note that the fractions of current blocked are similar to previous findings (Brown et al. 1994), suggesting that the Mn²⁺ present did not selectively block of any one type of Ca²⁺ channel. Neither -CgTX nor nimodipine abolished the response to DHPG, but each reduced the magnitude of the inhibition (i.e., the absolute amount of Ca²⁺ current inhibited), suggesting that each blocked some of the channels modulated by DHPG. Nimodipine reduced the DHPG-sensitive current by 36 ± 7.9%, whereas -CgTX reduced it by 30 ± 7.2% (Fig. 8C). After both -CgTX and nimodipine were applied, the remaining current sensitive to DHPG represented 34 ± 8.2% of the original (baseline) current inhibited by DHPG.

	DISCUSSION

Abstract Introduction Methods Results Discussion References

The effectiveness of DHPG implicates group I mGluRs in a slow pathway for Ca²⁺ current inhibition in neocortical pyramidal neurons. The finding that L-CCG-I was active at 10 µM is consistent with involvement of group I receptors because, although this agonist would be expected to selectively activate group II receptors at 1 µM, it also activates group I mGluRs at higher concentrations (Hayashi et al. 1992; Pin et al. 1994). In addition, the results fit with the rank order of potency previously found for Ca²⁺ current inhibition in these neurons (quisqualate > glutamate > trans-ACPD) (Sayer et al. 1992), a profile that is now considered characteristic of group I mGluRs.

The use of CHPG and (S)-4CPG to distinguish subtypes of group I mGluRs produced variable results. In the majority of neurons tested, the Ca²⁺ currents were inhibited by 1 mM CHPG, implicating mGluR5. The smaller amplitude of the inhibition, compared with 200 µM DHPG, is consistent with CHPG being a low-potency (and/or partial) agonist at this receptor (Doherty et al. 1997). However the results were complicated by the biphasic nature of the responses and, in a minority of cases, by pure increases of the current. Without further information on the specificity of CHPG, an action on non-group I receptors cannot be ruled out; such activity could converge on the same pathways or Ca²⁺ channels to oppose the group I mGluR-mediated inhibition. Alternatively, variable expression of group I subtypes (or their splice variants) between cells might account for the findings. Both mGluR1 and mGluR5 have been localized to neocortex by immunohistochemistry, although mGluR1 expression is low relative to mGluR5 (Baude et al. 1993; Martin et al. 1992; Romano et al. 1995; Shigemoto et al. 1993).

The same possibilities could underlie the findings for (S)-4CPG. The lack of antagonist action for some neurons would be consistent with mediation by mGluR5 (Brabet et al. 1995), whereas the slight enhancement of the current and block of the inhibition in other cells might reflect antagonism of an mGluR1 with a small degree of tonic activity. Alternatively, there may have been interference from agonist actions at non-group I mGluRs (Bedingfield et al. 1995).

The lack of response to 10 µM DCG-IV or 1 µM L-CCG-I suggests that group II mGluRs are rarely or never involved in the slow inhibition of Ca²⁺ current in these neurons. It is unlikely that the ineffectiveness of DCG-IV was due to the concentration being too low, given that the EC₅₀s for the action of DCG-IV on group II mGluRs expressed in oocytes were submicromolar (Hayashi et al. 1993), and 5 µM DCG-IV evoked a fast inhibition of Ca²⁺ channels in 75% of neocortical neurons tested when Ba²⁺ was used as the charge carrier (Choi and Lovinger 1996). The results of Choi and Lovinger (1996) indicate that group II mGluRs are present in these cells and that they have modulatory influence on Ca²⁺ channels, which can be revealed under certain conditions. That at least a subset of neocortical projection neurons produce group II mGluRs is also suggested by the presynaptic depression induced by DCG-IV at corticostriatal synapses (Lovinger and McCool 1995), although this is not necessarily mediated by regulation of Ca²⁺ channels.

The present findings for the group III agonist L-AP4 do not support a role for group III mGluRs in the slow inhibitory pathway for Ca²⁺ current inhibition, at least not under the conditions of these experiments. Instead, L-AP4 induced a small enhancement of Ca²⁺ current in the majority of neurons. Choi and Lovinger (1996) likewise reported only small effects of group III agonists on Ba²⁺ current in neocortical neurons (mean 3% inhibition with 100 µM D,L-AP4). More significant inhibition of Ba²⁺ current (mean 22% with 100 µM L-AP4) was observed by Stefani et al. (1996a, 1998) in the same type of neuron, but in older rats [aged 1-2 mo instead of 9-21 days for Choi and Lovinger (1996) and this study]. There is evidence for group III mGluRs being involved in suppression of excitatory synaptic transmission in the neocortex (Burke and Hablitz 1994), although again the inhibition of Ca²⁺ channels in presynaptic terminals is just one of a number of possible mechanisms by which this could occur.

Similar to the findings for the fast inhibition of Ba²⁺ current in neocortical neurons (Choi and Lovinger 1996), (+)-MCPG produced a partial and variable block of Ca²⁺ current inhibition at submaximally effective concentrations of 1S,3R-ACPD. The findings of this study are consistent with (+)-MCPG acting as an antagonist, albeit of low potency, at group I mGluRs. Note that (+)-MCPG has previously been shown to antagonize responses to 1S,3R-ACPD in cells expressing mGluR1a and in cells expressing mGluR5a (Brabet et al. 1995).

A number of features distinguished this form of modulation from the typical fast inhibition of Ca²⁺ channels that is associated with a shift in the voltage dependence of channel gating. The responses to 1S,3R-ACPD and DHPG were slow, taking tens of seconds to reach a maximum. The time courses were an order of magnitude slower than those observed for mGluR agonists in other studies in neocortical neurons where Ca²⁺ was absent and Ba²⁺ was used as the charge carrier (Choi and Lovinger 1996; Stefani et al. 1996a, 1998). No evidence was found for the Ca²⁺ current inhibition evoked by DHPG being voltage dependent. There was no obvious slowing of current onset, there were no shifts in the activation curves, and the inhibition was not relieved by strongly positive prepulses. This is in marked contrast to the voltage-dependent inhibition seen in the same cells in the absence of Ca²⁺ (Choi and Lovinger 1996) and that observed in many examples of Ca²⁺ channel inhibition in other types of neuron (reviewed by Dolphin 1995). However, the voltage-independent pathway for inhibition of N- and L-type channels in sympathetic neurons, activated by muscarinic and angiotensin II receptors, is slow in onset, sensitive to intracellular Ca²⁺ chelation, and involves a diffusible messenger (Hille 1994). This appears similar to the mechanism of inhibition described here, although the experiments required to establish the involvement of a diffusible messenger have not yet been performed in neocortical neurons.

The experiments with Ca²⁺ channel antagonists showed that the mGluRs activated by DHPG regulate multiple subtypes of Ca²⁺ channel. -CgTX and nimodipine each caused a reduction in the amount of Ca²⁺ current inhibited by DHPG, suggesting that N- and L-type Ca²⁺ channels were inhibited in the baseline (initial) response to DHPG. The Ca²⁺ current that remained after block by -CgTX and nimodipine was probably mostly conducted by P-type channels, along with some through channels resistant to organic blockers (~30 and 10% of the original current, respectively) (Brown et al. 1994; Choi and Lovinger 1996). Therefore the residual response to DHPG after block by -CgTX and nimodipine probably involved inhibition of P-type Ca²⁺ channels, although modulation of the resistant ("R-type") channels cannot be ruled out. In the previous study (Sayer et al. 1992), all of the Ca²⁺ current inhibition evoked by trans-ACPD appeared to be accounted for by modulation of L-type channels. Two factors that might have allowed additional channel modulation to be revealed in this study were the higher temperature and the reduced Ca²⁺ buffering. Using different conditions again (21-24°C, Ba²⁺ substituted for Ca²⁺), Choi and Lovinger (1996) suggested that the fast inhibitory pathway also regulated more than one type of Ca²⁺ channel, with evidence that N-type and possibly P-type Ca²⁺ channels were modulated.

It remains to be explained why, if fast and slow mechanisms for mGluR-mediated Ca²⁺ channel inhibition exist in neocortical neurons, both are not observed under the same experimental conditions. Even when currents were recorded every 2 s, there was no evidence for a fast component to the Ca²⁺ current inhibition. This argues against the possibility that a fast pathway was activated, but then desensitized within the first 20-s interval of drug application, between the evoked currents of the standard experimental protocol. The important variable may be Ca²⁺; when it is used as the charge carrier, only the slow inhibition is observed (Sayer et al. 1992 and this study), but when Ba²⁺ is substituted for Ca²⁺, only the fast inhibition is apparent (Choi and Lovinger 1996). The effect of Ca²⁺ may be intracellular, as a similar slowing of Ca²⁺ channel inhibition by 1S,3R-ACPD was found in cultured hippocampal neurons when intracellular Ca²⁺ concentration was buffered at 100 nM instead of <1 nM (Sahara and Westbrook 1993). However, strong intracellular Ca²⁺ buffering does not necessarily produce Ca²⁺ channel modulation identical to that seen when Ba²⁺ is used as the charge carrier (Jones et al. 1992; Sayer et al. 1992). Of possible relevance to this issue is the recent finding that group I mGluRs are activated by extracellular Ca²⁺ and that other divalent ions are less effective (Kubo et al. 1998, Kubokawa et al. 1996). However, the implications of mGluR sensitivity to Ca²⁺ for the mechanisms modulating Ca²⁺ channels are not yet clear.

In neocortical neurons it seems that two separate (fast and slow) mechanisms for Ca²⁺ channel inhibition exist rather than one that is accelerated or slowed under different conditions. They differ in the subtypes of receptors involved: both group I and group II mGluRs for the fast pathway but only group I mGluRs for the slow pathway. Although both pathways target multiple types of Ca²⁺ channel, the mechanisms of inhibition are different, involving a shift in the voltage dependence of gating for the fast pathway but not for the slow one, and it appears that each pathway is inhibited or occluded under conditions that favor the other.

	ACKNOWLEDGEMENTS

  I thank Dr. W. C. Abraham for comments on the manuscript.

  This work was supported by the New Zealand Lottery Grants Board, the Health Research Council of New Zealand, and the Otago Medical Research Foundation.

	FOOTNOTES

  Address for reprint requests: Dept. of Physiology, School of Medical Sciences, P.O. Box 913, Dunedin, New Zealand.

  Received 20 February 1998; accepted in final form 9 July 1998.

	REFERENCES

Abstract Introduction Methods Results Discussion References

• BAKER, S. R., GOLDSWORTHY, J., HARDEN, R. C., SALHOFF, C. R., SCHOEPP, D. D. Enzymatic resolution and pharmacological activity of the enantiomers of 3,5-dihydroxyphenylglycine, a metabotropic glutamate receptor agonist. Bioorg. Med. Chem. Lett. 5: 223-228, 1995.
• BAUDE, A., NUSSER, Z., ROBERTS, J.D.B., MULVIHILL, E., MCILHINNEY, R. A. J., AND SOMOGYI, P. The metabotropic glutamate receptor (mGluR1) is concentrated at perisynaptic membrane of neuronal subpopulations as detected by immunogold reaction. Neuron 11: 771-787, 1993.[Medline]
• BEDINGFIELD, J. S., KEMP, M. C., JANE, D. E., TSE, H.-W., ROBERTS, P. J., WATKINS, J. C. Structure-activity relationships for a series of phenylglycine derivatives acting at metabotropic glutamate receptors (mGluRs). Br. J. Pharmacol. 116: 3323-3329, 1995.[Medline]
• BRABET, I., MARY, S., BOCKAERT, J., PIN, J.-P. Phenylglycine derivatives discriminate between mGluR1- and mGluR5-mediated responses. Neuropharmacology 34: 895-903, 1995.[Medline]
• BROWN, A. M., SAYER, R. J., SCHWINDT, P. C., CRILL, W. E. P-type calcium channels in rat neocortical neurones. J. Physiol. (Lond.) 475: 197-205, 1994.[Abstract/Free Full Text]
• BURKE, J. P., HABLITZ, J. J. Presynaptic depression of synaptic transmission mediated by activation of metabotropic glutamate receptors in rat neocortex. J. Neurosci. 14: 5120-5130, 1994.[Abstract]
• CHOI, S., LOVINGER, D. M. Metabotropic glutamate receptor modulation of voltage-gated Ca²⁺ channels involves multiple receptor subtypes in cortical neurons. J. Neurosci. 16: 36-45, 1996.[Abstract/Free Full Text]
• CONN, P. J., PIN, J. P. Pharmacology and functions of metabotropic glutamate receptors. Annu. Rev. Pharmacol. Toxicol. 37: 205-237, 1997.[Medline]
• DOHERTY, A. J., PALMER, M. J., HENLEY, J. M., COLLINGRIDGE, G. L., JANE, D. E. (RS)-2-Chloro-5-hydroxyphenylglycine (CHPG) activates mGlu₅, but not mGlu₁, receptors expressed in CHO cells and potentiates NMDA response in the hippocampus. Neuropharmacology 36: 265-267, 1997.[Medline]
• DOLPHIN, A. C. Voltage-dependent calcium channels and their modulation by neurotransmitters and G proteins. Exp. Physiol. 80: 1-36, 1995.[Medline]
• GEREAU, R. W., CONN, P. J. Roles of specific metabotropic glutamate receptor subtypes in regulation of hippocampal CA1 pyramidal cell excitability. J. Neurophysiol. 74: 122-129, 1995.[Abstract/Free Full Text]
• HARRISON, S. M., BERS, D. M. The effect of temperature and ionic strength on the apparent Ca-affinity of EGTA and the analogous Ca-chelators BAPTA and dibromo-BAPTA. Biochim. Biophys. Acta 925: 133-143, 1987.[Medline]
• HAYASHI, Y., MOMIYAMA, A., TAKAHASHI, T., OHISHI, H., OGAWA-MEGURO, R., SHIGEMOTO, R., MIZUNO, N., NAKANISHI, S. Role of a metabotropic glutamate receptor in synaptic modulation in the accessory olfactory bulb. Nature 366: 687-690, 1993.[Medline]
• HAYASHI, Y., SEKIYAMA, N., NAKANISHI, S., JANE, D. E., SUNTER, D. C., BIRSE, E. F., UDVARHELYI, P. M., WATKINS, J. C. Analysis of agonist and antagonist activities of phenylglycine derivatives for different cloned metabotropic glutamate receptor subtypes. J. Neurosci. 14: 3370-3377, 1994.[Abstract]
• HAYASHI, Y., TANABE, Y., ARAMORI, I., MASU, M., SHIMAMOTO, K., OHFUNE, Y., NAKANISHI, S. Agonist analysis of 2-(carboxycyclopropyl)glycine isomers for cloned metabotropic glutamate receptor subtypes expressed in Chinese hamster ovary cells. Br. J. Pharmacol. 107: 539-543, 1992.[Medline]
• HILLE, B. Modulation of ion-channel function by G-protein-coupled receptors. Trends Neurosci. 17: 531-536, 1994.[Medline]
• ISHIDA, M., SAITOH, T., SHIMAMOTO, K., OHFUNE, Y., SHINOZAKI, H. A novel metabotropic glutamate receptor agonist: marked depression of monosynaptic excitation in the newborn rat isolated spinal cord. Br. J. Pharmacol. 109: 1169-1177, 1993.[Medline]
• JONES, S., ROBBINS, J., BROWN, D. A. Neurotransmitter modulation of calcium channels is dependent on the charge carrier used in the recording of currents. Neurosci. Lett. 145: 153-156, 1992.[Medline]
• KUBO, Y., MIYASHITA, T., MURATA, Y. Structural basis for a Ca²⁺-sensing function of the metabotropic glutamate receptors. Science 279: 1722-1725, 1998.[Abstract/Free Full Text]
• KUBOKAWA, K., MIYASHITA, T., NAGASAWA, H., KUBO, Y. Cloning and characterization of a bifunctional metabotropic receptor activated by both extracellular calcium and glutamate. FEBS Lett. 392: 71-76, 1996.[Medline]
• LESTER, R.A.J., JAHR, C. E. Quisqualate receptor-mediated depression of calcium currents in hippocampal neurons. Neuron 4: 741-749, 1990.[Medline]
• LOVINGER, D. M., MCCOOL, B. A. Metabotropic glutamate receptor-mediated presynaptic depression at corticostriatal synapses involves mGluR2 or 3. J. Neurophysiol. 73: 1076-1083, 1995.[Abstract/Free Full Text]
• MARTIN, L. J., BLACKSTONE, C. D., HUGUANIR, R. L., PRICE, D. L. Cellular localization of a metabotropic glutamate receptor in rat brain. Neuron 9: 259-270, 1992.[Medline]
• PIN, J.-P., JOLY, C., HEINEMANN, S. F., BOCKAERT, J. Domains involved in the specificity of G protein activation in phospholipase C-coupled metabotropic glutamate receptors. EMBO J. 13: 342-348, 1994.[Medline]
• ROBERTS, P. J. Pharmacology tools for the investigation of metabotropic glutamate receptors (mGluRs): phenylglycine derivatives and other selective antagonistsan update. Neuropharmacology 34: 813-819, 1995.[Medline]
• ROMANO, C., SESMA, M. A., MCDONALD, C. T., O'MALLEY, K., VAN DEN POL, A. N., OLNEY, J. W. Distribution of metabotropic glutamate receptor mGluR5 immunoreactivity in rat brain. J. Comp. Neurol. 355: 455-469, 1995.[Medline]
• SAHARA, Y., WESTBROOK, G. L. Modulation of calcium currents by a metabotropic glutamate receptor involves two kinetic components in cultured hippocampal neurons. J. Neurosci. 13: 3041-3050, 1993.[Abstract]
• SAYER, R. J., SCHWINDT, P. C., CRILL, W. E. High- and low-threshold calcium currents in neurons acutely isolated from rat sensorimotor cortex. Neurosci. Lett. 120: 175-178, 1990.[Medline]
• SAYER, R. J., SCHWINDT, P. C., CRILL, W. E. Metabotropic glutamate receptor-mediated suppression of L-type calcium current in acutely isolated neocortical neurons. J. Neurophysiol. 68: 833-842, 1992.[Abstract/Free Full Text]
• SHIGEMOTO, R., NOMURA, S., OHISHI, H., SUGIHARA, H., NAKANISHI, S., MIZUNO, N. Immunohistochemical localization of a metabotropic glutamate receptor, mGluR5, in the rat brain. Neurosci. Lett. 163: 53-57, 1993.[Medline]
• STEFANI, A., SPADONI, F., BERNARDI, G. L-AP4 inhibits high voltage-activated Ca²⁺ currents in pyramidal cortical neurones. Neuroreport 7: 421-424, 1996a.[Medline]
• STEFANI, A., SPADONI, F., BERNADI, G. Group III metabotropic glutamate receptor agonists modulate high voltage-activated Ca²⁺ currents in pyramidal neurons of the adult rat. Exp. Brain Res. 119: 237-244, 1998.[Medline]
• STEFANI, A., PISANI, A., MERCURI, N. B., CALABRESI, P. The modulation of calcium currents by the activation of mGluRs. Mol. Neurobiol. 13: 81-95, 1996b.[Medline]
• SWARTZ, K. J., BEAN, B. P. Inhibition of calcium channels in rat CA3 pyramidal neurons by a metabotropic glutamate receptor. J. Neurosci. 12: 4358-4371, 1992.[Abstract]

0022-3077/98 $5.00 Copyright ©1998 The American Physiological Society