HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Allen, T. G. J. Search for Related Content PubMed PubMed Citation Articles by Allen, T. G. J.

The role of N-, Q- and R-type Ca²⁺ channels in feedback inhibition of ACh release from rat basal forebrain neurones

T. G. J. Allen

MS 8424 Received 1 July 1998; accepted after revision 5 November 1998.

	ABSTRACT

Top Abstract Introduction Methods Results Discussion References

1. The Ca²⁺ channel subtypes controlling ACh release from basal forebrain neurones and the ionic basis underlying muscarinic receptor-mediated autoinhibition were studied using skeletal myoballs to detect ACh release from individual rat basal forebrain neurones in culture.

2. Somatic Ca²⁺ currents evoked using a simulated action potential waveform revealed that Ca²⁺ entry was primarily through N-, Q- and to a lesser extent R-, T- and L-type Ca²⁺ channels.

3. Muscarine (10 µM) inhibited N- and Q- but not R-, T- or L-type somatic Ca²⁺ channels. Agonist inhibition was totally blocked by pre-treatment with pertussis toxin (500 ng ml^-1).

4. ACh release from discrete sites along basal forebrain neurites (1·2 mM extracellular Ca²⁺) could be largely abolished by blocking Ca²⁺ entry through either N-type or Q-type Ca²⁺ channels. Inhibition of Ca²⁺ entry through L- or T-type channels had no effect upon release. Following inhibition of either N- or Q-type Ca²⁺ channels, release could be restored to near control levels by raising [Ca²⁺]_o. After selectively blocking N-, Q-, L- and T-type channels, low levels of release could still be evoked as a result of Ca²⁺ entry through R-type Ca²⁺ channels.

5. Muscarinic receptor activation reversibly inhibited ACh release due to Ca²⁺ entry through N-, Q- and R-type Ca²⁺ channels. In contrast, inhibition of inwardly rectifying K⁺ channels using Ba²⁺ (3-10 µM) or substance P (0·03-0·1 µM), or block of SK or BK Ca²⁺-activated K⁺ channels with apamin (100 nM) or charbydotoxin (100 nM) respectively, had no effect upon either ACh release or its modulation by muscarinic agonists.

6. These results show that ACh release from individual release sites on basal forebrain neurones is controlled by multiple Ca²⁺ channel subtypes with overlapping Ca²⁺ microdomains and that autoinhibition of release results from M₂ muscarinic receptor-mediated inhibition of these presynaptic Ca²⁺ channels rather than as a consequence of K⁺ channel activation.

	INTRODUCTION

Top Abstract Introduction Methods Results Discussion References

Correlation between a decline in cholinergic innervation of the cortex and the pathophysiology of Alzheimer's disease is well documented (Iversen, 1997; Kasa et al. 1997). As a consequence of this, large numbers of studies have been carried out to find ways of compensating for this loss with the aim of relieving some of the symptoms of dementia (Schorderet, 1995). One of the most common approaches has been to try and promote the effectiveness of the remaining cholinergic pathways, either by inhibiting degradation of ACh using anticholinesterases (Robbins et al. 1997) or by attempting to stimulate postsynaptic M₁ muscarinic receptors directly (Avery et al. 1997; Iversen, 1997). Unfortunately, this type of approach has met with only limited success. In part, this is due to our lack of understanding of the precise sites and ionic mechanisms responsible for controlling ACh release, that in some cases has led to drug treatments that have had conflicting actions upon different parts of the cholinergic pathway.

A wide variety of studies have demonstrated that ACh release within the cortex and hippocampus can be strongly inhibited by activation of presynaptic M₂ muscarinic receptors (see for example Quirion et al. 1994) and that there is selective loss of m2 receptors in Alzheimer's disease (Quirion et al. 1989). There is also limited evidence of a role for M₁-type receptors (Suzuki et al. 1988). Direct localization of muscarinic receptor mRNA has revealed that cholinergic basal forebrain neurones express, almost exclusively, m2 receptors (Vilaro et al. 1992), although there is some evidence of low levels of m4 receptor mRNA (Suygaya et al. 1997). Furthermore, the individual m2 receptor proteins have been localized to the cholinergic fibres innervating the cortex (Levey et al. 1991). However, whilst the m2 receptor is the most likely candidate for the presynaptic autoreceptor, it is not clear how its effects are mediated, and whether inhibition of release results from activation of K⁺ channels, inhibition of voltage-gated Ca²⁺ channels, or through a direct effect upon the release mechanism itself.

Exogenous application of ACh to acutely isolated rat cholinergic basal forebrain neurones has been shown to couple, via M₂ muscarinic receptors, to inhibition of multiple components of the somatic Ca²⁺ current (Allen & Brown, 1993) and also to activate a sustained outward K⁺ current in a population of basal forebrain neurones in the rat nucleus basalis (Harata et al. 1991). In addition, ACh and muscarinic agonists have also been reported to inhibit the late spike after-hyperpolarization in guinea-pig medial septal and diagonal band neurones (Sim & Griffith, 1991). However, in all of these reports, it was the action of cholinergic agonists upon somatic rather than presynaptic channels that was investigated and thus they provide us with only indirect evidence as to the true presynaptic autoinhibitory mechanism.

In a previous attempt to gain a clearer understanding of the presynaptic effects of ACh upon the release process, we employed nicotinic receptor-rich membrane patches to detect and measure ACh release directly from basal forebrain neurones in culture (Allen & Brown, 1996). Using this technique, we were able to demonstrate that spike-evoked Ca²⁺-dependent ACh release from discrete sites on individual basal forebrain neurones could be reversibly inhibited through the activation of M₂ muscarinic receptors (Allen & Brown, 1996). In the present study, the Ca²⁺ channel subtypes controlling release from these sites were investigated, and the ionic basis of this autoinhibitory feedback mechanism was examined. These experiments required much greater long-term recording stability than could be achieved using excised membrane patches as detectors. Therefore, to overcome this limitation, amphotericin perforated-patch myoballs were substituted for excised patches, thereby enabling the maintenance of stable recordings from individual release sites for in excess of 1 h. A preliminary report of this work has been given in abstract form to The Physiological Society (Allen, 1998).

	METHODS

Top Abstract Introduction Methods Results Discussion References

Preparation of dissociated basal forebrain cultures

Animals were placed into a chamber containing a saturated atmosphere of chloroform vapour and anaesthetized prior to decapitation according to Home Office guidelines. Details of the precise methods used for preparing dissociated mixed cell cultures of postnatal (11-14 day old) rat basal forebrain neurones have been described previously (see Allen et al. 1993).

Preparation of myoballs

The basic method used for preparing myoballs was similar to that previously described for the preparation of myotubes (Allen & Brown, 1996). However, in order to obtain an enriched population of small mononucleate myoballs, cultures were treated with 50 nM colchicine after 6-10 days in vitro in order to promote myoball rather than myotube formation. To obtain myoballs that could be easily detached from the substrate, cultures were briefly treated with 0·125 % trypsin for 5-8 min to round-up and detach a proportion of the myoballs 12-36 h prior to recording. The detached myoballs were then re-plated onto 13 mm diameter acid-washed glass coverslips. The resulting replated cultures contained an enriched population of loosely adherent myoballs which were used within 36 h of plating.

Electrophysiological studies

Ca²⁺ current. Whole-cell Ca²⁺ current recordings were made from acutely dissociated cells using intracellular and extracellular solutions the compositions of which have been detailed elsewhere (Allen et al. 1993; Allen & Brown, 1993). Ca²⁺ currents evoked with conventional voltage steps/ramps were recorded using a discontinuous single-electrode voltage clamp (Axoclamp-2, Axon Instruments). Ca²⁺ currents generated using a simulated action potential waveform were recorded with a patch-clamp amplifier (Axopatch 200A, Axon Instruments). When utilizing the patch-clamp amplifier, series resistance compensation ranged between 75 and 90 % and resulted in typical voltage errors of < 5 %. To evoke Ca²⁺ currents with a simulated action potential waveform, an action potential was recorded from a basal forebrain neurone and then scaled and used as the command waveform for the voltage-clamp amplifier. To minimize errors due to small fluctuations in series resistance when using such large voltage commands (140 mV) on-line leak subtraction was carried out using scaled subpulses. Additional compensation for any non-linear leak was made off-line in the same way as for step/ramp-generated currents, by subtracting the residual leak current remaining after blocking all Ca²⁺ channels with CdCl₂ (200 µM).

Transmitter release. Basal forebrain neurones were maintained in culture for 7-28 days prior to use in detection experiments. Attempts to detect ACh release were confined exclusively to magnocellular (20-25 µM) basal forebrain neurones (see Allen & Brown, 1996). At the start of each experiment, a coverslip bearing the re-plated myoballs was transferred into the Petri dish in which the basal forebrain culture was grown and the combined preparation was then mounted on the stage of an inverted microscope equipped with phase-contrast optics (Nikon TMS) and superfused at a rate of 4-6 ml min^-1 with Krebs solution of the following composition (mM): NaCl, 121; MgCl₂, 1·2; CaCl₂, 2·5; NaHCO₃, 25; Hepes, 5; and D-glucose, 11; adjusted to pH 7·4 (where necessary) and gassed with 95 % O₂-5 % CO₂. Switching between solutions was achieved via a multiway, zero dead space tap, connected to a series of independently oxygenated, drug-containing reservoirs. All experiments were carried out at room temperature (22-25°C). Electrophysiological recording from basal forebrain neurones was made using the whole-cell variant of the patch-clamp technique, details of which have been described previously (Allen & Brown, 1996). The composition of the intracellular pipette solution was (mM): potassium acetate, 108; KCl, 15·6; MgCl₂, 1; Hepes, 40; EGTA, 3; MgATP, 4; and Na₂GTP, 0·1; the solution was adjusted to pH 7·25 with NaOH, and had a final osmolarity of 307 mosmol l^-1. The free calcium concentration of the internal solution was calculated and adjusted to 30 nM. After rupture of the gigaseal to achieve whole-cell recording conditions, basal forebrain neurones were held close to their resting potential (-70 to -75 mV) and stimulated to fire action potentials by passing brief pulses of depolarizing current through the recording electrode in standard 'Bridge' recording mode (Axoclamp-2). Prior to patching a myoball, it was first detached from the substrate by gently rolling it loose using a heavily fire-polished patch electrode. Recording was carried out using the amphotericin perforated-patch technique (Rae et al. 1991) and an Axopatch 200A amplifier. Briefly, 1 mg of amphotericin B was dissolved (by ultrasonication) in 20 µl of DMSO and 1·5 µl of the resulting solution was then dissolved by ultrasonication in 1 ml of pipette filling solution of identical composition to that given above. To enable formation of a membrane gigaseal, electrodes were backfilled by dipping the tip of the electrode into filtered pipette solution, containing no amphotericin, for 30-60 s before conventional filling with amphotericin-containing solution. Once patched, the cell was voltage clamped (VH, -80 mV) and compensation was made for series resistance (typically 75-90 % compensation). The cell was then picked up and moved to the area of the composite culture containing the basal forebrain neurones.

Electrodes. In all cases, recording electrodes were fabricated from borosilicate glass capillaries (1·5 mm diameter; Clark Electromedical Instruments) using a List Medical (model L/M-3P-A) electrode puller, fire polished using a microforge (Nikon MF-9) and coated to within 100 µm of their tips with Sylgard (Dow Corning) to reduce transluminal capacitance. Electrode resistances ranged between 4 and 9 M for whole-cell recording and 2·5 and 4·5 M for perforated-patch recording. The reference ground electrode for both recording amplifiers consisted of a glass bridge containing 4 % agar-saline, one end of which was placed in the culture dish, the other in a 3 M KCl side chamber connected to the ground via a Ag-AgCl pellet.

Myoball detector calibration

In order to be able to relate changes in nicotinic receptor current to changes in ACh release, it was necessary to correct for the non-linear dose-response relationship of the nicotinic ACh receptors (nAChRs) in the myoball detection system. We previously devised a system for calibrating excised detector patches that enabled measurements to be made of the percentage change in the amount of ACh released (see Allen & Brown, 1996). Unlike patches, when a myoball is used as the detector, only a small, unknown proportion, of the total cell membrane actually participates in the detection process. Therefore, unlike a patch, the concentration of ACh to which the detector is exposed cannot be determined by comparing the amplitude of the detected response with that of the whole myoball exposed to the same concentration of ACh. Notwithstanding, examination of the dose-response relationship for nicotinic receptors (see Fig. 1) reveals that at low agonist concentrations, receptor current is almost linearly related to concentration. Within this linear region [ACh] I/10ⁿH - 1 where I is the peak receptor current and n_H is the Hill coefficient (where n_H = 2·1; see Allen & Brown, 1996; Allen, 1997). This linear relationship is maintained for all ACh concentrations that elicit a response that is less than or equal to approximately one-sixth of I_max. Therefore, provided the concentration of ACh to which individual receptors in the myoball are exposed does not exceed this region of linearity, then changes (reductions) in receptor current can be directly related to changes in ACh concentration using the above relationship in the same way as for patches. To ensure that recordings were confined to this region of linearity, a check was made to ensure that, under control conditions, the amplitude of the current responses recorded from individual release sites could be enhanced at least 6-fold by, for example, raising the extracellular Ca²⁺ concentration (see Fig. 1). This being so, then changes in the myoball receptor current could still be related to changes in ACh concentration using the same correction procedure as employed by Allen & Brown (1996). Note, in initial experiments, a direct check was made in each case to determine whether the amplitude of the current was less than or equal to approximately one-sixth of the maximum. After carrying out this procedure for some time it became clear that this condition was met in almost all cases provided the amplitude of the initial control current response in 1·2 mM Ca²⁺ did not exceed 2 nA. Thereafter, this check was only routinely carried out when the size of the initial control response exceeded this value.

		View larger version [in this window] [in a new window]
Figure 1. Calibration of myoball detectors A, log-log plot of mean peak agonist-induced current versus ACh concentration for a typical excised myoball membrane patch. The curve was constructed according to a binding scheme of the form I/Imax = 1/(1 + K/A)nH, where A is the agonist concentration, nH is the Hill coefficient, K is the equilibrium constant and I is the response amplitude. The specific values of Imax, K and nH used to construct the curve were mean values obtained from 11 patches reported by Allen & Brown (1996) with Imax, K and nH being 640 pA, 5·3 µM and 2·1, respectively. From the form of this curve, it can be seen that provided I <= 1/6 Imax (i.e. it lies within the roughly linear region of the curve), [ACh] I/10nH - 1. In the case of a myoball, this means that provided the amplitude of the detector response from a given site does not exceed 1/6 of maximum, then assuming the binding constants of the receptors in the excised patches and whole myoball are the same, the same linear relationship can be applied to relate changes in detector current to changes in ACh concentration. B, the response to detection of ACh release from a discrete site on a basal forebrain neurone in response to a single action potential (i) under control conditions, 1·2 mM extracellular Ca2+/Mg2+, and (ii) in elevated (5 mM) Ca2+ and 0 mM Mg2+. From this it can be seen that under the control conditions used in the present study, the peak response meets the criterion of being 1/6 Imax. Note, the individual records show the mean and S.D. values for 5-7 repetitions of single action potential-evoked release.

Data acquisition and analysis

Voltage command pulses were simultaneously generated and sampled on-line using a Labmaster (TL-1-125) DMA interface (Axon Instruments) connected to an IBM PC clone (Dell 466/MX) computer. Whole-cell data from basal forebrain neurones were filtered at 10 kHz. Myoball nAChR currents were filtered at 2 kHz using an 8-pole Bessel filter before acquisition at a sampling rate of 20 kHz using Axotape software (Axon Instruments). Subsequent analysis was carried out using the pCLAMP 6 suite of software (Axon Instruments). Curve fitting and additional data analysis were carried out using Origin (Microcal Software Inc., Northampton, MA, USA), Graphpad Prism (Graphpad, CA, USA) or CorelDraw (Corel graphics, Corel Corporation, Ontario, Canada). Unless stated otherwise, all data are expressed as means ± S.E.M. Note, the significance of individual drug effects upon both the Ca²⁺ current and release in individual cells was assessed using a Mann-Whitney test, with a P value of < 0·05.

Drugs

All drugs were applied via the superfusing Krebs solution. Acetylcholine chloride (ACh), ATP, amiloride hydrochloride, 4-aminopyridine, apamin, CdCl₂, charbydotoxin, GTP, NiCl₂, nifedipine and kynurenic acid were from Sigma. Tetrodotoxin and -conotoxin GVIA were from Calbiochem. (±)Muscarine chloride and Bay K 8644 were from Semat Technical Ltd (Herts, UK). -Agatoxin IVA and -conotoxin MVIIC were from The Peptide Institute (Osaka 562, Japan). Bicuculline methochloride was from Tocris.

	RESULTS

Top Abstract Introduction Methods Results Discussion References

Somatic Ca²⁺ channel subtypes

In a previous study of the somatic Ca²⁺ current in cholinergic basal forebrain neurones we reported that these cells express T-, N- and a small number of L-type Ca²⁺ channels (Allen et al. 1993). In addition, a substantial component of the high threshold Ca²⁺ current was found to be insensitive to any of the selective Ca²⁺ channel blockers available at that time. Other high voltage-activated Ca²⁺ channel subtypes, namely P-, Q- and R-types, have subsequently been described (Mintz et al. 1992; Wheeler et al. 1994; Randall & Tsien, 1997). The present study began with an investigation of which of these additional Ca²⁺ channel subtypes are expressed by basal forebrain neurones. Ca²⁺ currents were evoked by stepping to 0 mV (close to the peak of the Ca²⁺ current) for 100 ms from a holding potential of -80 mV in 2·5 mM extracellular Ca²⁺ (Ca) (see Fig. 2A and B). Selective block of P-type Ca²⁺ channels using low concentrations of -agatoxin IVA (-Aga IVA; 20 nM) inhibited only a small fraction of the peak Ca²⁺ current (9·4 ± 2·2 %, n = 10 cells). By contrast, a higher concentration of -Aga IVA (200 nM), which blocks additional Q-type channels, inhibited 30·5 ± 2·6 % (n = 16) of the peak high voltage-activated current (see Fig. 2A-C). In agreement with previous reports (Allen et al. 1993; Margeta-Mitrovic et al. 1997; Williams et al. 1997b), inhibition of N-type channels with -conotoxin GVIA (-CgTX; 100 nM) reduced I_Ca by approximately one-third (34·6 ± 2·1 %, n = 22), whilst inhibition of L-type Ca²⁺ channels with nifedipine (3 µM) inhibited 10·1 ± 0·8 % of peak I_Ca (n = 20; see Fig. 2). In the presence of all of these blockers, plus amiloride (300 µM) to block low voltage-activated T-type channels, there was a residual component of current (26·3 ± 2·5 %, n = 17). This current was also resistant to block by -conotoxin MVIIC (3 µM, n = 4) and thus was presumed to result from Ca²⁺ entry through R-type channels (see Fig. 2).

		View larger version [in this window] [in a new window]
Figure 2. Somatic Ca2+ channel subtypes expressed by cholinergic basal forebrain neurones A, plot of the time course of the block of the Ca2+ current by -Aga IVA, -CgTX and nifedipine in a typical basal forebrain neurone. B, data from the same cell as in A, showing somatic Ca2+ currents evoked by stepping from a VH of -80 mV to 0 mV for 100 ms (2·5 mM Ca2+o) under control conditions and in the presence of various selective Ca2+ channel blockers: low (20 nM) and high (200 nM) concentrations of -Aga IVA to block P-type and P/Q-type channels, respectively; -CgTX (100 nM) to block N-type channels; and nifedipine (3 µM) to block L-type Ca2+ channels. In each case, currents have been corrected for leak, by subtracting the current recorded in the presence of 200 µM Cd2+. C, cumulative data showing the relative contributions played by the different Ca2+ channels to the peak somatic Ca2+ current. All values are means ± S.E.M. for the number of cells indicated above the individual histogram bars.

Muscarinic modulation of the different Ca²⁺ channel subtypes

ACh and muscarinic agonists inhibit a substantial component of the high threshold Ca²⁺ current in basal forebrain neurones (see Fig. 3A, and Allen & Brown, 1993). This effect is mediated by M₂ muscarinic receptors with about half of the observed inhibition (54 %) being attributable to a reduction in Ca²⁺ entry through N-type Ca²⁺ channels. At the time of our original study, we were unable to determine the identity of the remaining agonist-sensitive Ca²⁺ current. This study examined whether this current results from muscarinic receptor-mediated inhibition of the Q- or R-type Ca²⁺ channels. The effects of ACh upon the individual high voltage-activated Ca²⁺ channel subtypes were examined after isolating them pharmacologically using a combination of selective Ca²⁺ channel blockers. In order to determine the effect of cholinergic agonists upon the R-type current, cells were pre-incubated for 30-60 min in saturating concentrations of -CgTX (200 nM), -Aga IVA (200 nM) and nifedipine (3 µM) to block N-, Q- and L-type Ca²⁺ channels. Pre-incubation was carried out prior to patching the cells in order to minimize any Ca²⁺ current run-down during the equilibration period. Under these conditions, application of ACh (100 µM) resulted in little or no inhibition of the R-current (9·52 ± 2·8 %, n = 5; see Fig. 3B), which equates to about 2·2 % of the control high threshold Ca²⁺ current. Modulation of L-type channels was examined by pre-incubating cells with -Aga IVA (200 nM) and -CgTX (200 nM) followed by Bay K 8644 (3 µM) to enhance the L-current. Under these conditions, ACh (100 µM) produced no discernable inhibition of the Bay K 8644-enhanced current (n = 4; see Fig. 3C). From these data it can be concluded that L- and R-type Ca²⁺ channels are not modulated to any significant degree by cholinergic agonists. Similarly, the low threshold T-type current was unaffected by cholinergic agonists (see Fig. 3F). This left only N- and Q-type channels as potential sites for cholinergic modulation. After blocking Q-type channels with -Aga IVA (200 nM), ACh (30 µM) inhibited the remaining current by 45·5 ± 4·1 % (n = 4). This equates to about 91 % inhibition of current flow through N-type channels (see Fig. 3D). Similarly, following incubation with -CgTX (200 nM) to block N-type channels, application of ACh (30 µM) inhibited the remaining Ca²⁺ current by 33·9 ± 4·2 % (n = 4) indicating that muscarinic agonists inhibit current flow through Q-type channels by approximately 75 % (Fig. 3E). Thus in basal forebrain neurones, muscarinic receptors couple to inhibition of N- and Q-type somatic Ca²⁺ channels, with the two channel subtypes contributing in roughly equal measure to the total agonist-sensitive current.

		View larger version [in this window] [in a new window]
Figure 3. Muscarinic receptor modulation of the different Ca2+ channel subtypes A, Ca2+ current I-V relationship generated using a voltage ramp from -80 to +60 mV (ramp rate, 1·4 V s-1) recorded under control conditions and in the presence of ACh (30 µM) or Cd2+ (200 µM). B, effect of ACh (30 µM) upon the R-type high threshold Ca2+ current after pre-incubation for 40 min prior to patching the cell, and continued application during the recording, of -Aga IVA (200 nM), -CgTX (200 nM) and nifedipine (3 µM). C, effect of ACh (30 µM) upon the Bay K 8644-enhanced (3 µM) L-type Ca2+ current evoked after inhibition of P-, Q- and N-type channels by pre-incubation of the cell for 45 min prior to patching the cell, and continued application during the recording, of -Aga IVA (200 nM) and -CgTX (100 nM). D, inhibition of the high threshold Ca2+ current by ACh (30 µM) after blocking Q-type channels by pre-incubating with -Aga IVA (200 nM) for 11 min. E, inhibition of the high threshold Ca2+ current by ACh (30 µM) after blocking N-type channels by pre-incubation with -CgTX (200 nM) for 7 min. F, lack of effect of ACh upon the T-type low threshold Ca2+ current evoked by stepping from -80 to -30 mV. Note, [Ca2+]o was 2·5 mM throughout except in B where it was 5·0 mM.

Ca²⁺ entry in response to simulated action potentials

Whilst conventional step voltage protocols allow us to determine which Ca²⁺ channel subtypes are expressed by basal forebrain cells, they tell us little about the contribution each makes to total Ca²⁺ influx following an action potential. In order to answer this, the relative contributions played by each Ca²⁺ channel subtype to total Ca²⁺ entry during a simulated action potential were measured. The desired waveform was obtained by recording a somatic action potential from a basal forebrain cell and then scaling it for use as the command potential for the voltage-clamp amplifier (see Methods). Selective Ca²⁺ channel blockers were then used in the same way as for the step protocol experiments in order to assess the contribution made by each channel subtype to Ca²⁺ entry (see Fig. 4). When using an action potential waveform, peak Ca²⁺ entry occurred during the repolarizing phase of the action potential at between -35 and -50 mV (see Fig. 4B). Furthermore, because of the greater electrochemical driving force, the peak amplitude of the current was larger than observed using step commands to 0 mV. The relative contributions played by each channel to total Ca²⁺ entry are shown in Fig. 4C. From this it can be seen that N- and Q-type channels play a dominant role, contributing 34·8 ± 3·1 % (n = 19) and 24·5 ± 2·5 % (n = 17), respectively, to total Ca²⁺ entry. In addition, there was significant Ca²⁺ entry due to influx through T-type channels. Selective inhibition of T-type channels with amiloride (300 µM) reduced total Ca²⁺ entry by between 8·7 and 41 % (18·1 ± 2·5 %, n = 12; see Fig. 4D). This wide variation in the contribution played by T-type channels to net Ca²⁺ influx was particularly striking using action potential waveforms. This was due to the fact that, as a result of their relatively slow deactivation kinetics, T-type channels act to prolong Ca²⁺ entry and thus make a disproportionately large contribution to total as compared with peak Ca²⁺ influx. In this way, small differences in the absolute number of T-type channels expressed by a cell can have a large effect upon total Ca²⁺ influx (see Fig. 4D). Furthermore, as a result of their lesser contribution to peak as opposed to total Ca²⁺ entry, muscarinic inhibition of high threshold N- and Q-type channels produced a significantly larger reduction in peak (45·94 ± 2·78 %, n = 8) as opposed to total Ca²⁺ influx (30·94 ± 3·9 %, n = 8).

		View larger version [in this window] [in a new window]
Figure 4. Ca2+ entry during a simulated somatic action potential A, plot of the time course of the block by amiloride (300 µM), -Aga IVA (20 and 200 nM), -CgTX (100 nM), nifedipine (3 µM) and CdCl2 (200 µM) of the peak whole-cell Ca2+ current. Individual currents were generated every 10 s using a simulated action potential waveform (2·5 mM Ca2+o). B, data from the same cell as in A, showing individual currents under control conditions and after equilibration with each of the different blockers. C, cumulative data showing the relative contribution played by each of the different Ca2+ channel subtypes to total Ca2+ entry following an action potential. All values are the means ± S.E.M. for the number of cells indicated above the individual histogram bars. D, the contribution of amiloride- sensitive (300 µM) T-type channels to peak and total Ca2+ influx in a cell expressing a larger number of somatic T-type channels. All currents shown in B and D have been corrected for leak by subtracting the current recorded in the presence of 200 µM Cd2+.

As observed using voltage step protocols (see earlier), block of Ca²⁺ entry through L-type channels with nifedipine (3 µM) produced only a small reduction in Ca²⁺ influx (10·9 ± 1·2 %, n = 18). In almost all cells, there was a substantial residual component of Ca²⁺ influx remaining after selective inhibition of N-, Q- , L- and T-type channels that resulted from Ca²⁺ entry through R-type Ca²⁺ channels (19·1 ± 3·6 %, n = 6).

Detection of transmitter release with myoballs

Figure 5 shows examples of ACh release from a discrete site on the neurite of a basal forebrain neurone detected using a skeletal myoball as a probe. The basal forebrain cell was held at resting potential and stimulated to fire by injecting brief (4-6 ms) DC current pulses through the recording electrode. Recording from myoballs was made using the perforated (amphotericin) patch-clamp technique, with the myoball voltage clamped at -80 mV. The myoball was placed in close contact with the neurite of a basal forebrain cell and slowly moved along until a site of release was detected (detection of a release site being associated with a brief latency, rapidly rising inward current due to the activation of nAChRs in the myoball in response to propagation of an action potential). Due to their relatively large size (15-20 µm diameter), myoballs would occasionally overlap more than one release site, resulting in the generation of a multipeak response. In such cases, the myoball was re-positioned until release from only one of the sites could be detected. Such multipeak responses did not appear to be the result of activation of recurrent synaptic inputs as they were unaffected by inhibiting GABAergic or glutamatergic receptors with kynurenic acid (3-10 mM) or bicuculline (10-30 µM). Secondly, the decay time constant of the myoball detector currents varied quite considerably between sites and, unlike events detected using excised membrane patches (Allen & Brown, 1996), were generally significantly longer than would be predicted from mean single channel open time. The most likely explanation for this is that, in the absence of endogenous cholinesterases to break down the released ACh rapidly, there may be significant lateral diffusion of ACh resulting in asynchronous activation of receptors over a greater area of the myoball.

		View larger version [in this window] [in a new window]
Figure 5. Effects of blocking N- or Q-type Ca2+ channels upon single action potential-evoked release A, upper panel, graph of peak myoball detector current generated in response to ACh release by single action potentials evoked every 60 s under control conditions and in the presence of -CgTX (100 nM). Middle and lower panels, original data from the same experiment. Under control conditions (1·2 mM Ca2+o; left-hand middle panel), stimulation of ACh release evoked a large rapidly activating and inactivating inward current due to activation of the nAChRs in the detector myoball. After incubation with -CgTX (100 nM) for 7 min to block Ca2+ entry through N-type channels, release was largely abolished (centre panel). Subsequently elevation of [Ca2+]o to 5 mM in the continued presence of -CgTX restored release to near control levels. B, similarly, inhibition of Ca2+ entry through Q-type channels by incubation of the cell with -Aga IVA (200 nM) for 20 min greatly reduced ACh release. In the continued presence of -Aga IVA release could again be restored to near control levels by elevating [Ca2+]o to 5 mM. Note, each of the records shown is the mean and S.D. of several repetitions (4-6) of single action potential-evoked responses.

Ca²⁺ channel subtypes underlying release

The relative importance of the different Ca²⁺ channel subtypes in triggering ACh release in response to an individual action potential was examined by studying the effects of various selective Ca²⁺ channel blockers upon the peak amplitude of the myoball detector current. Unless otherwise stated, all percentage inhibition values are for application of blockers to naive cells in the presence of 1·2 mM Ca, and are given after correction for the non-linear concentration- response relationship of the nAChRs (see Methods).

Block of Ca²⁺ entry through either N- or Q-type Ca²⁺ channels using -CgTX (100 nM) or -Aga IVA (200 nM) reduced ACh release by 82·7 ± 3·4 % (n = 6) and 63 ± 4·6 % (n = 15), respectively (see Figs 5 and 8). Application of low concentrations of -Aga IVA (20 nM) to inhibit P-type Ca²⁺ channels selectively had no significant effect upon release (P > 0·05, n = 4). Similarly, application of Ni²⁺ (10-30 µM, n = 13), which selectively inhibits T-type Ca²⁺ channels (EC₅₀, 5 µM; Allen et al. 1993), or modulation of Ca²⁺ entry through L-type channels with nifedipine (3 µM, n = 15) or Bay K 8644 (10 µM, n = 6) had no significant effect (P > 0·05) upon release (see Figs 6 and 8). Note, Ni²⁺ rather than amiloride was used to block presynaptic T-type channels as amiloride had a direct inhibitory effect upon the nAChRs in the myoball detector. R-type channels in some preparations have been reported to be relatively sensitive to Ni²⁺ (IC₅₀, 86 µM; Yu & Shinnick-Gallagher, 1997) whilst in rat cerebellar granule cells there appear to be multiple types of R-type channel with low and high affinities for Ni²⁺ (IC₅₀, 4 and 153 µM, respectively; see Tottene et al. 1996). In basal forebrain neurones, Ni²⁺ at concentrations between 10 and 30 µM subtantially inhibited the T-type somatic current but had little or no effect upon the R-type high threshold component of the current (data not shown). However, unlike either of the R-type channels expressed by cerebellar neurones, the R-type channels underlying release from basal forebrain cells were not inactivated by membrane depolarization to -50 mV, suggesting that they are different from either of these subtypes (see Fig. 7E).

		View larger version [in this window] [in a new window]
Figure 6. Effect of inhibiting L- and T-type Ca2+ channels upon evoked release A, the effect of applying nifedipine (3 µM) to block L-type channels. B, the effect of inhibiting T-type channels with Ni2+ (10 µM). In each case, records show the release evoked in response to single action potentials in 1·2 mM Ca2+o before and after application of the blockers. Note, each of the records illustrated is the mean and S.D. of 4-8 individual responses evoked at 1 min intervals.

		View larger version [in this window] [in a new window]
Figure 7. Transmitter release triggered by Ca2+ entry through R-type Ca2+ channels A, control ACh release in response to single action potentials in 1·2 mM Ca2+o. B, release after 11 min incubation with -Aga IVA (200 nM) to block Q-type channels. C, raising [Ca2+]o to 5 mM in the continued presence of -Aga IVA restores release to near control levels. D, after an additional 9 min incubation in -Aga IVA (200 nM) plus -CgTX (100 nM) with 5 mM Ca2+o a small component of release could still be detected. E, after subsequent addition of nifedipine (3 µM) to block L-type channels and depolarization of the cell to -50 mV using elevated K+ to inactivate T-type channels, release could still be detected as the result of Ca2+ entry through residual 'R-type' Ca2+ channels (7·5 mM Ca2+o). Note, each of the records shown is the mean and S.D. of 4-9 individual responses evoked at 1 min intervals.

It is clear from the magnitude of the observed inhibition of release produced by blocking Ca²⁺ entry through either N- or Q-type channels that, under normal conditions, significant ACh release can only be evoked if there is simultaneous Ca²⁺ entry through the two channel subtypes. However, after blocking entry via one of these routes, ACh release could be restored to near control levels simply by elevating [Ca²⁺]_o to between 5 and 7·5 mM, thereby increasing Ca²⁺ entry through the remaining N- or Q-type channels (see Fig. 5A and B). High levels of [Ca²⁺]_o have been reported to reduce toxin binding and this could potentially contribute to the Ca²⁺-dependent restoration of release observed in elevated Ca²⁺ (Witcher et al. 1993). This possibility was tested for by measuring the inhibition of the peak somatic Ca²⁺ current produced by -CgTX (100 nM) for extracellular Ca²⁺ concentrations over the range 1·2-7·5 mM, but no significant difference in inhibition of peak I_Ca at the different [Ca²⁺]_o values was observed (P > 0·05, n = 3). Thus, it would appear that release is determined by the net Ca²⁺ influx through the two channel subtypes acting in concert at individual release sites.

In the presence of all of the above selective Ca²⁺ channel blockers, a small amount of release could also be detected as a result of Ca²⁺ entry through R-type Ca²⁺ channels (see Fig. 7). Therefore, ACh release from basal forebrain cells appears to be mediated by Ca²⁺ influx through N-, Q- and to a much lesser degree R-type Ca²⁺ channels.

Cholinergic modulation of release

Activation of M₂ muscarinic receptors leads to inhibition of ACh release from basal forebrain neurones (Allen & Brown, 1996). However, the underlying ionic basis of this inhibitory mechanism is unknown. A variety of different neurotransmitters have been shown to exert their actions on presynaptic terminals and transmitter release by way of G-protein-linked receptors coupled to various voltage-dependent Ca²⁺ and K⁺ channels (Fletcher & Chiappinelli, 1993; Yawo et al. 1994). Therefore the role of these channels in cholinergic feedback inhibition of ACh release from basal forebrain neurones was examined.

Voltage-gated Ca²⁺ channels. The coupling of muscarinic receptors to the different Ca²⁺ channel subtypes was examined after pharmacologically isolating the components of release evoked by the different Ca²⁺ channel subtypes. In all cases, irrespective of [Ca²⁺]_o (range, 1·2-7·5 mM), ACh release triggered by Ca²⁺ entry through N-, Q- or R-type Ca²⁺ channels could be reversibly inhibited by the application of muscarine (10 µM; see Fig. 9). The mean inhibitions produced by muscarine (10 µM) on the N-, Q- and R-type channel-evoked release were 86·5 % (n = 2), 86·2 % (n = 3) and 94·8 % (n = 4), respectively. Note, muscarine (10 µM) had no effect upon the response of the myoball to exogenously applied ACh (1-3 µM). Cholinergic modulation of release due to each channel subtype could be totally inhibited by pre-treating the cell with pertussis toxin (500 ng ml^-1, n = 4).

		View larger version [in this window] [in a new window]
Figure 8. Effect of selective Ca2+ channel blockers upon control release Data shown are of the inhibition of single action potential-evoked ACh release produced by exogenous application of -CgTX (100 nM), -Aga IVA (200 nM), nifedipine (3 µM) or Ni2+ (30 µM) to naive cells in 1·2 mM Ca2+o in order to block Ca2+ entry through N-, Q-, L- and T-type Ca2+ channels, respectively. The histogram bars on the left of the graph show the percentage inhibition in terms of a change in measured myoball detector current, whilst those on the right show the same data expressed in terms of the change in ACh concentration after correcting for the non-linear concentration-response characteristics of the nAChRs (see Methods). All data are the means ± S.E.M. for the number of cells indicated above each bar.

		View larger version [in this window] [in a new window]
Figure 9. Muscarinic receptor-mediated inhibition of N-, Q- and R-type Ca2+ channel-evoked release A, single action potential-evoked release due to Ca2+ entry through N-type channels (5·0 mM Ca2+o) recorded under control conditions (after 10 min pre-incubation in 200 nM -Aga IVA, 3 µM nifedipine and 30 µM Ni2+; left-hand panel) and after application of 30 µM muscarine (right-hand panel). B, ACh release due to Ca2+ entry through Q-type channels under control conditions (after 10 min pre-incubation with 100 nM -CgTX, 3 µM nifedipine and 30 µM Ni2+; left-hand panel) and after application of 30 µM muscarine. C, release arising from Ca2+ entry through R-type channels following 12 min pre-incubation with all of the above blockers (7·5 mM Ca2+o; left-hand panel) and after application of 30 µM muscarine. Note, each of the records shown is the mean and S.D. of 4-9 individual responses evoked at 40 s intervals.

Inward rectifier K⁺ channels. Basal forebrain neurones express a G-protein-coupled somatic inward rectifier current that is partially activated at resting potential and helps to regulate neuronal excitability (Stanfield et al. 1985). These channels can be inhibited by substance P, neurotensin and low concentrations of Ba²⁺ (Stanfield et al. 1985; Farkas et al. 1994; Sim & Allen, 1998). In addition, substance P has been reported to enhance the release of ACh within the cortex (Feuerstein et al. 1996). However, application of substance P (10-100 nM, n = 3) or Ba²⁺ (10 µM, n = 4) had no effect (P > 0·05) upon either ACh release following a single action potential or subsequent cholinergic inhibition of release (see Fig. 10A). Mean inhibition by 10 µM muscarine in the presence of substance P or Ba²⁺ was 89·3 ± 5·4 % (n = 4) compared with a value of 91·4 ± 3·0 % (n = 5) under control conditions.

Ca²⁺-activated K⁺ channels. Cholinergic basal forebrain neurones express a variety of different somatic Ca²⁺-activated K⁺ conductances which underlie the spike after-hyperpolarization (Gorelova & Reiner, 1996; Williams et al. 1997b). However, application of apamin and charybdotoxin (100 nM) to block SK and BK channels, respectively, had no significant effect upon control release evoked in response to a single action potential (Fig. 10B and C). The mean depression by the two toxins was 0·16 ± 4·17 % (n = 5) and 1·7 ± 4·65 % (n = 6), respectively. Similarly, neither blocker affected the degree of inhibition produced by muscarine (10 µM; mean inhibition, 94·9 ± 3·3 %, n = 5).

		View larger version [in this window] [in a new window]
Figure 10. Effects of selective K+ channel blockers upon evoked ACh release A, application of Ba2+ (10 µM) to selectively inhibit KIR channels in magnocellular basal forebrain neurones had no effect upon either control release (middle panel; 1·2 mM Ca2+o) or its modulation by 10 µM muscarine (right-hand panel). B and C, similarly, incubation (5 min) with 100 nM apamin or charybdotoxin to inhibit selectively SK- and BK-type Ca2+-activated K+ channels, respectively, had no effect upon control release in response to individual action potentials (centre panels) or its modulation by exogenous application of 10 µM muscarine (right-hand panels). Note, all traces are the mean and S.D. of between 4 and 9 individual action potential-evoked release events generated every 40 s.

	DISCUSSION

Top Abstract Introduction Methods Results Discussion References

This paper reports on the Ca²⁺ channel subtypes underlying ACh release from basal forebrain neurones and their modulation by muscarinic receptors. In common with many other central neurones, basal forebrain neurones express a wide variety of different Ca²⁺ channel subtypes. Of the six native subtypes that have been described (Dolphin, 1995) all have been reported in basal forebrain neurones from various species (Allen et al. 1993; Williams et al. 1997b; Margeta-Mitrovic et al. 1997). In the present study of rat neurones, T-, N-, L-, Q-, and R-type somatic Ca²⁺ channel subtypes were distinguished. However, of these, only the N- and Q-type channels were found to be subject to modulation by cholinergic agonists and in each case this was mediated via M₂ muscarinic receptors coupled to a pertussis toxin-sensitive G-protein.

In this study, an attempt was also made to determine the relative importance of the different channel subtypes to the net influx of Ca²⁺ that results from a relevant physiological stimulus such as an action potential. Previous attempts to address this question, by using simulated action potential waveforms to activate the different Ca²⁺ channels in other neurones, have reported a number of important differences in both the magnitude of the agonist-induced inhibition (Penington et al. 1992) and the relative contributions of the different subtypes to total and peak Ca²⁺ influx (McCobb & Beam, 1991; Scroggs & Fox, 1992; Wheeler et al. 1996; Randall & Tsien, 1997). In the present study, similar differences were observed. In particular, it was noted that, as a consequence of their relatively slow activation and deactivation kinetics, T-type channels contributed proportionally more to total Ca²⁺ influx than to peak Ca²⁺ entry following an action potential. The increased contribution made by low threshold T-type channels under these conditions is also likely to explain the smaller agonist-induced inhibition of total compared with peak Ca²⁺ influx (31 and 46 %, respectively) as T-type channels unlike N- and Q-type channels were not subject to inhibition by muscarinic agonists.

With a few notable exceptions, the Ca²⁺ channels involved in transmitter release from the presynaptic terminals of central neurones are inaccessible to direct study using conventional electrophysiological techniques. However, in many instances indirect evidence as to their role can be obtained by using brain slice preparations combined with selective Ca²⁺ channel blockers to look at their effects upon synaptic transmission. Unfortunately, cholinergic transmission between the basal forebrain and their cortical targets is mediated by slow G-protein-linked muscarinic receptors, where the postsynaptic responses are difficult to evoke and prone to run-down. Furthermore, autoinhibition of release is also mediated by muscarinic receptors and it is very difficult to inhibit these receptors selectively without also affecting those in the postsynaptic membrane. However, by substituting myoball nAChRs for the normal postsynaptic muscarinic receptors many of these problems can be overcome, thus making it possible to isolate and study presynaptic muscarinic receptor-mediated modulation of Ca²⁺ channels upon ACh release from individual sites on basal forebrain neurones. Using this technique it was possible to demonstrate here that ACh release involves Ca²⁺ entry through N- and Q-type channels with a small additional contribution from R-type channels. Transmitter release involving N-, P- and Q-type Ca²⁺ channels is fairly widespread throughout the central nervous system (Wheeler et al. 1994; Mintz et al. 1995) and there have been several reported examples of a similar combination of N- and Q-type channels regulating release at other synapses (Wheeler et al. 1994). The present results indicate that N-type channels play a slightly more dominant role than Q-type channels in controlling ACh release from basal forebrain neurones. However, in isolation, the two subtypes were capable of triggering roughly comparable levels of ACh release. This finding is largely consistent with the relative contribution each subtype makes to the somatic Ca²⁺ current. However, in marked contrast to the clear additive nature of their contributions to total Ca²⁺ influx, the effect on the release process of selectively antagonizing these channels was clearly non-additive, with inhibition of Ca²⁺ entry through N- or Q-type channels reducing levels of ACh release by 82·7 and 63 %, respectively. This is similar to results reported for cerebellar synapses (Mintz et al. 1995); it indicates that calcium ions entering through the different Ca²⁺ channel subtypes co-operate, and that the intracellular Ca²⁺ microdomains of the individual channels overlap to influence release at a given release site. The consequences of this for basal forebrain neurones are that, at normal physiological [Ca²⁺]_o (1·2- 1·5 mM; Moghaddam & Bunney, 1989; Puka-Sundvall et al. 1994), there appears to be a requirement for simultaneous Ca²⁺ entry through N- and Q-type channels before significant amounts of ACh release can be evoked. Furthermore, the presence of multiple Ca²⁺ channel subtypes interacting at a given release site coupled with the highly co-operative effect of calcium ions upon transmitter release (Dodge & Rahaminoff, 1967) may allow for exquisite fine-tuning of synaptic transmission by subtle modulation of any of the associated Ca²⁺ channels.

In addition to N- and Q-type channels it was observed that R-type channels were also capable of triggering small amounts of ACh release from basal forebrain neurones. In contrast to the commonly reported role of N-, P- and Q-type channels in neurotransmission, R-type channels have only been shown to be involved in transmitter release at the calyx synapse of the rat medial nucleus of the trapezoid body (Wu et al. 1998). At this synapse, they are responsible for 26 % of total presynaptic Ca²⁺ influx, a value which is similar to the contribution these channels make to somatic Ca²⁺ influx (approximately 20 %) following an action potential in basal forebrain neurones. However, even assuming R-type channels are expressed in similar numbers at the presynaptic site, the relatively minor role they play in triggering ACh release from basal forebrain cells indicates either that they are less efficiently coupled than N- or Q-type channels or that they serve some other as yet undetermined role. In contrast to N-, Q- and R-type channels, neither L- nor T-type channels were observed to contribute to ACh release. In the case of L-type channels, this is perhaps not all that surprising if their expression level at the release site mirrors that of the soma. Furthermore, even in neurones that express greater numbers of L-type channels, there are only a few reports of them playing a direct role in transmitter release (Perney et al. 1986; Huston et al. 1990; Tachibana et al. 1993). Likewise, there is only limited evidence of a direct role for T-type channels in release. However, based upon the ability of T-type channels to inject significant amounts of Ca²⁺ over prolonged periods it has been suggested that they may play a modulatory role similar to NMDA receptors (Christie et al. 1997). Alternatively, their involvement in the generation of low threshold spikes and the production of brief intense bursts of firing (Suzuki & Rogawski, 1989) could result in an indirect modulation of transmitter release by transiently relieving voltage-dependent inhibition of presynaptic Ca²⁺ channels (Williams et al. 1997a).

The release of ACh from cortical and subcortical structures innervated by basal forebrain neurones can be modulated by muscarinic agonists acting on what are believed to be M₂ muscarinic autoreceptors (Quirion et al. 1994). Consistent with this, it was shown that ACh release from discrete sites along the neurites of cultured basal forebrain neurones can also be inhibited by activation of M₂ muscarinic receptors, and that these receptors couple to inhibition of various components of the somatic Ca²⁺ current (Allen et al. 1993; Allen & Brown, 1996). In the present study it was observed that application of muscarinic agonists largely abolished all ACh release triggered by Ca²⁺ entry through N-, Q- and R-type Ca²⁺ channels. The observed inhibition of N- and Q-type channel-evoked release is consistent with the actions of muscarinic agonists upon the somatic channels. However, whilst a substantial R-type channel contribution to the somatic current was observed, R-type channels were not inhibited to any significant degree by muscarinic agonists, indicating that there may be differences in the receptor coupling at the two sites.

Alternatively, the apparently promiscuous coupling of muscarinic receptors to N-, Q- and R-type presynaptic Ca²⁺ channels could indicate that the primary site of action of muscarinic agonists is not the Ca²⁺ channels but for example K⁺ channels. One potential site of action could be through activation of K_IR channels. However, whilst a variety of different neurotransmitters have been reported to inhibit K_IR channels in these cells (Stanfield et al. 1995; Margeta-Mitrovic et al. 1997), there have been no reports of their activation by agonists. Muscarinic agonists have been reported to modulate various other K⁺ channels in these cells, although in each case these effects would again be predicted to be excitatory rather than inhibitory (Harata et al. 1991; Sim & Griffith, 1991). In the present study, a variety of different K⁺ channel blockers were investigated including Ba²⁺, substance P, apamin and charybdotoxin, but in each case no effect was observed upon either control levels or receptor-mediated modulation of ACh release. It was possible to enhance ACh release using less selective K⁺ channel blockers such as 4-aminopyridine and tetraethylammonium (author's unpublished data). However, the actions of these compounds were hard to interpret with any degree of confidence as they both antagonized the muscarinic receptors on the basal forebrain cells (Allen & Brown, 1993) and also the nAChRs expressed by the myoball. Notwithstanding, at relatively low concentrations, the release evoked in the presence of these compounds could still be inhibited by muscarine, albeit at high concentrations, further arguing against a role for K⁺ channels as the target for muscarinic receptor-mediated inhibition of ACh release.

In conclusion, ACh release from magnocellular basal forebrain neurones involves Ca²⁺ entry through N-, Q- and to a lesser extent R-type Ca²⁺ channels. Furthermore, there appears to be significant overlap in the intracellular Ca²⁺ microdomains of the different channel subtypes at a single release site. As a consequence of this, at normal physiological Ca²⁺ concentrations, there is a requirement for simultaneous Ca²⁺ entry through N- and Q-type Ca²⁺ channels before significant ACh release can be evoked. Finally, the muscarinic receptor-mediated autoinhibitory mechanism regulating ACh release from these sites appears to involve presynaptic inhibition of voltage-gated Ca²⁺ rather than K⁺ channels. However, although this may be considered to be the most likely mechanism, at present the possibility of a direct inhibitory effect upon the release machinery itself cannot be ruled out.

	REFERENCES

Top Abstract Introduction Methods Results Discussion References

Allen, T. G. J. (1997). The 'sniffer-patch' technique for detection of neurotransmitter release. Trends in Neurosciences 20, 192-197	[Medline]
Allen, T. G. J. (1998). Ca2+ channel subtypes underlying ACh release from rat basal forebrain neurones and their modulation by cholinergic agonists. The Journal of Physiology 509.P, 61P.
Allen, T. G. J. & Brown, D. A. (1993). M2 muscarinic receptor mediated inhibition of the Ca2+ current in rat magnocellular cholinergic basal forebrain neurones. The Journal of Physiology 466, 173-189	[Abstract]
Allen, T. G. J. & Brown, D. A. (1996). Detection and modulation of acetylcholine release from neurites of rat basal forebrain cells in culture. The Journal of Physiology 492, 453-466	[Abstract]
Allen, T. G. J., Sim, J. A. & Brown, D. A. (1993). The whole-cell calcium current in acutely dissociated magnocellular cholinergic basal forebrain neurones of the rat. The Journal of Physiology 460, 91-116	[Abstract]
Avery, E. E., Baker, L. D. & Asthana, S. (1997). Potential role of muscarinic agonists in Alzheimer's disease. Drugs and Aging 11, 450-459.	[Medline]
Christie, B. R., Schexnayder, L. K. & Johnston, D. (1997). Contribution of voltage-gated Ca2+ channels to homosynaptic long-term depression in CA1 region in vitro. Journal of Neurophysiology 77, 1651-1655	[Abstract/Full Text]
Dodge, F. A. & Rahaminoff, R. (1967). Co-operative action of calcium ions in transmitter release at the neuromuscular junction. The Journal of Physiology 193, 419-432	[Medline]
Dolphin, A. C. (1995). Voltage-dependent calcium channels and their modulation by neurotransmitters and G proteins. Experimental Physiology 80, 1-36	[Medline]
Farkas, R. H., Nakajima, S. & Nakajima, Y. (1994). Neurotensin excites basal forebrain cholinergic neurons: ionic and signal-transduction mechanisms. Proceedings of the National Academy of Sciences of the USA 91, 2853-2857	[Abstract]
Feuerstein, T. J., Gleichauf, O. & Landwehrmeyer, G. B. (1996). Modulation of cortical acetylcholine release by serotonin: The role of substance P interneurons. Naunyn-Schmiederberg's Archives of Pharmacology 354, 618-626.
Fletcher, G. H. & Chiappinelli, V. A. (1993). The actions of the Kappa 1 opioid agonist U-50, 488 on presynaptic nerve terminals of the chick ciliary ganglion. Neuroscience 53, 239-250	[Medline]
Gorelova, N. & Reiner, P. B. (1996). Role of the afterhyperpolarization in control of discharge properties of septal cholinergic neurons in vitro. Journal of Neurophysiology 75, 695-706	[Medline]
Harata, N., Tateishi, N. & Akaike, N. (1991). Acetylcholine receptors in dissociated nucleus basalis of Meynert neurons of the rat. Neuroscience Letters 130, 153-156	[Medline]
Huston, E. R., Scott, R. H. & Dolphin, A. C. (1990). A comparison of the effect of calcium channel ligands and GABAB antagonist on transmitter release and somatic calcium channel currents in cultured neurones. Neuroscience 38, 721-729	[Medline]
Iversen, S. D. (1997). Behavioural evaluation of cholinergic drugs. Life Sciences 60, 1145-1152	[Medline]
Kasa, P., Rakonczay, Z. & Gulya, K. (1997). The cholinergic system in Alzheimer's disease. Progress in Neurobiology 52, 511-535	[Medline]
Levey, A. I., Kitt, C. A., Simonds, W. F., Price, D. L. & Brann, M. R. (1991). Identification and localization of muscarinic autoreceptor proteins in brain with subtype-specific antibodies. Journal of Neuroscience 11, 3218-3226	[Abstract]
McCobb, D. P. & Beam, K. G. (1991). Voltage-clamp commands reveal striking differences in calcium entry via low and high voltage-activated calcium channels. Neuron 7, 119-127	[Medline]
Margeta-Mitrovic, M., Grigg, J. J., Koyano, K., Nakajima, Y. & Nakajima, S. (1997). Neurotensin and substance P inhibit low- and high-voltage-activated Ca2+ channels in cultured newborn rat nucleus basalis neurons. Journal of Neurophysiology 78, 1341-1352	[Abstract/Full Text]
Mintz, I. M., Sabatini, B. L. & Regehr, W. G. (1995). Calcium control of transmitter release at a cerebellar synapse. Neuron 15, 675-688	[Medline]
Mintz, I. M., Venema, V. J., Swiderek, K. M., Lee, T. D., Bean B. P. & Adams, M. E. (1992). P-type calcium channels blocked by the spider toxin -Aga-IVA. Nature 355, 827-829	[Medline]
Moghaddam, B. & Bunney, B. S. (1989). Ionic composition of microdialysis perfusing solution alters the pharmacological responsiveness and basal outflow of striatal dopamine. Journal of Neurochemistry 53, 652-654	[Abstract]
Penington, N. J., Kelly, J. S. & Fox, A. P. (1992). Action potential waveforms reveal simultaneous changes in ICa and IK by 5-HT in rat dorsal raphe neurons. Proceedings of the Royal Society B 248, 171-179	[Medline]
Perney, T. M., Hirning, L. D., Leeman, S. E. & Miller, R. J. (1986). Multiple calcium channels mediate neurotransmitter release from peripheral neurons. Proceedings of the National Academy of Sciences of the USA 83, 6656-6659	[Medline]
Puka-Sundvall, M., Hagberg, H. & Andine, P. (1994). Changes in extracellular calcium concentration in the immature rat cerebral cortex during anoxia are not influenced by MK-801. Brain Research and Developmental Brain Research 77, 146-150.
Quirion, R., Aubert, I., Lapchak, P. A., Schaum, R. P., Teolis, S., Gauthier, S. & Araujo, D. M. (1989). Muscarinic receptor subtypes in human neurodegenerative disorders: focus on Alzheimer's disease. Trends in Pharmacological Sciences Supplement, 80-84.
Quirion, R., Richard, J. & Wilson, A. (1994). Muscarinic and nicotinic modulation of cortical acetylcholine release monitored by in vivo microdialysis in freely moving adult rats. Synapse 17, 92-100	[Medline]
Rae, J., Cooper, K., Gates, P. & Watsky, M. (1991). Low access resistance perforated patch recordings using amphotericin-B. Journal of Neuroscience Methods 37, 15-26	[Medline]
Randall, A. D. & Tsien, R. W. (1997). Contrasting biophysical and pharmacological properties of T-type and R-type calcium channels. Neuropharmacology 36, 879-893	[Medline]
Robbins, T. W., McAlonan, G., Muir, J. L. & Everitt, B. J. (1997). Cognitive enhancers in theory and practice: studies of the cholinergic hypothesis of cognitive deficits in Alzheimer's disease. Behavioural Brain Research 83, 15-23	[Medline]
Schorderet, M. (1995). Alzheimer's disease: fundamental and therapeutic aspects. Experientia 51, 99-105	[Medline]
Scroggs, R. E. & Fox, A. P. (1992). Multiple Ca2+ currents elicited by action potential waveforms in acutely isolated adult rat dorsal root ganglion neurons. Journal of Neuroscience 12, 1789-1802	[Abstract]
Sim, J. A. & Allen, T. G. J. (1998). Morphological and membrane properties of rat magnocellular basal forebrain neurons maintained in culture. Journal of Neurophysiology 80, 1653-1669	[Abstract/Full Text]
Sim, J. A. & Griffith, W. H. (1991). Muscarinic agonists block a late-afterhyperpolarization in medial septum/diagonal band neurons in vitro. Neuroscience Letters 129, 63-68	[Medline]
Stanfield, P. R., Nakajima, Y. & Yamaguchi, K. (1995). Substance P raises neuronal membrane excitability by reducing inward rectification. Nature 315, 498-501.
Sugaya, K., Clamp, C., Bryan, D. & McKinney, M. (1997). mRNA for the m4 muscarinic receptor subtype is expressed in adult rat brain cholinergic neurons. Brain Research and Molecular Brain Research 50, 305-313.
Suzuki, S. & Rogawski, M. A. (1989). T-type calcium channels mediate the transition between tonic and phasic firing in thalamic neurons. Proceedings of the National Academy of Sciences of the USA 86, 7228-7232	[Medline]
Suzuki, T., Fujimoto, K., Oohata, H. & Kawashima, K. (1988). Presynaptic M1 muscarinic receptor modulates spontaneous release of acetylcholine from rat basal forebrain slices. Neuroscience Letters 84, 209-212	[Medline]
Tachibana, M., Okada, T., Arimura, T., Kobayashi, K. & Piccolino, M. (1993). Dihydropyridine-sensitive calcium current mediates neurotransmitter release from bipolar cells of the goldfish retina. Journal of Neuroscience 13, 2898-2909	[Abstract]
Tottene, A., Moretti, A. & Pietrobon, D. (1996). Functional diversity of P-type and R-type calcium channels in rat cerebellar neurons. Journal of Neuroscience 16, 6353-6363	[Abstract/Full Text]
Vilaro, M. T., Weiderhold, K. H., Palacios, J. M. & Mengod, G. (1992). Muscarinic M2 receptor mRNA expression and receptor binding in cholinergic and non-cholinergic cells in the rat brain: a correlative study using in situ hybridization histochemistry and receptor autoradiography. Neuroscience 47, 367-393	[Medline]
Wheeler, D. B., Randall, A. D. & Tsien, R. W. (1994). Roles of N-type and Q-type Ca2+ channels in supporting hippocampal synaptic transmission. Science 264, 107-111	[Medline]
Wheeler, D. B., Randall, A. D. & Tsien, R. W. (1996). Changes in action potential duration alter reliance of excitatory synaptic transmision on multiple types of Ca2+ channels in rat hippocampus. Journal of Neuroscience 16, 2226-2237	[Abstract]
Williams, S., Serafin, M., Muhlethaler, M. & Bernheim, L. (1997a). Facilitation of N-type calcium current is dependent on the frequency of action potential-like depolarizations in dissociated cholinergic basal forebrain neurons of the guinea-pig. Journal of Neuroscience 17, 1625-1632	[Abstract/Full Text]
Williams, S., Serafin, M., Muhlethaler, M. & Bernheim, L. (1997b). Distinct contributions of high- and low-voltage-activated calcium currents to afterhyperpolarizations in cholinergic nucleus basalis neurons of the guinea-pig. Journal of Neuroscience 17, 7307-7315	[Abstract/Full Text]
Witcher, D. R., DeWaard, M. & Campbell, K. P. (1993). Characterization of the purified N-type Ca2+ channel and the cation sensitivity of -conotoxin GVIA binding. Neuropharmacology 32, 1127-1139	[Medline]
Wu, L. G., Borst, J. G. G. & Sakmann, B. (1998). R-type Ca2+ currents evoke transmitter release in a rat central synapse. Proceedings of the National Academy of Sciences of the USA 95, 4720-4725	[Abstract/Full Text]
Yawo, H., Chuhma, N. & Endo, K. (1994). Modulation of presynaptic calcium channels by neurotransmitters. Biomedical Research S1, 9-16.
Yu, B. & Shinnick-Gallagher, P. (1997). Dihydropyridine- and neurotoxin-sensitive and -insensitive calcium currents in acutely dissociated neurons of the rat central amygdala. Journal of Neurophysiology 77, 690-701	[Abstract/Full Text]

Acknowledgements

I would like to thank Professor D. A. Brown for providing laboratory facilities and also for his helpful comments and criticisms during the preparation of this manuscript. My thanks also go to Mrs B. Browning and Mrs M. Hall for preparing the myoball cultures used in these experiments. This work is supported by a grant from the Medical Research Council (UK).

Correspondence

T. G. J. Allen: Department of Pharmacology, University College London, Gower Street, London WC1E 6BT, UK.

Email: tallen{at}ucl.ac.uk

This article has been cited by other articles:

				T. G.J. Allen, F. C. Abogadie, and D. A. Brown Simultaneous Release of Glutamate and Acetylcholine from Single Magnocellular "Cholinergic" Basal Forebrain Neurons J. Neurosci., February 1, 2006; 26(5): 1588 - 1595. [Abstract] [Full Text] [PDF]

				K. Melliti, U. Meza, and B. Adams Muscarinic Stimulation of alpha 1E Ca Channels Is Selectively Blocked by the Effector Antagonist Function of RGS2 and Phospholipase C-beta 1 J. Neurosci., October 1, 2000; 20(19): 7167 - 7173. [Abstract] [Full Text] [PDF]

This Article Abstract Full Text (PDF) Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Allen, T. G. J. Search for Related Content PubMed PubMed Citation Articles by Allen, T. G. J.