Combinatorial synthesis of omega-conotoxin MVIIC analogues and their binding with N- and P/Q-type calcium channels

Omega-conotoxin MVIIC (MVIIC) blocks P/Q-type calcium channels with high affinity and N-type calcium channels with low affinity, while the highly homologous omega-conotoxin MVIIA blocks only N-type calcium channels. We wished to obtain MVIIC analogues more selective for P/Q-type calcium channels than MVIIC to elucidate structural differences among the channels, which discriminate the omega-conotoxins. To prepare a number of MVIIC analogues efficiently, we developed a combinatorial method which includes a random air oxidation step. Forty-seven analogues were prepared in six runs and some of them exhibited higher selectivity for P/Q-type calcium channels than MVIIC in binding assays.     

Effects of the blockade of high voltage-activated calcium channels on in vitro pineal melatonin synthesis

The presence of high voltage-activated calcium channels in the rat pineal gland is well known. However, their role in pineal metabolism is not completely understood and is even controversial. Better to understand this matter, we investigated the effects of L-, N- or P/Q-type calcium channel blockers (nifedipine, omega-conotoxin GVIA, omega-agatoxin IVA, respectively) on melatonin content and arylalkylamine-N-acetyltransferase activity of denervated rat pineal glands kept for 48 h in culture and stimulated with norepinephrine. Melatonin was measured by high performance liquid chromatography with electrochemical detection and arylalkylamine-N-acetyltransferase activity was quantified by radiometric assay. Pre-incubation with any of these high voltage-activated calcium channel blockers reduced the melatonin production induced by norepinephrine although arylalkylamine-N-acetyltransferase activity was reduced only by the N-type calcium channel antagonist, omega-conotoxin GVIA. The results indicate that calcium influx through L-, N- or P/Q-type of high voltage-activated calcium channels is necessary for the full expression of the metabolic process leading to melatonin synthesis in the rat pineal glands. However, the mechanisms involved in this process are different for the L- or P/Q- and N-type calcium channels.     

Binding of six chimeric analogs of omega-conotoxin MVIIA and MVIIC to N- and P/Q-type calcium channels

Replacement of the N-terminal half of omega-conotoxin MVIIC, a peptide blocker of P/Q-type calcium channels, with that of omega-conotoxin MVIIA significantly increased the affinity for N-type calcium channels. To identify the residues essential for subtype selectivity, we examined single reverse mutations from MVIIA-type to MVIIC-type in this chimeric analog. A reverse mutation from Lys(7) to Pro(7) decreased the affinity for both P/Q- and N-type channels, whereas that from Leu(11) to Thr(11) increased the affinity for P/Q-type channels and decreased the affinity for N-type channels. The roles of these two residues were confirmed by synthesizing two MVIIC analogs in which Pro(7) and Thr(11) were replaced with Lys(7) and Leu(11), respectively.     

Co-expression of metabotropic glutamate receptor 7 and N-type Ca(2+) channels in single cerebrocortical nerve terminals of adult rats

The modulation of calcium channels by metabotropic glutamate receptors (mGluRs) is a key event in the fine-tuning of neurotransmitter release. Here we report that, in cerebrocortical nerve terminals of adult rats, the inhibition of glutamate release is mediated by mGluR7. In this preparation, the major component of glutamate release is supported by P/Q-type Ca2+ channels (72.7%). However, mGluR7 selectively reduced the release component that is associated with N-type Ca2+ channels (29.9%). Inhibition of P/Q channels by mGluR7 is not masked by the higher efficiency of these channels in driving glutamate release when compared with N-type channels. Thus, activation of mGluR7 failed to reduce the release associated with P/Q channels when the extracellular calcium concentration, ([Ca2+]o), was reduced from 1.3 to 0.5 mm. Through Ca2+ imaging, we show that Ca2+ channels are distributed in a heterogeneous manner in individual nerve terminals. Indeed, in this preparation, nerve terminals were observed that contain N-type (31.1%; conotoxin GVIA-sensitive) or P/Q-type (64.3%; agatoxin IVA-sensitive) channels or that were insensitive to these two toxins (4.6%). Interestingly, the great majority of the responses to l-AP4 (95.4%) were observed in nerve terminals containing N-type channels. This specific co-localization of mGluR7 and N-type Ca2+-channels could explain the failure of the receptor to inhibit the P/Q channel-associated release component and also reveal the existence of specific targeting mechanisms to localize the two proteins in the same nerve terminal subset.     

Calcium channels involved in neurotransmitter release at adult,neonatal and P/Q-type deficient neuromuscular junctions (Review)

Different types of voltage-dependent calcium channels (VDCCs) have been recognized based on their molecular structure as well as their pharmacological and biophysical properties. One of these, the P/Q type, is the main channel involved in nerve evoked neurotransmitter release at neuromuscular junctions (NMJs) and many central nervous system synapses. However, under particular experimental or biological conditions, other channels can be involved. L-type VDCC presence at the NMJ has been demonstrated by the contribution to the perineural calcium currents (I Ca ) at adult mice Bapta-loaded NMJs. This is probably a result of a reduction in Ca 2+ inactivation. The L-type current was not coupled to neurotransmitter release, but became coupled, as demonstrated by the release of acetylcholine, after the inhibition of serine/threonine protein phosphatases with okadaic acid (OA). Thus, under these conditions, L-type channels were unmasked at Bapta- but not at Egta-loaded NMJs. This suggests that the speed, not the capacity, of the calcium chelator was decisive in preventing Ca 2+ -inactivation and facilitating the contribution to neurotransmitter release. At neonatal rat NMJs, N-type VDCCs were involved early during development whereas P/Q-type VDCCs play a main role at all stages of development. Furthermore, P/Q-type VDCCs were more efficiently coupled to neurotransmitter release than N-type VDCCs. This difference could be accounted for by a differential location of these channels at the release site. Neuromuscular transmission in P/Q-type calcium channel knock out ataxic mice jointly depends on both N-type and R-type channels and shows several altered properties including low quantal content. Thus, calcium channels may be recruited to mediate neurotransmitter release with a functional hierarchy where the P/Q channel seems to be the channel most suited to mediate exocytosis at NMJs.     

Calcium channels involved in neurotransmitter release at adult,neonatal and P/Q-type deficient neuromuscular junctions (Review)

Different types of voltage-dependent calcium channels (VDCCs) have been recognized based on their molecular structure as well as their pharmacological and biophysical properties. One of these, the P/Q type, is the main channel involved in nerve evoked neurotransmitter release at neuromuscular junctions (NMJs) and many central nervous system synapses. However, under particular experimental or biological conditions, other channels can be involved. L-type VDCC presence at the NMJ has been demonstrated by the contribution to the perineural calcium currents (Ica) at adult mice Bapta-loaded NMJs. This is probably a result of a reduction in Ca(2+) inactivation. The L-type current was not coupled to neurotransmitter release, but became coupled, as demonstrated by the release of acetylcholine, after the inhibition of serine/threonine protein phosphatases with okadaic acid (OA). Thus, under these conditions, L-type channels were unmasked at Bapta- but not at Egta-loaded NMJs. This suggests that the speed, not the capacity, of the calcium chelator was decisive in preventing Ca(2+)-inactivation and facilitating the contribution to neurotransmitter release. At neonatal rat NMJs, N-type VDCCs were involved early during development whereas P/Q-type VDCCs play a main role at all stages of development. Furthermore, P/Q-type VDCCs were more efficiently coupled to neurotransmitter release than N-type VDCCs. This difference could be accounted for by a differential location of these channels at the release site. Neuromuscular transmission in P/Q-type calcium channel knock out ataxic mice jointly depends on both N-type and R-type channels and shows several altered properties including low quantal content. Thus, calcium channels may be recruited to mediate neurotransmitter release with a functional hierarchy where the P/Q channel seems to be the channel most suited to mediate exocytosis at NMJs.     

Voltage-dependent P/Q-type calcium channels at the frog neuromuscular junction

It is well known that antagonists of N-type voltage-gated calcium channels inhibit the evoked quantal release of acetylcholine in amphibian neuromuscular synapses. This, however, does not exclude the functional expression of other types of voltage-gated calcium channels in these nerve terminals. Using immunocytochemistry, we detected the expression of the alpha1A subunit of P/Q-type calcium channels (that is otherwise typical of mammalian motor nerve endings) in the frog neuromuscular junction. In addition, we demonstrated that the P/Q-type channel blocker omega-agatoxin IVA (20 nM) reduced the action potential-induced calcium transient and significantly decreased both spontaneous and evoked mediator release. Our data indicates the functional expression of P/Q-type calcium channels in the frog motor nerve ending which participate in acetylcholine release.     

Custom distinctions in the interaction of G-protein beta subunits with N-type (CaV2.2) versus P/Q-type (CaV2.1) calcium channels

Inhibition of N- (Cav2.2) and P/Q-type (Cav2.1) calcium channels by G-proteins contribute importantly to presynaptic inhibition as well as to the effects of opiates and cannabinoids. Accordingly, elucidating the molecular mechanisms underlying G-protein inhibition of voltage-gated calcium channels has been a major research focus. So far, inhibition is thought to result from the interaction of multiple proposed sites with the Gbetagamma complex (Gbetagamma). Far less is known about the important interaction sites on Gbetagamma itself. Here, we developed a novel electrophysiological paradigm, "compound-state willing-reluctant analysis," to describe Gbetagamma interaction with N- and P/Q-type channels, and to provide a sensitive and efficient screen for changes in modulatory behavior over a broad range of potentials. The analysis confirmed that the apparent (un)binding kinetics of Gbetagamma with N-type are twofold slower than with P/Q-type at the voltage extremes, and emphasized that the kinetic discrepancy increases up to ten-fold in the mid-voltage range. To further investigate apparent differences in modulatory behavior, we screened both channels for the effects of single point alanine mutations within four regions of Gbeta1, at residues known to interact with Galpha. These residues might thereby be expected to interact with channel effectors. Of eight mutations studied, six affected G-protein modulation of both N- and P/Q-type channels to varying degrees, and one had no appreciable effect on either channel. The remaining mutation was remarkable for selective attenuation of effects on P/Q-, but not N-type channels. Surprisingly, this mutation decreased the (un)binding rates without affecting its overall affinity. The latter mutation suggests that the binding surface on Gbetagamma for N- and P/Q-type channels are different. Also, the manner in which this last mutation affected P/Q-type channels suggests that some residues may be important for "steering" or guiding the protein into the binding pocket, whereas others are important for simply binding to the channel.     

Voltage-Operated Calcium Channel Heterogeneity in Pancreatic β Cells: Physiopathological Implications

Voltage-operated calcium channels play crucial roles in stimulus-secretion coupling in pancreatic beta cells. A growing body of evidence indicates that these channels in beta cells are heterogeneous. In particular, not all the high-threshold calcium channels expressed belong to the best known L-type. In rat insulinoma cells, for example, L, N, and P/Q-type channels are present, while in human beta cells L-type and P/Q-type dominate. Where present, N-type and P/Q-type channels participate, alongside with the dominant L-type, in the control of sugar- or depolarization-induced hormone release. Distinct biophysical properties and selective modulation of the channel subtypes are likely to play important physiological roles. T-type channels are involved in beta cell apoptosis, while calcium channel autoantibodies recognizing high-threshold channels in beta cells, have been described both in neurological and diabetic patients. Subtype-selective calcium channel drugs have the potential for being beneficial in beta cell pathological states.     

Interactions between proteins implicated in exocytosis and voltage-gated calcium channels

Neurotransmitter release from synaptic vesicles is triggered by voltage-gated calcium influx through P/Q-type or N-type calcium channels. Purification of N-type channels from rat brain synaptosomes initially suggested molecular interactions between calcium channels and two key proteins implicated in exocytosis: synaptotagmin I and syntaxin 1. Co-immunoprecipitation experiments were consistent with the hypothesis that both N- and P/Q-type calcium channels, but not L-type channels, are associated with the 7S complex containing syntaxin 1, SNAP-25, VAMP and synaptotagmin I or II. Immunofluorescence confocal microscopy at the frog neuromuscular junction confirmed that calcium channels, syntaxin 1 and SNAP-25 are co-localized at active zones of the presynaptic plasma membrane where transmitter release occurs. Experiments with recombinant proteins were performed to map synaptic protein interaction sites on the alpha 1A subunit, which forms the pore of the P/Q-type calcium channel. In vitro-translated 35S-synaptotagmin I bound to a site located on the cytoplasmic loop linking homologous domains II and III of the alpha 1A subunit. This direct link would target synaptotagmin, a putative calcium sensor for exocytosis, to a microdomain of calcium influx close to the channel mouth. Cysteine string proteins (CSPs) contain a J-domain characteristic of molecular chaperones that cooperate with Hsp70. They are located on synaptic vesicles and thought to be involved in modulating the activity of presynaptic calcium channels. CSPs were found to bind to the same domain of the calcium channel as synaptotagmin, and also to associate with VAMP. CSPs may act as molecular chaperones in association with Hsp70 to direct assembly or dissociation of multiprotein complexes at the calcium channel.     

Physical Link and Functional Coupling of Presynaptic Calcium Channels and the Synaptic Vesicle Docking/Fusion Machinery

N- and P/Q-type calcium channels are localized in high density in presynaptic nerve terminals and are crucial elements in neuronal excitation–secretion coupling. In addition to mediating Ca²⁺ entry to initiate transmitter release, they are thought to interact directly with proteins of the synaptic vesicle docking/fusion machinery. As outlined in the preceding article, these calcium channels can be purified from brain as a complex with SNARE proteins which are involved in exocytosis. In addition, N-type and P/Q-type calcium channels are co-localized with syntaxin in high-density clusters in nerve terminals. Here we review the role of the synaptic protein interaction (synprint) sites in the intracellular loop II–III (L_II–III) of both _1B and _1A subunits of N-type and P/Q-type calcium channels, which bind to syntaxin, SNAP-25, and synaptotagmin. Calcium has a biphasic effect on the interactions of N-type calcium channels with SNARE complexes, stimulating optimal binding in the range of 10–20 M. PKC or CaM KII phosphorylation of the N-type synprint peptide inhibits interactions with native brain SNARE complexes containing syntaxin and SNAP-25. Introduction of the synprint peptides into presynaptic superior cervical ganglion neurons reversibly inhibits EPSPs from synchronous transmitter release by 42%. At physiological Ca²⁺ concentrations, synprint peptides cause an approximate 25% reduction in transmitter release of injected frog neuromuscular junction in cultures, consistent with detachment of 70% of the docked vesicles from calcium channels based on a theoretical model. Together, these studies suggest that presynaptic calcium channels not only provide the calcium signal required by the exocytotic machinery, but also contain structural elements that are integral to vesicle docking, priming, and fusion processes.     

Role of Thr(11) in the binding of omega-conotoxin MVIIC to N-type Ca2+ channels

As replacement of Thr(11) of omega-conotoxin MVIIC with Ala significantly reduced the affinity for both N- and P/Q-type calcium channels, we examined the effect of substitution at this position with other residues. Binding assays using rat cerebellar P2 membranes showed that the affinity is in the order of Leu>Val, aminobutyric acid, Thr>Asn&z.Gt;Ser, Ala, Asp, Phe, Tyr for N-type channels and Thr>Leu, Val, aminobutyric acid, Asn, Ser>Ala&z.Gt;Asp, Phe, Tyr for P/Q-type channels, suggesting that aliphatic amino acids with longer side chains are favorable for block of N-type channels. The effects of substitution were examined electrophysiologically in BHK cells expressing N-type Ca2+ channels. Inhibition of Ba2+ current by the analogs did not completely correlate with binding affinity, although binding to BHK cells was comparable to rat cerebellar membranes.     

Modulating modulation: crosstalk between regulatory pathways of presynaptic calcium channels

The activity of some voltage-gated calcium channels (VGCCs) can be inhibited by specific G protein beta subunits. Conversely, in the case of N-type VGCCs, protein kinase C can relieve Gbeta-dependent inhibition by phosphorylating at least one specific site on the calcium channel. A recent publication describes a newly identified method of intracellular regulation of specific VGCCs. Wu et al. have uncovered that VGCC activity can be regulated by phosphatidylinositol-4',5'-bisphosphate (PIP2). Whereas PIP2 is important for maintaining the activity (open state) of Cav2.1 (N-type) and Cav2.2 (P/Q-type) channels, the enzymatic breakdown of PIP2 leads to the inactivation of these channels. Additionally, PIP2 can cause changes in voltage-dependent activation of Cav2.2 (P/Q-type) channels that make it more difficult for these channels to open (from the closed state). Furthermore, protein kinase A activity can circumvent PIP2-mediated inhibition. Thus, the PIP2-mediated regulation of VGCCs is tightly controlled by the functions of kinases (and phosphatases), as well as phospholipases. Wu et al. stress that because PIP2 can be found at synapses, PIP2-dependent control of VGCCs "could have profound consequences on synaptic transmission and plasticity."     

Synthesis and biological evaluation of nonpeptide mimetics of omega-conotoxin GVIA

A benzothiazole-derived compound (4a) designed to mimic the C(alpha)-C(beta) bond vectors and terminal functionalities of Lys2, Tyr13 and Arg17 in omega-conotoxin GVIA was synthesised, together with analogues (4b-d), which had each side-chain mimic systematically truncated or eliminated. The affinity of these compounds for rat brain N-type and P/Q-type voltage gated calcium channels (VGCCs) was determined. In terms of N-type channel affinity and selectivity, two of these compounds (4a and 4d) were found to be highly promising, first generation mimetics of omega-conotoxin. The fully functionalised mimetic (4a) showed low microM binding affinity to N-type VGCCs (IC(50)=1.9 microM) and greater than 20-fold selectivity for this channel sub-type over P/Q-type VGCCs, whereas the mimetic in which the guanidine-type side chain was truncated back to an amine (4d, IC(50)= 4.1 microM) showed a greater than 25-fold selectivity for the N-type channel.     

Interactions between Presynaptic Calcium Channels and Proteins Implicated in Synaptic Vesicle Trafficking and Exocytosis

Monoclonal antibodies were generated by immunizing mice with chick brain synaptic membranes and screening for immunoprecipitation of solubilized conotoxin GVIA receptors (N-type calcium channels). Antibodies against two synaptic proteins (p35--syntaxin 1 and p58--synaptotagmin) were produced and used to purify and characterize a ternary complex containing N-type channels associated with these two proteins. These results provided the first evidence for a specific interaction between presynaptic calcium channels and SNARE proteins involved in synaptic vesicle docking and calcium-dependent exocytosis. Immunoprecipitation experiments supported the conclusion that syntaxin 1/SNAP-25/VAMP/synaptotagmin I or II complexes associate with N-type, P/Q-type, but not L-type calcium channels from rat brain nerve terminals. Immunofluorescent confocal microscopy at the frog neuromuscular junction was consistent with the co-localization of syntaxin 1, SNAP-25, and calcium channels, all of which are predominantly expressed at active zones of the presynaptic plasma membrane facing post-synaptic folds rich in acetylcholine receptors. The interaction of proteins implicated in calcium-dependent exocytosis with presynaptic calcium channels may locate the sensor(s) that trigger vesicle fusion within a microdomain of calcium entry.     

Role of high-voltage-activated calcium channels in glucose-regulated beta-cell calcium homeostasis and insulin release

High-voltage-activated (HVA) calcium channels are known to be the primary source of calcium for glucose-stimulated insulin secretion. However, few studies have investigated how these channels can be regulated by chronically elevated levels of glucose. In the present study, we determined the level of expression of the four major HVA calcium channels (N-type, P/Q-type, L(C)-type, and L(D)-type) in rat pancreatic beta-cells. Using quantitative real-time PCR (QRT-PCR), we found the expression of all four HVA genes in rat insulinoma cells (INS-1) and in primary isolated rat islet cells. We then determined the role of each channel in insulin secretion by using channel-selective antagonists. Insulin secretion analysis revealed that N- and L-type channels are both involved in immediate glucose-induced insulin secretion. However, L-type was preferentially coupled to secretion at later time points. P/Q-type channels were not found to play a role in insulin secretion at any stage. It was also found that long-term exposure to elevated glucose increases basal calcium in these cells. Interestingly, chronically elevated glucose decreased the mRNA expression of the channels involved with insulin secretion and diminished the level of stimulated calcium influx in these cells. Using whole cell patch clamp, we found that N- and L-type channel currents increase gradually subsequent to lower intracellular calcium perfusion, suggesting that these channels may be regulated by glucose-induced changes in calcium.     

Unified mechanisms of Ca2+ regulation across the Ca2+ channel family

L-type (CaV1.2) and P/Q-type (CaV2.1) calcium channels possess lobe-specific CaM regulation, where Ca2+ binding to one or the other lobe of CaM triggers regulation, even with inverted polarity of modulation between channels. Other major members of the CaV1-2 channel family, R-type (CaV2.3) and N-type (CaV2.2), have appeared to lack such CaM regulation. We report here that R- and N-type channels undergo Ca(2+)-dependent inactivation, which is mediated by the CaM N-terminal lobe and present only with mild Ca2+ buffering (0.5 mM EGTA) characteristic of many neurons. These features, together with the CaM regulatory profiles of L- and P/Q-type channels, are consistent with a simplifying principle for CaM signal detection in CaV1-2 channels-independent of channel context, the N- and C-terminal lobes of CaM appear invariably specialized for decoding local versus global Ca2+ activity, respectively.     

Signaling Complexes of Voltage-Gated Calcium Channels and G Protein-Coupled Receptors

Activation of opioid or opioid-receptor-like (ORL1 a.k.a. NOP or orphanin FQ) receptors mediates analgesia through inhibition of N-type calcium channels in dorsal root ganglion (DRG) neurons (). Unlike the three types of classical μ, δ, and κ opioid receptors, ORL1 mediates an agonist-independent inhibition of N-type calcium channels. This is mediated via the formation of a physical protein complex between the receptor and the channel, which in turn allows the channel to effectively sense a low level of constitutive receptor activity (). Further inhibition of N-type channel activity by activation of other G protein-coupled receptors is thus precluded. ORL1 receptors, however, also undergo agonist-induced internalization into lysosomes, and channels thereby become cointernalized in a complex with ORL1. This then results in removal of N-type channels from the plasma membrane and reduced calcium entry (). Similar signaling complexes between N-type channels and GABA_B receptors have been reported (). Moreover, both L-type and P/Q-type channels appear to be able to associate with certain types of G protein-coupled receptors (). Hence, interactions between receptors and voltage-gated calcium channels may be a widely applicable means to optimize receptor channel coupling.     

Signaling complexes of voltage-gated calcium channels and G protein-coupled receptors

Activation of opioid or opioid-receptor-like (ORL1 a.k.a. NOP or orphanin FQ) receptors mediates analgesia through inhibition of N-type calcium channels in dorsal root ganglion (DRG) neurons (1, 2). Unlike the three types of classical mu, delta, and kappa opioid receptors, ORL1 mediates an agonist-independent inhibition of N-type calcium channels. This is mediated via the formation of a physical protein complex between the receptor and the channel, which in turn allows the channel to effectively sense a low level of constitutive receptor activity (3). Further inhibition of N-type channel activity by activation of other G protein-coupled receptors is thus precluded. ORL1 receptors, however, also undergo agonist-induced internalization into lysosomes, and channels thereby become cointernalized in a complex with ORL1. This then results in removal of N-type channels from the plasma membrane and reduced calcium entry (4). Similar signaling complexes between N-type channels and GABA(B) receptors have been reported (5). Moreover, both L-type and P/Q-type channels appear to be able to associate with certain types of G protein-coupled receptors (6, 7). Hence, interactions between receptors and voltage-gated calcium channels may be a widely applicable means to optimize receptor channel coupling.     

Subtype-specific expression of group III metabotropic glutamate receptors and Ca2+ channels in single nerve terminals

The release properties of glutamatergic nerve terminals are influenced by a number of factors, including the subtype of voltage-dependent calcium channel and the presence of presynaptic autoreceptors. Group III metabotropic glutamate receptors (mGluRs) mediate feedback inhibition of glutamate release by inhibiting Ca(2+) channel activity. By imaging Ca(2+) in preparations of cerebrocortical nerve terminals, we show that voltage-dependent Ca(2+) channels are distributed in a heterogeneous manner in individual nerve terminals. Presynaptic terminals contained only N-type (47.5%; conotoxin GVIA-sensitive), P/Q-type (3.9%; agatoxin IVA-sensitive), or both N- and P/Q-type (42.6%) Ca(2+) channels, although the remainder of the terminals (6.1%) were insensitive to these two toxins. In this preparation, two mGluRs with high and low affinity for l(+)-2-amino-4-phosphonobutyrate were identified by immunocytochemistry as mGluR4 and mGluR7, respectively. These receptors were responsible for 22.2 and 24.1% reduction of glutamate release, and they reduced the Ca(2+) response in 24.4 and 30.3% of the nerve terminals, respectively. Interestingly, mGluR4 was largely (73.7%) located in nerve terminals expressing both N- and P/Q-type Ca(2+) channels, whereas mGluR7 was predominantly (69.9%) located in N-type Ca(2+) channel-expressing terminals. This specific coexpression of different group III mGluRs and Ca(2+) channels may endow synaptic terminals with distinct release properties and reveals the existence of a high degree of presynaptic heterogeneity.