Activation of metabotropic glutamate receptors enhances synaptic transmission at the Drosophila neuromuscular junction

Abstract

We examined the effects of activation of metabotropic glutamate receptors (mGluRs) on glutamatergic synaptic transmission at the neuromuscular junction of newly hatched Drosophila larvae. In nominally Ca²⁺-free solutions puff-application of low concentrations of glutamate evoked a transient frequency increase of miniature synaptic currents (mSCs). The mean amplitude of mSCs was unaffected, suggesting that this effect was presynaptic. Similar alterations of the mSC frequency were obtained using the mGluR agonists, (S)-4C3HPG, DCG-IV, or (1S,3S)-ACPD, but not when using agonists for ionotropic glutamate receptors, NMDA, AMPA or kainate. An mGluR antagonist, MCCG-I, blocked the effect of agonists on the mSC frequency. An adenylate cyclase activator, forskolin, and a cAMP analog, CPT-cAMP, mimicked the effects of mGluR activation. Meanwhile, an adenylate cyclase inhibitor, SQ22,536, blocked the mGluR agonist-induced effects, and in rutabaga, an adenylate-cyclase-defective mutant, the effect of the agonist was greatly reduced. In the presence of external Ca²⁺, (S)-4C3HPG decreased the failure rate and increased the mean amplitude of stimulus-evoked SCs, while MCCG-I decreased the amplitudes. We suggest that at the larval Drosophila neuromuscular junction endogenous glutamate released at the terminal potentiates synaptic transmission via a process involving cAMP.

Introduction

Metabotropic glutamate receptors (mGluRs) were found to be abundantly expressed in rat hippocampus and cerebellum, suggesting their important roles in brain function (Masu et al., 1991). To date, eight types of mGluRs (mGluR1 through mGluR8) have been reported and pharmacologically classified into three groups. Among a variety of responses resulting from activation of these mGluRs, changes in the cyclic AMP (cAMP) level are implicated when Group II receptors (including mGluR2 and mGluR3) are activated. In the majority of cases an activation of Group II mGluRs leads to a decrease of the cAMP level (Pin and Duvoisin, 1995). In some cases, however, they are positively coupled to adenylate cyclase. For example, in neonatal rat hippocampal neurons an activation of mGluRs leads to an increase of the frequency of spontaneous synaptic currents through a pathway involving cAMP and protein kinase A (Sciancalepore et al., 1995).

Recently Drosophila mGluR (DmGluRA) was cloned and expressed in a cell line, and its basic pharmacological properties have been described (Parmentier et al., 1996). Based on the sequence homology and pharmacological properties DmGluRA was classified as Group II, an activation of which leads to a modulation of synaptic transmission through the pathway involving cAMP rather than through the phospholipase C cascade. The expression of DmGluRA in the CNS was culminated at stage 14–17 which coincides with the period of the formation of the neuromuscular junction (Broadie and Bate, 1993), suggesting that DmGluRA may play a role in early synapses.

In Drosophila mutants defective in learning and memory, synaptic plasticity at the glutamatergic neuromuscular junction is altered (Zhong and Wu, 1991). Namely, in dunce larvae, in which cAMP-specific phosphodiesterase II is lacking, and consequently a high level of cAMP is demonstrated (Livingstone et al., 1984), the amplitudes of nerve-evoked synaptic currents are larger than those in the wild-type, and both synaptic facilitation during a tetanus and post-tetanic potentiation are absent, probably due to a high level of synaptic transmission at the resting state. In another mutant, rutabaga, in which Ca²⁺/calmodulin dependent adenylate cyclase is defective, both synaptic facilitation and post-tetanic potentiation are impaired. These results suggest that cAMP is an important mediator of synaptic plasticity at this synapse. However, the physiological pathways linking extracellular signals to this cAMP cascade are still unknown. In Aplysia an activation of serotonin receptors leads to an elevation of cAMP which closes K⁺ channels causing a long lasting depolarization. This depolarization in turn activates voltage-gated Ca²⁺ channels. This sequence of events is presumed to be the basis for short term synaptic plasticity (Kandel and Schwartz, 1982). In an analogy, mGluRs in Drosophila larvae may play a role in linking physiological signals to the cAMP cascade.

It has been reported that in a few preparations cAMP is working directly on the fusion machinery independent from Ca²⁺ entry to enhance synaptic transmission (Trudeau et al., 1996, Chen and Regehr, 1997, Chavis et al., 1998). Thus it is desirable to separate a direct effect of cAMP on the vesicle fusion machinery from an indirect effect via changes in Ca²⁺ influx accompanied by a membrane depolarization and subsequent activation of voltage-gated Ca²⁺ channels. For this purpose the majority of our experiments were carried out in nominally Ca²⁺-free external saline. However, to assess the role of mGluRs in more physiological conditions, we also examined the effects of mGluR activation on nerve-evoked synaptic currents in the presence of external Ca²⁺. The effects could be the compound effects through multiple pathways after activation of mGluRs, since the influx of Ca²⁺ may facilitate synaptic transmission not only through an effect on vesicle fusion but also through a facilitating effect on adenylate cyclase. Some of these results have been briefly reported elsewhere (Zhang et al., 1999).