Metabotropic glutamate receptors (mGluRs) are a type of G-protein coupled receptor that modulate protein expression, neurotransmission and neuronal excitability throughout the CNS [1-4]. mGluR1 is a member of a Group I mGluR that is expressed in cerebellar Purkinje cells and in many other postsynaptic neurons. Activation of mGluR1 induces several cellular events in cerebellar Purkinje cells like the activation of PLC-β-IP₃ signal cascade [5,6] or slow excitatory postsynaptic currents (EPSCs) [7,8]. Slow currents are known to be mediated by transient receptor potential canonical (TRPC) channels [8]. Slow currents are non-selective cation currents which include Na⁺ and Ca²⁺ influxes [9] so it may elevate intracellular Ca²⁺ concentration in cerebellar Purkinje cells. The function of such slow currents is still unknown, however.

Delta-9-tetrahydrocannabinol (Δ9-THC), the principal psychoactive constituent of marijuana affects brain function by activating the G-protein coupled receptor, cannabinoid receptor-1 (CB1R) [10,11], which is expressed throughout the brain [12]. Two major endogenous ligands of the receptor, 2-arachidonylglycereol (2-AG) and anandamide are known so far [13,14]. Endocannabinoids function as a retrograde messenger that contributes to long-term and short-term plasticity in the CNS [15-17]. It is reported that strong depolarization of postsynaptic neurons [18,19] or administration of mGluR agonists [20,21] can suppress presynaptic transmitter release and diminish EPSCs. It is found that the depression of EPSCs has an endocannabinoid-dependent mechanism and three different models for the mechanism of endocannabinoid release have been suggested. Endocannabinoids are released by large elevation of calcium concentration in the postsynaptic neuron that is usually induced by strong depolarization (depolarization-induced suppression of inhibition/excitation, DSI/DSE), strong activation of G protein-coupled receptors (receptor-driven endocannabinoid release, RER), or simultaneous small elevations of calcium concentration and weak receptor activation (Ca²⁺-assisted RER, Ca-RER) and acts presynaptically to suppress neurotransmitter release. Unlike DSE/DSI, Ca-RER can be induced by weak elevation of calcium, which can be mediated by calcium sources other than the voltage-operated calcium channel (VOCC) like the Endoplasmic reticulum (ER) calcium store or TRPC. It was reported that ER calcium release do not have a role in the induction of Ca-RER [22], indicating it is more worthwhile to investigate the role of TRPC in endocannabinoid release. Here, we investigated whether endocannabinoid-induced retrograde signaling is regulated by the TRPC-induced slow current. We induced transient depression of synaptic transmission by PF burst stimulation in current clamp mode and associative stimulation of parallel fiber (PF) burst and Purkinje cell (PC) depolarization in voltage clamp mode; however, neither was affected by TRPC blockers. This suggests TRPC mediated slow currents and calcium transients do not play an important role in the endocannabinoid signaling and that calcium sources other than TRPC is needed for Ca-RER.

Parasagittal brain slices of the cerebellar vermis (250 µm thick) were prepared from P15-P20 Sprague-Dawley rats using a vibrating tissue slicer and ice-cold standard artificial cerebrospinal fluid (ACSF) containing 124 mM NaCl, 2.5 mM KCl, 1 mM Na₂HPO₄, 1.3 mM MgCl₂, 2.5 mM CaCl₂, 26.2 mM NaHCO₃, and 20 mM D-glucose, bubbled with 95% O₂ and 5% CO₂. After cutting, slices were kept for 30 min at 35℃ and then for up to 8 hrs. at 25℃ in ACSF.

After a recovery period, the slices were placed in a submerged chamber that was perfused at a rate of 2 ml/min with ACSF supplemented with either 100 µM picrotoxin to block GABA_A receptors. Somatic whole-cell recording were obtained by using an EPC8 amplifier (HEKA Instruments Inc., Bellmore, NY). The recording electrodes (resistance 2~4 MΩ) were filled with a solution containing 130 mM K-gluconate, 10 mM KCl, 10 mM HEPES, 0.2 EGTA, 10 mM Na₂-phosphocreatine, 3 mM Mg₂ATP, 0.3 mM Na₃GTP (pH 7.25) for current-clamp recording and 135 mM Cs-methanesulfonate, 10 mM CsCl, 10 mM HEPES, 4 mM Mg₂ATP, 0.4 mM Na₃GTP, and 0.2 mM EGTA (pH 7.25) for voltage-clamp recordings. Currents were filtered at 1 kHz, digitized at 5 kHz, and acquired using Pulse (HEKA Instruments). For parallel fiber stimulation, standard patch pipettes were used that were filled with external saline and placed in the middle third of the molecular layer. Synaptic responses were evoked every 30 s using 20~40 µA pulses (100 µs duration). When burst stimulation was employed, the interpulse interval was 10 ms. Recordings were performed at RT in ACSF.

All group data are shown as mean±SEM. Comparisons were made using Student's t tests. All drugs were purchased from Sigma (St. Louis, MO) except for CPCCOEt, NBQX, AM251, SKF96365 (Tocris Cookson, Ballwin, MO).

Several studies demonstrated that brief burst of parallel fiber can trigger retrograde inhibition of the PF-PC synapse [16,22]. This phenomenon is termed as Ca-RER and is mGluR-PLCβ4 dependent. To mimic this Ca-RER, we made whole-cell patch clamp recordings from cerebellar Purkinje cells in cerebellar slices and recorded parallel fiber-Purkinje cell (PF-PC) excitatory postsynaptic potentials (EPSPs) every 2 s, then applied short burst PF stimulation (100 Hz×10) (Fig. 1A). Subsequent EPSCs were depressed transiently (Fig. 1B, C, D; 52.05±10.57%, n=5), and this short-term depression was inhibited by 3 µM of cannabinoid receptor 1 blocker AM251 (Fig. 1B, C, D; -23.69±15.20%, n=5). Furthermore, in another set of experiments, short-term depression (Fig. 2; 49.73±25.25%, n=5) was also inhibited by 100 µM of CPCCOEt, an mGluR inhibitor (Fig. 2; -12.37±17.65%, n=5). This data shows that the depression induced by burst stimulation of PF is CB1R and mGluR1-dependent RER.

TRPC is known to mediate slow inward currents activated by mGluR1 in cerebellar Purkinje cells [8,23]. TRPC is a calcium-permeable non-selective cation channel, thus it is possible that TRPC functions as a calcium source in calcium-dependent cellular events. To investigate whether TRPC-mediated calcium influx contributes to Ca-RER, we applied SKF96365, a TRP channel inhibitor. The short-term depression (Fig. 3; 38.42±11.10%, n=5) was inhibited by 30 µM of SKF96365 (Fig. 3; -13.62±4.99%) significantly. Regehr and his coworkers reported that associative stimulation can induce endocannabinoid signaling more effectively [22]. It is suggested that PF burst can cause a massive release of glutamate which can activate mGluR1 on the perisynaptic site of Purkinje cell dendritic spines, and climbing fiber (CF) can make global depolarization of Purkinje cell that may open VOCC, increasing the intracellular calcium concentration of the Purkinje cell. To mimic this associative short-term depression and investigate the role of TRPC in the depression, we made whole-cell patch clamp recordings from cerebellar Purkinje cells in cerebellar slices and recorded PF-PC EPSCs, then applied associative stimulation consisting of short burst PF stimulation (100 Hz×10) followed by 50 ms of depolarization of PC after 50 ms (Fig. 4A). The associative protocol induced short-term depression effectively (Fig. 4B, C, D; 50.03±14.84%, n=7), and SKF96365 decreased this depression partially (Fig. 4B, C, D; 14.69±11.72%, n=7).

Although we observed the effect of a TRP channel blocker, we had to confirm that TRPC specifically contributed to short-term plasticity since SKF96365 is known for its non-specificity [24,25], especially to VOCCs. To investigate whether TRPC-mediated calcium influx contribute to RER, we applied BTP2, another TRPC inhibitor. 1 µM of BTP2 blocked most TRPC-induced currents in cultured cells in previous studies [26], however, the concentration of BTP2 to block TRPC effectively in slices has not been reported. To find out the optimum concentration of BTP2 to block slow currents in brain slices, we induced TRPC-mediated slow current with parallel fiber burst (100 Hz×10 times) while applying NBQX (2.5 µM) to block the overlaying AMPAR current (Fig. 5A), and examined the concentration-dependent effect of BTP2 on the amplitude of the slow current. 50 µM of BTP2 blocked the slow current in slice, while 25 µM or lower concentrations of BTP2 could not (Fig. 5B). Thus, we investigated whether 50 µM of BTP2 changed the amplitude of transient depression of EPSC in cerebellar slices. In cells which showed the depression by 10 times of parallel fiber burst on current clamp mode, 50 µM of BTP2 did not affect the amplitude of transient depression (Fig. 5C, D), although it was enough to block TRPC-mediated slow current almost completely. Furthermore, BTP2 did not affect short-term depression induced by the associative protocol in the voltage clamp mode (Fig. 5E, F). This data indicates that TRPC-mediated slow current does not contribute to RER by the associative protocol we used and that the calcium influx through TRPC is overwhelmed by the calcium influx through VOCC or calcium release from intracellular calcium stores.

SKF96365 has been reported to elevate intracellular calcium ion levels [27] and inhibit VOCCs [24,28] - especially T-type calcium channels in cerebellar Purkinje cells [25], potassium channels [29,30], cytochrome P450 [31,32], and calcium ATPase [28,33]. SKF96365 blocked the endocannabinoid-induced depression effectively; we assumed the SKF96365 inhibited endocannabinoid release by reducing postsynaptic calcium increase through TRPC. However, BTP2, another blocker of TRP channel did not block RER, even though same concentration of BTP2 could block the most of the slow current. This means the action of SKF96365 was a nonspecific effect, although the mechanism by which SKF96365 inhibited RER is unknown. Drugs like SKF96365 or ruthenium red have been used as TRP channel blockers so far, but now they are known to have nonspecific effects on Endoplasmic reticulum (ER) calcium stores or channels other than TRPC. BTP2 has been known to function as a store-operated Ca²⁺ channel/Ca²⁺ release-activated Ca²⁺ channel blocker [34,35], and inhibit overexpressed TRPC channels in HEK293 cells. BTP2 almost fully blocks TRPC at about 1 µM, and had little effect on ER, K⁺ channel or VOCC in cultured single cell. However, a concentration of BTP2 that blocks TRPC effectively in brain slices has not been reported yet. We found that 50 µM of BTP2 is needed to block the slow current; it is much larger than the concentration that blocks TRPC in cultured single cell, and specificity of BTP2 to TRPC at this concentration is not clear yet. However, based on the fact that BTP2 did not affect endocannabinoid release, we conclude that TRPC does not contribute to endocannabinoid release although 50 µM of BTP2 may be nonspecific. The results described above suggest that BTP2 can be an alternative or additional option as a pharmacologic blocker of TRP channels.

The role of VOCC or ER calcium stores as a calcium source is well known, while that of calcium influx through TRPC is unclear. We investigated a possibility that TRPC plays a role as a novel calcium source other than VOCC or ER calcium store in Ca-RER, however, TRPC did not affect endocannabinoid release. It was also revealed previously that ER calcium store has little effect on the elevation of calcium concentration and endocannabinoid release [22]. Based on these findings, there is a large possibility that VOCC acts as the critical and sole calcium source. We could not induce transient depression by only PF burst on voltage clamp mode at -70 mV, unlike associative stimulation or PF burst on current clamp mode, which VOCC can activate by; this data is consistent with the idea that VOCC is the most important calcium source for endocannabinoid release.

Ca-RER can be induced by 1 µM or higher concentration of calcium [22]. Therefore, if calcium increase induced by VOCC was more than 1 µM and hence sufficient to mediate endocannabinoid release, TRPC-mediated calcium increase could be overwhelmed by the activation of VOCC. Thus, a possibility that TRPC can contribute to endocannabinoid release in the condition that VOCC activates to a lesser degree still remains. However, the amplitude of endocannabinoid-induced depression increases in a calcium dependent manner to almost 80% of control, up to calcium concentrations of 5 µM [22]. The amplitude of attenuation of EPSC we observed was about 50%, which is not maximal degree of endocannabinoid-mediated depression. If TRPC contributes to endocannabinoid release, BTP2 should inhibit the depression of EPSC at least partially in spite of VOCC. Therefore, our data suggest TRPC is not involved in endocannabinoid release.