Regulatory effects of anandamide on intracellular Ca²⁺ concentration increase in trigeminal ganglion neurons

Abstract

Activation of cannabinoid receptor type 1 on presynaptic neurons is postulated to suppress neurotransmission by decreasing Ca²⁺ influx through high voltage-gated Ca²⁺ channels. However, recent studies suggest that cannabinoids which activate cannabinoid receptor type 1 can increase neurotransmitter release by enhancing Ca²⁺ influx in vitro. The aim of the present study was to investigate the modulation of intracellular Ca²⁺ concentration by the cannabinoid receptor type 1 agonist anandamide, and its underlying mechanisms. Using whole cell voltage-clamp and calcium imaging in cultured trigeminal ganglion neurons, we found that anandamide directly caused Ca²⁺ influx in a dose-dependent manner, which then triggered an increase of intracellular Ca²⁺ concentration. The cyclic adenosine and guanosine monophosphate-dependent protein kinase systems, but not the protein kinase C system, were involved in the increased intracellular Ca²⁺ concentration by anandamide. This result showed that anandamide increased intracellular Ca²⁺ concentration and inhibited high voltage-gated Ca²⁺ channels through different signal transduction pathways.

Introduction

N-arachidonoylethanolamine (anandamide) is referred to as the first ‘endocannabinoid’ described, and anandamide signaling has been reported to be an inhibitor of axon regeneration. Anandamide acts primarily on cannabinoid receptor type 1 (CB₁)[1], cannabinoid receptor type 2 (CB₂)[2], and ionotropic receptors, including the transient receptor potential (TRP) vanilloid type 1 (TRPV1)[3,4,5], TRP ankyrin type 1 (TRPA1) and TRP melastatin type 8 (TRPM8) channels[6]. Anandamide can also directly modulate various other ion channels[5,7,8,9,10].

CB₁ and CB₂ are members of the subfamily of G-protein-coupled receptors and predominantly couple to Gi/o[1,11] to produce multiple cellular effects, such as the inhibition of adenylate cyclase and of voltage-gated calcium channels[12,13,14], regulation of potassium currents[15,16,17], and increase of Ca²⁺ influx via G_s[18] and G_q[19]. TRPV1 is a polymodal sensor of noxious stimuli including heat, hydrogen ions and capsaicin[20]. CB₁[21] and TRPV1, but not CB₂[21] are expressed in presynaptic primary neurons such as those in the trigeminal ganglion. CB₁ is localized to presynaptic nerve terminals[22] and contributes to the regulation of neuronal excitability and neurotransmitter release by modulating Ca²⁺ signals. TRPV1 activation causes Ca²⁺ to enter into the cell and promotes neurotransmitter release. Studies have demonstrated that anandamide can activate TRPV1 by binding to cytosolic sites[23]. The co-expression and close distribution of CB₁ and TRPV1 in primary sensory neurons[24,25,26,27] allows cross-talk between these two receptors, which further complicates the role of anandamide in nociception and antinociception. The present study will determine which receptor contributes to the effect of low and high concentrations of anandamide in small trigeminal ganglion neurons.

It is a widely accepted general hypothesis that endocannabinoids inhibit neurotransmitter release from primary afferent neurons by reducing Ca²⁺ influx via an inhibitory action on high voltage-gated Ca²⁺ channels. This hypothesis is supported by evidence that CB₁ activation by anandamide on primary nociceptive neurons causes antinociception by reducing high voltage-activated Ca²⁺ channel activity[12,13,14] and Ca²⁺ influx[28], thus inhibiting neurotransmission. In contrast, some studies suggest that cannabinoids can promote Ca²⁺ influx and hence increase neurotransmitter release in vitro[19,25,29,30,31]. How endocannabinoids may cause such opposing effects in intracellular Ca²⁺ concentration and neuron excitability is unclear. High voltage-gated calcium channels and ligand-gated channels, two important contributors to intracellular Ca²⁺ concentration, will also be investigated. The present study tests the above hypothesis by determining the effect of anandamide on high voltage-activated Ca²⁺ currents (I_HVA), Ca²⁺ influx and intracellular Ca²⁺ concentration, as well as the underlying mechanisms, in small trigeminal ganglion neurons.

Results

Anandamide-evoked inward currents

Whole-cell patch-clamp recordings were carried out in rat trigeminal ganglion neurons with bath perfusion of 0.1, 0.3, 1, 3, 10, 30 and 100 μmol/L anandamide at a range of holding potentials from −60 mV to +60 mV (Figure 1). Reversal potentials were around 0 mV. Anandamide-evoked inward and outward currents at holding potentials of −60 mV and +60 mV, respectively, were dose-dependent (Figure 2).

Figure 1.

Anandamide-evoked currents in rat trigeminal ganglion neurons.

(A) Typical traces of 10 µmol/L anandamide-induced currents at different holding potentials (−60 to 60 mV) applied to a neuron. Bar indicates duration of stimulus. (B) Current-voltage relationship of 10 µmol/L anandamide-induced currents at different holding potentials. s: Second.

Figure 2.

Dose-dependent anandamide-evoked currents (nA) in trigeminal ganglion neurons, measured by the voltage-ramp method from –60 to +60 mV.

n numbers represent individual experiments for each data point.

Inhibition of I_HVA by anandamide

Bath application of anandamide (0.001, 0.01, 0.1, 1, 10, 30 μmol/L) inhibited I_HVA in a concentration-dependent manner, by 4.62 ± 0.88%, 6.24 ± 2.57%, 18.17 ± 0.99%, 31.30 ± 3.21%, 64.73 ± 1.95% and 55.19 ± 2.06%, respectively (n = 4–11). The effect of anandamide was partially reversed after washout. The Hill equation was applied to the dose-response curve and revealed that the half-maximal inhibitory concentration (IC₅₀) of anandamide was 0.92 μmol/L. To confirm whether other CB₁ agonists mimicked the inhibition of anandamide, we tested the effect of WIN 55,212-2 (10 μmol/L) on I_HVA. The inhibition of I_HVA by 10 μmol/L WIN 55,212-2 was 55.78 ± 6.07%. Anandamide at 1 μmol/L reduced current amplitudes (Figure 3A, B) but did not cause a significant shift of the activation curve (Figure 3C). However, a hyperpolarization shift of almost 14 mV (n = 8; P < 0.05) was observed in the h-infinity curve (Figure 3E).

Figure 3.

Inhibition of high voltage-activated Ca²⁺ currents (I_HVA) in rat trigeminal ganglion neurons by exposure to anandamide (1 μmol/L) for 3 minutes.

(A) I_HVA reduced from −2.42 nA to −1.67 nA, returning to −2.29 nA after 3 minutes of washout. Traces were evoked by 450 ms step depolarization from −50 mV to 40 mV in 10 mV increments. (B) Current-voltage curve of I_HVA. Anandamide perfusion for 3 minutes reduced the peak current in-tensity of I_HVA from −51.50 ± 0.85 pA/pF to −32.73 ± 1.97 pA/pF (mean ± SEM, n = 8; paired t-test, anandamide, vs. control, P < 0.001). (C) As the amplitude of I_HVA reduced, the activation curve was not affected. The Boltzmann function was fitted to the activation curve. Before and after 1 µmol/L anandamide application, V_0.5 values were −3.73 ± 4.36 mV and −6.47 ± 3.02 mV, respectively (n = 7, paired t-test, P > 0.05), and k values (slope) were 8.74 ± 0.75 and 10.10 ± 0.58, respectively (n = 7, paired t-test, P > 0.05). (D) Effect of 1 µmol/L anandamide on the h-infinity curve. The steady-state inactivation-voltage protocol consisted of 3 seconds of preconditioned pulses ranging from −80 mV to 20 mV followed by a 200 ms test pulse depolarizing to 20 mV. (E) Anandamide hyperpolarization shifted the h-infinity curve. After fitting the Boltzmann function, V_0.5 values were −30.93 ± 2.75 mV and −45.32 ± 5.61 mV (n = 8, paired t-test, P < 0.05), and k values were −15.78 ± 1.33 and −15.85 ± 2.58 (n = 8, paired t-test, P > 0.05) before and after anandamide administration. s: Second.

Roles of cannabinoid and vanilloid receptors in the inhibition of I_HVA by anandamide

Since anandamide activates TRPV1, CB₁ and CB₂ receptors, we tested whether capsazepine, AM251 and AM630, selective antagonists at the three receptors, respectively, could reverse the effect of anandamide on I_HVA. Capsazepine (10 μmol/L), AM251 (10 μmol/L), AM630 (10 μmol/L), or anandamide (1 and 10 μmol/L) were added to the bath solution during whole-cell patch clamp recordings in rat trigeminal ganglion neurons. First, we examined how pre-incubation of the cells for 3 minutes with capsazepine, a competitive antagonist at TRPV1, would affect the inhibition induced by anandamide (Figure 4A, D). Pre-incubation with capsazepine did not affect the inhibition of I_HVA by 1 μmol/L anandamide (n = 10, P > 0.05), but abolished the effect of 10 μmol/L anandamide (n = 12, P < 0.05). These data indicate that TRPV1 activation is involved in the inhibition of I_HVA by high-concentration anandamide. AM251 was co-applied with anandamide (1 or 10 μmol/L) (Figure 4B, E). Low-dose (n = 7, P < 0.05) and high-dose (n = 4, P < 0.05) anandamide-induced inhibition was reversed by AM251 (Figure 4G), indicating that activation of CB₁ is essential for the negative modulation of I_HVA by anandamide in rat trigeminal ganglion neurons.

Figure 4.

Cannabinoid and TRPV1 receptor involvement in the inhibition of high voltage-activated Ba2+ currents (I_HVA) by 1 and 10 μmol/L anandamide in rat trigeminal ganglion neurons.

Pre-incubation with capsazepine (A) or AM630 (C) did not affect I_HVA inhibition by 1 µmol/L anandamide. (B) Pre-incubation with AM251 abolished I_HVA inhibition by 1 µmol/L anandamide. The decrease in I_HVA induced by 10 µmol/L anandamide was partially attenuated by pre-infusion with capsazepine (D) or AM251 (E). Pre-incubation with AM630 (F) had no effect on 10 µmol/L anandamide-induced inhibition of II_HVA. (G) Summary data of the effect of cannabinoid and TRPV1 receptor antagonists on the inhibition of I_HVA by 1 and 10 µmol/L anandamide. ^aP < 0.01, vs. inhibition induced by 1 µmol/L anandamide (unpaired Student's t-test). ^bP < 0.01, vs. inhibition induced by 10 µmol/L anandamide (unpaired Student's t-test). Data are expressed as mean ± SEM. n: number of neurons tested. CPZ: Capsazepine; AEA: anandamide.

Pre-incubation for 3 minutes with AM630 did not affect I_HVA inhibition induced by 1 μmol/L (Figure 4C, G; P > 0.05) or 10 μmol/L (Figure 4F, G; P > 0.05) anandamide, indicating a lack of CB2 receptors in rat trigeminal ganglion neurons[21].

Inhibition of I_HVA by capsaicin

As described above, inhibition of I_HVA by high-concentration anandamide was reversed by the TRPV1 antagonist capsazepine, suggesting that TRPV1 is involved in this effect. We further tested whether I_HVA could be directly inhibited by the TRPV1 agonist, capsaicin. Capsaicin and capsazepine were applied by bath perfusion. I_HVA was reversibly inhibited by capsaicin in a dose-dependent manner (Figure 5). Similar to the effect observed with high-concentration anandamide, the decrease in I_HVA by capsaicin was reversed by 10 μmol/L capsazepine. In addition, I_HVA was not inhibited by 10 μmol/L capsazepine (Figure 6).

Figure 5.

Cannabinoid and transient receptor potential (TRP) vanilloid type 1 receptor involvement in the inhibition of high voltage-activated Ca²⁺ currents (I_HVA) by 1 and 10 μmol/L anandamide in rat trigeminal ganglion neurons.

(A) I_HVA currents were generated in a neuron with a membrane resistance of 878 MΩ. Application of 0.3 µmol/L capsaicin induced a maximum inward current of 0.6 nA, which desensitized to 0.3 nA after about 60 seconds (sec). At this time, the membrane resistance was 164 MΩ and I_HVA was reduced from 2.4 to 0.4 nA. After washing the cell for 20 seconds (sec), the capsaicin-induced inward current returned to baseline and the membrane resistance increased to 856 MΩ. I_HVA recovered to 0.8 nA. Current-voltage curves are shown in the absence and presence of capsaicin as well as for the 20 sec and 3-minute (min) washes. (B) Dose-dependent inhibition of IHVA by capsaicin in capsaicin-sensitive trigeminal ganglion neurons. The Hill equation was fitted to the dose-response curve, with IC₅₀ = 0.21 µmol/L. n: Times of individual experiments carried out at each concentration. P < 0.05, vs. before capsaicin (one-way analysis of variance).

Figure 6.

Capsazepine (CPZ) inhibited the capsaicin-induced decrease of Ca²⁺ current (I_HVA) in trigeminal ganglion neurons.

(A) Representative traces showing the effects of co-application of capsazepine and capsaicin in a capsaicin-sensitive neuron. (B) Summary data of capsazepine blockade of capsaicin-induced inhibition of I_HVA. ^aP < 0.01, vs. capsaicin alone by unpaired t-test. “n” indicate the times of neurons tested. After I_HVA was obtained, neurons were incubated for 3 minutes (min) with 10 µmol/L capsazepine. Capsazepine did not affect the I_HVA. With a subsequent application of 0.3 µmol/L capsaicin plus 10 µmol/L capsazepine, and after a 3-minute wash, I_HVA remained unchanged. An inward current was induced by 10 µmol/L capsaicin, suggesting it was a capsaicin-sensitive neuron. On average, in the presence of 0.3 µmol/L capsaicin, I_HVA was inhibited by 70.0 ± 9.9% (n = 8). In the presence of 10 µmol/L capsazepine, IHVA was inhibited by 2.8 ± 1.8% (n = 6). In the presence of 0.3 µmol/L capsaicin plus 10 µmol/L capsazepine, the inhibition of capsaicin was reduced to 15.5 ± 8.9% (P < 0.01). s: Second.

Characterization of signal transduction pathways mediating the inhibition of I_HVA via cannabinoid type-1 receptor activation by low-concentration anandamide

Three important signal transduction systems were examined using specific agonists and antagonists to test whether these systems participated in the inhibition of I_HVA by CB1 receptors activated by 1 μmol/L anandamide.

Whole-cell patch clamp measurements in trigeminal ganglion neurons revealed that application of 1 μmol/L KT5720, an inhibitor of cyclic adenosine monophosphate -dependent protein kinase A, significantly attenuated the inhibition of I_HVA by 1 μmol/L anandamide (Figure 7). Incubation with an antagonist of protein kinase C, bisindolylmaleimide (1 μmol/L for 10 minutes) and an inhibitor of cyclic guanosine monophosphate (cGMP)-dependent protein kinase, Rp-8-Br-cGMP (1 μmol/L for 10 minutes) also significantly attenuated anandamide-induced inhibition of I_HVA.

Figure 7.

The role of different signaling pathways in the inhibition of 1 μmol/L anandamide on Ca²⁺ currents (I_HVA) in rat trigeminal ganglion neurons.

Sample traces of I_HVA currents pre-treated with KT5720 (1 µmol/L for 10 minutes in the bath solution) (A) and responding to co-application with 1 µmol/L anandamide for 3 minutes (B). (C) Averaged effects of 8-Br-cAMP alone and KT5720 co-application with 1 µmol/L anandamide on I_HVA. Typical traces of I_HVA currents pre-treated with BIM (10 µmol/L for 10 minutes in the bath solution) (D) and responding to co-application with 1 µmol/L anandamide for 3 minutes (E, F). Typical traces of I_HVA currents in neurons pre-treated with Rp-8-Br-cGMP (1 µmol/L for 10 minutes in the bath solution) (G) and anandamide (1 µmol/L) for 3 minutes (H). (I) Mean (± SEM) effects on I_HVA of pCPT-cGMP alone and Rp-8-Br-cGMP co-application with anandamide (1 µmol/L). P < 0.01, vs. anandamide (unpaired t-test). Numbers in bars indicate the number of neurons tested. IHVA: High voltage-activated Ca²⁺ currents; cGMP: cyclic guanosine monophosphate; cAMP: cyclic adenosine monophosphate; AEA: anandamide.

Anandamide increased intracellular Ca²⁺ concentration via a direct influx of extracellular calcium into rat trigeminal ganglion neurons

Calcium imaging was used to examine the effect of anandamide on intracellular Ca²⁺ concentration in the presence of solutions containing either 0 or 2 mmol/L Ca²⁺ (Figure 8A). In the absence of extracellular calcium (buffered by 10 mmol/L ethylene glycol bis(alpha-aminoethyl ether)-N,N′-tetraacetic acid to decrease extracellular Ca²⁺ concentrations), the rise in intracellular Ca²⁺ concentration induced by anandamide was notably decreased (Figure 8B; anandamide([Ca²⁺]₊): 20.33 ± 2.10%, n = 19; anandamide([Ca²⁺]₋): 4.22 ± 1.06%, n = 48; P < 0.01). Thus, the anandamide-induced rise in intracellular Ca²⁺ concentration is dependent on Ca²⁺ influx from the extracellular space.

Figure 8.

Effects of anandamide on intracellular Ca²⁺ concentration ([Ca²⁺]_i) in rat trigeminal ganglion neurons.

(A) Concentration-response curve for increases in ([Ca²⁺]_i) with anandamide application. (B) Different responses in [Ca²⁺]_i evoked by 1 µmol/L anandamide in the presence and absence of extracellular Ca²⁺ ([Ca²⁺]_o). A significant effect was seen with 2 mmol/L Ca²⁺ present in the extracellular environment (n = 19) compared with the Ca²⁺-free condition (n = 48), ^aP < 0.01, vs. presence of extracellular Ca²⁺ (unpaired Student's t-test). (C) Responses (mean ± SEM) induced by 1 and 10 µmol/L anandamide and their co-application with AM251 and capsazepine. Pre-incubation with 10 µmol/L AM251 reversed the rises evoked by 1 and 10 µmol/L anandamide (^bP < 0.01, vs. 1 µmol/L anandamide alone; one-way analysis of vari-ance; ^cP < 0.01 vs. 10 µmol/L anandamide alone; one-way analysis of variance). 1 and 3 µmol/L AM251 facilitated the increase induced by 1 µmol/L anandamide. Pre-application of 10 µmol/L capsazepine blocked the 10 µmol/L anandamide-induced increase in intracellular Ca²⁺ concentration (n = 14, ^cP < 0.01, vs. 10 µmol/L anandamide alone), while the increase evoked by 1 µmol/L anandamide was not affected by pre-treatment with 10 µmol/L capsazepine (P > 0.05, vs. 1 µmol/L anandamide alone). (D) Mean (±SEM) responses of preperfusion with KT5720, staurosporine and KT5823 and followed by co-application of 1 µmol/L anandamide. The rise evoked by 1 µmol/L anandamide was blocked by KT5720 (^dP < 0.01, vs. 1 µmol/L anandamide alone) and KT5823 (^dP < 0.01, vs. anandamide alone) but not staurosporine (P > 0.05, vs. anandamide alone, one-way anal-ysis of variance). Numbers in parentheses indicate the number of neurons tested. CPZ: Capsazepine; AEA: anandamide.

Cannabinoid / vanilloid type-1 receptor involvement in the mechanism underlying increased intracellular Ca²⁺ concentration evoked by anandamide

Because AM630 had no effect on the inhibition of anandamide on I_HVA, in the present test only AM251 and capsazepine were used, to determine whether they would affect the increase of anandamide on the intracellular Ca²⁺ concentration, using calcium imaging. The 1 μmol/L anandamide-induced rise in intracellular Ca²⁺ concentration was abolished by co-application with 10 μmol/L AM251 Figure 8C, while lower doses of AM251 (1 and 3 μmol/L) facilitated the increase of 1 μmol/L anandamide Figure 8C. This suggests that co-incubation with 1 μmol/L anandamide and low-dose AM251 facilitates intracellular Ca²⁺ influx via TRPV1 while at the same time inhibiting it via CB₁ receptors.

Similar to the effect of 10 μmol/L anandamide observed on I_HVA, the increase in intracellular Ca²⁺ concentration was also blocked by both AM251 and capsazepine (Figure 8C), indicating that different receptors participate in the responses of intracellular Ca²⁺ concentration to low and high concentrations of anandamide in rat trigeminal ganglion neurons.

Signaling pathways mediating the increase in intracellular Ca²⁺ concentration after cannabinoid receptor activation by low-concentration anandamide

Calcium imaging was used to evaluate the possible contribution of cAMP-dependent protein kinase, protein kinase C, and cGMP-dependent protein kinase systems to the observed effect on intracellular Ca²⁺ concentration after CB1 activation. The effects of 1 μmol/L anandamide on intracellular Ca²⁺ concentrations in trigeminal ganglion neurons pre-treated with KT5720 (an inhibitor of cAMP-dependent protein kinase), staurosporine (an inhibitor of protein kinase C) and KT5823 (an inhibitor of cGMP-dependent protein kinase) were investigated. Figure 8D showed that staurosporine (10 μmol/L for 15 minutes) had no effect on the rise in intracellular Ca²⁺ concentration after anandamide application (n = 13, P > 0.05). Pretreatment with KT5720 (10 μmol/L for 10 minutes) and KT5823 (10 μmol/L for 10 minutes) markedly attenuated the response of 1 μmol/L anandamide Figure 8D. These data suggest that the cAMP-dependent protein kinase and cGMP-dependent protein kinase systems, but not the protein kinase C system, are involved in the increase in intracellular Ca²⁺ concentration with 1 μmol/L anandamide.

Discussion

It is well established that presynaptic CB₁ receptor activation has an inhibitory role on high voltage-activated Ca²⁺ channels, resulting in a decrease in Ca²⁺ influx and an inhibition of neurotransmitter release. In contrast, results from the present study show that anandamide causes a dose-dependent elevation of intracellular Ca²⁺ concentration, triggered by anandamide-evoked currents, but not modulation of high voltage-activated Ca²⁺ channels by anandamide. Furthermore, anandamide-induced inhibition of the high voltage-activated Ca²⁺ channels and the increase in the intracellular Ca²⁺ concentration are demonstrated to be two separate events differing in signal transduction pathways but having similar effects via the same receptors. Thus, our data suggest that anandamide, as an endocannabinoid, might exert an excitatory effect by increasing intracellular Ca²⁺ concentration, triggered by anandamide-evoked currents.

It has been well documented that the endocannabinoid anandamide, and its synthetic analogue methanandamide, have dual effects. At low concentrations, anandamide selectively activates CB₁ receptors, while at high concentrations it activates excitatory TRPV1 receptors[24]. In pre-contracted strips of rat hepatic artery, rat small mesenteric artery and guinea-pig basilar artery, vasorelaxation induced by high-concentration anandamide is antagonized by the TRPV1 antagonist capsazepine, but not by the CB₁ receptor antagonist SR141716 (0.3 μmol/L)[4]. Further evidence indicates that anandamide not only acts on transfected TRPV1 receptors to produce membrane currents and increase the intracellular Ca²⁺ concentration[4,32], but also acts on naturally-expressed TRPV1 receptors in neonatal rat dorsal root ganglia to produce membrane currents[33]. Here, we present strong evidence that anandamide selectively activates the CB₁ receptor at low concentrations in small trigeminal ganglion neurons. At high concentrations, anandamide activates both CB₁ and TRPV1 receptors, inhibiting I_HVA. To confirm that high-concentration anandamide inhibited I_HVA via TRPV1 receptor activation, we compared the effects of high-concentration anandamide and the TRPV1 agonist capsaicin. We found that the inhibition induced by high-concentration anandamide and by capsaicin shared similar characteristics. First, both inhibited I_HVA in a dose-dependent manner and were reversed by the TRPV1 receptor antagonist capsazepine. Second, like capsaicin, high-concentration anandamide induced inward currents in cultured trigeminal ganglion neurons. Similar to capsaicin-induced inward current, the reversal potential of anandamide-induced current was also around 0 mV. Anandamide had different effects at low and high concentration via the activation of different receptors. By comparing the contribution of CB₁ and TRPV1 activation on anandamide-induced increases in intracellular Ca²⁺ concentration, we conclude that anandamide-induced inhibition of high voltage-activated Ca²⁺ channels and an increase in intracellular Ca²⁺ concentration have the same dual effects as activation of the receptor pathway.

Inhibition of high voltage-activated Ca²⁺ channels and modulation of Ca²⁺ influx are important for neuronal activity and neurotransmitter release induced by endocannabinoids. However, evidence also suggests that cannabinoids can couple to G_s[18] and G_q[19] and activate TRPV1 receptors to increase Ca²⁺ influx. In addition, the response of cannabinoids on intracellular Ca²⁺ concentration differs across cell types[34]. To elucidate the complex mechanisms underlying the intracellular Ca²⁺ response, we further tested the effect of anandamide on intracellular Ca²⁺ concentration. We found that anandamide increased intracellular Ca²⁺ and simultaneously inhibited IHVA in the same population of trigeminal ganglion neurons, providing strong evidence that an increase in the intracellular Ca²⁺ concentration is not due to I_HVA. We found that anandamide had dual effects on the inhibition of I_HVA and elevation of intracellular Ca²⁺ concentration. However, by comparing the contribution of second messenger system, we found that cAMP- and cGMP-dependent protein kinase antagonists both reversed the anandamide-induced increase in intracellular calcium concentration and I_HVA inhibition. Application of a protein kinase C antagonist reversed the inhibition of I_HVA but the anandamide-induced increase in intracellular Ca²⁺ concentration was not blocked. From this, we can conclude that the increase in intracellular Ca²⁺ concentration and the I_HVA inhibition observed with anandamide are separate processes that share the same receptors but differ in intracellular signal transduction pathways.

CB₁ receptor activation by endocannabinoids has been mostly reported to cause depression of neuronal excitability and neurotransmitter release in presynaptic primary neurons[10,13,25,28,35,36]. Accumulating evidence indicates that cannabinoids may contribute to the potentiation of neurotransmission. Endocannabinoid release potentiates synaptic transmission via CB₁ receptor activation and dopamine release in the goldfish Mauthner cell[37]. Moreover, endocannabinoids may potentiate hippocampal synaptic transmission via astrocytic CB₁ activation[30]. In the suprachiasmatic nucleus, CB₁ receptor activation can increase excitability of circadian clock neurons[38]. It is widely accepted that CB₁ receptor activation suppresses neurotransmission by inhibiting high voltage-activated Ca²⁺ channels, and decreases Ca²⁺ influx. In contrast to this general hypothesis, we found that endocannabinoids caused elevation of the intracellular Ca²⁺ concentration in trigeminal ganglion neurons, which was triggered by endocannabinoid-evoked inward current around the resting membrane potential (−60 mV). Furthermore, endocannabinoid-induced increase in intracellular Ca²⁺ concentration and decrease of high voltage-activated Ca²⁺ channels were two independent events, which had the same dual effects via similar receptor pathways, but differed in the second messenger transduction pathway. Thus, our results suggest that presynaptic endocannabinoid release might potentiate neurotransmission via endocannabinoid-evoked current, increased Ca²⁺ influx, and increased intracellular Ca²⁺ concentration.

In summary, anandamide dose-dependently causes an increase in intracellular Ca²⁺ concentration, mediated by Ca²⁺ influx via anandamide-evoked currents, but not high voltage-activated Ca²⁺ channels. In the same population of trigeminal ganglion neurons, CB₁ receptor activation-induced inhibition of I_HVA and increase in high voltage-activated Ca²⁺ channels are two separate processes that share the same (CB₁) receptors, but differ in signal transduction pathway. Thus, the effects of anandamide on high voltage-activated Ca²⁺ channels ultimately depend on the balance between the increased Ca²⁺ influx via anandamide-evoked currents and inhibition of high voltage-activated Ca²⁺ channels, suggesting a complex role of anandamide on inhibitory or excitatory neuromodulation. Contrary to the general hypothesis of endocannabinoid depression on neurotransmission, this study elucidates possible mechanisms underlying endocannabinoid-induced potentiation of neurotransmission via Ca²⁺ signaling modulation.

Materials and Methods

Design

A cytological in vitro study.

Time and setting

All data were collected at the Departments of Pharmacology and Physiology, Tongji Medical College, Huazhong University of Science and Technology in China from September 2003 to April 2008. Data analysis and manuscript writing were performed at Baylor Medical College, USA, from April 2011 to October 2012.

Materials

180 male Sprague-Dawley rats weighing 180–200 g and aged 6–8 weeks were used. All animal protocols were approved by the faculty of Laboratory Animal Science, Huazhong University of Science and Technology (license No. SYXK (E) 2009-0049). All the experiments followed the Guidance Suggestions for the Care and Use of Laboratory Animals, formulated by the Ministry of Science and Technology of China[39].

Methods

Cell dissociation

Trigeminal ganglion neurons from Sprague-Dawley rats were cultured as described previously[40]. Briefly, trigeminal ganglia were dissected aseptically and washed with cold (4°C) modified Hank's balanced salt solution containing NaCl 130 mmol/L, KCl 5 mmol/L, KH₂ PO₄ 0.3 mmol/L, NaHCO₃ 4.0 mmol/L, NaH₂ PO₄ 0.3 mmol/L, D-glucose 5.6 mmol/L, and ethylene glycol bis(alpha-aminoethyl ether)-N, N′-tetraacetic acid 10 mmol/L, hydroxyethyl piperazine ethanesulfonic acid 10 mmol/L, at pH 7.4. The ganglia were chopped into small pieces, and then incubated in 3 mL modified Hank's solution with 0.1% collagenase (type XI–S) for 20–40 minutes at 37°C. Individual cells were dissociated by triturating them through a fire-polished glass pipette, followed by incubating with 10 μg/mL DNase I (type lV) in F12 medium (Life Technologies, Gaithersburg, MD, USA) for 10 minutes at 37°C, before centrifuging for 5 minutes at 1,000 r/min. After centrifuging three times, the cells were cultured in F12 supplemented with 10% fetal bovine serum. The cells were planted on poly-D-lysine pre-coated glass coverslips (15 mm diameter) and cultured no more than 12 hours at 37°C in a water saturated atmosphere with 5% CO₂.

Patch-clamp recording

The cells were placed in a recording chamber mounted on an inverted microscope (Leica Inc., Solms, Germany) and perfused with extracellular solution at room temperature (21–22°C). Whole-cell patch-clamp experiments were carried out using an Axopatch 200B amplifier (Axon Instruments, Foster City, CA, USA) and the output was digitized with a Digidata 1332A converter (Axon Instruments) and program pCLAMP 9.02 (Axon Instruments). Data were acquired at a sampling rate of 2 KHz. Cell membrane capacitance and series resistance were measured and compensated (> 90%). Data obtained from neurons in which uncompensated series resistance resulted in voltage-clamp errors > 5 mV were not taken into further analysis. The cell diameters were measured with a calibrated eyepiece under phase contrast illumination.

The resistance of the microelectrode was 2–4 MΩ when filled with the pipette solution. The microelectrode was made from G85150T-4 glass pipettes (Warner Instruments Inc., Hamden, CT, USA). The external solution contained Choline-Cl 110 mmol/L, TEACl 20 mmol/L, BaCl₂ 10 mmol/L, MgCl₂ 2.0 mmol/L, hydroxyethyl piperazine ethanesulfonic acid 10 mmol/L, and D-glucose 20 mmol/L, adjusted to pH 7.4 with CsOH. Ba²⁺ was used as the charge carrier when recording I_HVA. Ba²⁺ was replaced by Ca²⁺ when recording anandamide-evoked currents. The pipette solution contained CsCl 120 mmol/L, CaCl₂ 0.1 mmol/L, MgCL₂ 2.0 mmol/L, ethylene glycol bis(alpha-aminoethyl ether)-N, N′-tetraacetic acid 10.0 mmol/L, hydroxyethyl piperazine ethanesulfonic acid 10.0 mmol/L and Tris-ATP 5.0 mmol/L, pH adjusted to 7.2 with CsOH. Small-sized cells were selected to perform the further experiments.

The volume of the recording chamber was about 1 mL and the local superfusion rate was 1 mL/min.

Calcium imaging

Drug-induced changes of high voltage-activated Ca²⁺ channels were measured using a confocal laser scanning imaging system (Fluoview FV500, Olympus, Tokyo, Japan). The trigeminal ganglion cells on glass cover slips were loaded with Fluo 2-AM by incubation with 1–5 μmol/L Fluo 2-AM in standard external solution and were maintained in the dark for 30–40 minutes. The standard external solution contained NaCL 140.0 mmol/L, KCL 5.0 mmol/L, CaCL₂ 2.0 mmol/L, MgCL₂ 1.0 mmol/L, glucose 10.0 mmol/L and hydroxyethyl piperazine ethanesulfonic acid 10.0 mmol/L, pH adjusted to 7.4 with NaOH. Ca²⁺ -free medium solution was identical except for the 2.0 mmol/L Ca²⁺ and additional 10 mmol/L ethyleneglycol bis(alpha-aminoethyl ether)-N, N′-tetraacetic acid to lower extracellular Ca²⁺ concentration. Small-sized trigeminal ganglion neurons (< 33 μmol/L) were selected to perform the further experiments.

Data analysis

The data were analyzed using pCLAMP 9.02 (Axon Instruments) and Sigmaplot 11.0 software (Systat Software Inc., San Jose, CA, USA). The amplitude of I_HVA was calculated as the peak current. Voltage-dependent activation for the study of changes on I_HVA was measured by a series of depolarized pulses (450 ms) from −50 mV to +40 mV, stepping by 10 mV with interval time of 5 seconds, at a holding potential of −80 mV. We fitted a Boltzmann function to the voltage-dependent activation curves, that is G/G_max =1/[1+exp (V_0.5–V_m)/k], where G_max is the maximum conductance, V_0.5 is the membrane potential at which 50% of activation was observed, and k is the slope of the function. Voltage-dependent inactivation was measured by a two pulse protocol in which the precondition pulses (3 seconds) ranged from −80 to +20 in 10 mV increments; following test pulse (200 millsecords) was +10 mV with an internal time of 6 seconds. The Boltzmann function was also fitted to the h-infinity curve, that is, I/I_max=1/[1+exp(V_0.5–V_m)/k], where V_0.5 is the membrane potential at which 50% of inactivation was observed, and k is the slope of the function. The dose-response curve was fitted by the Hill equation, in which, I_peak=I_peakmax/[1+(IC₅₀/C)ⁿ], with IC₅₀ as the concentration producing 50% inhibition and n as the Hill coefficient.

Statistical analysis

Data were presented as mean ± SEM. For all experiments, data were examined for Gaussian distribution first, and then analyzed for statistical significance using the paired or unpaired t-test and one-way analysis of variance by using Sigma plot 11.0 software (Systat Software Inc.). A value of P < 0.05 was considered statistically significant.