Chemical Determinants Involved in Anandamide-induced Inhibition of T-type Calcium Channels*

Anandamide, originally described as an endocannabinoid, is the main representative molecule of a new class of signaling lipids including endocannabinoids and N-acyl-related molecules, eicosanoids, and fatty acids. Bioactive lipids regulate neuronal excitability by acting on G-protein-coupled receptors (such as CB1) but also directly modulate various ionic conductances including voltage-activated T-type calcium channels (T-channels). However, little is known about the properties and the specificity of this new class of molecules on their various targets. In this study, we have investigated the chemical determinants involved in anandamide-induced inhibition of the three cloned T-channels: CaV3.1, CaV3.2, and CaV3.3. We show that both the hydroxyl group and the alkyl chain of anandamide are key determinants of its effects on T-currents. As follows, T-currents are also inhibited by fatty acids. Inhibition of the three CaV3 currents by anandamide and arachidonic acid does not involve enzymatic metabolism and occurs in cell-free inside-out patches. Inhibition of T-currents by fatty acids and N-acyl ethanolamides depends on the degree of unsaturation but not on the alkyl chain length and consequently is not restricted to eicosanoids. Inhibition increases for polyunsaturated fatty acids comprising 18–22 carbons when cis-double bonds are close to the carboxyl group. Therefore the major natural (food-supplied) and mammalian endogenous fatty acids including γ-linolenic acid, mead acid, and arachidonic acid as well as the fully polyunsaturated ω3-fatty acids that are enriched in fish oil eicosapentaenoic and docosahexaenoic acids are potent inhibitors of T-currents, which possibly contribute to their physiological functions.


Anandamide, originally described as an endocannabinoid, is the main representative molecule of a new class of signaling lipids including endocannabinoids and N-acyl-related molecules, eicosanoids, and fatty acids. Bioactive lipids regulate neuronal excitability by acting on G-protein-coupled receptors (such as CB1) but also directly modulate various ionic conductances including voltage-activated T-type calcium channels (T-channels). However, little is known about the properties and the specificity of this new class of molecules on their various targets. In this study, we have investigated the chemical determinants involved in anandamide-induced inhibition of the three cloned T-channels: Ca V 3.1, Ca V 3.2, and Ca V 3.3. We show that both the hydroxyl group and the alkyl chain of anandamide are key determinants of its effects on T-currents. As follows, T-currents are also inhibited by fatty acids. Inhibition of the three Ca V 3 currents by anandamide and arachidonic acid does not involve enzymatic metabolism and occurs in cell-free inside-out patches. Inhibition of T-currents by fatty acids and N-acyl eth-
anolamides depends on the degree of unsaturation but not on the alkyl chain length and consequently is not restricted to eicosanoids. Inhibition increases for polyunsaturated fatty acids comprising 18 -22 carbons when cis-double bonds are close to the carboxyl group. Therefore the major natural (food-supplied) and mammalian endogenous fatty acids including ␥-linolenic acid, mead acid, and arachidonic acid as well as the fully polyunsaturated 3-fatty acids that are enriched in fish oil eicosapentaenoic and docosahexaenoic acids are potent inhibitors of T-currents, which possibly contribute to their physiological functions.
Voltage-dependent calcium channels comprise three families: the L-type channels (Ca V 1 family), the neuronal N-, P/Q-, and R-type channels (Ca V 2 family), and the T-type channels (Ca V 3 family). The electrophysiological features of T-type Ca 2ϩ channels (T-channels) are low voltage-activated Ca 2ϩ currents, low unitary conductance, fast inactivation and slow deactivation kinetics, and strong steady-state inactivation at physiological resting potentials (1,2). Three T-channel subunits have been identified: Ca V 3.1 (or ␣ 1G ), Ca V 3.2 (or ␣ 1H ), and Ca V 3.3 (or ␣ 1I ) (2). In mammalian expression systems, the Ca V 3.1 and Ca V 3.2 currents share typical properties of native T-channels, while Ca V 3.3 currents display unusually slow inactivation kinetics, which are characteristic of specific neurons (1)(2)(3). These three subunits are also subject to alternative splicing, which influences their electrophysiological properties (4 -6). The Ca V 3.1 and Ca V 3.2 subunits are widely expressed in various tissues, mainly in the heart and in the nervous system, whereas Ca V 3.3 is restricted to the central nervous system (2). In the nervous system, T-channels generate low threshold spikes and participate in spontaneous firing (1,2). They are involved in slow wave sleep (7), in absence epilepsy (8,9), and in the pain perception (10 -12). In heart cells, T-channels are involved in cardiac pacemaker (13,14) and cardiac hypertrophy (15). In nonexcitable cells, they participate in hormone secretion (16) and fertilization (17). They are also suspected to induce cellular differentiation of many cell types (18 -21). However, the lack of specific blockers and endogenous ligands of the T-channels has precluded the elucidation of their functions.
We recently identified that anandamide, an endogenous signaling molecule that modulates neuronal excitability (22,23), is a direct blocker of T-type calcium channels (24). Anandamide, primarily described as an endocannabinoid acting on CB1 receptors (25), also modulates various ionic conductances (26,27). Anandamide (N-arachidonoyl ethanolamide, AEA) 2 belongs to a new major class of small lipid messengers, including endocannabinoids and N-acyl-related molecules, eicosanoids, and fatty acids (22, 28 -32). These bio-active lipids are involved in a multitude of physiological and pathophysiological events, including neuronal excitability (22,23,31,33), sleep (34), epilepsy and neuroprotection (34 -37), inflammation and pain (28,29,32,38,39), as well as cardiovascular modulation (40 -42), fertilization, and cell cycle progression (38,43,44). Despite the fact that both small lipid messengers and T-channels display similar physiological functions (but with opposite effects), little is known about the properties and the specificity of this new class of molecules on the three Ca V 3 channels. In the present study, we provide the first detailed analysis of the chemical determinants involved in anandamide-induced inhibition of Ca V 3 T-channels.

MATERIALS AND METHODS
Cell Culture and Transfection Protocols-TsA-201 cells were cultivated in Dulbecco's modified Eagle's medium supplemented with glutamax and 10% fetal bovine serum (Invitrogen). Transfection was performed using jet-PEI (QBiogen) according to the manufacturer protocol (4 l of jet-PEI for ϳ1.5 g of DNA) with a DNA mix containing 0.5% of a green fluorescent protein plasmid and 99.5% of either of the pcDNA3 plasmid constructs that code for human Ca V 3.1a, Ca V 3.2, and Ca V 3.3 T-channel isoforms. Two days after, cells were then dissociated with trypsin, and electrophysiological recordings were performed during the next 2 days.
Electrophysiological Recordings-Macroscopic currents were recorded at room temperature (ϳ24°C) by the whole-cell patch clamp technique using an Axopatch 200B amplifier (Axon Instruments). Extracellular solution contained (in mM): 135 NaCl, 20 tetraethylammonium chloride, 2 CaCl 2 , 1 MgCl 2 , and 10 HEPES (pH adjusted to 7.4 with KOH, ϳ330 mosM). Borosilicate glass pipettes have a typical resistance of 1.5-2.5 M⍀ when filled with an internal solution containing (in mM): 140 CsCl, 10 EGTA, 10 HEPES, 3 Mg-ATP, 0.6 GTPNa, and 2 CaCl 2 (pH adjusted to 7.2 with KOH, ϳ315 mosM). For cell-free inside-out experiments the bath solution contained (in mM): 145 KCl, 5 EGTA, 1 MgCl 2 , and 10 HEPES (pH adjusted to 7.4 with KOH, ϳ310 mosM) and the intra-pipette solution contained (in mM): 100 BaCl 2 and 10 HEPES (pH adjusted to 7.4 with KOH, ϳ310 mosM). Recordings were filtered at 2 kHz. Data were analyzed as described previously (24) using pCLAMP9 (Axon Instruments) and GraphPad Prism (GraphPad Inc.) software. One-way analysis of variance combined with a Student-Newman-Keuls post-test were used to compare the different values and were considered significant at p Ͻ 0.05. Results are presented as the mean Ϯ S.E., and n is the number of cells used.
Chemical Reagents-Fatty acids, N-acyl ethanolamides, as well as others compounds (from Sigma and Cayman Chemical) were dissolved in ethanol at a concentration of 10 -100 mM. Stock solutions were briefly sonicated, aliquoted, sealed under argon, and kept at Ϫ80°C. These aliquots were dissolved daily in the extracellular solution, which was sonicated on ice for 5 min before use. Control experiments were carried out using the solvent alone. Drugs were applied by a gravity-driven homemade perfusion device (mostly using Teflon tubing), controlled by solenoid valves.

RESULTS
Anandamide (AEA) inhibits the three cloned T-channels expressed in tsA-201 cells, as illustrated for Ca V 3.3 (Fig. 1A). The inhibition is potent (ϳ90% for a concentration of 10 M, Fig. 1B), occurs in the minute range (maximal after a 2-3-min perfusion), and is partially relieved by perfusion of a bovine serum albumin solution (3 mg/ml) (data not shown).
We first looked for the principal chemical motifs involved in anandamide inhibition. As schematized in Fig. 1A, we identified that both the hydroxyl group of the anandamide head (indicated in blue in Fig. 1A) as well as its alkyl chain (indicated in green in Fig.  1A) are crucial for anandamide effects. Indeed, arachidonic acid (cis-5,8,11,14-eicosatetraenoic acid, AA), which contains a carboxyl group and the same alkyl chain as anandamide, mimics anandamide effects. As observed with anandamide, the inhibition of the three cloned T-channels by arachidonic acid occurs in the minute range (maximal after a 2-3-min perfusion) and is partially relieved by perfusion of a bovine serum albumin solution (data not shown). As previously described for lower anandamide concentrations (24), arachidonic acid (10 M) preferentially inhibits Ca V 3.2 currents (ϳ66%) compared with Ca V 3.3 (ϳ53%) and Ca V 3.1 (ϳ47%) currents (Fig. 1B, p Ͻ 0.05). However, the arachidonic acid inhibition is less potent than those induced by anandamide on the three Ca V 3 currents (Fig. 1B, p Ͻ 0.001) indicating that although the ethanolamide group of anandamide does not affect per se T-currents (as assessed with ethanolamine (EA) and acetyl ethanolamide (acetyl-EA) (Fig. 1A)), its presence potently increases the T-current inhibition. Furthermore, arachidonic acid methyl ester (AA-methyl ester), which lacks both the hydroxyl and the amide group of anandamide, has no significant effect on the three Ca V 3 currents (Fig. 1, A and B). It should be noted that the hydroxyl group of AEA seems very important to confer strong T-current inhibition since arachidonyl 2Ј-chloroethylamide (ACEA), which contained a chloride instead a hydroxyl group, poorly inhibits the three Ca V 3 currents ( Fig. 1, A and B). Indeed, anandamide derivatives possessing the amide linkage (indicated in red in Fig.  1A) but lacking the hydroxyl group (as arachidonamide, arachidonic acid N-methylamide (ANMA), and ACEA) inhibit the three T-currents (ϳ30%) but weakly compared with arachidonic acid (ϳ55%) and anandamide derivatives (ϳ95%) that possess both the hydroxyl and the amide groups (as AEA and R1-and R2-methanandamide) (Fig. 1B, p Ͻ 0.001).It is important to note that ANMA and ACEA are potent cannabinoid receptor CB1 agonists. In addition, the carbonyl group of AEA seems less crucial since arachidonyl alcohol is as potent as AA to inhibit the three Ca V 3 currents (Fig. 1, A and B). It should be noted that both R1and R2-methanandamide produce maximal inhibition of the three Ca V 3 currents (Fig. 1, A and B).
Similar to that described for anandamide (24), arachidonic acid has pronounced effect on T-channel inactivation (Fig. 6). Indeed, arachidonic acid accelerates inactivation kinetics of T-currents as illustrated for Ca V 3.3 (Fig. 6A). Inactivation kinetics () are ϳ2 times faster in the presence of arachidonic acid for the three Ca V 3 currents (p Ͻ 0.01; n ϭ 8 -14). In addition, arachidonic acid induces a hyperpolarizing shift of the steady-state inactivation curve of T-currents (ϳ12 mV, p Ͻ 0.01 for the three Ca V 3 currents; n ϭ 8 -12; Fig. 6B) without a corresponding effect on the activation curve (Fig. 6, C-F). Therefore, in the presence of arachidonic acid, no current is recorded for a resting potential of Ϫ70 mV, whereas ϳ50% of the current remained in control condition (Fig. 6B). Also, the window current, resulting from the overlap of the activation and inactivation curve, is almost completely abolished in the presence of arachidonic acid (Fig. 6, D-F).
Because the fatty acids and the N-acyl ethanolamides that potently inhibit T-currents contained at least a cis-1,4-pentadiene unit (™C¢C™C™C¢C™) and therefore are potentially metabolized in mammalian cells by lipoxygenase, cyclooxygenase, and cytochrome P450 epoxygenase and -hydrolase pathways (see schematic signaling cascade in Fig. 7A), we investigated whether these latter enzymes could be implicated in AA and AEA-induce T-current inhibition (Fig. 7A). Specific inhibition of cyclooxygenase with indomethacin (10 M) as well as of lipoxygenase with 5,8,11-eicosatriynoic acid (10 M) does not altered T-current inhibition by AA and AEA (Fig. 7A). Similar findings were obtained using 17-octadecynoic acid, which inhibits P450 epoxygenase and -hydrolase enzymes (Fig. 7A). Furthermore, ETYA (10 M), which acts as a nonspecific blocker of all AA-metabolizing enzymes, does not suppress T-current inhibition by either AA or AEA (Fig. 7A). Finally, AEA hydrolysis into AA by fatty acid amide hydrolase (FAAH) is not involved in T-current inhibition induced by AEA, as assessed with phenylmethanesulfonyl fluoride (100 M, Fig. 7A). Because inhibition of T-currents could involve another mechanisms (as G-protein receptor and protein kinase activation), we performed cell-free inside-out patches in the presence of AA and AEA (Fig. 7B) using an "intracellular" medium lacking both GTP and ATP. In this configuration, AA potently inhibits the three Ca V 3 currents (ϳ50% inhibition of Ca V 3.1 and Ca V 3.3 currents and ϳ70% inhibition of Ca V 3.2 currents). Similar results are obtained with AEA, which inhibits ϳ85% of the three Ca V 3 currents (Fig. 7B).

DISCUSSION
In this study, we have identified the important chemical determinants of anandamide involved in the inhibition of Ca V 3 T-currents. We show that both the alkyl chain of anandamide and the hydroxyl present in its head group are primarily responsible for T-current inhibition. Indeed, anandamide derivatives lacking the hydroxyl group are weak inhibitors of T-currents (ϳ30%), as assessed with arachidonamide, arachidonic acid N-methyl amide, and arachidonyl 2Ј-chloroethylamide. In addition, the amide linkage also contributes to increase anandamide effects since arachidonic acid (that contains a carboxyl group and the same alkyl chain as anandamide but not the amide linkage) potently inhibits T-currents (ϳ60%) but to a lesser extent than anandamide (ϳ90%). Thus, arachidonic acid methyl ester, which lacks both the reactive hydroxyl group and the amide linkage, has no effect on T-currents. In addition, it should be noted that arachidonic alcohol has the same potency as arachidonic acid to inhibit T-currents, suggesting that the carbonyl group is not crucial for anandamide effects. Overall, these results suggest that anandamide and fatty acids could act on T-channels primarily using their -OH group, presumably via hydrogen bonds. However, acetyl ethanolamide and ethanolamine, which both lack the alkyl chain of anandamide, have no effect on T-currents, indicating that the fatty acid moiety of anandamide is necessary for the inhibition. Indeed, the inhibition of T-currents mainly depends on the degree of unsaturation of the alkyl chain but not on the chain length and increases with the number of cis-double bonds. Therefore the inhibition of T-currents is not restricted to eicosanoids (20 carbons) and also occurs with fatty acids and N-acyl ethanolamides containing 18 -22 carbons. Furthermore, inhibition is also increased for unsaturated fatty acids when cis-double bonds are close to the carboxyl group. By contrast, fatty acids in trans-configuration or containing cis-triple-bonds (as linolenelaidic acid and ETYA, respectively) have no effect. Therefore the major natural and endogenous fatty acids, including ␥-linolenic acid (18:3 6), mead acid (20:3 9), and arachidonic acid (20:4 6), as well as the fully polyunsaturated 3 fatty acids stearidonic acid (18: 4), EPA (20:5), and DHA (22:6) are potent inhibitors of T-currents (see Fig. 3). Similar results were obtained with the N-acyl related ethanolamides (see Fig. 4).
As observed in our study on Tchannels, most of the other voltagedependent anandamide-sensitive ionic channels are also modulated by fatty acids (reviewed in Ref. 27), including voltage-activated sodium channels (45)(46)(47)(48), K V potassium channels (49,50), and L-type calcium channels (51)(52)(53). By contrast, TRPV1 and TASK-1 channels, which are not voltage-sensitive, are modulated by anandamide but weakly by arachidonic acid (54 -57). In addition, anandamide and arachidonic acid activate (but not inhibit) voltage-insensitive BK potassium channels (58 -60). Also, by contrast with BK channel activation (58), inhibition of voltage-activated channels (including T-channels) occurs with unsaturated but not with the following saturated fatty acids: myristic (14:0), palmitic (16:0), and stearic acid (18:0) (45-47, 49, 51, 52). However, it is important to note that inhibition of T-channels by fatty acids displays specific features when compared with other voltage-activated anandamide-sensitive channels. T-channels are insensitive to long chain mono-unsaturated fatty acids such as palmitoleic (16:1) and oleic (18:1) acids, which potently inhibit L-type calcium channels (52). Furthermore, contrary to T-channel inhibition, the inhibition of L-type calcium channels does not increase with polyunsaturated fatty acids (51,52). In addition, the inhibition of sodium channels by polyunsaturated fatty acids comprising 18 -22 carbons is different from the T-channel inhibition since it does not increase with the number of double bonds and therefore linoleic acid Inactivation values were obtained by monoexponential fits. B, effects of AA on steady-state inactivation of Ca V 3.3 currents. Currents were elicited by a Ϫ30 mV test pulse (450-ms duration) applied from holding potentials ranged from Ϫ110 mV to Ϫ50 mV (10-s duration, 5-mV increments). C, effects of AA on current-voltage (I-V) curves of Ca V 3.3 currents. I-V curves were obtained from experiments illustrated in A. D, AA shifts the steady-state inactivation (hϱ) curve of Ca V 3.3 currents toward negative potentials without the corresponding effect on the steady-state activation (mϱ) curve. The steady-state inactivation (hϱ) curve was obtained from experiments illustrated in B, while the steady-state activation (mϱ) curve were deduced from the fit of the I-V curves presented in C. For clarity, the corresponding symbol of the mϱ curve was omitted. E and F, AA shifts the steady-state inactivation (hϱ) curves of Ca V 3.1 (E) and Ca V 3.2 currents (F) toward negative potentials without the corresponding effect on the steady-state activation (mϱ) curves. Steady-state activation (mϱ) and inactivation (hϱ) curves of Ca V 3.1 and Ca V 3.2 currents were deduced from the same experiments as in A and B except that the test pulses were of 200-ms duration.
(18:2) produces the same inhibitory effects than the fully polyunsaturated DHA (22:6) and EPA (22:5) (45,46). In contrast, inhibition of Kv channels occurs with polyunsaturated fatty acids containing four or more double bonds (as arachidonic acid and DHA) but not with linoleic (18:2) and linolenic fatty acid (18:3) (49), which weakly inhibit T-channels. Also, contrary to T-channels, inhibition of K V channels is maximal with ETYA (49), which contains four triple bonds. It should also be noted that TRPV1 is fully activated by N-acyl ethanolamides possessing at least one double bond and therefore linoleoyl ethanolamide, which weakly inhibits T-channels, is a full activator of TRPV1 (61). Overall, our structure-activity study of fatty acids on T-channels reveals unique features among anandamide-sensitive channels.
We also have demonstrated that the position of the double bonds in fatty acids is important for T-current inhibition. Inhibition is increased for unsaturated fatty acids when cis-double bonds are proximal to the carboxyl group. Therefore natural and endogenous fatty acids containing three double bonds near to the carboxyl group, such as ␥-linolenic acid (18:3 cis-6,9,12-octadecatrienoic acid) and mead acid (20:3 cis-5,8,11-eicosatrienoic acid), produce maximal inhibition equivalent to those obtained with the fully polyunsaturated fatty acids, including stearidonic acid (18:4), EPA (20:5), and DHA (22:6). This observation suggests that the fully polyunsaturated 3 fatty acids stearidonic acid, EPA, and DHA produce their effects not because they contain cis-double bonds in the 3 position but rather because they are fully polyunsaturated and therefore contain double bonds near to the carboxyl group. It should be noted that the importance of the position of the double bonds in 3 fatty acids has not been systematically tested on ionic currents, and most studies have concluded that stearidonic acid, EPA, and DHA produce maximal effect because they are 3 fatty acids, providing molecular explanation for their beneficial physiological effects, as in neuroprotection and cardioprotection (31, 34 -36, 40, 42). Furthermore, contrary to 3 and 6 fatty acids, which come from diet (from green plants, algae, and especially from fish oil) and are not endogenously produced in mammals, bio-active lipids acting on T-channels are synthesized in mammalian cells (which possess the unsaturase enzymes allowing double bond insertions near the carboxyl group, i.e. after the 6 position) and therefore are likely to operate as signal transduction molecules.
Fatty acids and N-acyl ethanolamides inhibiting T-channels possess at least a cis-1,4-pentadiene unit (™C¢C™C™C¢C™) and therefore are potentially metabolized in mammalian cells by several enzymes (see schematic signaling pathways for anandamide and arachidonic acid in Fig. 7A). We demonstrated with different enzyme inhibitors that the inhibition of the three T-channels by anandamide and arachidonic acid does not involve lipoxygenase, cyclooxygenase, as well as cytochrome P450 pathways. Furthermore, anandamide hydrolysis into arachidonic acid by FAAH is not involved in anandamide effects since T-current inhibition is resistant to phenylmethanesulfonyl fluoride and also occurs with R1-and R2-methanandamide, which are resistant to FAAH hydrolysis. We also demonstrated that anandamide and arachidonic acid inhibit the three Ca V 3 T-channels in cell-free inside-out patches, suggesting that both anandamide and arachidonic acid exert their action by binding directly to T-channels or by acting on their near lipid environment. For instance, it has been reported that fatty acids possibly modulate BK and TREK-1 potassium channels by acting directly on the membrane, as these channels are sensitive to membrane stretch (57)(58)(59). It should be noted, however, that in these two later cases, fatty acids and membrane stretch induce activation rather than inhibition of the channels. We have tested the membrane hypothesis by mimicking the membrane effects of arachidonic acid with unrelated molecules trinitrophenol and dinitriphenol that also are anionic amphipathic molecules (62,63). These molecules, like arachidonic acid, have been previously shown to cause membrane crenation and opening of the mechano-sensitive TREK-1 channels (57,62,63). We found that these compounds had very weak effects on T-currents (ϳ5% increase) even at a concentration of 400 M (data not shown). Furthermore, a hypo-osmotic solution (ϳ150 mosM), which also mimics arachidonic acid crenator effects on cell membrane and TREK-1 channels (57), induced a slight increase of T-currents (ϳ20%, data not shown) but no inhibition. In addition, by contrast with our data on T-channel inhibition, membrane effects of fatty acids mainly depend on the chain length and the degree of unsaturation rather than the position of double bonds. Although we cannot completely exclude a membrane effect of anandamide and fatty acids on T-channels, our results are overall in favor of a direct binding of these molecules to T-channels.
We have previously shown that anandamide accelerates the inactivation kinetics and shifts the steady-state inactivation curve of the three Ca V 3 currents toward more negative potentials (ϳ10 mV) (24). We demonstrated in this study that arachidonic acid induces similar changes in T-current properties, confirming a common molecular mechanism (64 -66). Also, similar effects were obtained with polyunsaturated fatty acids on inactivation properties of other anandamide-sensitive voltage-dependent channels such as sodium channels (45)(46)(47), K V potassium channels (49,50), and L-type calcium channels (51,52), indicating that the mechanism of inhibition by anandamide and fatty acids might be conserved among voltage-dependent channels.

CONCLUSIONS
Polyunsaturated fatty acids and N-acyl ethanolamides are signaling molecules implicated in a multitude of physiological and pathophysiological events, including cardiac and neuronal excitability (31,33,40,42), cardioprotection and neuroprotection from ischemic and epileptic episodes (31, 34 -36, 40), inflammation, and pain (32,39). Considering the role of T-channels in cardiac and neuronal pacemaker (1,2,13,14), cardiac hypertrophy (15), neuroprotection, and absence epilepsy (8,9,67) as well as in pain perception (10 -12), it is attractive to suggest that T-channel inhibition may contribute to polyunsaturated fatty acid and N-acyl ethanolamide effects. In addition, it is important to note that both anandamide and arachidonic acid also act as intracellular second messengers (68,69) and therefore could potentially participate in the intracellular signal transduction cascades by which neurotransmitters inhibit T-currents (2,70). It will be of great interest to investigate whether endogenously produced anandamide and arachidonic acid can modulate T-currents as recently demonstrated for TRPV1 receptors (69).