Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

⁹-THC is the major psychoactive component in marijuana. There are two known cannabinoid receptors: CB1, primarily in the central nervous system (Felder et al., 1993), and its amino-terminal variant, the CB1A receptor (Shire et al., 1995); and the CB2 receptor found on cells of the immune system (Munro et al., 1993). ⁹-THC produces psychoactive effects through binding to CB1 receptors (Ledent et al., 1999; Zimmer et al., 1999; Buckley et al., 2000) that have been cloned (Matsuda et al., 1990; Gerard et al., 1991; Munro et al., 1993). CB1 and CB2 receptors have specific antagonists (Rinaldi-Carmona et al., 1994, 1998).

CB1 and CB2 receptors are G protein-coupled receptors (GPCRs) linked to a G_i/o protein, which when activated inhibits the activity of adenylyl cyclase (Howlett and Fleming, 1984). Upon agonist binding, the subunit dissociates from the  subunit of the G_i/o protein (Childers and Deadwyler, 1996). The  subunit inhibits adenylyl cyclase, whereas the subunit has been linked to activation of other cellular events, such as activation of tyrosine kinases (TKs).

Tolerance develops to the in vivo and in vitro pharmacological effects of cannabinoids (Martin, 1985; Dill and Howlett, 1988; Mason and Welch, 1999). Receptor down-regulation is a possible mechanism of ⁹-THC tolerance (Oviedo et al., 1993; Rodriguez de Fonseca et al., 1994). Studies indicate that the cannabinoid receptor is rapidly internalized after binding of an agonist (Hsieh et al., 1999). However, Abood et al. (1993), found no alterations in cannabinoid receptor mRNA or protein levels in mouse whole brain homogenates after a chronic injection paradigm. Thus, the effects of long-term administration of ⁹-THC on receptor down-regulation are unclear. It is proposed that phosphorylation enhances the down-regulation of the CB1 receptor. We hypothesized that modification of intracellular phosphorylations with several kinase inhibitors might attenuate the expression of THC-induced tolerance.

cAMP-Dependent Protein Kinase (PKA). Acute administration of ⁹-THC decreases cAMP formation by inhibiting adenylyl cyclase and decreases PKA activity. Conversely, chronic cannabinoid exposure enhances the adenylyl cyclase activity, and increases cAMP levels and PKA activity in the same areas that CB1 receptor down-regulation is observed (i.e., cerebellum, striatum, and cortex) (Rubino et al., 2000). Thus, the adenylyl cyclase cascade seems to become constitutively active during tolerance.

PKG. The cannabinoid levonantradol, but not dextronantradol, decreases basal and isoniazid-induced increases in cGMP (Leader et al., 1981). Thus, cannabinoids could alter cyclic-GMP formation (and thus PKG activity) in tolerance expression.

Protein Kinase C (PKC). ⁹-THC increases the activity of brain PKC in vitro (Hillard and Auchampach, 1994; De Petrocellis et al., 1995). PKC seems to directly affect CB1 receptors. Phosphorylation of the CB1 receptor with PKC suppresses the modulation of calcium channels by cannabinoids (Garcia et al., 1998). Neurotransmitters that activate PKC restore the neuronal excitability and synaptic activity inhibited by cannabinoids. We hypothesized that PKC inhibitors might reverse ⁹-THC tolerance.

TKs. CB1 receptor activation of the subunit of G proteins can stimulate TKs. One target of activation by the subunit is Src tyrosine kinase that has been shown to activate Ras, activating mitogen-activated protein kinase. CB1 and CB2 stimulation increases the activation of MAPK (Rinaldi-Carmona et al., 1998), which becomes tyrosine-phosphorylated in cannabinoid-treated cells, an effect blocked by TK inhibitors (Bouaboula et al., 1995). We tested the Src tyrosine kinase inhibitor PP1 (Daub et al., 1997) in mice for its effects on THC-induced antinociception and reversal of tolerance.

PI3-K. PI3-K is an early intermediate of the G-mediated MAPK signaling pathway (Daub et al., 1997). Therefore, we proposed to block PI3-K and reduce the subunit-mediated tyrosine phosphorylation of MAPK.

G Protein-Coupled Receptor Kinase (GRK). -Adrenergic receptor desensitization involves rapid PKA and GRK phosphorylation. GRK phosphorylation in turn promotes -Arrestin binding and receptor internalization (Seibold et al., 1998). We inhibited -adrenergic receptor kinase (-ARK), a type of GRK, with low molecular weight heparin (LMWH) to evaluate the chronic affects of ⁹-THC after inhibition of -ARK.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animal Model of ⁹-THC Tolerance. All studies using the tail-flick test were performed on male ICR mice. The mice were kept on a 12-h light/dark cycle and received food and water ad libitum. In the acute studies, mice weighed 16 to 25 g; in chronic studies, mice weighed 25 to 34 g upon testing. Mice were rendered tolerant to ⁹-THC over 7 days. The mice received twice daily s.c. injections of ⁹-THC (20 mg/kg) for 6 days and on day 7 just received the morning dose. On the morning of day 8, mice were challenged with an ED₈₀ dose (i.t.) of ⁹-THC for determination of tolerance. Rats were used for the spinal cord release of dynorphin to obtain sufficient cerebrospinal fluid for testing. Male Sprague-Dawley rats, weighing between 350 and 400 g, obtained from Harlan (Indianapolis, IN), were housed in plastic cages, two rats per cage, and maintained on a fixed 12-h light cycle at a temperature of 22 ± 2°C. Water and food (Harlan Rat Chow) were provided ad libitum. Rats were rendered tolerant to ⁹-THC using the doses and time course as for mice and were also challenged on day 8 with the ED₈₀ dose of ⁹-THC (i.t.) for tolerance determination.

Intrathecal Injections. Intrathecal injections were performed in mice following the protocol of Hylden and Wilcox (1983). Unanesthetized mice were injected with 5 µl of drug between the L5 and L6 area of the spinal cord with a 30-gauge needle. In studies using rats, the pentobarbital-anesthetized rats were placed in stereotaxis, and an incision was made on the atlanto-occipital membrane to expose the cisterna magna. A catheter of polyethylene-10 tubing was inserted through the exposed cisternal cavity, caudally, into the subarachnoid space of the spinal cord. The catheter contained an artificial cerebrospinal fluid, composed of 125 mM Na⁺, 2.6 mM K⁺, 0.9 mM Mg²⁺, 1.3 mM Ca²⁺, 122.7 mM Cl, 21.0 mM HOC, 2.4 mM HOP, 0.5 mg/ml bovine serum albumin, bacitracin (30 mg/ml), and 0.01% Triton X, and effervesced with 95% O₂ and 5% CO₂. Positioned as such, the catheter extended caudally 8.5 cm passing through the thoracolumbar region to an area just above the sacral enlargement. After catheter implantation, animals were allowed to acclimate approximately 30 min on a heating pad. After acclimation, baseline tail-flick latency was assessed. Only animals exhibiting normal tail-flick response to noxious stimuli greater than 1.5-s but less than 4-s latency were used to avoid use of any rat with spinal cord damage. Drug or vehicle (1:1:18; ethanol/Emulphor/saline) was administered in a 20-µl bolus over 1 min using a peristaltic pump.

Measurement of Dynorphin. Measurement of dynorphin A(1-17) was accomplished using a dynorphin A(1-17)-specific radioimmunoassay kit obtained from Peninsula Laboratories (Belmont, CA). The reconstituted samples were analyzed in duplicate. The manufacturer reports cross-reactivity of dynorphin A(1-17) antibody as 100% versus dynorphin A(1-24), a parent compound, and less than 2% versus smaller peptide fragments. We found no cross-reactivity of the antibody to dynorphin A(1-8), dynorphin A(1-13), dynorphin B, anandamide, or morphine. Only the linear portion of the radioimmunoassay standard curve, between 0.1 and 64 pg/ml of the standard dynorphin peptide, was used to calculate dynorphin concentration. The cerebrospinal fluid from individual rats was analyzed for dynorphin concentrations using at least six rats per test group. The rats were evaluated in the tail-flick test before the removal of cerebrospinal fluid for testing. Thus, the behavioral effects of each rat can be compared with the dynorphin levels in that individual rat's spinal fluid.

Tail-Flick Test. Mice and rats were tested for antinociception by the tail-flick procedure (D'Amour and Smith, 1941). Reaction times of 1.5 to 4 s were used for the control, whereas a time of 10 s was used for the cutoff to prevent tissue damage. Antinociception was quantified as the percentage of maximum possible effect (% MPE) formula: % MPE = 100 × [(test  control)/(10  control)] (Harris and Pierson, 1964). % MPE values were calculated for each animal, using at least six animals per dose, for which mean effect and S.E.M. were calculated for each dose. At least three doses of each test drug or combination of drugs were used to generate dose-response curves.

Materials. Doses for all drugs used were predetermined in naïve animals using the maximal dose without toxicity. Time points were determined in naïve animals to ascertain the point at which each drug had its peak effect. Dimethyl sulfoxide (100%, DMSO) was purchased from Sigma-Aldrich (St. Louis, MO). KT5720, purchased from Calbiochem (La Jolla, CA), was prepared in 100% DMSO and was injected i.t. at a dose of 2.7 µg/mouse 15 min before drug or vehicle (i.t.). The tail-flick test was then conducted 15 min after the second injection. KT5823, purchased from Calbiochem, was prepared in 100% DMSO and injected i.t. at a dose of 2.5 µg/mouse 15 min before drug or vehicle (i.t.). The tail-flick test was then conducted 15 min after the second injection. Dibutyryl-cAMP (10 µg/mouse) and dibutyryl-cGMP (5 µg/mouse) were purchased from Calbiochem and were prepared in distilled water (dH₂O) and injected i.t 15 min before the i.t. injection of drug or vehicle. Fifteen minutes later, the tail-flick test was conducted. ⁹-THC was obtained from the National Institute on Drug Abuse and was prepared in 100% DMSO for acute tests and in 1 part ethyl alcohol (Aaper Alcohol and Chemical, Shelbyville, KY), 1 part Emulphor, and 18 parts 0.9% normal saline (Baxter, Deerfield, IL) (1:1:18) for tolerance studies. 2-(4-Morpholinyl)-8-phenyl-4H-1-benzopyran-4-one (LY294002) was purchased from BIOMOL Research Laboratories (Plymouth Meeting, PA) and was prepared in 100% DMSO and injected i.t. 15 min before drug or vehicle (i.t.). The tail-flick test was then conducted 15 min after the second injection. Bisindolymaleimide I, HCl (bis; Calbiochem) was prepared in dH₂O and injected i.t. (5 µg/mouse) or (0.5 µg/mouse) 15 min before drug or vehicle (i.t.). The tail-flick test was then conducted 15 min after the second injection. LMWH was purchased from Sigma-Aldrich and was prepared in dH₂O and injected i.t. (30 µg/mouse) 15 min preceding the i.t. injection of drug or vehicle. The tail-flick test was then conducted 15 min after the second injection. PP1 purchased from Alexis Corporation (Läufelfingen, Switzerland) was prepared in 100% DMSO and injected i.t. 10 min before the i.t. injection of drug or vehicle, with the administration of the tail-flick test 15 min after the second i.t. injection.

Statistical Analysis. Analysis of variance was used to determine significant differences between control and treatment animal groups followed by Dunnett's t test. These calculations were performed using StatView, version 512+ (BrainPower, Inc., Agoura Hills, CA). p values of less than 0.05 were deemed significant. Parallelism of the dose-response curves was determined by the methods of Tallarida and Murray (1987). Potency ratios were determined using the methods of Colquhoun (1971).

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

The i.t. administration of KT5720, a PKA inhibitor, at a dose of 2.7 µg/mouse in 100% DMSO vehicle (i.t.), significantly (p < 0.05) reversed ⁹-THC antinociceptive tolerance in a dose-dependent manner, as determined by the tail-flick test. There was a leftward shift of the dose-response curve. The ED₅₀ in the ⁹-THC-tolerant mice was shifted from 80 µg/mouse (95% confidence limits, 62-102) to 8.6 µg/mouse (95% confidence limits, 4.7-16) in the KT5720-treated mice. The lines were parallel and the potency ratio was 9.3 (Fig. 1).

View larger version (12K): [in this window] [in a new window]   Fig. 1.   Reversal of 9-THC-induced tolerance by the PKA inhibitor KT5720. The graph shows a significant rightward shift in the 9-THC (i.t.) dose-effect curve, indicating the development of tolerance in mice pretreated chronically with 9-THC, as described under Materials and Methods versus those pretreated chronically with vehicle. Also indicated is a significant leftward shift of the dose-effect curve of 9-THC (i.t.) in mice chronically treated with 9-THC and pretreated with KT5720 (2.7 µg/mouse, i.c.v.) before testing. KT5720 did not shift the 9-THC dose-effect curve in chronic vehicle-pretreated (nontolerant) mice. Each point represents an n of at least eight mice per dose. The effects of i.c.v. administration of 100% DMSO (5 µl/mouse) alone in the tolerant and nontolerant mice were less than 20% MPE. , nontolerant mice receiving vehicle (i.c.v. DMSO) + 9-THC (i.t.); , nontolerant mice receiving KT5720 (i.c.v.) + 9-THC (i.t.); , tolerant mice receiving vehicle (i.c.v. DMSO) + 9-THC (i.t.); and , tolerant mice receiving KT5720 (i.c.v.) + 9-THC (i.t.).

The PKG inhibitor KT5823, at a dose of 2.5 µg/mouse in 100% DMSO vehicle (i.t.), had no effect on ⁹-THC antinociceptive tolerance (2% MPE in the tolerant mice compared with 7% in the tolerant animals treated with KT5823; Fig. 2 A). A higher dose of KT5823 (5 µg/mouse) had no greater effect. Pretreatment with 10 µg/mouse produced lethality.

View larger version (24K): [in this window] [in a new window]   Fig. 2.   Lack of reversal of 9-THC-induced tolerance by various inhibitors KT5823, bis, LY294002 (LY), and LMWH. Mice were pretreated chronically with 9-THC or vehicle, as described under Materials and Methods. Mice were evaluated for tolerance to 9-THC via i.t. administration of 9-THC (i.t., 20 µg/mouse) + DMSO (i.c.v.). A, mice were tested for reversal of tolerance via a challenge dose of 9-THC (20 µg/mouse, i.t.) + KT5823 (2.5 µg/mouse, i.c.v.). B, mice were tested for reversal of tolerance via a challenge dose of 9-THC (20 µg/mouse, i.t.) + bis (0.5 µg/mouse, i.c.v.). C, mice were tested for reversal of tolerance via a challenge dose of 9-THC (20 µg/mouse, i.t.) + LY (20 µg/mouse, i.c.v.). D, mice were tested for reversal of tolerance via a challenge dose of 9-THC (20 µg/mouse, i.t.) + LMWH (30 µg/mouse, i.c.v.). The effects of i.c.v. administration of 100% DMSO (5 µl/mouse) + veh (i.t.) (veh + veh) alone in the tolerant and nontolerant mice were less than 20% MPE in all cases (A-D). [inhibitor (i.c.v.) + veh (i.t.)] produced less than 25% MPE in both tolerant and nontolerant mice in all cases (A-D). In nontolerant mice veh + 9-THC produced nearly 100% MPE and was significantly reduced to less than 15% MPE in 9-THC-tolerant mice (*, p < 0.01, A-D). Inhibitor + 9-THC in nontolerant mice was 98% MPE or greater and was reduced to less than 22% MPE in tolerant mice (*, p < 0.01). Thus, all the inhibitors in Fig. 2 were inactive in that they failed to reverse tolerance to 9-THC. , nontolerant mice; and , tolerant mice.

The PKC inhibitor bis, at a dose of 0.5 µg/mouse administered i.t., did not affect the antinociceptive tolerance in mice. The % MPEs in the tolerant groups treated with bis compared with vehicle-treated mice were not significantly different (20 ± 15 versus 14 ± 6.0, respectively) (Fig. 2B). At an increased dose of 5 µg/mouse, there was not a significant shift in the ED₅₀ values of the tolerant mice treated with bis compared with tolerant mice treated with vehicle [41 (32-51) versus 80 (50-129); Fig. 3], although clearly there was a trend toward a rightward shift in the dose-effect curve. The dose-effect curves for mice pretreated with bis versus those pretreated with vehicle were not parallel in the tolerant mice. Interestingly, in the nontolerant mice, there was an attenuation of the antinociceptive effect of ⁹-THC. There was a significant (p < 0.05) rightward shift in the dose-response curve of ⁹-THC. The ED₅₀ was shifted from 7 (5-11) in the vehicle-treated nontolerant animal to 26 (16-45) in the bis-treated nontolerant animal. The dose-effect curves were parallel and the potency ratio was 3.6. 

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Attenuation of 9-THC-induced antinociception by the PKC inhibitor bis (5 µg/mouse, i.c.v.). The graph shows a significant rightward shift in the 9-THC (i.t.) dose-effect curve, indicating the development of tolerance in mice pretreated chronically with 9-THC, as described under Materials and Methods versus those pretreated chronically with vehicle. Also indicated is a significant leftward shift of the dose-effect curve of 9-THC (i.t.) in mice chronically treated with vehicle and pretreated with bis (5 µg/mouse, i.c.v.) before testing. bis did not shift the 9-THC dose-effect curve significantly in chronically 9-THC-pretreated (tolerant) mice, although there is a trend toward attenuation of the effects of 9-THC. Each point represents an n of at least eight mice per dose. The effects of i.c.v. administration of 100% DMSO (5 µl/mouse) alone in the tolerant and nontolerant mice were less than 20% MPE. , nontolerant mice receiving vehicle (i.c.v. DMSO) + 9-THC (i.t.); , nontolerant mice receiving bis (5 µg/mouse, i.c.v.) + 9-THC (i.t.); , tolerant mice receiving vehicle (i.c.v. DMSO) + 9-THC (i.t.); and , tolerant mice receiving bis (5 µg/mouse, i.c.v.) + 9-THC (i.t.).

LY294002, a phosphatidylinositol-3-kinase inhibitor, administered i.t. in 100% DMSO vehicle at a dose of 0.1 µg/mouse, did not significantly alter ⁹-THC antinociceptive tolerance in mice (Fig. 2C). The LY294002-treated tolerant mice had a % MPE of 10 ± 10 compared with the tolerant vehicle-treated mice who had a % MPE of 2 ± 1. At higher doses, LY294002 produced an antinociceptive effect that confounded use for tolerance reversal studies.

LMWH, which inhibits -ARK, at a dose of 30 µg/mouse administered i.t., did not affect the antinociceptive tolerance in the mice. The % MPE in the ⁹-THC-tolerant group treated with LMWH (5 ± 2) was not significantly different from the vehicle-treated THC-tolerant group (14 ± 6.0) (Fig. 2D).

We also evaluated PP1, a Src family tyrosine kinase inhibitor. Because the subunit of the cannabinoid receptor interacts with tyrosine kinase to activate MAPK, we wanted to determine what would happen if this pathway was disrupted. At a dose of 0.0001 µg/mouse, in 100% DMSO vehicle administered i.t., PP1 significantly (p < 0.05) reversed ⁹-THC antinociceptive tolerance in mice. The 0.0001 µg/mouse dose was shown to be inactive (% MPE 4 ± 1) in the tail-flick test in naïve mice, and in the nontolerant group the % MPE (44 ± 20) did not differ from that of vehicle-pretreated mice (% MPE, 45 ± 19) (Fig. 4). At doses of 0.001 µg/mouse and higher, PP1 shows a variable antinociceptive affect that confounded studies at the higher doses.

View larger version (12K): [in this window] [in a new window]   Fig. 4.   Reversal of 9-THC-induced tolerance by the TK inhibitor PP1. Mice were pretreated chronically with 9-THC or vehicle, as described under Materials and Methods. Mice were evaluated for tolerance to 9-THC via i.t. administration of 9-THC (20 µg/mouse, i.t) + DMSO (i.c.v.). Alternatively, mice were tested for reversal of tolerance via a challenge dose of 9-THC (20 µg/mouse, i.t.) + PP1 (0.0001 µg/mouse, i.c.v.). Each point represents an n of at least eight mice per dose. The effects of i.c.v. administration of 100% DMSO (5 µl/mouse) + veh (i.t.) (veh + veh) alone in the tolerant and nontolerant mice were less than 30% MPE. [PP1 (i.c.v.) + veh (i.t.)] produced less than 30% MPE in both tolerant and nontolerant mice. In nontolerant mice veh + 9-THC produced 100% MPE and was significantly reduced to 2% MPE in 9-THC-tolerant mice (*, p < 0.01). PP1 + 9-THC in nontolerant mice was 93% MPE and remained 93% MPE in tolerant mice. That is, PP1 reversed tolerance to 9-THC. , nontolerant mice; and , tolerant mice.

We also evaluated enhancement of tolerance. Because PKA inhibition reversed tolerance, we evaluated the ability of a cAMP analog to enhance tolerance. The challenge dose of ⁹-THC was increased to 100 µg/mouse i.t. to get approximately 50% MPE in the tolerant mice. Dibutyryl cyclic-GMP at 5 µg/mouse administered i.t. did not significantly enhance tolerance (36 ± 20% MPE in the tolerant animals compared with 46 ± 23% MPE in the db-cGMP-treated animals) (data not shown). Higher doses (30 and 50 µg/mouse) of db-cGMP had intrinsic antinociceptive effects. Dibutyryl cyclic-AMP at doses of 10 to 50 µg/mouse administered i.t. also did not enhance ⁹-THC antinociceptive tolerance (55 ± 14% MPE in the tolerant animals compared with 46 ± 23% MPE in the db-cAMP-treated animals for the 10 µg/mouse group). db-cAMP did not produce intrinsic antinociception at the doses tested.

A study was performed to address the hypothesis that KT5720 reversal of tolerance might restore dynorphin release by 300 µg/rat ⁹-THC to levels observed in nontolerant rats. Doses and time points for KT5720 administration and tail-flick testing were as those in the mouse. Figure 5A indicates THC-stimulated dynorphin release was significantly depressed in THC-tolerant rats (25 ± 7 pg/ml in non-THC-tolerant rats versus 6 ± 3 pg/ml in THC-tolerant rats). Administration of KT5720 before ⁹-THC did not alter dynorphin release in nontolerant rats (22 ± 6 pg/ml) compared with the nontolerant ⁹-THC only group. However, KT 5720 significantly reversed the ⁹-THC-stimulated decrease in dynorphin observed in ⁹-THC-tolerant rats (levels of dynorphin were raised to 32 ± 8 pg/ml). Figure 5B shows the behavioral response (tail-flick test) in the same rats. The rats were tolerant to ⁹-THC as indicated by "a". KT5720 significantly reversed tolerance as indicated by "b". Thus, as the behavioral response returned to nontolerant levels, the release of dynorphin increased to nontolerant levels.

View larger version (30K): [in this window] [in a new window]   Fig. 5.   Reversal of tolerance to 9-THC by KT5720 in the rat restores dynorphin A release. A study was performed to address the hypothesis that KT5720 reversal of tolerance might restore dynorphin release by 300 µg/rat 9-THC to levels observed in nontolerant rats. Doses and time points for KT5720 (KT) administration and tail-flick testing were as those in the mouse. A, dynorphin release (picograms per milliliter ± S.E.M.). B, behavioral response (tail-flick test, average % MPE ± S.E.M.) in the same rats. As the behavioral response returned to nontolerant levels, the release of dynorphin increased to nontolerant levels. n = 6 rats/group. In A, a indicates p < 0.05 from "THC-nontolerant" group and b indicates p < 0.05 from "THC-tolerant" group. In B, a indicates p < 0.05 from "THC-nontolerant" group and b indicates p < 0.05 from "THC-tolerant" group.

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

We sought to address the role of various kinases in ⁹-THC antinociceptive tolerance. We evaluated kinases that were downstream from the cannabinoid receptor (PKA, PI3-K, and TK), that may interact directly with the receptor (PKA, -ARK, and PKC), and others that act in different pathways (PKC and PKG). When a ligand binds to a GPCR, as the cannabinoid receptor, there is a decreased affinity between the  and subunits of the G protein, and they separate from one another. In the acute model of ⁹-THC exposure, the  subunit will produce a decrease in adenylate cyclase, decreases in cAMP, and decreases in PKA activation. There is also an associated opening of low-voltage potassium channels, leading to an efflux of potassium, and a modulation of calcium channels, leading to decreased calcium conductance. In animals chronically treated with THC, there is a compensatory increase in adenylate cyclase, cAMP, and PKA activation. The intrathecal administration of the protein kinase A inhibitor KT5720 reversed the antinociceptive tolerance to ⁹-THC in both mice and rats. Of significance, inhibition of PKA also reversed a biochemical correlate of tolerance, namely, dynorphin release. Thus, our data indicate that protein kinase A plays a role in the mechanism of ⁹-THC antinociceptive tolerance. The role of PKA in THC-induced tolerance is unknown. However, several possibilities for the site of action of PKA exist. PKA could be responsible for phosphorylating the CB1 receptor upon binding of a ligand to the receptor. PKA also could be increasing potassium conductance through phosphorylation of the potassium channel, causing the cell to become hyperpolarized. Other possible roles of PKA in THC-mediated tolerance include the possibility that PKA is rapidly and continuously phosphorylating the CB1 receptor if and when it is down-regulated into the cytosol in tolerant animals. The CB1 receptor seems to be rapidly internalized upon exposure to cannabinoids (Hsieh et al., 1999). Once the cell is "tolerant", we propose that there will be a compensatory increase in production of PKA. Higher levels of PKA could be responsible for a continuous phosphorylation of the CB1 receptor while in the cytosol. This continued phosphorylation might facilitate the receptor down-regulation. Upon inhibition of PKA by KT5720, the receptor would in theory no longer remain phosphorylated and could therefore be recycled to the membrane where it would be capable of binding to the ligand again. If the phosphorylation of the receptor was maintained, and the receptor remained down-regulated, eventually it would be degraded, requiring mRNA for new protein synthesis.

Because PKA inhibition reverses cannabinoid antinociceptive tolerance, we hypothesized that a cAMP analog might enhance tolerance. In a nontolerant neuron exposed to cannabinoids, cAMP is decreased, but in the presence of forskolin, which increases cAMP, or Cl-cAMP, a cAMP analog, antinociception is attenuated (Cook et al., 1995). However, dibutyryl-cAMP did not enhance tolerance. Perhaps we were not able to increase levels of c-AMP to a level consistent with enhancement of tolerance.

We also evaluated kinases involved in the G-mediated signaling pathway. PI-3 kinase and tyrosine kinase work downstream from the subunit of the GPCR, and these kinases are generally associated with growth and differentiation. With the membrane-destabilizing activity of cannabinoids and release of free arachidonic acid, such kinases might play a role in antinociception. LY294002 is a specific PI3-K inhibitor. PI3-K is an enzyme implicated in growth factor signal transduction by associating with receptor and nonreceptor tyrosine kinases (Vlahos et al., 1994). Pertussis-sensitive GPCRs and TKs may converge or share a common pathway upstream from the activation of MAPK. Our goal was to determine whether by blocking a kinase or kinases in the pathway leading to MAPK activation, we could reverse tolerance. Upon blocking PI3-K, tolerance was not affected. One caveat in our results was that we could not use higher doses of LY294002 due to its intrinsic analgetic effects. Thus, our results with PI3-K blockade are inconclusive. However, the blockade of the Src family tyrosine kinase with PP1 reversed tolerance. Thus, by blocking a tyrosine kinase, we may be inhibiting downstream actions of the subunit. The subunit may be necessary to maintain a tolerant state by activation of MAPK. Further studies need to be conducted looking for the role of the subunit and tyrosine kinases and their role in central cannabinoid effects.

In addition to kinases downstream to the  and subunits, we also tested other kinases involved in a variety of cellular processes. -ARK is known to phosphorylate the ₂-adrenergic receptor and is a potential candidate for phosphorylation of the cannabinoid receptor before internalization. Hsieh et al. (1999) noted that the CB1 receptor is internalized after a pathway grossly similar to the one used by the ₂-adrenergic receptor. If such were the case, blocking -ARK with LMWH might be expected to prevent receptor phosphorylation and possibly desensitization or down-regulation and reverse tolerance. Because LMWH did not reverse tolerance, it seems that the cannabinoid receptor is may not be phosphorylated by -ARK. However, due to solubility and toxicity issues we may not have been able to increase the dose of LMWH to levels needed to block -ARK.

PKC may act directly on the CB1 receptor and/or downstream from the receptor. It has been shown that cannabinoids increase brain protein kinase C activity in vitro (Hillard and Auchampach, 1994). Our data indicate that using two different doses of the PKC inhibitor did not alter tolerance to cannabinoids. However, we did observe at the higher dose of PKC inhibitor that the effects of ⁹-THC in nontolerant animals were significantly attenuated. Hillard and Auchampach (1994) showed that cannabinoids increase the levels of PKC in rat brain and that these increased levels are responsible for reestablishing neuronal excitability. Because inhibiting PKC attenuates the effects of cannabinoid-induced antinociception, it is likely that increased levels of PKC may be at least partially responsible for cannabinoid-induced antinociception.

In summary, the data presented indicate that by inhibiting PKA and Src tyrosine kinase, ⁹-THC antinociceptive tolerance can be reversed. It seems likely that these two kinases work independently of one another. The other kinase inhibitors for PKG, PKC, PI3-K, and -ARK, did not alter tolerance at the doses testable. One has to take these negative data in the context of the possibility that the drug did not achieve high enough levels in the whole animal to inhibit the kinases. Thus, positive data as with PKA and Src TK indicate potential sites for ⁹-THC modulation, whereas negative data remain rather inconclusive. However, the higher dose of PKC inhibitor was shown to attenuate the antinociceptive effects of ⁹-THC in the nontolerant mice, which indicates that the inhibitor very likely reaches its site of action at a concentration that is active. However, although negative results must be interpreted with caution, positive results also need to be interpreted with caution as to the site of action of the drugs used. Our data indicate that the PKA and tyrosine kinase inhibition have prominent roles in tolerance to cannabinoids. The drugs used to inhibit such kinases (KT5720 and PP1, respectively) are the most selective drugs available. Such inhibitors likely act on PKA and TK at various sites intracellularly. It is intriguing that reversal of tolerance is a rapid process and occurs using drugs that do not alter the acute effects of THC. A similar effect has been shown for KT5720-induced reversal of tolerance to morphine (Bernstein and Welch, 1998). It takes several days to develop tolerance. The ability to reverse tolerance so rapidly is surprising and opens the door to a plethora of questions to be answered as to the plasticity of the neuronal system during the tolerance process. Certainly, the ability to reverse tolerance to any drug has profound clinical implications. A future direction would be to evaluate the phosphorylation state of the receptor. If PKA were responsible for the initial desensitization or maintaining the down-regulated state of the receptor, we would expect to see the receptor in the phosphorylated state in tolerant animals. We would also expect to see a dephosphorylated receptor in tissue that had been treated with the PKA inhibitor immediately before harvest. Additionally, and of greater clinical importance, will be to determine the duration of the reversal of tolerance with the kinase inhibitors.