Role of PLC in morphine tolerance
To our knowledge, this is the first report demonstrating the reversal of morphine toleerance by selective inhibitors of PLC. ET-18-OCH3 is a highly selective ether lipid analogue inhibitor of PtdIns-PLC, which converts PtdIns (4,5)P2 into equimolar concentrations of DAG and IP3. ET-18-OCH3 does not antagonize PtdCholine-PLC or phospholipase D (for a review see Powis & Phil, 1994). It is thought that ET-18-OCH3 disturbs phospholipid metabolism by accumulating in the membrane (Helfman et al., 1983). Our results indicate that both products of hydrolysis, IP3 and DAG, contribute to the expression of tolerance (Figure 1a, Table 1). It is noteworthy that antagonists of metabotropic glutamate receptors mGluR1 and mGluR2, positively linked to activation of the PtdIns system, attenuated naloxone precipitated withdrawal in rats (Fundytus et al., 1997). Thus, both IP3 and DAG may be important in the maintenance of opioid tolerance and physical dependence. The results with heparin, as discussed later, indicates a role for IP3-sensitive Ca2+ pools in morphine tolerance. In addition, DAG production leads to the translocation and activation of PKC, which appears from this study to contribute to tolerance.
The possibility that PKC activation contributes to morphine tolerance was supported by the reversal of tolerance with D609, a PtdCholine-PLC inhibitor (Figure 1b, Table 1). The hydrolysis of PtdCholine by PtdCholine-PLC provides another source of DAG, without concomitant production of IP3. Stimulation of this pathway leads to the translocation and activation of PKC (Billah & Anthes, 1990; Dennis et al., 1991). Therefore, the results with both ET-18OCH3 and D609 indicate that phospholipid metabolism may be more active in neurons crucial to the expression of opioid tolerance. As a consequence, it is predicted that both PKC activity and the levels of IP3 would also be increased, as discussed later.
It remains to be revealed in which brain regions the PtdIns and PtdCholine systems may be affected, since all agents were administered i.c.v. The periaqueductal gray (PAG), lining the aqueduct between the third and fourth ventricle, is readily accessible to drugs injected into the lateral ventricles of the brain. This region, besides possessing mu-opioid receptors and receiving β-endorphin-containing terminals from cell bodies in the arcuate nucleus, plays an important role in modulating the activity of bulbospinal monoaminergic antinociceptive systems (Aston-Jones et al., 1991; Reichling et al., 1988). These descending pathways have been shown to inhibit the transmission of nociception at the level of the spinal dorsal horn. Mu-opioid agonists injected directly into the PAG elicits not only dose-dependent antinociception, but significantly reduces the levels of intra-neuronal Ca2+ (Zhang et al., 1992). Alternatively, phorbol ester injection i.c.v. antagonizes opioid antinociception (Narita et al., 1997; Ohsawa & Kamei, 1997). Since the PAG is readily accessible to inhibitors of the phospholipid pathway, it is tempting to speculate that this region may be affected in morphine tolerance.
The issue of whether a stress component from free-hand i.c.v. injection could be modulated by these inhibitors should be addressed. It cannot be completely ruled out that i.c.v. injections are somewhat stressful. Recent evidence indicates that i.c.v. injection of saline in awake mice may be mildly stressful, since the slight increase in hot-plate latency threshold was blocked by nociceptin/Orphanin FQ, an endogenous ligand of opioid receptor like 1 receptors (Suaudeau et al., 1998). However, handling and injection stress was minimized by anaesthetizing the animals, as required by IACUC procedures. Furthermore, the direct influence of these phospholipid pathway inhibitors was examined in non-tolerant animals. First, the inhibitors did not alter the threshold for nociception in placebo pellet-implanted animals injected with vehicle s.c. Second, the potency of morphine s.c. was nearly identical in placebo pellet-implanted mice injected with vehicle or inhibitor i.c.v.
Role of IP3 in tolerance
It is notable that low molecular weight heparin injected i.c.v. significantly reversed morphine tolerance. Yet interpreting these results is difficult without supportive in vitro evidence. Heparin is a potent and selective IP3 receptor antagonist that prevents Ca2+ release from intracellular Ca2+ pools (Jonas et al., 1997). Yet heparin must be injected into cells or perfused onto permeabilized cells because of its high molecular weight (i.e., 12,000–30,000 Da) and lack of membrane permeability. Some evidence indicates that the low molecular weight heparin (i.e., 6000 Da) used in this study is membrane permeable. Perfusion of low molecular weight heparin over a non-permeabilized cerebellar slice preparation attenuated glutamate-stimulated increases in free intracellular Ca2+ (Jonas et al., 1997). Obviously, other studies measuring intracellular Ca2+ release are needed to confirm its membrane permeable properties. However, it is interesting to speculate that heparin may have reversed tolerance by antagonizing IP3 receptors. In similar fashion, we have shown that ryanodine also reversed morphine tolerance (Smith et al., 1999), presumably by blocking Ca2+ release from Ca2+/caffeine-sensitive intracellular pools (Friel & Tsien, 1992; Smith & Stevens, 1995).
Alternatively, heparin may act through other cellular mechanisms. Heparin is a potent inhibitor of G protein-coupled receptor kinases (GRKs) (Kunapuli et al., 1994). GRKs regulate the responsiveness of mu- and delta-opioid receptors through agonist-specific receptor phosphorylation, desensitization and internalization (Kunapuli et al., 1994; Morikawa et al., 1998; Zhang et al., 1998). For example, heparin significantly reduced the magnitude and rate of delta-opioid receptor desensitization in cultured SK-N-BE and NG108-15 cells (Hasbi et al., 1998; Morikawa et al., 1998). Thus, it is equally plausible that heparin reversed tolerance by transiently inhibiting GRK. Finally, if low molecular weight heparin did not permeate the cells, extracellular sites should be considered. Heparin can act on cell-surface heparin sulphate proteoglycan sites which were shown to be crucial in enabling the expression of long-term potentiation in the hippocampus (Lauri et al., 1999). The function of proteoglycan sites on other neuron types remains to be investigated.
Role of PKC in morphine tolerance
Our data is consistent with the hypothesis that selective inhibitors of PKC reverse morphine tolerance. Others have prevented or reversed acute tolerance to mu- or delta-opioid agonists with PKC inhibitors such as chelerythrine chloride, H7 and calphostin C (Fundytus & Coderre, 1996; Bilsky et al., 1996; Narita et al., 1995; 1996). Both chelerythrine chloride and calphostin C are selective inhibitors of PKC. Our data with bisindolylmalemide I, Go 7874 and sangivamycin further supports the role for PKC in tolerance. Furthermore, both acute tolerance and tolerance resulting from chronic opioid administration appear to be mediated, in part, by changes in PKC levels or activity. Others have shown that Ca2+ sensitive PKC enzyme activity is increased in the pons/medulla, but not the midbrain, of morphine tolerant rats (Narita et al., 1994). Alternatively, measurements of immunoreactive PKC have yielded mixed results. Chronic morphine, heroin or methadone administration significantly reduce PKC-alpha/beta immunoreactivity in rat cerebral cortex, brainstem and hypothalamus (Ventayol et al., 1997). In opposite fashion, PKC-gamma immunoreactivity is significantly increased in spinal cord laminae I and II of morphine tolerant rats (Mao et al., 1995). At the very least, these studies, combined with the behavioural data, indicate that PKC plays a major role in expression of opioid tolerance. The recently developed PKC-gamma knock-out mouse may provide new insights into the role of one PKC isoform in tolerance (Malmberg et al., 1997).