Opioid drugs, such as morphine, are among the most effective analgesics available. However, their utility for the treatment of chronic pain is limited by side effects including tolerance and dependence. Morphine acts primarily through the mu-opioid receptor (MOP-R) , which is also a target of endogenous opioids. However, unlike endogenous ligands, morphine fails to promote substantial receptor endocytosis both in vitro, and in vivo. Receptor endocytosis serves at least two important functions in signal transduction. First, desensitization and endocytosis act as an "off" switch by uncoupling receptors from G protein. Second, endocytosis functions as an "on" switch, resensitizing receptors by recycling them to the plasma membrane. Thus, both the off and on function of the MOP-R are altered in response to morphine compared to endogenous ligands. To examine whether the low degree of endocytosis induced by morphine contributes to tolerance and dependence, we generated a knockin mouse that expresses a mutant MOP-R that undergoes morphine-induced endocytosis. Morphine remains an excellent antinociceptive agent in these mice. Importantly, these mice display substantially reduced antinociceptive tolerance and physical dependence. These data suggest that opioid drugs with a pharmacological profile similar to morphine but the ability to promote endocytosis could provide analgesia while having a reduced liability for promoting tolerance and dependence.

Morphine and methadone are both high affinity, potent mu opioid peptide (MOP) receptor analgesics. Here we compared the antinociceptive potencies of these two drugs when administered subcutaneously (s.c.), intrathecally (i.t.) or intracerebroventricularly (i.c.v.) in both rat and mouse using the tail-flick assay. We found that both morphine and methadone were potently antinociceptive when the drugs were administered s.c., showing comparable AD50 values in both species. However, the antinociception produced by methadone, when it was administered centrally, was much weaker than that produced by centrally-administered morphine. Specifically, the AD50 value for methadone antinociception was greater than 30-fold higher at both the i.t. and i.c.v. site in mouse, and not measurable in rat. Naloxone methiodide (NLX-M), a peripherally-restricted antagonist, was used to further examine the relative contribution of central versus peripheral sites to morphine and methadone antinociception. NLX-M, when administered s.c., blocked the antinociceptive effect of either systemically- or centrally-administered methadone, but had little effect on the antinociception produced by centrally-administered morphine. Furthermore, centrally-administered NLX-M significantly blocked antinociception produced by centrally-administered morphine, but not that produced by centrally-administered methadone. Together, these results suggest that methadone antinociception is significantly dependent on an action of the drug at peripheral sites, and could provide novel insight into the neural mechanisms that distinguish morphine versus methadone antinociception.

No abstract