In search of a role for the morphine metabolite morphine-3-glucuronide.

Abstract

M orphine is the most commonly used opioid analgesic, and while extremely effective in relieving moderate to severe acute pain, it produces a number of well-characterized, undesirable side effects that include respiratory depression, sedation, dysphoria, decreased gastric motility, nausea, and vomiting. These analgesic and unwanted actions of morphine are mediated predominately by -opioid receptors. Some clinical reports indicate that large doses of morphine and other opioid analgesics may produce altered pain behaviors such as hyperalgesia and allodynia, as well as motor excitation and seizures. The prevalence of these symptoms in patients treated with even large doses of morphine is not well established, and has not been systematically studied. Estimates from clinical reports range from 1% (1,2) to at least 25% (3). The molecular basis for opioid-induced neuroexcitation in humans is not established, but in animals large doses of various opioids can produce excitation and seizures via both opioid receptor-dependent and independent mechanisms (4 – 6). -Opioid receptor-mediated disinhibition of neuronal subpopulations through suppression of gamma-aminobutyric acid (GABA) release is commonly observed throughout the brain and spinal cord (7,8) and probably accounts for the opioid receptor-mediated excitatory effects of -opioids. The nonopioid receptor-mediated excitation produced by opioid alkaloids may be mediated via a similar mechanism, because these compounds are low-affinity antagonists of postsynaptic ionotropic glycine and GABA receptors (9,10). In humans, morphine is metabolized predominantly into morphine-3-glucuronide (M3G) and to a lesser extent into morphine-6-glucuronide (M6G). Relatively large cerebrospinal fluid (CSF) concentrations of M3G and M6G may occur in patients who are treated with chronic, large doses of morphine (11). Like morphine, M6G is an agonist at -opioid receptors and presumably acts there to produce analgesia. In contrast, M3G does not have significant affinity for any opioid receptor (12) and has generally been considered to be an inactive metabolite. However, some investigators have reported excitatory effects of M3G, including seizures, after injection into the central nervous system of animals (13,14). The putative neuroexcitation produced by M3G has been proposed not only to underlie frank toxicity associated with morphine, but also to contribute to the development of tolerance to the analgesic effects of morphine, by functionally opposing the actions of morphine (15,16). However, other animal studies have not been able to demonstrate M3G-induced reductions in morphine analgesia (17,18), and no molecular mechanism for the actions of M3G has been identified. Furthermore, small-scale clinical trials in healthy humans have reported no toxic or other significant effects with acute plasma concentrations of M3G up to about 5 M and no interference by M3G with the analgesia produced by either morphine or M6G (19,20). Smith et al. (21,22) have been strong proponents of the hypothesis that M3G (and other glucuronide metabolites of alkaloid opioids) is a major cause of the toxicity and tolerance associated with large-dose morphine. They have previously documented the excitatory effects produced by large-dose morphine or M3G injection in rats and have contributed to excluding several sites of action of M3G, including the -opioid receptor (13). The study by Hemstapat et al. (23) in the current issue of Anesthesia & Analgesia examines the effect of M3G on calcium oscillations in an in vitro preparation of cultured embryonic rat hippocampal neurons. They report that large concentrations of M3G (5–500 M) produce an increase in calcium oscillations in the neurons, which is largely unaffected by the opioid antagonist naloxone. Interestingly, these results are similar to a previous study that found that the -opioid agonist DAMGO produced similar oscillations in intracellular calcium concentration, as well as neuronal excitation in a similar preparation (24). These two studies indicate that activation of the -opioid Accepted for publication May 7, 2003. Address correspondence and reprint requests to Dr. C. W. Vaughan, Pain Management Research Institute, Northern Clinical School, University of Sydney, NSW 2006, Australia. Address e-mail to chrisv@med.usyd.edu.au.