Low Sensitivity of the Positron Emission Tomography Ligand [C]Diprenorphine to Agonist Opiates

Previously, we reported minimal opioid receptor occupancy following a clinical dose of the -opioid agonist, methadone, measured in vivo using positron emission tomography (PET) with [C]diprenorphine and subsequently used rats to obtain experimental data in support of a high receptor reserve hypothesis (Melichar et al., 2005). Here, we report on further preclinical studies investigating opioid receptor occupancy with oxycodone ( and -receptor agonist), morphine ( -receptor agonist), and buprenorphine (partial agonist at the -receptor and antagonist at the and -receptors), each given at antinociceptive doses. In vivo binding of [C]diprenorphine was not significantly reduced after treatment with the full agonists but was reduced by 90% by buprenorphine. In addition, given that [C]diprenorphine is a non-subtype-specific PET tracer, there was no regional variation that might feasibly be interpreted as due to differences in opioid subtype distribution. The data support minimal competition between the high-efficacy agonists and the non-subtype-selective antagonist radioligand and highlight the limitations of [C]diprenorphine PET to monitor in vivo occupancy. Alternative means may be needed to address clinical issues regarding opioid receptor occupancy that are required to optimize treatment strategies. Positron emission tomography (PET) with the radioligand, [C]diprenorphine has been used for over a decade to assess or measure opioid receptor function in human brain. Specific binding of this ligand in vivo can be blocked by the opioid receptor antagonist, naloxone (13 g/kg naloxone resulted in 50% reduction in binding; Melichar et al., 2003) and is locally reduced in conditions expected to result in an increased concentration of endogenous agonist; for example, pain (for review, see Sprenger et al., 2005). We had therefore used PET with [C]diprenorphine to quantify opioid receptor occupancy in methadone-maintained opiate-dependent patients. Our aim was to address clinical issues in treatment of opioid dependence such as agonist dose-to-outcome, with a view to optimizing treatment strategies but, in the end, found no statistically significant reduction in specific binding compared with control (Melichar et al., 2005). A likely explanation was that clinically relevant doses (18–90 mg daily) of methadone result in low levels of occupancy of opioid receptors. Other factors considered were receptor trafficking, agonist concentration at the site of action, and, since the predominant form of the opioid receptor in vivo is a low agonist affinity state, the limited ability of agonists to block highaffinity binding of an antagonist radiotracer. Opioid receptors are classified into at least three subtypes ( , , ) and are G-protein coupled, with activation by both native peptides and structurally distinct nonpeptide alkaloid ligands (Minami and Satoh, 1995). In the continued presence of agonists, acute actions are followed by diverse regulatory processes, including rapid internalization, that modulate the number and functional activity of the receptors (e.g., Ko et al., 1999). The oripavine, diprenorphine has a high affinity to each subtype (Richards and Sadée, 1985) but, although acting as a partial agonist at (Toll et al., 1998), likely partial agonist at -receptors, is an antagonist at the -subtype (Lee et al., 1999). Its high lipophilicity (Shapira et al., 2001) and lack of sensitivity to sodium ions (Lee et al., 1999) facilitates binding of [C]diprenorphine to internalized (Shapira et al., 2001), but not down-regulated (Ko et al., 1999), opioid receptors. The efficacy of diprenorphine at the different subtypes is important when interpreting its use as a PET tracer, because the likelihood of it being displaceable by an endogenous or exogenous agonist is less likely if diprenorphine is an antagonist. If a PET tracer is sensitive to exogenous or en1 Current affiliation: Department of Palliative Medicine, Royal Marsden National Health Science Foundation Trust, Surrey, United Kingdom. Article, publication date, and citation information can be found at http://jpet.aspetjournals.org. doi:10.1124/jpet.107.121749. ABBREVIATIONS: PET, positron emission tomography; SBI, specific binding index; ANOVA, analysis of variance. 0022-3565/07/3222-661–667$20.00 THE JOURNAL OF PHARMACOLOGY AND EXPERIMENTAL THERAPEUTICS Vol. 322, No. 2 Copyright © 2007 by The American Society for Pharmacology and Experimental Therapeutics 121749/3229886 JPET 322:661–667, 2007 Printed in U.S.A. 661 at A PE T Jornals on A ril 0, 2017 jpet.asjournals.org D ow nladed from dogenous compounds, increasing their levels will result in reduced radioactive tracer binding. Such sensitivity is most widely exploited in human imaging of the dopaminergic system, e.g., [C]raclopride and dopamine. To help explore why in our human [C]diprenorphine PET study we were unable to detect occupancy by methadone and to inform future human [C]diprenorphine PET studies, we investigated whether acute doses of an alternative opioid agonist used in opioid addiction or treatment of dependence (morphine, oxycodone, or buprenorphine) were able to block [C]diprenorphine binding in rat brain. Relative intrinsic efficacies at the -opioid receptor subtype of those tested were methadone oxycodone morphine buprenorphine (Adams et al., 1990; Duttaroy and Yoburn, 1995; Cowan et al., 1977). If higher intrinsic activity at the different receptor subtypes is related to a higher receptor reserve in vivo, we hypothesized that regional differences in [C]diprenorphine blockade (receptor occupancy) might be detected with the different opiates. As with human studies, in vivo rat PET scanning had already demonstrated blockade of [C]diprenorphine binding with the antagonist, naloxone (Rajeswaran et al., 1991), and ex vivo autoradiographic findings were consistent with [H]diprenorphine binding being a sensitive index of endogenous agonist concentration (Seeger et al., 1984; Gerrits et al., 1999). In the first rat study of the current series, we determined that specific binding of [C]diprenorphine was not reduced by acute pretreatment with methadone, at antinociceptive i.v. doses, and this finding was published alongside the human study, in support of the high receptor reserve hypothesis (Melichar et al., 2005). Here, we report results from acute treatments with the remaining compounds, namely, oxycodone, a (Chen et al., 1991) and putative (Nielsen et al., 2000) receptor agonist, morphine, a potent agonist acting primarily via the -receptor, and buprenorphine, a partial agonist at the -receptor and antagonist at the and -receptors (Lee et al., 1999; Neilan et al., 2004). All doses were chosen from those reported in the literature as being effectively antinociceptive. Materials and Methods The study was carried out by licensed investigators in accordance with the UK Home Office’s Guidance on the Operation of the Animals (Scientific Procedures) Act 1986 (Her Majesty’s Stationery Office, February, 1990) and used 36 adult male Sprague-Dawley rats (mean S.E.; body weight, 288 7 g). Animals were from Harlan UK Ltd. (Bicester, UK). Oxycodone hydrochloride, quinine hydrochloride, and morphine sulfate were purchased from Sigma-Aldrich (Poole, UK). Buprenorphine hydrochloride (Temgesic; Reckitt and Coleman) was from Central Biological Services (Hammersmith Campus, Imperial College, London, UK). [C]Diprenorphine was supplied by Hammersmith Imanet on site for human PET, according to the method described by Luthra et al. (1994), and aliquots dispensed for biological studies. The biodistribution studies were begun within 10 min of end-ofsynthesis and used three to four rats per preparation. The radioligand was given i.v. to awake but lightly restrained rats, via a previously catheterized tail vein. The mean S.E. injected radioactivity was 11.7 0.4 MBq, with a specific activity of 42 4 MBq/ nmol (decay-corrected to the first injection for each experiment). The coinjected stable diprenorphine was 1.3 0.1 nmol/kg, expected to have a minimal “occupancy” effect (Richards and Sadée, 1985; Cunningham et al., 1991). At 60 min after injection of [C]diprenorphine, rats were euthanized by i.v. Euthatal, and a blood sample ( 1 ml) was collected from the ventricle of the heart. The brains were removed within 2 min, and 16 regions were sampled, as listed in Table 1. These were chosen to represent not only a range of opioid receptor density but also a differential distribution of the three major classes. Table 1, compiled from various sources, summarizes the relative proportion of the subtypes in each of the regions sampled. Note that -sites are not prevalent in rat brain, representing an estimated 10 to 12% of the total opioid receptor binding capacity, compared with 37% in human brain (Mansour et al., 1988). Radioactivity was measured using a Wallac gamma counter, with automatic correction for radioactive decay. Results were normalized for the amount of radioactivity injected and for body weight, giving “uptake units” (counts per minute per gram wet weight tissue)/ (injected counts per minute per gram body weight). A more detailed description and the rationale for the 60-min endpoint are published in Melichar et al. (2005). To account for group differences in clearance of radioactivity from the peripheral circulation, the tissue concentrations were also expressed relative to both individual plasma and cerebellum concentrations. The latter was assumed to approximate to nonspecific binding of the radioligand (Seeger et al., 1984; Schadrack et al., 1999), with the tissue/cerebellum uptake ratio at 60 min giving a measure of the specific binding index (SBI). Seven groups of animals were used. One group received only the radioligand (controls, n 5). Six further groups were predosed 5 min before [C]diprenorphine as follows: 1) i.v. oxycodone HCl at 0.5 mg/kg in physiological saline (0.9% sodium chloride), n 5; 2)