RAPID COMMUNICATION Evidence for an active transport of morphine-6-b-D- glucuronide but not P-glycoprotein-mediated at the blood–brain barrier Fanchon Bourasset,* Salvatore Cisternino,* Jamal Temsamani  and Jean-Michel Scherrmann* *INSERM U26, Ho ˆpital Fernand Widal, Paris, France  Synt:em, Parc scientifique Georges Besse, Nı ˆmes, France Abstract Morphine-6-b-D-glucuronide (M6G) is an active metabolite of morphine with high analgesic potency despite a low blood–brain barrier (BBB) permeability. The aim of the study was to elucidate its transport mechanism across the BBB. We first checked if M6G was effluxed by the P-glycoprotein (P-gp), as previously reported by others. Second, we investigated the role of anionic transporters like the multidrug resistance-associated protein mrp1 and the glucose transporter GLUT-1. The brain uptake of [ 14 C]M6G was measured by the in situ brain perfusion technique in wild-type and deficient mice [mdr1a(–/–) and mrp1(–/–)], with and without pro- benecid, digoxin, PSC833 or D-glucose. No difference was found between P-gp and mrp1 competent and deficient mice. The brain uptake of [ 14 C]M6G co-perfused with probenecid in wild-type mice was not significantly different from that found in group perfused with [ 14 C]M6G alone. The co-perfusion of [ 14 C]M6G with digoxin or PSC833 was responsible of a threefold decrease of its uptake in mdr1a competent and deficient mice, suggesting that another transporter than P-gp and sensitive to digoxin and PSC833, may be involved. The co-perfusion of [ 14 C]M6G with D-glucose revealed a threefold decrease in M6G uptake. In conclusion, P-gp and mrp1 are not involved in the transport of M6G at the BBB level in contrast to GLUT-1 and a digoxin-sensitive transporter (prob- ably oatp2), which can actively transport M6G but with a weak capacity. Keywords: blood–brain barrier, GLUT-1, morphine-6-b-D- glucuronide, mrp1, P-glycoprotein. J. Neurochem. (2003) 86, 1564–1567. Morphine-6-b-D-glucuronide (M6G) is an active metabolite of morphine (Abbott and Palmour 1988; Paul et al. 1989) which could be a more attractive analgesic than morphine because of its higher potency (Paul et al.1989;Gong et al.1991;Osborne et al.1992)andlowerrespiratory depressive effects (Thompson et al.1995;Lo ¨tschandGeisslinger2001). Its analgesic activity was extremely potent following intracerebro- ventricular administration but markedly lower after systemic adminis- tration because of its poor blood–brain barrier (BBB) permeability. Several hypotheses have been advanced to explain its low BBB permeability. The first one is related to its weak capacity to cross the BBB by passive diffusion because of its hydrophilicity with an apparent log octanol/water partition coefficient at )2.3 (Murphey and Olsen 1994).M6Gwasalsoreportedtobe7.5timeslesspermeablethroughthe rat BBB than morphine (Yoshimura et al. 1973) and similar to that of sucrose, which is considered as not crossing BBB, by an internal carotid artery study in the rat (Bickel et al. 1996). Another explanation results from more recent studies suggesting that the BBB permeability of M6G is restricted by the implication of an efflux mechanism (Bouw et al. 2001; Lo ¨tsch et al. 2002a). Several in vitro (Huwyler et al. 1996, 1998) and in vivo studies (Lo ¨tsch et al. 2002a) have suggested that M6G could interact with P-glycoprotein (P-gp), an ABC transporter expressed at the luminal membrane of the brain microvessel endothelial cells forming the BBB. P-gp is known to transport a large number of lipophilic and cationic substrates, but not, in theory, anionic compounds like M6G. This intriguing role of P-gp was also questioned by a pharmacodynamic study comparing the analgesic effects of M6G in wild-type and P-gp-deficient mice which showed no difference in the anti-nociceptive effect between the two strains of mice (Lo ¨tsch et al. 2000), whereas several experiments with morphine, reported as being a substrate of P-gp, gave significant differences in analgesia after inhibiting P-gp by chemical modulators (Letrent et al. 1999), antisense (King et al. 2001) or using knock-out mice (Xie et al. 1999; Thompson et al. 2000). All these results emphasize the confusing role that P-gp could play at the level of BBB on the control of M6G brain kinetics and activity. The aims of our study were, first, to control if P-gp restricts the brain transport of M6G through the BBB and, second, to elucidate if other transporters are involved. Because of the anionic and osidic properties of M6G, we have checked if transporters belonging to the multidrug resistance-associated proteins (MRP) like mrp1 or to the organic anion transporting polypeptide (oatp) family which are known to transport anionic substrates, could transport M6G at the BBB, as well as GLUT-1, the glucose transporter. Our experimental design was based on measuring the brain transport at BBB of [ 14 C]M6G using the in situ brain perfusion technique (Dagenais et al. 2000) in wild-type and deficient mice [mdr1a(–/–) and mrp1(–/–)] and several brain transport modulators or competitors of P-gp, GLUT-1 and organic anions transporters. Received June 11, 2003; accepted June 24, 2003. Address correspondence and reprint requests to Fanchon Bourasset, INSERM U26 – Ho ˆpital F. Widal. 200 rue du Fbg St-Denis, 75475 Paris cedex 10, France. E-mail: Fanchon.Bourasset@fwidal.inserm.fr Abbreviations used: BBB, blood–brain barrier; DMSO, dimethyl sulfoxide; i.p., intraperitoneal; M6G, morphine-6-b-D-glucuronide; MRP, multidrug resistance- associated protein; oatp, organic anion transporting polypeptide. Journal of Neurochemistry , 2003, 86, 1564–1567 doi:10.1046/j.1471-4159.2003.01990.x 1564 Ó 2003 International Society for Neurochemistry, J. Neurochem. (2003) 86, 1564–1567