The N-phenethyl analogue (compound 4) of the μ-selective opioid antagonist cyprodime (1) has been prepared and evaluated in radioligand binding, bioassays (guinea-pig ileum myenteric plexus and mouse vas deferens preparations) and the rat tail flick test. It was found to possess remarkable affinity and high preference for μ opioid receptors and high antinociceptive potency which is comparable to its N-methyl analogue 3. The nonpeptide, competitive μ opioid receptor antagonist cyprodime (1) has become a valuable tool in opioid research. Recently a new and more efficient synthesis of cyprodime has been reported. Cyprodime has been tritium labelled and has become available for radioligand binding assays especially since it is commercially available. An extensive study on cyprodime-related compounds revealed that several changes to the cyprodime molecule (e. g. an additional methoxy group at C-3, different substituents at C-4, a 14-ethoxy group, a N-allyl group) yielded compounds with either less μ selectivity or partial agonist activity. Increasing the chain length at C-4 (compound 2) resulted in higher affinity for μ receptors (ca. 2-fold) but very little change in either selectivity or intrinsic activity. A N-methyl group instead of the N-cyclopropylmethyl group afforded a compound (3) with very high antinociceptive potency. In the series of morphinans having a 4,5-epoxy bridge it was found that a N-phenethyl group could increase the antinociceptive potency ca. 10-fold in comparison to their N-methyl counterparts. In an effort to investigate the role of a N-phenethyl group in morphinans lacking the 4,5-epoxy bridge we synthesized 4,14 dimethoxy-17-(2-phenylethyl)-morphinan-6-one (4) and performed in vitro and in vivo studies. The synthesis of compound 4 was accomplished starting from 4,14-dimethoxy-17-methylmorphinan-6 one (3) which is available from oxymorphone in 6 steps. In modification of the original procedure, compound 3 was N-demethylated by an improved procedure using 1-chloroethyl chloroformate instead of 2,2,2-trichloroethyl chloroformate. Thus, treatment of 3 with 1-chloroethyl chloroformate gave carbamate 5 as an oil which was not further purified and characterized. Refluxing this oil in MeOH afforded the N-normorphinan 6 as the hydrochloride salt. 6.HCl was alkylated with 2-phenylethyl bromide in DMF at elevated temperature to yield compound 4 which was isolated as the hydrobromide salt. The in vitro studies performed were radioligand binding and bioassays (guinea-pig ileum myenteric plexus preparation (GPI) and mouse vas deferens preparation (MVD)). The inhibitory effects of 4.HBr and cyprodime were assessed in homogenates of guinea-pig brain in Tris.HCl buffer (50 mM, pH 7.4) employing [³H]DAMGO (μ agonist), [³H]DPDPE (δ agonist) and [³H]U69593 (κ agonist) as radioligands (Table 1). Introduction of a N-phenethyl group to the cyprodime molecule greatly enhanced μ affinity and afforded a compound with 50-fold preference for μ over δ receptors and 700-fold preference for μ over κ receptors. A similar μ affinity was observed with the N-methyl analogue 3 but without showing preference for μ over κ receptors. The bioassays (GPI and MVD) were performed as described previously. Compound 4.HBr was a potent agonist in both GPI and MVD. In both tissues the ability of naloxone to antagonize the responses to 4.HBr was similar to values determined against the μ agonist DAMGO, but different to the Kₑ for naloxone determined against the κ agonist U69593 and the δ agonist DPDPE, confirming that the compound was acting as a μ agonist in both tissues (Table 2). The mouse vas deferens has high δ receptor reserve but considerably lower μ receptor population. The μ activity of 4 in this tissue therefore confirms the high selectivity of the compound for μ over δ opioid receptors. Compound 4.HBr was tested in vivo in the rat tail flick test and exhibited remarkable antinociceptive potency which is comparable to its N-methyl analogue 3. When compared to morphine, 4.HBr shows approximately 50 times the potency of this reference drug (Table 3). Replacement of the N-cyclopropylmethyl moiety of cyprodime with phenethyl to afford 4.HBr or with methyl to afford 3 markedly improved μ opioid receptor affinity in spite of the lack of a 3-OH function, without altering δ affinity thus resulting in improved μ/δ selectivity. Unlike 3 however substitution with phenethyl led to a loss of affinity for κ receptors such that 4.HBr had a very high preference for μ over κ receptors, resulting in a compound with much improved overall μ selectivity. In marked contrast to cyprodime, 4.HBr was a potent agonist in both bioassays (GPI and MVD) and was found to exhibit a high preference for μ opioid receptors in these tissues. 4.HBr possessed also high agonist potency in vivo in the rat tail flick test comparable to its N-methyl analogue 3 as might be expected from the similar affinities of the compounds for the μ opioid receptor.