A series of 17 analogues were developed on the basis of the general formula H-Dmt-Tic-NH-CH(R)-R' (denotes chirality; R = charged, neutral, or aromatic functional group; R' = -OH or -NH(2)). These compounds were designed to test the following hypothesis: the physicochemical properties of third-residue substitutions C-terminal to Tic in the Dmt-Tic pharmacophore modify delta-opioid receptor selectivity and delta-opioid receptor antagonism through enhanced interactions with the mu-opioid receptor. The data substantiate the following conclusions: (i) all compounds had high receptor affinity [K(i)(delta) = 0.034-1.1 nM], while that for the mu-opioid receptor fluctuated by orders of magnitude [K(i)(mu) = 15.1-3966 nM]; (ii) delta-opioid receptor selectivity [K(i)(mu)/K(i)(delta)] declined 1000-fold from 22,600 to 21; (iii) a C-terminal carboxyl group enhanced selectivity but only as a consequence of the specific residue; (iv) amidated, positive charged residues [Lys-NH(2) (6), Arg-NH(2) (7)], and a negatively charged aromatic residue [Trp-OH (11)] enhanced mu-opioid affinity [K(i)(mu) = 17.0, 15.1, and 15.7 nM, respectively], while Gly-NH(2) (8), Ser-NH(2) (10), and His-OH (12) were nearly one-tenth as active; and (v) D-isomers exhibited mixed effects on mu-opioid receptor affinity (2' << 3' < 4' < 1' < 5') and decreased delta-selectivity in D-Asp-NH(2) (1') and D-Lys(Ac)-OH (5'). The analogues exhibited delta-opioid receptor antagonism (pA(2) = 6.9-10.07) and weak mu-opioid receptor agonism (IC(50) > 1 microM) except H-Dmt-Tic-Glu-NH(2) (3), which was a partial delta-opioid receptor agonist (IC(50) = 2.5 nM). Thus, these C-terminally extended analogues indicated that an amino acid residue containing a single charge, amino or guanidino functionality, or aromatic group substantially altered the delta-opioid receptor activity profile (selectivity and antagonism) of the Dmt-Tic pharmacophore, which suggests that the C-terminal constituent plays a major role in determining opioid receptor activity as an "address domain".