A series of mu-agonist DAMGO analogs were synthesized and pharmacologically characterized to test the 'biousian' hypothesis of membrane hopping. DAMGO was altered by incorporating moieties of increasing water solubility into the C-terminus via carboxamide and simple glycoside additions. The hydrophilic C-terminal moieties were varied from glycinol in DAMGO (1) to l-serine amide (2), l-serine amide beta-d-xyloside (3), l-serine amide beta-d-glucoside (4), and finally to l-serine amide beta-lactoside (5). Opioid binding and mouse tail-flick studies were performed. Antinociceptive potency (intravenous) increased, passing through a maximum (A(50) approximately 0.2 micromol/kg) for 2 and 3 as membrane affinity versus water solubility became optimal, and dropped off (A(50) approximately 1.0 micromol/kg) for 4 and 5 as water solubility dominated molecular behavior. Intravenous A(50) values were plotted versus hydrodynamic values (glucose units, g.u.) for the glycoside moieties, or the hydrophilic/hydrophobic Connolly surface areas (A(50) versus e(-Awater/Alipid)), and provided either a V-shaped or a U-shaped curve, as predicted by the 'biousian' hypothesis. The mu-selective receptor profile was maintained (K(i)'s = 0.66-1.3 nm) upon modifications at the C-terminus. The optimal 'degree of glycosylation' for the DAMGO peptide message appears to be between 1.25 and 1.75 g.u. (hydrodynamic g.u.), or 0.75 and 0.90 in terms of the surface-derived amphipathicity values.