The effects of the novel mixed peptidase inhibitor, kelatorphan [N-(R)-3-(N-hydroxyaminocarbonyl-2-benzyl-1-oxopropyl)-L-alanine], were compared to those of a combination of the potent "enkephalinase" inhibitor thiorphan and the nonselective aminopeptidase inhibitor bestatin, on the catabolism of [3H]Met-enkephalin and on the release of endogenous Met-enkephalin by the rat spinal cord in vitro and in vivo. At 20 microM, kelatorphan almost prevented completely the degradation of exogenous [3H] Met-enkephalin by slices of the dorsal zone of the lumbar enlargement. Similarly, the addition of 20 microM kelatorphan to a [3H] Met-enkephalin-containing artificial cerebrospinal fluid superfusing the whole spinal cord of halothane-anesthetized rats efficiently protected the exogenous peptide from enzymatic degradation. In contrast, in the same in vitro and in vivo models, thiorphan (1 microM) or bestatin (20 microM) alone was inactive, and only their combination induced a significant protection of the exogenous peptide. In vitro and in vivo, kelatorphan (20 microM) increased markedly the spontaneous outflow of endogenous Met-enkephalin-like material as well as the peptide overflow due to K+-induced depolarization (in vitro and in vivo) or noxious stimulation (in vivo). Under similar conditions, thiorphan (1 microM) plus bestatin (20 microM) also enhanced the efflux of Met-enkephalin-like material, but generally to a lower extent than kelatorphan. Compared to thiorphan plus bestatin, kelatorphan exerts additional inhibitory effects on dipeptidylaminopeptidase activity and the present results could indicate that this enzyme also may be involved in the inactivation of extracellular Met-enkephalin at the spinal level in rats.