In this study we compare the effects of treatment with external sodium adenosine 5'-triphosphate (ATP) with the effects of L-buthionine-SR-sulfoximine (BSO) on B16 melanoma growth and on the modulation of the cytotoxic antimelanoma activity of cyclophosphamide (CY). We evaluated the in vitro effects of treatment with ATP or BSO on intracellular glutathione (GSH) levels, mitochondrial membrane potential (delta psi(m)) and the proliferation rate of the B16F10 melanoma cell line. Compared with untreated cells, delta psi(m) and GSH levels were already significantly decreased (25% and 57% reduction, respectively) after the first hour of incubation in culture cells exposed to 3 mM ATP. After 24 and 48 h a major reduction was observed in delta psi(m) (nearly 30%). GSH levels were also maximally depleted at 24 h (approximately 75%) and partially recovered (up to 37% of levels of control) after ATP was removed from the medium. At 24 and 48 h, the proliferation rate was decreased 1.4- and 1.7-fold, respectively, compared with control cells. Treatment with 50 microM BSO produced a time-dependent decrease in GSH levels (0.5, 21, 48 and 97.3% reduction at 1, 4, 8 and 24 h, respectively), but up to 54% of the levels of control cells was recovered after BSO was removed from the medium. In contrast to ATP, neither delta psi(m) nor proliferation rate was significantly modified in the first 24 h with BSO treatment. At 48 h, delta psi(m) was reduced by nearly 27%, and cell proliferation decreased 1.2-fold compared with controls. When the in vitro cytotoxic effect of low dose acrolein (an active metabolite of CY) in combination with BSO or ATP was analysed, a synergistic effect was found between BSO and acrolein, with a dose modification factor (DMF) of 1.98, but the antiproliferative effects of acrolein plus ATP were only approximately additive (DMF = 1.05). In addition, in in vivo studies differential effects were found between ATP and BSO. Specifically, whereas BSO alone significantly increased the survival time of mice bearing B16 melanoma liver metastases, and enhanced the cytotoxic effect of CY on this tumour model, no therapeutic benefits could be observed with ATP treatment, either alone or in combination with diethyl maleate (a GSH-depleting agent) and CY. In conclusion, our findings show that in our experimental system, both extracellular ATP and BSO have growth-inhibitory properties against B16 melanoma in vitro. In vivo, however, only BSO produces a chemosensitizing effect, whereas ATP has not proved useful as a biological modifier of chemotherapy.