A spontaneously originated murine mammary adenocarcinoma (16C), selected for its sensitivity to agents active against breast cancer in women, and one of the very few experimental solid tumor models responsive to Adriamycin (ADR) was used to study the mechanism of induced ADR resistance in vivo. A resistant variant of the tumor was obtained from the explant of a regrown tumor following a dose of ADR (12 mg/kg) that caused complete tumor repression but not cure. Progressive refractoriness to ADR was observed following up to six repeated cycles of treatment, regression and regrowth. However, beyond the sixth treatment, no further degree of resistance could be obtained. The cell line so established, designated 16C/ADRR, has a glutathione (GSH) content 1.67 times greater than the parent 16C line. Depletion of GSH by buthionine sulfoximine (BSO) enhanced the cytoxicity of ADR in both cell lines. The sensitization effect appeared to be dependent on the degree of GSH depletion, requiring a threshold level of depletion to approximately 30% of control. The resistance of 16C/ADRR, however, appeared not to be directly related to the increased absolute GSH level per se since reduction of the GSH content of the 16C/ADRR line to levels similar to that of the parent 16C line did not restore the original sensitivity to ADR. However, the activities of two important elements in the GSH detoxification system, GSH peroxidase and S-transferase, were found to be elevated in resistant cells by factors of 2.4 and 4.7-5.6 respectively. In vivo studies with a diverse spectrum of antineoplastic drugs revealed a pattern of cross-resistance consistent with the idea that elevated GSH S-transferase and peroxidase activities may be responsible for the decreased (2.8- to 5.3-fold) sensitivity to ADR. 16C/ADRR exhibited cross-resistance with melphalan (MEL), but none with vincristine (VCR), vinblastine (VBL) or etoposide (VP-16). These results clearly demonstrate non-adherence by the 16C/ADRR tumors to the well characterized multidrug resistance (mdr) phenotype. Further affirmation of this conclusion was obtained by immunochemical and pharmacological studies. When a monoclonal antibody prepared against the mdr associated, 170 kD P-glycoprotein (170 P-gp), was used, the presence of the 170 kD P-gp in both the sensitive and resistant 16C lines could not be detected, although the presence of a lower molecular weight form of P-gp could not be ruled out entirely. High performance liquid chromatographic measurement of ADR accumulation and elimination also failed to reveal any significant differences between the sensitive and resistant variants.(ABSTRACT TRUNCATED AT 400 WORDS)