The effects of carminomycin and its 11-methyl ether analog were characterized and compared with those of Adriamycin in several systems both in vivo and in vitro. When comparing maximum effective doses, carminomycin was found to be approximately 10- to 20-fold more potent in vivo than was Adriamycin against mouse L1210 leukemia, and the latter agent was approximately equipotent with carminomycin-11-methyl ether. Similarly, 50% inhibitory concentrations of 0.09, 0.39, and 0.53 µm were obtained for carminomycin, Adriamycin, and carminomycin-11-methyl ether, respectively, using in vitro colony survival studies against Novikoff hepatoma ascites cells. This ranking was not repeated for the other systems tested. The 50% inhibitory concentrations of Adriamycin for whole cellular nucleic acid syntheses were 2- to 3-fold lower than those of carminomycin and over 10- to 15-fold lower than those of carminomycin-11-methyl ether. The apparent binding constants obtained for calf thymus DNA and salmon sperm DNA for Adriamycin were 3.67 × 106 and 11.68 × 106m−1, respectively. Those obtained for carminomycin were 0.26 × 106 and 0.15 × 106m−1, respectively, and no detectable binding was observed for carminomycin-11-methyl ether. These findings were confirmed by analysis of the effects of these anthracyclines on superhelical PM-2 DNA by agarose gel electrophoresis. Increasing concentrations of Adriamycin of up to 200 µm progressively decreased the superhelicity of PM-2 DNA in a manner typical of an intercalative binding agent, and concentrations 2- to 5-fold and 20- to 50-fold higher for carminomycin and its analog, respectively, were required to obtain equivalent results. These results demonstrate that DNA-binding and nucleic acid synthesis-inhibitory effects do not correlate with the antitumor action of carminomycin and its 11-methyl ether analog. This suggests the importance of other subcellular targets, possibly distinct from those of Adriamycin, which may be important in the cytotoxicity of carminomycin and its 11-methyl ether analog.

1

This work was partially supported by a grant from Bristol Laboratories.

This content is only available via PDF.