Raymond F. Novak

Bis(alkylamino)anthracenedione antineoplastic agent metabolic activation by NADPH-cytochrome P-450 reductase and NADH dehydrogenase: Diminished activity relative to anthracyclines☆

Stimulation of the rates of NAD(P)H oxidation, superoxide generation, and hydrogen peroxide formation by three anthracenedione antineoplastic agents in the presence of NADPH-cytochrome P-450 reductase, NADH dehydrogenase, or rabbit hepatic microsomes was studied and the results compared with those obtained for the anthracyclines Adriamycin and daunorubicin. In all cases the anthracenediones, including mitoxantrone and ametantrone, were significantly (5- to 20-fold) less effective than the anthracyclines in stimulating NAD(P)H oxidation, superoxide formation, or hydrogen peroxide production. Of the three anthracenediones studied, the ring-monohydroxylated compound showed the greatest activity followed by the ring-dihydroxylated derivative (mitoxantrone). In contrast, the non-ring-hydroxylated anthracenedione (ametantrone) was a relatively ineffective electron acceptor and inhibited the reduction of more effective acceptors such as Adriamycin. Michaelis-Menten kinetic constants were determined by analysis of the rates of NADPH oxidation. NADP+ and 2-AMP inhibited the reduction of the ring-hydroxylated anthracenediones and anthracyclines, demonstrating the enzymatic nature of the reaction. The non-ring-hydroxylated anthracenedione inhibited the reduction of Adriamycin by both P-450 reductase and NADH dehydrogenase with 50% inhibition achieved at approximately 300 microM. Thus, there appears to exist a structural relationship between anthracenedione ring hydroxylation and metabolic activation. These results also suggest that the relative inability of the anthracenediones to function as artificial electron acceptors in comparison to the anthracyclines may be correlated with diminished anthracenedione cardiotoxicity.

Inhibition of angiogenesis by the antineoplastic agents mitoxantrone and bisantrene

The effects of mitoxantrone and bisantrene on angiogenic responses induced by tumor cell-conditioned media in the avascular cornea of rat eye have been evaluated. Both mitoxantrone and bisantrene effectively inhibited, in a concentration-dependent manner, angiogenesis induced by conditioned media obtained from either a hamster buccal pouch carcinoma cell line or P388D1 murine macrophage-like cells. Whereas vessel ingrowth in corneas containing tumor cell-conditioned media was detected as early as day 2 or 3 and was maximal by day 7, inclusion of mitoxantrone or bisantrene in the conditioned media at a 1:1 ratio (160 microM mitoxantrone or 32 microM bisantrene) resulted in complete inhibition of angiogenesis throughout the 14-day evaluation period. When concentrations of 64 and 32 microM mitoxantrone or 13 and 6.4 microM bisantrene were employed there was a marked delay in the appearance of capillary blood vessels (day 5 to 7) and a reduction in the intensity of angiogenic responses. No untoward toxicity to the tissue was observed at the concentrations of mitoxantrone or bisantrene employed.

Mitoxantrone: Propensity for free radical formation and lipid peroxidation — implications for cardiotoxicity

SummaryResults of comparative studies on stimulation of the rates of cofactor consumption, superoxide generation and hydrogen peroxide production by mitoxantrone (Novantrone®; dihydroxyanthracenedione; MXN), ametantrone (AM), doxorubicin (DOX) and daunorubicin (DNR) in the presence of NADPH-cytochrome P-450 reductase, NADH dehydrogenase, or rabbit hepatic microsomes have been reported. MXN and AM were substantially less effective in stimulating the rate of cofactor oxidation, superoxide formation or hydrogen peroxide production relative to the anthracyclines. In the presence of P-450 reductase, the rate of NADPH oxidation or superoxide generation produced by 100 μM MXN or AM was only 15% and 2% respectively of that produced by 100 μM anthracycline.The effects of MXN and AM on lipid peroxidation in hepatic microsomes, cardiac sarcosomes and cardiac mitochondria were determined and compared with those produced by ADM. MXN and AM at 50 μM inhibited the basal rate of NADPH-dependent rabbit liver microsomal lipid peroxidation by 50%; in contrast, DOX enhanced the rate of hepatic microsomal lipid peroxidation by 2-and 2.5-fold at 100 and 200 μM, respectively. Rabbit cardiac sarcosomal NADPH-dependent lipid peroxidation was inhibited completely at 100 μM anthracenedione. NADH-dependent lipid peroxidation in cardiac mitochondria was diminished by 50 μM MXN and AM, whereas 50 μM DOX produced a 2-fold stimulation in lipid peroxidation. The anthracenediones also effectively inhibited DOX-stimulated lipid peroxidation with 50% inhibition occurring at 4 μM (MXN) and 6 μM (AM). Moreover, both MXN and AM potently inhibited iron (100 μM)-stimulated lipid peroxidation in rabbit hepatic microsomes with 80% inhibition produced by 15 μM anthracenedione.These results are consistent with the diminished cardiotoxicity of mitoxantrone and ametantrone relative to DOX or DNR and may require a reassessment of the role of lipid peroxidation in the mechanism(s) of quinone antineoplastic agent-mediated cardiotoxicity.

Inhibition of adriamycin-stimulated microsomal lipid peroxidation by mitoxantrone and ametantrone, two new anthracenedione antineoplastic agents

Abstract Ametantrone and mitoxantrone, two new anthracenedione antineoplastic agents, produced a concentration-dependent inhibition of hepatic microsomal lipid peroxidation. Malondialdehyde production was diminished from 10.6 nmoles/mg/60 min to 3.3 and 5.4 nmoles/mg/60 min, in the presence of 100 μM mitoxantrone and ametantrone, respectively. Under similar conditions, Adriamycin stimulated lipid peroxidation over twofold. In addition, both mitoxantrone and ametantrone inhibited Adriamycin-stimulated lipid peroxidation, with 50% inhibition occurring at concentrations of 4 and 6 μM, respectively. Microsomal superoxide production was not significantly inhibited at anthracenedione concentrations which markedly decreased lipid peroxidation, suggesting that inhibition of lipid peroxidation was not the result of inhibition of superoxide generation. These results correlate with the lack of anthracenedione cardiotoxicity and also demonstrate anthracenedione inhibition of lipid peroxidation at micromolar concentrations; an observation with potential therapeutic significance.

The molecular basis for complexation of adriamycin with flavin mononucleotide and flavin adenine dinucleotide.

Abstract The interaction of Adriamycin with FMN or FAD was characterized using proton nuclear magnetic resonance chemical shift, linewidth, and T 1 relaxation rate measurements. Adriamycin produced saturable, concentration-dependent upfield shifts of the proton signals of both FMN and FAD. Differential changes in the magnitude of the upfield chemical shifts of the various proton signals of the flavin nucleotides in the presence of Adriamycin were also observed. The greatest change in chemical shift produced by Adriamycin occurred for the overlapping H-6, H-9 proton signals of the isoalloxazine ring of FMN followed by the 8α, 7α methyl group and aliphatic side chain signals, respectively. Qualitatively similar results were observed for FAD where changes in chemical shift in the presence of Adriamycin were such that H-6 > H-9 > 8α,7α-CH 3 > adenosine H-8 > adenosine H-1′. Concomitant with the Adriamycin-induced changes in chemical shift were increases in the linewidth ( ν 1 2 ) and longitudinal relaxation rate, T 1 −1 . Variable temperature studies demonstrated the reversibility of Adriamycin-flavin nucleotide interaction and the presence of rapid exchange between free and complexed species. These results suggest complex formation between Adriamycin and FMN or FAD via π-π ring stacking interactions occurring between the anthracycline ring of Adriamycin and the isoalloxazine ring of the flavins. Analysis of spectroscopic changes revealed the predominance of 1:1 stoichiometry of complex formation and allowed estimation of association constants and thermodynamic parameters. The results indicate that Adriamycin is effective in forming specific complexes with each flavin nucleotide and with comparable association constants.

Induction of rabbit hepatic microsomal cytochrome P-450 by imidazole: Enhanced metabolic activity and altered substrate specificity☆

Pretreatment of rabbits with imidazole resulted in a twofold increase in hepatic microsomal cytochrome P-450 content, with the apparent induction of two or more distinct forms of the cytochrome [K. K. Hajek and R. F. Novak (1982) Biochem. Biophys. Res. Commun. 108, 664-672]. The metabolic properties of imidazole-induced microsomes have been compared to those of uninduced, phenobarbital- and beta-naphthoflavone-induced preparations. Metabolic activity was enhanced as a consequence of increased P-450 content and as a result of the presence of different forms of the cytochrome. When rates were expressed per nanomole P-450 the following were observed: (a) p-nitroanisole O-demethylation was comparable in all preparations; (b) N,N-dimethylaniline N-demethylation was comparable in imidazole- and beta-naphthoflavone-induced, and uninduced microsomes; (c) polycyclic aromatic hydrocarbon hydroxylase activity was approximately twofold greater in imidazole-induced relative to phenobarbital-induced microsomes, but was only one-half that of beta-naphthoflavone-induced microsomes; and (d) metabolism of N,N-dimethylnitrosamine was enhanced fivefold, alcohol oxidation increased three- to fivefold, and aniline hydroxylation was threefold greater in imidazole-induced microsomes compared to phenobarbital- or beta-naphthoflavone-induced preparations. Eadie-Scatchard analysis yielded a single Km value for dimethylnitrosamine N-demethylase activity in imidazole-induced microsomes; in contrast, both high- and low-Km values were obtained for phenobarbital- or beta-naphthoflavone-induced microsomal preparations. Dimethylnitrosamine N-demethylase activity was P-450 dependent; neither flavin monooxygenase nor monoamine oxidase appeared to contribute significantly to dimethylnitrosamine metabolism. Dimethyl sulfoxide was a competitive inhibitor of dimethylnitrosamine N-demethylase activity in imidazole-, phenobarbital-, and beta-naphthoflavone-induced microsomes. Dimethyl sulfoxide competitively inhibited ethanol oxidation in imidazole-induced microsomes; it was a noncompetitive inhibitor of ethanol oxidation in phenobarbital- or beta-naphthoflavone-induced microsomes.

Characterization of hydrazine-stimulated proteolysis in human erythrocytes☆

The ability of hydrazine, acetylphenylhydrazine, methylhydrazine, and phenylhydrazine to stimulate proteolysis in red cells has been characterized. All four hydrazines effectively stimulated proteolysis in red cells and in hemolysate as evidenced by a two- to threefold increase in the rate of tyrosine release. The rate of tyrosine release varied linearly with time, increased with increasing concentration of hydrazine, and also increased as a function of hematocrit. The rank order for stimulation of proteolysis in red cells was phenylhydrazine greater than methylhydrazine greater than hydrazine approximately equal to acetylphenylhydrazine. Inhibitors of glycolysis in red cells only minimally (13-27%) decreased the rate of tyrosine release stimulated by the different hydrazines. Agents which diminished electron transport decreased the rate of tyrosine release. NADP inhibited the rate of tyrosine release stimulated by hydrazine, methylhydrazine, and acetylphenylhydrazine by approximately 36 to 41%; 2-AMP was less effective. The rate of tyrosine release resulting from insult by the hydrazines was increased slightly by methylene blue, moderately inhibited (approximately 10 to 27%) by the chelator o-phenanthroline and inhibited approximately 30 to 40% by N-ethylmaleimide. Use of an oxygen-depleted atmosphere (N2) increased slightly the rate of tyrosine release stimulated by the hydrazines; in contrast, carbon monoxide decreased proteolysis stimulated by hydrazine, methylhydrazine, and acetylphenylhydrazine by approximately 50%. Although the antioxidants dimethylfuran, dimethylthiourea, and methylsulfoxide failed to diminish proteolysis stimulated by the hydrazines, N-acetylcysteine exerted a protective effect, decreasing hydrazine-stimulated tyrosine release in red cells approximately 30 to 50%. Inclusion of 3-amino-1,2,4-triazole in the incubation failed to increase further the rate of hydrazine-stimulated proteolysis. These data suggest that more reactive free radicals generated from the hydrazine are responsible for protein damage, that damaged protein (hemoglobin) is degraded via proteolysis, and that an ATP-independent process primarily participates in the degradation of abnormal proteins in the red cell. Thus, proteolytic enzymes present in the erythrocyte appear to exert a protective effect against cellular damage through the removal of abnormal proteins generated as a consequence of xenobiotic insult. The ability of proteolytic enzymes to recognize and degrade abnormal proteins may be of importance in using protein (hemoglobin)-xenobiotic adducts to assess exposure to toxic agents (risk assessment).

Spectral and metabolic properties of liver microsomes from imidazole-pretreated rabbits.

Abstract Imidazole or phenylimidazole, when administered in vivo , elevates rabbit liver microsomal cytochrome P-450 levels approximately 2.0- and 1.5-fold, respectively, as compared to controls. SDS-polyacrylamide gel electrophoresis revealed protein bands of enhanced intensity occurring at the approximate positions of LM 2, LM 3, LM 4 and possibly, LM 6. Imidazole- and phenylimidazole-induced microsomes exhibit N,N-dimethylaniline and p -nitroanisole demethylase activities which paralleled the increase in cytochrome P-450 content. Dimethylnitrosamine N-demethylase activity, however, when expressed per nmole P-450, was 2- to 8-fold greater in imidazole-induced microsomes than in control, phenobarbital-, or β-naphthoflavone-induced microsomes, and was inhibited by carbon monoxide or ethyl isocyanide. Dimethylsulfoxide inhibited dimethylnitrosamine N-demethylase activity 62% in imidazole-induced microsomes, but only 11% and 26% in phenobarbital- or β-naphthoflavone-induced microsomes, respectively. Binding of imidazole to cytochrome P-450 in imidazole-induced microsomes was monophasic, in contrast to the biphasic binding observed in phenobarbital or β-naphthoflavone-induced preparations. In addition, the spectral dissociation constant (Kd) for imidazole binding to imidazole-induced microsomes was 10- to 100-fold less (i.e. 10- to 100-fold greater affinity) than that measured in phenobarbital or β-naphthoflavone-induced preparations.

Effects of anthrapyrazole antineoplastic agents on lipid peroxidation

The effects of three anthrapyrazoles and an aminoacridine derivative on doxorubicin- and iron-stimulated lipid peroxidation in rabbit hepatic microsomes have been characterized. Two anthrapyrazoles, CI-937 and CI-942, were potent inhibitors of lipid peroxidation with 15 microM drug inhibiting the rate of peroxidation 70 to 90%. In contrast CI-941 was relatively ineffective in inhibiting lipid peroxidation with only 35% inhibition occurring at 100 microM drug. CI-921, an aminoacridine derivative, diminished lipid peroxidation by 65% at 15 microM. All four drugs failed to decrease the rate of doxorubicin-stimulated NADPH oxidation at concentrations less than 50 microM, suggesting that inhibition of lipid peroxidation was not the result of diminished enzyme activity. CI-937 formed a 2:1 complex with ferric ion, KD = 47 microM, which was reversible with EDTA.

Spectroscopic evidence for anthracenedione antineoplastic agent self-association and complex formation with flavin nucleotides☆

Dihydroxyanthraquinone (DHAQ) and ametantrone (anthraquinone) are two new anthracenedione antineoplastic agents which were found by proton NMR spectroscopy to self-associate in aqueous media. Self-association was consistent with a bimolecular model, with average association constant values of 3400 and 2900 M-1 determined for DHAQ and ametantrone, respectively. Both anthracenediones interacted with the flavin nucleotides FMN and FAD to produce concentration-dependent upfield shifts of the flavin isoalloxazine ring proton signals, as observed by proton NMR spectroscopy. Average association constant values obtained for FMN-DHAQ, FAD-DHAQ, FMN-ametantrone, and FAD-ametantrone complexation were 5100, 2600, 4300, and 1600 m-1, respectively. Optical difference spectroscopy confirmed FMN-DHAQ complexation, which resulted in a hyperchromic, bathochromic shift of the DHAQ spectrum following addition of FMN. These results were consistent with the formation of a pi-pi bimolecular ring-stacking complex. Information obtained on anthracenedione self-association and complexation with flavins may be of consequence in the interpretation of anthracenedione-DNA binding data and flavoprotein-mediated anthracenedione metabolic activation.