Usefulness of statistic experimental designs in enzymology: Example with recombinant hCYP3A4 and 1A2

First, the effects of 10 incubation factors were screened altogether on nifedipine dehydrogenase (NIF) and methoxyresorufin O-deethylase (MROD) activities catalyzed by recombinant human CYP3A4 and 1A2, respectively. Using a statistic experimental design, only 36 assays were needed to be exhaustive. Eight factors influenced CYP3A4-mediated NIF activity: buffer type, pH, temperature, Mg/EDTA, cytochrome b5, NADPH-P450 reductase, NADH, and solvent. Two factors had no significant effect: total ionic concentration and catalase/ reduced glutathione. Six factors influenced CYP1A2-mediated MROD rates: buffer type, pH, temperature, Mg/ EDTA, NADH, and glycerol. Four factors had no significant effect: total ionic concentration, cytochrome b5, reductase, and NAD+. Secondly, the combined effects of ionic strength and Mg concentration on NIF/ CYP3A4 were studied and easily modeled using another statistic experimental design. The effect of Rig was strong at a constant ionic strength of 100 mM and negligible at a constant ionic strength of 500 mM. Thirdly, the effects of influencing factors and their interactions on MROD/CYP1A2 were modeled after 40 assays using a third statistic experimental design. Later experiments confirmed the predictivity of the models and the efficiency of optimized conditions. This approach can be applied to other biochemistry areas. (C) 1999 Academic Press.