The standard method to evaluate CYP3A inhibition is to study the conversion of the specific CYP3A probe testosterone to its 6 beta-hydroxy metabolite in human liver microsomes, in the absence and presence of potential inhibitors. Quantification of the 6 beta-hydroxy metabolite is achieved by HPLC resulting in a tedious and time-consuming assay. In order to increase the P450 inhibition throughput, efforts were made to find a CYP3A probe that would produce a fluorescent metabolite. This paper reports the discovery of DFB as a potential CYP3A fluorescent probe. DFB was significantly metabolized in human microsomes (approximately 1-2 nmol/(min. mg protein)) to give the fluorescent compound DFH. The involvement of CYP3A in the metabolism of DFB was determined using multiple approaches. First, incubations conducted with microsomes made from cell lines expressing single CYPs (Gentest Supersomes) indicated that CYP3A played a major role in the metabolism of DFB. Secondly, immunoinhibition studies conducted with CYP3A antibody resulted in >95% inhibition of DFB metabolism in HLM. Thirdly, inhibition studies with specific CYP1A1, 1A2, 2C8/9, 2C19, 2D6, and 2E1 chemical inhibitors did not suppress DFB activity in HLM. However, ketoconazole, miconazole, nicardipine, and nifedipine, all known CYP3A inhibitors, completely abolished the formation of DFH in HLM. The potency of several inhibitors determined using DFB and testosterone as CYP3A probes was consistent (R = 0.98). Finally, a good agreement was obtained for the formation of DFH and production of 6 beta-hydroxytestosterone when DFB and testosterone were incubated separately with various human liver microsome preparations (R = 0.94, N = 11). In order to use DFH as a fluorescent CYP3A marker in a 96-well plate format, it was important to remove the excess of NADPH at the end of the incubation because the fluorescence of NADPH interferes with DFH detection. This was achieved by adding oxidized glutathione and glutathione reductase to convert NADPH to NADP(+) which is not fluorescent. The liquid-handling steps were fully automated in a 96-well plate format and a template was designed to generate IC(50) curves and to address potential fluorescent interferences from the test compounds. The assay was found to be reproducible (intraday variability <10% and interday variability indicated less than a 2-fold variation in the IC(50) values) and is now routinely used in our laboratory to evaluate CYP3A inhibition of NCEs.
Copyright 1999 Academic Press.