The source of hepatocytes, whether via autopsy from accident victims or via
biopsy from diseased livers, may affect the induction of CYP3A4. Several
studies have indicated that the expression of CYP3A4 is decreased in certain
liver diseases such as hepatic cirrhosis, hepatitis B or C, and hepatic tumors
(Kirby et al., 1996; Luo et al., 2004; Philip et al., 1994; Sotaniemi et al., 1995;
Yang et al., 2003). Therefore, hepatocytes obtained from these patients will
probably have lower constitutive or basal CYP3A4 activity compared to
hepatocytes from accident victims without liver diseases. These studies also
showed that human hepatocytes with lower basal activity exhibit a greater fold
induction of CYP3A4 activity (LeCluyse et al., 2000, 2001b; Luo et al., 2002).
Because of high interindividual differences in induction potential, experiments
should be conducted with hepatocytes obtained from at least three individual
donor livers.
Experimental results, including CYP3A4-dependent metabolic activity,
mRNA and protein levels, are usually expressed as fold induction, namely
(results from test drug treated cells)/(results from vehicle control cells), or as
the percentage of positive control, namely [100% (results from test-drug
treated cells)/(results from positive control cells)]. A test compound is
considered to be a potential CYP inducer in humans if it produces more
than twofold induction, or if its induction is more than 40% of that produced
in the positive control. For example, in experiments with human hepatocytes
obtained from four donor livers, carbamezepine, clotrimazole, dexamethasone,
dexamethasone-t-butylacetate, phenobarbital, phenytoin, rifampin, sulfadimi-
dine, and sulfinpyrazone induced CYP3A4 activity more than twofold, or their
induction was more than 40% of positive control (rifampin at 10mM), while
methotrexate, probenecid, ritanovir, taxol, and troleadomycin did not show
significant induction in primary human hepatocyte cultures (Luo et al., 2002).
An alternative method of expression of induction potential is the use of
EC 50 or percentage ofYmax.EC 50 (effective concentration of test compound at
which 50% maximal induction is achieved) can be used to compare the potency
of test compounds, but EC 50 alone does not provide sufficient information
about induction. In addition, the experimental design and the concentration-
response function for the inducer can significantly affect the accuracy of the
EC 50 value. The index %Ymaxexpresses the percentage of maximal induction
by a positive control (e.g., rifampin) that a test compound can theoretically
produce. However,Ymaxis most accurate when estimated using concentrations
of test compound that correspond to likely therapeutic concentrations.
In general, the estimates of fold induction obtained by measuring mRNA,
protein, and enzyme activity are consistent with each other. However, for
certain test compounds, results obtained by measuring mRNA or protein
directly more accurately represent CYP3A4 induction than those obtained by
measuring enzyme activity. Such compounds, including mifepristone, ritona-
vir, tamoxifen, troleandomycin, and DPC 681, are both inducers and
mechanism-based inactivators of CYP3A4, and direct inhibition of drug
metabolism will partially mask the induced increase in CYP3A4 expression
556 TESTING DRUG CANDIDATES FOR CYP3A4 INDUCTION