Abstract
A refined cytochrome P450 (P450) enzyme IC50 shift assay for more accurately screening CYP3A time-dependent inhibitors (TDIs) is presented. In contrast to the regular IC50 shift assay, in which only one pair of P450 inhibition curves is generated, this modified method generates two pairs of inhibition curves; one pair of curves is created from human liver microsomal incubations with the test article in the presence or absence of NADPH (curves 1 and 2) (same as the traditional assay), and the other pair is created from new microsomal incubations with extract (compound/metabolites) of previous incubations (curves 3 and 4). To assess the true CYP3A time-dependent inhibition, we propose a new parameter, the vertical IC50 curve shift (VICS), represented by vertical shift difference between the two sets of curves divided by inhibitor concentration at which maximal vertical shift of curves 1 and 2 is observed. A shift in the curves 1 and 2 could mean a time-dependent inhibition or formation of a more active inhibitory metabolite(s). The new method provides more reliable characterization of the shift as a result of a true TDI- or metabolite-mediated reversible inhibition. Nine known TDI drugs were evaluated using this refined shift assay. The derived VICS values correlated well with the reported kinact/KI values derived via the conventional dilution assay method. Thus, the refined assay can be used to identify a true TDI and quantitatively assess the inactivation potential of TDIs in a high-throughput fashion. This assay can be invaluable to screen for true P450 TDIs in the early drug discovery.
Introduction
Reports of adverse, sometimes fatal, events in patients on multiple drug therapies are high. A significant number of these events is due to pharmacokinetic drug-drug interactions (DDIs) leading to elevated exposure to drugs with toxicity. Inhibition of cytochrome P450 (P450) enzymes is one of the most common mechanisms that can lead to DDIs. P450 inhibition can be categorized as reversible (competitive, uncompetitive, or noncompetitive) or irreversible (mechanism-based inactivation). Adverse pharmacokinetic DDIs can occur via both reversible and irreversible processes. For example, mibefradil, a potent mechanism-based inhibitor of CYP3A (Prueksaritanont et al., 1999), led to a fatal interaction when coadministered with substrates of CYP3A, such as HMG CoA reductase inhibitors (Krayenbühl et al., 1999), eventually leading to withdrawal of mibefradil from the U.S. market in 1998. Irreversible inactivations generally involve metabolism of inactivators to reactive metabolites, which covalently modify the P450 enzymes and can result in loss of P450 activities (Silverman, 1996; VandenBrink and Isoherranen, 2010). When the mechanism of the time-dependent inhibitor (TDI) is confirmed by additional studies, such as identification of P450-bound species, determination of P450 activity recovery after equilibrium dialysis, modulation of TDI by GSH, and/or measurement of inactivation kinetics, the TDI, a phonotypical term, could then be characterized as a mechanism-based inhibitor (MBI), a mechanistic term of enzyme inactivation. Thus, all MBIs are TDIs. As opposed to reversible inhibitors, P450 enzymatic activities can only be restored through de novo protein synthesis in the case of MBIs, and the duration of P450 inactivation in vivo can be prolonged even after elimination of the inactivators from the body. Hence, failure to take time-dependent inhibition of P450s into consideration can lead to gross underestimation of the potential of pharmacokinetic DDIs (Bjornsson et al., 2003; Grimm et al., 2009). Mechanism-based inactivation of P450s may lead to covalent binding to protein, which sometimes has been thought to be associated with idiosyncratic toxicities (Tucker et al., 2001). CYP3A is the most abundant enzyme among P450s in the liver and gut and is known to metabolize approximately 50% of the commercially available small molecule drugs. A number of important drugs have also been identified as time-dependent inhibitors of CYP3A. For the aforementioned reasons, many pharmaceutical companies have established screening paradigms to assess time-dependent inhibition of major human P450s by new chemical entities at an early drug discovery stage. A perspective of the Pharmaceutical Research and Manufacturers of America to address time-dependent inhibition of drug-metabolizing enzymes was published previously (Grimm et al., 2009).
In the conventional two-step dilution assay, widely used by the pharmaceutical industry to characterize time-dependent inhibition, the inactivation kinetic parameters depend on the fold dilution, and thus it may result in biased parameters (Maurer and Fung, 2000; Yang et al., 2005). Although much effort has been made to improve the throughput of the conventional dilution method, it is still a cumbersome method for the pharmaceutical industry to use at an early drug discovery phase. As an alternative to the traditional dilution assay, an IC50 shift assay was introduced (Walsky and Obach, 1995; Obach et al., 2007), in which a shift to a lower IC50 value for the inhibition curve in the presence of NADPH compared to that in the absence of NADPH might indicate CYP3A inactivation. Later, this CYP3A IC50 shift assay method was successfully applied to CYP2D6 and CYP2C9 inactivation assessments (Berry and Zhao, 2008). The predictive utility of the two in vitro methods, measuring empirical IC50 shift or mechanistic kinact/KI, was compared and evaluated by Burt et al. (2010).
It was suggested that the empirical IC50 approach was less accurate than the mechanistic kinact/KI model for in vivo DDI predictions, and thus the use of the empirical IC50 shift approach was limited to preliminary investigations. The IC50 shift assay clearly offers throughput advantage over conventional dilution assay, but ambiguity still prevails over inhibition because of the potential of a potent competitive inhibition by a metabolite or a true enzyme inactivation. Our refined IC50 shift method, which was first reported in 1995 (Li et al., 1995), has a similar throughput as this IC50 shift assay but also addresses the above concern. It can quantitatively assess inactivation potential of TDIs, similar to the conventional two-step dilution assay, and differentiate true P450 inactivation from potent reversible inhibition by a metabolite(s), in a single study.
Materials and Methods
Reagents.
Human liver microsomes were purchased from XenoTech, LLC (Lenexa, KS); mibefradil, diltiazem, raloxifene, verapamil, fluoxetine, ritonavir, erythromycin, quinidine, testosterone, and the reduced form of NADPH were purchased from Sigma-Aldrich (St. Louis, MO); bergamottin, 6′,7′-dihydroxybergamottin, and 6β-hydroxytestosterone were purchased from BD Gentest (Woburn, MA).
Refined P450 IC50 Shift Assay.
Test compounds were preincubated for 60 min, at 10 different concentrations, with human liver microsomes (1 mg/ml) at 37°C in the presence or absence of 2 mM NADPH and 3 mM MgCl2 in 100 μl of 0.1 M phosphate buffer (pH 7.4). Half of the incubates (50 μl) was transferred into 96-well plates containing 50 μl of testosterone (100 μM; final concentration, 50 μM) and NADPH in 0.1 M potassium phosphate buffer (pH 7.4) after a 1-h incubation. The samples were incubated for an additional 10 min at 37°C, and the formation of 6β-hydroxytestosterone, a CYP3A-mediated metabolite of testosterone, was analyzed by liquid chromatography/mass spectrometry/mass spectrometry. The other half of incubates was transferred into 96-well plates containing two volumes of acetonitrile to precipitate microsomes, followed by centrifugation (3000g for 10 min). The supernatants were then transferred to new 96-well plates and dried down under nitrogen. The dried residues were reconstituted in 0.1 M phosphate buffer with 1% acetonitrile (pH 7.4) and incubated with 0.5 mg/ml fresh human liver microsomes, NADPH, and testosterone in 50 μl for 10 min at 37°C for CYP3A activity measurement. The percentage of CYP3A activities remaining in all incubations, i.e., preincubations with or without NADPH and extract (compound/metabolites) from preincubation with or without NADPH, were plotted against test compound concentrations. The concentration that caused maximal vertical shift (CImax) was identified with an Excel macro program. The determination of vertical IC50 curve shift (VICS) parameter is illustrated in Fig. 1. The two curves in Fig. 1a (top), curves 1 and 2, represent the inhibition curves after a 60-min preincubation of microsomes with the CYP3A inactivator mibefradil in the presence and absence of NADPH. The vertical dashed line represents maximal change in the percentage of CYP3A activity with mibefradil concentration at CImax. The two curves in Fig. 1b, curves 3 and 4, represent the inhibition curves of mibefradil/metabolites in the extracts from the 60-min preincubation of microsomes in the presence and absence of NADPH. A basic assumption made is that the extraction of compound/metabolites from incubates was not an issue. The vertical dashed line represents the change in the percentage of control activity when microsomes were incubated with the inactivator at CImax.
Results
Nine known CYP3A TDIs (mibefradil, diltiazem, raloxifene, verapamil, fluoxetine, ritonavir, bergamottin, 6′,7′-dihydroxybergamottin, and erythromycin) and one non-TDI (quinidine) were selected as model compounds to validate the modified P450 IC50 shift assay. Figure 2, A1–A8, shows curves 1 and 2, the first pair of inhibition curves involving 60-min preincubations with and without NADPH, and Fig. 2, B1–B8, shows curves 3 and 4, the second pair of inhibition curves from the extracts (compounds/metabolites) of the above preincubations. As illustrated in Figs. 1a and 2, A1–A8, inhibition curves generated from all nine TDIs except ritonavir after a 60-min preincubation with NADPH showed a left shift compared with the curves without NADPH. Quinidine, which is not a TDI, showed no left shift (Fig. 1a, bottom). The shift in the inhibition curves generated from incubations with extracts in contrast was compound dependent (Figs. 1b and 2, B1–B8). Mibefradil showed a left shift, erythromycin showed no shift, and ritonavir, on the other hand, showed a right shift. In the cases in which a left shift was observed, the degree of the shift from curves 3 and 4 was always less than that of curves 1 and 2. The new empirical parameter VICS used for quantitatively assessing inhibition potential of a TDI is intuitively expressed as follows:
CImax is the concentration of inhibitor where the maximal vertical shift between curves 1 and 2 is observed, being directly proportional to inactivation and inversely proportional to concentration. A “positive” number of verticalshift max,curve3&4 is used for a left shift, and a negative number of verticalshift max,curve3&4 is used to represent a right shift. The correlation plot between the inactivation potentials derived from the conventional dilution method (kinact/KI) and the inactivation potentials derived from the refined IC50 shift assay (VICS parameter) for nine TDIs are depicted in Fig. 3.
Discussion
Mibefradil, diltiazem, raloxifene, verapamil, fluoxetine, ritonavir, bergamottin, 6′,7′-dihydroxybergamottin, and erythromycin are known MBIs. The inhibition curves 1 and 2 of eight of the nine TDIs (ritonavir as an exception) showed appreciable left shift, whereas quinidine, as expected, showed no clear shift. The shift between curves 1 and 2 from the preincubation experiment resulted from both CYP3A inactivation and reversible inhibition by the remaining parent drug or metabolites generated in the 60-min preincubation. The shift between the inhibition curves 3 and 4 generated from the second incubation (using extracts from the first experiment) reflects the CYP3A inhibition difference between “parent drug” and “parent drug plus metabolites” generated during the 60-min preincubation. Thus, left or right shift indicates that CYP3A inhibition potential of the metabolites generated during the 60-min preincubation is higher or lower than that of the parent drug, respectively. Similarly, no shift suggests that CYP3A inhibition potential of metabolites is similar to that of the parent drug. For example, curves 3 and 4 of ritonavir showed appreciable right shift, suggesting that its metabolites are less potent CYP3A inhibitors than ritonavir; the inhibition curves 3 and 4 of erythromycin showed no clear shift, suggesting that the CYP3A inhibition potential of the metabolites/parent is similar to that of erythromycin. Characterization of metabolites of MBIs for their reversible and irreversible inhibition potential using conventional methods were reported (Zhao et al., 2002; Wang et al., 2005).
In contrast to the regular shift assay in which IC50 shift was used for assessing the CYP3A inactivation, we propose to use VICS instead, an analog for net enzyme activity loss due to inactivation. In addition, the concentration at which the maximum vertical shift is observed (CImax) would be valuable to relate to plasma or predicted organ Cmax to assess the degree of inactivation in vivo. Because the contribution of metabolites to inhibition curve shift was taken into account, the proposed parameter VICS, expressed in eq. 1, is considered to provide a true characterization of a TDI. The VICS value for all nine CYP3A inactivators was calculated. The VICS values were found to correlate well with their kinact/KI from the conventional dilution method (Zhou et al., 2004), with the coefficient of determination (r2) of 0.96 (Fig. 3). The average 6′,7′-dihydroxybergamottin data were omitted from the correlation analysis because the kinact/KI value of 6′,7′-dihydroxybergamottin reported from two separate laboratories varied 10-fold. The r2 was still respectable at 0.86 when 6′,7′-dihydroxybergamottin data were included. This correlation demonstrated the value of this refined assay for reliably identifying a TDI and, hence, a possible mechanism-based CYP3A inhibitor. The new method can readily be used for rank ordering compounds based on the VICS parameter for any P450. It is realized that the data set described in this report is currently small, but further research in the future would help define wider applicability and limitations of this approach. The extraordinary correlation found between VICS and kinact/KI would also help design a mathematical link between the two parameters. In addition, its high throughput and ability to differentiate true CYP3A inactivation from that confounded by reversible inhibitory metabolites in a single study makes it an ideal primary tool for the pharmaceutical industry to screen P450 TDIs in the early drug discovery stage.
Authorship Contributions
Participated in research design: Li, Lu, Balani, and Gan.
Conducted experiments: Li.
Performed data analysis: Li and Lu.
Wrote or contributed to the writing of the manuscript: Li, Balani, Lu, and Gan.
Footnotes
↵1 Drug Metabolism and Pharmacokinetics, Preclinical and Clinical Development Sciences, Biogen Idec, Inc., 14 Cambridge Street, Cambridge, MA 02142.
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
doi:10.1124/dmd.111.038208.
-
ABBREVIATIONS:
- DDI
- drug-drug interaction
- TDI
- time-dependent inhibitor
- MBI
- mechanism-based inhibitor
- VICS
- vertical IC50 curve shift
- CImax
- inhibitor concentration that produced maximal vertical shift in the IC50 curves
- P450
- cytochrome P450.
- Received January 14, 2011.
- Accepted March 10, 2011.
- Copyright © 2011 by The American Society for Pharmacology and Experimental Therapeutics
References
DMD articles become freely available 12 months after publication, and remain freely available for 5 years.Non-open access articles that fall outside this five year window are available only to institutional subscribers and current ASPET members, or through the article purchase feature at the bottom of the page.
|