Abstracts


[Back to top]
A Multi-Endpoint Evaluation of Cytochrome P450 1A2, 2B6 and 3A4 Induction Response in Human Hepatocyte Cultures After Treatment with β - Naphthoflavone, Phenobarbital and Rifampicin
J. G. Zhang, Thuy Ho, Alanna L. Callendrello, Charles L. Crespi and David M. Stresser

US FDA and EMEA guidance recommend that the preferred in vitro model for cytochrome P450 induction testing is human hepatocytes coupled with acceptable inducers as controls. However, there are surprisingly few published studies characterizing this model system for dose and time-dependence response to model inducing compounds. The concentration-dependent response and time-course for the induction of CYP1A2, CYP2B6 and CYP3A4 by inducing agents β-naphthoflavone, phenobarbital and rifampicin, respectively were examined in two or more donors using multiple end-points (mRNA, enzyme activity and Western blot analysis). Depending on the endpoint, exposure time for maximal response of CYP induction potential for the three enzymes ranged from 24 to 72 hours. Of the concentrations of BNF, PB and RIF tested, those which gave the maximal response were found to be 33 µM, > 2 mM and 10 µM, respectively.

[Back to top]
In Vitro Screening of Metabolic Clearance Using Two Concentration Points
Eric T. Williams, Nadia Farah, Robert D. Pelletierand W. George Lai

For high throughput screens, the quickest methodology possible is desirable, but a substantial amount of potentially useful information is lacking since most screens for metabolic stability are conducted at one concentration, and sometimes at one time point. Information that would benefit projects during the discovery phase are to know the metabolic rate linearity (K_m value) and projected hepatic clearance (CL_h value), which is possible by the addition of one more concentration. This study used the FDA-preferred probe cytochrome P450 substrates to determine K_m, V_max, and CL_int values. The results showed that compounds with relatively high metabolic rates produced more accurate and reproducible results that match well with predicted K_m values according to the FDA. On the other hand, compounds with relatively low metabolic rates yielded more variable results. Thus, the use of two substrate concentrations should be useful with screening assays for assessing the kinetic values for other compounds.

[Back to top]
Allosteric Modulation of SULT2A1 by Celecoxib and Nimesulide: Computational Analyses
Emine Bihter Yalcin, Scott M. Struzik and Roberta S. King

We used protein-ligand docking and minimization to identify celecoxib as an allosteric modulator of SULT2A1-catalyzed estradiol sulfonation. Subsequent to celecoxib docking and complex minimization, conformational changes in SULT2A1 allowed estradiol docking to an alternative binding region with predicted preference for 17β-OH-E₂ sulfonation over 3-OH-E₂ sulfonation.

[Back to top]
Decreased Valproate Level Caused by VPA-glucuronidase Inhibition by Carbapenem Antibiotics
Yutaka Nakamura, Keiko Nakahira and Takaharu Mizutani

The serum concentration of valproic acid (VPA) in epilepsy patients decreased in the administration of carbapenem antibiotics (CP), such as meropenem, panipenem, biapenem or imipenem, to a sub-therapeutic level. The liver is the key organ for the decrease of VPA concentration by CP, because it has been reported that no decrease of the VPA level by CP was found in hepatectomized rats. This effect was also shown with monkey and rat liver slices. In this report, we show the results of in vitro inhibition of VPA-glucuronidase in human liver microsomes and cytosol by CP. We found the highest metabolic activity of VPA-glucuronide in human liver cytosol. The level in liver cytosol was 149 pmol/min/mg protein. The level in human liver microsomes (HLM) was one-fifth of that in cytosol and the level in serum was negligible. We found that this hydrolysis depends on VPA-glucuronidase in cytosol, because digestion was inhibited by D-saccharic acid 1,4-lactonemonohydrate of a specific inhibitor of β-glucuronidase, but not by phenylmethylsulfonylfluoride of an esterase inhibitor. We also found the inhibition of VPA-glucuronidase in cytosol by CP, and the maximum inhibition was found with panipenem (IC₅₀ = 3 μM). We also found inhibition of VPA-glucuronidase in HLM by meropenem. These results showed that the inhibition in liver slices depended on the inhibition of VPA-glucuronidase by CP. We considered that the inhibition of VPA-glucuronidase by CP in cytosol is a key factor to decrease the plasma VPA level.

[Back to top]
Evaluation of Cytochrome P450 Inhibition Assays Using Human Liver Microsomes by a Cassette Analysis /LC-MS/MS
Serenella Zambon, Stefano Fontana and Mahmud Kajbaf

In vitro inhibition assays are used to screen new chemical entities (NCEs) for inhibition studies by using human liver microsomes. High-throughput inhibition assays using pooling methods have been developed to keep pace with screening requirements at the lead ADME stage. This method can determine IC₅₀ NCEs using microsomes from various organs from any species. A cassette analysis inhibition assay for five of the major CYP enzymes (phenacetin for CYP1A2, diclofenac for CYP2C9, (S)-mephenytoin for CYP2C19, bufurarol for CYP2D6 and midazolam, nifedipine and atorvastatin for CYP3A4) in pooled human liver microsomes using ultraperformance liquid chromatography tandem mass spectrometry (LC/MS/MS) were developed. The major metabolites of seven CYP-specific probe substrates for the five P450 isoforms were monitored and quantified to determine IC₅₀ values. Human liver microsomal incubation samples at each test compound concentration were combined and analyzed simultaneously by the LC/MS/MS method. The incubation was performed using the selective CYP inhibitors for each isoform: fluvoxamine (CYP1A2), sulphaphenazole (CYP2C9), ticlopidine (CYP2C19), quinidine (CYP2D6), and Ketoconazole (CYP3A4). Similar IC₅₀ results were obtained using the cassette analysis and discrete analysis method.  The IC₅₀ values determined for typical CYP inhibitors were reproducible and consistent with those reported in the literature. The assay was fully automated in 96 well plate formats using Microlab 4000 series (Hamilton) coupled with two termomixer comfort (Eppendorf). An overall 70% time savings was achieved by pooling four isoforms samples (1A2, 2C9, 2C19 and 2D6) into one sample  and also  by pooling   three CYP 3A4 substrate samples (MDZ, ATR, NIF) into one sample. Cassette analysis minimized the number of injections on LC/MS/MS analysis which results in a decrease in the LC/MS/MS analysis time.

[Back to top]
Interaction Between Clopidogrel and Proton Pump Inhibitors: Hypothesis to Explain Multifactorial CYP2C19 Inhibition
Huixia Zhang, Isabelle Ragueneau-Majlessi and Rene H. Levy

Clopidogrel is an antiplatelet drug that requires bioactivation to its active metabolite to demonstrate its antiplatelet effect. Formation of the active metabolite involves multiple cytochrome P450 enzymes, with CYP2C19 playing an important role. Clopidogrel is often co-administered with proton pump inhibitors (PPIs) to decrease GI-tract bleeding, and decreased antiplatelet effect has been observed in these patients. This observation cannot be explained by the weak inhibitory effect of PPIs on CYP2C19. A hypothesis is proposed to interpret the phenomenon of PPI inhibition based in part on the finding that clopidogrel is itself an inhibitor of CYP2C19.

[Back to top]
Investigations Into the Drug-Drug Interaction Potential of Tapentadol in Human Liver Microsomes and Fresh Human Hepatocytes
Christa Kneip, Rolf Terlinden, Horst Beier and Genfu Chen

The new analgesic tapentadol was evaluated for induction and inhibition of several cytochrome P450 enzymes in vitro, and protein binding was assessed. It was concluded that no clinically relevant drug–drug interactions are likely to occur through either mechanism.

[Back to top]
P-Glycoprotein, but not Multidrug Resistance Protein 4, Plays a Role in the Systemic Clearance of Irinotecan and SN-38 in Mice
Michael Tagen, Yanli Zhuang, Fan Zhang, K. Elaine Harstead, Jun Shen, Paula Schaiquevich, Charles H. Fraga, John C. Panetta, Christopher M. Waters and Clinton F. Stewart

The ATP-binding cassette transporters P-glycoprotein (ABCB1, MDR1) and multidrug resistance protein 4 (MRP4) efflux irinotecan and its active metabolite SN-38 in vitro, and thus may contribute to system clearance of these compounds. Mdr1a/b^-/-, Mrp4^-/-, and wild-type mice were administered 20 or 40 mg/kg irinotecan, and plasma samples were collected for 6 hours. Irinotecan and SN-38 lactone and carboxylate were quantitated and data were analyzed with nonlinear mixed-effects modeling. Mdr1a/b genotype was a significant covariate for the clearance of both irinotecan lactone and SN-38 lactone. Exposures to irinotecan lactone and SN-38 lactone after a 40 mg/kg dose were 1.6-fold higher in Mdr1a/b^-/- mice compared to wild-type mice. Plasma concentrations of irinotecan lactone, irinotecan carboxylate, and SN-38 lactone in Mrp4^-/- mice were similar to the wild-type controls. These results suggest that P-gp plays a role in irinotecan and SN-38 elimination, but Mrp4 does not affect irinotecan or SN-38 plasma pharmacokinetics.

[Back to top]
Reversible Covalent Binding of Neratinib to Human Serum Albumin In Vitro
Appavu Chandrasekaran, Li Shen, Susan Lockhead, Aram Oganesian, Jianyao Wang and JoAnn Scatina

Neratinib (HKI-272), an irreversible inhibitor of Her 2 tyrosine kinase, is currently in development as an alternative for first and second line therapy in metastatic breast cancer patients who overexpress Her 2. Following incubation of [¹⁴C]neratinib in control human plasma at 37°C for 6 hours, about 60% to 70% of the radioactivity was not extractable, due to covalent binding to albumin. In this study, factors that could potentially affect the covalent binding of neratinib to plasma proteins, specifically to albumin were investigated. When [¹⁴C]neratinib was incubated at 10 µg/mL in human serum albumin (HSA) or control human plasma, the percent binding increased with time; the highest percentages of binding (46 and 67%, respectively) were observed at 6 hours, the longest duration of incubation examined. Binding increased with increasing temperature; the highest percentages of binding to HSA or human plasma (59 and 78%) were observed at 45°C, the highest temperature tested. The binding also increased with increasing pH of incubation; the highest percentages of binding (56 and 65%) were observed at pH 8.5, the highest pH value tested. The percentages of binding were similar (53% to 57%) when a wide range of concentrations of [¹⁴C]neratinib (50 ng/mL to 10 µg/mL) were incubated with human plasma at 37°C for 6 hours, indicating that the binding was independent of the substrate concentration, especially in the therapeutic range (50 to 200 ng/mL). When human plasma proteins containing covalently bound [¹⁴C]neratinb were suspended in a 10 fold volume of phosphate buffer at pH 4.0, 6.0, 7.4, and 8.5, and further incubated at 37°C for ~ 16 hours, about 45%, 44%, 32%, and 12% of the total radioactivity, respectively, was released as unchanged [¹⁴C]neratinib, indicating that the binding is reversible in nature, with more released at pH 7.4 and below. In conclusion, the covalent binding of neratinib to serum albumin is pH, time and temperature dependent, but not substrate concentration dependent, especially in the therapeutic range. Acidification and incubation of human plasma proteins that contained covalently bound [¹⁴C]neratinib leads to the release of the drug, indicating that the binding is reversible in nature. It is reasonable to speculate that the release of neratinib from human serum albumin provides a transport system leading to release of neratinib in the more acidic environment of the tumor.

[Back to top]
Two Branched Polar Groups and Polar Linker Moieties of Thiophene Amide Derivatives Are Essential for MRP2/ABCC2 Recognition
Yiding Hu, Kathleen E. Sampson, Li Xing, Na Li, Bruce R. Heyde, Jin Xie and Yurong Lai

Previously we demonstrated that the torsion angle between two biphenyl rings forming a three-dimensional conformation is the determinant factor for multi-drug resistance protein 2 (Mrp2/Abcc2) interaction [1]. More recently, we reported a heterocyclic compound, 1-(1-(4-bromophenyl)-3-carbamoyl-1H-pyrazol-4-yl) urea that shares the polar head groups with the biphenyl-substituted heterocycles, is highly secreted from bile by Mrp2/Abcc2 [2]. Collectively we hypothesized that the two branched polar groups and linkers might be essential with proposed Mrp2/Abcc2 recognition fitting in two primarily positive regions deep in the binding site. To test the hypothesis, a discovery lead compound (Compound 1) was examined to confirm the Mrp2/Abcc2 involvement resulting in hepatobiliary secretion in rats. The structural requirement of Mrp2/Abcc2 recognition was further explored in a series of thiophene amides derivatives divided into eight structural classes, with structural changes focused on the amide linker orientation or substitution from amide and sulfonamide to alkene, alkane, or alkyne linkers. In Caco-2 cell bidirectional transport assays and Mrp2/Abcc2 membrane vesicle uptake assays, the involvement of Mrp2/Abcc2 mediated transport was confirmed in structural classes 1 – 5, which contains polar amide or sulfonamide linker, but not in classes 6 – 8 with non-polar aliphatic linker. The Mrp2/Abcc2 recognition showed strong correlation with structural descriptors in predictive Bayesian model, as well as with polar surface area and lipophilicity (LogP). The result provided valuable information for predicting transporter recognition in silico, for improved predictions of transporter-involved ADME in early drug discovery.