We developed a template-based system for predicting the regioselectivity of CYP2B6-mediated oxidation of chemicals. Two planar templates consisting of hexagonal blocks (A and B) were deduced from the overlapping regions of polyaromatic hydrocarbon-type substrates. Substrate atoms were placed only on the corners of the hexagonal blocks in the simulated interaction with the CYP2B6 templates. Through the application of various substrates having non-planar structures to Templates A and B, these templates were found to link to each other at specific positions (pinching points). The occupancy rates at each position of the templates were evaluated using more than 40 substrates, and a heavily utilized area (trigger region) was identified on Template A. This CYP2B6 template system is suggested to interact with substrates in at least three positions (trigger, pinching/bending and oxidation sites). In the present method, chemicals drawn as two or three-dimensional structures were directly overlaid on the templates to verify the feasibility of the simulated interaction. Both CYP2B6 substrates and non-substrates were applied to assess the validity of this template system. Results were consistent between the predicted and the in vitro experimental data with high accuracy, indicating the potential use of this system for studies on drug metabolism and new drug development.
The constitutive androstane receptor (CAR) is an orphan nuclear receptor which has been shown to participate in the activation of human CYP3A4, which metabolizes more than 50% of clinically used drugs. We investigated the effects of an array of compounds isolated from herbal medicines such as Rheum palmatum (Da Huang), Peucedanum praeruptorum Dunn (Qian Hu), Cortex Mori Radicis (Sang Bai Pi), Radix Asteris (Zi Wan), Salvia miltiorrhiza (Dan Shen), Polygonum cuspidatum Sieb. et Zucc (Hu Zhang), and Ginkgo biloba (Yin Xing) on the CAR-mediated transactivation of CYP3A4. The effect of herbal compounds on CYP3A4 expression was measured using a CYP3A4 luciferase reporter gene assay in transiently transfected human intestinal LS174T cells. The gene expression, protein expression, and catalytic activity of CYP3A4 in LS174T cells transfected with CAR were determined by using real-time PCR, Western blot analysis, and LC-MS/MS-based substrate assay. The study found that in CAR-transfected cells, praeruptorin A, C, and D significantly induced CYP3A4 luciferase activity, mRNA expression, and functional activity through the CAR-mediated pathway; conversely, induction was not found in untransfected cells. Our findings suggest that these herbal compounds can significantly up-regulate the CYP3A4 gene via the CAR-mediated pathway, which has important implications in herb–drug interactions.
Ferulic acid (FA), a member of the hydroxycinnamate family, is an abundant dietary antioxidant that may offer beneficial effects against cancer, cardiovascular disease, diabetes, osteoarthritis and Alzheimer's disease. In this study, evidence for sulfation and glucuronidation of FA was investigated upon incubation with human liver microsomes and cytosol. Two main glucuronides, M1 (ether O-glucuronide) and M2 (ester acylglucuronide), were formed with a similar affinity (apparent Km 3.53 and 5.15 mM, respectively). A phenol sulfoconjugate was also formed with a higher affinity (Km 0.53 mM). Identification of the UDP-glucuronosyltransferase (UGT) isoforms involved in FA glucuronidation was investigated with 12 human recombinant enzymes. FA was mainly glucuronidated by UGT1A isoforms and by UGT2B7. UGT1A4, 2B4, 2B15 and 2B17 failed to glucuronidate the substance. Examination of the kinetic constants revealed that FA was mainly glucuronidated by UGT1A1 at the two nucleophilic groups. UGT1A3 was able to glucuronidate these two positions with the same, but low, efficiency. UGT1A6 and 1A8 were involved in the formation of the ether glucuronide only, whereas UGT1A7, 1A10 and 2B7 preferentially glucuronidated the carboxyl group. Moreover, octyl gallate, a marker substrate of UGT1A1, competitively inhibited FA glucuronidation mediated by this isoform. Altogether, the results suggest that FA glucuronidation is primarily mediated by UGT1A1.
To predict CYP4A-mediated reactions, we developed a two-dimensional template scoring system based on published data. The system predicts the order of occurrence among multiple oxidation sites, as well as the regioselectivity. The template has a linearly arranged honeycomb shape and an adjacent area. Molecules are overlaid on the template with the locations of the atoms restricted to the corners of hexagonal blocks. The overlaid conformers are then checked to determine whether they reside within the template area, and their position occupancy and position function scores are calculated. The position occupancy score is determined based on occupation of the respective positions on the template. The functional and steric properties are reflected in the position function score. The sum of these scores is compared among possible conformers, and the conformer with the highest total score is predicted to be preferentially metabolized. In the present study, prediction of sites of CYP4A-mediated oxidation and classification into substrates and non-substrates were performed for collected compounds, and agreement between predicted and experimental data exceeded 95% for substrates and non-substrates. The template scoring system can be easily linked to databases of two-dimensional chemical structures, and thus this system may be useful for drug development and studies of drug metabolism.
The enzyme kinetic profiles of the formation of resveratrol-3-O-glucuronide (R3G) and resveratrol-4′-O-glucuronide (R4′G) by liver microsomes from humans, dogs, and rodents were investigated. Glucuronidation by human and dog liver microsomes to R3G and R4′G occurred for about 65% of applied resveratrol, and was significantly reduced to 10% when substrate concentration was increased 10-fold. In contrast, rodent microsomes glucuronidated about 90% of applied resveratrol independently of substrate concentration. Furthermore, in mouse and rat liver microsomes, resveratrol was almost exclusively conjugated at position 3, whereas human and dog livers also glucuronidated resveratrol at position 4′ (ratio R3G:R4′G = 5:1). Interspecies differences were also found when calculating the enzyme kinetic profiles of both conjugates. Formation of R4′G in human and dog microsomes followed Michaelis-Menten kinetics, while R3G showed substrate inhibition at higher resveratrol concentrations. In mouse and rat microsomes, however, both R3G and R4′G formation exhibited auto-activation kinetics. Formation of R3G and R4′G by recombinant UGT1A1 also showed substrate inhibition kinetics that led to decreased intrinsic clearance values, while UGT1A9-catalyzed glucuronidation demonstrated substrate inhibition kinetics at position 3 and Hill kinetics for the formation of R4′G. In conclusion, resveratrol glucuronidation exhibited species-dependent differences, with the dog as the animal model that most closely represents humans in terms of this process.
The present study was undertaken to identify genetic polymorphisms of multidrug resistance-associated protein 3 (MRP3, gene name ABCC3), an ATP-binding cassette transporter that mediates the transport of substrates across the basolateral membrane into the blood, and to investigate their effects on ABCC3 expression and MRP3 function. We identified genetic polymorphisms of ABCC3 and evaluated the effects by (1) a luciferase reporter gene assay, (2) measuring mRNA levels, and (3) a human pharmacogenomics study with 4-methylumbelliferone glucuronide (4-MUG). Overall, 61 genetic variants were identified in three ethnic populations; of these variants 17 were novel (7 were non-synonymous: 61Arg>Cys, 132Gln>Stop, 221Trp>Stop, 270His>Gln, 548Leu>Gln, 600Lys>Arg, and 1324Arg>His). However, these mutations occurred at very low frequencies (max. 4.7%). The observed allele frequencies showed considerable inter-ethnic differences. The reporter gene assay indicated a significant reduction of transcriptional activity with the −1767G>A allele compared to the wild-type allele; however, a decreased expression of ABCC3 mRNA was not detected in human liver samples. A human pharmacokinetic study showed that the ABCC3 genotype in the promoter region was not associated with changes in the pharmacokinetics of 4-MUG, a substrate of MRP3. This is the first study to assess the effects of ABCC3 polymorphisms on human pharmacokinetics; however, further investigations are needed to complete the picture.
Clinical studies have revealed that some fluoroquinolones may cause severe adverse effects when co-administered with substrates of CYP1A2. Our previous study showed antofloxacin (ATFX) was responsible for mechanism-based inhibition (MBI) of the metabolism of phenacetin in rats. In the clinical setting, ATFX is likely to be administrated with theophylline (TP), which is mainly metabolized by CYP1A2. The aim of the present study was to investigate the possible mechanism of TP/ATFX interaction. In vitro studies showed that the inhibitory effect of ATFX on the formation of three TP metabolites depended on NADPH, the pre-inhibition time, and ATFX concentration, i.e., factors which characterize MBI. In vivo studies demonstrated that single-dose ATFX (20 mg/kg) did not affect the pharmacokinetic behavior of TP, but multidose ATFX (20 mg/kg b.i.d. for 7.5 days) significantly increased the AUC of TP, decreased the amount of three TP metabolites in urine, and suppressed hepatic microsomal activity. A physiologically based pharmacokinetic (PBPK) model characterizing MBI of the three TP metabolites was developed for predicting TP/ATFX interaction in rats; this model was further extrapolated to humans. The predicted results were in good agreement with observed data. All the results indicated that ATFX was responsible for MBI of the metabolism of TP, and the PBPK model characterizing MBI may give good prediction of TP/ATFX interaction.
The widely used analgesic-antipyretic drug acetaminophen (APAP) is known to cause serious liver necrosis at high doses in man and experimental animals. For studies of toxic processes, 1H NMR spectroscopy of biofluids allows monitoring of endogenous metabolite profiles that alter characteristically in response to changes in physiological status. Herein, a 1H NMR metabolomics approach was applied to the investigation of APAP toxicity in rats and the effect of phenobarbital (PB) on APAP-induced hepatotoxicity. Metabolite differences due to hepatotoxicity were observed in 1H NMR spectra of serum and urine, and enhanced APAP hepatotoxicity by pretreatment with PB was clearly shown by a principal components analysis of the spectral data. NMR spectra of APAP-dosed rat urine provided profiles of APAP-related compounds together with endogenous metabolites. By comparison of endogenous and APAP-related metabolite spectra with those from rats pretreated with PB, it was possible to show the importance of oxidative metabolism of APAP to N-acetyl-p-benzoquinone, an essential step in APAP hepatotoxicity.
Cytochrome P450s (P450s) contribute to carcinogenesis by activating procarcinogens and also metabolize anti-cancer drugs. The activity and protein levels of P450s are important in cancer risk and in cancer therapy. In this study, we found that overexpression of CYP3A4 induced growth of a human hepatoma cell line, Hep3B. Overexpression of CYP2D6, by comparison, decreased cell growth. An inhibitor of CYP3A4, ketoconazole, significantly suppressed the growth of Hep3B cells overexpressing CYP3A4, but an inhibitor of CYP2D6, quinidine, did not restore Hep3B cell growth to baseline levels. Overexpression of CYP3A4 increased the production of reactive oxygen species, but this was not the cause of the CYP3A4-induced growth. Previously, we showed that CYP3A4 can produce epoxyeicosatrienoic acids (EETs) from arachidonic acid. The CYP3A4-enhanced cell growth was attenuated by a putative EET receptor antagonist, 14,15-EEZE. CYP3A4 promoted progression of the cell cycle from the G1 to the S phase. CYP3A4 also induced a hypoxic response of Hep3B cells, detected as enhanced erythropoietin gene expression (a typical hypoxic response). The cell growth promoted by CYP3A4 was inhibited by PI3K inhibitor LY294002. These results suggest that CYP3A4 plays an important role in tumor progression, independent of the activation of carcinogens and metabolism of anti-cancer drugs.
The present study was undertaken to clarify the possible association between nicotine intake/cigarette smoking and detrusor instability. For pharmacokinetic characterization of nicotine and cotinine (a major and pharmacologically less active metabolite of nicotine), a rapid ultra-performance liquid chromatography/electrospray ionization-mass spectrometry (UPLC/ESI-MS) method was developed that requires only a small amount of sample and simple pretreatment. The UPLC/ESI-MS method was validated with a focus on specificity, sensitivity (limit of detection, 2.5 ng/mL; limit of quantification, 5 ng/mL), linearity (r > 0.998), accuracy (97.2–102.8%), precision (relative standard deviation <8%) and robustness in accordance with ICH guidelines (Q2B Validation of Analytical Procedures: Methodology). The developed method was successfully applied to determine nicotine and cotinine levels in rat biological samples such as plasma, urine and several tissues. After subcutaneous administration of nicotine ditartrate (2 mg/kg of body weight) in rats, the absorbed nicotine was rapidly and extensively metabolized into cotinine. However, nicotine was found to be predominant in cortex and bladder, where nicotinic acetylcholine receptors were expressed for neuronal control of voiding function. Repeated administration of nicotine led to a ca. 3-fold higher accumulation of nicotine than that of cotinine in rat urine. The results of the pharmacokinetic study using the UPLC/ESI-MS method further support the possible involvement of nicotine in increased risk of urinary dysfunction in smokers.
Interspecies allometric scaling is a useful tool for calculating human pharmacokinetic (PK) parameters from data in animals. In this study, in order to determine the scaling exponent in a simple allometric equation that can predict human clearance (CL) and distribution volume at steady state (Vss) of monoclonal antibodies (mAbs) from monkey data alone, PK data of 24 mAbs were collected and analyzed according to the types of targeted antigens (soluble or membrane-bound antigens). Based on the observed PK data in humans (at clinical doses) and monkeys (at >1 mg/kg), where the PK is expected to be linear, the mean scaling exponents in the allometric equation for CL and Vss, respectively, against body weight were calculated to be 0.79 and 1.12 [95% confidence intervals (CIs): 0.69–0.89 and 0.96–1.28] for soluble antigens, and 0.96 and 1.00 (95% CIs: 0.83–1.09 and 0.87–1.13) for membrane-bound antigens. Using these exponents and monkey PK data (at >1 mg/kg) alone, both human CL and Vss of mAbs can be predicted with reasonable accuracy, i.e., within 2-fold of the observed values. Compared with traditional allometric scaling using PK data from three or more preclinical species, this approach is simple, quick, resource-saving, and useful in drug discovery and development.
The human aryl hydrocarbon receptor (AHR) is a protein for which there is little evidence of polymorphic variability of functional consequence. It has been hypothesized that potential variability in dioxin sensitivity may be due to polymorphisms in AHR-associated proteins, such as the human AHR-interacting protein (AIP). There are limited data on AIP single nucleotide polymorphisms (SNPs) with potential functional consequences. We sequenced 103 human DNA samples within the open reading frames of the AIP locus using samples from six ethnic populations to further characterize AIP SNPs. Eight exonic SNPs were identified at the AIP locus, including three novel SNPs: T48T, L212L, and V302V. Combined with prior reports, there are now a total of 14 exonic SNPs that have been identified within AIP. Of these, six are non-synonymous and are therefore of potential functional importance, though only two of these (Q228K and A276V) were detected in the current study. The functional consequences of Q228K and A276V are unknown, although functional evidence from AIP SNPs associated with congenital pituitary tumors suggests that such amino acid changes are likely to have no effect or to decrease, rather than increase, sensitivity to dioxins. To date, no non-synonymous SNPs have been detected in the AHR-binding region of AIP.
Drug Metabolism and Pharmacokinetics Vol. 26(1): 47–59 (2011) The name of one of the authors of this paper was printed incorrectly. The correct author list is as follows: Xiaochu Zhang, Feng Liu, Xin Chen, Xu Zhu and Jack Uetrecht