Tissue-specific regulation of canine intestinal and hepatic phenol and morphine UDP-glucuronosyltransferases by beta-naphthoflavone in comparison with humans

The Past, Present and Future of Intestinal In Vitro Cell Systems for Drug Absorption Studies

The intestinal epithelium acts as a selective barrier for the absorption of water, nutrients and orally administered drugs. To evaluate the gastrointestinal permeability of a candidate molecule, scientists and drug developers have a multitude of cell culture models at their disposal. Static transwell cultures constitute the most extensively characterized intestinal in vitro system and can accurately categorize molecules into low, intermediate and high permeability compounds. However, they lack key aspects of intestinal physiology, including the cellular complexity of the intestinal epithelium, flow, mechanical strain, or interactions with intestinal mucus and microbes. To emulate these features, a variety of different culture paradigms, including microfluidic chips, organoids and intestinal slice cultures have been developed. Here, we provide an updated overview of intestinal in vitro cell culture systems and critically review their suitability for drug absorption studies. The available data show that these advanced culture models offer impressive possibilities for emulating intestinal complexity. However, there is a paucity of systematic absorption studies and benchmarking data and it remains unclear whether the increase in model complexity and costs translates into improved drug permeability predictions. In the absence of such data, conventional static transwell cultures remain the current gold-standard paradigm for drug absorption studies.

AhR regulates the expression of human cytochrome P450 1A1 (CYP1A1) by recruiting Sp1

Cytochrome P450 1A1 (CYP1A1) is abundant in the kidney, liver, and intestine and is involved in the phase I metabolism of numerous endogenous and exogenous compounds. Therefore, exploring the regulatory mechanism of its basal expression in humans is particularly important to understand the bioactivation of several procarcinogens to their carcinogenic derivatives. Site-specific mutagenesis and deletion of the transcription factor binding site determined the core cis-acting elements in the human CYP1A1 proximal and distal promoter regions. The proximal promoter region [overlapping xenobiotic-responsive element (XRE) and GC box sequences] determined the basal expression of CYP1A1. In human hepatocellular carcinoma cells (HepG2) with aryl hydrocarbon receptor (AhR) or specificity protein 1 (Sp1) knockdown, we confirmed that AhR and Sp1 are involved in basal CYP1A1 expression. In HepG2 cells overexpressing either AhR or Sp1, AhR determined the proximal transactivation of basal CYP1A1 expression. Via DNA affinity precipitation assays and ChIP, we found that AhR bound to the promoter and recruited Sp1 to transactivate CYP1A1 expression. The coordinated interaction between Sp1 and AhR was identified to be DNA mediated. Our work revealed a basal regulatory mechanism of an interesting human gene by which AhR interacts with Sp1 through DNA and recruits Sp1 to regulate basal CYP1A1 expression.

Utilizing In Vitro Dissolution-Permeation Chamber for the Quantitative Prediction of pH-Dependent Drug-Drug Interactions with Acid-Reducing Agents: a Comparison with Physiologically Based Pharmacokinetic Modeling

For many orally administered basic drugs with pH-dependent solubility, concurrent administration with acid-reducing agents (ARAs) can significantly impair their absorption and exposure. In this study, pH-dependent drug-drug interaction (DDI) prediction methods, including in vitro dissolution-permeation chamber (IVDP) and physiologically based pharmacokinetic (PBPK) modeling, were evaluated for their ability to quantitatively predict the clinical DDI observations using 11 drugs with known clinical pH-dependent DDI data. The data generated by IVDP, which consists of a gastrointestinal compartment and a systemic compartment separated by a biomimic membrane, significantly correlated with the clinical DDI observations. The gastrointestinal compartment AUC ratio showed strong correlation with clinical AUC ratio (R=0.72 and P=0.0056), and systemic compartment AUC ratio showed strong correlation with clinical C_max ratio (R=0.91 and P=0.0003). PBPK models were also developed for the 11 test compounds. The simulations showed that the predictions from PBPK model with experimentally measured parameters significantly correlated with the clinical DDI observations. Future studies are needed to evaluate predictability of Z-factor-based PBPK models for pH-dependent DDI. Overall, these data suggested that the severity of pH-dependent DDI can be predicted by in vitro and in silico methods. Proper utilization of these methods before clinical DDI studies could allow adequate anticipation of pH-dependent DDI, which helps with minimizing pharmacokinetic variation in clinical studies and ensuring every patient with life-threatening diseases receives full benefit of the therapy.

Assessment of UDP-glucuronosyltransferase catalyzed formation of Picroside II glucuronide in microsomes of different species and recombinant UGTs

This study compared the hepatic glucuronidation of Picroside II in different species and characterized the glucuronidation activities of human intestinal microsomes (HIMs) and recombinant human UDP-glucuronosyltransferases (UGTs) for Picroside II. The rank order of hepatic microsomal glucuronidation activity of Picroside II was rat > mouse > human > dog. The intrinsic clearance of Picroside II hepatic glucuronidation in rat, mouse and dog was about 10.6-, 6.0- and 2.3-fold of that in human, respectively. Among the 12 recombinant human UGTs, UGT1A7, UGT1A8, UGT1A9 and UGT1A10 catalyzed the glucuronidation. UGT1A10, which are expressed in extrahepatic tissues, showed the highest activity of Picroside II glucuronidation (K(m) = 45.1 μM, V(max) = 831.9 pmol/min/mg protein). UGT1A9 played a primary role in glucuronidation in human liver microsomes (HLM; K(m) = 81.3 μM, V(max) = 242.2 pmol/min/mg protein). In addition, both mycophenolic acid (substrate of UGT1A9) and emodin (substrate of UGT1A8 and UGT1A10) could inhibit the glucuronidation of Picroside II with the half maximal inhibitory concentration (IC(50)) values of 173.6 and 76.2 μM, respectively. Enzyme kinetics was also performed in HIMs. The K(m) value of Picroside II glucuronidation was close to that in recombinant human UGT1A10 (K(m) = 58.6 μM, V(max) = 721.4 pmol/min/mg protein). The intrinsic clearance was 5.4-fold of HLMs. Intestinal UGT enzymes play an important role in Picroside II glucuronidation in human.

Species comparison of oral bioavailability, first-pass metabolism and pharmacokinetics of acetaminophen

Species differences in oral bioavailability, first-pass metabolism and pharmacokinetics of biopharmaceutics classification system (BCS) class I compound acetaminophen were studied. The absolute bioavailability was 42.2%, 39.0%, 44.5%, 75.5% and 91.0% in chickens, turkeys, dogs, pigs and horses, respectively. After hydrolysis of metabolites by beta-glucuronidase/sulfatase, apparent bioavailability increased significantly in all species (turkeys: 72.4%, dogs: 100.5%, pigs: 102.2%), except horses (91.6%). Mean metabolic ratios of [acetaminophen glucuronide]/[acetaminophen] between 0 and 1h were significantly higher after oral dosing in turkeys, dogs and pigs, revealing the role of first-pass metabolism in incomplete bioavailability. Evidence of species differences in acetaminophen metabolism is provided by differences in plasma clearance, which was inversely proportional to bioavailability. In conclusion, differences in BA appeared to originate predominantly from differences in first-pass metabolism, demonstrating that the BCS high permeability classification of acetaminophen is consistent across the mammalian species studied. In turkeys, however, incomplete absorption additionally seemed to contribute to the low BA.

Gastrointestinally Distributed UDP-glucuronosyltransferase 1A10, Which Metabolizes Estrogens and Nonsteroidal Anti-inflammatory Drugs, Depends upon Phosphorylation

Among gastrointestinal distributed isozymes encoded at the UGT1 locus, UDP-glucuronosyltransferase 1A10 (UGT1A10) metabolizes a number of important chemicals. Similar to broad conversion of phytoestrogens (Basu, N. K., Ciotti, M., Hwang, M. S., Kole, L., Mitra, P. S., Cho, J. W., and Owens, I. S. (2004) J. Biol. Chem. 279, 1429-1441), UGT1A10 metabolized estrogens and their derivatives, whereas UGT1A1, -1A3, -1A7, and -1A8 differentially exhibited reduced activity toward the same. UGT1A10 compared with UGT1A7, -1A8, and -1A3 generally exhibited high activity toward acidic nonsteroidal anti-inflammatory drugs and natural benzaldehyde derivatives, while UGT1A3 metabolized most efficiently aromatic transcinnamic acids known to be generated from flavonoid glycosides by microflora in the lower gastrointestinal tract. Finally UGT1A10, -1A7, -1A8, and -1A3 converted plant-based salicylic acids; methylsalicylic acid was transformed at high levels, and acetylsalicylic (aspirin) and salicylic acid were transformed at moderate to low levels. Atypically UGT1A10 transformed estrogens between pH 6 and 8 but acidic structures preferentially at pH 6.4. Furthermore evidence indicates UGT1A10 expressed in COS-1 cells depends upon phosphorylation; UGT1A10 versus its single, double, and triple mutants at three predicted protein kinase C phosphorylation sites incorporated [(33)P]-orthophosphate and showed a progressive decrease with no detectable label or activity for the triple T73A/T202A/S432G-1A10 mutant. Single and double mutants revealed either null/full activity or null/additive activity, respectively. Additionally UGT1A10-expressing cultures glucuronidated 17beta-[(14)C]estradiol, whereas cultures containing null mutants at protein kinase C sites showed no estrogen conversion. Importantly UGT1A10 in cells supported 10-fold higher glucuronidation of 17beta-estradiol than UGT1A1. In summary, our results suggest gastrointestinally distributed UGT1A10 is important for detoxifying estrogens/phytoestrogens and aromatic acids with complementary activity by UGT1A7, -1A8, -1A3, and/or -1A1 evidently dependent upon phosphorylation.

HUMAN UDP-GLUCURONOSYLTRANSFERASES SHOW ATYPICAL METABOLISM OF MYCOPHENOLIC ACID AND INHIBITION BY CURCUMIN

Although the promising immunosuppressant, mycophenolic acid (MPA), has many desirable properties and is widely prescribed for organ transplant recipients, its high oral dosage requirement is not understood. Whereas previous Northern blot analysis by Basu and colleagues (2004) located the mRNAs encoding MPA primary metabolizers, UDP-glucuronosyltransferases (UGTs) 1A7, 1A8, 1A9, and 1A10, in human gastrointestinal (GI) tissues, in situ hybridization analysis of mRNAs found that the isozymes were restricted to the mucosal layer of various GI organs. Concomitantly, MPA was glucuronidated by microsomes isolated from normal adjoining specimens. Microsomal studies showed the highest relative rates of metabolism in esophagus, ileum, duodenum, colon, and stomach at pH 6.4; only esophagus showed high pH 7.6 activity. Properties of the recombinant UGTs indicate that MPA is metabolized with pH 6.4 or 7.6 optimum. Activity for 1A7 and 1A9 increased with increasing concentrations up to 2.4 mM, with parallel production of both ether- and acylglucuronides; similarly, 1A8 and 1A10 reached plateaus at 1.6 mM, producing both glucuronides. K(m) values were 250 to 550 microM. Between 400 and 1600 microM MPA, isozymes generated between 15 and 42% of the acylglucuronides. In effect, high K(m) values (MPA) are associated with high concentrations to achieve saturation kinetics. Finally, transient inhibition of UGTs in human LS180 colon cells and mouse duodenum by the dietary agent, curcumin, has implications for in vivo pretreatment to reduce MPA glucuronidation to increase the therapeutic index.

AN ASSESSMENT OF UDP-GLUCURONOSYLTRANSFERASE INDUCTION USING PRIMARY HUMAN HEPATOCYTES

Uridine diphosphate glucuronosyltransferases (UGTs) catalyze the glucuronidation of a wide range of xenobiotics and endogenous substrates. However, there is a lack of information concerning the response of human UGTs to inducers, and this observation prompted the current investigation. The glucuronidation of estradiol (3- and 17-positions), naphthol, propofol, and morphine (3- and 6-positions) was assessed against a battery of recombinant human UGTs to determine selective glucuronidation reactions for induction studies. The potential induction of the glucuronidation of estradiol at the 3-position, naphthol, propofol, and morphine at the 3-position was subsequently investigated in cultured primary human hepatocytes against a range of prototypic inducers including dexamethasone, 3-methylcholanthrene (3-MC), phenobarbital, rifampicin, and omeprazole. Treatment with 3-MC induced estradiol-3-glucuronidation (up to 2.5-fold) in four of five donors investigated. Statistically significant increases in naphthol glucuronidation (up to 1.7-fold) were observed following treatment with carbamazepine. UGT1A9-mediated propofol glucuronidation was induced by phenobarbital (up to 2.2-fold) and rifampicin (up to 1.7-fold). However, treatment with alpha-naphthoflavone and tangeretin resulted in a decrease in propofol glucuronidation (30% of control values). Statistically significant induction of morphine-3-glucuronidation was observed in at least three donors following treatment with phenobarbital, rifampicin, and carbamazepine. Each UGT isoform investigated displayed a distinct induction profile. Although statistically significant increases in glucuronidation were observed for each reaction studied, the level of induction was less than that observed for CYP1A2 or CYP3A4 and exhibited a large interdonor variability. The clinical relevance of the induction responses obtained in this study is unclear.

Differential and special properties of the major human UGT1-encoded gastrointestinal UDP-glucuronosyltransferases enhance potential to control chemical uptake

UDP-glucuronosyltransferase (UGT) isozymes detoxify metabolites, drugs, toxins, and environmental chemicals via conjugation to glucuronic acid. Based on the extended UGT1 locus combined with Northern blot analysis and in situ hybridization, we determined the distribution of UGT1A1 and UGT1A7 through UGT1A10 mRNAs and found them for the first time segmentally distributed in the mucosal epithelia layer of the gastrointestinal tract. Biochemically, recombinant isozymes exhibited pH optima of 5.5, 6.4, 7.6, 8.5, and/or a broad pH range, and activities were found to be unaffected or progressively inhibited by increasing substrate concentrations after attaining Vmax for certain chemicals. Under different optimal conditions, all exhibited wide substrate selections for dietary and environmentally associated chemicals. Evidence also suggests tandem effects of isozymes in the time for completion of reactions when comparing short- and long-term incubations. Moreover, treatment of colon cells with certain diet-associated constituents, curcumin and nordihydroguaiaretic acid, reversibly targets UGTs causing inhibition without affecting protein levels; there is no direct inhibition of control UGT using curcumin as substrate in the in vitro assay. In summary, we demonstrate that UGTs are located in gastrointestinal mucosa, have vast overlapping activities under differential optimal conditions, and exhibit marked sensitivity to certain dietary substrates/constituents, representing a first comprehensive study of critical properties concerning glucuronidating isozymes in alimentary tissues. Additionally, the highly dynamic, complex, and variable properties necessarily impact absorption of ingested chemicals and therapeutic drugs.

Cloning and characterisation of the first drug-metabolising canine UDP-glucuronosyltransferase of the 2B subfamily

Glucuronidation is a major route of clearance for a diverse set of both drug and endogenous substrates. The present study was undertaken to redress the lack of molecular information currently available on drug glucuronidation by the dog, a species widely used in metabolism studies by the pharmaceutical industry. A novel dog uridine diphosphate glucuronosyltransferase (UGT), designated UGT2B31 (GenBank Accession Number: AY135176), has been isolated from a dog cDNA library, expressed in V79 cells and characterised using various methods: (i) UGT2B31 sequence has been compared with mammalian UGT sequences using both sequence alignments and phylogenetic analysis; and (ii) the substrate specificity of UGT2B31 has been determined using functional analysis and compared with that obtained using UGT2B7 and dog liver microsomes. The following results were obtained: (i) sequence alignments between UGT2B31 and UGT2B15 gave the greatest degree of identity (76%); however, human UGT2B4, human UGT2B7, monkey UGT2B9 (all 75%), and rat UGT2B1 (73%) also gave a high degree of identity; (ii) phylogenetic analysis determined UGT2B31 to be most closely related to rat UGT2B1; (iii) UGT2B31 displayed a substrate specificity similar to human UGT2B7 and rat UGT2B1, catalysing the glucuronidation of phenols, opioids, and carboxylic acid-containing drugs; and (iv) UGT2B31 only formed morphine-3-glucuronide; however, kinetic analysis determined the K(m) of this reaction to be similar to that observed with UGT2B7 (both approximately 1300 microM). The results suggest that UGT2B31 plays a crucial role in drug detoxification by the dog and may be the canine equivalent of human UGT2B7.

Glucuronides of tea catechins: enzymology of biosynthesis and biological activities

(-)-Epigallocatechin gallate (EGCG) and (-)-epigallocatechin (EGC) are major green tea catechins with antioxidant and anticancer activities. In this study, we characterized the glucuronidation of EGCG and EGC in human, mouse, and rat microsomes and by nine different human UGT 1A and 2B isozymes expressed in insect cells. Six EGCG and EGC glucuronides were biosynthesized, and their structures were identified for the first time. (-)-EGCG-4"-O-glucuronide was the major EGCG glucuronide formed in all incubations. The catalytic efficiency (V(max)/K(m)) for (-)-EGCG-4"-O-glucuronide formation followed the order: mouse intestine > mouse liver > human liver > rat liver >> rat small intestine. The UGT-catalyzed glucuronidation of EGC was much lower than that of EGCG. The V(max)/K(m) for (-)-EGC-3'-O-glucuronide followed the following order: mouse liver > human liver > rat liver > rat and mouse small intestine. Human UGT1A1, 1A8, and 1A9 had high activities with EGCG. UGT1A8, an intestine-specific UGT, had the highest V(max)/K(m) for EGCG but low activity with EGC. Mice appeared to be more similar to humans than rats to humans in the glucuronidation of EGCG and EGC. Some of these catechin glucuronides retained the activities of their parent compounds in radical scavenging and in inhibiting the release of arachidonic acid from HT-29 human colon cancer cells. These results provide foundations for understanding the biotransformation and biological activities of tea catechins.