Quaternary ammonium-linked glucuronidation of tamoxifen by human liver microsomes and UDP-glucuronosyltransferase 1A4

Characterization of tamoxifen and 4-hydroxytamoxifen glucuronidation by human UGT1A4 variants

Introduction

Tamoxifen (TAM) is an antiestrogen widely used in the treatment and prevention of breast cancer in women. One of the major mechanisms of metabolism of TAM and one of its major active metabolites, 4-hydroxytamoxifen (4-OH-TAM), is via glucuronidation. In the present study, the glucuronidating activities of three common variant isoforms encoded by the human UDP-glucuronosyltransferase (UGT) 1A4 gene were examined against TAM, trans-4-OH-TAM and cis-4-OH-TAM.

Methods

HPLC was used to detect glucuronide conjugates in microsomes from UGT1A4-overexpressing HK293 cells. The UGT1A4 wild-type cDNA was synthesized by RT-PCR using normal human liver total RNA. The UGT1A4^24Thr/48Leu and UGT1A4^24Pro/48Val variants were generated by site-directed mutagenesis of the pcDNA3.1/V5-His-TOPO plasmid expressing wild-type UGT1A4^24Pro/48Leu. Levels of UGT1A4 expression in UGT-overexpressing cell lines were measured by western blot analysis.

Results

Microsomes from wild-type UGT1A4^24Pro/48Leu-overexpressing HK293 cells exhibited significant levels of activity against TAM, trans-4-OH-TAM and cis-4-OH-TAM, forming exclusively the tamoxifen quaternary ammonium glucuronide (TAM-N⁺-glucuronide) and the 4-hydroxytamoxifen quaternary ammonium glucuronides (trans-4-OH-TAM-N⁺-glucuronide and cis-4-OH-TAM-N⁺-glucuronide) with apparent K_m values of 2.0 μM, 2.2 μM, and 2.1 μM, respectively. Higher glucuronidation activities were found by kinetic analysis for microsomes from the variant UGT1A4^24Pro/48Val-overexpressing cell line as compared with microsomes from wild-type UGT1A4^24Pro/48Leu-overexpressing cells against TAM and against both the trans and cis isomers of 4-OH-TAM. A significantly (P < 0.02) lower K_m value (~1.6-fold to 1.8-fold) was observed for both 4-OH-TAM isomers, while a near-significant (P = 0.053) decrease in K_m was observed for TAM for the UGT1A4^24Pro/48Val variant as compared with wild-type UGT1A4. The V_max/K_m ratio for the UGT1A4^24Pro/48Val variant was significantly (P ≤ 0.005) higher than that observed for the wild-type UGT1A4 isoform for both the trans and cis isomers of 4-OH-TAM after normalization for UGT1A4 expression by western blotting. No significant effect on enzyme kinetics was observed for the UGT1A4^24Thr/48Leu variant against either isomer of 4-OH-TAM or with TAM.

Conclusion

These data suggest that the UGT1A4 codon 48 Leu>Val polymorphism significantly alters glucuronidation rates against TAM and its active hydroxylated metabolites, and that this polymorphism may play an important role in individual pharmacological response to TAM therapy.

On-Line Drug Metabolism in Capillary Electrophoresis. 1. Glucuronidation Using Rat Liver Microsomes

A rat liver microsome pseudostationary phase has been used for the on-line capillary electrophoresis monitoring of glucuronidation. Uridine diphosphate glucuronosyltransferase (EC 2.4.1.17) containing microsomes was isolated from rat liver and directly injected onto neutrally coated capillary containing polymeric replaceable gels followed by injection of the substrate mixture. On-line glucuronidation was observed within 15 min without any sample preparation. The factors affecting the separation of glucuronides and parent compounds were investigated by varying the applied electric fields and the size (length and internal diameter) of capillary. The Michaelis-Menten parameters (Km and Vmax) for the glucuronidation of 4-methyl-7-hydroxy coumarin and 4-nitrophenol were determined using the CE method and by off-line microsomal incubation. No significant differences were observed for Km and Vmax values for 4-methyl-7-hydroxycoumarin and 4-nitrophenol between on-line and off-line glucuronidation of these two compounds. This method was also used to determine the inhibition constant (IC50 value) for the competitive inhibition of morphine glucuronidation by codeine, IC50 (on-line) = 170 vs 580 microM (off-line). The results demonstrate that this method can be used to screen for the glucuronidation of test compounds and should reduce the time required for this screening process.

Uridine diphosphoglucuronosyltransferase pharmacogenetics and cancer

The uridine diphosphoglucuronosyltransferases (UGTs) belong to a superfamily of enzymes that catalyse the glucuronidation of numerous endobiotics and xenobiotics. Several human hepatic and extrahepatic UGT isozymes have been characterized with respect to their substrate specificity, tissue expression and gene structure. Genetic polymorphisms have been identified for almost all the UGT family members. A wide variety of anticancer drugs, dietary chemopreventives and carcinogens are known to be conjugated by members of both UGT1A and UGT2B subfamilies. This review examines in detail each UGT isozyme known to be associated with cancer and carcinogenesis. The cancer-related substrates for several UGTs are summarized, and the functionally relevant genetic polymorphisms of UGTs are reviewed. A number of genotype-phenotype association studies have been carried out to characterize the role of UGT pharmacogenetics in several types of cancer, and these examples are discussed here. In summary, this review focuses on the role of the human UGT genetic polymorphisms in carcinogenesis, chemoprevention and cancer risk.

In vitro methods in human drug biotransformation research: implications for cancer chemotherapy

Anticancer drugs have a complex pharmacological and toxicological profile with a narrow therapeutic index. It is therefore critical to understand the factors that contribute to the marked intersubject variability in the pharmacokinetics and pharmacodynamics often observed with anticancer compounds. Since hepatic and extra-hepatic drug metabolism represents a major drug disposition pathway, extensive efforts are made to thoroughly investigate metabolism of anticancer compounds during the pre-clinical and clinical development phases as well as to address issues encountered during the clinical use of an approved drug. In recent years there has been a significant paradigm shift in pre-clinical/non-clinical drug metabolism studies. Most importantly, this has included a reduced reliance on animal models and increased use of human tissues (i.e. human liver microsomes and other cellular fractions, primary culture of human hepatocytes, cDNA expressed human-specific enzymes and cell-based reporter assays). Typically, experiments are performed using these tools to identify the phase I and/or phase II enzymes involved in metabolism of the drug/investigational agent and for metabolic fingerprinting. Additionally, issues pertaining to the rate, extent and mechanism(s) of the inhibition or induction of the metabolic pathways are also investigated. These studies provide important clues about various aspects of the disposition of a therapeutic agent including first-pass metabolism, elimination half-life, overall bioavailability and the potential for drug-drug interactions. The methodologies used for in vitro assessment of drug metabolism and their applications to drug development and clinical therapeutics with special emphasis on anticancer drugs are reviewed in this manuscript.

Quaternary ammonium-linked glucuronidation of trans-4-hydroxytamoxifen, an active metabolite of tamoxifen, by human liver microsomes and UDP-glucuronosyltransferase 1A4

Tamoxifen (TAM), a nonsteroidal antiestrogen, is the most widely used drug for chemotherapy of hormone-dependent breast cancer in women. Trans-4-hydroxy-TAM (trans-4-HO-TAM), one of the TAM metabolites in humans, has been considered to be an active metabolite of TAM because of its higher affinity toward estrogen receptors (ERs) than the parent drug and other side-chain metabolites. In the present study, we found a new potential metabolic pathway of trans-4-HO-TAM and its geometrical isomer, cis-4-HO-TAM, via N-linked glucuronic acid conjugation for excretion in humans. N+-Glucuronides of 4-HO-TAM isomers were isolated along with O-glucuronides from a reaction mixture consisting of trans- or cis-4-HO-TAM and human liver microsomes fortified with UDP-glucuronic acid and identified with their respective synthetic specimens by high performance liquid chromatography-electrospray ionization time-of-flight mass spectrometry. Although N- and O-glucuronidating activities of human liver microsomes toward trans-4-HO-TAM were nearly comparable, O-glucuronidation was predominant for cis-4-HO-TAM conjugation. Only UGT1A4 catalyzed the N-linked glucuronidation of 4-HO-TAM among recombinant human UGT isoforms (UGT1A1, UGT1A3, UGT1A4, UGT1A6, UGT1A7, UGT1A8, UGT1A9, UGT1A10, UGT2B4, UGT2B7, UGT2B15, and UGT2B17) expressed in insect cells. In contrast, all UGT isoforms, except for UGT1A3 and UGT1A4, catalyzed O-glucuronidation of 4-HO-TAM. Although O-glucuronidation of 4-HO-TAM greatly decreased binding affinity for human ERs, 4-HO-TAM N+-glucuronide still had binding affinity similar to 4-HO-TAM itself, suggesting that N+-glucuronide might contribute to the biological activity of TAM in vivo.

Association of genetic variation in tamoxifen-metabolizing enzymes with overall survival and recurrence of disease in breast cancer patients

Tamoxifen has been a mainstay of adjuvant therapy for breast cancer for many years. We sought to determine if genetic variability in the tamoxifen metabolic pathway influenced overall survival in breast cancer patients treated with tamoxifen. We examined functional polymorphisms in CYP2D6, the P450 catalyzing the formation of active tamoxifen metabolites, and UGT2B15, a Phase II enzyme facilitating the elimination of active metabolite in a retrospective study of breast cancer patients. We also examined whether the combination of variant alleles in SULT1A1 and UGT2B15 had more of an impact on overall survival in tamoxifen-treated patients than when the genes were examined separately. We conducted a retrospective study using archived paraffin blocks for DNA extraction and data from pathology reports and hospital tumor registry data for information on clinical characteristics, treatment, and outcomes (162 patients receiving tamoxifen and 175 who did not). Genotypes for CYP2D6 and UGT2B15 were obtained and Cox proportional hazards modeling was performed. After adjusting for age, race, stage of disease at diagnosis, and hormone receptor status, we found no significant association between CYP2D6 genotype and overall survival in either group of breast cancer patients. Tamoxifen-treated patients with UGT2B15 high activity genotypes had increased risk of recurrence and poorer survival. When UGT2B15 and SULT1A1 'at-risk' alleles were combined, women with two variant alleles had significantly greater risk of recurrence and poorer survival than those with common alleles. These studies indicate that genetic variation in Phase II conjugating enzymes can influence the efficacy of tamoxifen therapy for breast cancer.

Heterologous expression of active human uridine diphosphate glucuronosyltransferase 1A3 in Chinese hamster lung cells

AIM: To obtain the active human recombinant uridine diphosphate glucuronosyltransferase 1A3 (UGT1A3) enzyme from Chinese hamster lung (CHL) cells.

METHODS: The full-length UGT1A3 gene was amplified by reverse transcription-polymerase chain reaction (RT-PCR) using total RNA from human liver as template. The correct fragment confirmed by sequencing was subcloned into the mammalian expression vector pcDNA3.1 (+), and the recombinant vector was transfected into CHL cells using a calcium phosphate method. Expressed UGT1A3 protein was prepared from CHL cells resistant to neomycin (G418). Then the protein was added into a reaction mixture for glucuronidation of quercetin. The glucuronidation activity of UGT1A3 was determined by reverse phase-high performance liquid chromatography (RP-HPLC) coupled with a diode array detector (DAD). The quercetin glucuronide was confirmed by hydrolysis with β-glucuronidase. Control experiments were performed in parallel. The transcriptions of recombinants were also determined by RT-PCR.

RESULTS: The gene was confirmed to be an allele (UGT1A3-3) of UGT1A3 by DNA sequencing. The fragment was introduced into pcDNA3.1 (+) successfully. Several colonies were obtained under the selection pressure of G418. The result of RT-PCR showed transcription of recombinants in mRNA level. Glucuronidation assay and HPLC analysis indicated UGT1A3 expressed heterologously in CHL cells was in an active form, and one of the gulcuronides corresponding to quercetin was also detected.

CONCLUSION: Correct sequence of UGT1A3 gene can be obtained, and active UGT1A3 enzyme is expressed heterologously in CHL cells.

Glucuronidation of 2-(4-Chlorophenyl)-5-(2-Furyl)-4-Oxazoleacetic Acid (TA-1801A) in Humans: Species Differences in Liver and Intestinal Microsomes

The metabolism of ethyl 2-(4-chlorophenyl)-5-(2-furyl)-4-oxazoleacetate (TA-1801), a potent hypolipidemic agent, was studied in humans after oral administration and compared with that found in rats, rabbits, and dogs previously. Hydrolysis of the ethyl ester to produce metabolite M1 (TA-1801 active form; TA-1801A) is the first metabolic step and the subsequent biotransformation includes the glucuronidation to form the metabolite M4 and the oxidation to form the metabolites M2 and M3. The metabolism of TA-1801 in humans was qualitatively similar to that in the experimental animals studied, although species differences were seen in the amount of metabolites. M4, the glucuronide of TA-1801A was the most abundant metabolite in human urine (24.3% of the dose). In vitro studies using human liver and jejunum microsomes indicated that the TA-1801A glucuronosyltransferase activity in human jejunum microsomes was 2-fold higher than that in liver microsomes. With regard to the interspecies differences in the TA-1801A glucuronosyltransferase activities, the intrinsic clearance for the TA-1801A glucuronidation in liver microsomes was in the following order: rabbit>monkey>human=rat=dog. In jejunum microsomes, the intrinsic clearance for the TA-1801A glucuronidation was in the following order: human>monkey>rabbit>rat=dog. These results suggest that the species differences in the intestinal TA-1801A glucuronidation contribute to the species differences in the excretion rate of TA-1801A glucuronide into the urine.

Substrate Specificity of Human Hepatic Udp‐Glucuronosyltransferases

Five human hepatic UDP-glucuronosyltransferases (UGTs) catalyze the facilitated excretion of more than 90% of drugs eliminated by glucuronidation. The substrate specificity of these UGTs has been examined using cloned expressed enzymes and liquid chromatography-mass spectrometry assays to determine the intrinsic clearance of drug glucuronidation in vitro. Specific substrates for the five individual UGTs have been identified. These five probe substrates could be used to predict the drug clearance catalyzed by individual UGTs in vivo.

Genetic Variations and Haplotypes of UGT1A4 in a Japanese Population

Nineteen genetic variations, including 11 novel ones, were found in exon 1 and its flanking region of the UDP-glucuronosyltransferase (UGT) 1A4 gene from 256 Japanese subjects, consisting of 60 healthy volunteers, 88 cancer patients and 108 arrhythmic patients. These variations include -217T>G and -36G>A in the 5'-flanking region, 30G>A (P10P), 127delA (43fsX22; frame-shift from codon 43 resulting in the termination at the 22nd codon, codon 65), 175delG (59fsX6), 271C>T (R91C), 325A>G (R109G), and 357T>C (N119N) in exon 1, and IVS1+1G>T, IVS1+98A>G and IVS1+101G>T in the following intron. Among them, 127delA and 175delG can confer early termination of translation, resulting in an immature protein that probably lacks enzymatic activity. Variation IVS1+1G>T is located at a splice donor site and thus may lead to aberrant splicing. Since we did not find any significant differences in the frequencies of all the variations among the three subject groups, the data were analyzed as one group. The allele frequencies of the novel variations were 0.006 for IVS1+101G>T, 0.004 for 30G>A (P10P) and 357T>C (N119N), and 0.002 for the 8 other variations. In addition, the two known nonsynonymous single nucleotide polymorphisms (SNPs), 31C>T (R11W) and 142T>G (L48V), were found at 0.012 and 0.129 frequencies, respectively. The SNP 70C>A (P24T), mostly linked with 142T>G (L48V) in German Caucasians, was not detected in this study. Sixteen haplotypes were identified or inferred, and some haplotypes were confirmed by cloning and sequencing. It was shown that most of 142T>G (L48V) was linked with -219C>T, -163G>A, 448T>C (L150L), 804G>A (P268P), and IVS1+43C>T, comprising haplotype *3a; haplotype *4a harbors 31C>T (R11W); 127delA (43fsX22) and 142T>G (L48V) were linked (haplotype *5a); 175delG (59fsX6) was linked with 325A>G (R109G) (*6a haplotype); and -219C>T, -163G>A, 142T>G (L48V), 271C>T (R91C), 448T>C (L150L), 804G>A (P268P), and IVS1+43C>T comprised haplotype *7a. Our results provide fundamental and useful information for genotyping UGT1A4 in the Japanese and probably Asian populations.