ANTICANCER RESEARCH 26: 2189-2196 (2006)

Glucuronidation of SN-38 and NU/ICRF 505 in Human Colon Cancer and Adjacent Normal Colon

JEFFREY CUMMINGS1, BRIAN T. ETHELL2, LESLEY JARDINE1 and BRIAN BURCHELL2

1Cancer Research UK, Edinburgh Oncology Unit, Western General Hospital, Edinburgh, EH4 2XR; 2Department of Molecular and Cellular Pathology, Ninewells Hospital and Medical School, Dundee, DD1 9SY, U.K.

Abstract. Background: Glucuronidation represents a novel poorly characterised (2-4). Glucuronidation, catalysed by mechanism of intrinsic drug resistance in colon cancer cells. the UDP-glucuronosyltransferases (EC 2.4.1.17, UGT) To safely reverse this mechanism in vivo, it is essential to superfamily of drug metabolising enzymes, has recently identify which isoforms of UDP-glucuronosyltransferases are been identified in vitro in human CRC cell lines as a novel responsible for catalysing this drug metabolism in tumour intrinsic drug resistance mechanism operative against the tissue. Materials and Methods: LC-MS was applied to measure topoisomerase I (topo I) inhibitors 7-ethyl 10-hydroxy- rates of glucuronidation of two anticancer compounds (SN-38 camptothecin (SN-38, the active metabolite of clinically and NU/ICRF 505) by patient colon cancer biopsies and administered irinotecan) and the novel N-tyrosine paired normal colon. Results: Three independent lines of conjugate of anthraquinone, NU/ICRF 505 (5-8). Selective enquiry indicated that, in the tumour specimens, SN-38 was inhibition of this drug metabolism pathway with the glucuronidated primarily by UGT1A1, the isozyme generally UGT1A9 isozyme specific substrate propofol demonstrated recognised as being responsible for hepatic detoxification of that glucuronidation could account de novo for, at least, this compound, while with NU/ICRF 505 two candidate five-fold resistance to NU/ICRF 505 and two-fold isoforms emerged – UGT1A8 and/or UGT1A10 – both of resistance to SN-38 and provided proof of principle that which are not normally expressed in the liver. Conclusion: this mechanism can be effectively reversed by UGT These data suggest that tumour selective modulation of this aglycones, at least in vitro (7). However, in man specific drug resistance mechanism in patients may be feasible with hepatic UGT isoenzymes contribute significantly to the NU/ICRF 505 but more difficult to realise with SN-38. detoxification and elimination of SN-38 via the bile (9-13). Thus, in order to consider glucuronidation a possible De novo drug resistance is recognised as contributing exploitable mechanism of drug resistance amenable to significantly to the poor response rates of colorectal cancer reversal in the clinic with modulators, it is essential to (CRC) to chemotherapy (1). Nonetheless, the underlying establish which isoform(s) of the enzyme are responsible mechanisms responsible for drug insensitivity remain for metabolising anticancer drugs in CRC. The UGT superfamily consists of 18 gene products that are subdivided into two major families: the UGT1 family comprising nine different proteins encoded by a single Abbreviations: CRC, colorectal cancer; UGT, UDP-glucuronosyl- genetic locus on chromosome 2 (UGT1A) and the UGT2 transferases; topo I, topoisomerase I; SN-38, 7-ethyl 10-hydroxy- family (including the UGT2A and UGT2B sub-families), camptothecin; C4-G, NU/ICRF 505 C4-O-‚-glucuronide; Tyr-G, also comprising nine different proteins but encoded by NU/ICRF 505 tyrosyl-O-‚-glucuronide; SN-38 C10-G, SN-38 C10- individual genes (with the exception of UGT2A1 and 2) O-‚-glucuronide; LC-MS, high-performance liquid chromatography with mass spectrometry; UDPGA, UDP-glucuronic acid; SIR, single located on chromosome 4 (14, 15). Differential expression ion recording. of individual UGTs occurs in different tissues (16-18) and certain isoforms can be predominately expressed by a single Correspondence to: Dr. Jeffrey Cummings at present address: organ (19). Therefore, the concept of tissue (or possibly Clinical and Experimental Pharmacology, Paterson Institute for even tumour) selective target protein expression is not Cancer Research, University of Manchester, Wilmslow Road, unprecedented. Individual UGT isoforms can also exhibit a Manchester, M20 4BX, U.K. Tel: 44-(0)-161-446-3149, Fax: 44-(0)- high degree of substrate specificity (20-23), while a large 161-446-3109, e-mail: [email protected] pool of natural products, food additives and proprietary Key Words: SN-38, NU/ICRF 505, UDP-glucuronosyltransferases, medicines is available from which to screen candidates as colon cancer. non-toxic, high affinity, isozyme specific inhibitors (24).

0250-7005/2006 $2.00+.40 2189 ANTICANCER RESEARCH 26: 2189-2196 (2006)

In this report, the glucuronidation of SN-38 and standards and running a full 7- to 10-point calibration curve with NU/ICRF 505 was studied by human colon cancer biopsies each assay of patient samples. The range of the calibration curves and adjacent normal colon specimens in situ, with the aim of extended from 30 pg to 20 ng on the column and each glucuronide yielded a linear calibration curve over this range with regression providing insights into the identity of the UGT isoenzymes correlation coefficients (r2) always better than 0.975. The reactions responsible for catalysing this drug metabolism activity. were stopped by the addition of methanol to precipitate the protein followed by direct injection, ensuring quantitative recovery of the Materials and Methods glucuronides from the incubation mixtures. Substrate concentrations were: propofol, 500 ÌM; 1-naphthol, 500 ÌM; morphine, 500 ÌM; Drug standards. NU/ICRF 505, its hydrolysis product NU/ICRF NU/ICRF 505, 150 ÌM and SN-38, 250 ÌM. 505/M, the anthraquinone ring C4-O-‚-glucuronide (C4-G) and The kinetics of inhibition of NU/ICRF 505 glucuronidation at tyrosyl-O-‚-glucuronide (Tyr-G) of NU/ICRF 505 were synthesised 10- and 150-ÌM substrate concentrations were also measured in the and purified to homogeneity as described in full previously (25-27). human colon cancer biopsies by adding increasing concentrations SN-38 was a gift from Aventis (Vitry, France). Purified SN-38 of propofol (0, 20, 50 100 150, 250 and 500 ÌM). The IC50 for C10-O-‚-glucuronide (C10-G) was synthesised as described above enzyme inhibition was defined as the concentration of propofol for the glucuronides of NU/ICRF 505 (27). 1-Naphthol glucuronide necessary to induce a 50% reduction in glucuronidation activity. was from Sigma, Poole, UK, and morphine 3 and 6 glucuronides The experiments were repeated on two to four separate were from Ultrafine Chemicals, Salford, UK. A standard of occasions and glucuronidation activity was calculated as pmol propofol glucuronide was generated by in vitro incubation of glucuronide formed/min/mg of protein (pmol/min/mg). propofol with V79 Chinese hamster fibroblasts transfected with human UGT1A9, as previously described in detail (28). Results Cell lines. The HT29 human colon adenocarcinoma cancer cell line, which was used as a positive control for glucuronidation activity, Preferential formation of the two different glucuronides of was from the ATCC (Manassas, VA, USA), while the HCT116 NU/ICRF 505 by individual UGT isozymes. NU/ICRF 505 is human colon cancer cell line, used as a negative control, was from metabolised to two different glucuronides: C4-G and Tyr-G the ECACC (European collection of cell culture, Salisbury, UK), (6). Utilising a panel of individual cDNA expressed UGT both being cultured as monolayers as previously described (29). isozymes, it was demonstrated that each active isoform Transfected cell lines expressing UGT isoforms. V79 Chinese hamster displayed a reproducible and characteristic preference fibroblasts and HEK293 fibroblasts stably expressing full-length towards the formation of the two different glucuronides human UGT cDNAs were as previously described (30, 31). (Figure 1A-E), where UGT1A1 generated both in similar Microsomes prepared from Sf-9 insect cells transfected with human proportion, UGT1A9 and 1A7 showed selectivity for the UGT1A7 and 1A10 were purchased from Oxford Biomedical tyrosine glucuronide, UGT1A10 showed selectivity for the Research (Oxford, MI, USA). ring C-4 glucuronide and UGT2B7 exhibited complete Clinical specimens and sample preparation. Nineteen colon cancer specificity for C4-G. biopsies, each from a different patient including nine with paired adjacent normal colon specimens (0.2-0.5), were kindly provided Relative levels of production of the two different glucuronides by Professor Malcolm Dunlop, Western General Hospital, of NU/ICRF 505 by the clinical specimens. The pattern of Edinburgh, UK. After resection, all the colon specimens were glucuronide formation observed in incubations of both immediately snap-frozen in solid carbon dioxide and then stored at –80ÆC until utilised. Written informed patient consent and local colon cancer biopsies and normal paired colon specimens committee ethical approval were both obtained to conduct the with NU/ICRF 505, while likely to be a composite measure, experimental investigations described in this report. produced a profile almost identical to that of UGT1A10. In this case, the C4-glucuronide was generated in large excess Glucuronidation activity of clinical specimens determined by LC-MS. of Tyr-G, by a factor of 5.1±1.7 SD in the tumour biopsies Crude lysates of the clinical specimen fragments (approximately and by 5.7±1.0 SD in the normal colon specimens 100 mg) were prepared by hand homogenisation in pre-chilled compared to 4.6 for UGT1A10. 0.25 M sucrose, 5 mM HEPES buffer pH 7.4 (50 Ìl per 10 mg of tissue) in a loose fitting 1-ml glass homogeniser and aliquoted into 100-Ìl fractions. Any remaining tissue was broken up further by Glucuronidation activities present in human colon cancer sonication (MSE Soniprep 150, Sanyo Gallenkamp, Loughborough, biopsies and paired normal colon specimens. Glucuronidation UK) by two 5-sec bursts with 1 min resting on ice between bursts. The activity was detected in 18 of the 19 colon cancer biopsies protein concentration was measured according to the Lowry method. and all nine adjacent paired normal colon specimens, but Glucuronidation activity was measured by incubating the clinical with marked inter-patient variations, ranging from <1 to specimen lysates (or expressed cell sonicates) at 37ÆC with the 300 pmol/min/mg depending on the substrate utilised. These substrate and co-factor over a period of up to 2 h in order to obtain a kinetic rate of glucuronidation by measuring glucuronide rates of activity were in the same range as that observed formation by LC-MS (21, 28, 32). Quantitation of the glucuronides with single UGT isozymes expressed in the cells in culture was achieved by calibrating the LC-MS with authentic glucuronide and, more significantly, were in the same range as that

2190 Cummings et al: Glucuronidation of Anticancer Drugs by Human Colon Cancer

Figure 1. Preferential formation of the two different glucuronide metabolites of NU/ICRF 505 (Tyr-G, 6.4 min and C4-G, 7.1 min) by cloned UGT isoforms. cDNA expressed cell sonicates or microsomes (100-200 Ìg of protein ) were incubated with 150 ÌM of NU/ICRF 505 for 2 h and the formation of glucuronide metabolites determined by single ion recording (SIR) LC-MS. For the purposes of comparison, all five chromatograms were normalised to the most abundant chromatographic peak: A, UGT1A10; B, UGT1A7; C, UGT2B7; D, UGT1A9; E, UGT1A1.

observed in HT29 cells where the UGTs were responsible tumour (patients 87, 94, 106, 125 and 461). Only low rates for the expression of intrinsic drug resistance to anticancer of activity were detected with the UGT2B7 substrate drugs (7, 8). morphine (normally <1 pmol/min/mg, Figure 2A), in The profiles of glucuronidation activity determined in the keeping with previous reports of more modest activity of the nine paired colon cancer and normal colon specimens UGT2 family in the gastrointestinal tract (17, 33, 34), while towards the three different UGT probe substrates are by far the highest level of activity was recorded with shown in Figure 2A-C. Interestingly, in four out of nine 1-naphthol (Figure 2B). In addition to UGT1A6, which we patients higher rates of glucuronidation activity were have shown not to be capable of metabolising both SN-38 measured in the tumour compared to the normal colon and NU/ICRF 505, UGT1A1 will accept 1-naphthol as a (patients 84, 95, 99 and 271, with the exception of patient substrate (31). Intermediate activity was recorded with the 271 with propofol), while in the remaining five patients UGT1A9 probe substrate propofol (1-10 pmol/min/mg, higher rates were recorded in the normal colon versus the Figure 2C) (21). By comparison, rates of glucuronidation

2191 ANTICANCER RESEARCH 26: 2189-2196 (2006)

Correlation between anticancer drug and isozyme probe substrate glucuronidation. When the tumour and normal colon profiles of glucuronide formation determined with NU/ICRF 505 and SN-38 were compared against that of the three probe substrates, interesting correlations emerged (Table I). The profile of formation of the C4-G of NU/ICRF 505 in the different colon cancer samples correlated strongly with the pattern of glucuronide formation obtained with propofol (see Table I). No other highly significant correlations were evident, suggesting that the isoform responsible for metabolising propofol was also responsible for generating the major glucuronide metabolite of NU/ICRF 505. There was also a very strong correlation between the formation of SN-38 C10-G, NU/ICRF 505 Tyr-G and 1-naphthol, again suggesting that the same isoform(s) were responsible for metabolising all three different compounds. In the normal colon a more complex picture emerged where formation of C4-G correlated with that of propofol glucuronide and Tyr-G; formation of Tyr-G correlated with that of propofol glucuronide, 1-naphthol glucuronide and C4-G, and formation of SN-38 glucuronide correlated with that of Tyr-G, 1-naphthol glucuronide and morphine 3-glucuronide (see Table I).

Kinetics of inhibition of NU/ICRF 505 glucuronide formation by propofol in human colon cancer biopsies. The kinetics of inhibition of NU/ICRF 505 glucuronide formation in the colon cancer biopsies by the UGT1A9 substrate propofol was investigated. Substrates of this isoform often show cross reactivity with UGT1A7 and UGT1A10 (33), due to close sequence homology (35). The IC50 for inhibition of NU/ICRF 505 glucuronide formation by cloned UGT1A9 was 99 ÌM for Tyr-G and 461 ÌM for C4-G (Figure 3). UGT1A9 showed a strong preference for the formation of Tyr-G (see Figure 1). In HT29 cells, where propofol has been demonstrated to effectively inhibit NU/ICRF 505 metabolism (5, 36), the IC50 for inhibition of Tyr-G was 94 ÌM, consistent with UGT1A9 catalysing this metabolism. However, the IC50 for inhibition of C4-G was 143 ÌM, indicative of participation of other closely related UGTs in the metabolism of this drug, such as UGT1A7, 1A10 or even 1A8 which were not investigated in this study. In the colon cancer biopsies, up to a concentration of 500 ÌM propofol Figure 2. Glucuronidation activity in a panel of patient colon cancer and was without effect on the formation of Tyr-G, powerful adjacent normal colon biopsies. Glucuronidation rates were determined evidence against the involvement of the UGT1A7-10 series by LC-MS using three different isoform selective probes and are presented of isoforms. as pmol of glucuronide formed per min per mg of protein: A, morphine as substrate; B, 1-naphthol as substrate; C, propofol as substrate. Propofol effectively prevented the formation of C4-G in six out of eight biopsies investigated with an IC50 ranging from 172-372 ÌM, indicating participation of the UGT1A7-10 activity with the anticancer drugs as substrates where series, but with either a significant contribution or the comparatively high: 5-25 pmol/min/mg for the C-4 exclusive participation of isozymes other than UGT1A9. In glucuronide of NU/ICRF 505. patient biopsies 271 and 461, propofol was without effect,

2192 Cummings et al: Glucuronidation of Anticancer Drugs by Human Colon Cancer

Table I. P values for the degree of correlation between the patterns of glucuronidation activity measured in a panel of biopsies obtained using NU/ICRF 505 (C4-G and Tyr-G) and SN-38 (SN-38 C10-G) as substrates compared to that obtained using 3 different UGT isozyme probes as substrates.

Probe Propofol Naphthol Morphine C4-G Tyr-G SN-38 C10-G (Isoform (1A9, 7, 10) (1A6, 1) (2B7) selectivity)

Human colon cancer biopsies

C4-G 0.0001* 0.06 0.03 - 0.17 0.02 Tyr-G 0.39 0.004 0.04 0.17 - 0.0004 SN-38 C10-G 0.07 0.002 0.02 0.02 0.0004 -

Adjacent normal colon specimens

C4-G 0.007* 0.03 0.09 - 0.00005 0.03 Tyr-G 0.006 0.002 0.03 0.00001 - 0.01 SN-38 C10-G 0.04 0.002 0.003 0.03 0.01 -

*Two data sets of glucuronidation activity obtained with two different substrates (e.g., SN-38 and 1-naphthol) consisting of 18 values for the tumour biopsies and nine values for the normal colon were graphed on an X,Y scatter plot using the computer programme Excel. The regression correlation coefficient was first determined and then the level of significance of the correlation calculated using Excel, and these latter data are presented in Table I.

highlighting a significant inter-patient variation in UGT to a lesser extent 1A10 (7, 10, 42-44); while, NU/ICRF 505 isozyme profiles or perhaps even the presence of a is metabolised predominately by UGT1A9 and 1A1, and to polymorphism, which has been reported for each member of a lesser extent 1A10 (7). this series (37-41). Though pattern recognition data supported UGT1A10 as the isoform solely responsible for NU/ICRF 505 metabolism Discussion in both the tumour and normal samples, the probe substrate data indicated a far more complex picture. UGT1A10 was In the present study, a functional approach was utilised in an originally reported to be expressed predominantly in the attempt to provide insights into the identity of the UGT colon (19), although it is now clear that this isoform is present isoforms responsible for metabolising anticancer drugs in throughout the gastrointestinal tract as well as the biliary colon cancer, utilising LC-MS and in situ incubations with canal (16, 17). UGT1A10 has also been demonstrated to be patient biopsies. Three different strategies were pursued. very active in the metabolism of the anticancer and Use was made of the fact that NU/ICRF 505 is converted immunosuppressant agent mycophenolic acid, while most into two different glucuronides (27) and that individual UGT other UGTs investigated had relatively little activity towards isoforms produce a unique signature in their preference for this drug (20). This protein is, therefore, a strong candidate as the formation of each metabolite. A comparison was then the UGT isoform primarily responsible for the expression of made to the pattern of NU/ICRF 505 glucuronide the intrinsic resistance to mycophenolic acid in human colon formation observed in the incubations with colon tumour cancer cells, including HT29 (45, 46) where intrinsic and normal tissue specimens. The next approach was to resistance to NU/ICRF 505 and SN-38 was first reported (5, screen patient samples with UGT probe substrates and then 27). However, recent reports on the glucuronidation of correlate the activity profiles to those obtained with the two mycophenolic acid suggested a more complicated picture anticancer compounds. Third, the identity(ies) of UGT implicating isoforms such as UGT1A9 and UGT2B7 (47, 48). isoform(s) catalysing anticancer glucuronidation in the colon In the studies with probe substrates, corroborative cancer biopsies was further clarified by studying the effect of evidence was obtained that a member of the UGT1A7-1A10 an inhibitor of NU/ICRF 505 metabolism. homologous series of isoforms (19) was responsible for the Data from a series of previous in vitro based analytical formation of the C4-G major metabolite of NU/ICRF 505 studies utilising cloned UGT isozymes demonstrated that in the colon cancer biopsies. Since UGT1A7 is not only a limited number of isoforms are capable of expressed in the colon (17, 49), this protein may be metabolising both compounds, thus limiting the potential discounted. However, with Tyr-G there was no correlation number of candidates in the present study. SN-38 is to the UGT1A7-10 substrate propofol, but a strong metabolised predominately by UGT1A1, 1A7 and 1A9 and correlation to the UGT1A6 (and 1A1) substrate 1-naphthol.

2193 ANTICANCER RESEARCH 26: 2189-2196 (2006)

Figure 3. IC50 values for inhibition of NU/ICRF 505 glucuronidation in colon cancer biopsies by propofol. Colon cancer biopsy homogenates were incubated with 150 ÌM of NU/ICRF 505 over 2 h in the presence of increasing concentrations of propofol and the rates of formation of the Tyr-G and C4-G glucuronide metabolites were determined by LC-MS. The IC50 values for inhibition were calculated by linear regression analysis. The numbers on ■ ■ top of the bars represent the IC50 values: , C4-G and , Tyr-G.

Likewise, in the colon tumour specimens there was also a variation in the spectrum of isoforms expressed in the strong correlation between SN-38 C10-G formation and 1- colon cancer biopsies. The second was that the isoform(s) naphthol. Since UGT1A6 does not accept either SN-38 or present behaved differently to that of UGT1A9 in their NU/ICRF 505 as a substrate, UGT1A1 may possibly emerge response to enzyme inhibition. Thus, by a process of as the lead candidate. While there are numerous reports elimination, UGT1A10 and 1A8 remain as the main that implicate UGT1A1 in the hepatic metabolism and candidates responsible for the formation of the major biliary clearance of SN-38 (9-11, 13, 50-52), the present metabolite of NU/ICRF 505 in colon cancer tissue. study is one of the first reports implicating UGT1A1 in the Although, the pattern recognition data tends to favour metabolism of the drug and possible expression of intrinsic UGT1A10, in the absence of analogous data on this drug resistance, by colon cancer tissue. However, the fact isoform, UGT1A8 cannot be discounted. Therefore, it is that the same isozyme may be responsible for both systemic not possible at this stage to distinguish between UGT1A8 drug detoxification and drug resistance might indicate that and UGT1A10. However, since both these isoforms are tumour selective modulation of this pathway would be more not expressed in the liver (17, 19, 49), selective modulation difficult to achieve. Nonetheless, selective modulation of of NU/ICF 505 glucuronidation (which is predominantly SN-38 resistance mediated by glucuronidation may be through C4-G) by colon tumour may be possible without possible in other tumour types. Recent evidence suggested adversely affecting systemic detoxification in the liver. The that UGT1A10 is responsible for the expression of drug data generated with the probe substrates in the normal resistance to SN-38 in human lung cancer (44, 53). colon specimens suggested that multiple forms of UGTs The data generated using propofol as an inhibitor of catalyse the metabolism of both SN-38 and NU/ICRF 505, NU/ICRF 505 glucuronidation by the tumour biopsies further supporting the contention that tumour-specific adds further support to the conclusion that one or more modulation of this drug may be achievable. members of the UGT1A8-10 trio produce C4-G but not In summary, it is suggested that, in human colon cancer, Tyr-G. However, the IC50 values for inhibition of C4-G by SN-38 is glucuronidated predominantly by UGT1A1 while propofol revealed two interesting observations. The first NU/ICRF 505 is mainly metabolised by UGT1A8 and/or was that there appeared to be a significant inter-patient UGT1A10.

2194 Cummings et al: Glucuronidation of Anticancer Drugs by Human Colon Cancer

References 14 Tukey RH, Strassburg CP and Mackenzie PI: Pharmacogenomics of human UDP-glucuronosyltransferases and irinotecan toxicity. 1 Sobrero A, Kerr D, Glimelius B, VanCutsem E, Milano G, Mol Pharmacol 62: 446-450, 2002. Pritchard DM, Rougier P and Aapro M: New directions in the 15 Mackenzie PI, Walter Bock K, Burchell B, Guillemette C, Ikushiro treatment of colorectal cancer: a look to the future. Eur J S, Iyanagi T, Miners JO, Owens IS and Nebert DW: Nomenclature Cancer 36: 559-566, 2000. update for the mammalian UDP glycosyltransferase (UGT) gene 2 Tomonaga M, Oka M, Narasaki F, Fukuda M, Nakano R, superfamily. Pharmacogenet Genomics 15: 677-685, 2005. Takatani H, Ikeda K, Terashi K, Matsuo I, Soda H, Cowan KH 16 Strassburg CP, Oldhaffer K, Manns MP and Tukey RH: and Kokno S: The multidrug resistance-associated protein gene Differential expression of the UGT1A locus in human liver, confers drug resistance in human gastric and colon cancer. Jpn biliary, and gastric tissue: identification of UGT1A7 and J Cancer Res 87: 1263-1270, 1996. UGT1A10 transcripts in extrahepatic tissue. Mol Pharmacol 52: 3 Meschini S, Calcabrini A, Monti E, DelBufalo D, Stringaro A, 212-220, 1997. Dolfini E and Arancia G: Intracellular P-glycoprotein expression is 17 Tukey RH and Strassburg CP: Genetic multiplicity of the associated with the intrinsic multidrug resistance phenotype in human UDP-glucuronosyltransferases and regulation in the human colon adenocarcinoma cells. Int J Cancer 87: 615-628, 2000. gastrointestinal tract. Mol Pharmacol 59: 405-414, 2001. 4 Pavillard V, Charasson V, Laroche-Clary A, Soubeyran I and 18 Gregory PA, Lewinsky RH, Gardner-Stephen DA and Mackenzie Robert J: Cellular parameters predictive of the clinical response PI: Regulation of UDP glucuronosyltransferases in the of colorectal cancers to irinotecan. A preliminary study. gastrointestinal tract. Toxicol Appl Pharmacol 199: 354-363, 2004. Anticancer Res 24: 579-585, 2004. 19 Mojarrabi B and Mackenzie PI: Characterization of two UDP 5 Cummings J, Boyd G, Ethell BT, Macpherson JS, Burchell B, glucuronosyltransferases that are predominantly expressed in Smyth JF and Jodrell DI: Glucuronidation as a de novo human colon. Biochem Biophys Res Commum 247: 704-709, 1998. mechanism of drug resistance in colon cancer cells. Clin Cancer 20 Mojarrabi B and Mackenzie PI: The human UDP Res 6 (suppl): 82, 2000. glucuronosyltransferase, UGT1A10, glucuronidates mycophenolic 6 Cummings J, Zelcer N, Allen JD, Yao D, Boyd G, Friedberg acid. Biochem Biophys Res Commum 238: 775-778, 1997. TH, Smith G, Wolf H, Ethell B, Burchell B, Smyth JF and 21 Ethell BT, Beaumont K, Rance DJ and Burchell B: Use of Jodrell DI: Glucuronidation as a mechanism of intrinsic drug cloned and expressed human UDP-glucuronosyltransferases for resistance in colon cancer cells: contribution of drug transport the assessment of human drug conjugation and identification of proteins. Proc Am Assoc Cancer Res 42: 278, 2001. potential drug interactions. Drug Metab Dispos 29: 48-53, 2001. 7 Cummings J, Ethell BT, Jardine L, Boyd G, Macpherson JS, Burchell 22 Shelby MK, Cherrington NJ, Vansell NR and Klaassen CD: Tissue B, Smyth JF and Jodrell DI: Glucuronidation as a mechanism of mRNA expression of the rat UDP-glucuronosyltransferase gene intrinsic drug resistance in human colon cancer: reversal of resistance family. Drug Metab Dispos 31: 326-333, 2003. by food additives. Cancer Res 63: 8443-8450, 2003. 23 Sorich MJ, Miners JO, McKinnon RA and Smith PA: Multiple 8 Cummings J, Zelcer N, Allen JD, Yao D, Boyd G, Maliepaard pharmacophores for the investigation of human UDP- M, Friedberg TH, Smyth JF and Jodrell DI: Glucuronidation glucuronosyltransferase isoform substrate selectivity. Mol as a mechanism of intrinsic drug resistance in colon cancer cells: Pharmacol 65: 301-308, 2004. contribution of drug transport proteins. Biochem Pharmacol 67: 24 Grancharov K, Naydenova Z, Lozeva S and Golovinsky E: 31-39, 2004. Natural and synthetic inhibitors of UDP-glucuronosyltransferase. 9 Iyer L, King CD, Whitington PF, Green MD, Roy SK, Tephly Pharmacol Ther 89: 171-186, 2001. TR, Coffman BL and Ratain MJ: Genetic predisposition to the 25 Cummings J, Meikle I, MacPherson JS and Smyth J: metabolism of irinotecan (CPT-11). Role of uridine diphosphate Accumulation of anthracenyl-amino acid topoisomerase I and glucuronosyltransferase isoform 1A1 in the glucuronidation of II inhibitors in drug sensitive and drug resistant human ovarian its active metabolite (SN-38) in human liver microsomes. J Clin cancer cell lines determined by HPLC. Cancer Chemother Invest 101: 847-854, 1998. Pharmacol 37: 103-109, 1995. 10 Gagne JF, Montminy V, Belanger P, Journault K, Gaucher G and 26 Cummings J, Meikle I, MacPherson JS and Smyth JF: Guillemette C: Common human UGT1A polymorphisms and the Characterisation of the major metabolite of the novel altered metabolism of irinotecan active metabolite 7-ethyl-10- topoisomerase I inhibitor NU/ICRF 505. Anticancer Drug hydroxycamptothecin (SN-38). Mol Pharmacol 62: 608-617, 2002. Design 11: 367-382, 1996. 11 Iyer L, Das S, Janisch L, Wen M, Ramirez J, Karrison T, 27 Cummings J, Boyd G, Ethell BT, Macpherson JSM, Burchell B, Fleming GF, Vokes EE, Schilsky RL and Ratain MJ: Smyth JF and Jodrell DI: Enhanced clearance of topoisomerase UGT1A1*28 polymorphism as a determinant of irinotecan I inhibitors from human colon cancer cells by glucuronidation. disposition and toxicity. Pharmacogenomics J 2: 43-47, 2002. Biochem Pharmacol 63(4): 607-613, 2002. 12 Innocenti F, Undevia SD, Iyer L, Chen PX, Das S, 28 Ethell BT, Anderson GD, Beaumont K, Rance DJ and Kocherginsky M, Karrison T, Janisch L, Ramirez J, Rudin CM, Burchell B: A universal radiochemical high-performance Vokes EE and Ratain MJ: Genetic variants in the UDP- liquid chromatographic assay for the determination of UDP- glucuronosyltransferase 1A1 gene predict the risk of severe glucuronosyltransferase activity. Anal Biochem 255: 142-147, neutropenia of irinotecan. J Clin Oncol 22: 1382-1388, 2004. 1998. 13 Mehra R, Murren J, Chung G, Smith B and Psyrri A: Severe 29 Cummings J, MacPherson JS, Meikle I and Smyth JF: irinotecan-induced toxicities in a patient with uridine Development of anthracenyl-amino acid conjugates as topo- diphosphate glucuronosyltransferase 1A1 polymorphism. Clin isomerase I and II inhibitors that circumvent drug resistance. Colorectal Cancer 5: 61-64, 2005. Biochem Pharmacol 52: 979-990, 1996.

2195 ANTICANCER RESEARCH 26: 2189-2196 (2006)

30 Fournel-Gigleux S, Sutherland L, Sabolovic N, Burchell B and Siest 44 Oguri T, Takahashi T, Miyazaki M, Isobe T, Kohno N, G: Stable expression of two human UDP-glucuronosyltransferase Mackenzie PI and Fujiwara Y: UGT1A10 is responsible for cDNAs in V79 cell cultures. Mol Pharmacol 39: 177-183, 1991. SN-38 glucuronidation and its expression in human lung 31 Burchell B, Brierley C and Rance D: Specificity of human cancers. Anticancer Res 24: 2893-2896, 2004. UDP-glucuroosyltransferases and xenobiotic glucuronidation. 45 Franklin T, Jacobs V, Jones G, Ple P and Bruneau P: Life Sci 57: 1819-1831, 1995. Glucuronidation associated with intrinsic resistance to 32 Ethell BT, Ekins S, Wang J and Burchell B: Quantitative mycophenolic acid in human colorectal carcinoma cells. Cancer structure activity relationships for the glucuronidation of simple Res 56: 984-987, 1996. phenols by expressed human UGT1A6 and UGT1A9. Drug 46 Franklin T, Jacobs V, Jones G and Ple P: Human colorectal Metab Dispos 30: 734-738, 2002. carcinoma cells in vitro as a means to assess the metabolism of 33 Strassburg CP, Strassburg A, Nguyen N, Li Q, Manns MP and analogs of mycophenolic acid. Drug Metab Dispos 25: 367- Tukey RH: Regulation and function of family 1 and family 2 370, 1997. UDP-glucuronosyltransferase genes (UGT1A, UGT2B) in 47 Picard N, Ratanasavanh D, Premaud A, Le Meur Y and human oesophagus. Biochem J 338(Pt 2): 489-498, 1999. Marquet P: Identification of the UDP-glucuronosyltransferase 34 Strassburg CP, Nguyen N, Manns MP and Tukey RH: UDP- isoforms involved in mycophenolic acid phase II metabolism. glucuronosyltransferase activity in human liver and colon. Drug Metab Dispos 33: 139-146, 2005. Gastroenterology 116: 149-160, 1999. 48 Miles KK, Stern ST, Smith PC, Kessler FK, Ali S and Ritter JK: 35 Gong QH, Cho JW, Huang T, Potter C, Gholami N, Basu NK, An investigation of human and rat liver microsomal Kubota S, Carvalho S, Pennington MW, Owens IS and Popescu mycophenolic acid glucuronidation: evidence for a principal NC: Thirteen UDP-glucuronosyltransferase genes are encoded role of UGT1A enzymes and species differences in UGT1A at the human UGT1 gene complex locus. Pharmacogenetics 11: specificity. Drug Metab Dispos 33: 1513-1520, 2005. 357-368, 2001. 49 Strassburg CP, Manns MP and Tukey RH: Expression of the 36 Cummings J, Macpherson JS, Boyd G, Ethell BT, Burchell B, UDP-glucuronosyltransferase 1A locus in human colon. Smyth JF and Jodrell DI: Novel mechanism of drug resistance Identification and characterization of the novel extrahepatic to the topoisomerase I (topo I) inhibitor NU/ICRF 505. Proc UGT1A8. J Biol Chem 273: 8719-8726, 1998. Am Assoc Cancer Res 41: 844, 2000. 50 Ando Y, Saka H, Ando M, Sawa T, Muro K, Ueoka H, 37 Strassburg CP, Vogel A, Kneip S, Tukey RH and Manns MP: Yokoyama A, Saitoh S, Shimokata K and Hasegawa Y: Polymorphisms of the human UDP-glucuronosyltransferase Polymorphisms of UDP-glucuronosyltransferase gene and (UGT) 1A7 gene in colorectal cancer. Gut 50: 851-856, 2002. irinotecan toxicity: a pharmacogenetic analysis. Cancer Res 60: 38 Miners JO, McKinnon RA and Mackenzie PI: Genetic 6921-6926, 2000. polymorphisms of UDP-glucuronosyltransferases and their 51 Ando M, Ando Y, Sekido Y, Shimokata K and Hasegawa Y: functional significance. Toxicology 181-182: 453-456, 2002. Genetic polymorphisms of the UDP-glucuronosyltransferase 39 Huang YH, Galijatovic A, Nguyen N, Geske D, Beaton D, 1A7 gene and irinotecan toxicity in Japanese cancer patients. Green J, Green M, Peters WH and Tukey RH: Identification Jpn J Cancer Res 93: 591-597, 2002. and functional characterization of UDP-glucuronosyltransferases 52 Paoluzzi L, Singh AS, Price DK, Danesi R, Mathijssen RH, UGT1A8*1, UGT1A8*2 and UGT1A8*3. Pharmacogenetics 12: Verweij J, Figg WD and Sparreboom A: Influence of genetic 287-297, 2002. variants in UGT1A1 and UGT1A9 on the in vivo 40 Jinno H, Saeki M, Tanaka-Kagawa T, Hanioka N, Saito Y, Ozawa glucuronidation of SN-38. J Clin Pharmacol 44: 854-860, 2004. S, Ando M, Shirao K, Minami H, Ohtsu A, Yoshida T, Saijo N 53 Chikamori M, Takigawa N, Kiura K, Tabata M, Shibayama T, and Sawada J: Functional characterization of wild-type and Segawa Y, Ueoka H, Ohnoshi T and Tanimoto M: Establishment variant (T202I and M59I) human UDP-glucuronosyltransferase of a 7-ethyl-10-hydroxy-camptothecin-resistant small cell lung 1A10. Drug Metab Dispos 31: 528-532, 2003. cancer cell line. Anticancer Res 24: 3911-3916, 2004. 41 Carlini LE, Meropol NJ, Bever J, Andria ML, Hill T, Gold P, Rogatko A, Wang H and Blanchard RL: UGT1A7 and UGT1A9 polymorphisms predict response and toxicity in colorectal cancer patients treated with capecitabine/irinotecan. Clin Cancer Res 11: 1226-1236, 2005. 42 Hanioka N, Ozawa S, Jinno H, Ando M, Saito Y and Sawada J: Human liver UDP-glucuronosyltransferase isoforms involved in the glucuronidation of 7-ethyl-10-hydroxycamptothecin. Xenobiotica 31: 687-699, 2001. 43 Jinno H, Tanaka-Kagawa T, Hanioka N, Saeki M, Ishida S, Nishimura T, Ando M, Saito Y, Ozawa S and Sawada J: Glucuronidation of 7-ethyl-10-hydroxycamptothecin (SN-38), an active metabolite of irinotecan (CPT-11), by human UGT1A1 variants, G71R, P229Q, and Y486D. Drug Metab Dispos 31: Received March 11, 2006 108-113, 2003. Accepted March 29, 2006

2196