Abstract
Interindividual variability in the glucuronidation of xenobiotics metabolized by UDP-glucuronosyltransferase 1A9 (UGT1A9) suggests the presence of functional UGT1A9 variants. The aim of this study was to evaluate whether the putative functionality of the UGT1A9 variants–118T9>10 (rs3832043), I399C>T (rs2741049), –275T>A (rs6714486), and–2152C>T (rs17868320) could be confirmed in an independent study. UGT1A9 genotypes and UGT1A9 activity (i.e., flavopiridol and mycophenolic acid glucuronidation) were determined in 46 Caucasian human livers. mRNA levels were quantitated by real-time polymerase chain reaction in 35 of these livers. In addition, samples from 60 unrelated Caucasians belonging to the HapMap Project were also genotyped to confirm the allele frequencies and linkage disequilibrium (LD) pattern observed in our Caucasian livers. The allele frequencies of the–118T9>10, I399C>T, –275T>A, and–2152C>T variants were 0.39, 0.39, 0.02, and 0.02 in the livers, respectively. The I399C>T variant was in complete LD (r2 = 1) with–118T9>10 (linked alleles: C and T9, respectively). Complete LD between these two variants was also found in the HapMap samples (frequencies of–118T9>10 and I399C>T = 0.38). I399C>T and–118T9>10 correlated with neither UGT1A9 activities nor mRNA levels. Because of the low frequencies of the–275T>A and–2152C>T variants, an effect on phenotype could not be evaluated. Our data demonstrate that the common I399C>T and–118T9>10 polymorphisms do not explain interindividual variation in hepatic UGT1A9 activity and mRNA expression and are in complete LD in the donor liver samples we studied.
UDP-glucuronosyltransferases (UGTs) are phase II enzymes involved in the metabolism and detoxification of a wide variety of compounds. UGT1A9, a member of the UGT1A family, plays an important role in the glucuronidation of flavopiridol (Hagenauer et al., 2001; Ramirez et al., 2002), mycophenolic acid (Mackenzie 2000; Basu et al., 2004; Bernard and Guillemette, 2004; Miles et al., 2005; Picard et al., 2005), propofol (Court, 2005), acetaminophen (Court et al., 2001) and propranolol (Sten et al., 2006). UGT1A9 is expressed in multiple tissues, including liver, kidney, colon, and esophagus (Albert et al., 1999; Tukey and Strassburg, 2000). The observed interindividual variability in UGT1A9 substrate clearance (Kuypers et al., 2005) and glucuronidation levels (coefficient of variation 72–99%) (Innocenti et al., 2000) suggests the presence of UGT1A9 variants. Polymorphisms in the promoter region and exon 1 of UGT1A9 have been identified (Desai et al., 2003). Coding variants UGT1A9*2 (C3Y) and UGT1A9*3 (M33T) appear to be relatively rare, with allele frequencies of 1 and 4% in Caucasians populations, respectively (Villeneuve et al., 2003; Ehmer et al., 2004). UGT1A9*3 had greatly reduced SN-38 glucuronidation compared with UGT1A9*1, but a similar effect on flavopiridol glucuronidation was not detected, indicating that this variant may have a substrate-dependent effect (Villeneuve et al., 2003). UGT1A9*4 (Y242X) and UGT1A9*5 (D256N) have been identified only in Japanese populations (Jinno et al., 2003; Saeki et al., 2003).
Additional promoter region polymorphisms have been described, but their relative contribution to the variability in UGT1A9 glucuronidation remains unclear. A variable promoter region repeat (–118T9>10) was reported to correlate with a 2.6-fold higher level of luciferase activity (Yamanaka et al., 2004). However, further studies were not able to confirm this finding (Girard et al., 2006). Additional single nucleotide polymorphisms (SNPs) in the 5′ region (–275T>A, –331C>T, –440T>C, –665C>T, and–2152C>T) are proposed to correlate with variability in UGT1A9 protein expression levels (Girard et al., 2004). In particular, the linked–275A and–2152T alleles have been found to be associated with both higher UGT1A9 protein expression and increased rates of propofol and mycophenolic acid glucuronidation compared with individuals homozygous for the wild-type genotypes (Girard et al., 2004). However, the low allele frequencies of–275A (4–6%) and–2152T (3–6%) in Caucasians may not be sufficient to explain the variability in pharmacokinetic parameters observed with some pharmaceutical compounds metabolized by UGT1A9 (Girard et al., 2004; Innocenti et al., 2005).
The I399C>T variant was recently identified as a novel and common polymorphism in the immediate downstream intron of the UGT1A9 exon and was reported as being in low linkage disequilibrium (LD) (r2 < 0.19) with variants at positions–2152, –275, and–118 (Girard et al., 2006). This SNP was also significantly correlated with increased UGT1A9 protein levels and activities.
To evaluate whether the putative functionality of UGT1A9 variants–118T9>10, I399C>T, –275T>A, and–2152C>T could be confirmed in an independent study, genotype-phenotype correlations were investigated in Caucasian human livers. A luciferase assay was additionally performed to further study the function of the–118T9>10 polymorphism. To confirm allelic frequencies and LD pattern, DNA samples from Caucasians belonging to the International HapMap Project were studied. Our results were then compared with previously published data regarding the functionality of UGT1A9 polymorphisms.
Materials and Methods
Study Populations. Samples (n = 106) from two cohorts of unrelated Caucasian individuals were studied, with approval of institutional review boards.
Cohort 1: Human Liver Samples. DNA, RNA, and human liver microsomes from 46 donors were processed through Dr. Mary Relling's laboratory at St. Jude Children's Research Hospital (Memphis, TN). The hepatic tissue was provided by the Liver Tissue Procurement and Distribution System (funded by Grant NO1-DK-9-2310) and by the Cooperative Human Tissue Network. For preparation of human liver microsomes, diced liver pieces were homogenized in ice-cold buffer (0.1 M KCl-0.1 M Tris-0.02 mM butylated hydroxytoluene-1 mM EDTA, pH 7.4). Microsomes were prepared by differential centrifugation methods and resuspended in 0.25 M sucrose-1 mM EDTA. The protein content was measured using the Bradford (1976) assay.
Cohort 2: CEPH Samples. The following sixty DNA samples belonging to the International Hapmap Project [CEPH (I): Utah residents with ancestry from Northern and Western Europe] were obtained from the Coriell Institute for Medical Research (Camden, NJ): NA12144, NA12145, NA12146, NA12239, NA06994, NA07000, NA07022, NA07056, NA07034, NA07055, NA06993, NA06985, NA12056, NA12057, NA07357, NA07345, NA12043, NA12044, NA11881, NA11882, NA11839, NA11840, NA11829, NA11830, NA11831, NA11832, NA12716, NA12717, NA11992, NA11993, NA11994, NA11995, NA12264, NA12234, NA12154, NA12155, NA12156, NA12248, NA12249, NA12003, NA12004, NA12005, NA12006, NA12750, NA12751, NA12760, NA12761, NA12762, NA12763, NA12812, NA12813, NA12814, NA12815, NA12872, NA12873, NA12874, NA12875, NA12891, and NA12892.
UGT1A9Genotyping. Cohort 1 had been previously genotyped for UGT1A9–118T9>10, –275T>A, and–2152C>T (Innocenti et al., 2005). These variants were genotyped in cohort 2 following the same methods as for cohort 1 (Innocenti et al., 2005). Genotyping of I399C>T in cohort 1 was performed by single base extension and denaturing high-performance liquid chromatography (Wolford et al., 2000) as described previously (Innocenti et al., 2004, 2005). The primer sequences for the polymerase chain reaction (PCR) were 5′-TCTTGAAGATATGTATTTATAAC-3′ (forward) and 5′-TGTATTGGTCTATAAAGGTATT-3′ (reverse). The probe sequence for single base extension was 5′-AACAAGATAGGTGTGAGAATTT-3′ (the underlined base represents a modified base from the reference sequence AF297093.1). Genotypes of I399C>T in cohort 2 are publicly available and were downloaded from the HapMap database (http://www.hapmap.org/index.html.en). Genotypes of–275T>A are also publicly available for the HapMap samples. Comparison of these–275T>A genotypes with the ones determined in our laboratory for cohort 2 matched perfectly.
UGT1A9 Glucuronidation Activity. UGT1A9 activities in 46 human liver microsomes were determined using two specific UGT1A9 substrates, mycophenolic acid (Bernard and Guillemette, 2004; Miles et al., 2005) and flavopiridol (Hagenauer et al., 2001; Ramirez et al., 2002). Optimal conditions for formation of mycophenolic acid phenyl glucuronide (MPAG) were determined by varying incubation times (0–60 min), microsomal protein (0.1–1 mg/ml), and substrate concentration (1–1000 μM). Final conditions included 0.5 mg/ml human liver microsomes, 3 mM uridine 5′-diphospoglucuronic acid (UDPGA), 10 mM magnesium chloride, 50 μg/mg alamethicin, 158 μM mycophenolic acid (Km), and 0.75 mM Tris-HCl buffer (pH 7.4) in a final volume of 200 μl. After preincubation for 5 min at 37°C, reactions were initiated by addition of UDPGA. The reactions were stopped after 20 min with 12 μl of concentrated perchloric acid. After the addition of 10 μl of internal standard (oxazepam, 500 μM), samples were centrifuged (20,817 relative centrifugal force for 15 min at 4°C) and injected (100 μl) into the high-performance liquid chromatography system (Hitachi High Technologies America, Inc., San Jose, CA). The mobile phase consisted of 60:40 (v/v) methanol-0.1% trifluoroacetic acid delivered at a flow rate of 1 ml/min. Separation was achieved using a μBondapak C18 column (10 μm, 3.9 × 300 mm; Waters Corp., Milford, MA) preceded by a Nova-Pak C18 Guardpak (Waters, Milford, MA). The eluate was monitored at 250 nm. Retention times of MPAG, mycophenolic acid, and oxazepam were 4.9, 10.9, and 8.2 min, respectively. Reactions were performed in duplicate. As standards of MPAG were not available, its formation was measured using mycophenolic acid standards, assuming they have similar UV absorbance at 250 nm. Standards were made by combining mycophenolic acid stock and incubation buffer, leaving them for 20 min at 37°C, and processing as described above for human liver microsomes. The quantitation limit was 5 μM (0.5 nmol/min/mg protein). Intra-assay reproducibility was determined by measuring MPAG in 10 separate incubations with the same pool of human liver microsomes. The coefficient of variation was 7.3%. Interassay reproducibility was determined by incubating three pooled human liver microsomes each day for 3 days. The coefficient of variation was 13.0%. To confirm the identity of MPAG (as opposed to the acyl glucuronide, which is specific for UGT2B7) (Picard et al., 2005), we performed two experiments. Screening to confirm specificity for UGT1A9 was performed with recombinant enzymes [UGT1A1, UGT1A3, UGT1A4, UGT1A6, UGT1A7, UGT1A8, UGT1A9, UGT1A10, UGT2B4, UGT2B7, UGT2B15, and UGT2B17 (BD Gentest, Woburn, MA)] and control Supersomes (BD Gentest) as described above. UGT1A9 was the only hepatic UGT forming the peak eluting at 4.9 min. Additionally, as MPAG is subject to enzymatic rather than alkaline hydrolysis, we treated two incubations of human liver microsomes with 2000 units of β-glucuronidase [type VII from Escherichia coli (Sigma-Aldrich, St. Louis, MO), dissolved in 0.1 M sodium phosphate buffer, pH 6.4], and two other incubations with 0.5 M sodium hydroxide. After incubation for 1.2 h at 37°C, perchloric acid was added, and samples were treated as described above. Control incubations containing no β-glucuronidase or sodium hydroxide were treated identically. Incubation with β-glucuronidase resulted in total disappearance of the peak eluting at 4.9 min. In contrast, the peak was intact in the incubations treated with sodium hydroxide, confirming we were looking at the phenol glucuronide. The MPAG peak was observed in the control incubations when β-glucuronidase and sodium hydroxide were absent (data not shown).
Incubation of human liver microsomes with flavopiridol was performed under linear conditions with regard to concentrations of microsomal protein, UDPGA, and substrate as well as incubation time, using the conditions described previously (Ramirez et al., 2002). Briefly, incubations containing 3 mg/ml microsomal protein, 3 mM UDPGA, 10 mM magnesium chloride, 500 μM flavopiridol (Km), and 50 mM Tris-HCl (pH 7.4) were performed for 3 h at 37°C. Samples were analyzed by high-performance liquid chromatography with UV detection (265 nm). Formation of flavopiridol-7-glucuronide (FLAVO-7-G) was expressed as the ratio of the peak heights of the glucuronide to flavone (internal standard) as standards were not available and our purpose was to compare the difference in glucuronide formation among livers rather than quantify the exact formation of glucuronide.
Real-Time PCR.UGT1A9 mRNA expression was determined in 35 liver samples by two-step real-time PCR using the Mx3000P system (Stratagene, Cedar Creek, TX). Briefly, cDNA was synthesized using the iScript cDNA Synthesis Kit (Bio-Rad Laboratories, Hercules, CA). Real-time PCR reactions were performed using IQSYBR Green Supermix (Bio-Rad, Hercules, CA) and the UGT1A9 specific primers 5′-GAGGAACATTTATTATGCCACCG-3′ (forward) and 5′-TGCCCAAAGCATCAGCAATT-3′ (reverse). 18S RNA was used as an internal control with primers 5′-CGATGCTCTTAGCTGAGTGT-3′ (forward) and 5′-GGTCCAAGAATTTCACCTCT-3′ (reverse). Reactions were performed in triplicate and included standard curves for both the target and control genes. Expressions of UGT1A9 and 18S RNA were calculated on the basis of the equations generated from their respective standard curves. The UGT1A9 mRNA levels were expressed as the ratio of UGT1A9 to the control gene, 18S, multiplied by a dilution factor of 200 as cDNA was diluted before quantification of 18S.
Construction ofUGT1A9Promoter Reporter Vectors. The UGT1A9 promoter region (–1 to–1040, according to the A in the translation initiation codon ATG) containing either–118T9 or–118T10 was subcloned into the pGL3-basic vector containing the firefly luciferase enzyme (Promega, Madison, WI). Briefly, a PCR was carried out with Proofstart DNA polymerase (QIAGEN, Valencia, CA) to generate the amplicon from a–118T9>10 heterozygote DNA with primer sequences UGT1A9KPN (5′-GATCAAGGTACCGTGTTCTTGCCGAGGCCTTCT-3′) and UGT1A9HIND (5′-ATGAATAAGCTTCAGAGAACTGCAGCTGAGAGCAA-3′). To facilitate the subcloning, the UGT1A9KPN and UGT1A9HIND primers were designed to contain a KpnI and a HindIII site, respectively. The modified nucleotides are underlined in the primer sequence listed above. After the PCR products were digested, the 1040- or 1041-bp products were cloned into the KpnI/HindIII site of the pGL3-basic vector to construct vectors designated as pGL3UGT1A9luc*9 or pGL3UGT1A9luc*10, respectively. Plasmid DNA was prepared, and all clones were sequenced to exclude PCR errors and ensure the orientation of the construct before transfection.
Cell Culture and Transient Transfection. HepG2 and HEK-293 cell lines were maintained in Dulbecco's Modified Eagle's Medium (Invitrogen, Carlsbad, CA) supplemented with 2 mM glutamine and 10% fetal bovine serum in a humidified atmosphere (5% CO2-95% air) at 37°C. To compare the relative expression of the luciferase gene, pGL3UGT1A9luc*9 and pGL3UGT1A9luc*10 were transfected into the HepG2 and HEK-293 cell lines. The pGL3-basic reporter was transfected into HEK-293 cell lines. Briefly, the cell lines were seeded into a 12-well plate 1 day before transfection. The cells were then cotransfected with 1 μg of pGL3UGT1A9luc and 0.1 μg of pRL-TK plasmids (Promega) and combined with 2.5 μl of Tfx-20 (Promega) in 400 μl of serum-free medium (Invitrogen), according to the manufacturer's instructions. The cells were incubated at 37°C under an atmosphere of 5% CO2 for 1 h. After incubation, the culture medium (2 ml) was added to the wells, and cells were incubated for 24 additional h. Luciferase activity was measured using a dual-luciferase reporter assay system (Promega), following the manufacturer's instructions. Relative luciferase activities were normalized to the activities of the internal control for Renilla luciferase activity. The transfections were performed in triplicate and repeated twice.
Statistics. The frequency distribution of the UGT1A9 mRNA levels (KS distance = 0.28, p = 0.009) and the MPAG formation rates (KS distance = 0.21, p = 0.04) were not normally distributed. They were log-transformed before statistical analyses, after which they passed the normality test (KS distance = 0.10, p > 0.10). Formation of FLAVO-7-G was apparently normal (KS distance = 0.09, p > 0.10). Pearson analysis was used to test the level of correlation between glucuronidation levels of mycophenolic acid and flavopiridol and between the activities and mRNA levels. Statistical analysis of the correlation between phenotypes and genotypes was performed using the one-way analysis of variance test or an unpaired t test. A chi-square test was used to determine the statistical significance of the relationship between the haplotype frequencies. The Mann-Whitney U test was used to compare the differences in relative luciferase activities between the T9 and T10 alleles. All statistical tests were done using GraphPad Prism (version 4.00 for Windows; GraphPad Software Inc., San Diego, CA; www.graphpad.com). Hardy-Weinberg equilibrium was checked with Arlequin software (version 2.0; University of Geneva, Geneva, Switzerland). LD was calculated using LDPlotter (https://innateimmunity.net/IIPGA2/Bioinformatics). Haplotypes were estimated with Phase (version 2.02; University of Washington, Seattle, WA). The threshold value for statistical significance was set at 0.05.
Results
UGT1A9mRNA Levels and Glucuronidation Activities. Interindividual variation in mRNA levels was determined in 35 livers. The median (not log-transformed) UGT1A9 mRNA levels were 448 (range, 11–9493). Formation of MPAG and FLAVO-7-G was measured in 46 liver samples. The apparent formation rates of MPAG varied approximately 11-fold, ranging from 0.8 to 8.4 nmol/min/mg protein. The median value was 2.5 nmol/min/mg protein. The FLAVO-7-G/flavone peak height ratio varied approximately 5-fold, ranging from 0.11 to 0.54. The mean value (±S.D.) and coefficient of variation were 0.30 ± 0.10 and 33%, respectively. The glucuronidation activities of both UGT1A9 substrates were significantly correlated (r = 0.62, p < 0.0001, n = 46). A significant correlation was also observed between the mRNA levels and the glucuronidation activities of both mycophenolic acid (r = 0.52, p = 0.001, n = 35) and flavopiridol (r = 0.43, p = 0.01, n = 35).
Age and gender did not have a significant effect on UGT1A9 phenotypes (p > 0.05, univariate linear regression, data not shown). A history of alcohol and drug intake (phenobarbital/dexamethasone) was available for 28 donors; the smoking history of 39 individuals was known. No statistically significant associations were found between intake of any of these UGT inducers and phenotypes (p > 0.05, univariate linear regression; data not shown). Information on hepatic disease was not available for the sample donors.
UGT1A9Variants. The allele frequencies of the four variants were similar in cohorts 1 and 2 (Table 1). The–118T9>10 and I399C>T variants were present at high frequency, whereas the–275T>A and–2152C>T variants were relatively rare. All alleles were in Hardy-Weinberg equilibrium. In both cohorts, complete LD (r2 = 1.00) was observed between 1)–118T9>10 and I399C>T and 2)–275T>A and–2152C>T. There was very low LD between–118 and–275 (r2 = 0.01 and 0.05 for cohorts 1 and 2, respectively). The haplotype structure and frequencies are presented in Table 2. The frequencies of haplotypes I and II were very similar in cohorts 1 and 2 (χ2 = 3.2, df = 2, p = 0.21).
Association between mRNA Levels, Activities, andUGT1A9Polymorphisms. Correlations between genotypes and phenotypes were only tested with–118T9>10 and I399C>T because of the low frequency of the–275T>A and–2152C>T polymorphisms in our liver samples (2%). No significant differences were observed between the–118T9/9, –118T9/10, and–118T10/10 genotypes for mRNA expression or formation of MPAG or FLAVO-7-G (Table 3). As the–118T9>10 and I399C>T polymorphisms were in perfect LD, associations identical to those observed with–118T9>10 were found between UGT1A9 phenotypes and I399C>T (Fig. 1).
Luciferase Reporter Gene Assay. Significant basal promoter activity of both T9 and T10 constructs was observed compared with the pGL3-basic empty vector (p < 0.01, data not shown). No significant differences in luciferase activity were observed between the–118T9 and–118T10 alleles (p > 0.05).
Discussion
Although clinical studies show interindividual variability in the clearance of mycophenolic acid (Kuypers et al., 2005) and in the capacity to glucuronidate flavopiridol (Innocenti et al., 2000), we find no evidence that variability in UGT1A9 expression or activity is related to previously identified common polymorphisms (Girard et al., 2006). We first focused on the high frequency variant–118T9>10. The functionality of the–118T9>10 polymorphism had been suggested by a reporter assay that used a small promoter construct (–152 to–171bp, according to the A in the translation initiation codon ATG) transfected into HepG2 cells (Yamanaka et al., 2004). As this construct included only the proximal promoter, we thought it might be insufficient to reflect the regulation of the promoter activity in vivo. To include the distal promoter, we constructed a longer subclone spanning more than 1 kilobase of the 5′-flanking sequence, a region that has been demonstrated to have full UGT1A9 promoter activity (Gregory et al., 2003, 2004). By using two different cell lines, we observed similar promoter activity between the two alleles, suggesting that the promoter polymorphism does not alter function. This is in agreement with others studies reporting only a marginal fold elevation (1.35–1.43) of reporter expression by the T10 allele, compared with the T9 allele, using 154- and 170-bp constructs (Girard et al., 2006) and showing that T insertions/deletions in the UGT1A8, UGT1A9, and UGT1A10 promoters do not affect gene transcription (Mackenzie et al., 2005).
For phenotypic assessment of UGT1A9 activity, we used flavopiridol and mycophenolic acid. Flavopiridol has been shown to be a relatively specific substrate for UGT1A9 in experiments using recombinant enzymes and propofol as enzyme inhibitor (Hagenauer et al., 2001; Ramirez et al., 2002). Even though we could not quantify the rate of formation of FLAVO-7-G using standards, we believe we can detect accurately differences in glucuronidation activities, as the activity measurements were done under linear assay conditions and using a substrate concentration equal to the Km determined under the experimental conditions reported in this study. In addition, a significant and modest correlation was observed between the glucuronidation of flavopiridol and mycophenolic acid, a commonly used UGT1A9 probe (Bernard and Guillemette, 2004; Girard et al., 2004; Miles et al., 2005). However, the lack of high correlation suggests that both drugs may be glucuronidated to a certain extent by other UGTs (Hagenauer et al., 2001; Ramirez et al., 2002; Bernard and Guillemette, 2004; Miles et al., 2005; Picard et al., 2005). Levels of correlations similar to those observed between the formation of FLAVO-7-G and MPAG in our study (r = 0.62) were observed between the glucuronidation rates of propofol and MPAG (r = 0.62) in another study (Girard et al., 2004). The lack of association observed between–118T9>10 and either the UGT1A9 mRNA level or activities is consistent with other studies (Girard et al., 2004; Innocenti et al., 2005). Overall, these results suggest that the–118 polymorphism is unlikely to regulate UGT1A9 expression.
We could not confirm in our livers either the low level of LD previously observed between I399C>T and the other three UGT1A9 variants (Girard et al., 2006), or the functionality of I399C>T. We found complete LD between I399 and–118 not only in our liver set but also in the HapMap CEU samples for which the I399 genotypes are publicly available. We could not confirm a previous report of higher glucuronidation rates of mycophenolic acid in individuals homozygous for TT compared with CC (Girard et al., 2006). To the best of our knowledge, that report did not control for multiple comparisons and might have detected a false-positive association.
As for the effect of the–2152C>T and–275T>A variants, we could not study their association with phenotypes because of their low frequency in our liver samples. Although heterozygous samples have been associated with higher UGT1A9 protein expression (Girard et al., 2006), 9% of the samples studied were of non-Caucasian ethnicities, which might have confounded the findings. In addition, as these SNPs are not common in Caucasians, their clinical effect might be difficult to assess in clinical trials that enroll subjects from predominantly Caucasian populations.
All variants in the present study were genotyped in a previous study of 42 Caucasians (Girard et al., 2006). This allows comparison of the allele and haplotype frequencies between the two studies. For simplicity, we will refer to this third sample set as cohort 3. Variant allele and haplotype frequencies are listed in Tables 1 and 2, respectively. We predicted three haplotypes for 106 samples (cohorts 1 and 2 combined). In contrast, six haplotypes were inferred for the 42 samples in Girard et al. (cohort 3). The haplotypes from cohort 3 were significantly different from those predicted for cohorts 1 (χ2 = 22.0, df = 5, p = 0.0005) and 2 (χ2 = 30.5, df = 5, p < 0.0001). A possible explanation for the difference in haplotype structure and LD pattern observed between the two studies might be population stratification in the samples used in cohort 3 (Girard et al., 2006). As our data are strikingly similar to those of the CEPH families from the HapMap, we believe that our estimates are likely to be highly representative of a Caucasian population.
As the UGT1A9 polymorphisms studied so far do not seem to contribute to the phenotypic variability of the gene, interindividual variability in UGT1A9 function might more likely be controlled by trans-acting factors. HNF1α, HNF4α, and Cdx2 have been shown to regulate the hepatic transcription of the UGT1A9 gene (Gregory et al., 2004; Barbier et al., 2005). Thus, comprehensive studies investigating the effect of trans-acting factors may lead to the discovery of new polymorphisms that might help elucidate the observed UGT1A9 functional variability.
Footnotes
-
This work was supported by the Pharmacogenetics of Anticancer Agents Research Group (http://pharmacogenetics.org) [National Institutes of Health (NIH)/National Institute of General Medical Sciences (NIGMS) Grant U01GM61393]. Data will be deposited into PharmGKB (supported by NIH/NIGMS U01GM61374, http://pharmgkb.org/).
-
J.R. and W.L. contributed equally to this work.
-
Article, publication date, and citation information can be found at http://dmd.aspetjournals.org.
-
doi:10.1124/dmd.107.015446.
-
ABBREVIATIONS: UGT, uridine diphosphate-glucuronosyltransferase; SN-38, 7-ethyl-10-hydroxycamptothecin; SNP, single nucleotide polymorphism; LD, linkage disequilibrium; CEPH, Centre d'Etude du Polymorphisme Humain; PCR, polymerase chain reaction; MPAG, mycophenolic acid phenyl glucuronide; UDPGA, uridine 5′-diphospoglucuronic acid; FLAVO-7-G, flavopiridol-7-glucuronide; bp, base pair(s); KS, Komolgorov-Smirnoff.
- Received February 27, 2007.
- Accepted August 29, 2007.
- The American Society for Pharmacology and Experimental Therapeutics