The Use of Allelic Imbalance to Ascertain Cis-Regulation for Human UGT2B7 in Vivo
Pin-Yi Wang,Dezheng Huo,Chang Sun,Olufunmilayo I. Olopade
DOI: https://doi.org/10.1007/s00228-013-1538-0
2013-01-01
European Journal of Clinical Pharmacology
Abstract:Dear Editor, Glucuronidation plays an important role in clearing numerous endobiotics and xenobiotics from the human body, including steroid hormones, bile acids, carcinogens and clinical drugs and is catalyzed by UDP-glucuronosyltransferases (UGT) [1–4]. Considering this and the fact that steroid hormones play a central role in breast cancer risk, it has long been proposed that UGTs are involved in breast cancer susceptibility [5,6]. Based on the chromosome location and homology, this gene family can be classified into two major subfamilies, UGT1 and UGT2, the latter of which can be further divided into UGT2A and UGT2B [3]. Among UGT2B family, UGT2B7 is most relevant due to its high expression level [7–9] and activity [10], and wide substrate spectrum [10,11]. As a consequence, the genetic variants in this gene, especially those that can affect the regulation of gene expression or alter the protein sequence and further enzyme activity, have received much attention in attempt to understand the inter-individual differences in drug response and susceptibility to breast cancer. Recently, by transient transfection, it has been shown that −900G/A (relative to translation start; rs7438135) could influence promoter activity in liver cell line in both basal [12] and oxidative stress conditions [13] in vitro. However, it still remains unclear whether this polymorphism can affect the regulation of UGT2B7 expression in vivo. Interestingly, two adjacent coding SNPs, 801A/T (rs7438284) and 802T/C (rs7439366, causing 268Y/H amino acid substitution), were found to be in complete linkage disequilibrium (LD) with this putative regulatory SNP, −900G/A, and were used to categorize two major haplotypes, UGT2B7*1 and *2 [12,14–16]. Therefore, if this SNP indeed controls UGT2B7 expression, differential allelic expression (DAE; or allelic imbalance, AI) should be observed between two haplotypes. To this end, we retrieved 31 normal liver (3 European American [CA] and 1 African American [AA], 27 unknown) tissue samples. All liver tissues came from biospecimen bank of University of Chicago and this study was approved by ethics committee (approval ID 10760B). RNA and DNA were isolated by RNeasy Lipid Tissue and QIAamp DNA Mini Kit (Qiagen, Valencia, CA), respectively. cDNA was synthesized by High Capacity Reverse Transcription Kit (Applied biosystems, Foster City, CA). Genotype of UGT2B7 haplotype was determined by a Taqman genotyping assay with primers GGAAAGCTGACGTATGGCTTATT and AAAGCCAACAAAATAAAACCAACA, and probes 6FAM-TCAGTTTCCTCATCCAC and VIC-TCAGTTTCCATATCCAC (target in bold; Applied Biosystems) according to the manufacturer’s protocol. Among all tissues, 11 liver samples were heterozygous for the two haplotypes. To evaluate the allelic imbalance in 801 and 802 positions, allele-specific real time PCR was performed in cDNA from heterozygous individuals by the Taqman Assay with primers GACAATGGGGAAAGCTGACG and GGAGTCCTCCAACAAAATCAACA and abovementioned probes. For normalization, a heterozygous genomic DNA sample was serially diluted as standard. All real time PCR assays were performed in triplicate for each sample and the AI ratio was expressed as Type2quantity/Type1quantity, in which type 1 and type 2 indicated 801T-802C and 801A-802T, and were in LD with haplotype 1 and 2, respectively. All real time PCR and genotyping reading were performed on a StepOne Plus Realtime PCR System (Applied Biosystems). To make sure that the two alleles had the same efficiency in our allele-specific real time PCR and that our AI assay was accurate, we first quantified the gDNA mixes by type 1 and 2 with known ratios (4:1, 2:1, 1:1, 1:2, and 1:4). The observed ratio presented a highly linear relation with the expected one (r2=0.991, P<10−17; result not shown), thus confirming that our assay was sensitive enough to differentiate two alleles, at least in the 0.25–4 ratio interval, which was consistent with our recent study [17]. As shown in Figure 1, nearly all heterozygous samples presented an excess of type 1 over type 2, with a significant deviation from 1:1 ratio (mean ± standard deviation [SD], 0.81±0.11; 95% confidence interval [CI] 0.73–0.88; t-test, P<0.001). This result showed that −900G/A can indeed influence UGT2B7 gene expression in liver. However, the proportion of haplotype 1 was only 13.6%–37.0% higher than that of haplotype 2 on average in vivo (see Fig 1), which was much lower than previous result (around one fold increasing) in vitro [12,13]. This discrepancy might hint the existence of one or more cis-regulatory variants that can hamper the effect of −900G/A and the importance of using both in vivo and in vitro approaches to investigate gene expression regulation. Moreover, the limited difference in gene expression in vivo might interpret the lack of association between haplotype and multiple drugs metabolism [18–20], especially when no enzyme activity difference on these drugs was observed between H and Y proteins. Indeed, considering the remarkable difference in promoter activity between two haplotypes in vitro, it could be easily deduced that the regulatory SNP in promoter should have some functional consequence in drug metabolism. However, multiple studies failed to obtain any difference between two haplotypes [18–20]. This result could be explained by our observation that the promoter activity difference in vivo is not as great as that in vitro. Moreover, the diverse metabolism efficiency on some drugs between two haplotypes in previous study [21–23] was more likely to be attributed to enzyme activity instead of promoter. Fig 1 The ratio of Type 2/Type 1 in liver. Each bar in x-axis indicates one individual while the height of each bar denotes the AI ratio, which is expressed as mean ± standard deviation (SD). Type 1 and 2 represent 801T-802C and 801A-802T, respectively. ...