RESPONSE: Re: Modification of Clinical Presentation of Prostate Tumors by a Novel Genetic Variant in CYP3A4

Abstract

We are pleased to see the work of Ando et al. that addresses the functional significance of the cytochrome P450-3A4 (CYP3A4) genetic variant originally reported by our group ( 1 ). We would like to comment on the findings reported in their correspondence and contribute additional data that complement the findings of those authors. We have previously reported that the frequency of the CYP3A4 variant allele differs substantially across ethnic groups ( 2 ). We estimated that the frequency of the variant allele among healthy, unrelated individuals was 53% in African-Americans (n = 70), 9% in U.S. Caucasians (n = 132), and 0% in Taiwanese (n = 130). The observation by Ando et al. of no variant alleles in 123 Japanese individuals provides further evidence that the CYP3A4 variant may be found only rarely, if at all, in Asians. However, the high frequency of variant alleles in African-Americans, and the significant differences in allele frequencies between the racial groups studied to date, could in part explain differences in CYP3A4-associated disease susceptibility or drug metabolism by race. The results of expression analyses presented by Ando et al. are similar to those recently performed in our laboratory. To evaluate the functional significance of the CYP3A4 variant, we generated two eukaryotic reporter constructs (Fig. 1, A). One construct contained the wild-type (W) CYP3A4 promoter sequence, and the other contained the variant (V) CYP3A4 promoter sequence. The expression of luciferase by these promoter constructs was studied in the hepatoma cell line HepG2 and in the breast cancer cell line MCF7. As shown in Fig. 1 , B, average luciferase expression was 1.4-fold higher for the V promoter compared with the W promoter in HepG2 cells (χ 21 = 18.04; P <.0001) and 1.9-fold higher for the V compared with the W promoter in MCF7 cells (χ 21 = 33.29; P <.0001). Both sets of experiments were repeated in triplicate nine times. These results suggest that the CYP3A4-V promoter may be associated with higher expression of CYP3A4 than the CYP3A4-W promoter. Ando et al. reported no statistically significant differences in nifedipine oxidation or CYP3A4 protein levels between wild-type and variant homozygotes. However, they do report 1.6-fold higher nifedipine oxidation and a 2.1-fold higher CYP3A4 protein levels in V versus W genotypes. As suggested by those authors, the lack of a statistically significant difference in these levels may reflect insufficient statistical power to detect the reported magnitude of effect. Taken together with our results, however, it appears that CYP3A4-V may be associated with CYP3A4 overexpression compared with CYP3A4-W. While the magnitude of these differences may be small, it is possible that the downstream effect of these genetic differences on some metabolic pathways could be physiologically relevant. Additional research will be required to define the regulatory features of this potential difference and to clarify the biologic mechanism by which CYP3A4 genotype may explain the observed epidemiologic associations of CYP3A4 with prostate cancer ( 1 ) and childhood leukemias ( 3 ). A) Diagram of the reporter-gene construct used to evaluate the effect of a 1-kilobase (kb)-segment of the CYP3A4 promoter on luciferase expression. The vector used is a pGL3 Enhancer vector (Promega Corp., Madison, WI) containing an SV40 enhancer sequence and the wild-type (W) or variant (V) CYP3A4 promoter driving luciferase expression. Genomic DNA from a homozygous wild-type (W/W) individual and a homozygous variant (V/V) individual was amplified by polymerase chain reaction (PCR) by use of a forward primer: 5′-TCCTGTTTCCAGACATGCAG-3′ and reverse primer: 5′-TCAGTGAGGCTGTT GGATTG-3′. 100 ng of genomic DNA was used as template in a 50-μL reaction containing 2.5 U of PWO Taq polymerase and 5 μL of 10 × PCR buffer without MgSO 4 (Boehringer Mannheim, Indianapolis, IN), 0.2 m M deoxynucleoside triphosphate, 1 m M MgSO 4 , and 0.5 μ M primers. Blunt-ended PCR products were ligated into the PCR-Blunt vector (Invitrogen, Carlsbad, CA). All CYP3A4 DNA sequences were confirmed after amplification by use of an ABI Prism 377 sequencer (The Perkin-Elmer Corp., Foster City, CA). The W and V constructs were generated by inserting restriction digested CYP3A4 5′-flanking DNA fragments corresponding to promoter sequence −14 to −1018, which were digested with XhoI and KpnI (New England Biolabs, Beverly, MA) into the pGL3-Enhancer vector (Promega Corp.). *denotes the approximate position of the CYP3A4 variant (an A to G transition in the nifedipine-specific element, or NFSE) located −290 bp from the transcription start site of CYP3A4. B) Luciferase reporter gene expression by CYP3A4 genotype. The expression of luciferase under the influence of either the wild-type (W) or variant (V) CYP3A4 promoter is reported relative to that of empty vector. Cell culture and transient transfections were undertaken by use of HepG2 and MCF-7 cell lines maintained in high glucose Dulbecco's minimal Eagle medium (Life Technologies, Inc. [ GIBCO BRL], Gaithersburg, MD) supplemented with 10% fetal bovine serum (Life Technologies, Inc.). Transient transfection was performed after plating 3 × 10 5 cells/well in six-well plates plated the day prior to transfection using Superfect Reagent (Qiagen, Valencia, CA). In general, the transfection mixes in each well contained 0.5 μg pCH110 β-galactosidase (Pharmacia, Piscataway, NJ) and 1.0 μg carrier plasmid. The cells were washed with 1 × phosphate buffer solution and covered with fresh medium after 3 hours incubation at 37 °C and 5% CO 2 . Nine triplicate repeat assays for β-galactosidase and luciferase enzyme activities were performed after 48 hours post-transfection incubation by use of the respective enzyme assay kits manufactured by Promega.