Functional genetic polymorphisms in the aromatase gene CYP19 vary the response of breast cancer patients to neoadjuvant therapy with aromatase inhibitors

Abstract

Aromatase (CYP19) is a critical enzyme in estrogen biosynthesis and aromatase inhibitors (AI) are employed widely for endocrine therapy in postmenopausal women with breast cancer. We hypothesized that single nucleotide polymorphisms (SNPs) in the CYP19 gene may alter the effectiveness of AI therapy in the neoadjuvant setting. Genomic DNA was obtained for sequencing from 52 women pre-AI and post-AI treatment in this setting. Additionally, genomic DNA obtained from 82 samples of breast cancer and 19 samples of normal breast tissue was subjected to resequencing. No differences in CYP19 sequence were observed between tumor and germ-line DNA in the same patient. A total of 48 SNPs were identified including 4 novel SNPs when compared with previous resequencing data. For genotype-phenotype association studies, we determined the levels of aromatase activity, estrone, estradiol, and tumor size in patients pre-AI and post-AI treatment. We defined two tightly linked SNPs (rs6493497 and rs7176005 in the 5'-flanking region of CYP19 exon 1.1) that were significantly associated with a greater change in aromatase activity after AI treatment. In a follow-up study of 200 women with early-stage breast cancer who were treated with adjuvant anastrozole, these same two SNPs were also associated with higher plasma estradiol levels in patients pre-AI and post-AI treatment. Electrophoretic mobility shift and reporter gene assays confirmed likely functional effects of these two SNPs on transcription of CYP19. Our findings indicate that two common genetic polymorphisms in the aromatase gene CYP19 vary the response of breast cancer patients to aromatase inhibitors.

Figure 1

Human CYP19 gene structure and polymorphisms. The figure shows a schematic representation of the human CYP19 gene. Exons are represented as rectangles, with black rectangles indicating the open reading frame (ORF) and white indicating untranslated regions (UTRs). Arrows indicate the locations of polymorphisms, with colors representing minor allele frequencies. Red arrows indicate polymorphisms with frequencies of >10%, while blue and black arrows indicate polymorphisms with frequencies from 1% to 10% and of <1%, respectively. “*” represents SNPs that were “novel” to this study.

Figure 2

Functional characterization of the CYP19 Arg128 variant allozyme. (A). Enzyme activity assay. CYP19 enzyme activity was measured with microsomal preparations from cells transfected with “empty” vector, WT and Arg128 constructs using [1β³H] androst-4-ene-3,17-dione as substrate. Error bars are mean ± SEM for 3 independent experiments. (B). Representative letrozole inhibition study of WT CYP19. Double inverse plots were created using 6 concentrations of [1β³H] androst-4-ene-3,17-dione in the presence of increasing concentrations of letrozole (0-0.8nM).

Figure 3

Genotype-phenotype association study. Quartile box plot was plotted between the genotypes of the two SNPs, rs6493497 (−588) and rs7176005 (−144), and change in aromatase activity measured in tumor tissue pre- and post-AI treatment.

Figure 4

Functional characterization of CYP19 rs6493497 (−588) and rs7176005 (−144). (A) Exon array analysis of CYP19 expression in 7 breast tumor samples. Exon array probe sets (1-30) for CYP19 are plotted against expression levels. The arrow indicates probe sets for exon 1.1. The red rectangular box indicates the coding region of the gene. The cutoff for the expression level is 3. (B) EMSA of SNPs at (−588) and (−144) upstream of exon 1.1. EMSA were performed using biotin-labeled probes containing WT or variant sequences using nuclear extract from SKBR03 Cells. Competition reactions were performed with 400-fold excess of unlabeled probes. (C) Reporter gene assays. Reporter gene constructs were created for WT (GC) and the observed haplotypes (AT and AC) for the rs6493497 and rs7176005 SNPs. Luciferase activity was corrected for Renilla luciferase activity and is expressed as a percent of the WT activity. Values are mean ± SEM for 3 independent experiments.