Abstract
Genome wide association studies have identified single nucleotide polymorphisms (SNP) within fibroblast growth factor receptor 2 (FGFR2) as one of the highest ranking risk alleles in terms of development of breast cancer. The potential effect of these SNPs, in intron two, was postulated to be due to the differential binding of cis-regulatory elements, such as transcription factors, since all the SNPs in linkage disequilibrium were located in a regulatory DNA region. A Runx2 binding site was reported to be functional only in the minor, disease associated allele of rs2981578, resulting in increased expression of FGFR2 in cancers from patients homozygous for that allele. Moreover, the increased risk conferred by the minor FGFR2 allele associates most strongly in oestrogen receptor alpha positive (ERα) breast tumours, suggesting a potential interaction between ERα and FGFR signalling. Here, we have developed a human cell line model system to study the effect of the putative functional SNP, rs2981578, on cell behaviour. MCF7 cells, an ERα positive breast cancer cell line homozygous for the wild-type allele were edited using a Zinc Finger Nuclease approach. Unexpectedly, the acquisition of a single risk allele in MCF7 clones failed to affect proliferation or cell cycle progression. Binding of Runx2 to the risk allele was not observed. However FOXA1 binding, an important ERα partner, appeared decreased at the rs2981578 locus in the risk allele cells. Differences in allele specific expression (ASE) of FGFR2 were not observed in a panel of 72 ERα positive breast cancer samples. Thus, the apparent increased risk of developing ERα positive breast cancer seems not to be caused by rs2981578 alone. Rather, the observed increased risk of developing breast cancer might be the result of a coordinated effect of multiple SNPs forming a risk haplotype in the second intron of FGFR2.