Fine-scale mapping of the 4q24 locus identifies two independent loci associated with breast cancer risk

Abstract

Background: A recent association study identified a common variant (rs9790517) at 4q24 to be associated with breast cancer risk. Independent association signals and potential functional variants in this locus have not been explored.

Methods: We conducted a fine-mapping analysis in 55,540 breast cancer cases and 51,168 controls from the Breast Cancer Association Consortium.

Results: Conditional analyses identified two independent association signals among women of European ancestry, represented by rs9790517 [conditional P = 2.51 × 10(-4); OR, 1.04; 95% confidence interval (CI), 1.02-1.07] and rs77928427 (P = 1.86 × 10(-4); OR, 1.04; 95% CI, 1.02-1.07). Functional annotation using data from the Encyclopedia of DNA Elements (ENCODE) project revealed two putative functional variants, rs62331150 and rs73838678 in linkage disequilibrium (LD) with rs9790517 (r(2) ≥ 0.90) residing in the active promoter or enhancer, respectively, of the nearest gene, TET2. Both variants are located in DNase I hypersensitivity and transcription factor-binding sites. Using data from both The Cancer Genome Atlas (TCGA) and Molecular Taxonomy of Breast Cancer International Consortium (METABRIC), we showed that rs62331150 was associated with level of expression of TET2 in breast normal and tumor tissue.

Conclusion: Our study identified two independent association signals at 4q24 in relation to breast cancer risk and suggested that observed association in this locus may be mediated through the regulation of TET2.

Impact: Fine-mapping study with large sample size warranted for identification of independent loci for breast cancer risk.

Figure 1. Regional plot of genetic variants associated with breast cancer risk at 4q24

A) The index SNP rs9790517 is plotted in diamond purple. The LD (r²) for the index SNP with each SNP was computed based on European ancestry subjects included in the 1000 Genome Mar 2012 EUR. P values were from the single-marker analysis based on logistic regression models after adjusted for age, study sites and the first six principal components plus one additional principal component for the LMBC in analyses of data from European descendants. The plot was generated using LocusZoom (50).

Figure 2. Functional annotation of SNPs association with breast cancer risk at 4q24

A) Epigenetic landscape at 4q24 risk locus for breast cancer. From top to bottom, RefSeq genes (TET2 and PPA2), layered H3K4Me1, H3K4Me3 and H3K27Ac histone modifications, DNase clusters, annotation using chromatin states on the ENCODE cell lines, and H3K27Ac histone modification in MCF-7, predicted enhancers reported in the Hnisz et al. study, regulatory elements of enhancers associated with TSS and TSSs from the FANTOM5 project and ChIA-PET interactions in MCF-7 cell (mediated by RNA polymerase 2) between enhancers and TET2 promoter are shown. The signals of different layered histone modifications from the same ENCODE cell line are shown in the same color (The detailed color scheme for each ENCODE cell line described in the UCSC genome browser). The red and orange colors in chromatin states refer to active promoter and strong enhancer regions, respectively (The detailed color scheme of the chromatin states described in the previous study (27)). For ChIA-PET track, black lines represented interactions with the promoter region (−1500/+500) of TET2, and gray lines represent chromatin interactions that do not involve the TET2 promoter region. Purple and green lines represent interactions within +/− 500pb of rs73838678 and rs62331150 variants, respectively. B) Epigenetic signals of two potential functional variants rs73838678 and rs62331150. From top to bottom, lanes showing that the variant mapped to TF predicted binding motifs, TF ChIP-Seq binding peaks and DNase I hypersensitivity sites. The corresponding location of the variant is indicated by dashed line. C) LD plot for breast cancer risk associated SNPs at 4q24. In the top lane, two SNPs representing independent association signals are indicated by the black arrows. The index SNP is indicated by the red arrow. In the bottom lane, two LD SNP blocks were shown based on r² values, which were computed based on the genotype data from the BCAC.

Figure 3. The association between SNP rs62331150 and TET2 expression in breast cancer tissues from TCGA

The reference allele G of rs62331150 is significantly associated with the increased gene expression relative to the risk allele T.