A multistage genome-wide association study in breast cancer identifies two new risk alleles at 1p11.2 and 14q24.1 (RAD51L1)

Abstract

We conducted a three-stage genome-wide association study (GWAS) of breast cancer in 9,770 cases and 10,799 controls in the Cancer Genetic Markers of Susceptibility (CGEMS) initiative. In stage 1, we genotyped 528,173 SNPs in 1,145 cases of invasive breast cancer and 1,142 controls. In stage 2, we analyzed 24,909 top SNPs in 4,547 cases and 4,434 controls. In stage 3, we investigated 21 loci in 4,078 cases and 5,223 controls. Two new loci achieved genome-wide significance. A pericentromeric SNP on chromosome 1p11.2 (rs11249433; P = 6.74 x 10(-10) adjusted genotype test, 2 degrees of freedom) resides in a large linkage disequilibrium block neighboring NOTCH2 and FCGR1B; this signal was stronger for estrogen-receptor-positive tumors. A second SNP on chromosome 14q24.1 (rs999737; P = 1.74 x 10(-7)) localizes to RAD51L1, a gene in the homologous recombination DNA repair pathway. We also confirmed associations with loci on chromosomes 2q35, 5p12, 5q11.2, 8q24, 10q26 and 16q12.1.

Figure 1. Results of Combined Stage 1 and 2

This figure includes the most promising SNP associations based on a combined analysis of Stage 1 and Stage 2. A joint analysis of the genotypes was performed using an age, study design and population stratification-adjusted logistic regression analysis (2 df test). Dashed vertical lines indicate loci previously reported in GWAS –,. The horizontal magenta line denotes the range of genome-wide significance (p < 5 × 10⁻⁷). Black vertical arrows indicate loci explored in Stage 3 chosen on the basis of the p value. The magenta vertical arrow point to rs3817198 in the LSP1 gene. Blue dots denote the results of the genotype test and red dots denote the trend test.

Figure 2. Linkage Disequilibrium plot of Two Novel Loci

Both panels present the LD plots (using D’) for novel loci based on SNPs with MAF > 5% using HapMap Stage II individuals of European background (n=60 unrelated individuals). Above the plots are the results of the three individual Stages and the combined analysis for the SNPs reaching genome-wide significance. Panel A. Chromosome 1 region marked by rs11249433 and bounded by SNPs between chr1:120,400,700 −121,060,765. Note that one side is closely anchored to the centromere while the region distal to the centromere is bounded by a “SNP desert” of approximately 220 kb. Panel B. Chromosome 14q24.1 region marked by rs999737 and the block resides in the intron between two exons, of which the last has been observed in one of the three splice variants observed. Note that the SNP is located in an intron exclusive to the longest predicted transcript of RAD51L1.

Figure 2. Linkage Disequilibrium plot of Two Novel Loci

Both panels present the LD plots (using D’) for novel loci based on SNPs with MAF > 5% using HapMap Stage II individuals of European background (n=60 unrelated individuals). Above the plots are the results of the three individual Stages and the combined analysis for the SNPs reaching genome-wide significance. Panel A. Chromosome 1 region marked by rs11249433 and bounded by SNPs between chr1:120,400,700 −121,060,765. Note that one side is closely anchored to the centromere while the region distal to the centromere is bounded by a “SNP desert” of approximately 220 kb. Panel B. Chromosome 14q24.1 region marked by rs999737 and the block resides in the intron between two exons, of which the last has been observed in one of the three splice variants observed. Note that the SNP is located in an intron exclusive to the longest predicted transcript of RAD51L1.

Figure 3. Forest plot for Overall, and ER+ and ER− Analysis, for rs 1124933 and rs999737

The results of the Overall Pooled analysis, and case-control analyses for ER+ cases, and ER−ve cases, were generated using a trend test with one degree of freedom. The figure includes per allele odd ratio (log additive/multiplicative model) for each study. For the overall analysis, the P-heterogeneity values are for rs1124933 P=0.44, and for rs999737 P=0.79. Data were available for estrogen-receptor status in 6,586 cases.