Table 1.
Characteristics of study participants and number of SNPs analyzed in each stage.
Figure 1.
Per-allele OR for rs4784227 in association with breast cancer risk by study and ethnic groups.
Table 2.
Results from Stages I to III for rs4784227 and Stages I and II for other three SNPs that showed promising associations with breast cancer risk in Stage II.
Figure 2.
SNP association at 16q12 with breast cancer risk.
In SBCS (A) and CGEMS (B). Top panel: Results (-log10P) are shown for the associations of breast cancer risk with directly genotyped (diamonds) and imputed (circles) SNPs located in the TOX3 and LOC643714 genes. P-values for rs4784227 are shown in red for Stage I data and in blue for the combined data. The three previously-reported SNPs (rs8051542, rs12443621, and rs3803662) are highlighted in black. Middle panel: Genomic view at 16q12.1. Gene locations are from the March 2006 UCSC genome browser assembly. Bottom panel: Estimates of pairwise LD (r2) for common SNPs (with MAF≥5%) from HapMap release 23a in the region from 10 kb downstream of rs8051542 to 10 kb upstream of rs4784227.
Table 3.
Linkage disequilibrium patterns among rs4784227 and the three previously-reported SNPs in 16q12.1.
Table 4.
Association of rs4784227 and the three previously-reported SNPs at 16q12with breast cancer risk among Chinese women in Stages I and II.
Table 5.
Association of rs4784227 and the three previously-reported SNPs at 16q12with breast cancer risk among European American women.
Figure 3.
In vitro functional characterization of SNP rs4784227.
(A) Effect of rs4784227 on luciferase reporter activity. HEK 293, MCF10A, MCF-7, and MDA-231 cells were transiently transfected with pGL3 promoter vector and the constructs carrying the reference allele (C) and risk allele (T) of rs4784227, respectively. Relative luciferase activities are shown as mean ± SD of four experiments (relative to empty vector). Statistical analysis was done using Student's t-test comparing C and T alleles (* P<0.05, ** P<0.01, n = 4). (B) Electrophoretic mobility shift assays. Nuclear protein extracts from MCF-7 (top panel), MCF10A (middle panel), and MDA-231 (bottom panel) cells were incubated with biotin-labeled probes corresponding to reference allele C (lanes 1–5) or the risk allele T (lanes 6–10) of rs4784227 in the absence or presence of competitors. Lanes 1 and 6, no nuclear extracts; lanes 2 and 7, unlabeled competitor in 200-fold molar excess; lanes 3 and 8 (5 mM MgCl2), lanes 4 and 9 (2.5 mM MgCl2), and lanes 5 and 10 (1.25 mM MgCl2), no competitor. I: nonspecific DNA-protein complex bands from MCF-7, MCF10A, and MDA-231 cells; II: specific DNA-protein complex bands; III: free biotin-labeled probes.