p53 haplotype determination in breast cancer.

Inheritance of certain germ line haplotypes consisting of three biallelic polymorphisms of p53 has been proposed as a risk factor for breast cancer and colorectal cancer [A. Själander et al., Carcinogenesis (Lond.), 17: 1313-1316, 1996, and Carcinogenesis (Lond.), 16: 1461-1464, 1995]. In their studies, pairwise haplotypes of these three polymorphisms were estimated. Extended haplotypes were further projected from the pairwise combinations. To overcome the necessity to estimate pairwise and extended haplotype frequencies, a PCR method has been developed to determine the absolute extended p53 haplotypes in diploid genomes. The method requires allele-specific PCR, confirmed by restriction analysis, and successive amplicon analysis. It has been applied to a nested case-control study of breast cancer (284 subjects; 99 cases and 185 controls; 182 Caucasians, 56 Hispanics, and 46 African-Americans). Evidence is presented that minor variants of the intron 3, codon 72, and intron 6 polymorphisms were moderately elevated in Caucasian breast cancer cases (intron 3, P = 0.03 for genotype and P = 0.01 for allelic frequency; codon 72, P = 0.07 for genotype and P = 0.054 for allelic frequency; and intron 6, P = 0.02 for genotype and P = 0.02 for allele frequency). Accordingly, analysis of haplotype distributions suggested an association of minor p53 haplotypes with breast cancer risk in Caucasians (P = 0.07). The relative allelic frequencies in breast cancer cases compared with controls also differed by age and menopausal status; the 1-2-1 haplotype was overrepresented in postmenopausal cases (P = 0.02) and cases older than 50 years (P = 0.02), whereas the other minor haplotypes (1-1-2 and rare variants) were overrepresented in premenopausal cases (P = 0.003) and cases 50 years of age and younger (P = 0.02). Genotype distributions at each locus and for all control groups were consistent with Hardy-Weinberg equilibria. Differences in haplotype distribution were associated with ethnicity (Caucasians versus African-Americans and Caucasians versus Hispanics, P < 0.001). The new haplotyping method may be useful in the study of gene-environment interactions.