Abstract
B51
There is increasing evidence that changes in methylation are important in breast carcinogenesis. Little is know about the etiology for changes in methylation. One-carbon metabolism involved in the transfer of methyl groups for DNA methylation and synthesis plays a role in breast carcinogenetic process, possibly in part by contributing to changes in DNA methylation. Folic acid is a critical substrate for this pathway, together with vitamins B6 and B12 as cofactors.
Ninety four healthy woemn with no history of cancer who were undergoing reduction mammoplasty were recruited to our study. All participants completed a structured interview and donated breast tissue and blood. p16INK hypermethylation and global methylation phenotypes were analyzed using pyrosequencing assays. LINE-1 elements were used to serve as a marker for global methylation. Plasma folate concentrations were quantified using the Immulite 1000 system. Breast folate levels were measured using the standard microbiological assay. We used Spearman correlation coefficients, t-tests and logistic regression models to describe the associations between folate levels and methylation phenotypes and other characteristics of these women.
p16 INKpromoter hypermethylation was identified in 31% of the breast tissues analyzed. Global methylation level range was 65.3-82.8% (mean=73.8%, median=73.3% and SD=4.6%). Plasma folate levels were significantly associated with age (p-value= 0.01), race (p-value <0.0001), prior pregnancy (p-value= 0.02), current smoking (p-value= 0.02), history of alcohol consumption (p-value= 0.03), history of cancer in the family (p-value= 0.0002) and also with global methylation phenotype (p-value= 0.04). Breast folate levels were significantly associated with prior pregnancy (p-value= 0.0001), history of alcohol consumption (p-value= 0.02), history of cancer in the family (p-value <0.0001) and weakly associated with p16 INKpromoter hypermethylation (p-value= 0.06). When the difference between breast and plasma folate concentrations was analyzed in relationship to the demographic characteristics we observed several significant associations with race (p-value <0.0001), pregnancy (p-value <0.0001), current smoking (p-value= 0.01), history of alcohol consumption (p-value= 0.02), history of cancer in the family (p-value <0.0001) and family history of breast cancer (p-value= 0.03).
Both systemic and tissue-specific folate levels appear to influence methylation phenotypes, especially in relationship with other risk factors such as age, race, smoking, drinking and history of cancer in the family. These findings may provide further insights in the understanding of gene-nutrient interactions in the determination of breast cancer susceptibility.
Supported by: DOD BC022346
Citation Information: Cancer Prev Res 2008;1(7 Suppl):B51.
Seventh AACR International Conference on Frontiers in Cancer Prevention Research-- Nov 16-19, 2008; Washington, DC