Soy, DNA Methylation and Breast Cancer

Institution: University of Southern California
Investigator(s): Anna Wu, Ph.D. -
Award Cycle: 2011 (Cycle 17) Grant #: 17IB-0017 Award: $244,524
Award Type: IDEA
Research Priorities
Etiology and Prevention>Prevention and Risk Reduction: ending the danger of breast cancer

Initial Award Abstract (2011)

Non-technical overview of the research topic and relevance to breast cancer:
There is now compelling epidemiologic evidence from case-control and prospective cohort studies that regular soy food consumers experience a lower risk of breast cancer development and recurrence. Before this finding can be translated into preventing breast cancer, we must have a better understanding of the underlying mechanisms by which soy may protect against breast cancer. There is growing experimental (e.g. non-human) evidence that soy and other dietary factors may influence breast cancer risk through its impact on gene expression, by influencing which genes are turned on and off (this is referred as “epigenetic mechanisms”). In this proposed study, we will investigate whether soy intake in breast cancer patients influences which genes are turned on and off in their breast tumor tissues (by measuring DNA methylation profiles of these genes). We have chosen to study 7 genes known to have an important role in breast cancer development and are commonly turned off in breast cancer patients; soy has been suggested to affect the activity of these genes.

The question(s) or central hypotheses of the research:
Does tumor tissue from breast cancer patients who were high soy consumers differ from tumor tissue from breast cancer patients who were low soy consumers in terms of DNA methylation patterns of selected genes important in breast cancer development? In other words, are there differences in which genes (or groups of genes) are turned on or off based on level of soy intake?

The general methodology:
The overall objective of this study is to explore the association between pre-diagnostic soy intake and DNA methylation profile in breast tumor tissues of 326 Asian American women with breast cancer. These women had participated as cases in a CBCRP funded population-based case-control study in Los Angeles County and for whom we have collected detailed information on lifetime soy food intake and other lifestyle factors. The DNA methylation study will be conducted using previously obtained breast tumor tissue from these patients. The laboratory methods to assess DNA methylation status in this proposed study have been developed by colleagues at the University of Southern California (USC) Epigenome Center.

Innovative elements of the project:
This project has been designed to bridge a knowledge gap in studies of soy and breast cancer. This study uses a new angle to investigate possible mechanisms by which soy may influence breast cancer risk using human data. To our knowledge, this is one of the first studies to test the soy-DNA methylation hypothesis in breast cancer patients with a wide range of soy food intake during adolescence and adult life so that exposure to soy as well as the timing of soy intake will be investigated. Epigenetic changes are potentially reversible, therefore if we learn that certain tumor suppressor genes are inappropriately silenced or oncogenes are inappropriately turned on by soy’s influence on DNA methylation, this would represent a novel chemo-preventive opportunity to reduce breast cancer development and/or recurrence.

Advocacy involvement and relevance to the human issues associated with breast cancer:
We will work closely with Diana Chingos, Chairperson of the Cancer Survivorship Council at the USC/Norris Comprehensive Cancer Center. Dr. Wu and Ms Chingos plan to have regular discussions of this study’s design and execution. Dr. Wu will also receive feedback from the members of the Cancer Survivorship Advisory Council at their monthly meeting. Ms Chingos is committed to disseminating the research findings to the breast cancer advocacy and lay community.

Final Report (2014)

The overall objective of this IDEA study is to obtain greater understanding of the underlying molecular and biochemical mechanisms that might explain why soy protects against breast cancer. We aimed to study whether soy intake during adolescence and/or adulthood can influence the methylation status of a few select genes involved in hormone- and receptor-mediated cell signaling (ESR1, ESR2, RARb2), DNA repair (BRCA1, PTEN, GSP1), and cell-cell adhesion (CDH1). The original study proposed to examine methylation status of these 7 genes using the well-established Methylight protocol on 326 patients for whom we already have formalin-fixed paraffin embedded (FFPE) archival sections. However, since our grant application submission, there had been substantial developments in the field of DNA methylation. Specifically, after conducting some pilot testing, we decided to use the Illumina 450 K Methylation Beadchip. This 450K Methylation platform covers over 480,000 CG dinucleotides involving ~14,500 genes, including the 7 genes proposed in our original grant application. Although the higher cost of the Beadchip compared to Methylight runs meant that we would be studying only 96 instead of 326 patients, we decided the large gain in information is a reasonable trade off and may reveal new information on soy and breast cancer.

We have completed preliminary analyses of the 450K Beadchip DNA methylation data on 96 breast cancer patients. After excluding samples from 4 duplicate subjects and 14 samples that did not pass final quality control, our analysis was based on 78 patients (i.e., 85% of patients). The HM450 array interrogates 482,421 CpG, 3,091 non-CpG sites, and 65 random SNPs. The final analysis was based on 410,622 probes (i.e., 85% of probes on Beadchip) that passed all quality control measures. We conducted various preliminary analyses to evaluate whether soy intake during adult life and adolescence influence DNA methylation at these 410,622 probes. We used both Beta (ß)- and M-value statistics as metrics of associations. None of the CpGs were associated with soy intake during adolescence when we conducted regression analysis with multiple comparison correction. However, 12,813 probes remained statistically significantly associated with adult soy intake after multiple adjustment using both beta and M-value statistics. Three genes (BRCA1, GSTP1, and RARb2) were statistically significantly associated with adult soy intake showing significantly lower DNA methylation level in association with highest quartile of soy intake. Using Interactive Pathway Analysis to further explore associations with these 12,813 probes, we have identified 5 pathways that remained significant after multiple test corrections. Interestingly, the top gene showing greater than 3-fold increases in DNA methylation levels in association with adult soy intake is CCND1. We are currently conducting additional analyses to verify this finding.