Off-campus UMass Amherst users: To download campus access dissertations, please use the following link to log into our proxy server with your UMass Amherst user name and password.

Non-UMass Amherst users: Please talk to your librarian about requesting this dissertation through interlibrary loan.

Dissertations that have an embargo placed on them will not be available to anyone until the embargo expires.

Author ORCID Identifier



Open Access Dissertation

Document Type


Degree Name

Doctor of Philosophy (PhD)

Degree Program

Molecular and Cellular Biology

Year Degree Awarded


Month Degree Awarded


First Advisor

Kathleen F. Arcaro

Subject Categories

Cancer Biology


Roughly two-thirds of all breast cancers are Estrogen Receptor a (ER)-positive and can be treated with an anti-estrogen such as Tamoxifen, however resistance occurs in 33% of women who take the drug for more than 5 years. In addition to this acquired antiestrogen resistance, de novo- or intrinsic-resistance occurs primarily in ER-negative tumors but also occasionally in ER-positive tumors. Aberrant DNA promoter methylation, a major epigenetic mechanism by which gene expression is altered in cancer, is thought to play a role in this resistance. To date, few studies have examined promoter methylation and Tamoxifen resistance in breast cancer. Of the studies conducted, one detected drug-specific promoter methylation and gene expression profiles in an ER-positive, Tamoxifen-selected MCF-7 derivative cell line. However, studies using both ER-positive and –negative, Tamoxifen-selected cell lines have not been described until now. To develop an understanding of Tamoxifen-resistance and identify novel pathways and targets of aberrant methylation, I first analyzed two Tamoxifen-resistant clones of MCF-7, one that retained expression of ER (TMX2-11) and one that lost expression of the gene (TMX2-28) after 6-months of Tamoxifen treatment, by Illumina HumanMethylation450 BeadChip (HM450BC). I found that prolonged treatment with Tamoxifen induced hypermethylation and hypomethylation throughout the genome. Compared to MCF-7, the ER-positive line, TMX2-11 had 4,000 hypermethylated sites, while the ER-negative line, TMX2-28 had over 33,000. Analysis of CpG sites in both TMX2-11 and TMX2-28 revealed that the two Tamoxifen-selected lines share 3,000 hypermethylated CpG sites with 21% of those sites being located in the promoter region. Promoter methylation and expression of two genes, MAGED1 and ZNF350, in both Tamoxifen-resistant cell lines demonstrated cell line-specific responses to treatment with 5-aza-2’deoxycitidine (5-Aza). Sixteen additional genes involved in signal transduction, cell adhesion, transcriptional repression, inflammatory response, cell proliferation and hormone response were chosen for further analysis based on their shared hypermethylation or their reduced expression in TMX2-28 as detected in a previously completed expression array. Five genes, RORA, THBS1, CAV2, TGFβ2, and BMP2 had decreased expression in TMX2-28, but not TMX2-11 as compared to MCF-7, and 5-Aza increased expression of the genes. This indicates that Tamoxifen is affecting a set of genes similarly in both the ER-positive and -negative breast cancer cell lines, however overall methylation changes are more pronounced in the ER-negative line. Our data as well as others suggest that DNA methylation may be contributing to Tamoxifen-resistance. I hypothesized that both ER-positive and ER-negative second human breast tumors occurring after anti-estrogen treatment would be hypermethylated. I characterized the methylation profiles of 70 human breast tumor samples using the HM450BC. These data confirm previous findings that ER-positive breast tumors have more hypermethylated CpG sites than ER-negative tumors. Stratification of the tumors by ER-positive first and second tumor sets shows that methylation is greater in first tumors.. Additionally, I saw that first tumors from ipsilateral pairs had higher methylation than the second tumors; in contrast, second tumors from contralateral pairs had higher methylation than in the first tumor. These data, together with the fact that tumor progression is associated with an increase in methylation, are consistent with the prediction that ipsilateral, not contralateral, tumors are more likely to be a true recurrence. Pathway analysis was conducted to provide insight into biomarkers associated with tumors that recur. Two pathways, ‘homophilic cell adhesion via plasma membrane adhesion molecules’ and ‘cell fate commitment’, were selected for further analysis. ER-positive first tumors that recurred as either ER-positive or ER-negative compared with non-recurrent tumors shared hypermethylated genes in the homophilic cell adhesion pathway. ER-positive first tumors that recurred as ER-negative compared with ER-positive first tumors that recurred as ER-positive were associated with a unique set of hypermethylated genes in the cell fate commitment pathway. To examine the association of methylation changes in my tumor data set with breast cancer patient survival data, Kaplan-Meier plots were created using TGCA breast cancer data available online. Expression of the genes only hypermethylated in each individual comparison group in the homophilic cell adhesion pathway was linked to overall survival. These data suggest that the genes hypermethylated only in ER-positive tumors recurring as ER-negative are a potential signature for poor survival. The underlying mechanisms of anti-estrogen resistance are poorly understood. Variable responses to breast cancer therapy highlights the need for biomarkers that can effectively guide treatment. The findings presented here underscore the potential use of breast tumor stratification based on methylation biomarkers in guiding treatment.