Date of Award

Summer 8-15-2017

Author's School

Graduate School of Arts and Sciences

Author's Department

Biology & Biomedical Sciences (Molecular Genetics & Genomics)

Degree Name

Doctor of Philosophy (PhD)

Degree Type



DNA methylation is known to silence gene expression in the context of imprinting, X-chromosome inactivation, and retrotransposon silencing. However, the role of DNA methylation in silencing gene expression outside of these contexts is not fully understood. This is especially true in diseases such as cancer, where normal DNA methylation patterns are significantly altered. In breast cancer as well as nearly all cancer types, most of the genome loses DNA methylation while small regions of the genome gain methylation. DNA methylation generally correlates with decreased gene expression when present at a gene promoter. Therefore, these regions of hypo- and hyper-methylation may contribute to cancer development and progression by activating oncogenes or silencing tumor suppressor genes. My work focuses on building tools to study the functional role of DNA methylation changes and exploring how methylation changes at a gene promoter promote resistance to treatment in breast cancer.

About 75% of breast cancers depend on estrogen signaling through the estrogen receptor (ERα). These tumors are effectively treated by aromatase inhibitors (AI) that prevent estrogen production. However, almost all advanced cases of ERα positive breast cancer develop resistance to AI therapy. I therefore sought to identify methylation changes that promote this resistance. I studied UCA1 and PTGER4, two genes identified by a screen for negatively correlated methylation and expression changes in a cell line model of AI resistance. UCA1 is a long non-coding RNA that promotes growth and metastasis in bladder cancer. PTGER4 encodes the prostaglandin E2 receptor 4 (EP4), which supports the progression of multiple cancer types by altering cell signaling. While my experiments did not indicate that UCA1 has a strong role in AI resistance, I found that hypomethylation of the PTGER4 promoter correlates with increased expression and EP4 signaling. My data further suggest that the downstream effector of EP4 signaling, CARM1, promotes endocrine therapy resistance by increasing the ligand-independent transcription activity of ERα.

The effects of local DNA methylation changes are most often identified by correlating the methylation and expression levels from two samples. To show that methylation causes the expression change, these studies rely on non-specific tools that demethylate the whole genome: DNA methyltransferase (DNMT) inhibitors or by DNMT knockout/knockdown. To build a tool capable of inducing site-specific DNA methylation changes, I fused the human DNMT3A catalytic domain to the RNA-guided nuclease Cas9 (the Cas9 is nuclease dead). I used this tool to induce up to 53% DNA methylation on individual cytosines within 50 bp of the target site. When multiple sites within the CDKN2A or ARF promoters were targeted, the induced DNA methylation decreased the expression of the targeted gene. To determine the optimal DNMT catalytic domain to use in this system, I created alternative DNMT fusions that included human DNMT1, a fusion of mouse Dnmt3a to mouse Dnmt3L, human DNMT3B, and the bacterial methyltransferase M.SssI. While the Dnmt3a-Dnmt3L fusion increased methylation relative to DNMT3A alone, it also induced more off-target methylation. The continued development of targeted DNA methylation technologies will increase our ability to identify functional methylation changes in tumors. As a result, we will learn the specific ways that methylation-induced gene expression changes contribute to cancer.


English (en)

Chair and Committee

John Edwards

Committee Members

Jason Weber, Ting Wang, Sheila Stewart, Grant Challen,


Permanent URL: