Karen Fancher

Date of Award


Level of Access Assigned by Author

Open-Access Dissertation

Degree Name

Doctor of Philosophy (PhD)


Interdisciplinary Program


Barbara B. Knowles

Second Committee Member

Gary A. Churchill

Third Committee Member

Keith W. Hutchison


Family history, reproductive factors, hormonal exposures, and subjective immunihistochemical evaluations of in situ lesions, and to a lesser extent age, remain the best clinical predictors of an individual's risk of developing breast cancer. Identification of early markers predictive of impending invasive breast cancer from in situ carcinoma is a long-term goal. The latent mammary cancer transgenic mouse model of human breast cancer, C57BL/6JTg(WapTag)1Knw (Waptag1), develops characteristic stages of tumorigenesis in a highly predictable manner: atypical hyperplasia advances to ductal carcinoma in situ (DCIS), which progresses to papillary adenocarcinomas and/or solid, invasive tumors. Microarray analyses of whole mammary glands and tumors across these stages, to detect transcriptional changes throughout tumorigenesis, revealed marked, phased stage-specific changes. In constrast, results from the laser capture microdissected tumor cells depict a moderately constant characteristic tumorigenic profile, irrespective of stage. Evaluation of differences in whole glands with those of microdissected samples suggests that paracrine signaling between tumor and stromal cells substantially alters the tumor microenvironment, early in progression. Strikingly, comparison of statistically significant microarray results between Waptag1 DCIS and human DCIS revealed 2,097 overlapping early transcriptional changes. When compared with species-specific controls, common abundant early gene alterations were associated with cell cycle, cell division, and DNA replication Gene Ontology categories, with a notable decrease in genes involved in aerobic energy metabolism, and significant increased transcription of retrotransposons and chromosome modification genes. Based on these initial experimental results, retrotransposons were identified as a potential marker for testing in several mouse models and in biopsies derived from breast cancer patients. Analysis of data from five independent mouse models of mammary cancer and five human breast cancer datasets revealed over expression of retrotransposons, mainly Class I and Class II LTR elements, as well as LINEs and SINEs, when compared with normal samples. Cross-species comparison of gene expression profiles suggests epigenetic alterations and chromatin remodeling changes coincide with retrotransposon over expression. Through validation of such mutual human and mouse changes in gene expression, these novel putative markers may allow earlier detection and therapeutic intervention, possibly reducing the incidence of invasion and metastases in patients with breast cancer.