Modeling clonal evolution in mouse xenograft models

Cancers develop as a result of somatic evolution. Deciphering the evolutionary dynamics behind this should provide a more accurate understanding of how cancers arise and enable more intelligent approaches toward anti-cancer therapies. However, this area receives almost no experimental attention, and our understanding of clonal evolution in cancers is very rudimentary. To address this deficiency, we have developed a mouse xenograft model of human breast cancers that allows us to follow dynamics of clonal competition in genetically heterogeneous tumors.

Intratumor heterogeneity and metastasis

Metastatic dissemination of cancer cells is the most prominent cause of death due to breast cancer. Recent work in this field has established that the progression of metastatic invasion from the primary tumor to distant locations (such as bone, lungs, and brain) depends on heterogeneous interactions of cancer cells with each other and with cells composing the microenvironment. We aim to elucidate some of the factors and mechanisms that influence metastatic co-operation between cancer cells and their environment in order to fully understand the metastatic cascade and aid in the development of therapies that address this phenomenon.

Diversity in human breast tumors

Intra-tumor genetic and phenotypic diversity may predict the risk of breast cancer progression and response to treatment. To deepen our understanding of these factors, we have been defining intra-tumor diversity using immuno-FISH and ecological models in breast tumors at different progression stages (i.e., in situ, invasive, metastatic), and before and after chemotherapy or targeted (e.g., antu-Her2) treatment.

46 views1 comment

Interrogating consequences of interactions between breast carcinoma cells and tumor fibroblasts

While it is becoming increasingly apparent that interactions between carcinoma cells and tumor stroma are an essential part of tumor biology, our understanding of this crosstalk is far from complete. Using organotypic 3D culture models, we are interrogating mutual changes in transcriptome, metabolome, and phospho-proteome that result from the interaction between breast carcinoma cells and primary breast tumor-associated fibroblasts.

Myoepithelial cells and leukocytes in DCIS

The progression from in situ to invasive carcinoma is a key but poorly understood step of breast tumorigenesis, characterized by loss of the myepithelial cell layer and basement membrane. We hypothesize that the differentiation of bipotential mammary epithelial progenitors to myoepithelial cells is progressively inhibited by signals coming from tumor epithelial cells and stromal cells, such as leukocytes, leading to their eventual disappearance. Project objectives include:

  • Defining normal myoepithelial cell differentiation and its abnormalities in DCIS

  • Characterizing the role of immune cells in myoepithelial cell differentiation during breast carcinoma progression using in vivo and in vitro model systems and human breast tissue

The completion of this project will increase our understanding of the role of myoepithelial and immune cells in breast cancer, and may also provide new targets for breast cancer treatment via abnormally expressed paracrine signaling in the tumor microenvironment.

35 views0 comments

Pregnancy study

Human epidemiological and experimental data in rodent models suggest that full-term pregnancy in early adulthood decreases the risk of estrogen receptor positive (ER+) breast cancer in post-menopausal women; however, the underlying mechanism is largely unknown. We hypothesized that the cancer-preventive effects of parity may be due to alterations in the number or properties of mammary epithelial progenitor/stem cells that are thought to be the cell-of-origin of breast cancer, rendering them less susceptible to oncogenesis. To test this hypothesis, we analyzed the relative frequency and comprehensive molecular profiles of four distinct cell types (CD24+ luminal, CD10+ myoepithelial, lin-/CD24-/CD44+ progenitor-enriched, and stromal fibroblasts) isolated from normal breast tissue of premenopausal nulliparous and parous women. Based on the comprehensive analysis of gene expression, DNA, and histone H3 K27 trimethylation profiles of these cell types, we determined that the most significant changes occurred in lin-/CD24-/CD44+ progenitor-enriched cells. The activity of many genes and pathways involved in development, differentiation, and cell cycle regulation are decreased in parous women that may contribute to their decreased breast cancer risk. We also identified a parity-associated gene signature that predicted clinical outcome in breast cancer patients diagnosed with ER+ tumors.

The role of DNA methylation in mouse mammary gland development

The mouse mammary gland is a useful model system for understanding factors that regulate mammary development. We are pursuing molecular characterization of the different cell types that comprise the mammary epithelium of the mouse. Based on the varying proportional distributions we observe in the mature, progenitor, and stem cell populations of the mammary gland during different life stages, we seek to understand the underlying molecular cues that maintain cell type identities and direct cellular distribution changes by studying the gene expression and epigenetic properties of distinct cell populations during puberty and pregnancy, stages during which there is dramatic tissue remodeling in the mammary gland. Furthermore, with the use of in vitro and in vivo mouse models for the functional characterization of maintenance DNA methylation, we are characterizing potential active roles of this important epigenetic mark in directing cell fate in the mammary gland.

Histone modifying enzymes as new therapeutic targets

The differentiation of normal stem cells and the development of normal tissue are controlled by epigenetic mechanisms. Abnormalities in these processes play a role in the initiation and progression of tumors and intra-tumor diversification of cancer cells. A number of histone-modifying genes were found to be mutated in breast and other cancers, implying that these genes may represent novel therapeutic targets and biomarkers. We have recently reported the characterization of cell-type specific patterning of histone and DNA methylation in normal breast tissues. We developed modified chromatin immunoprecipitation combined with high-throughput sequencing (ChIP-Seq) protocol which enables us to investigate the epigenetic status genome-wide, using limited numbers of cells purified from human breast tissue samples. Currently, we are using various genomic profiling and functional studies to validate several histone demethylases as potential therapeutic targets in breast cancer.

Determinants of basal-like and luminal breast cancer cell phenotypes

Basal-like and luminal breast tumors have distinct molecular profiles and clinical behavior, yet the mechanisms underlying these differences are poorly defined. We investigated the potential role of genetic factors in determining these distinct phenotypes and their inheritance pattern by generating somatic cell fusions between basal-like and luminal breast cancer cells and analyzing their molecular profiles and functional characteristics. Based on the molecular profiles, we identified candidate key transcriptional and epigenetic determinants of basal-like and luminal cell phenotypes. We are further characterizing these genes using functional genomics approaches.

18 views0 comments

Dana-Farber Cancer Institute

450 Brookline Ave. SM-1070

Boston, MA 02215


Tel: (617) 632-(2106)​


The Polyak Lab is dedicated to the molecular analysis of human breast cancer