UC Berkeley

The mammary gland can be thought of as an "organism" with very complicated biology. The normal breast is comprised of a milieu of different cell types and extracellular matrix (ECM) components, the sum of which yield a complex microenvironment and series of biological interactions that are essential for normal development and functional differentiation. Many of these events go awry in breast cancer, a disease best characterized by the loss of tissue structure, morphology, and function - in essence, a loss of microenvironment homeostasis.

Perhaps reflecting the complexity of the normal mammary gland, breast cancer cannot be simply described as a single disease, and is instead thought to be comprised of at least 5 molecular subtypes which have distinct clinical features. Evidence for this is noted in the histological, molecular, intra- and intertumor heterogeneity found in the clinic, the differences in patient response to treatment, and correspondingly, the differences in clinical outcome observed between subgroups of patients. Improving our understanding of the molecular basis of breast tumor heterogeneity can potentially lead to improved diagnosis of breast cancer at earlier stages, identify which patients would benefit most from current treatments or combinations of treatments, and unravel novel biomarkers that, if therapeutically targeted, would improve the clinical management of breast cancer. This would represent a significant advance over current therapeutic strategies that are known to be ineffective against certain subtypes of breast cancer.

Basal-like breast cancer, which accounts for 15-20% of invasive breast carcinomas, represents one subtype of breast cancer that is unresponsive to conventional therapies such as tamoxifen and Herceptin, and which has a particularly aggressive clinical course characterized by poor 5-year survival. This breast cancer subtype presents the most challenging case for current clinical management, and major effort has been placed towards understanding the molecular basis of basal-like breast cancer progression, origin, and diagnosis. Towards that goal, several animal and culture models of basal breast cancer progression and tumor heterogeneity have been developed, and are currently employed in numerous basic science, translational, and preclinical studies.

To investigate the molecular basis of breast cancer progression, previous work in the Bissell lab has characterized an isogenic culture model of tumor progression, called HMT-3522, which recapitulates many features of basal-like breast tumor progression. Proteomic screening of the cell lines by two-dimensional gel electrophosesis and mass spectrometry identified increased 14-3-3sigma protein level during malignant progression in this model. 14-3-3sigma belongs to a family of molecular scaffolds known to integrate signaling pathways by binding to phosphorylated serine/threonine motifs found in a spectrum of proteins and regulating their subcellular distributions. There are numerous reports implicating 14-3-3sigma as a tumor suppressor in the breast and other organs, yet this has not been sufficiently demonstrated in vivo. In addition, there are many conflicting reports suggesting that 14-3-3sigma is not a tumor suppressor, and instead may be related to poor clinical outcome in a subset of breast and other cancers. It is clear that the complicated role for 14-3-3sigma in cancer progression or suppression is context-dependent, and the overarching goal of my research has been to characterize the function of 14-3-3sigma in breast cancer progression and clinical outcome, particularly within the basal-like breast cancer subtype.

Here, I describe a novel function for 14-3-3sigma in promoting basal-like breast cancer cell motility and tumor invasion by regulating cytoskeletal dynamics. Using the HMT-3522 series and an independent isogenic model of basal-like breast cancer progression (the MCF10 series), I discovered 14-3-3sigma expression increased as cells transition towards malignancy, and upon shRNA-mediated knockdown, cancer cells with perturbed 14-3-3sigma expression show decreased motility and invasion through ECM in culture and diminished tumor invasion in vivo. Immunohistochemical analysis of a tissue microarray showed that 14-3-3sigma expression preferentially identified basal-like tumors among 245 invasive breast carcinomas, and in independent breast cancer patient cohorts, 14-3-3sigma correlated with metastasis and poor clinical outcome. These data demonstrate that 14-3-3sigma expression identifies a subset of aggressive breast carcinomas, namely basal-like breast tumors, and furthermore, that 14-3-3sigma functions in these tumors by regulating invasion and migration.

I additionally unravelled a mechanism, supported by several lines of evidence, by which 14-3-3sigma regulates basal-like breast cancer migration and invasion. Using a combination of live cell imaging, confocal microscopy, biochemical assays, and transgenic cell lines, I discovered that 14-3-3sigma regulates cytoskeletal homeostasis in cells by interacting with G-actin directly and sequestering it from growing F-actin filaments, thus inhibiting its polymerization. Through this interaction, 14-3-3sigma retains a soluble, "bioavailable" form of actin within cells to allow for regulated incorporation into growing filaments, ensuring polarized cell migration can occur. In contrast, basal-like breast cancer cells deficient for 14-3-3sigma have elevated rates of actin polymerization, have decreased pools of soluble actin available to be incorporated in a regulated fashion, and as a consequence, have decreased motility and invasion. These phenotypes can be rescued by overexpressing actin, but not a mutant form that is unable to be incorporated into existing actin filaments. This mechanism could explain how 14-3-3sigma, by defining the balance of actin polymerization within cells, regulates tumor invasion in basal-like breast cancer.

My data showing that 14-3-3sigma is a "functional" marker of basal-like breast tumors and poor clinical outcome does not dispute, but rather reconciles, the literature of whether 14-3-3sigma is a tumor suppressor in the breast; I propose that the expression and function of 14-3-3sigma in breast cancer is contingent on whether tumors show basal-like differentiation, or are of another molecular subtype. My data implicate 14-3-3sigma as a potential therapeutic target against the progression of basal-like breast cancer to metastastic disease, which may be of clinical importance following further validation studies. Furthermore, as the regulation of actin cytoskeletal homeostasis by 14-3-3sigma was characterized largely in vitro, the functional interaction of 14-3-3sigma and actin may be of importance not only in basal-like breast cancer, but additionally during normal development or during malignant progression in tumors originating from other organs.