During development, Six1 plays a critical role in the expansion of progenitor cell populations via regulating proliferation, survival, migration, and invasion. While not expressed in most differentiated tissues, Six1 is re- expressed in 50% of primary and 90% of metastatic breast lesions. In animal models, Six1 overexpression causes both mammary tumorigenesis and metastasis. Because Six1 is a transcription factor and therefore not amenable to targeting for anti-cancer therapy, we must understand the mechanism by which it mediates tumor progression in an effort to identify means to target the pathways it activates. We have found that Six1 induces an epithelial to mesenchymal transition (EMT), a tumor initiating cell (TIC) phenotype, and metastasis via its ability to induce TGF? signaling. While inhibition of TGF? signaling in Six1-overexpressing breast cancer cells reverses metastasis, inhibition of TGF? signaling in control cells promotes metastasis. These data indicate that Six1 is able to shift TGF? signaling from tumor suppressive to promotional, and may provide a novel mechanism to explain the elusive "TGF? paradox". Although Six1 is expressed in all subtypes of breast cancer, high expression of Six1 significantly correlates with adverse outcomes particularly in the aggressive and hard to treat luminal B subtype. Similarly, when breast cancer datasets are queried with a signature that examines only the tumor suppressive arm of TGF? signaling, low expression of this signature is associated with poor prognosis specifically in ER+ (luminal) breast cancers, and the luminal B subtype expresses the lowest levels of the TGF? tumor suppressive signature. These data imply that tumors within this subgroup escape the growth suppressive effects of TGF?. Importantly, TGF? inhibitors are currently in clinical trials, yet because TGF? signaling can be tumor suppressive or promotional, one of the greatest concerns surrounding their use in cancer is how to predict which patients will benefit. In this proposal we focus on a novel mechanism by which Six1 may mediate the switch in TGF? signaling, particularly in luminal B tumors: activation of the miR106b-25 cluster of microRNAs (miRs). This cluster of miRs is known to overcome the growth suppressive effects of TGF? signaling, and we demonstrate that the same cluster can also activate the tumor promotional arm of TGF? signaling. Thus, we will test the hypothesis that induction of the miR106b-25 cluster by Six1 provides a mechanism by which Six1 executes the switch in TGF? signaling from tumor suppressive to promotional, thus leading to increased EMT and TIC capacity and increased metastasis, and that inhibition of miR106b-25 may thus be a means to therapeutically target luminal B breast cancers. We will also test whether Six1/miR-106b-25 expression may provide a means to distinguish between patients who will benefit rather than be harmed by treatment with TGF? inhibitors. To address these hypotheses, we will utilize cell culture, xenograft, and transgenic mouse models in which the Six1/miR106b-25/ TGF? axis is modulated by genetic means and/or drug treatment, as well as by experimentation on human breast cancer samples.