Vitamin K: Body Pools and Function in Breast Cancer

This proposal focuses on the divergent effects of the two major dietary forms of vitamin K on breast cancer. K vitamins act as cofactors for gamma-glutamyl carboxylase (GGCX), which post-translationally introduces γcarboxyglutamate residues into proteins. Although most of the 17 known γ-carboxylated proteins function in coagulation and bone homeostasis, the presence of GGCX in most tissues (including mammary gland) suggests more extensive physiological roles for vitamin K. We have demonstrated that triple negative breast cancer (TNBC) cell lines express GGCX and produce γ-carboxylated proteins in response to vitamin K1 (phylloquinone), the major dietary form. In TNBC cells, K1 treatment enriches for the stem cell marker aldehyde dehydrogenase 1 (ALDH1) and promotes mammosphere formation. These data suggest that K1 sustains GGXC mediated γcarboxylation to drive aggressive breast cancer phenotypes. Through analysis of genomic cancer datasets, we find that ~25% of breast tumors express GGCX and the vitamin K oxidoreductase (VKOR) genes required for its activity. Patients with such tumors have poorer survival than those whose tumors do not express these genes at high levels. Patients with this subtype of tumor would be candidates for therapies that limit K1 availability and/or inhibit GGCX. Surprisingly, we found that vitamin K2 (menaquinone-4), another naturally occurring form present in diet, does not stimulate γ-carboxylation or stem cell phenotypes in TNBC cells, but instead strongly suppresses cell growth, migration and energy metabolism. These provocative data indicate that K1 and K2 exert distinct effects on breast cancer cells, with K1 promoting and K2 suppressing aggressive phenotypes. We also found that expression of the vitamin K2 biosynthesis enzyme UbiA Prenyltransferase Domain Containing 1 (UBIAD1) is undetectable in TNBC, suggesting altered cellular handling of vitamin K. In Aim 1 we will dissect the effects of K1 and K2 in vitro, evaluate the role of UBIAD1 and conduct feeding studies to measure accumulation of K1 and K2 in TNBC xenografts and host mammary gland in relation to tumor growth. In Aim 2 we will determine whether deletion of GGCX from TNBC cells impacts γ-carboxylated protein synthesis and aggressive phenotypes in vitro and in vivo. Aim 3 will identify relevant γ-carboxylated GGCX substrate proteins that mediate the effects of K1. We anticipate that growth of tumors with high GGCX activity and low UBIAD1 will be stimulated by high dietary K1 and inhibited by high dietary K2. These findings would identify GGCX as an oncogene and the vitamin K pathway as a therapeutic target in a subset of patients with advanced breast cancer.