Publication Date




Embargo Period


Degree Type


Degree Name

Doctor of Philosophy (PHD)


Biochemistry and Molecular Biology (Medicine)

Date of Defense


First Committee Member

Joyce M. Slingerland

Second Committee Member

Feng Gong

Third Committee Member

Ralf Landgraf

Fourth Committee Member

Barry Hudson

Fifth Committee Member

Nipun Merchant


Increasing evidence indicates that cancer stem cells (CSCs) comprise a distinct self-renewing sub-population that can generate cancerous progeny with reduced replicative potency. CSCs are important therapeutic targets: they may not only initiate tumors but also mediate recurrence and metastasis. Most anti-cancer drugs kill the bulk cancer population. CSCs either proliferate too slowly for targeting by cycle active drugs, or escape chemotherapy by drug expulsion or greater DNA repair, leading to recurrence. The work in this thesis was undertaken to identify novel mechanisms leading to CSCs expansion. My first project identified a novel mechanism whereby vascular endothelial growth factor A (VEGFA) promotes CSC expansion. VEGFA is widely known as an angiogenic factor. Here, we investigated the role of VEGFA as an ovarian cancer stem cell factor and mechanisms underlying its actions. We showed VEGFA mediated cancer stem cell actions via VEGFR2-dependent Src activation to upregulate Bmi1 and ALDH1 activity in primary human OVCA culture and OVCA lines. VEGFA stimulated sphere formation only in the ALDH1+ subpopulation, and increased OVCA-initiating cells and tumor formation in vivo through Bmi1. DNA methyl transferase 3A (DNMT3A) played a pro-oncogenic role downstream of Src to methylate miR-128-2, upregulating Bmi1 to increase stem like cells. DNMT3A knockdown prevented VEGFA-driven miR-128-2 loss, and the increase in Bmi1 and tumor spheres. Analysis of over 1,300 primary human OVCAs revealed an aggressive subset in which high VEGFA is associated with miR-128-2 loss. The second CSC pathway identified is C-terminally phosphorylated p27 dependent. p27 is known as a cell cycle inhibitor and a tumor suppressor. However, mis-regulated p27 has oncogenic effects. Previously, our group showed that double phosphomimetic p27 (p27CK-T157D/T198D) induced epithelial-mesenchymal transition (EMT) and metastatic potential of cancer cell lines. We also observed that C-terminally phosphorylated p27 increased CSC properties, including sphere formation, CSC markers and colony formation in soft agar in vitro. Transduction of a mutant p27 that cannot bind cyclin-CDKs and that bears phophomimetic mutations converting T157 and T198 to D (p27CK-DD) increased the expression of several embryonic stem cell transcription factors. These are known to drive embryonic stem cell self-renewal and to promote CSC expansion. A human phospho-kinase array showed that Pyk2 is activated by p27CK-DD. We found that Pyk2 activation and its binding to p27 are dependent on phosphorylation of p27 at T198 and T157. Treatment with a Pyk2 inhibitor, and PYK2-knockdown by siRNA, CRISPR or ShRNA revealed that Pyk2 is a key mediator of the increase in tumor spheres, ALDH1 activity and ES-TFs expression and of the increased abundance of tumor initiating stem cells in vivo in cancer cells expressing abundant C-terminally phosphorylated p27. These data reveal a novel mechanism whereby p27-driven Pyk2 activation promotes CSC expansion and tumor progression.


p27; Cancer stem cell; Breast cancer; Pyk2