This proposal pursues a basis for our striking observation that oral cancer cells transit between two states with dichotomous PI3K/Akt signals and responses to anti-PI3K drugs. Specifically, maximal Akt activation is found in only a small, slow-cycling fraction of tumor cells, whereas PI3K/Akt signals are suppressed in a distinct, G0- like fraction. These G0-like cells are non-dividing but retain high tumorigenic potential and show broad therapy resistance, including escape from PI3K inhibitors. Exiting the G0-like state is regulated by the H3K4me3 demethylase JARIDB, which represses transcription at select promoters but also has ill-defined roles, including ones independent of its catalytic function. JARID1B upregulation in G0-like cells reprograms them into the slow cycling, Akt-hyperactive phenotype, which is sensitive to PI3K inhibitors. The stem cell-like features and enhanced tumorigenicity of this JARID1Bhigh state implicate it in allowing quiescent cells to exert their malignant potential. When JARID1Bhigh cells are depleted by targeting PI3K, G0-like cells could still sustain tumor growth by an alternate, JARID1B-independent pathway for exiting quiescence. By this pathway, G0-like cells may become the critical locus for the compensatory MAPK pathway activation known to underlie PI3K inhibitor resistance. Thus our hypothesis is that G0-like cells are driven by JARID1B to a PI3K-dependent, stem cell-like state with aggressive features but also bypass this state under PI3K inhibition to maintain a central role in sustaining cancer growth. To determine JARID1B's contribution to G0-like cells exiting quiescence, we will delineate its demethylase dependent vs. independent functions and interaction with oncogenic PI3K signals in promoting G1-S transition (aim 1). To pursue JARID1B's contribution to the G0-like state's malignant potential, we will examine JARID1B's potential to promote aggressive tumor behavior by prolonging S/G2 while driving mesenchymal-like differentiation (aim 2). The roles of G0-like and JARID1Bhigh cells will also be defined under PI3K inhibition, where JARID1Bhigh cells may return to a G0-like state that sustains oral cancer progression via MAPK pathway compensation in a JARID1B-independent manner (aim 3). Understanding the mechanisms of transit between the dual states, which together comprise an aggressive, treatment refractory segment of the tumor cell pool, will define new strategies to address the epigenetic plasticity underlying treatment resistance in oral cancer.

Public Health Relevance

More effective oral cancer therapies can be achieved by understanding the non-genetic differences among tumor cells in a single cancer that allow some cells to evade treatment. Here we define the actions of certain non-dividing oral cancer cells that are therapy resistant and retain high malignant potential after exposure to modern drugs targeting key growth signals in this disease. Understanding the nature of such non-dividing cells and the mechanisms used by them to resume growth will create opportunities for new oral cancer therapies.

National Institute of Health (NIH)
National Institute of Dental & Craniofacial Research (NIDCR)
Research Project (R01)
Project #
Application #
Study Section
Tumor Cell Biology Study Section (TCB)
Program Officer
Wang, Chiayeng
Project Start
Project End
Budget Start
Budget End
Support Year
Fiscal Year
Total Cost
Indirect Cost
University of Pennsylvania
Schools of Medicine
United States
Zip Code