Dissect governing factors for tumor stem cell dormancy in biomimetic vascular/GBM model

SUMMARY Glioblastoma (GBM), the most common and malignant type of brain tumor, has a dismal prognosis due to recurrence despite treatment. Tumor cell dormancy (quiescence) is a major root for tumor relapse, as conventional therapies target mostly proliferating cells. Moreover, quiescent cells harbor a privileged subpopulation of stem-like cells in special niches, which can be reawakened to spawn tumor re- propagation. Dissecting determining factors for tumor dormancy remains challenging due to difficulty to track this population. We have developed a doxycycline-inducible Histone2B-GFP quiescence reporter. In pulse- chase paradigms, quiescent tumor cells retain H2B-GFP label, while proliferative cells sequentially dilute GFP. Our pilot studies in intracranial GBM transplants showed that quiescent cells preferentially reside in close proximity to vasculature. This suggests that perivascular niche may promote glioma stem cell (GSC) dormancy. To test this hypothesis, we take a multidisciplinary approach to develop a bioengineered vascular GBM organoid model. Patient-derived glioma stem cells (GSC) are used to generate GBM organoids, which are then embedded in a multi-scale vascular network with perfusion. Our pilot data demonstrated infiltrative growth of GSC along vasculature, recapitulating a key clinical feature of GBM. In Aim 1, we will analyze the H2B-GFPhigh vs. H2B-GFPlow populations in our organoid model to define the link between perivascular niches and GSC dormancy. Based on our pilot transcriptome profiling that showed upregulation of matrix modifier genes in dormant cells, we will test top candidates to determine their function in promoting GSC dormancy. Parallel in vivo transplant experiments will be carried out in SCID mice using identical GSC lines to verify the biomimetic nature of our model. In Aim 2, we will take advantage of the unique features of our vascular model that allow regulation of perfusion speed and oxygen tension to test the hypothesis that a metabolic stressor such as hypoxia promotes GSC dormancy. To reveal metabolic heterogeneity of GBM cells, we will deploy dual reporters to simultaneously detect hypoxia (RFP) and quiescence (H2B-GFP) to address whether perivascular niches confer a hypoxic microenvironment to dormant GSC. We will also test in our model the efficacy of a novel pro-drug (TH-302), which is activated by hypoxia and crosslinks DNA, to target hypoxic dormant GSC. Our preliminary study showed compelling evidence of the link between GSC quiescence and radiation (XRT) resistance. In Aim 3, we will test if perivascular niche and hypoxia further enhance XRT- resistance of quiescent GSC. Mechanistically, we will determine if DNA repair pathways are activated by hypoxia in dormant GSC. Lastly, we will test potential synergy of the hypoxia-activated drug together with XRT in killing dormant GBM cells in hypoxic state. In sum, our proposal introduces a novel 3D vascular GBM model with perfusion to step-wise interrogate governing factors that enhance GSC dormancy. Our biomimetic model provides a powerful platform for testing gene function and novel drugs to target GBM dormancy.