William Weiss, MD, PhD, Investigator and Co-Leader of Project 2
Anders Persson, PhD, Co-Leader of Project 2
Figure 1. Association of NSCs and transit-amplifying progenitors (TAPs) with the vascular niche in normal brain and in malignant glioma. (A) Quiescent NSCs (white) in the SVZ are in close association with the ependymal layer. In contrast, proliferating NSCs and TAPs reside in the vascular niche and express EGFR, 1-integrin, and CXCR4. (B) We hypothesize that a similar situation exists during tumor progression in malignant glioma, where activated NSC-like tumor cells (blue) and TAPs (green), but not quiescent NSC-like cells (white), are associated with the vasculature (red) and recruited immune cells (grey) that release cytokines (orange). Radiotherapy disrupts the vasculature and target proliferative tumor cells, leading to massive infiltration of immune cells and cell cycling of previously dormant tumor cells. (C) We hypothesize that mechanoreceptor activate Notch1/2 and down-stream targets N-CoR1/2 and Hes1/5 in quiescent GBM stem cells. Radiotherapy lead to disruption of vascular niche, infiltration of immune cells, and altered mechanical forces that promote activation of Notch1/2 signaling in GBM stem cells and promote relapse.
Introduction to Project 2
Glioma, the most common primary brain tumor, typically relapses after treatment. Nonetheless, this recurrence is poorly understood. We have demonstrated a glial progenitor origin for human oligodendroglioma, while others have demonstrated either progenitor or neural stem cells (NSC) origin for human glioblastoma multiforme (GBM) brain tumors. Subpopulations of GBM cells expressing the NSC marker podoplanin are radio-resistant and cluster within the mesenchymal (NSC-derived) group, whereas GBMs that express progenitor markers cluster with oligodendroglioma tumors in the proneural group. We have characterized two distinct transgenic mouse models for glioma that mirror NSC- or progenitor-derived gliomas. Tumor-free sections of human brains show scattered immune cells expressing the macrophage/microglial marker Iba1 in a resting state, with both human and murine data suggesting NSC-derived tumors to have a more dense immune infiltrated, compared with progenitor derived brain tumors.
Project 2 focuses on how radiation-induced changes in the tumor microenvironment differentially drives recurrence in glioma as a function of transcriptomal subclass, specifically comparing gliomas arising from different neural precursor cells.
We hypothesize that
1). Radiation therapy, the standard-of-care for all brain tumors, disrupts the vascular niche, leading to infiltration of activated microglia and altering tissue compliance, both of which contribute to recurrence and radiation resistance.
2). Recurrence in NSC-derived glioma is modulated by the interplay among extracellular matrix stiffness, interstitial compression and infiltrating activated microglia, which direct NSC-like glioma to re-establish the vascular stem cell niche.
3). Altered tissue stiffness and interstitial compression preferentially drives recurrence in NSC-derived gliomas (in contrast to progenitor derived glioma), through activation of b1-integrin signaling.
In Aim 1, we will study the proliferation and differentiation of glioma cells (as a function of NSC- or progenitor-origin) in response to a disrupted tumor microenvironment. We will administer intracranial g-irradiation to mice with tumors in both transgenic murine glioma models and in orthotopic xenografts of human gliomas. In collaboration with Dr. Tracy McKnight, we will use MRI to study g-irradiation-induced changes in diffusion (reflecting vascularization), as an indirect but translatable measure of mechanical force, to determine how radiation influences the viscoelasticity of tumor tissue. Successful completion of this aim clarifies how g-irradiation-induced vascular collapse, remodeling of extracellular matrix, and infiltration of immune cells, differentially promotes relapse in progenitor- versus radiation resistant NSC-like glioma subtypes.
Our goal in Aim 2 is to test the idea that mechanical force and microglial cells collaboratively drive relapse of mesenchymal rather than progenitor-derived gliomas. Murine and human glioma of the proneural and mesenchymal subtypes will be cultured on collagen, laminin, and fibronectin-laminated polyacrylamide gels from low to high rigidit8. Cells cultured on these substrates will be grown in the absence and presence of conditioned media from activated microglia as well as co-cultures of microglia:tumor cells of glioma subtypes. Our goal is to analyze the combined effects of substrate stiffness and microglial-derived cytokines on glioma subtypes. We will establish co-cultures of microglia cells with glioma cells to determine how the cytokines released from these cultures will affect tumor cells. We will assay survival, proliferation, differentiation and migration in response to variations in substrate stiffness. We will analyze proliferation, apoptosis, differentiation, and self-renewal.
In Aim 3, we will investigate how b1-integrin signaling regulates cell cycling of TAP-like fast-dividing and NSC-like glioma cells during tumor progression and after g-irradiation. We will culture proneural and mesenchymal gliomas using 2D and 3D matrices and study the cellular responses after inhibition of the b1-integrin signaling, or varying substrate stiffness and interstitial pressure. Within the mesenchymal glioma subtype, we will separately analyze NSC- and their TAP-like derivative tumor cells. In subsequent experiments, we aim to identify intermediates that link mechanosensor b1-integrin signaling, Notch signaling, and downstream effectors, as a pathway that mediates effects leading to tumor recurrence. Finally, we will transplant proneural and mesenchymal gliomas, expressing tamoxifen-inducible b1-integrin shRNA constructs, into recipient mice and study the role of b1-integrin on tumor progression and recurrence after g-irradiation.
Figure 2. Podoplanin (PDPN)-expressing human GBM cells show stem cell-properties and are radio-resistant. (A) In vivo section of human GBM exemplifies tumor heterogeneity with a large cell population of fast-dividing cells expressing MCM2 and Ribonucleotide reductase and a smaller quiescent population expressing the NSC-marker Musashi-1. (B) In vitro experiments show that Podoplanin is a more reliable than CD133 as a marker of sphere-formation cells, a measure of self-renewal and a property of NSCs. (C) Podoplanin is also a marker of radio-resistant GBM cells cultured as neurospheres. (D) Increasing doses of g-irradiation enrich for Podoplanin-expressing GBM cells xenografted into NOD/SCID mice.