2005 Project Summaries
Recipients of Three-year Awards
Arturo Alvarez-Buylla, Ph.D. Astrocytes have been considered the likely source of glioblastomas for some time, but the origin of brain tumors remains unresolved. Our primary goal is to identify the cells that give rise to gliomas and to learn more about the function and properties before and after malignant transformation. Our previous work showed that some astrocytes in the subventricular zone (SVZ) of adult animals function as stem cells. Recent work indicates that tumor cells share many properties with stem cells, including the capacity for extensive self-renewal and the ability to give rise to progeny with various states of differentiation, some of which continue to divide. Moreover, it has been shown that brain tumors contain CD133+ brain tumor stem cells (BTSCs) that are required for tumor propagation, and which form multipotent neurospheres when cultured in vitro. Although it has been widely hypothesized that astrocytes with stem cell properties are the cell-of-origin of gliomas, the subtype of astrocytes that function as stem cells and their potential to generate tumors have not been determined. We have identified a subpopulation of astrocytes in the SVZ that express the receptor for platelet-derived growth factor (PDGF) and hyperproliferate when stimulated with PDGF. The PDGF ligand and cognate receptor are overexpressed in nearly 50% of human astrocytomas of all grades and in >80% of oligodendrogliomas, suggesting that upregulation of this pathway may be involved in tumor initiaiton. Given the prevalence of PDGF pathway activation in human gliomas, we hypothesized that these PDGFR+ cells in the SVZ may play an important role in tumor initiation and progression. These PDGFR expressing astrocytes may be a subpopulation of stem cells that are the source of some gliomas. We propose: 1) to ascertain the fate of PDGFR+ SVZ astrocytes in the normal brain and during tumor growth; 2) to determine if PDGF stimulation of non-stem cell astroyctes in non-germinal regions of the brain hyperproliferate in response to PDGF, whether PDGF stimulation uncovers latent stem cell properties in these astrocytes and whether these cells can give rise to tumors; 3) to determine if PDGFR+ astrocytes are the stem cells of the adult brain and of some gliomas, and as such possible targets for therapeutic intervention. An understanding of the cell-of-origin of brain tumors will lead to new methods of early diagnosis and treatment.
Frank Furnari, Ph.D. Despite in depth knowledge of glioblastoma mutliforme (GBM) pathogenic lesions and decades of advances in neurosurgery, radiation therapy and clinical trials, little improvement in the medium 12-month patient survival has been achieved. A central issue that confounds successful treatment is the heterogeneous nature of this aggressive tumor. As a result, multiple and spatially distinct clonal populations exist within a GBM, making any lesion- or pathway-specific therapy less effective. The overall goal of this project is to dissect the mechanisms whereby GBM heterogeneity drives tumor aggressiveness and therapeutic resistance. As an entry point to this study, we will examine amplification of the epidermal growth factor receptor, a hallmark mutation present in 50% of cases, often in a heterogeneous manner and frequently associated with structural alterations. The most common of these alterations (EGFRvIII, de2-7EGFR, deltaEGFR; herein referred to as EGFR*) results in a constitutively active mutant receptor with tumor enhancing capability; a function lacking from amplified wtEGFR despite its highly pervasive tumor expression in comparison to focally occurring EGFR*. This disconnect between tumorigenic potential and inverse frequencies of wt and mutant EGFR led us to discover that EGFR* augments via a paracrine fashion the tumorigenic growth of neighboring cells expressing amplified wt receptor. We propose this finding will allow the modeling of GBM heterogeneity to determine biological significance and therapeutic responsiveness with respect to the tumor’s most common genetic lesion. To achieve this, 3 aims are proposed: 1) we will determine the cell intrinsic, EGFR*-mediated mechanism(s) conferring enhanced proliferation to wtEGFR-expressing cells by using EGFR* phosphotyrosine docking site mutants; 2) we will determine and genetically manipulate the EGFR*-secreted factors which enhance proliferation of wtEGFR-expressing cells and; 3) we will determine the clinical value of functionally validated molecules in Aims 1 and 2 through concordant expression with wtEGFR/EGFR* in clinical samples. Subsequently, the most highly correlative findings will be targeted in preclinical therapeutic approaches using tumor models recapitulating receptor heterogeneity. Recipients of One-year Awards
Robert Bachoo, M.D., Ph.D. In the absence of truly robust astrocyte-specific promoters, the central question of ‘glioma cell of origin’ will remain unanswered. We have developed experimental and computational methods to define the astrocyte transcriptome and identify rigorously validated specific markers of mature astrocytes. These in situ validated astrocyte specific genes provide a foundation upon which to test the capacity of various promoter/enhancer elements for targeting of genetic elements to the mature astrocyte compartment. I seek support to generate and test promoter elements of 3 astrocyte-specific genes for targeting expression of Cre and CreER-T2 recombinase. The spatiotemporal expression pattern of the promoters will be analyzed by monitoring both cell type-specific transgene expression and LacZ signal and in the embryonic and adult brain and across multiple tissue compartments. The generation of mouse models that faithfully recapitulate human GBM would enable genetic and biological dissection of disease initiation and progression and facilitate the systematic evaluation of targeted therapy.
Maria-Magdalena Georgescu, M.D., Ph.D. The broad objective of this project is to characterize the role of the PDGFR-NHERF-PTEN complex in the migration of glioblastoma cells. Both loss of PTEN tumor suppressor and overexpression of platelet-derived growth factor receptor (PDGFR) or of its ligands are common genetic modifications in gliomas. We found that the adaptor-type molecules EBP50/NHERF1 (ezrin-radixin-moesin (ERM)-binding phosphoprotein 50; Na+/H+ exchanger regulatory factor 1) and its homologue NHERF2, couple PTEN to PDGFR. Our hypothesis is that NHERF proteins are involved in the migration of astrocytes and that a dysfunction of either component of the PDGFR-NHERF-PTEN complex leads to glioma cell invasiveness. The project will cover the following topics: (1) molecular and dynamic characterization of the PDGFR-NHERF-PTEN complex upon stimulation and inhibition of PGDFR in glioblastoma cell lines and in EBP50(+/+) and (-/-) mouse astrocytes; (2) characterization of the migratory phenotype of glioblastoma cells and astrocytes with controlled genetic alterations of EBP50 and PTEN by in vitro real-time imaging and in vivo cross-hemisphere migration of labeled cells; and (3) detection of NHERF proteins in glioblastoma cell lines and brain tumor samples by Western blot analysis and immunohistochemistry. This study will be instrumental for the understanding of the mechanism of PTEN membrane recruitment and activation in order to suppress the signaling of PDGFR and the motility and invasiveness of glioblastoma cells.
Qing Richard Lu, Ph.D. The most common human brain tumors are thought to be of glial cell origin. Brain tumors (gliomas) are the fourth leading cause of cancer-related deaths in the United States. Despite the enormous clinical impact of brain tumors, the biological mechanisms of glioma formation and those that sustain the disease states are not well understood. Recently, I have identified a pair of basic helix-loop-helix (bHLH) transcription factors Olig1/2, which are specifically expressed in oligodendroglial lineage cells in the adult rodent brain. Olig2 is essential for oligodendroglial cell fate specification from neural stem cells in the developing CNS. Interestingly, human OLIG genes are expressed strongly in oligodendrogliomas and glioblastomas. Currently, it is not clear whether OLIG genes have an oncogenic role in glioma formation. In the proposed studies, I will establish a transgenic model to determine if overexpression of the bHLH gene Olig2 can lead to brain tumorigenesis in combination with oncogene activation and/or tumor suppressor mutation. In addition, to circumvent the neonatal lethality of Olig2 null mice, I have generated floxed Olig2 conditional knockout mice to study Olig2 function in adult mice. The floxed Olig2 allele is susceptible to the Cre recombinase-mediated recombination in a cell type- and time-controlled manner. Thus, I will determine whether conditional ablation of Olig2 in neural progenitor cells will prevent brain tumor formation in adult mice using a mouse glioma model. These studies should contribute significantly to the molecular mechanisms of glioma formation and offer new avenues for better diagnosis and treatment of human brain cancers.
William A. Weiss, M.D., Ph.D. Amplification of EGFR, deletion of PTEN, and activating mutations in PI3-kinase itself represent common mechanisms through which PI3-kinase is activated in glioma. Although PI3-kinase inhibitors should be useful in this disease, the small molecule PI3-kinase inhibitors are not viable clinical candidates due to their limited target selectivity, poor chemical stability (wortmannin), and low potency (LY294002). We hypothesize that selective inhibitors of PI3-kinase isoforms will show decreased toxicity, ultimately allowing these compounds to be used in the treatment of patients with glioma. To address this issue, Zachary Knight and Kevan Shokat at UCSF have recently synthesized a novel series of isoform-selective inhibitors of the particular PI3-kinase (p110) isoforms that contribute most prominently to cancer. These agents represent the first new tools available in a decade for analysis of PI3-kinase signaling, and are available to us through an ongoing collaboration with the Shokat lab. We screened these compounds against glioma cell lines, and show here that the PI3-kinase p110α inhibitor ZK75 has remarkable efficacy against glioma. We hypothesize that ZK75 will be effective in the therapy of glioma. The specific aims of this project are therefore: 1. To assess the importance of p110α-blockade, and inhibition of PI3-kinase related kinases ATM and ATR, in contributing to the overall potency of ZK75. The p110α inhibitor ZK75 is unique among other p110α inhibitors in its efficacy against glioma. Apoptosis mediated by ZK75 is dependent on p110 inhibition, whereas arrest at G2M occurs independently of p110 inhibition. We therefore hypothesize that the efficacy of ZK75 results both from inhibition of p110α, and from inhibition of a PI3-kinase related kinase (PIKK). In this manner, ZK75 functions as PI3-kinase inhibitor, and also akin to an alkylating agent (like temozolomide) that arrests cells through activating a DNA-damage checkpoint. ATM (ataxia-telangiectasia-mutated) and ATR (ATM-Rad3-related) are DNA damage-signaling PIKKs that regulate cell cycle checkpoint activation and DNA repair. In this aim, we will characterize the impact of ZK75 on ATM and ATR to determine how PIKK blockade contributes to the efficacy of ZK75. 2. To test ZK75 in-vivo, and to determine how the efficacy of therapy is impacted by mutation at EGFR or at PTEN. In-vivo experiments will involve treatment of both xenografted cell lines, and of primary tumor xenografts that maintain amplification of EGFR, and in which PTEN status is known. We will interrogate xenografts wild-type or mutant (amplified in the case of EGFR) at one or both loci, in order to better understand how these genetic alterations impact response to ZK75. |
 |
 |
