Dimitris G. Placantonakis - US grants

DESCRIPTION (provided by applicant): Glioblastoma multiforme (GBM) is an incurable brain malignancy with limited treatment options. Within GBM, cells with stem-like properties (GBM stem cells or GSCs) initiate and propagate tumors, and are highly resistant to conventional chemoradiotherapy. A major obstacle in understanding the biology of GSCs and developing therapies that directly target them has been the lack of molecular markers that universally identify them. Previous observations indicated that inhibition of Notch signaling, a pathway that regulates fate decisions in neuroglial development, attenuates but does not completely block the self-renewal of GSCs. These findings raise the possibility that Notch activation may be critical to some but not all GSCs, suggesting functional and molecular heterogeneity within the GSC compartment. Prior to Hurricane Sandy, we initiated a set of experiments aiming to clarify the identity of GBM cells in which Notch signaling is activated and understand their contribution to tumor heterogeneity during tumor growth and in response to chemotherapy. Using primary human GBM cultures genetically engineered to express fluorescent reporters in response to activation of Notch signaling, we discovered that, under in vitro conditions that favor GSC self-renewal, Notch is activated in cells that do not express CD133, a well-established cell surface marker of GSCs. Furthermore, we found that in vitro induction of differentiation increases the fraction of cells with activated Notch signaling. These findings raise important questions about the role of the Notch pathway in the cellular hierarchy of GBM: Does Notch signaling identify stem cells with tumor initiating properties or a different type of progenitor cells within GBM? What are the lineages that descend from cells in which Notch signaling is active in vivo? And how do these cells respond to chemotherapy treatment? Unfortunately, the storm inflicted substantial damage to our laboratory, including loss of our primary human GBM cultures. In addition, we had to re-establish our mouse colony in a new animal facility within NYU. Through this funding opportunity, we are requesting funds to extend our preliminary work to further understand the function on the Notch pathway in GBM. Our experiments will assess whether cells with active Notch signaling have tumor- initiating properties and what cell lineages they generate in vivo, including after treatment with the chemotherapeutic agent temozolamide, a mainstay in clinical management of GBM. We anticipate that our findings will generate a critical mass of data that will lead to a publication and a successful submission of an R01 proposal. Importantly, our findings will shed light on an important area of research within neuro-oncology and will facilitate the design of novel and informed therapeutic strategies.

DESCRIPTION (provided by applicant): The prognosis for patients with glioblastoma multiforme (GBM) remains extremely poor despite decades of research. The current standard of care for newly diagnosed glioblastoma is surgical resection to the extent feasible, followed by adjuvant radiotherapy and temozolomide chemotherapy. GBM tumor cells in situ are considered to be radioresistant, which is classically thought to be a cell-intrinsic property. However, recent studies point to the contribution of two non-classical mechanisms that contribute to radiation resistance in GBM: glioblastoma stem cells (GSCs) and the tumor microenvironment (TME). While GSCs employ defined molecular mechanisms that lead to radioresistance, these mechanisms are dramatically potentiated in vivo, suggesting a strong TME influence. Given that non-classical radioresistance in GBM is modulated by the TME, it follows that testing of radiosensitizing agents cannot be performed in cell culture. Instead, novel testing platforms are required that both provide appropriate biological context that takes into account the TME, as well as allow for rapid drug testing. Such testing is further complicated by intertumoral heterogeneity in GBM. The identification of four major molecular GBM subtypes that have different prognoses motivates concerns that such heterogeneity may confound drug testing if specific radiosensitizing agents are efficacious in one subtype but not others. Here we propose to implement a novel approach to screen contextual GBM response to radiosensitizers using organotypic culture of human GBM operative specimens to evaluate the molecular and cellular response to radiation in situ. Based on our preclinical studies and the knowledge of current GBM clinical trials, we propose to evaluate TGF¿ inhibition as a means to increase GBM radiosensitivity to validate this testing platform. The proposed experiments are based on the hypothesis that response to radiation is enhanced by inhibition of TGF¿ in the form of decreased recognition and repair of radiation-induced double-stranded DNA breaks (DSBs). We predict that inhibition of TGF¿ signaling will prevent the observed radiation-induced increase in the prevalence of GSCs in organotypic cultures, as measured by functional clonogenic assays and tumor initiation potential. Importantly, we will evaluate the relative efficacy of TGF¿ inhibitors as radiosensitizers in human GBM specimens representing all molecular subtypes previously described. We posit that this approach, which preserves TME and GSC contributions to GBM radiobiology in an ex vivo setting, will allow for efficient drug screening by incorporating both cellular and functional readouts for drug efficacy, as well as by examining drug effects in distinct molecular subtypes of GBM. Importantly, we envision this approach becoming a paradigm for discovery of radiosensitizing agents that can be applied to other brain tumors.

? DESCRIPTION (provided by applicant): Low-grade gliomas (LGGs) are primary brain tumors affecting young adults. Although they initially grow relatively slowly, they eventually transform t more aggressive high-grade gliomas, leading to neurologic deterioration and death. The questions of cell of origin and significance of known mutations in early low-grade gliomagenesis remain unanswered, thus limiting the ability to develop sensitive detection methods and new therapies. In up to 80% of LGGs, gain-of-function mutations are found in the gene encoding the cytosolic isoform of Isocitrate DeHydrogenase (IDH), usually due to an arginine to histidine substitution at position 132 (R132H). While the wild-type enzyme normally functions to convert isocitrate to a-ketoglutarate (aKG), R132H-IDH1 catalyzes the production of R-2-hydroxyglutarate (2HG), which leads to genome-wide epigenetic modifications that may be related to tumorigenesis. The putative role of IDH1 mutations in gliomagenesis has been supported by in vitro observations in human astrocytes and glioma cells, as well as the fact that patients with Ollier disease, which is due to mosaic germline mutations of IDH1, occasionally develop gliomas. LGGs bearing IDH1 mutations are predominantly located in the frontal lobes in close proximity to the frontal horns of the ventricular system. Because the subventricular zone around the lateral ventricles is an area of active neurogenesis, we postulate that the cell of origin in IDH1-mutated LGGs is a component of the neurogenic niche in the subventricular zone. Mouse models of brain-specific mutant IDH1 expression suffer from perinatal lethality and fail to show tumorigenesis. To test the hypothesis that mutant IDH1 represents a driver alteration in LGG initiation and to investigate which brain cells are predisposed by mutant IDH1 to undergo oncogenic transformation, we propose a new approach that overcomes limitations associated with in vitro and in vivo mouse models. Our strategy makes use of human embryonic stem cells to inducibly express R132H-IDH1 in specific cell types within the human neural lineage: neural stem cells, neuroblasts, astrocytes and oligodendrocytes. We propose to test the hypothesis that inducible expression of R132H-IDH1 alters the self-renewal, differentiation potential, proliferation rate, metabolome and epigenetic/transcriptional profile of neural stem cells or other components of the human neural lineage in vitro. Furthermore, to test the hypothesis that mutant IDH1 expression in specific human neural cell types contributes to initiation of LGG formation, we will transplant these target cells expressing R132H-IDH1 into the mouse brain and assess their ability to form invasive tumors. The proposed research will address the important question of whether IDH1 mutations in human embryonic stem cell-derived neural lineages alter cellular physiology and facilitate oncogenic transformation. Successful completion of this project will lead to a disease model, which will allow detailed analysis of the metabolome, epigenome and transcriptome of early human LGGs. Furthermore, such a model can be used for translational applications, such as high-throughput drug screening and biomarker identification.

PROJECT SUMMARY/ABSTRACT  Glioblastoma  (GBM)  is  a  deadly  primary  brain  malignancy  with  limited  therapeutic  options.  Tumor  progression  is  thought  to  be  driven  by  stem  cell-­like  cells  that  evade  conventional  chemoradiotherapy  and  anti-­ angiogenic treatment. Indeed, anti-­angiogenic therapy and subsequent worsening of tumor oxygenation promote  hypoxia-­resistant  stem  cell  phenotypes  that  lead  to  further  tumor  progression.  However,  our  understanding  of  mechanisms  that  underlie  both  GBM  stem  cell  (GSC)  behavior  and  its  regulation  by  oxygen  tension  remains  incomplete. In our effort to identify novel targetable mediators of the GSC phenotype, we recently discovered that  GPR133 (ADGRD1), an orphan member of the adhesion family of G protein-­coupled receptors, is necessary for  initiating  tumor  growth  in  vitro  and  in  vivo,  both  GSC  properties,  in  part  by  triggering  signaling  mechanisms  that  increase  cytoplasmic  cAMP  and  lead  to  transcription  of  genes  necessary  for  ?stemness?.  While  GPR133  is  absent  from  normal  brain  tissue,  it  is  expressed  with  full  penetrance  in  all  GBM  specimens  tested,  regardless  of  molecular subtype. On the basis of these findings, we hypothesize that GPR133 is a critical component of tumor  growth by supporting the GSC phenotype. We, therefore, believe that GPR133 inhibition represents a novel and  appealing therapeutic strategy in GBM that merits further testing and development.  We  now  seek  to  expand  on  our  published  findings  and  use  patient-­derived  GBM  models  to  elucidate  basic  mechanisms of action of GPR133. Aim 1 will test the hypotheses that GPR133 identifies GBM stem cells and its  knockdown  in  tumor  xenografts  slows  tumor  growth  and  prolongs  survival.  Aim  2  will  build  on  our  finding  that,  within  each  tumor,  GPR133  expression  is  highest  in  the  most  hypoxic  regions,  suggesting  regulation  by  oxygen  tension.  More  specifically,  we  will  determine  the  effect  of  intratumoral  fluctuations  in  oxygenation  on  GPR133  expression  by  correlating  mRNA  and  protein  levels  with  tumor  vascularity  and  oxygenation  using  targeted  intraoperative  biopsies  of  patient  tumors.  In  addition,  Aim  2  will  determine  whether  GPR133  knockdown  synergizes  with  cediranib,  an  anti-­angiogenic  agent,  to  prevent  tumor  progression  after  aggravation  of  tumor  hypoxia.  Finally,  Aim  3  will  determine  the  relative  contribution  of  canonical  G  protein  signaling  initiated  by  GPR133 and transduced by cAMP and its effectors RAP and PKA, and adhesion mediated by GPR133?s long N-­ terminal ectodomain, to the transcriptional regulation of genes that support the GSC phenotype.  The  proposed  studies  will  mechanistically  clarify  GPR133?s  role  in  tumor  progression,  including  in  hypoxia  exacerbated  by  anti-­angiogenic  therapy.  The  results  of  these  studies  will  complement  our  ongoing  small  molecule  inhibitor  and  physiological  ligand  discovery  efforts.  We  envision  GPR133  inhibition  as  a  testable  novel  approach in GBM, either by itself or as a powerful ?one-­two punch? when combined with anti-­angiogenic therapy,  that can target both hypoxia-­vulnerable and hypoxia-­resistant tumor cells.