2007 Project Summaries
Recipients of Three-year Awards
Ronald DePinho, M.D. Glioblastomas (GBM) are among the most biologically aggressive and therapeutically challenging cancers due to their rapid growth and tendency to spread throughout the brain. Alterations of EGFR, PDGFR and phosphatidylinositol 3-kinase (PI3K) signaling pathway components, coupled with loss of p16RB and ARF-p53 pathway function, play direct causal roles in the development of GBM. Recent high-resolution genomic analyses of GBM tumors have also revealed hundreds recurrent copy number aberrations (CNAs), pointing to the presence of many yet-to-be-discovered glioma-relevant oncogenes and tumor suppressor genes. In this proposal, we will exploit our detailed atlas of recurrent copy number alterations in human primary and secondary GBM and low-grade astrocytomas (LGAs) as well as extensive copy number alteration datasets of diverse human cancer types and various mouse models of cancer. The integration of these datasets will serve as a starting point to identify and validate GBM oncogenes that are acquired during progression from LGA to secondary GBM, shared between primary and secondary GBM, and present in other types of cancer. To this end, we employ tumorigenesis screens as a first-pass means of identifying potential GBM progression oncogenes, followed by in-depth mechanistic and clinicopathological studies using mouse and human model systems. Lastly, we will employ stem-transgenesis system to investigate the role of these novel progression related oncogenes on GBM maintenance. Using our stringently filtered list of oncogene candidates and rigorous functional validation systems, the long-term deliverable of this proposal will be able to expand the list of prime targets that may be enlisted into more productive drug discovery and development efforts.
Jeremy Rich, M.D. Glioblastomas are highly lethal cancers for which conventional therapies are essentially palliative. We recently demonstrated that a subset of glioblastoma cells that share characteristics with somatic neural stem cells (called cancer stem cells, tumor-initiating cells, or stem cell-like glioma cells) are resistant to radiation through preferential activation of the DNA damage checkpoint. In another study, we demonstrated that cancer stem cells also promote tumor angiogenesis through VEGF expression. These results and those from several other laboratories suggest that cancer stem cells may be important determinants of the overall behavior of glioblastomas. Therapy directed against cancer stem cells may be more effective in controlling glioblastoma growth but these therapies may have increased toxicities against somatic stem cells; of particular concern are neural and hematopoietic stem cells. We have initiated research efforts to identify molecular targeted therapies that may be effective against cancer stem cells. Screening transcriptomes of glioma cancer stem cells derived from patient specimens and animal models (genetically engineered models, xenografts) to determine genes overexpressed in cancer stem cells relative to neural stem cells and other cancer cells, we have identified a protein that we have termed glioma progenitor cell antigen 1 (GPCAl) that is relatively overexpressed in glioblastoma cancer stem cells. GPCAl is a cell surface protein permitting prospective identification of viable cells expressing GPCAl. Targeting GPCAl expression by shRNA ablates cancer stem cells and slows growth of established tumors suggesting that GPCAl is a potential molecular target in cancer stem cell biology. Based on these preliminary data, we now hypothesize that we can identify molecular targets and therapies against glioblastoma cancer stem cells that can reverse therapeutic resistance of these cancer stem cells while minimizing toxicity against somatic stem cells.
Devanand Sarkar, MBBS, Ph.D. Malignant glioma is the most fatal of all brain cancers. The tumor invades into the surrounding tissue thus limiting complete removal by surgical resection resulting in recurrence. Identifying molecules involved in glioma invasion is an important step to develop targeted effective therapies. We have demonstrated that the expression of Astrocyte elevated gene-1 (AEG-1) is increased in malignant glioma and inhibition of AEG-1 significantly decreases invasion and migration properties of malignant glioma cells. AEG-1 exerts its function by activating the NF-kB signaling pathway. In the nucleus AEG-1 interacts with the p65 subunit of NF-kB as well as with CBP, an activator of transcription that augments NF-kB transcriptional activity. Thus AEG-1 functions as a co-activator of transcription. AEG-1 does not contain any classical DNA-binding domain or transcription activation domain indicating that it exerts its effects predominantly by interaction with other proteins. The long-term objective of the present proposal is to unravel the molecular mechanism of malignant glioma generation and progression so that the garnered information might be exploited to develop novel therapeutic strategies for the more effective management of malignant diffuse glioma tumors. The immediate objective of the present proposal is to authenticate the role of AEG-1 in in vivo regulation of glioma invasion, elucidate in detail the molecular mechanism of AEG-l function, especially in the context of regulation of NF-kB activity, and identify critical AEG-1-downstream genes required for migration and invasion of malignant glioma cells. Our proposed studies are innovative because we aim at understanding the functions of a novel gene AEG-1 that plays an essential role in malignant glioma progression. A direct outcome of the present studies will be the generation of baseline information for developing a therapeutic approach for malignant glioma that selectively targets this molecule. Successful completion of the proposed studies will significantly extend the mission of the Goldhirsh Foundation by generating novel insights into malignant glioma pathogenesis with potential to develop an effective therapy for this aggressive and frequently fatal cancer. Recipients of One-year Awards
Donald Durden, M.D., Ph.D. Malignant brain tumors (high grade gliomas, HGG) represent one of the most frequent neoplasms diagnosed in adults and children. What HGG tumors represent to the adult and pediatric oncologist is a real challenge to current therapy. Importantly, these tumors utilize a common and shared group of signaling proteins to communicate within the cell. This fact identifies a potential target for attack, an "Achilles heal". If we can develop novel drugs which hit this critical control point, we can develop more effective drugs to treat this disease. This observation and other reports in the literature, led us to develop a drug to shut off an important cancer target, the PI-3 kinase signaling pathway. The drug is termed SF1126, which is targeted to the tumor site where it is released like "smart bomb" to hit the tumor and stromal compartment. This drug has now received FDA approval to enter Phase I clinic trials in adult cancer. The next step is to obtain data which proves it will be useful in diffuse malignant glioma. Preliminary results from our laboratory suggest strong activity of this agent inhibits against HGG tumor grown in experimental animals. In our proposal, we seek to generate additional proof that this new drug can help patients afflicted with HGG. We also will determine what pharmacokinetic (PK) and pharmacodynamic (PD) parameters will be useful in predicting efficacy in vivo. If successful, we will bring the SF1126 drug into clinical trials in patients with relapsed diffuse malignant glioma. Finally and perhaps most importantly, we will use our previous drug discovery experience to identify more potent inhibitors against the same target through drug discovery efforts.
Isabelle Germano, M.D. Malignant gliomas are the most frequent primary brain tumors in adult age patients. The prognosis for patients with malignant gliomas remains only one year after diagnosis even with the most aggressive surgical, radiation, and chemotherapeutic treatments. New strategies focused on targeting brain tumor cells while sparing normal tissue include cells and/or gene therapeutic approaches. Recent literature supports the evidence that embryonic stem cells (ESC) can be promising gene therapy vectors for brain tumors. Our recent work shows that we can generate a pure population of astrocytes from mouse ESC. In contrast to other cancer cells, most malignant brain tumor cells are resistant to chemotherapy and radiation therapy that activate cell-programmed death (apoptosis) mechanisms. In a short time frame, the mda-7/IL-24 gene has progressed from a laboratory discovery to a potential therapy for cancer. In particular, mda-7/IL-24 has selective pro-apoptotic antitumor effects, anti-angiogenic properties, a unique ability to radiosensitize tumor cells, and it is safe in the clinical settings. Our recent work shows that we can generate a pure population of transgenic ESC-derived astrocytes with a highly regulated, robust conditional expression of mda-7/IL-24. Based on our recent work, we hypothesize that ESC-derived astrocytes conditionally expressing mda-7/IL-24 can represent an effective and novel approach for the treatment of human malignant gliomas. In this proposal, we will test our hypothesis in the following specific aims: To determine the behavior of mouse ESC-derived astrocytes conditionally expressing mda-7/IL-24 in situ after transplantation into the mouse brain; To determine the pro-apoptotic effects of ESC-derived astrocytes conditionally expressing mda-7/IL-24 in vitro, in vivo, and in situ; To investigate the effects of transgenic mda-7/IL-24 to enhance the pro-apoptotic effects of conventional ionizing radiation (IR) and/or DNA-alkylating chemotherapy in malignant brain tumor cells resistant to conventional therapy. A unique and novel aspect of our proposal is the emphasis that ESC-derived astrocytes can be used for powerful and selective gene transfer in therapeutic strategies for human malignant gliomas. If the approach proposed in our submission were successful, it would not only prove meaningful for the treatment of patients with malignant gliomas but of wide spectrum for other neurological diseases and malignancies.David Sabatini, M.D., Ph.D. Malignant gliomas, like other types of cancer, have been shown to contain a subpopulation of cancer stem cells. Like normal neural stem cells, glioma stem cells possess extensive proliferative potential, long-term self-renewal, and pluripotency – the ability to differentiate into multiple mature cell types. Cancer stem cells are also highly tumorigenic and believed to possess enhanced drug resistance compared to other cancer cells. Thus, brain tumor stem cells appear to be crucial drivers of glioma initiation and recurrence. To develop curative therapies against this cancer, the question that must be answered is what genes expressed in brain tumor stem cells allow them to self-renew, differentiate, and initiate tumor formation. To systematically address this question, we have initiated a collaboration with Dr. William Hahn, a molecular oncology expert at the Dana-Farber Cancer Institute (DFCI), and Drs. Angelo Vescovi, Howard Fine, and Harley Kornblum, three world experts in neuro-oncology. Drs. Vescovi, Fine, and Kornblum have provided us with several cell lines that were isolated from patient brain tumors and enriched for the cancer stem cell subpopulation. To identify genes that control brain tumor “stemness” in these cell lines, we will perform a comprehensive loss-of-function screen using a library of approximately 13,000 shRNA hairpins expressed from lentiviral vectors that we have developed. The unique lentiviral shRNA libraries are able to achieve robust, stable knockdown of thousands of human genes in various cell types and can be used to perform arrayed screens, in which each shRNA is tested in an individual well of a multiwell plate. The main accepted assays for brain tumor “stemness,” namely neurosphere formation and marker expression, are image-based, and we will combine the arrayed shRNA screen with automated fluorescence microscopy and high-throughput image analysis, using software that was also developed here, to ascertain candidate drivers of cancer stem cell functions. By the end of the one-year funding period, we will identify a list of hits, which will subsequently be retested, stringently validated, and investigated in follow-up in vitro studies and in vivo tumorigenesis experiments. Eventually these studies may identify potential therapeutic targets for malignant glioma. Khalid Shah, M.Sc., Ph.D. Primary gliomas present one of the greatest challenges in oncology. Novel therapies to treat gliomas are urgently needed as traditional treatments which rely on nonspecific, cytotoxic approaches have had marginal impact on patient survival. The recognition that embryonic stem cell (ES)-derived neural stem cells (eNSC) can incorporate into the cyto-architecture of brain following transplantation has unveiled new roles for their use as therapeutic agents for the treatment of diseases of central nervous system. In prior research we have established that NSC: a) migrate extensively to sites of cerebral pathology; and b) can be engineered to express therapeutic protein, secreted tumor necrosis factor receptor apoptosis inducing ligand (S-TRAIL), which selectively induces apoptosis in proliferating glioma cells. In close collaboration with: a) Kwang-Soo Kim lab (McLean Hospital, Boston) who has successfully employed ES derived neural stem cells in mouse models of neuro-degeneration; and b) Rich lab (B&W Hospital, Boston) who has recently developed DT resistant mammalian cells, we will initially develop DT resistant eNSC and arm them with IL-13R and EGFR-targeted forms of DT. Differently armed eNSC will be tested for their therapeutic efficacy in vitro and in mouse glioma models. A number of studies have shown synergistic anti-tumor effects in which cytokine TRAIL has been combined with other therapies. In order to explore the combined effect of S-TRAIL and targeted-DT, we will further engineer DT resistant eNSC with S-TRAIL and test the efficacy of combination therapy targeting both IL-13R/EGFR and death receptors (DR)-4/5 in glioma cells. We hypothesize that simultaneous activation of death pathways and inhibition of protein synthesis in glioma cells would result in enhanced eradication of gliomas. The proposed combination strategies are expected to have a major impact in developing more efficient stem cell therapies for brain tumors that will be compatible with clinical trials.
Irving Weissman, M.D. Extensive molecular characterization of Brain tumors done in recent years has proven valuable for tumor classification, risk stratification and outcome prediction for current treatment modalities. To design new therapeutic strategies however, it is critical to isolate and investigate the relatively few cells within a tumor that are directly responsible for its continuous growth, that is, the tumor-initiating cells or cancer stem cells(CSC). The proposed study is designed to improve the definition and functional characterization of brain tumor stem cells in Glioblastoma Multiforme (GBM). Our method of CSC isolation from primary GBM samples will utilize novel candidate cell surface markers. To identify GBM stem cell-specific antigens we have generated a monoclonal antibody library against human GBM samples. Freshly isolated human GBM tumor cells will be fractionated based on the presence of cell surface antigens identified by these antibodies. Select cell populations will then be tested for their clonogenic and tumorigenic capacities. The purification of GBM Brain Tumor Stem Cells (BTSC) will be followed by the characterization of mutations, chromosomal aberrations, active signaling pathways and unique surface antigens. This knowledge is necessary to guide the development of highly specific and more effective anti-GBM therapies.
Despite diagnostic and therapeutic advances, the outcomes of patients with high grade gliomas continue to account for significant morbidity and mortality. These neoplasms demonstrate marked invasion and neovascularization that, to a great extent, account for the failure of current therapeutic attempts. Invasion and neovascularization have been associated with the presence of tumor hypoxia, which triggers an enhanced expression of many genes under the control of the transcription factor hypoxia-inducible factor (HIF)-l in the tumor cells and/or tumor-associated endothelium. These genes include vascular endothelial growth factor (VEGF), VEGF receptor (VEGFR)-2, the chemokine receptor CXCR4 and its ligand stromal cell-derived factor (SDF)-l. While the process of neovascularization was first thought to occur primarily via sprouting and remodeling of existing vessels (angiogenesis), accumulating evidence suggests that formation of blood vessels via trafficking of bone marrow-derived CXCR4-expressing endothelial progenitor cells (EPCs) to the tumor site and differentiation into endothelial cells in the tumor vascular bed (vasculogenesis) contributes to neovascularization. Antiangiogenic therapy has emerged as a promising new approach, but selection of its optimal biologic dose requires alternative, non-invasive biomarker(s) of tumor response. Whether CXCR4 inhibition could be used to interrupt both the invasion and neovascularization of CNS tumors has not been studied. Furthermore, whether EPCs contribute to the neovascularization of these tumors and whether they could be used to assess prognosis and monitor neovascularization and its therapeutic response have not been examined. Thus, we have decided to combine the efforts from two experienced laboratories for: Aim I: To test CXCR4 antagonist AMD3100 on the migration ofGL261 glioma cells in vitro. Aim II: To test the effect ofCXCR4 and VEGFR2 inhibition, alone or in combination, on the growth, invasion and neovascularization in the GL261 glioma murine model in vivo. We will also evaluate the levels of EPCs in peripheral blood as a potential surrogate marker of neovascularization and anti-angiogenic drug efficacy.
|
 |
 |
