Bakhos A. Tannous, PhD
Associate Professor, Department of Neurology, Harvard Medical School
Associate Neuroscientist, Neurology, Massachusetts General Hospital
DF/HCC Program AffiliationTranslational Pharmacology and Early Therapeutic Trials
Lab WebsiteViral Vector Development FacilityExperimental Therapeutics and Molecular Imaging Lab
Research AbstractGlioblastoma tumors are highly heterogeneous and there is a complex interaction among different types of tumor cells and stromal cells within the tumor. Recently it has been shown that the majority of tumor cells do not have the capacity to recapitulate a phenocopy of the original tumor and that only a small subpopulation of cells in the tumor, called cancer stem cells, have that ability upon xenotransplantation in nude mice. These cancer stem cells appear to be more resistant to conventional therapy (chemotherapy and radiation) as compared to the non-cancer stem cells. Following current therapy for high-grade glioma tumors, most patients die within a year from a new secondary tumor foci forming within one centimeter of the resected area. These foci are enriched for cancer stem cells, and it is likely that they are responsible for tumor recurrence. One aspect of our work is to develop novel bioluminescent reporter systems to monitor self-renewal, differentiation or death of GBM stem cells simultaneously and use these reporters to screen for drugs which can increase therapeutic efficacy for GBM stem cells by either: (1) eradicating these cells or (2) revert GBM stem cells into a more differentiated state which can then response to conventional therapies.
Another aspect of our research is to target brain tumors using different gene transfer technologies. In collaboration with Dr. Ralph Weissleder, we have engineered a reporter protein which is metabolically biotinylated and selectively presents biotin on the tumor cell surface. This reporter allows sensitive imaging of glioma cells expressing it using labeled streptavidin moieties and different imaging modalities, such as positron emission tomography and magnetic resonance. We are currently extending this work and developing a gene delivery system that can change the surface of a GBM cell so that it presents a molecular “beacon” (biotinylated receptors) to attract diagnostic and therapeutic agents to it. This strategy is designed to make tumor cells a selective, “easy target” for different biotinylated-imaging agents (for diagnostics) or biotinylated-toxins (for therapy) complexed to streptavidin which binds with extremely high affinity to biotin (Kd = 10-15) and can be internalized into tumor cells. A critical component of our strategy is to facilitate delivery of these biotinylated therapeutic/diagnostic agents from the vasculature into gliomas by targeting endothelial cells lining the blood vessels in the brain which normally facilitate passage of molecules across the blood-brain barrier.
Since the last decade, a largely secluded source of gene expression data has accumulated, and is publicly available, comparing human cancer samples with normal tissue. This encouraged us to develop a strategy to order glioblastoma gene families by level and frequency of overexpression in order to identify putative treatment targets in the glioblastoma kinase gene family. In collaboration with Dr. Wurdinger at the VU Medical Center, the Netherlands, we identified a GBM-specific kinase expression profile that suggests signal transduction pathways to be involved in gliomagenesis. Out of the identified glioblastoma-specific overexpressed kinases, we selected WEE1 kinase as a putative treatment target based on its potential function as a mitotic gatekeeper and its specific overexpression in GBM. We showed that GBM tumors are treated efficiently upon inhibition of WEE1 kinase using a drug or small interfering RNAs (siRNAs) in combination with conventional therapy. We are currently working on identify the role of WEE1 tyrosine kinase as well as different members of the TP53 pathway mutated in over 80% of GBM and known to affect WEE1 inhibition activity.
Pre-clinical analysis of any diagnostic/therapeutic strategy using clinically relevant animal models is very important for translational research. We use primary GBM stem-like cells dissociated from patient tissue sections and cultured as neural spheres which recapitulate a phenocopy of the patient tumor with infiltrating/migrating capability upon sterotactic injection in the brain of nude mice.