Tumors are caused by a genetic, epigenetic, and developmental defects. Each may trigger deviation from the original cellular developmental program and subsequent imbalances in patterns of growth, maturation, and proliferation may produce tumorigenesis and tumor progression. Once uncontrolled growth has developed into a clinically apparent tumor mass, amplifying cells often independently generate additional genetic mutations and deletions, with a rapid increase in the number of genetic and epigenetic changes (?genetic heterogeneity?). At the same time, tumor cells interact with genetically intact, reactive host tissues, with variable histopathological results (?morphologic heterogeneity?). The combination of genetic and morphologic heterogeneity of tumors, in association with developmental deviations, constrain optimal analysis of the complex tumor state. However, recent advances in molecular and pathological techniques provide tools for gaining greater insight into these problems. We are taking several approaches to gain insight into tumorigenetic mechanisms: we are applying molecular diagnostic tools to study tumorigenesis in several inherited and sporadic human tumor conditions and we are developing new technologies that are tailored to the analysis of complex morphologic arrangements as well as technology for diagnostic and therapeutic applications. Our focus is on translational research that links basic science with the clinic. Germline mutation versus germline deletion in VHL disease. In contrast to patients with germline mutation, patients with germline deletion consistently revealed wild-type mutations as ?second hits?. The unique genetic mechanism associated with the ?second hit? in germline deletion cases is not a random event. It is related to a genetic selection process that is required to maintain integrity of at least one allele at that gene locus for cell survival. Based on this finding, we have developed a mechanistic hypothesis, a hypothesis that also applies to other tumor suppressor gene syndromes, that the milder clinical course in the VHL germline deletion patients results from selection of the rare sporadic genetic hit as the ?second hit?. Glioma progression and glioma differentiation. Selective tissue microdissection was used to obtain pure populations of tumor cells, which we studied using two-dimensional protein gel electrophoresis (2-DGE) and protein sequencing of select target proteins expressed differentially among tumors to distinguish between the two main categories of glioblastomas (GBMs), de novo and secondary GBMs. We isolated and sequenced 11 unique proteins that were differentially expressed in the primary and secondary GBMs and that produced two distinctive proteomic patterns. Thus, the two patterns of GBMs, primary versus secondary, previously distinguished by clinical and genetic differences, can be recognized at the protein level, and may have implications for prognosis and treatment options. We are using the same approach to identify differentially expressed proteins in different stages of glioma formation, as well as within gliomas with different phenotypic expression. N-CoR pathway targeting induces brain tumor stem cell differentiation. By comparative proteomic screening, we identified nuclear receptor co-repressor (N-CoR) as an overexpressed protein in GBM. Additional studies indicated that the subcellular localization and presence of N-CoR is related to the differentiation state of GBM cells. Cells expressing nuclear N-CoR do not express glial fibrillary acidic protein (GFAP), a marker of astrocytic differentiation. Cells with cytoplasmic N-CoR or loss of N-CoR express GFAP. Some cells with nuclear localization of N-CoR also express CD133, a marker of neural stem cells. When brain tumor stem cells are treated with a differentiation inducing cytokine, the cytoplasmic fraction of N-CoR increases, indicating translocation. We are excited about this finding because the N-CoR pathway represents a novel therapeutic target based on a differentiation strategy. Our studies show a striking synergistic growth inhibition in a glioma cell line treated with two different compounds intended to promote phosphorylation and subsequent cytoplasmic translocation of N-CoR. We believe these findings provide information for enhancing current understanding of glioma biology which also may lead to new therapeutic options for patients. Pathogenesis of malignant glial tumors. We previously identified a primate model for glioma tumorigenesis using a radiation dose that is sufficient to initiate intracerebral tumors, but low enough to allow long-term survival. Morphological analysis revealed the tumors to be de novo glioblastomas, which often occurred multifocally. Brain and tumor tissue from these Rhesus monkeys provided clues to understanding of primary glioma tumorigenesis. We have recently identified tumor precursor cells in many normal white matter areas. Developmental biology and tumorigenesis. Analagous to other TSG syndromes, tumorigenesis in VHL disease is most commonly initiated by a VHL wild-type deletion in susceptible cells. Several key questions, however, remain unexplained in most, if not all, tumor suppressor gene syndromes: 1) in any organ, only one specific type of tumor occurs; 2) tumorigenesis is restricted to specific sets of organs; and, 3) there is no obvious association between tumor suppressor gene function and tumorigenesis. Our recent studies on the histogenesis of hemangioblastomas revealed evidence that hemangioblastomas represent developmentally-arrested tissue. We have further established developmental effects of pVHL deficiency on central nervous system tissues by the discovery of numerous mesenchymal precursors that precede hemangioblastoma formation. In analogy to embryonal blood island differentiation, blood island formation in hemangioblastomas is associated with transient expression of the erythropoietin (Epo) receptor (EpoR). EpoR expression coincides with expression of Epo secondary to VHL deficiency, which we have recently proposed as a mechanism of tumor progression. This mechanism of tumor progression appears to be applicable to other VHL disease-associated tumors, including renal clear cell carcinoma and endolymphatic sac tumor. We worked on an expansion of the developmental concept of VHL tumorigenesis by analyzing endolymphatic sac tumors and tumor-free samples of endolymphatic sac and duct and found evidence of abundant microscopic epithelial precursor structures. In spite of their microscopic size, we could consistently demonstrate VHL gene inactivation in these structures by their upregulation of HIF1 and HIF2 as well as other HIF targets. We are currently looking for further supportive evidence in VHL target structures outside the CNS, e.g. epididymis (preliminary evidence shows an abundant number of VHL-deficient precursor structures in human epididymis), pancreas or kidney. Other current projects on VHL disease-associated hemangioblastomas include a) expression of developmental hemangioblast-associated genes in hemangioblastomas, b) detailed characterization of hemangioblastoma precursor material including differential upregulation of hypoxia-inducible proteins 1 and 2 and HIF target VEGF by in-situ hybridization, c) detailed analysis of angiogenesis and vasculogenesis in hemangioblastoma which is frequently associated with tumor cell differentiation into red blood cells, d) closer characterization of the evolution of VHL tumorigenesis in cerebellum and brainstem e) investigation of the pathomechanisms of hemangioblastoma cyst formation to exactly define different stages of hemangioblastoma progression. Representative tissues from different stages are investigated with newl proteomic techniques for specific stage-associated protein expression.
Showing the most recent 10 out of 79 publications
