Abstract LB-191: Glioblastomas can originate from neurons in the CNS

Abstract

Abstract Malignant gliomas remain one of the most aggressive tumors of the central nervous system. Different interpretations have been proposed about the nature of the neural cell type that is targeted by transformation and results in tumorigenesis. The identification of the cellular origin of gliomas presents an opportunity for improving our understanding of this type of cancer. We recently developed a mouse glioma model using Cre-inducible lentiviral vectors that faithfully recapitulate the pathophysiology of human glioblastoma multiforme (GBM). Injection of a single lentiviral vector expressing H-RasV12-shp53 in the cortex of Synapsin I-Cre (SynI-Cre) mice led to tumor formation after 6-8 weeks of injection. SynI-Cre mice primarily and specifically express the Cre recombinase transgene in differentiated neurons. Tumors were also obtained when CamK2a-Cre mice, also expressing Cre specifically in neurons, were injected with the same virus. We also aim to target astrocytes by injecting the virus either in the cortex or the stratium of GFAP-Cre mice, and tumors presenting the classical characteristics of GBM developed, suggesting that astrocytes can also serve as the glioma cell of origin. We made sections of these brains at various time points following injection of the lentiviral vector and, using high resolution large-scale mosaic imaging, we examined the expression of different markers. Notably, tumors start out to be GFAP+, but by eight weeks are largely Nestin+ and Sox2+. We believe that either astrocytes or neurons can be reprogrammed by the introduction of oncogenes/tumor suppressors to form cancer iPS-like stem cells that can give rise to all the cell lineages and heterogeneity observed in GBM. To further explore this hypothesis, we transduced primary cortical astrocytes and neurons obtained from GFAP-Cre and SynI-Cre mice, respectively. The transduced cells when switched to neural stem cell (NSC) media displayed: i) neurosphere-like structures, ii) robust NSC marker expression (Nestin and Sox2), iii) self-renewal capacity, iv) strong tumor initiating capacity, v) expression of reprogramming factors, and vi) capacity to differentiate into different lineages. Finally, we assessed the human relevance of our findings by comparing the transcriptome profile of tumors in our model with the molecular signatures of human glioma samples. The data from the molecular signatures and histopathology of tumors originating in the cortex where the primary target is astrocytes in the GFAP cre mice and Neurons in the Synapsin Cre mice show both are mesenchymal GBM subtype. We obtained Neural subtype mostly when the virus was injected in the hippocampus of Nestin-Cre mice (aim to target NSC/progenitor cells). Together, our results suggest that any cell in the brain, whether terminally differentiated or neural stem cell, can be the glioma cell of origin and the biological behavior of these tumors depends on the dysregulation of specific genetic elements. Citation Format: {Authors}. {Abstract title} [abstract]. In: Proceedings of the 103rd Annual Meeting of the American Association for Cancer Research; 2012 Mar 31-Apr 4; Chicago, IL. Philadelphia (PA): AACR; Cancer Res 2012;72(8 Suppl):Abstract nr LB-191. doi:1538-7445.AM2012-LB-191