GENT-53. SELF-ORGANIZATION OF GLIOMAS: GENETIC RODENT MODELS, GENOMIC NETWORKS, AND MATHEMATICAL MODELING

Abstract

The existence of large scale organization in gliomas remains unknown. We here demonstrate that gliomas are indeed self-organized. We do so by uncovering the cellular, genomic, and mathematical networks that underlie glioma self-organization and aggressive behavior. Glioma self-organization was studied in genetically induced gliomas using the Sleeping Beauty Transposase system expressing/down-modulating N-Ras, large T ag from SV40, TP53, PDGF, ATRX, and IDH1-R132H. We discovered the existence of onco-streams within gliomas, the glioma equivalent of self-organizing flocks of birds. Onco-streams are 10-20 glioma cells wide, of variable length, formed by fusiform cells, and are distributed throughout tumors. Onco-streams participate in glioma growth, invasion, and lethality. Onco-streams play various roles: they enhance migration of slower moving glioma stem cells, and block immune cell invasion of gliomas. Both in human and in rodent gliomas the prevalence of onco-streams correlates positively with tumor aggressiveness. Secondly, we modeled gliomas’ onco-streams using agent based computational modeling. Cellular interactions in our model were exhaustively studied using numerical investigations of our agent- based model. Our model predicts that changes in cell shape affect the population behavior of glioma cells, induce the formation of onco-streams, and thereby increase tumor aggressiveness. These predictions are now being tested experimentally. Thirdly, we analyzed the genomic structure of gliomas’ self-organization. Utilizing RNA-seq and bioinformatic analyses we uncovered a gene expression network that underlies onco-stream formation, and aggressive glioma behavior. One of the most highly connected gene nodes was selected for inhibition. Specific pharmacological and genetic targeting of this network demonstrated strong inhibitory effects only of the most malignant glioma models, as predicted by our analysis. We conclude that gliomas are self-organizing structures. This is supported by experimental, mathematical, and genomic studies. Finally, this approach will lead to the discovery of hitherto unsuspected essential therapeutic targets as is illustrated herein.