CSIG-11. ONCOSTREAMS: NOVEL STRUCTURES THAT SPECIFY GLIOMAS’ SELF-ORGANIZATION ARE ANATOMICALLY DISCRETE FUNCTIONALLY UNIQUE AND MOLECULARLY DISTINCT

摘要

Whether tumors in general and gliomas in particular are self-organized has not been determined. Brain tumor self-organization would be revealed through individual recognizable structures. Here we demonstrate that brain tumors are indeed self-organized. Brain tumor self-organization is evidenced by the existence of oncostreams, multicellular structures composed of elongated glioma cells. Oncostream structure, function and molecular makeup was studied in several glioma models. In experimental mouse models, and in human GBM tumors, oncostreams were demarcated anatomically as multicellular arrangements of fusiform cells, 10–20 cells wide, of various lengths and distributed throughout the tumors. Using genetically engineered glioma models we found a negative correlation between oncostream density and survival, suggesting that oncostreams contribute to tumor aggressiveness. As oncostreams also reached tumor borders, it is likely that they participate in glioma invasion. Further, by placing slow moving glioma stem cells into oncostreams we detected that these function as streams that distribute glioma cells throughout the tumors. Importantly, we performed the molecular analysis of oncostreams using a Laser Microdissection Microscope to dissect oncostreams from surrounding tumor. The transcriptome landscape was analyzed by next generation sequencing using RNA-Seq and advanced bioinformatics algorithms. The molecular analysis showed that oncostreams are defined by a distinctive transcriptome landscape. We hereby show that oncostreams: (i) are anatomically discrete structures that reveal the existence of glioma self-organization, (ii) are functionally unique as they regulate glioma growth, invasion and spread, and (iii) molecularly distinct, their genetic landscape and associated gene networks differentiate oncostreams from surrounding glioma. Understanding glioma self-organization and its molecular basis will contribute to uncover novel targets for future translational development.