Imaging the spatial organization of healthy human brain and glioblastoma

Abstract

Multicellular biological systems have a single genetic plan that contains the set of instructions to shape the three-dimensional architecture of specialized tissues and organs. The genetic plan is contained in the DNA, and information is relayed through RNA to produce proteins that grant cells with the power to carry out different functions. As cells specialize giving rise to differentiated cell types of a particular tissue, they express different RNA molecules that will reshape their protein content and grant them with a certain phenotype. The development of scRNA-seq technologies at the start of the 2010s allowed researcher to quantify RNA expression at high gene throughput, enabling a deep molecular characterization of cell types.

However, scRNA-seq methods requires dissociation of tissues and organs into cell suspensions, and the spatial organization is lost. The present thesis features the need for high-throughput spatial transcriptomics methods, to enable the characterization of cell types and the spatial architecture of the adult human healthy brain and adult diffuse high-grade gliomas. The thesis is composed of three parts: first, the evolution of the spatial transcriptomics methods and field; second, cell types of the human healthy brain; and third, the biology of adult high- grade diffuse gliomas, and more specifically glioblastoma.

In paper I we present Enhanced Electric Fluorescent In Situ Hybridization, or EEL- FISH, a high-throughput spatial transcriptomics methods that enables to obtain the expression profile of up to 448 RNA species per fluorescent channel over large tissue areas. The ability speedily to image over large tissue area was achieved by enhancing the RNA capture on a surface, which reduced unnecessary imaging time through the tissue thickness (z-axis). Moreover, EEL-FISH enabled imaging of human tissue sections, which are typically challenging for smFISH due to their high autofluorescence. EEL FISH reduces tissue autofluorescence because the tissue is removed after RNA has been captured. Paper II focuses on the direct application of single cell technologies to characterize the cell types of over 100 regions of the human adult brain of healthy individuals. To this end, we employed scRNA-seq to create a census of cell types and combined with EEL-FISH to characterize the spatial organization of cells in specific brain areas. Lastly, we used EEL-FISH to understand the spatial organization of adult high- grade gliomas, more specifically glioblastoma, in paper III. We characterized the expression of 888 RNA species, by using two fluorescent channels, of 57 sections from 27 different high-grade glioma patients (2 oligodendroglioma and 25 glioblastoma). We generalized a model of the spatial organization of glioblastoma, which organizes across a gradient of hypoxia and wound response that are activated on top of a glial-like tumour cell.

Along this thesis, I will review the literature of spatial transcriptomics, and the biology of the adult human brain and glioblastoma from a single cell perspective, and discuss the results of the research conducted in papers I, II and III with respect to these topics.