Spatially resolved and single cell transcriptomics

In recent years, massive parallel sequencing has revolutionized the field of biology and has provided us with a vast number of new discoveries in fields such as neurology, developmental biology and cancer research. A significant area is deciphering gene expression patterns, as well as other aspects of transcriptome information, such as the impact of splice variants and mutations on biological functions and disease development. By applying RNA-sequencing, one can extract this type of information in a large-scale manner. The most recent approaches include high-resolution techniques such as single cell sequencing and in situ methods in order to circumvent the problems with gene expression averaging in homogenized samples, and loss of spatial information. The research in this thesis is focused on the development of a novel genome-wide spatial transcriptomics method. The technique is used for analysis of intact tissue sections as well as single cells from solution, with the aim to combine gene expression and morphological information. In Paper I, the method is described in detail, and it is shown that the method is able to generate spatial high quality data from mouse olfactory bulb tissue sections (a part of the forebrain) as well as from tissue sections from breast cancer samples. In Paper III, we adapt the library preparation method in order to be able to execute it on a robotic workstation, thus increasing the reproducibility and the throughput, and decreasing the hands-on time. In Paper IV, we generate 3D-data from breast cancer samples by serial sectioning. We show that the gene expression can be highly variable along all three axes of a tumor, and we track pathways with specific spatial activity, as well as perform subtype classification with three-dimensional resolution. In Paper II, we present a high-throughput method for single cell transcriptomics of cells in solution. The method is based on the same type of solid surface capture as the tissue protocol described in Papers I, III and IV. Again, we show that we can generate high-quality gene expression data, and connect this to morphological characteristics of the analyzed single cells; both using cultured cells and samples from patients with leukemia.QC 20170125