Recent advancements in single-cell RNA sequencing have transformed the way we classify cells. These innovations have provided an unprecedented view into the complexities of cell types within the human body and enabled the creation of comprehensive cellular atlases for various tissues and organs. While these techniques have revolutionized biology, they are limited in that they lack the necessary information to examine the architecture of complex tissues. This limitation has led to the development of spatially resolved transcriptomic (SRT) techniques, giving rise to the field of spatial biology. The rapid growth of spatial biology has largely been driven by academic research, though commercial entities have recently begun to emerge in the market.

In situ sequencing (ISS) is one such SRT technique that enables the mapping of hundreds of transcripts directly in tissue samples. ISS utilizes padlock probes, which hybridize to a target. Ligation of the probes allows for their circularization. Once circularized, the probes are enzymatically amplified, and their identity can be decoded using fluorescent probes and a microscope. 

In Paper I, we build upon the first iteration of ISS to create a hybridization-based sequencing approach  that allows for more genes to be targeted. Additionally, the improved signal-to-noise ratio enabled more robust signal detection, facilitating the application to aged human brain tissue, which is challenging to analyze due to autofluorescence.

In Paper II, we use the technology developed in Paper I to map out 75 transcriptomically defined cell types in the human cortex. In total, we pinpointed the location of 59,816 cells in their native positions and abundances. We looked at both the within- and across-layer distribution of these cell types.  

In Paper III, we profiled the development of oligodendrocyte lineage cells in mouse spinal cord and brain. Over different time points, we could map out the location of cell types and model the development of these cellular lineages, uncovering the neighboring preferences of these cells. The data revealed spatial heterogeneity of oligodendrocyte lineage progression in the brain and spinal cord. 

In Paper IV, we modeled the development of neuroinflammatory lesions in the multiple sclerosis model of experimental autoimmune encephalomyelitis (EAE). We built a comprehensive atlas of the spatio-temporal dynamics of the disease development by collecting tissues from different time points and different regions of the CNS. We uncovered the dynamic nature of disease-associated glial subtypes, being induced globally at peak and then reverting back to their homeostatic gene expression signature at the reduced inflammatory state. Moreover, the temporal analysis of the lesion development allowed for uncovering the intricate structure of these lesions and how they propagate over time. Human tissue sections confirmed the cell-driven approach to lesion identification and the induction of disease-associated glial subtypes. 

Taken together, the work presented in this thesis serves to showcase how one can use ISS to create comprehensive atlases, but more importantly, move beyond cellular atlases towards understanding disease.