The central nervous system is a complex, yet well-organised, often laminated, tissue. This robust organisation is evident in the architecture of the retina: consisting of 5 different neuronal types organised into distinct layers: Retinal Ganglion Cell (RGC), Amacrine Cell (AC), Bipolar Cell (BP), Horizontal Cell (HC) and Photoreceptor cell (PR) layers. This remarkable organisation is evolutionarily conserved in vertebrates, yet little is known about the mechanisms by which these cells form the correct layers. Live imaging has revealed overlapping periods of birth and extensive inter-digitation followed by cells sorting out into their appropriate positions, suggesting cell-cell interactions are important. To investigate possible cellular and molecular mechanisms responsible for the establishment of the tissue architecture I developed an organoid culture system for zebrafish retinal cells.

To identify the cells in culture I used a Spectrum of Fates fish line which is a multiply transgenic line in which each retinal cell type can be identified based on expression of a combination of fluorescently tagged cell fate markers. The development of the protocol by which I cultured the cells and observed their cell-cell interactions involved establishing the best methods to dissociate and culture zebrafish retinal cells in a non-adhesive environment, then imaging the resulting reaggregates to examine the position of the different retinal cell types. By doing this I observed their inherent self-organising properties, in the absence of extrinsic cues or scaffolds. These cells appeared to be arranged in an inside-out layering, although all cell types are layered in the same relative order as they are in vivo.

To analyse the organization in these aggregates I developed a Matlab script in collaboration with Leila Muresan which analyses the relative positioning of cells in concentric rings from the periphery to the centre of the aggregates according to the cell fate-tagged fluorescent markers. The script then fits this data as an empirical cumulative distribution function for different groups of cells to determine how spatially distinct populations of cells are. This gave me my measure of organisation.

I then investigated the cell-cell interactions involved in this self-organisation by genetically or pharmacologically removing individual cell types and assaying the resulting organisation of the reaggregated, cell-type deficient, retinal organoids. I revealed that Müller Glia are important for retinal cell self-organisation. I also investigated the role of Retinal Pigment Epithelial (RPE) cells and Retinal Ganglion Cells and found they had no impact on the ability of the remaining cell types to organize. I began to investigate the role of Amacrine Cells but found that retinas void of ACs were susceptible to disaggregating in our dissection setup, preventing me from collecting the material needed for culture. I also investigated the role of candidate molecules in this system and revealed that R-Cognin is critical for retinal cells to reaggregate.

Not only can I remove cells or molecules from the system, but I show how it can also be manipulated to replace molecules of interest such as laminin, by coating beads with the substance of choice and placing it amongst the cells to see if their organisational behaviour is affected. In summary, I have developed a system which provides a simple and easy platform to manipulate in various ways to help us potentially reveal some of the important players in neuronal patterning.