Development of a novel in vitro model of the cornea to investigate cell dynamics within the mammalian limbal niche

Abstract

The cornea, the transparent tissue overlying the front of the eye, is vital for protecting the delicate internal ocular structures and transmitting light onto the retina.

Our understanding of how the cornea functions and develops, as well as its response to damage, has been derived historically from non-human studies, but more recently, this has transitioned to in vitro equivalents. However, most equivalents poorly recapitulate the cornea, and in particular the cornea’s tissue-specific adult stem cell niche, the limbus.

This thesis therefore begins with a thorough characterisation of the rabbit limbal niche in vivo, with the findings then contextualising and further developing a novel 3D microscale model of the rabbit cornea. The corneal model is based on a hybrid of the hanging drop spheroid and organoid techniques and consists of a core of primary stromal fibroblasts surrounded by an envelope of primary limbal epithelial cells. 3D culture enables the core to produce fibrillar collagen, providing early indications of physiological function. Interactions between the core and epithelium also enable the epithelial population to self-stratify, forming a complex layer which extensively recapitulates the limbal niche and which matches several in vivo observations as made earlier.

The highest quality corneal models can be characterised by a CK12+3- and CK14+ basal epithelial layer, and recapitulation of the vimentin+ and Pax6(low) basal epithelial phenotype, as this thesis has also identified in vivo. This phenotype has not been reported in vivo before and may represent a crucial species-specific difference between the rabbit and human.

This model could have a range of end use applications, given its detailed level of complexity. Whilst preliminary work to assess how it responds to ocular irritants was ineffective, the method which was used does require optimisation before any further conclusions can be made. However, the model’s detailed complexity may also facilitate investigations beyond ocular irritancy, for example to study basic biological questions, limbal stem cell regulation, or as a tool to assess epithelial quality within clinical contexts.