Patterning the Vertebrate Retina - Molecular Mechanisms Underlying the Development of Retinal High Acuity Area

Abstract

The vertebrate retina is not a uniform sheet of neurons but is patterned into distinct domains, with the central region being uniquely specialized. The ability to see in fine detail relies on this small region specialized for high-acuity vision. Known as the high-acuity area (HAA), this domain is defined by several striking features: a high density of retinal ganglion cells, which serve as the output neurons of the retina; an enrichment of cone photoreceptors, which mediate daylight and color vision; and a complete absence of rod photoreceptors, which instead support dim-light vision. This architecture allows signals from a small number of cones to be transmitted directly to individual ganglion cells, reducing convergence and enabling exceptional spatial resolution. Together, these specializations endow the HAA with its critical role in enabling sharp central vision. The human HAA, known as the fovea, is distinguished by its characteristic pit-shaped morphology and by being a rod-free, cone-dense region that supports high-resolution tasks such as reading and facial recognition. However, very little is known about the molecular and developmental mechanisms that give rise to this specialized region. This gap in knowledge is compounded by the fact that commonly used mammalian models, such as mouse and rat, do not possess an HAA at all, thus, limiting their use for developmental and disease studies. Interestingly, certain birds possess an HAA that shares many structural and functional features of the primate fovea. Unlike in humans, the chick retina offers ready access to embryonic tissue and is highly amenable to experimental perturbation, making it a powerful system for probing the developmental logic of high-acuity specialization. In this dissertation, we use the chick retina to investigate how the HAA is generated, to define the molecular patterning events that localize and specify this specialized territory, and to understand how these principles extend across species. First, we characterized the morphological development and cellular composition of the chick HAA, defining when specialized photoreceptor and ganglion cell patterns emerge and how they differ from peripheral retina. Using single molecule fluorescent in situ hybridization (smFISH) in combination with classical histological methods, we identified Fgf8 as a robust molecular marker of the HAA that persists throughout embryonic development. We showed that the ganglion cell layer (GCL) was consistently thickest at the HAA beginning during the period of neurogenesis. This pattern indicates that elevated production of retinal ganglion cells, rather than selective survival or death, accounts for their enrichment in this region. In contrast, rods photoreceptors were entirely absent from the HAA, and analysis of apoptotic markers showed no evidence of rod elimination, suggesting instead that rods fail to be generated there. Together, these findings show that distinct developmental processes (differential neurogenesis, delayed cone accumulation, and restricted rod genesis) combine to build the specialized architecture of the HAA. We next asked how positional information is encoded to localize the HAA within the naso-central retina. To do so, we developed an integrated framework combining multiplexed gene expression imaging with single-cell transcriptomics, enabling quantitative 2D reconstruction of molecular expression domains. This approach revealed sharp boundaries in signaling pathways, identified novel HAA-enriched candidates, and provided a reproducible spatial atlas of the developing chick retina. By anchoring spatial reconstructions to experimental landmarks such as the Fgf8 expression domain, we demonstrated how expression domains of retinoic acid-Fgf8 signaling pathway components correlate with classical dorso-ventral and naso-temporal patterning axes to establish a unique retinal territory. Finally, we extended this framework across vertebrate species. By comparing chick, human, and mouse single-cell datasets, we identified conserved axis-based programs as well as species-specific spatial patterns. Both chick and human retinas contained distinct domains consistent with an HAA, whereas mouse retinas resolved primarily into broad DV and NT axes, consistent with their lack of an HAA. Additional comparisons across birds, and reptiles highlighted both conserved and divergent molecular strategies underlying the evolution of high-acuity vision. Together, our work provides an integrated cellular and molecular framework for HAA development. We investigated how gene expression boundaries, and distinct developmental programs potentially converged to create a specialized retinal territory and further used comparative approaches to understand these findings in an evolutionary context. More broadly, our integration of multiplexed imaging with single-cell transcriptomics offers a generalizable strategy for reconstructing spatial patterning logic in developing tissues. This study not only advances our understanding of human foveal development but also offers new tools to study the origins of tissue specialization across systems and species.