Zebrafish Research at UCL

The eyes, which are part of the central nervous system, enable organisms to visually perceive their surroundings. During embryogenesis, the eyes originate as outpocketings of the brain; our work aims to elucidate the key stages of eye development including induction of the eye field, morphogenesis of the optic vesicles, choroid fissure closure, and differentiation of retinal neurons. In addition to deepening our basic understanding of eye development, our studies provide information about the aetiology of eye diseases and hereditary conditions that include coloboma and retinal degenerations.

Below, we describe our main projects; you can also download our publications and if you don't have a scientific background, you can read summaries of our papers written for a general audience. Our lab members also describe their projects on their personal pages.

A-B) During early embryogenesis all of the cells destined to form the eyes are positioned as a single field of cells spanning the midline of the neural plate (A; ef, eyefield). As development proceeds, midline neuroepithelial tissue moves anteriorly in the neural plate (B; hy, prospective hypothalamus). This, combined with the movement of the eye field cells themselves helps to drive cells into the prospective left and right optic vesicles while the cells closest to the midline contribute to the optic stalk (C; os, optic stalk). As the optic vesicle evaginates, it contacts the overlying epidermis and induces the formation of the lens. Then the optic vesicle invaginates with the tissue closest to the lens becoming the neural retina (C; nr, neural retina) and the more distant tissue becoming pigmented epithelium (C; pe, pigmented epithelium). The neural retina begins as a pseudostratified epithelium that matures into a highly organized laminated structure (D) composed of different classes of neurons and one type of glia. The optic stalk provides a path for the retinal axons as they exit the retina to form the optic nerve (D; on, optic nerve).

OUR CURRENT PROJECTS
Eye Specification & Morphogenesis
The induction of the eye field within the anterior neural plate is influenced by a number of signalling pathways and the Wnts are probably the most studied signals during this process (Wilson and Houart, Developmental Cell 6, 167-181., 2004). We have shown that too much Wnt/β-catenin pathway activity in the anterior neural plate suppresses eye formation converting eye field cells into more posterior diencephalic fates (Heisenberg et al., Genes Dev. 15, 1427-1434 (2001); Houart et al., Neuron 35, 255-65 (2002)).

Our most recently published work on eye specification (Cavodeassi et al., Neuron 47, 43-56 (2005)) revealed an essential role for Wnt11 and its likely receptor Fz5 in the coordination of eye field specification with morphogenesis (Fig 1). Manipulation of Wnt11 activity leads to defects in the morphogenetic movements that split the eye field in two domains to give rise to the optic vesicles. Our work to date suggests that these defects may be due to disruption of cell-polarity and cell-adhesion properties of the eye field cells. Importantly, these findings uncover a novel link between the Wnt signalling pathway, fate determination, and morphogenesis of the eye providing us with a working model to test how Wnts contribute to specification and morphogenesis of the eyes.

Figure 1
The combination of two Wnt signals, Wnt8b and Wnt11, is essential for specification of the eye field. Wnt8b is required for diencephalic specification and represses eye field fate. Wnt11 helps to segregate diencephalic and eye field domains by antagonising the role of Wnt8b in more anterior regions of the neural plate, and allowing the formation of the eye field in these regions.

In addition, Wnt11 function within the eye field is essential for the subsequent morphogenesis of this domain and optic vesicle evagination.

We are now extending these investigations by analyzing the behaviour and organization of individual cells in the eye field during the transition from specification to evagination. In addition, we are manipulating Wnt signalling as well as altering the adhesive properties of eye field cells to understand how cell-cell junctions contribute to evagination of the optic vesicles. A screen for Wnt pathway genes required for eye formation and maintenance is in progress in the lab, and this is sure to uncover new gene activities required for optic vesicle evagination.

After the optic cup forms (through invagination of the optic vesicles), the optic stalk and retina are remodelled so that the two ventral sides of the optic cup close around the optic nerve and blood vessels. Previously, our lab examined the developmental changes that occur in the optic stalk and ventral retina during eye development by investigating the regulation and function of three transcription factors (Pax2, Vax1, and Vax2) that are induced within the optic stalks (Macdonald et al., Development, 121, 3267-3278 (1995); 124, 2397-2408 (1997)). These studies suggested that pax2, vax1 and vax2 act downstream of Hh and Fgf signals during formation of the optic stalk as well as during later morphogenetic events in the ventral retina (Take-Uchi et al., Dev 130, 955-968 (2003)).

Currently, our interests in ventral retina morphogenesis are focused on choroid fissure closure. The choroid fissure is a transient opening on the ventral side of the optic cup which provides an entry site for newly forming blood vessels and an exit for newly born retinal axons. Failure of the choroid fissure to close during eye development results in ocular colobomas (Figure 3). Although work from our lab and others has identified a number of signalling pathways and transcription factors important for choroid fissure closure, how ventro-temporal and ventro-nasal cells move together to close the choroid fissure remains unknown. We are employing both forward and reverse genetic approaches as well as using advanced imaging techniques in living fish to elucidate the cellular mechanisms that underlie choroid fissure closure. We are also developing collaborations with human geneticists to help use our work in fish to understand human congenital eye malformations.

Figure 3
(A) Cartoon of a developing eye. The choroid fissure extends along the ventral part of the optic cup. Figure taken from Drews, U. Color Atlas of Embryology (Thieme Medical Publishers, New York, 1995). (B-C) Coloboma phenotype in humans (B, image source: unknown) and zebrafish (C), detected as an opening at the ventral side of the eye (arrows). Because of the rapid development of zebrafish, we can easily observe the morphogenetic movements that close the choroid fissure (D, click below to play movie).

Neurogenesis - links between proliferation and differentiation
While the optic cup is undergoing complex morphogenetic movements during development, neurogenesis in the retina begins at the interface between the optic stalk and the neural retina (Masai et al., Neuron 27, 251-263 (2000)). From this position, waves of neurogenesis subsequently spread across the nasal and temporal retina generating all the various retinal neurons. This process can be nicely visualised in transgenic lines in which the neurons that are being generated are labelled with GFP. Here are two such examples.

Presently, we are analysing several mutants that exhibit defects is cell cycle progression and retinal differentiation. Our studies of the flotte lotte (flo) mutant, (Figure 4), highlight the importance of coordinating intrinsic (cell cycle progression) and extrinsic factors (secreted signals) during retinal neurogenesis. Another mutant that illustrates the importance of cell cycle progression on formation of a functional retina is eisspalte (ele). re_get_image (8,5); The eyes of this mutant exhibit a variety of deficiencies resulting from cell cycle defects including reduced neurogenesis, aberrant axonal projection, and delayed choroid fissure closure. We have now cloned this mutation and find that it disrupts a gene involved in processing cell-cycle regulated RNAs.

As well as studying these mutants, we are also screening for new mutants with specific defects in proliferation, differentiation, and morphogenesis in collaboration with the current screen team.

Figure 4
(A) flo gene expression in the proliferating cells of the retinal stem cell niche (arrows).

(B-C) flo mutant cells cycle more slowly. Embryos injected with BrdU (red) were then stained with antiserum to PH3 (green) to assess cell cycle progression. Not only do flo eyes contain more PH3+ cells, they contain cells that are only PH3+, indicating that they have taken longer than WT to progress from S phase to mitosis.

(D-F) flo differentiation defects are rescued by a WT environment. Transplantation of flo cells into a WT eye or WT cells into a flo eye shows that flo cell neuronal differentiation, which is normally severely effected, occurs normally when WT extrinsic signals are present.

In addition to studying early development, we have started to analyse stem cells and proliferation during larval stages. For this work, we are studying both stem cells in the eye and in the central target of the retinal projections, the midbrain tectum. There is a separate web page describing this work.

FUNDING AND COLLABORATORS
We have ongoing collaborations with Brian Link, Michael Brand, Nicky Ragge, Guiseppe Lupo and Alexander Picker. Our current work is/has been supported by a project grant from the MRC, Fellowship support from Telethon (GG), Damon Runyon Cancer Research Foundation (KLC) and European Community, and all of our projects are underpinned by Programme grant support from the Wellcome Trust.