Skip to main content

Cdon-Hh Interaction: functional in vivo analysis in zebrafish

Periodic Reporting for period 1 - CHI-ZEF (Cdon-Hh Interaction: functional in vivo analysis in zebrafish)

Reporting period: 2017-07-01 to 2019-06-30

Patterning of the vertebrate optic vesicle into proximal/optic stalk and distal/neural retina involves midline-derived Hh signalling, which promotes stalk specification. In the absence of Hh signalling, the stalk is not specified, forming a single cyclopic eye. Work from the hosting laboratory has shown that the cell adhesion molecule Cdon presents a complementary expression pattern to the canonical Hh receptor Ptc and acts as a negative Hh signalling regulator during the formation of the eye in zebrafish. At the neural retina/optic stalk border Cdon binds Hh, serving as a decoy receptor to protect the neural retina from Hh activity, likely preventing its diffusion and limiting its long-range signalling. How Cdon controls Hh dispersion shaping the gradient responsible for the correct P-D patterning of the eye remains unknown. Furthermore and although cell-to-cell communication mediated by morphogen signalling has been widely studied, there is surprisingly little information about the mechanistic insights of its spatial-temporal occurrence. This project attempted to explore these mechanisms during early stages of eye development, with the aim of providing a better understanding of Hh dispersion in the nervous system. In particular, we sought to focus on the impact that membrane bound molecules have in shaping Hh gradients and its consequence on tissue morphogenesis. Using zebrafish, CRISPR/Cas9 technology and advanced imaging we attempted generate cutting-edge tools to follow in vivo Hh/Cdon interaction during zebrafish eye formation. We envisioned that a better knowledge of tissue morphogenesis would immediately revert into the field of regenerative medicine, which attempts to recreate tissue and organs in non-developmental contexts, a goal that requires a deep understanding of developmental principles. The successful development of the project should have also provided key technical developments and useful information for clinical research towards prevention and treatment of congenital ocular malformations. Although not lethal, eye pathologies are severely impairing diseases with an important socio-economical charge for the society. Their prevention and treatment would thus represent an important step forward in the life of many people.
To investigate how loss of cdon affects the patterning of the eye we have generated cdon mutant lines using the CRISPR/Cas9 technology. The general morphology of the mutants appeared normal and no evident eye defects were found. However, cdon-/- mutants showed a reduction in the expression levels of pax2, a gene essential for patterning of the eye. Despite this decrease, the expression domain of the optic stalk was expanded, indicating that Cdon is required to limit the distal region of the eye, well in line with previous work from the host lab. Furthermore, the mutant embryos were more sensitive to Hh signaling inhibitors and showed changes in the expression levels of gli genes, Hh downstream targets, suggesting that the absence of Cdon impairs Hh signaling. To test if compensation by related genes was responsible for the mild phenotype observed in the generated cdon-/- mutants, we carried out qPCR and ISH analysis of boc and gas1 in cdon-/- embryos and injected guide RNAs against both genes in cdon-/- background. Crossing cdon-/- and boc+/- (available at the host lab) mutant lines was also performed. There was no up-regulation or changes in the expression patterns of boc in cdon-/- mutants and boc inactivation had no consequences on the cdon-/- phenotype. In contrast, the expression of gas1 was up-regulated and with a wider expression domain but we did not succeeded in proving that this up-regulation could compensate cdon function because the selected guide RNAs against gas1 were rather inefficient and did not lead to gene inactivation. To visualize Shh protein in vivo and determine the role that Cdon has on the dispersion of Hh protein we planned to generate a zebrafish shh:kaede knock-in line. Despite the use of different strategies and tools, we did not succeeded in generating the corresponding zebrafish line. To circumvent this problem, we turned to the medaka fish. We attempted to generate a shh:gfp knock-in using the reported biotin 5’ ends donor vector modification strategy. We are currently genotyping the injected generation to identify founders. In parallel, we are also generating gas1 and cdon mutants in medaka fish to pursue our aim. To visualize Cdon-Hh interaction, we planned to use BiFC assay and to outcross shh:kaede line to cdon-/- mutant. We have generated Venus-Shh and Venus-Cdon fusion proteins and evaluated VN-Shh – VN-Cdon, VN-Shh – VC-Cdon, VC-Shh – VN-Cdon and VC-Shh – VC-cdon interactions. Unfortunately, we could not establish the best combination of Shh/Venus and Cdon/Venus fusion proteins nor generate the shh:kaede knock-in in zebrafish, as already mentioned. We hope that the use of the medaka fish lines will provide the right tools to address this question.
As already mentioned in this report, we did not succeed in generating the planned knock-in lines in zebrafish. To overcome these technical problems we turned to the medaka fish to generate a shh:gfp knock-in. GFP expression within the shh expression domain was observed in the injected generation and the gfp gene and shh-gfp fusion products were amplified in the positive embryos, providing evidence that the approach might provide the desired results. In parallel, and to achieve the main goal of this proposal we also generated the medaka cdon and gas1 mutants. We are currently genotyping the injected generation to identify founders for all lines. We hope that our goal will be finally feasible in a different model system, the medaka, as we have explained in our report. Although not initially planned, the development of this proposal lead to the unexpected demonstration of an Hh unrelated role of Cdon in kidney development. Mutations in children affected by kidney defects (in cases associated to eye malformations) have been found in humans, suggesting that Cdon function might be relevant to renal development. We found that the potentially pathogenic CDON variants impair matrix-cell adhesion as well as cell-to-cell contacts with a mechanism that implicated alterations in β1-integrin and N-cadherin interactions. In parallel, we showed that cdon is expressed in the embryonic kidney of the zebrafish, the pronephros. Furthermore, truncation of cdon results in kidney with developmental abnormalities. By using ISH, immunofluorescence and confocal microscopy in the cldnb::gfp;cdon-/- line , we showed that the proximal and distal tubules of the pronephros had a decreased diameter, an altered cell polarity and cilia organization together with a reduced and abnormally shaped lumen of the glomerulus and tubules. Analysis of kidney morphology in cdon-/- adult fishes with H&E staining showed an abnormal kidney with cysts formation and tubules dilation. These data demonstrate a novel and unexpected role of Cdon in zebrafish and likely human kidney formation. This finding provides novel basic information and a valuable candidate for congenital kidney malformations for clinicians.