Dorsoventral patterning of the Xenopus eye involves differential temporal changes in the response of optic stalk and retinal progenitors to Hh signalling

Abstract

Background: Hedgehog (Hh) signals are instrumental to the dorsoventral patterning of the vertebrate eye, promoting optic stalk and ventral retinal fates and repressing dorsal retinal identity. There has been limited analysis, however, of the critical window during which Hh molecules control eye polarity and of the temporal changes in the responsiveness of eye cells to these signals.

Results: In this study, we used pharmacological and molecular tools to perform stage-specific manipulations of Hh signalling in the developing Xenopus eye. In gain-of-function experiments, most of the eye was sensitive to ventralization when the Hh pathway was activated starting from gastrula/neurula stages. During optic vesicle stages, the dorsal eye became resistant to Hh-dependent ventralization, but this pathway could partially upregulate optic stalk markers within the retina. In loss-of-function assays, inhibition of Hh signalling starting from neurula stages caused expansion of the dorsal retina at the expense of the ventral retina and the optic stalk, while the effects of Hh inhibition during optic vesicle stages were limited to the reduction of optic stalk size.

Conclusions: Our results suggest the existence of two competence windows during which the Hh pathway differentially controls patterning of the eye region. In the first window, between the neural plate and the optic vesicle stages, Hh signalling exerts a global influence on eye dorsoventral polarity, contributing to the specification of optic stalk, ventral retina and dorsal retinal domains. In the second window, between optic vesicle and optic cup stages, this pathway plays a more limited role in the maintenance of the optic stalk domain. We speculate that this temporal regulation is important to coordinate dorsoventral patterning with morphogenesis and differentiation processes during eye development.

Figure 1

PMP treatments cause stage dependent effects on eye DV polarity. (A) Lateral views of heads of st. 33 embryos treated with DMSO (mock) or 300 to 600 μM PMP from the indicated stages and hybridized with probes for Pax2, Vax1b, Vax2 or Tbx3. In mock-treated embryos, wild-type expression patterns of these genes are detectable: Pax2 and Vax1b staining is restricted to the OS, Vax2-positive region covers the OS and VR, Tbx3 expression identifies the DR. PMP treatments affect gene expression domains to various degrees depending on the stage of delivery (st. 8 to 10, blastula-early gastrula; st. 12.5 to 14, late gastrula-early neurula; st. 19 to 22, late neurula-early optic vesicle; st. 26, mid optic vesicle). For Pax2 and Vax1b, embryos were grouped into 0 to 3 scores (numbers at the bottom right corner of each image) as explained in the ‘Results’ section and representative eyes for each score group are shown. See text for details. The broken yellow circles highlight the eye region. Scale bar, 200 μm. (B) Quantification of the percentages of embryos stained for Pax2 or Vax1b with 0 to 3 scores in each treatment condition. Embryos stained for Vax2 or Tbx3 were grouped according to the DV extent of Vax2/Tbx3 expression domain (more or less than 90% of the eye for Vax2; more or less than 10% of the eye for Tbx3). The percentages of embryos with strong eye reductions are also indicated (S, small eyes). The number of experiments performed for each probe and treatment condition is indicated on top of the corresponding histogram bar. At least 20 eyes/10 embryos were analysed for each experiment. (C) Histological sections of eyes of st. 33 embryos treated as in (A) and (B), and hybridized with the indicated probes, confirming stage dependent alterations in the expression domains of Pax2, Vax1b, Vax2 and Tbx3 as detected in whole mount views. Scale bar, 100 μm.

Figure 2

Quantification of gene expression changes following PMP treatments. Real-time PCR quantification of gene expression in st. 33 dissected heads following treatments with 600 μM PMP or DMSO from the indicated stages, shown as the mean ratio between PMP and DMSO conditions in four independent experiments. Error bars show standard deviations. *P < 0.05; **P < 0.01; ***P < 0.001; ns, non-significant (P ≥ 0.05) according to two-tailed Student’s t-test.

Figure 3

Grafts of ShhC25II-soaked beads reproduce the stage-dependent effects of PMP treatments on eye DV patterning. (A) Lateral views of heads of st. 33 embryos that received a graft of control or ShhC25II-soaked beads next to the optic vesicle at the indicated stages and hybridized with probes for Pax2, Vax1b or Vax2. Compared to controls, embryos grafted with ShhC25II beads show stage-dependent increase in the expression domains of ventral eye genes. Scale bar, 100 μm. (B) Quantification of the percentages of embryos with different effects on gene expression domains or eye reductions (S) in each treatment condition. The number of experiments performed for each probe and treatment condition is indicated on top of the corresponding histogram bar. (C) Histological sections of eyes of st. 33 embryos treated as in (A) and (B) and hybridized with the indicated probes, confirming stage dependent effects on DV eye patterning as detected in whole mount views. Triangles point to ectopic Pax2 expression in the dorsal marginal zone and arrows to expanded ventral expression domains of Pax2, Vax1b and Vax2, in embryos grafted ShhC25II beads. Stars indicate the position of the beads. Scale bar, 100 μm.

Figure 4

Overexpression of VP16-Gli1-GR causes stage dependent effects on eye DV polarity. (A) Lateral views of heads of st. 33 embryos that were unilaterally injected with 250 pg of VP16-Gli1-GR mRNA at the eight-cell stage, treated with dex from the indicated stages and hybridized with probes for Pax2, Vax1b, Vax2 and Tbx3. Compared to the control side (uninj.), stage-dependent alterations in gene expression domains are detectable on the injected side. Light-blue β-gal staining identifies the injected side. Scale bar, 200 μm. (B) Quantification of the percentages of embryos with different effects on gene expression domains or eye reductions (S) in each treatment condition. The number of experiments performed for each probe and treatment condition is indicated on top of the corresponding histogram bar. (C) Histological sections of eyes of st. 33 embryos treated as in (A) and (B), and hybridized with the indicated probes, confirming stage dependent gene expression changes as detected in whole mount views. Scale bar, 100 μm.

Figure 5

Quantification of gene expression changes following VP16-Gli1-GR overexpression. Real-time PCR quantification of gene expression in st. 33 heads dissected from controls or from embryos injected with VP16-Gli1-GR mRNA and treated with dex from the indicated stages. Results are shown as the mean ratio between VP16-Gli1-GR-injected and control embryos in six independent experiments. Error bars show standard deviations. *P < 0.05; **P < 0.01; ns, non-significant (P ≥ 0.05) according to two-tailed Student’s t-test.

Figure 6

CPM treatments cause stage dependent effects on eye DV polarity. (A) Lateral views of heads of st. 33 embryos treated with ethanol (mock) or 100 μM CPM from the indicated stages and hybridized with probes for Pax2, Vax1b, Vax2 or Tbx3. Compared to controls, CPM-treated embryos show stage-dependent changes in gene expression domains along the anteroposterior (AP; Pax2, Vax1b) or the DV (Vax2, Tbx3) axes of the eye. Scale bar, 200 μm. (B) Quantification of the mean AP (Pax2, Vax1b) or DV (Vax2, Tbx3) width/height of gene expression domains, normalized to total width/height of the eye, in the eyes of embryos treated as in (A). The number of eyes analysed for each probe and treatment condition is indicated within the corresponding histogram bar. Error bars show standard deviations. *P < 0.05; ns, non-significant (P ≥ 0.05) according to two-tailed Student’s t-test. (C) Histological sections of eyes of st. 33 embryos treated with ethanol or CPM from st. 13 and hybridized with the indicated probes. CPM-treated eyes show a reduction of Pax2, Vax1b and Vax2 expression domains along the eye proximodistal axis. Yellow brackets highlight the proximodistal extent of the whole ventral eye region. Scale bar, 100 μm.

Figure 7

Proposed model of the stage dependent effects of Hh signalling in the DV patterning of the Xenopus eye. (A) During gastrula/neurula stages of development, Hh signalling controls the specification of both the OS and VR domains by promoting expression of OS (Pax2, Vax1b) and VR (Vax2) genes and repressing expression of DR genes (Tbx3, Tbx5). (B) During optic vesicle stages, the expression domains of Vax2 and Tbx3/5 become less dependent on Hh signalling (dashed lines), which is required to maintain Pax2 and Vax1b expression and proper OS development. At both stages, activation of Hh signalling is likely to depend on the Shh ligand, acting through inhibition of Gli repressor proteins (Gli-R) and increase of Gli activator proteins (Gli-A) [40]. Vax2 was previously shown to self-activate its own expression at optic vesicle stages [29].