Fig. 1. Dystroglycan gene expression during early Xenopus laevis retinogenesis. Panels A�D are sections of in situ whole-mount hybridized embryos from stage 19
to stage 28. (A) Tangential section showing DG mRNA expression localized in the anterior neural plate at stage 19. (B�D) Transversal sections showing X-DG
mRNA expression at stage 24 (B), at stage 26 (C) and at stage 28 (D) in the optic stalk, in the presumptive pigmented epithelium and in the lens vesicle (C, D).
Panels E� IV are cryostat sections showing. X-DG immunolocalization. (E) X-DG detection at stage 28 in the neural tube and in the retina at the level of the
presumptive pigmented epithelium, the optic stalk and all around the sensorial layer of the ectoderm. (F, G) High magnification of retina sections at stage 32 showing
the presence of X-DG in the endfeet of retinal precursors in the ciliary marginal zone and in the inner limiting membrane at the vitreal border of the retina (purple
arrowheads in panel G). Green arrowheads indicate X-DG detection in the lens vesicle. (H� IV) Transverse sections of Xenopus retina at stage 45 showing X-DG
immunoreactivity at the level of retinal pigmented epithelium (white arrowhead), inner limiting membrane (purple arrowhead) and lens epithelium (green
arrowhead). Punctuated distribution of X-DG in the OPL is evidenced at high magnification in panels I, IV (same section, red detection) between the two nuclear
layers, highlighted in panel IV by Hoechst staining. Abbreviations: anp, anterior neural plate; OS, optic stalk; ppe, presumptive pigmented epithelium; nr, neural
retina; nrp, neural progenitors; lv, lens vesicle; d, dorsal; v, ventral; nt, neural tube; sle, sensorial layer of ectoderm; cmz, ciliary marginal zone; gcl, ganglion cell
layer; inl, inner nuclear layer; onl, outer nuclear layer; opl, outer plexiform layer; ipl, inner plexiform layer; ilm, inner limiting membrane.
Fig. 2. Results of DG loss-of-function experiments. (A) Stage 45 Xenopus
embryo injected at four cells stage with 8 ng of Mo-X-DG plus 250 pg of GFP
cRNA. The injected side is indicated by the green staining; white brackets
highlight reduction of the eye size in the injected side (left) compared to the
uninjected, control side (right). (B, C) Transversal section of a stage 32 injected
embryo in which specific DG immunostaining (in red) is lost in neural tube (nt)
and eye (E) of the Mo-X-DG-injected side and in some cells of the other side,
where diffusion of the injected material sporadically happens. Fluorescent
green labeling shows the injected side of the embryo in panel C.
Fig. 3. Expression of eye markers in X-DG morphants. Xotx2 (A, C, E, EV), Xrx1 (B, D, F, FV) and b-crystallin (G, GV) mRNA expression was detected by BMP purple
(blue staining) in Mo-X-DG-injected embryos at stage 14 (A, B), 25 (C, D) and 28 (E�GV). Nuclear beta-gal staining (red) was used to trace the side of injection. Panels
A�D show both injected and uninjected sides in frontal views. Panels E�GV show lateral view of uninjected (E, F, G) and injected (EV, FV, GV) sides, respectively. No
changes in the expression domains of Xotx2 (A) and Xrx1 (B) are detectable between the two sides at stage 14. White brackets highlight the decrease of Xotx2 (C) and
Xrx1 (D) expression in the eye of the injected sides at stage 25. Decrease is evident also at stage 28 (compare panel E to EV and F to FV); beta-crystallin expression is also
dramatically reduced in the injected side (compare panel G to GV). Mo-X-DG injection completely abolishes also Xotx2 rostral expression domain in panels E to EV.
Fig. 4. Cryostat sections of Xenopus retina immunolabeled with polyclonal antibody against Lam1 (A�D and F: green labeling); monoclonal antibody against h1-
integrin (A�G: red labeling) and polyclonal antibody against aPKC (H�K, red labeling). Control side (A, D, E, H, I) and injected side (B, C, F, G, J, K) of retinas at
stage 32. The injected side in panels J and K was traced by GFP. Laminin and integrin were co-detected (yellow signal) in the Bruch�s membrane (BrM) at the bases
of pigmented epithelium (PE), at the border of vitreal cavity in the inner limiting membrane (ILM) and in the lamina surrounding the lens. Severe disorganization of
these laminae was detected in the injected sides of morphants (arrows in panels B, C, F). Integrin staining, which showed a radial-like pattern in the control sides (A,
E) was affected in the injected sides of morphants (B, C, G). High magnifications (D�G) highlight the histology of the retina at its most marginal side near the lens.
Localization of aPKC at the apical side of the retinal neural precursors (nrp), underneath the pigmented epithelium (PE) was visible either in control (H, I) or in Moinjected
eye (J, K). At higher magnification, the aPKC staining of retinas in the control (I) and injected (K) sides was comparable.
Fig. 5. Retinal delayering in DG morphants. (A�E) control sides; (AV�EV) Mo-X-DG-injected sides. Expression of markers for specific cell types in retinal sections
from stage 45 embryos: (A, AV) Xotx5 mRNA; (B, BV) Xotx2 mRNA; (C, CV) hermes mRNA; (D, DV) N-tubulin. (E, EV) Hoechst nuclear staining of sections shown in
panels D and DV. In panel A, blue and white arrowheads indicate photoreceptors and bipolar cells. Green arrowheads show rosette-like structures. Red arrowhead
indicates ectopic hermes expression in panel CV. Yellow arrowheads indicate ectopic nervous fibers in panel DV and the corresponding position in panel EV.
Fig. 6. Effects of the unilateral injection of Mo-X-DG (8 ng) at stage 4 cells on
the apoptosis and cell proliferation of the eye. (A�F) Eye sections of stage 32
embryos. Pe: pigmented epithelium. Detection of GFP, which was used as
tracer, is shown in panels B, D, F. Panels A�D show BrdU immunodetection
(red nuclei) in the control side (A, C) and Mo-injected side (Mo, B, D) of the
embryo. Hoechst counterstaining labels all nuclei in panels C, D. Panels E and
F show TUNEL detection of apoptotic nuclei (brown staining) in control (E)
and Mo-X-DG-injected side (F), respectively. Sections of Mo-X-DG injected
eyes always displayed a significantly higher number of TUNEL-positive cells
than sections of uninjected, contralateral eyes. Eyes injected with control-Mo
had the same number of TUNEL-positive cells than uninjected eyes (not
shown). The numbers of BrdU-positive nuclei in control and Mo-X-DGinjected
(Mo) retinas, expressed as percent over the total number of cells
analyzed (control: n = 550, Mo: n = 514), are compared in panel G. The
average number of TUNEL-positive cells per eye section (n = 120 sections) in
control and Mo-X-DG-injected retinas is compared in panel H. Bars indicate
SE in panel G and SEM in panel H. Differences between control and Mo are
significant in panel G (t test: P = 0.008904109) and in panel H (t test: P =