plxna2 morphants exhibit small eye vesicles. (A-D) RNA in situ hybridization for plxna2 viewed in lateral wholemounts (A,B), and in sagittal (C) and transverse (D) sections. At 4-12 somites, plxna2 mRNA is in the ventral, but not dorsal, eye (e) vesicle (outlined in white), as well as in the mesenchyme (m), somites (s) and brain (br). op, olfactory placode. Red dotted line separates the eye vesicle into approximate dorsal and ventral domains. (E) Transverse (t) and coronal (c) planes of section. (F) RT-PCR indicates the expected mis-splicing of plxna2 mRNA in the morphants. (G) A dorsal view of the events of eye vesicle formation. (H-O) Dorsal views of whole-mount RNA in situ hybridization of control (H,J,L,N) and plxna2 morphant (I,K,M,O) embryos. rx3 (H,I), six3 (J,K) and meis1 (L,M) labeling. Arrows indicate the hypothalamus; the arrowhead marks meis1+ cells that bridge the eye vesicles in the plxna2 morphant. (N,O) nkx2.1 labeling of hypothalamus. (P,Q) Quantitation of the length (P) and area (Q) of the meis1+ domain. n=2. Error bars indicate s.e.m. ***P<0.001 unpaired Student′s t-test.

Plxna2 is required for eye tissue cohesion. (A-F) Dorsal confocal images of (A,B) 20- and (E,F) 12-somite rx3:GFP transgenic eye vesicles (e) in control (A,E) and plxna2 morphant (B,F). Scale bar: 50μm in A,B; 100μm in E,F. GFP+ cells are seen ectopic to (arrowheads), emerging from (arrows) and bridging (red arrow, F) the eye vesicles. hy, hypothalamus. Schematics highlight ectopic GFP+ cells in morphants (D), but not in control (C). (G,H) Mean rx3:GFP eye vesicle area (dorsal) (G) and the numbers of ectopic eye cells (H). n=3. Error bars indicate s.e.m. ***P<0.001 unpaired Student′s t-test. (I,J) Dorsal views of 14-somite rx3:GFP control (I) and ATG plxna2 morphant (J). (K-N) Dorsosagittal views of 14-somite rx3:GFP control (K) and e3i3 plxna2 morphant (L), and two embryos co-injected with human PLXNA2 mRNA and e3i3 plxna2 MO (M,N). (O) Quantitation of the dorsosagittal GFP+ eye vesicle area in control, e3i3 plxna2 morphants and e3i3 morphants injected with human PLXNA2 mRNA. The latter have significantly larger eye vesicles than plxna2 morphants, although their eyes are still smaller than control (P<0.001) and some ectopic cells are evident. n=3. Error bars indicate s.e.m., ***P<0.001, one-way ANOVA, Newman-Keuls test.

Retinal progenitors leave the eye vesicle with Plxna2 knockdown. Frames from time-lapse confocal imaging of control (A,C) and plxna2 morphant (B,D) rx3:GFP eyes viewed dorsally between 10 and 12 somites (A,B) and laterally between 8 and 14 somites (C,D). In the plxna2 morphant the eye field separates, but some cells leave the eye vesicle (arrowhead), whereas others fail to enter the eye vesicle (*). (D2,D3) Higher magnification view of the areas boxed in D, showing eye cells moving out of the eye (D2, arrows and arrowheads) or into the eye vesicle ventricle (ve) (D3, arrowheads). hy, hypothalamus.

plxna2 morphant cells leave the eye epithelium and undergo apoptosis. (A-D) Lateral whole-mount view (A,B) and coronal sections (C,D) of meis1 label of the eye vesicle (e) of a control (C) and a plxna2 morphant (A,B,D). Ectopic GFP immunoreactive cells (arrowheads) are indicated, as well as an ectopic meis1+ cell (arrow). (E-L) Transverse sections of 12-somite rx3:GFP control (E,G,I) or plxna2 morphant eye vesicles (F,H,J,K,L) immunolabeled with an anti-laminin antibody (E,F,L). The basal lamina (white arrows) is present in both sets of embryos, but is not complete (red arrows). GFP+ cells are found ectopic to the eye vesicle (arrowheads), and within (asterisk) the ventricle (outlined; ve), of the plxna2 morphant (K,L). (M,M2) A GFP+ cell in a plxna2 morphant sits at the vesicle edge (M) with a crenated nucleus (DAPI; M2). (N-Q) Anti-activated caspase 3 immunolabeling (N,P,Q) of transverse sections of control (N) and plxna2 morphant (O-Q; vesicle outlined by dots). Merge with GFP epifluorescence (N,Q). Apoptotic GFP+ cells (arrowheads; Q) are present in the eye vesicle, ventricle and mesenchyme (arrow; P) of the morphant. Scale bar: 20μm for A-L. br, brain; m, mesenchyme.

Plxna2 and Sema6a function together in eye cohesion. (A-D) RNA in situ hybridization for sema6a viewed in lateral wholemounts (A-C) and in a transverse section (D). At 4 (A), 6 (B) and 12 somites (C,D), sema6a is expressed in the brain (br), and both the dorsal and ventral vesicle (e), but not the mesenchyme (m; arrows). (E) Schematic of a transverse view of eye vesicle depicting sema6a (blue) throughout and plxna2 (red) restricted ventrally. (F) RT-PCR indicates a knockdown of sema6a mRNA in the morphants because of mis-splicing of exon 1. (G-J) Dorsal images of 12-somite rx3:GFP transgenic eyes for control (G), sema6a morphant (I), full-length zebrafish sema6 mRNA-injected embryo (H) and sema6a morphant rescued with sema6a mRNA (J). (K-M) Dorsal views of 12-somite control (K), plxna2 morphant (L) and plxna2 and sema6a double morphant (M) rx3:GFP eye vesicles. Ectopic GFP+ cells are seen around, and at the midline between, the eye vesicles (arrows). (N) Mean area of the 12-somite GFP+ eye vesicles of control, and embryos co-injected with the sema6a MO, with or without sema6a mRNA. n=3. Numbers of embryos in brackets. ***P<0.001, one-way ANOVA, Newman-Keuls test. (O,P) Percentage of ectopic GFP+ cells (O) and the GFP+ eye vesicle area (P) in control and morphant embryos. Error bars indicate s.e.m., numbers of embryos are in brackets. n=3. (O) Chi-square ***P<0.001; (P) ***P<0.001, **P<0.01 unpaired one-way ANOVA, Bonferroni′s test.

Sema6a/Plxna2 mediate eye tissue aggregation. (A) Explant assay. (B,C) Explants with mesenchyme removed treated with (C) or without (B) Sema6a-Fc. (D-F) Control (D), plxna2 (E) and sema6a (F) morphant explants without mesenchyme. (G,H) Numbers of GFP+ cells outside explants. (G) n=3, ***P<0.001, unpaired Student′s t-test; (H) n=2, ***P<0.001, one-way ANOVA, Dunnett′s test.

Sema6a/Plxna2 are necessary for proper regionalization of the eye vesicle. Twelve-somite control eyes (A,E,E2,K,M), and plxna2 (B,F,F2,L), sema6a (C,G) and sema6a/plxna2 double (D,H,N) morphant eyes processed with riboprobes for foxg1a (A-D), vax2 (E-H; posterior extent of expression indicated by red dots) and tbx5 (K-N). (A-D) Transverse sections. (E-H,K,L) Lateral views. (M,N) Dorsal views. A/P orientation in E applies. (A-D) foxg1a label is present ventrally in morphants (arrowheads) but not in control (arrow). In E-H, arrows indicate in situ label showing through from the opposite eye vesicle. Mesenchyme (*) is stuck to the eye vesicle in L. (E2,F2) Coronal sections show vax2 mRNA expansion into the posterior eye vesicle in a plxna2 morphant. (I,J) Analysis method (I) and the ratio of vax2 domain height (Ev) normalized to eye vesicle height (Eh) (J). ***P<0.001, unpaired Student′s t-test. Numbers of embryos in brackets (n=2); error bars indicate s.e.m.

plxna2 morphants form retinas with the appropriate cell types. (A-D) Transverse sections of 21hpf control (A,B) and plxna2 (C,D) morphant rx3:GFP optic cups immunolabeled with anti-ZO-1. (B,D) Merge. a, apical; b, basal. (E-P) Analysis of 72hpf retinas. Transverse sections (F,H-N). (E-H) Control (E) and plxna2 morphant (G) indicating smaller eye (bracket) and pericardial edema (*) in the morphant that likely reflects cardiac Plxna2 expression (Toyofuku et al., 2004). (F,H) Hematoxylin and Eosin staining. Optic nerve head is indicated by arrows. (I-N) Immunolabeling of ZN8+ retinal ganglion cells (RGCs) (I,J), Pax6+ RGCs and amacrine cells (K,L), and rhodopsin+ rods (M,N). Scale bar: 50μm for I-L; 20μm for M,N. (O,P) Eye height (O) (***P<0.001; one-way ANOVA, Dunnett′s test, n=2) and % of total nuclei in each layer of the central one-third of a central retinal section (P) (control, n=6; plxna2 morphant, n=8). Error bars indicate s.e.m. inl, inner nuclear layer; l, lens; ipl, inner plexiform layer; nr, neural retina; opl, outer plexiform layer; pe, pigment epithelium; pr, photoreceptors.

plxna2 and sema6a expressed in the 72 hpf zebrafish retina. (A,B) RNA in situ hybridization of plxna2 (A) and sema6a (B) viewed in transverse retinal sections at 72 hpf. inl, inner nuclear layer; rgc, retinal ganglion cell layer.

Apoptosis but not regionalization is rescued in the plxna2 and sema6a double morphant. A-D: Transverse sections showing anti-activated Caspase3 immunolabeling of the eye vesicle (outlined by dots) of a sema6a (A-C) and a sema6a/plxna2 double (D) morphant at 14 somites. D, dorsal; V, ventral.

Reduced proliferation in sema6a and plxna2 morphant eyes. (A-D) Dorsal views of wholemount control (A,C) and plxna2 morphants (B,D) processed for rx3 mRNA in situ hybridization at 8 (A,B) and 16 (C,D) somites. rx3 is downregulated by 16 somites in a plxna2 morphant, specifically in the eye (arrows). (E-J) Dorsal views of GFP+ eye vesicles immunostained with anti-pHH3 (red) in control (E,F), plxna2 (G,H), and sema6a (I,J) morphant at 20 somites. (K) Quantitation of the numbers of pHH3+ cells within the eye vesicles of control and morphant embryos. N=2. Error bars represent s.e.m (***, P<0.001; One way ANOVA, Dunnett’s test).

(A,B) GFP expression in differentiating RGCs initiates at 30 hpf in both a control (A) and plxna2 morphant (B) isl2b:GFP transgenic eye (white arrows), indicating no apparent delay in the timing of RGC differentiation in plxna2 morphants.