August 2018
Volume 59, Issue 10
Open Access
Retinal Cell Biology  |   August 2018
Temporal Requirement of Mab21l2 During Eye Development in Chick Reveals Stage-Dependent Functions for Retinogenesis
Author Affiliations & Notes
  • Soufien Sghari
    Umeå Centre for Molecular Medicine, Umeå University, Umeå, Sweden
  • Lena Gunhaga
    Umeå Centre for Molecular Medicine, Umeå University, Umeå, Sweden
  • Correspondence: Lena Gunhaga, Umeå Centre for Molecular Medicine, Building 6M 4th Floor, Umeå University, Umeå 901 87, Sweden; lena.gunhaga@umu.se
Investigative Ophthalmology & Visual Science August 2018, Vol.59, 3869-3878. doi:10.1167/iovs.18-24236
  • Views
  • PDF
  • Share
  • Tools
    • Alerts
      ×
      This feature is available to authenticated users only.
      Sign In or Create an Account ×
    • Get Citation

      Soufien Sghari, Lena Gunhaga; Temporal Requirement of Mab21l2 During Eye Development in Chick Reveals Stage-Dependent Functions for Retinogenesis. Invest. Ophthalmol. Vis. Sci. 2018;59(10):3869-3878. doi: 10.1167/iovs.18-24236.

      Download citation file:


      © ARVO (1962-2015); The Authors (2016-present)

      ×
  • Supplements
Abstract

Purpose: Different missense mutations in the single exon gene Mab21l2 have been identified in unrelated families with various bilateral eye malformations, including microphthalmia, anophthalmia, and coloboma, but the molecular function of Mab21l2 during eye development still remains largely unknown.

Methods: We have established an in vivo Mab21l2-deficient eye development model in chick, by using a Mab21l2 RNA interference construct that we electroporated in ovo in prospective retinal cells. In addition, we designed a Mab21l2 gain-of-function electroporation vector. Mab21l2-modulated retinas were analyzed on consecutive sections in terms of morphology, and molecular markers for apoptosis, cell proliferation, and retinogenesis.

Results: Our Mab21l2-deficient chick model mimics human ocular phenotypes. When Mab21l2 is downregulated prior to optic vesicle formation, the embryos develop anophthalmia, and Mab21l2 inhibition by optic cup stages results in a microphthalmic colobomatous phenotype. Our results show that inhibition of Mab21l2 affects cell proliferation, cell cycle exit, and the expression of Atoh7/Ath5, NeuroD4/Ath3, Isl1, Pax6, AP-2α, and Prox1. In addition, Mab21l2 overexpression hampers cell cycle exit and differentiation of retinal progenitor cells (RPCs).

Conclusions: Our results highlight the importance of a regulated temporal expression of Mab21l2 during eye development: At early stages, Mab21l2 is required to maintain RPC proliferation and expansion of cell number; before retinogenesis, a decrease in Mab21l2 expression in proliferating RPCs is required for cell cycle exit and differentiation; during retinogenesis, Mab21l2 is chronologically upregulated in RGCs, followed by differentiated horizontal and amacrine cells and cone photoreceptor cells.

During eye development, defects in proliferation, migration, and differentiation can cause various diseases leading to partial or complete loss of vision. Ocular coloboma is a congenital visual impairment affecting structural eye formation with an estimated prevalence of 1 per 5000 live births.13 Coloboma describes a large spectrum of eye malformations caused by failure of optic fissure closure during embryogenesis. Coloboma belongs to a group of eye disorders that also includes anophthalmia and microphthalmia, characterized by the absence of and a marked reduction in size of the eye, respectively. The most common genes associated with the structural eye malformations microphthalmia, anophthalmia, and coloboma (MAC) have been reviewed and include among others, SOX2, OTX2, PAX6, RAX, VSX2, BMP4/7, ALDH1A3, and RARB.4,5 
Recently, different missense mutations in the single exon gene Mab21l2 have been identified in unrelated families with bilateral eye malformations. These mutations are either de novo dominant heterozygous causing amino acid substitutions (R51C, R51H, R51G, E49K) or recessive homozygous causing amino acid (R247Q) substitution.68 Interestingly, these Mab21l2 mutations cause a large spectrum of major structural eye malformations including MAC, microcornea, cataract, and eso/exotropia.68 Other extraocular Mab21l2-deficient phenotypes are identified as intellectual disability, rhizomelic skeletal dysplasia, and hand and foot syndactyly.68 The human Mab21l2 gene encodes a 41-kDa (kilodalton) nuclear protein similar to C. elegans mab-21 cell fate determining factor.9 Several studies have shown that Mab21l2 is conserved between human and other vertebrate species, and is expressed in the eye, midbrain, branchial arches, limb buds, body wall, and umbilical cord.1012 Parallel studies have tried to establish Mab21l2-deficient animal models to study the function of Mab21l26,12,13; however, embryonic lethality in Mab21l2 knockout mouse and zebrafish lines have prohibited conclusive results. 
Here, by using double-stranded (ds)RNA interference, we have established an in vivo Mab21l2-deficient eye development model in chick that mimics the human ocular phenotypes. In combination with Mab21l2 overexpression studies, our results indicate that Mab21l2 has different functions depending on phases of eye development. When Mab21l2 is downregulated at early stages, the embryos develop anophthalmia with decreased cell proliferation. Downregulation of Mab21l2 at somewhat later stages results in a microphthalmic colobomatous phenotype. In addition, our results provide evidence that inhibition of Mab21l2 affects cell cycle exit, the initiation of retinogenesis, and the specification and maturation of horizontal cells (HCs) and amacrine cells (ACs). 
Methods
Embryos
Fertilized white Lohman chicken eggs were obtained from Strömbäcks Ägg, Vännäs, Sweden. Chick embryos were staged according to the Hamburger and Hamilton stage (HH).14 
Design of Mab21l2 Constructs
For Mab21l2 loss-of-function experiments, a 680-bp (base pair) double-stranded (ds)RNA was designed from cDNA. Gain of-function experiments were carried out using the pCAG-Mab21l2-P2A-EGFP-m5 vector. For details see Supplementary Methods
In Ovo Electroporation
Stage 8 to 12 chick embryos were electroporated in the optic vesicle region. To get restricted targeted and viable embryos, microelectroporation was performed.15 The constructs used were pCAβ-EGFPm516 at 1 μg/μL, together with either dsMab21l2 at 0.4 μg/μL, or pCAG-Mab21l2-P2A-EGFP-m5 at 1 μg/μL; 0.01% Fast Green FCF (Sigma-Aldrich, Corp., St. Louis, MO, USA) was added as contrast to control the site of injection. The vectors were electroporated by applying three square pulses (9–12 volts, 25-ms duration, 0.5-second interval) (T830 BTX; Genetronics, Holliston, MA, USA). After electroporation, eggs were sealed and incubated at 38°C until reaching the stage of interest. 
Fixation and Histology
After fixation with 4% paraformaldehyde (PFA), embryos were transferred gradually to 7%, 15%, and 25% sucrose, embedded, and stored at −80°C until cryosectioned at 10 μm on consecutive slides. Sections were stained with hematoxylin and eosin (H&E) and examined under a light microscope for histoarchitectural data. 
Immunohistochemistry
Sections were processed for immunohistochemistry using standard protocols17 including blocking with 10% fetal calf serum (FCS) before primary antibody incubation. The Table presents primary and secondary antibodies used. Nuclei were detected using 4′,6-diamidino-2-phenylindole (DAPI) (1:400; Sigma Aldrich, Corp.). 
Table
 
Primary and Secondary Antibodies Used (Dilution, Species, Company or Gift Reference, and Catalogue Number)
Table
 
Primary and Secondary Antibodies Used (Dilution, Species, Company or Gift Reference, and Catalogue Number)
In Situ Hybridization
In situ RNA hybridization was performed essentially as previously described.18 Applied chick digoxigenin-labeled probes were Mab21l2, Mab21l1, Atoh7/Ath5, NeuroD4/Ath3 (for cloning details see Supplementary Methods), and Vsx2 (previously known as Chx10).19 
Statistics and Imaging
Phosphorylated histone H3 (pHH3)+ cells were quantified on five sections from five dsMab21l2-electroporated embryos. The graphs represent the mean number ± SEM as a percentage of the total cell number. Student's t-test was used to determine significance, and P values of <0.05 or <0.0001 were accepted as statistically significant. Images were produced using a Nikon Eclipse E800 microscope (Nikon Instruments, Inc., Melville, NY, USA) for simultaneous Epifluorescence/DIC observations, equipped with a CCD camera connected to a computer (Nikon Imaging Software NIS-Elements). Images were processed with Photoshop CS2 (Adobe, San Jose, CA, USA). 
Results
Early Inhibition of Mab21l2 Leads to Anophthalmia
The spectrum of eye malformations identified in families with Mab21l2 mutations prompted us to investigate whether inactivation of Mab21l2 at different developmental stages can cause these various phenotypes. To address this, we established an in vivo chick model assay in which Mab21l2 was downregulated by the use of RNA interference. First, we electroporated embryos at HH 8-10 in one of the prospective optic vesicle regions, with a control vector expressing green fluorescent protein (GFP )16 alone or together with a 680-bp ds part of the Mab21l2 RNA. The nonelectroporated optic vesicle was used as control. The electroporated embryos were cultured to approximately embryonic day (E) 2-6 (HH 13–HH 29), and embryos with GFP staining in the targeted retina were collected for analyses. 
All embryos electroporated with the GFP control vector alone displayed normal morphology of the developing optic vesicle and retina (see Supplementary Figs. S1A–D). Moreover, the expression of Mab21l2 and microtubule associated protein (MAP), labeled with RA4 antibody, indicating retinal progenitor cell (RPC) differentiation, was not affected in the GFP -electroporated cells of the eye (Supplementary Figs. S1A–D). In contrast, embryos electroporated with the dsMab21l2 construct at HH 8-10 and cultured to E6 exhibited a rudimentary retina and a significantly smaller but morphologically unaffected lens in the electroporated side, compared to the control side (Figs. 1A–D; see Supplementary Fig. S2A). Although hemorrhages occasionally were observed in the electroporated side, maybe due to disturbed vascularization, the development of the midbrain appeared mostly unaffected (Supplementary Figs. S2A, S2B). Downregulation of Mab21l2 was verified by reduced Mab21l2 expression in embryos cultured to E2.5 (Figs. 2A, 2D; Supplementary Fig. S2B), and also led to downregulated Vsx2 expression (Figs. 2C, 2F), which is in agreement with previous results in mouse.12 Notably, the expression of Mab21l1 was not altered by the dsMab21l2 construct, verifying the specificity of the dsMab21l2 construct (Figs. 2B, 2E). 
Figure 1
 
Inhibition of Mab21l2 at HH 8-10 causes anophthalmia. (A, C) Lateral whole-mount views of E6 control (A) and dsMab21l2 anophthalmic (C) sides after electroporation at HH 8-10 (n = 4/4). (B, D) H&E staining of transversal sections of E6 control (B) and dsMab21l2-electroporated (D) sides of the embryo shown in (A, C) indicates that the retina is rudimentary and the lens is smaller in the dsMab21l2-electroporated side (n = 4/4). (EH) Analyses of embryos electroporated at HH 8-10 cultured to E2 (n = 4). GFP indicates the dsMab21l2-targeted area in the eye (G), and in situ hybridization shows downregulation of Mab21l2 in the targeted region (H) (n = 4/4). L, lens; R, retina. Scale bar: 100 μm (B, DH).
Figure 1
 
Inhibition of Mab21l2 at HH 8-10 causes anophthalmia. (A, C) Lateral whole-mount views of E6 control (A) and dsMab21l2 anophthalmic (C) sides after electroporation at HH 8-10 (n = 4/4). (B, D) H&E staining of transversal sections of E6 control (B) and dsMab21l2-electroporated (D) sides of the embryo shown in (A, C) indicates that the retina is rudimentary and the lens is smaller in the dsMab21l2-electroporated side (n = 4/4). (EH) Analyses of embryos electroporated at HH 8-10 cultured to E2 (n = 4). GFP indicates the dsMab21l2-targeted area in the eye (G), and in situ hybridization shows downregulation of Mab21l2 in the targeted region (H) (n = 4/4). L, lens; R, retina. Scale bar: 100 μm (B, DH).
Figure 2
 
Suppressed Mab21l2 results in reduced proliferation and disturbed retina morphology. (AP) Analyses on transversal sections of dsMab21l2-electroporated embryos at HH 8-10 and cultured to E2 (KP; n = 5), E2.5 (AF; n = 3), and E3 (GJ; n = 4). (AF) In situ hybridization shows specific downregulation of Mab21l2 (D), while Mab21l1 remains unaffected (E), compared to the control side (A, B) (n = 3/3). Vsx2 expression is suppressed in the Mab21l2-downregulated eye (F) (n = 3/3). (GJ) Immunolabeling indicates an increase in HuC/D+ postmitotic neurons in the electroporated side (I) compared to the control side (G) (n = 4/4). Prox1 indicates normal lens patterning in both sides (E, J) (n = 4/4). (KP) GFP indicates the dsMab21l2-targeted areas in the eye (N), and immunolabeling detected aCaspase3 and pHH3 (L, M, O, P). aCaspase3 expression is not altered between the control and Mab21l2-deficient retinas, whereas pHH3+ cells are significantly decreased (n = 5/5). (O) Histogram presenting reduced number of pHH3+ cells in the dsMab21l2-electroporated optic vesicles compared to controls at stage HH 13/14 (n = 5). ***t-test; P < 0.0001. Scale bar: 100 μm (AP).
Figure 2
 
Suppressed Mab21l2 results in reduced proliferation and disturbed retina morphology. (AP) Analyses on transversal sections of dsMab21l2-electroporated embryos at HH 8-10 and cultured to E2 (KP; n = 5), E2.5 (AF; n = 3), and E3 (GJ; n = 4). (AF) In situ hybridization shows specific downregulation of Mab21l2 (D), while Mab21l1 remains unaffected (E), compared to the control side (A, B) (n = 3/3). Vsx2 expression is suppressed in the Mab21l2-downregulated eye (F) (n = 3/3). (GJ) Immunolabeling indicates an increase in HuC/D+ postmitotic neurons in the electroporated side (I) compared to the control side (G) (n = 4/4). Prox1 indicates normal lens patterning in both sides (E, J) (n = 4/4). (KP) GFP indicates the dsMab21l2-targeted areas in the eye (N), and immunolabeling detected aCaspase3 and pHH3 (L, M, O, P). aCaspase3 expression is not altered between the control and Mab21l2-deficient retinas, whereas pHH3+ cells are significantly decreased (n = 5/5). (O) Histogram presenting reduced number of pHH3+ cells in the dsMab21l2-electroporated optic vesicles compared to controls at stage HH 13/14 (n = 5). ***t-test; P < 0.0001. Scale bar: 100 μm (AP).
Next, we examined whether changes in proliferation and/or cell death could explain the anophthalmic phenotype in the Mab21l2-deficient retina by analyzing pHH3 and activated (a) Caspase3 expression, respectively, 20 to 24 hours after electroporation. During these conditions, no difference in apoptosis was observed, whereas a significant decrease in proliferation was noted in the electroporated region of the retina compared to the control side (Figs. 2K–Q). To examine whether the decrease in proliferation was coupled to premature differentiation, we examined dsMab21l2-electroporated retinas at E3. An increase in postmitotic HuC/D neurons was detected in the malformed Mab21l2-deficient retina (Figs. 2G, 2I), indicating premature differentiation, whereas the differentiation of lens fiber cells, detected by Prox1, appeared normal (Figs. 2H, 2J). Together these data suggest that the anophthalmic phenotype is caused by specific downregulation of Mab21l2, which is required for proliferation of RPCs in the optic vesicle. 
Inhibition of Mab21l2 at the Onset of Optic Cup Formation Leads to Microphthalmic Coloboma
To investigate the impact of Mab21l2 downregulation at later stages, we blocked Mab21l2 by dsMab21l2 electroporations in the optic vesicle at HH 11-12. The embryos exhibited a microphthalmic colobomatous phenotype of varying degree (Figs. 3A, 3B, 3F, 3G), and histologic analysis revealed a reduction in ocular and lens size (Figs. 3C–E, 3H–J). In addition, the optic disc in the dsMab21l2-electroporated side exhibited an excavated shape with large diameter, compared to the normal funnel-shaped optic disc in the control side (Figs. 3H, 3J). Moreover, the optic nerve in the electroporated side appeared to be hypoplasic (Fig. 3J). Thus, at the stage of optic cup formation in chick, downregulation of Mab21l2 results in microphthalmic coloboma. 
Figure 3
 
Blocking Mab21l2 at HH 11-12 results in microphthalmic coloboma at E6. (A, B, F, G) Lateral whole-mount views of E6 control (A, B) and dsMab21l2 (F, G) sides (n = 4/4) after electroporation at HH 11/12. The arrowheads indicate the coloboma of the optic fissure (F, G). (CE, HJ) H&E-stained transversal sections. The dsMab21l2-electroporated side exhibits a reduced size of the eye (H) and a smaller lens (I) compared to the control side (C, D). The Mab21l2-deficient optic disc exhibits a bowl-shaped excavation with large diameter (blue line in J) associated with a hypoplasic optic nerve (J) compared to the funnel-shaped control optic disc (E). Scale bar: 100 μm (CE, HJ).
Figure 3
 
Blocking Mab21l2 at HH 11-12 results in microphthalmic coloboma at E6. (A, B, F, G) Lateral whole-mount views of E6 control (A, B) and dsMab21l2 (F, G) sides (n = 4/4) after electroporation at HH 11/12. The arrowheads indicate the coloboma of the optic fissure (F, G). (CE, HJ) H&E-stained transversal sections. The dsMab21l2-electroporated side exhibits a reduced size of the eye (H) and a smaller lens (I) compared to the control side (C, D). The Mab21l2-deficient optic disc exhibits a bowl-shaped excavation with large diameter (blue line in J) associated with a hypoplasic optic nerve (J) compared to the funnel-shaped control optic disc (E). Scale bar: 100 μm (CE, HJ).
Broad Mab21l2 Expression in the Optic Vesicle Becomes Restricted to Specific Retinal Cells
The temporal variations of ocular phenotypes obtained after Mab21l2 downregulation during eye development encouraged us to study Mab21l2 expression in detail during different developmental stages. Mab21l2 expression is upregulated in the optic vesicle and future lens ectoderm around HH 9, and maintains a similar broad expression pattern at optic cup stages, HH 13 and 14 (Figs. 4A–D). By stage HH 18, at the onset of retinogenesis, Mab21l2 expression becomes restricted to the vitreal side of the optic cup, known as the ganglion cell layer (GCL), of the center of neural retina (Figs. 4D, 4E). In addition, Mab21l2 expression was also detected radially in the periphery of the retina, and in the future retinal pigmented epithelium (RPE) (Fig. 4E). Also, at HH 18, MAP expression was detected in the GCL and in radial extensions, indicating postmitotic neuronal precursors (Fig. 4I).20 From E6 (HH 29), MAP expression is detected only in the GCL (Figs. 4J–L), which at E6 also is marked by Mab21l2 expression (Fig. 4F). At E8 (HH 32), Mab21l2 expression was detected in three different retinal layers: GCL, inner nuclear layer (INL), and outer nuclear layer (ONL), which become more segregated at stage E11 (HH 37) (Figs. 4G, 4H). Thus, at early stages of optic vesicle formation Mab21l2 is expressed broadly, and during stages of retinogenesis it becomes restricted to the GCL, INL, and ONL. 
Figure 4
 
Expression pattern of Mab21l2 along chick eye development. (AH) In situ hybridization of Mab21l2 in transversal sections. (AC) Onset of Mab21l2 expression around HH 9/10 in the entire optic vesicle and surface ectoderm until stages HH 13 and 14 (∼E2). (D, E) At stage HH 18 (∼E3.5), Mab21l2 is detected in the GCL in the center of the retina, indicated by arrowheads in (D), a magnification of marked region in (E), in the periphery of the retinal epithelium and all the retinal pigmented epithelium. (FH) Around E6 (F), Mab21l2 is expressed only in the GCL, and by E8 (G) and E11 (H), Mab21l2 is detected in the GCL, INL, and ONL. (IL) Immunohistochemistry analysis of transversal sections using RA4 antibody against chicken MAP. At stage HH 18, MAP is detected in the GCL and retinal neuronal precursors (I). By E6 and until E11, MAP becomes restricted to the GCL (JL). Black (G) and white (K) arrows indicate the GCL. oh, optic head. Scale bars: 100 μm (AC, EL); 25 μm (D).
Figure 4
 
Expression pattern of Mab21l2 along chick eye development. (AH) In situ hybridization of Mab21l2 in transversal sections. (AC) Onset of Mab21l2 expression around HH 9/10 in the entire optic vesicle and surface ectoderm until stages HH 13 and 14 (∼E2). (D, E) At stage HH 18 (∼E3.5), Mab21l2 is detected in the GCL in the center of the retina, indicated by arrowheads in (D), a magnification of marked region in (E), in the periphery of the retinal epithelium and all the retinal pigmented epithelium. (FH) Around E6 (F), Mab21l2 is expressed only in the GCL, and by E8 (G) and E11 (H), Mab21l2 is detected in the GCL, INL, and ONL. (IL) Immunohistochemistry analysis of transversal sections using RA4 antibody against chicken MAP. At stage HH 18, MAP is detected in the GCL and retinal neuronal precursors (I). By E6 and until E11, MAP becomes restricted to the GCL (JL). Black (G) and white (K) arrows indicate the GCL. oh, optic head. Scale bars: 100 μm (AC, EL); 25 μm (D).
The Progression of Retinogenesis Is Blocked After Mab21l2 Downregulation
To explore the role of Mab21l2 in retinogenesis, we incubated HH 11-12 dsMab21l2-electroporated embryos until E6, and analyzed the expression of Isl1 and MAP, indicative of retinal ganglion cell (RGC) genesis. At E6, Mab21l2 expression was still disrupted (see Supplementary Figs. S3A, S3D), and Vsx2 expression is suppressed, indicating that the RPC pool is reduced compared to the control side (Supplementary Figs. S3B, S3E). The patterning of the lens, indicated by Prox1 expression, appeared to be unaffected although reduced in size (Supplementary Figs. S3C, S3F). In addition, downregulation of Mab21l2 caused a drastically reduced number of Isl1+ RGCs in the electroporated side (Figs. 5A, 5D). Moreover, the expression pattern of MAP was disturbed, and showed an immature radial pattern through the width of the retina compared to the restricted GCL expression in the control side (Figs. 5B, 5E). We also observed a malformed optic head with an excavated shape in the dsMab21l2-electroporated side (Figs. 5A–F; see Supplementary Fig. S4). To further examine the growth of the optic head, we analyzed neurofilament medium (NF-M), normally expressed in the cell body and axons of the RGCs.20 NF-M expression was clearly reduced in the optic head region in the dsMab21l2-electroporated side (Figs. 5C, 5F). These data highlight the requirement of Mab21l2 for RGC genesis, and indicate that downregulated Mab21l2 expression results in disrupted maturation of postmitotic neurons. In addition, in conditions with reduced Mab21l2, the few differentiated RGCs failed to undergo normal axon extension to form a funnel-shaped optic head and instead their axons were aggregated. With all these observations taken together, downregulation of Mab21l2 results in blocked retinogenesis. 
Figure 5
 
Downregulation of Mab21l2 at HH 11-12 results in optic nerve hypoplasia. (AF) Immunohistochemical labeling of transversal sections at E6 of control (AC) and dsMab21l2 (DF) sides after electroporation at HH 11/12 shows a disturbed development of the optic head and optic nerve hypoplasia (n = 4/4). (A, D) The generation of Isl1+ RGCs is decreased in the dsMab21l2-electroporated optic vesicle (D) compared to control side (A) (n = 4/4). (B, E) MAP expression is detected in a radial pattern throughout the retina width in the dsMab21l2-electroporated optic vesicle (E), compared to GCL-restricted MAP expression in the control side (B) (n = 4/4). (C, F) NF-M is expressed in the GCL and optic nerve head in the control side (C), whereas NF-M expression is inhibited in the dsMab21l2-electroporated side (F) (n = 4/4). oh, optic head. Scale bar: 100 μm (AF).
Figure 5
 
Downregulation of Mab21l2 at HH 11-12 results in optic nerve hypoplasia. (AF) Immunohistochemical labeling of transversal sections at E6 of control (AC) and dsMab21l2 (DF) sides after electroporation at HH 11/12 shows a disturbed development of the optic head and optic nerve hypoplasia (n = 4/4). (A, D) The generation of Isl1+ RGCs is decreased in the dsMab21l2-electroporated optic vesicle (D) compared to control side (A) (n = 4/4). (B, E) MAP expression is detected in a radial pattern throughout the retina width in the dsMab21l2-electroporated optic vesicle (E), compared to GCL-restricted MAP expression in the control side (B) (n = 4/4). (C, F) NF-M is expressed in the GCL and optic nerve head in the control side (C), whereas NF-M expression is inhibited in the dsMab21l2-electroporated side (F) (n = 4/4). oh, optic head. Scale bar: 100 μm (AF).
GCL and INL Patterning Is Disturbed After Mab21l2 Downregulation
We continued to examine how GCL genesis was affected at E8, after long-term disruption of Mab21l2 function, by analyzing Atoh7 (also called Ath5) and MAP expression in the RGCs. After dsMab21l2 electroporation at HH 11-12 and culture to E8, Mab21l2 expression was still not recovered in all layers in the electroporated side (Figs. 6A, 6D). In the Mab21l2- disrupted side we noticed reduced levels and not clearly segregated expression of Atoh7 in the developing retina compared to the control side (Figs. 6B, 6E), indicating blocked retinogenesis. Consistently, MAP expression was still observed in a radial pattern in newly born RGCs, compared to mature GCL-restricted MAP expression in the control side (Figs. 6C, 6F). 
Figure 6
 
Downregulation of Mab21l2 impairs GCL and INL patterning at E8. (AS) Analysis of RGC and INL genesis in embryos electroporated at HH 11/12 and cultured until E8 (n = 3), by in situ hybridization (A, B, D, E, GI) and immunohistochemistry (C, F, JS). (A, B, D, E) Mab21l2 expression is still inhibited in the dsMab21l2-electroporated side (D) compared to expression in GCL, INL, and ONL in the control side (A), and Atoh7 expression is reduced in the GCL together with disturbed patterning (E) compared to the control side (B) (n = 3/3). (C, F) MAP is still expressed in a radial pattern in the dsMab21l2-electroporated side (F), compared to normal GCL-restricted MAP expression in the control side (C) (n = 3/3). At E6, NeuroD4 is expressed throughout the retina, apart from the GCL (arrowheads) (H), demarcating INL precursor cells. At E8, NeuroD4 is expressed in the ventricular surface of the control retina (G), while a NeuroD4 radial pattern is observed in the dsMab21l2-electroporated side (I) (n = 3/3). (JS) At E8, Isl1, Pax6, AP-2α, and Prox1 are expressed in the segregated GCL and INL in the control side, clearly viewed in magnified inserted boxes (L, M) (n = 3/3). In the dsMab21l2-electroporated side, the generation of Isl1+ RGCs and Pax6+ cells in the GCL and INL is decreased, and shows an intermingled pattern (OQ) (n = 3/3). The generation of AP-2α+ ACs is reduced in the dsMab21l2-electroporated side, with only a few remaining cells positioned in the INL (R), compared to the control side (M) (n = 3/3). The generation of Prox1+ cells is completely inhibited in the dsMab21l2-electroporated side (S) compared to the control side (N) (n = 3/3). Scale bar: 100 μm (AS).
Figure 6
 
Downregulation of Mab21l2 impairs GCL and INL patterning at E8. (AS) Analysis of RGC and INL genesis in embryos electroporated at HH 11/12 and cultured until E8 (n = 3), by in situ hybridization (A, B, D, E, GI) and immunohistochemistry (C, F, JS). (A, B, D, E) Mab21l2 expression is still inhibited in the dsMab21l2-electroporated side (D) compared to expression in GCL, INL, and ONL in the control side (A), and Atoh7 expression is reduced in the GCL together with disturbed patterning (E) compared to the control side (B) (n = 3/3). (C, F) MAP is still expressed in a radial pattern in the dsMab21l2-electroporated side (F), compared to normal GCL-restricted MAP expression in the control side (C) (n = 3/3). At E6, NeuroD4 is expressed throughout the retina, apart from the GCL (arrowheads) (H), demarcating INL precursor cells. At E8, NeuroD4 is expressed in the ventricular surface of the control retina (G), while a NeuroD4 radial pattern is observed in the dsMab21l2-electroporated side (I) (n = 3/3). (JS) At E8, Isl1, Pax6, AP-2α, and Prox1 are expressed in the segregated GCL and INL in the control side, clearly viewed in magnified inserted boxes (L, M) (n = 3/3). In the dsMab21l2-electroporated side, the generation of Isl1+ RGCs and Pax6+ cells in the GCL and INL is decreased, and shows an intermingled pattern (OQ) (n = 3/3). The generation of AP-2α+ ACs is reduced in the dsMab21l2-electroporated side, with only a few remaining cells positioned in the INL (R), compared to the control side (M) (n = 3/3). The generation of Prox1+ cells is completely inhibited in the dsMab21l2-electroporated side (S) compared to the control side (N) (n = 3/3). Scale bar: 100 μm (AS).
The INL becomes more pronounced around stage E8, and can be defined by NeuroD4 (also called Ath3), a bHLH transcription factor implicated in horizontal (HC) and amacrine (AC) cell specification.21 During normal development of the retina at E6, NeuroD4 is expressed throughout the neuroepithelium, defining INL precursor cells, but is not expressed in the GCL (Fig. 6H). At E8, NeuroD4 expression becomes restricted to the ventricular surface of the retina (Fig. 6G). However, after disrupting Mab21l2, NeuroD4 expression failed to restrict and remained expressed throughout the neuroepithelium (Fig. 6I), raising the question whether INL genesis is blocked at a precursor stage in the absence of Mab21l2 activity. To investigate this, we studied expression of a set of transcription factors implicated in INL specification: Pax6, AP-2α, and Prox1.2224 
In the control retina at E8, Pax6 is expressed in ACs and coexpressed with Isl1 in RGCs (Figs. 6J–L), which defines the two segregated INL and GCL, respectively. In contrast, in the Mab21l2-disrupted side a reduction in the number of both Pax6+ and Isl1+ cells was observed (Figs. 6O, 6P), and without clear segregation of the INL and GCL (Fig. 6Q). AP-2α is also implicated in AC specification,25 and was expressed in the INL in the control E8 retina (Fig. 6M). Strikingly, in the dsMab21l2-electroporated side, AP-2α was severely reduced, with only a few remaining AP-2α+ cells (Fig. 6R). At E8, the generation of HCs is ongoing in parallel with ACs. To examine whether Mab21l2 is required also for HC specification we studied the expression of Prox1, a transcription factor required for HC genesis.26 At this stage in the control retina, Prox1 expression is visible in the GCL and INL layers (Fig. 6N), whereas in the Mab21l2-disrupted side the Prox1 expression was completely absent (Fig. 6S). Together these data highlight that Mab21l2 appears to play a crucial role in HC and AC specification, as well as in the segregation of the GCL and INL of the developing retina. 
Temporal Regulation of Mab21l2 Is Required for the Initiation of Retinogenesis
To get further insights about the function of Mab21l2, we overexpressed Mab21l2 in the optic vesicle at HH 8-10, around the onset of endogenous Mab21l2 expression. We designed an expression construct of Mab21l2 linked to GFP (pCAG-Mab21l2-P2A-EGFP) to generate the same stoichiometric quantity of both Mab21l2 and GFP. In embryos cultured to E2, increased Mab21l2 expression was observed in the optic vesicle, and ectopic Mab21l2 expression was also detected in the forebrain, which overlapped with the GFP+ regions (Figs. 7A–C). No change in pHH3+ proliferative cells or the expression pattern of Vsx2, indicative of RPCs, was detected (Supplementary Fig. S5), suggesting that at this stage Mab21l2 promotes proliferation by inhibiting differentiation. We continued to analyze Mab21l2 overexpressed electroporated embryos cultured to E3, around the onset of retinogenesis. The generation of postmitotic retinal precursor cells, as well as RGCs reaching the GCL, was examined by expression of MAP. Strikingly, in the electroporated GFP+ regions, no expression of MAP, either in retinal precursors cells or in the GCL, was detected, in contrast to the normal MAP pattern in the control side (Figs. 7D–F). Thus, our data suggest that continuous ectopic expression of Mab21l2 at the onset of retinogenesis hampers cell cycle exit and differentiation of RPCs into retinal neurons. 
Figure 7
 
Continuous Mab21l2 expression in the retina results in suppressed retinogenesis. (AC) Mab21l2-P2A-EGFP-electroporated embryos at HH 9/10 and cultured to E2 (n = 4). GFP immunofluorescence indicates the electroporated area in the eye and forebrain (A). In situ hybridization on transversal sections shows overexpression of Mab21l2 in the eye and ectopic Mab21l2 expression in the forebrain side (B) compared to normal Mab21l2 in the eye in the control side (C) (n = 4/4). (DF) Mab21l2-P2A-EGFP-electroporated embryos at HH 9/10 and cultured to E3 (n = 5), and analyzed for MAP immunohistochemistry. GFP expression indicates the electroporated area (D). Overexpression of Mab21l2 results in loss of MAP expression in the electroporated side (E) compared to the control side with GCL and radial MAP expression (F) (n = 5/5). Scale bar: 100 μm (AF).
Figure 7
 
Continuous Mab21l2 expression in the retina results in suppressed retinogenesis. (AC) Mab21l2-P2A-EGFP-electroporated embryos at HH 9/10 and cultured to E2 (n = 4). GFP immunofluorescence indicates the electroporated area in the eye and forebrain (A). In situ hybridization on transversal sections shows overexpression of Mab21l2 in the eye and ectopic Mab21l2 expression in the forebrain side (B) compared to normal Mab21l2 in the eye in the control side (C) (n = 4/4). (DF) Mab21l2-P2A-EGFP-electroporated embryos at HH 9/10 and cultured to E3 (n = 5), and analyzed for MAP immunohistochemistry. GFP expression indicates the electroporated area (D). Overexpression of Mab21l2 results in loss of MAP expression in the electroporated side (E) compared to the control side with GCL and radial MAP expression (F) (n = 5/5). Scale bar: 100 μm (AF).
Discussion
In this study, we have established a novel vertebrate in vivo model to study how the specific loss of Mab21l2 in the optic vesicle affects eye development. Briefly, our results show that a temporal regulation of Mab21l2 in the early developing retina is crucial for cell cycle exit, the initiation of retinogenesis, and the specification of HCs and ACs, as well as the positioning of ACs in the INL. 
The range of eye phenotypes observed, including MAC, in several human cases of Mab21l2 mutations,4,7,8 suggests different functions of Mab21l2 during development. Consistent with this assumption, our data show that the expression pattern of Mab21l2 in chick changes during development, from an initial broad expression throughout the newly formed optic vesicle to restricted expression in the GCL, INL, and ONL during stages of retinogenesis, and the development of ACs and HCs. These expression patterns in chick resemble reported Mab21l2 expression in mouse and zebrafish.6,12,27 
Previous attempts to unravel the function of Mab21l2 in eye development using mouse Mab21l2 knockout and zebrafish homozygous Mab21l2 in-frame deletion and frameshift truncation have been challenged by early embryonic lethality due to disrupted heart formation.6,12,13 To overcome this issue, we established an in vivo chick model, in which Mab21l2 was inhibited specifically in the optic vesicle by the use of targeted electroporation of dsMab21l2. In addition, our chick Mab21l2-deficient model has the advantage that it can be used to determine the temporal requirement of Mab21l2 during early optic vesicle development and during stages of retinogenesis. Thus, our study further emphasizes the advantages of using chick as model for a large spectrum of human ocular diseases recently reviewed.28 
Our findings, using inhibition of Mab21l2 at specific time points, indicate that Mab21l2 plays different important roles during separate phases of eye development. An early inhibition of Mab21l2 in the prospective optic vesicle region, when RPCs proliferate and normally expand the retinal cell population, leads to a clear decrease in proliferation. In addition, the development of the retina is arrested at optic vesicle stages, which consequently results in an anophthalmic phenotype. This resembles the bilateral anophthalmia observed in human individuals with the identified Mab21l2 R51C mutation.6,8 In addition, our results are in accordance with previous studies of Mab21l2-disrupted zebrafish and mouse embryos, which also exhibit MAC phenotypes.6,12,27 Inhibition of Mab21l2 by optic cup stages resulted in the formation of microphthalmic coloboma, a phenotype observed in Mab21l2 R51H and E49L human mutations.8 Moreover, in human Mab21l2 mutations, the congenital malformations of the optic disc, associated with significantly impaired visual function, are excavation of the optic disc and optic nerve hypoplasia.68 Both of these features are observed in our chick Mab21l2-deficient model. Moreover, the Mab21l2-deficient optic head axons show decreased NF-M expression, whose function is required for axon growth and guidance.29 Thus, our study provides evidence that in the absence of Mab21l2 in the GCL, the optic head axons are aggregated and fail to extend to form a normal funnel-shaped optic disc. A previous study has shown that newly postmitotic RGCs express MAP protein, and that the initial broad expression pattern through the retina later becomes confined to the GCL.20 Our data showing persistent expression of MAP throughout the Mab21l2-deficient retina suggests that in the absence of Mab21l2, retinal neuronal precursors cannot differentiate properly into RGCs, which further explains the optic nerve hypoplasia. 
During normal development, RPCs give rise to various neurons in a chronological order, first the RGCs followed by HCs and ACs, which are regulated by a number of transcription factors.30 In the chick, lamination of the retina into specific cell layers becomes pronounced from E7.23,24 Our results at E8 show that downregulation of Mab21l2 leads to a decrease and/or disruption in the pattern of expression of the RGC-specific gene Atoh7 and transcription factor Isl1. Atoh7 is known to determine the competence state of RGC progenitors, and also acts upstream of Brn3b to promote RGC differentiation.31,32 Now our data suggest that Mab21l2 acts upstream of Atoh7 to determine the competence of newly born retinal precursor neurons in order for correct migration to the GCL and further maturation. Moreover, in the Mab21l2-deficient retina, NeuroD4 expression remains expanded throughout the thickness of the retina, indicating INL precursor cells, instead of becoming restricted to the ventricular surface of the retina. Thus, similar to our MAP data, these results suggest that the differentiation and/or migration of cells to the INL is arrested at a precursor stage. 
Our results show a reduction or loss of Pax6, AP-2α, and Prox1 expression after Mab21l2 inhibition, which implies that in the absence of Mab21l2 the generation of both ACs and HCs is inhibited, with only a few remaining ACs. We also noted a failure of segregation of the GCL and INL. Around E8, AP-2α is expressed in the INL, shown both in this study and by others,33 and has been implicated in HC and AC differentiation downstream of Ptf1a.25,34 Moreover, FoxN4 has also been suggested to regulate the generation of ACs and HCs by acting upstream of Ptf1a.3537 Strikingly, our results show that in the dsMab21l2-electroporated retina, AP-2α is severely reduced, with only a few remaining AP-2α+ cells. Our finding is in accordance with data observed in Ptf1a-, Tcfap2a/b-, and FoxN4-deficient mouse retinas, with a striking loss of HCs and reduction of ACs.34,36,37 Taken together, these data indicate that Mab21l2 might affect the FoxN4, Ptf1a, and/or AP-2 pathways, regulating the generation of both ACs and HCs. It remains to be determined whether Mab21l2 acts upstream of, in combination with, or separately from FoxN4, Ptf1a and/or AP-2 to regulate AC and HC differentiation. 
Since the normal Mab21l2 expression pattern changes from an initial broad expression to a restricted pattern during stages of retinogenesis, we hypothesized that sustained expression of Mab21l2 repressed RPC differentiation. Indeed, overexpression of Mab21l2 in the center of the retina and throughout the neuroepithelium until E3 (HH 20), when Mab21l2 normally is restricted to the GCL, hampered MAP expression, indicative of blocked cell cycle exit of RPCs and subsequent differentiation. Moreover, our data show that Mab21l2 overexpression has no effect on proliferation, indicating that Mab21l2 affects the process of differentiation rather than cell cycle rate. Previous studies have suggested that micro (mi)RNAs regulate the temporal order of retinogenesis by inhibiting transcription factors required for RPC differentiation.38,39 Similar to our findings, these miRNAs appear to promote differentiation downstream of the cell cycle setting.38 Interestingly, it has been shown that Mab21l2 protein binds to single-stranded (ss)RNA, but not ssDNA,8 implying that Mab21l2 might bind to miRNA. In such a scenario, Mab21l2 binding to miRNA could result in either stabilization of miRNA function to inhibit RPC differentiation, or a direct block of miRNA inhibition, which would promote differentiation. These two context-dependent scenarios need to be confirmed by additional experiments. Taken together our results highlight the importance of a regulated temporal expression of Mab21l2 during eye development: (1) Mab21l2 is required to maintain RPC proliferation and expansion of cell number; (2) just before the onset of retinogenesis, a decrease in Mab21l2 expression in the proliferating RPCs is required for cell cycle exit and differentiation, (3) during retinogenesis, Mab21l2 is chronologically upregulated in RGCs, followed by differentiated HCs and ACs in the INL, and then in cone photoreceptor cells in the ONL. 
Acknowledgments
The authors thank Cedric Patthey for help with cloning of Mab21l1 and Mab21l2, Connie Cepko for providing the Vsx2 plasmid, Steven McLoon for sharing the RA4 antibody, and members of the Gunhaga lab for helpful discussions. 
Supported by the Swedish Research Council, Umeå University Sweden, Ögonfonden, and Kronprinsessan Margaretas (KMA) foundation. 
Disclosure: S. Sghari, None; L. Gunhaga, None 
References
Morrison D, FitzPatrick D, Hanson I, et al. National study of microphthalmia, anophthalmia, and coloboma (MAC) in Scotland: investigation of genetic aetiology. J Med Genet. 2002; 39: 16–22.
Shah SP, Taylor AE, Sowden JC, et al. Anophthalmos, microphthalmos, and coloboma in the United Kingdom: clinical features, results of investigations, and early management. Ophthalmology. 2012; 119: 362–368.
Shah SP, Taylor AE, Sowden JC, et al. Anophthalmos, microphthalmos, and typical coloboma in the United Kingdom: a prospective study of incidence and risk. Invest Ophthalmol Vis Sci. 2011; 52: 558–564.
Reis LM, Semina EV. Conserved genetic pathways associated with microphthalmia, anophthalmia, and coloboma. Birth Defects Res C Embryo Today. 2015; 105: 96–113.
Williamson KA, FitzPatrick DR. The genetic architecture of microphthalmia, anophthalmia and coloboma. Eur J Med Genet. 2014; 57: 369–380.
Deml B, Kariminejad A, Borujerdi RHR, Muheisen S, Reis LM, Semina EV. Mutations in MAB21L2 result in ocular coloboma, microcornea and cataracts. PLoS Genet. 2015; 11: e1005002.
Horn D, Prescott T, Houge G, et al. A novel oculo-skeletal syndrome with intellectual disability caused by a particular MAB21L2 mutation. Eur J Med Genet. 2015; 58: 387–391.
Rainger J, Pehlivan D, Johansson S, et al. Monoallelic and biallelic mutations in MAB21L2 cause a spectrum of major eye malformations. Am J Hum Genet. 2014; 94: 915–923.
Baldessari D, Badaloni A, Longhi R, Zappavigna V, Consalez GG. MAB21L2, a vertebrate member of the male-abnormal 21 family, modulates BMP signaling and interacts with SMAD1. BMC Cell Biol. 2004; 5: 48.
Mariani M, Baldessari D, Francisconi S, et al. Two murine and human homologs of mab-21, a cell fate determination gene involved in Caenorhabditis elegans neural development. Hum Mol Genet. 1999; 8: 2397–2406.
Wong YM, Chow KL. Expression of zebrafish mab21 genes marks the differentiating eye, midbrain and neural tube. Mech Dev. 2002; 113: 149–152.
Yamada R, Mizutani-Koseki Y, Koseki H, Takahashi N. Requirement for Mab21l2 during development of murine retina and ventral body wall. Dev Biol. 2004; 274: 295–307.
Saito Y, Kojima T, Takahashi N. Mab21l2 is essential for embryonic heart and liver development. PLoS One. 2012; 7: e32991.
Hamburger V, Hamilton HL. A series of normal stages in the development of the chick embryo. J Morphol. 1951; 88: 49–92.
Momose T, Tonegawa A, Takeuchi J, Ogawa H, Umesono K, Yasuda K. Efficient targeting of gene expression in chick embryos by microelectroporation. Dev Growth Differ. 1999; 41: 335–344.
Yaneza M, Gilthorpe JD, Lumsden A, Tucker AS. No evidence for ventrally migrating neural tube cells from the mid- and hindbrain. Dev Dyn. 2002; 223: 163–167.
Wittmann W, Schimmang T, Gunhaga L. Progressive effects of N-myc deficiency on proliferation, neurogenesis, and morphogenesis in the olfactory epithelium. Dev Neurobiol. 2014; 74: 643–656.
Wilkinson DG, Nieto MA. Detection of messenger RNA by in situ hybridization to tissue sections and whole mounts. Methods Enzymol. 1993; 225: 361–373.
Chen CM, Cepko CL. Expression of Chx10 and Chx10-1 in the developing chicken retina. Mech Dev. 2000; 90: 293–297.
Gutierrez C, McNally M, Canto-Soler MV. Cytoskeleton proteins previously considered exclusive to ganglion cells are transiently expressed by all retinal neuronal precursors. BMC Dev Biol. 2011; 11: 46.
Inoue T, Hojo M, Bessho Y, Tano Y, Lee JE, Kageyama R. Math3 and NeuroD regulate amacrine cell fate specification in the retina. Development. 2002; 129: 831–842.
Belecky-Adams T, Tomarev S, Li HS, et al. Pax-6, Prox 1, and Chx10 homeobox gene expression correlates with phenotypic fate of retinal precursor cells. Invest Ophthalmol Vis Sci. 1997; 38: 1293–1303.
Boije H, Edqvist P-HD, Hallböök F. Temporal and spatial expression of transcription factors FoxN4, Ptf1a, Prox1, Isl1 and Lim1 mRNA in the developing chick retina. Gene Expr Patterns. 2008; 8: 117–123.
Edqvist PH, Lek M, Boije H, Lindbäck SM, Hallböök F. Axon-bearing and axon-less horizontal cell subtypes are generated consecutively during chick retinal development from progenitors that are sensitive to follistatin. BMC Dev Biol. 2008; 8: 46.
Jin K, Jiang H, Xiao D, Zou M, Zhu J, Xiang M. Tfap2a and 2b act downstream of Ptf1a to promote amacrine cell differentiation during retinogenesis. Mol Brain. 2015; 8: 28.
Dyer MA, Livesey FJ, Cepko CL, Oliver G. Prox1 function controls progenitor cell proliferation and horizontal cell genesis in the mammalian retina. Nat Genet. 2003; 34: 53–58.
Kennedy BN, Stearns GW, Smyth VA, et al. Zebrafish rx3 and mab21l2 are required during eye morphogenesis. Dev Biol. 2004; 270: 336–349.
Wisely CE, Sayed JA, Tamez H, et al. The chick eye in vision research: an excellent model for the study of ocular disease. Prog Retin Eye Res. 2017; 61: 72–97.
Eyer J, Peterson A. Neurofilament-deficient axons and perikaryal aggregates in viable transgenic mice expressing a neurofilament-β-galactosidase fusion protein. Neuron. 1994; 12: 389–405.
Ohsawa R, Kageyama R. Regulation of retinal cell fate specification by multiple transcription factors. Brain Res. 2008; 1192 (suppl C): 90–98.
Liu W, Mo Z, Xiang M. The Ath5 proneural genes function upstream of Brn3 POU domain transcription factor genes to promote retinal ganglion cell development. Proc Natl Acad Sci U S A. 2001; 98: 1649–1654.
Yang Z, Ding K, Pan L, Deng M, Gan L. Math5 determines the competence state of retinal ganglion cell progenitors. Dev Biol. 2003; 264: 240–254.
Bisgrove DA, Godbout R. Differential expression of AP-2α and AP-2β in the developing chick retina: repression of R-FABP promoter activity by AP-2. Dev Dyn. 1999; 214: 195–206.
Bassett EA, Korol A, Deschamps PS, et al. Overlapping expression patterns and redundant roles for AP-2 transcription factors in the developing mammalian retina. Dev Dyn. 2012; 241: 814–829.
Boije H, Shirazi Fard S, Ring H, Hallböök F. Forkheadbox N4 (FoxN4) triggers context-dependent differentiation in the developing chick retina and neural tube. Differentiation. 2013; 85: 11–19.
Fujitani Y, Fujitani S, Luo H, et al. Ptf1a determines horizontal and amacrine cell fates during mouse retinal development. Development. 2006; 133: 4439–4450.
Li S, Mo Z, Yang X, Price SM, Shen MM, Xiang M. Foxn4 controls the genesis of amacrine and horizontal cells by retinal progenitors. Neuron. 2004; 43: 795–807.
Decembrini S, Bressan D, Vignali R, et al. MicroRNAs couple cell fate and developmental timing in retina. Proc Natl Acad Sci U S A. 2009; 106: 21179–21184.
Ji HP, Xiong Y, Song W-T, et al. MicroRNA-28 potentially regulates the photoreceptor lineage commitment of Muller glia-derived progenitors. Sci Rep. 2017; 7: 11374.
McLoon SC, Barnes RB. Early differentiation of retinal ganglion cells: an axonal protein expressed by premigratory and migrating retinal ganglion cells. J Neurosci. 1989; 9: 1424–1432.
Figure 1
 
Inhibition of Mab21l2 at HH 8-10 causes anophthalmia. (A, C) Lateral whole-mount views of E6 control (A) and dsMab21l2 anophthalmic (C) sides after electroporation at HH 8-10 (n = 4/4). (B, D) H&E staining of transversal sections of E6 control (B) and dsMab21l2-electroporated (D) sides of the embryo shown in (A, C) indicates that the retina is rudimentary and the lens is smaller in the dsMab21l2-electroporated side (n = 4/4). (EH) Analyses of embryos electroporated at HH 8-10 cultured to E2 (n = 4). GFP indicates the dsMab21l2-targeted area in the eye (G), and in situ hybridization shows downregulation of Mab21l2 in the targeted region (H) (n = 4/4). L, lens; R, retina. Scale bar: 100 μm (B, DH).
Figure 1
 
Inhibition of Mab21l2 at HH 8-10 causes anophthalmia. (A, C) Lateral whole-mount views of E6 control (A) and dsMab21l2 anophthalmic (C) sides after electroporation at HH 8-10 (n = 4/4). (B, D) H&E staining of transversal sections of E6 control (B) and dsMab21l2-electroporated (D) sides of the embryo shown in (A, C) indicates that the retina is rudimentary and the lens is smaller in the dsMab21l2-electroporated side (n = 4/4). (EH) Analyses of embryos electroporated at HH 8-10 cultured to E2 (n = 4). GFP indicates the dsMab21l2-targeted area in the eye (G), and in situ hybridization shows downregulation of Mab21l2 in the targeted region (H) (n = 4/4). L, lens; R, retina. Scale bar: 100 μm (B, DH).
Figure 2
 
Suppressed Mab21l2 results in reduced proliferation and disturbed retina morphology. (AP) Analyses on transversal sections of dsMab21l2-electroporated embryos at HH 8-10 and cultured to E2 (KP; n = 5), E2.5 (AF; n = 3), and E3 (GJ; n = 4). (AF) In situ hybridization shows specific downregulation of Mab21l2 (D), while Mab21l1 remains unaffected (E), compared to the control side (A, B) (n = 3/3). Vsx2 expression is suppressed in the Mab21l2-downregulated eye (F) (n = 3/3). (GJ) Immunolabeling indicates an increase in HuC/D+ postmitotic neurons in the electroporated side (I) compared to the control side (G) (n = 4/4). Prox1 indicates normal lens patterning in both sides (E, J) (n = 4/4). (KP) GFP indicates the dsMab21l2-targeted areas in the eye (N), and immunolabeling detected aCaspase3 and pHH3 (L, M, O, P). aCaspase3 expression is not altered between the control and Mab21l2-deficient retinas, whereas pHH3+ cells are significantly decreased (n = 5/5). (O) Histogram presenting reduced number of pHH3+ cells in the dsMab21l2-electroporated optic vesicles compared to controls at stage HH 13/14 (n = 5). ***t-test; P < 0.0001. Scale bar: 100 μm (AP).
Figure 2
 
Suppressed Mab21l2 results in reduced proliferation and disturbed retina morphology. (AP) Analyses on transversal sections of dsMab21l2-electroporated embryos at HH 8-10 and cultured to E2 (KP; n = 5), E2.5 (AF; n = 3), and E3 (GJ; n = 4). (AF) In situ hybridization shows specific downregulation of Mab21l2 (D), while Mab21l1 remains unaffected (E), compared to the control side (A, B) (n = 3/3). Vsx2 expression is suppressed in the Mab21l2-downregulated eye (F) (n = 3/3). (GJ) Immunolabeling indicates an increase in HuC/D+ postmitotic neurons in the electroporated side (I) compared to the control side (G) (n = 4/4). Prox1 indicates normal lens patterning in both sides (E, J) (n = 4/4). (KP) GFP indicates the dsMab21l2-targeted areas in the eye (N), and immunolabeling detected aCaspase3 and pHH3 (L, M, O, P). aCaspase3 expression is not altered between the control and Mab21l2-deficient retinas, whereas pHH3+ cells are significantly decreased (n = 5/5). (O) Histogram presenting reduced number of pHH3+ cells in the dsMab21l2-electroporated optic vesicles compared to controls at stage HH 13/14 (n = 5). ***t-test; P < 0.0001. Scale bar: 100 μm (AP).
Figure 3
 
Blocking Mab21l2 at HH 11-12 results in microphthalmic coloboma at E6. (A, B, F, G) Lateral whole-mount views of E6 control (A, B) and dsMab21l2 (F, G) sides (n = 4/4) after electroporation at HH 11/12. The arrowheads indicate the coloboma of the optic fissure (F, G). (CE, HJ) H&E-stained transversal sections. The dsMab21l2-electroporated side exhibits a reduced size of the eye (H) and a smaller lens (I) compared to the control side (C, D). The Mab21l2-deficient optic disc exhibits a bowl-shaped excavation with large diameter (blue line in J) associated with a hypoplasic optic nerve (J) compared to the funnel-shaped control optic disc (E). Scale bar: 100 μm (CE, HJ).
Figure 3
 
Blocking Mab21l2 at HH 11-12 results in microphthalmic coloboma at E6. (A, B, F, G) Lateral whole-mount views of E6 control (A, B) and dsMab21l2 (F, G) sides (n = 4/4) after electroporation at HH 11/12. The arrowheads indicate the coloboma of the optic fissure (F, G). (CE, HJ) H&E-stained transversal sections. The dsMab21l2-electroporated side exhibits a reduced size of the eye (H) and a smaller lens (I) compared to the control side (C, D). The Mab21l2-deficient optic disc exhibits a bowl-shaped excavation with large diameter (blue line in J) associated with a hypoplasic optic nerve (J) compared to the funnel-shaped control optic disc (E). Scale bar: 100 μm (CE, HJ).
Figure 4
 
Expression pattern of Mab21l2 along chick eye development. (AH) In situ hybridization of Mab21l2 in transversal sections. (AC) Onset of Mab21l2 expression around HH 9/10 in the entire optic vesicle and surface ectoderm until stages HH 13 and 14 (∼E2). (D, E) At stage HH 18 (∼E3.5), Mab21l2 is detected in the GCL in the center of the retina, indicated by arrowheads in (D), a magnification of marked region in (E), in the periphery of the retinal epithelium and all the retinal pigmented epithelium. (FH) Around E6 (F), Mab21l2 is expressed only in the GCL, and by E8 (G) and E11 (H), Mab21l2 is detected in the GCL, INL, and ONL. (IL) Immunohistochemistry analysis of transversal sections using RA4 antibody against chicken MAP. At stage HH 18, MAP is detected in the GCL and retinal neuronal precursors (I). By E6 and until E11, MAP becomes restricted to the GCL (JL). Black (G) and white (K) arrows indicate the GCL. oh, optic head. Scale bars: 100 μm (AC, EL); 25 μm (D).
Figure 4
 
Expression pattern of Mab21l2 along chick eye development. (AH) In situ hybridization of Mab21l2 in transversal sections. (AC) Onset of Mab21l2 expression around HH 9/10 in the entire optic vesicle and surface ectoderm until stages HH 13 and 14 (∼E2). (D, E) At stage HH 18 (∼E3.5), Mab21l2 is detected in the GCL in the center of the retina, indicated by arrowheads in (D), a magnification of marked region in (E), in the periphery of the retinal epithelium and all the retinal pigmented epithelium. (FH) Around E6 (F), Mab21l2 is expressed only in the GCL, and by E8 (G) and E11 (H), Mab21l2 is detected in the GCL, INL, and ONL. (IL) Immunohistochemistry analysis of transversal sections using RA4 antibody against chicken MAP. At stage HH 18, MAP is detected in the GCL and retinal neuronal precursors (I). By E6 and until E11, MAP becomes restricted to the GCL (JL). Black (G) and white (K) arrows indicate the GCL. oh, optic head. Scale bars: 100 μm (AC, EL); 25 μm (D).
Figure 5
 
Downregulation of Mab21l2 at HH 11-12 results in optic nerve hypoplasia. (AF) Immunohistochemical labeling of transversal sections at E6 of control (AC) and dsMab21l2 (DF) sides after electroporation at HH 11/12 shows a disturbed development of the optic head and optic nerve hypoplasia (n = 4/4). (A, D) The generation of Isl1+ RGCs is decreased in the dsMab21l2-electroporated optic vesicle (D) compared to control side (A) (n = 4/4). (B, E) MAP expression is detected in a radial pattern throughout the retina width in the dsMab21l2-electroporated optic vesicle (E), compared to GCL-restricted MAP expression in the control side (B) (n = 4/4). (C, F) NF-M is expressed in the GCL and optic nerve head in the control side (C), whereas NF-M expression is inhibited in the dsMab21l2-electroporated side (F) (n = 4/4). oh, optic head. Scale bar: 100 μm (AF).
Figure 5
 
Downregulation of Mab21l2 at HH 11-12 results in optic nerve hypoplasia. (AF) Immunohistochemical labeling of transversal sections at E6 of control (AC) and dsMab21l2 (DF) sides after electroporation at HH 11/12 shows a disturbed development of the optic head and optic nerve hypoplasia (n = 4/4). (A, D) The generation of Isl1+ RGCs is decreased in the dsMab21l2-electroporated optic vesicle (D) compared to control side (A) (n = 4/4). (B, E) MAP expression is detected in a radial pattern throughout the retina width in the dsMab21l2-electroporated optic vesicle (E), compared to GCL-restricted MAP expression in the control side (B) (n = 4/4). (C, F) NF-M is expressed in the GCL and optic nerve head in the control side (C), whereas NF-M expression is inhibited in the dsMab21l2-electroporated side (F) (n = 4/4). oh, optic head. Scale bar: 100 μm (AF).
Figure 6
 
Downregulation of Mab21l2 impairs GCL and INL patterning at E8. (AS) Analysis of RGC and INL genesis in embryos electroporated at HH 11/12 and cultured until E8 (n = 3), by in situ hybridization (A, B, D, E, GI) and immunohistochemistry (C, F, JS). (A, B, D, E) Mab21l2 expression is still inhibited in the dsMab21l2-electroporated side (D) compared to expression in GCL, INL, and ONL in the control side (A), and Atoh7 expression is reduced in the GCL together with disturbed patterning (E) compared to the control side (B) (n = 3/3). (C, F) MAP is still expressed in a radial pattern in the dsMab21l2-electroporated side (F), compared to normal GCL-restricted MAP expression in the control side (C) (n = 3/3). At E6, NeuroD4 is expressed throughout the retina, apart from the GCL (arrowheads) (H), demarcating INL precursor cells. At E8, NeuroD4 is expressed in the ventricular surface of the control retina (G), while a NeuroD4 radial pattern is observed in the dsMab21l2-electroporated side (I) (n = 3/3). (JS) At E8, Isl1, Pax6, AP-2α, and Prox1 are expressed in the segregated GCL and INL in the control side, clearly viewed in magnified inserted boxes (L, M) (n = 3/3). In the dsMab21l2-electroporated side, the generation of Isl1+ RGCs and Pax6+ cells in the GCL and INL is decreased, and shows an intermingled pattern (OQ) (n = 3/3). The generation of AP-2α+ ACs is reduced in the dsMab21l2-electroporated side, with only a few remaining cells positioned in the INL (R), compared to the control side (M) (n = 3/3). The generation of Prox1+ cells is completely inhibited in the dsMab21l2-electroporated side (S) compared to the control side (N) (n = 3/3). Scale bar: 100 μm (AS).
Figure 6
 
Downregulation of Mab21l2 impairs GCL and INL patterning at E8. (AS) Analysis of RGC and INL genesis in embryos electroporated at HH 11/12 and cultured until E8 (n = 3), by in situ hybridization (A, B, D, E, GI) and immunohistochemistry (C, F, JS). (A, B, D, E) Mab21l2 expression is still inhibited in the dsMab21l2-electroporated side (D) compared to expression in GCL, INL, and ONL in the control side (A), and Atoh7 expression is reduced in the GCL together with disturbed patterning (E) compared to the control side (B) (n = 3/3). (C, F) MAP is still expressed in a radial pattern in the dsMab21l2-electroporated side (F), compared to normal GCL-restricted MAP expression in the control side (C) (n = 3/3). At E6, NeuroD4 is expressed throughout the retina, apart from the GCL (arrowheads) (H), demarcating INL precursor cells. At E8, NeuroD4 is expressed in the ventricular surface of the control retina (G), while a NeuroD4 radial pattern is observed in the dsMab21l2-electroporated side (I) (n = 3/3). (JS) At E8, Isl1, Pax6, AP-2α, and Prox1 are expressed in the segregated GCL and INL in the control side, clearly viewed in magnified inserted boxes (L, M) (n = 3/3). In the dsMab21l2-electroporated side, the generation of Isl1+ RGCs and Pax6+ cells in the GCL and INL is decreased, and shows an intermingled pattern (OQ) (n = 3/3). The generation of AP-2α+ ACs is reduced in the dsMab21l2-electroporated side, with only a few remaining cells positioned in the INL (R), compared to the control side (M) (n = 3/3). The generation of Prox1+ cells is completely inhibited in the dsMab21l2-electroporated side (S) compared to the control side (N) (n = 3/3). Scale bar: 100 μm (AS).
Figure 7
 
Continuous Mab21l2 expression in the retina results in suppressed retinogenesis. (AC) Mab21l2-P2A-EGFP-electroporated embryos at HH 9/10 and cultured to E2 (n = 4). GFP immunofluorescence indicates the electroporated area in the eye and forebrain (A). In situ hybridization on transversal sections shows overexpression of Mab21l2 in the eye and ectopic Mab21l2 expression in the forebrain side (B) compared to normal Mab21l2 in the eye in the control side (C) (n = 4/4). (DF) Mab21l2-P2A-EGFP-electroporated embryos at HH 9/10 and cultured to E3 (n = 5), and analyzed for MAP immunohistochemistry. GFP expression indicates the electroporated area (D). Overexpression of Mab21l2 results in loss of MAP expression in the electroporated side (E) compared to the control side with GCL and radial MAP expression (F) (n = 5/5). Scale bar: 100 μm (AF).
Figure 7
 
Continuous Mab21l2 expression in the retina results in suppressed retinogenesis. (AC) Mab21l2-P2A-EGFP-electroporated embryos at HH 9/10 and cultured to E2 (n = 4). GFP immunofluorescence indicates the electroporated area in the eye and forebrain (A). In situ hybridization on transversal sections shows overexpression of Mab21l2 in the eye and ectopic Mab21l2 expression in the forebrain side (B) compared to normal Mab21l2 in the eye in the control side (C) (n = 4/4). (DF) Mab21l2-P2A-EGFP-electroporated embryos at HH 9/10 and cultured to E3 (n = 5), and analyzed for MAP immunohistochemistry. GFP expression indicates the electroporated area (D). Overexpression of Mab21l2 results in loss of MAP expression in the electroporated side (E) compared to the control side with GCL and radial MAP expression (F) (n = 5/5). Scale bar: 100 μm (AF).
Table
 
Primary and Secondary Antibodies Used (Dilution, Species, Company or Gift Reference, and Catalogue Number)
Table
 
Primary and Secondary Antibodies Used (Dilution, Species, Company or Gift Reference, and Catalogue Number)
Supplement 1
×
×

This PDF is available to Subscribers Only

Sign in or purchase a subscription to access this content. ×

You must be signed into an individual account to use this feature.

×