SEARCH

SEARCH BY CITATION

Keywords:

  • cadherin-11;
  • differentiation;
  • retina;
  • visual system development;
  • antisense oligonucleotide

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Background: Cadherins orchestrate tissue morphogenesis by controlling cell adhesion, migration and differentiation. Various cadherin family members are expressed in the retina and other neural tissues during embryogenesis, regulating development of these tissues. Cadherin-11 (Cdh11) is expressed in mesenchymal, bone, epithelial, neural and other tissues, and this cadherin was shown to control cell migration and differentiation in neural crest, tumor and bone cells. Our previous studies characterized Cdh11 expression and function in zebrafish. Results: Here, we report effects of Cdh11 loss-of-function on visual system development using morpholino oligonucleotide knockdown methods. Cdh11 is expressed in the retina and lens during retinal differentiation. Cdh11 loss-of-function produced defects in retinal differentiation and lens development. Cdh11 loss-of-function also reduced retinotectal axon projection and organization, consistent with known Cdh11 function in cell migration. Conclusion: Cdh11 expression in the developing visual system and Cdh11 loss-of-function phenotype illustrates the critical role for differential cadherin activity in visual system differentiation and organization. Developmental Dynamics 241:442–454, 2012. © 2012 Wiley Periodicals, Inc.


INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Differential cell adhesion activity is necessary for embryo development by regulating tissue morphogenesis and differentiation (Halbleib and Nelson, 2006; Steinberg, 2007; Suzuki and Takeichi, 2008). During neurogenesis, neural tissue segregates from the ectoderm, neurons differentiate and neuronal processes extend, events that depend on differential adhesion mechanisms. Similarly, during retina development, neural tissue segregates and forms laminae with distinct cell types (photoreceptors, bipolar cells, amacrine cells, ganglion cells, etc.) that differentiate as organ development proceeds (Malicki, 2004; Wallace, 2011). In addition, retinal ganglion cells elaborate axonal processes that target specific areas of the brain (Reese, 2011). Functional development of the visual system depends on proper coordination of these processes. Cadherin cell adhesion molecules are important players in the coordinated regulation of neural tissue sorting, differentiation and connection into functional circuits (Suzuki and Takeichi, 2008).

Cadherin genes are grouped into a large superfamily that includes several subgroups that encode membrane proteins with evolutionarily conserved extracellular domain amino acid repeat sequences (Nollet et al., 2000). Type I and type II cadherin molecules are membrane proteins with a very well conserved organization of tandem cadherin repeats in the extracellular domain, a transmembrane domain and a cytoplasmic domain (Nollet et al., 2000). Cadherin extracellular domain repeats bind to one another on apposing cells, forming an intercellular molecular adhesion complex, which is generally homotypic (Halbleib and Nelson, 2006; Steinberg, 2007). Type I and type II cadherin cytoplasmic domains assemble with catenins, which regulate cadherin interaction with the membrane-cytoskeleton (Nollet et al., 2000). Cell biological studies show that cadherin activity regulates cell migration, proliferation, survival, differentiation, and various other processes (Halbleib and Nelson, 2006; Steinberg, 2007; Suzuki and Takeichi, 2008).

The neural retina differentiates from precursor cells within an undifferentiated neuroepithelium, eventually forming laminae that contain retinal cell types that alternate with synaptic layers (Malicki, 2004; Wallace, 2011). Retinal ganglion cells are within the innermost lamina, and photoreceptors are within the outermost lamina. The inner nuclear layer has bipolar, amacrine, horizontal, and Müller glial cells. In the zebrafish, the retinal ganglion cells are the first cells to exit the cell cycle (28 hours postfertilization; hpf), and these new retinal ganglion cells quickly begin to grow processes (Schmitt and Dowling, 1996). Other retinal cell types differentiate in a sequential series of steps orchestrated by an incompletely understood developmental induction process (Schmitt and Dowling, 1996; Malicki, 2004; Wallace, 2011). Synaptic connections are made, which produces functional integration of the neural retina (Sanes and Zipursky, 2010). Retinal ganglion cell projections exit the eye in a bundle called the optic nerve, whose migration and pathfinding to retinorecipient brain regions is controlled by a variety of chemoaffinity signals (Ruthazer and Cline, 2004). At the midline, ganglion cell axons intersect with axons from the opposite eye and migrate to the contralateral optic tectum, and the ganglion cell axons arborize and form connections with central neurons within the neuropil (Stuermer, 1988; Burrill and Easter, 1994).

Using the zebrafish, various laboratories have examined genes that affect visual system development (Malicki et al., 2002; Avanesov and Malicki, 2004), including the roles for cadherin cell adhesion molecules (Liu et al., 1999a,b, 2002, 2006, 2007, 2008, 2009, 2010; Erdmann et al., 2003; Malicki et al., 2003; Masai et al., 2003; Babb et al., 2005; Seiler et al., 2005; Yamaguchi et al., 2010; Lewis et al., 2011). Differential expression of cadherin cell adhesion molecules during retina development indicates that cadherins regulate cell sorting, cell recognition and morphogenesis. Functional analysis showed that cadherins control histogenesis in the zebrafish visual system (Erdmann et al., 2003; Malicki et al., 2003; Masai et al., 2003; Babb et al., 2005; Raymond et al., 2006; Liu et al., 2008; Lewis et al., 2011). For example, Cdh2 (N-cadherin) loss-of-function did not prevent robust differentiation of retinal cell types, but retina lamination was severely disrupted in Cdh2 loss-of-function embryos with mixing of cell types into various retinal laminae (Erdmann et al., 2003; Malicki et al., 2003; Masai et al., 2003). In contrast, Cdh4 (R-cadherin) and Cdh6 loss-of-function inhibited differentiation of retinal cell types, but when retinal cells differentiated, cells were correctly distributed within appropriate laminae (Babb et al., 2005; Liu et al., 2008).

Retinal ganglion cell axon projections to the optic tectum extend during the second day of development (Stuermer, 1988). In the zebrafish, these retinotectal projections normally only extend to the contralateral side of the brain (Burrill and Easter, 1994). In Cdh2 loss-of-function embryos, there were relatively normal numbers of ganglion cell axons, but errors in retinotectal axonal pathfinding at the midline were observed: many axons extended to the ipsilateral optic tectum in cdh2 mutant embryos (parachute mutants) (Masai et al., 2003). In contrast, Cdh4 loss-of-function embryos showed reduced ganglion cell differentiation, but all retinotectal projections extended to the contralateral optic tectum (Babb et al., 2005). In severely affected Cdh4 loss-of-function embryos, some retinotectal axons overshot the optic tectum neuropil (Babb et al., 2005). Thus, different cadherin family members produce distinctive phenotypes, strongly indicating that balanced, coordinated cell adhesion mechanisms control visual system development.

In our previous analysis of Cdh11 loss-of-function phenotype, phenotypic effects of Cdh11 loss-of-function were categorized into a series of severity on body length, tail curl, pigmentation, hindbrain, eyes, ears, and otoliths, and these categories of phenotypic severity correlated with Cdh11 protein expression (Clendenon et al., 2009). In the slightly affected category, embryos displayed near normal body length, no tail curl, reduced pigmentation, hindbrain defects. In the moderately affected category, embryos have short body length, slight tail curl, small eyes, small otoliths. In the severely affected category, embryos showed very short body length, tail extremely curled, very small eyes, very small, and sometimes absent otoliths.

Small eye phenotype was noted in Cdh11 (cadherin-11) loss-of-function embryos (Clendenon et al., 2009). In the current study, Cdh11 expression in the visual system, and the consequences of Cdh11 loss-of-function were examined in the zebrafish. Morpholino oligonucleotide injection was used to specifically knockdown Cdh11 expression, as previously characterized (Clendenon et al., 2009). We find that Cdh11 is required for proper visual system development.

RESULTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Experiments were performed to directly examine eye development and retinotectal projections. Morpholino oligonucleotides to knockdown Cdh11 expression were previously described and characterized (Clendenon et al., 2009). Three independent sequences that block mRNA splicing and maturation displayed the same overall phenotype and inhibit Cdh11 expression. Defects in retina and lens formation were observed in cdh11 knockdown embryos before retina differentiation (Clendenon et al., 2009). In severely affected cdh11 knockdown embryos, the optic cup was abnormally shaped, and retinal cells formed abnormal rosettes (Fig. 1H, arrowhead). Retinal tissue organization was disrupted and lamination was severely reduced or not apparent due to Cdh11 loss-of-function (Fig. 1B–D,F–H). Lenses were also smaller and disorganized in cdh11 knockdown embryos at these stages (Fig. 1B–D,F–H) relative to control embryos (Fig. 1A,E). These data indicate that Cdh11 activity controls retina and lens development.

thumbnail image

Figure 1. Eyes of cdh11 morphants are small and lamina organization is disrupted at 2 days postfertilization (dpf). A–H: Zebrafish embryos were injected with either standard control morpholino oligonucleotide (MO; control, A and E) or cdh11 MO (B–D,F–H). A–D: Ventral views eyes of living embryos were imaged using differential interference contrast microscopy (A–D). E–H: BODIPY-ceramide labeled living 2 dpf embryos were imaged using confocal microscopy. Anterior is up and lateral is right. All the major layers characteristic of the adult retina are seen in control embryos, including the retinal ganglion cell layer (rgl), inner plexiform layer (ipl; arrow), the inner nuclear layer (inl), and the outer nuclear layer (onl). B–D,F–H: Increasing severity of Cdh11 loss-of-function phenotype were observed: B and F, in slightly affected embryos, eyes were similar sizes but retinal lamina were reduced; C and G, in moderately affected embryos, the eyes were smaller in size, the inner plexiform layer was sparse, the distance between the inner plexiform layer and the lens was not uniform, and the outer nuclear layer was sparse or absent; D and H, eyes and lenses of severely affected embryos were small, rosettes were evident (arrowhead) and laminae nearly absent. Scale bar = 20 μm.

Download figure to PowerPoint

To examine whether Cdh11 expression correlates with defects observed in the zebrafish visual system, whole mount immunostaining and laser scanning confocal microscopy was performed on zebrafish embryos. Previous studies showed that cdh11 mRNA was expressed in numerous tissues throughout development (Franklin and Sargent, 1996; Liu et al., 2003; Clendenon et al., 2009). Our analysis specifically examined eye, retinotectal projections, and retinorecipient brain regions. An affinity purified Cdh11 antibody reagent was previously characterized and shown to be highly specific for immunostaining experiments (Clendenon et al., 2009).

Cdh11 expression was detected in the retina before retinal differentiation. At 18 hpf, Cdh11 was detected in the forming lens and neural retina (Fig. 2A–D). Immunofluorescence signal was detected in punctate spots on the cell surface and cytoplasmic puncta, which differs from the typical cadherin subcellular distribution that accumulate at sites of cell–cell contact. At 48 hpf, when retina differentiation is more complete, Cdh11 expression was widespread, but Cdh11 protein accumulated in the lens (Fig. 2E–H) and a subset of cells in the retina (Fig. 2I–L), including ganglion cells and associated with their processes (optic nerve; Fig. 2I–L). Cdh11 expression was also detected in the brain through various developmental stages. By labeling Cdh11 and neural processes using acetylated-tubulin antibody, Cdh11 accumulation was detected in association with the optic nerve within the retina (Fig. 2I–L); along its trajectory in the brain and at the chiasm (Fig. 2M–P); and, within the presumptive neuropil of the optic tectum (Fig. 2Q–T). Close association of Cdh11 and ganglion cell axons was observed in single laser scanning confocal microscope optical sections (Fig. 2L, inset, and 2P). In addition to expression in the ganglion cell axons, it is also possible that Cdh11 was expressed in cells along the retinotectal projection path to help guide ganglion cell axons. Because Cdh11 may be exported from the cells expressing this cadherin, it is very difficult to unequivocally determine which cells express Cdh11 in a particular brain region. Together, these expression data suggest that Cdh11 participates in visual system development.

thumbnail image

Figure 2. Distribution of Cdh11 in the developing lens, retina, and retinotectal projections. Laser scanning confocal microscope images are shown. Panels A–H, L (inset), and P are single optical sections, and other images are projection images of several optical sections to show neural projections through the tissues. Embryos at 18 hours postfertilization (hpf) were double labeled using antibodies that specifically recognize Cdh11 (green, panels A, C, and D) and TO-PRO-3 (nuclear stain, blue, panels B and C). Cdh11 and nuclear label images were merged (panel C). The boxed region in panel A contains the optic vesicle (ov), showing details of the retina and lens, are shown in panel D. A–D: Dorsal views are shown, and anterior is up. Embryos at 48 hours postfertilization (hpf) were triple labeled using antibodies that specifically recognize Cdh11 (green, panels E, I, M, and Q); antibodies that specifically recognize acetylated tubulin (red, panels F, J, N, and R); and TO-PRO-3 that labels nuclei (blue, panels G, K and S). Merged images (panels H, L, O, P, and T) are shown for comparison of expression patterns. Cdh11 was detected in the retina and lens at 18 and 48 hpf. E–H,M–T: Ventral views are shown, and anterior is up. I–L: Anterior is left, and lateral is up. Acetylated tubulin labels neuronal processes, including the optic nerve (on, panels J and L); the optic chiasm (oc, panels N, O and P); and the retinotectal projections arriving in the optic tectum (ot, panels R and T). Views of the eye (I and L) and brain (M, O, P, Q, and T) illustrate Cdh11 staining is associated with the optic nerve and optic chiasm. Scale bars = 20 μm.

Download figure to PowerPoint

Small eye phenotype could be due to effects of cdh11 knockdown on early events in eye development like eye induction and specification. To examine this possibility, expression of early eye induction (sonic hedgehog gene, shh) and specification (rx1 gene) markers were assayed by in situ hybridization in control and cdh11 morpholino oligonucleotide injected embryos. At 21 hpf, rx1 expression was strongly detected in the neural retina of control embryos (Fig. 3A). In cdh11 knockdown embryos, rx1 expression was detected in all embryos, including severely affected embryos (Fig. 3B). At 30 hpf, cdh11 morphant and control eyes also displayed rx1 expression (Fig. 3C,D). Sonic hedgehog gene (shh) expression is a marker for the optic stalk. Like rx1 gene expression, there was no detectable difference in shh expression in cdh11 knockdown relative to control embryos (Fig. 3E,F), but the shape or size of the optic stalk was different (arrows, Fig. 3E,F).

thumbnail image

Figure 3. Expression of early retinal specification markers in cdh11 morphant embryos. Zebrafish embryos were injected with either standard control morpholino oligonucleotide (MO, Control) or cdh11 MO. Whole-mount in situ hybridization labeling of embryos at different stages was performed using probes for markers retinal specification. Probes are shown in the bottom right corner of each panel. A,B:rx1 labeling of 21 hours postfertilization (hpf) control and cdh11 morphant embryos. Anterior views from a ventral perspective; a, anterior embryo border; d, dorsal surface. rx1 labeling was detected in control (A) and cdh11 morphant embryos (B). C,D:rx1 labeling was strongly detected in 30 hpf control and cdh11 morphant embryos. Dorsal view of 30 hpf embryos: anterior is top. E,F: Expression of shh, a marker of retinal stalk (arrows) differentiation, was strongly detected in 28 hpf control and cdh11 morphant embryos; lateral views, anterior is up.

Download figure to PowerPoint

Reduced eye size may be a consequence of reduced cell proliferation or increased cell death. Mitotic rate was measured using anti-phosphohistone H3 antibody labeling, which detects mitotic nuclei. Numbers of anti-phosphohistone H3 antibody positive nuclei per unit area were compared between control and cdh11 knockdown 48 hpf embryos. Mitosis in retinas of cdh11 knockdown 48 hpf embryos was increased in frequency (Fig. 4A,B; see statistical results in legend). Thus, mitotic rate does not explain the small eye phenotype seen in cdh11 knockdown embryos.

thumbnail image

Figure 4. Microphthalmia is due to increased cell death in cdh11 knockdown embryos. Zebrafish embryos were injected with either standard control morpholino oligonucleotide (MO, Control) or cdh11 MO (Slight, Moderate, Severe). A: Mitotic cells were visualized in eyes of 48 hours postfertilization (hpf) embryos using histone-H3 immunostaining. Most dividing cells were found in peripheral lamina of the retina, in both control and knockdown embryos. Apoptosis was detected using acridine orange staining in eyes of 48 hpf embryos. Fluorescent nuclei were detected using confocal microscopy, and fluorescent images were overlayed on differential interference contrast transmitted light images. Increasing numbers of fluorescent nuclei were detected in embryos with increasing severity of cdh11 knockdown phenotype. Anterior is up, and lateral is left. B: Quantitative analysis of H3-positive nuclei per unit area (arbitrary units) showed a significant increased mitosis rate in cdh11 knockdown eyes. Graph shows data separated into slightly, moderately and severely affected categories using criteria outlined in the text. Error bars represent standard deviation of the mean. Control, n = 10; slight, n = 9; moderate, n = 10; severe, n = 9. Analysis of variance (ANOVA) comparisons of control vs. pooled data from all groups and control vs. slight, moderate or severe groups were significant (*P < 0.05), but the mitosis rate differences between slight, moderate or severe groups were not statistically significant. Quantitative analysis of acridine orange stained nuclei showed a significant increased apoptosis rate in cdh11 knockdown eyes. Graph shows data separated into slightly, moderately and severely affected categories using criteria outlined in the text. Error bars represent standard deviation of the mean. Control, n = 10; slight, n = 12; moderate, n = 8; severe, n = 5. ANOVA comparisons of control vs. pool data from all groups and control vs. slight, moderate or severe groups were significant (*P < 0.001), and the apoptosis rate differences between slight, moderate or severe groups were statistically significant (**P < 0.001).

Download figure to PowerPoint

In our previous studies, increased apoptosis was observed in cadherin knockdown embryos (Babb and Marrs, 2004; Babb et al., 2005). Acridine orange staining was used to measure apoptosis in 48 hpf embryos. Apoptosis was dramatically increased specifically in the retina, and apoptosis frequency increased directly proportional to phenotype severity in cdh11 morphants (Fig. 4A,C; see statistical results in legend). Increased apoptosis frequency also correlated with the distribution of Cdh11 expression, rather than a general effect of morpholino oligonucleotide injection. Together, the data indicate that Cdh11 provides a survival signal during retinal differentiation.

Reduced differentiation events during retina development could also be responsible for the observed eye size reduction in cdh11 morphant embryos. To determine the consequences of Cdh11 loss-of-function on cell type specification and retinal histogenesis, control and cdh11 morpholino oligonucleotide injected embryos were assayed by immunofluorescence using antibodies that detect ganglion cells, amacrine cells and photoreceptors. Retinas from 48 hpf embryos were stained with zn-5 monoclonal antibody (which detects DM- GRASP/neurolin present on retinal ganglion cells and other neurons). Normal embryos have a uniform layer of zn-5 positive retinal ganglion cells, but in cdh11 morphant embryos, increasing severity of phenotype corresponded with disorganized retinal ganglion cell layer, reduced zn-5 immunolabeling in the ganglion cell layer and forming optic nerve (Fig. 5A), indicating that Cdh11 participates in retinal ganglion cell differentiation.

thumbnail image

Figure 5. cdh11 knockdown inhibits retinal differentiation. A–C: Zebrafish embryos were injected with standard control morpholino oligonucleotide (MO, Control columns) or cdh11 MO (Slight, Moderate, and Severe columns). A: In 2 day postfertilization (dpf) embryos, retinal ganglion cells were visualized using whole-mount zn-5 monoclonal antibody staining and confocal imaging. Ventral view projections of sections through the optic nerve show retinal ganglion cells and their axon processes. In cdh11 morphant embryos, the retinal ganglion cell layer was progressively reduced in thickness and numbers of ganglion cells detected. B: In 3 dpf embryos, cryostat sections were immunolabeled using antibodies that detect retinal layers: HuC/D (HU), which detects retinal ganglion cells and inner nuclear layer neurons; zn-5 (Zn5), which detects retinal ganglion cells and their axons; Pax6 (Pax6), which detects cells in the ganglion cell layer and brightly labeled amacrine cells within the inner nuclear layer (arrowhead in Control image); and zpr-1 (Zpr1), which detects double cone photoreceptors. Cdh11 loss-of-function results in a progressive reduction in immunostaining markers for differentiated cell types in the retina. C: In 4 dpf embryos, cryostat sections were immunolabeled using antibodies that detect retinal layers: Pax6 (Pax6), which detects a variety of cells including brightly labeled amacrine cells within the inner nuclear layer (arrowhead in Control image); and zpr-1 (Zpr1), which detects double cone photoreceptors. The older embryos maintain the Cdh11 loss-of-function phenotype, showing reduced labeling of differentiated cell types. Scale bar = 100 μm.

Download figure to PowerPoint

In 3 day postfertilization (dpf) embryos, effects of cdh11 knockdown on retinal ganglion, amacrine and photoreceptor cell differentiation was compared with control embryos. Using HuC/D antibody (which detects neuronal RNA-binding proteins) or zn-5 antibody, cdh11 knockdown embryos showed less retinal ganglion cell differentiation than control embryos (Fig. 5B). Pax6 antibody were used to detect brightly staining amacrine cells within the inner nuclear layer (there is also labeling in the ganglion cell layer). Reduced Pax6 immunostaining was seen in cdh11 morphants relative to control embryos, and in severely affected morphants, very little Pax6 staining was detected (Fig. 5B). Using a monoclonal antibody that detects red/green double cone photoreceptors (zpr-1), reduced photoreceptor differentiation was observed as a consequence of cdh11 morpholino injection, as compared with control embryos (Fig. 5B). In severely affected cdh11 morphants very few zpr-1 positive cells were detected (Fig. 5B).

Previous studies of cdh4 morphant embryos showed recovery of retinal cell differentiation at 4 dpf (Babb et al., 2005). To determine whether the retinal differentiation phenotype persists, 4 dpf embryos were examined for effects of cdh11 knockdown on retinal cell differentiation. Pax6 antibody staining was reduced within the inner nuclear layer relative to control embryos (Fig. 5C). Red/green double cone photoreceptor (zpr-1) differentiation was also reduced relative to control embryos (Fig. 5C). Together, these findings indicate that Cdh11 activity is needed for appropriate retinal differentiation during embryogenesis.

Our immunofluorescence experiments showed that Cdh11 is expressed in the retinal ganglion cells and in brain regions along the retinotectal path (Fig. 2), retinotectal projections were examined in cdh11 morpholino oligonucleotide injected embryos and compared with control embryos. Acetylated tubulin antibody staining labels neuronal processes, including the optic nerve (Fig. 6). In moderately affected cdh11 knockdown embryos, axons extend from the eye and reach the brain, forming the chiasm, but the optic nerve is thinner (Fig. 6B), as compared with control embryos (Fig. 6A).

thumbnail image

Figure 6. Retinotectal projections are reduced in cdh11 knockdown embryos. In 2 days postfertilization embryos, acetylated tubulin labeling detects retinal ganglion cell processes (optic nerve) and the optic chiasm (arrow). Image volumes were collected using 2-photon microscopy. A: Control embryo. B:cdh11 morphant embryo. In cdh11 knockdown embryos, retinal ganglion cell processes were present, but the optic nerve was thinner. C–F: Anterograde labeling of retinotectal projections are reduced in cdh11 knockdown embryos. Zebrafish embryos were injected with either standard control morpholino oligonucleotide (MO, C and C′) or cdh11 MO (D,D′,E,F). Embryos were fixed at 3 days postfertilization (dpf), and retinotectal projections were traced by injecting eyes with the lipophilic dye DiI (1,1-dioctadecyl-3,3,3,3-tetramethylindocarbo- cyanine perchlorate), and image volumes of the retinotectal projections were collected using 2-photon microscopy. In both controls and cdh11 knockdowns, retinal ganglion cell axons project toward the contralateral optic tectum. In control embryos (C,C′, where C′ is a rotation of the volume shown in C), upon reaching the optic tectum, retinal ganglion cells projections arborized extensively within the tectal neuropil. In slightly affected cdh11 knockdowns (D,D′), the optic nerve arborized extensively within the tectal neuropil, but there were projections that extended beyond the tectal neuropil (arrowheads). Also, there ectopic projections were detected at the optic chiasm (arrows in D and D′, where D′ is a rotation of the volume shown in D). In moderately affected embryos (E) and severely affected embryos (F), retinotectal projections were reduced, but those that reached the tectum arborized within the neuropil. In these embryos, there were also frequently projections that extended beyond the neuropil (arrowheads). Scale bars = 100 μm in A,B, 50 μm in C–F.

Download figure to PowerPoint

Normally, ganglion cell axons project from the eye, through a specific path to retinorecipient brain targets. In fish, all retinotectal projections extend to the contralateral optic tectum. Retinal ganglion cell axons ramified within the neuropil of the optic tectum, and these axon arbors are normally confined to the neuropil region (Fig. 6C,C′). In cdh11 knockdown embryos, the retinotectal projections were reduced and disorganized (Fig. 6D,D′,E,F). Axons from cdh11 morphant retinas ramified within the neuropil (note: nearly normal pattern in slightly affected embryo, Fig 6D,D′), but a subset of retinotectal projections in cdh11 morphants extended beyond the neuropil into the cellular layers of the tectum (Fig. 6D,D′,E,F). In more severely affected morphants, axons that overshoot the neuropil were proportionally more frequent (Fig. 6E,F). In moderately affected cdh11 morphants, small disorganized axonal projections at the optic chiasm were observed (Fig. 6D,D′), but these axons did not extend to the ipsilateral optic tectum.

DISCUSSION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Differential cell adhesion is necessary for organ development, and cadherin adhesion molecules were previously shown to be necessary for visual system development (Malicki, 2004; Pujic and Malicki, 2004; Halbleib and Nelson, 2006; Steinberg, 2007). There are numerous cadherin cell adhesion molecules that are expressed in the developing and adult retina (Honjo et al., 2000). Zebrafish has been a particularly useful model for examining functional roles of cadherin adhesion in visual system development, and expression patterns of various cadherin and protocadherin cell adhesion molecules were studied in zebrafish (Liu et al., 1999a,b, 2002, 2006, 2007, 2008, 2009, 2010; Erdmann et al., 2003; Malicki et al., 2003; Masai et al., 2003; Babb et al., 2005; Seiler et al., 2005; Yamaguchi et al., 2010; Lewis et al., 2011). The loss-of-function consequences for zebrafish Cdh2, Cdh4 and Cdh6 were examined (Erdmann et al., 2003; Malicki et al., 2003; Masai et al., 2003; Babb et al., 2005; Liu et al., 2008). In this study, the developmental consequences of Cdh11 loss-of-function were characterized.

Previous studies showed that Cdh11 participates in differentiation and cell migration (Simonneau and Thiery, 1998; Vallin et al., 1998; Borchers et al., 2001; Kawaguchi et al., 2001; Lee et al., 2007). First characterized in osteoblasts (Okazaki et al., 1994), Cdh11 is necessary for appropriate bone development (Kawaguchi et al., 2001). Cdh11 has also been called the mesenchymal cadherin because it is expressed in migratory embryonic cells before differentiation and in metastatic tumor cells (Kimura et al., 1995; Simonneau and Thiery, 1998; Vallin et al., 1998; Borchers et al., 2001; Chu et al., 2008; Nakajima et al., 2008; Huang et al., 2010). Cdh11 is also expressed in neural crest cells, and migratory behaviors of neural crest cells are regulated by metalloprotease degradation of Cdh11 (Simonneau and Thiery, 1998; Vallin et al., 1998; Borchers et al., 2001; McCusker et al., 2009). In the nervous system, Cdh11 is expressed in numerous neuronal cell types (Franklin and Sargent, 1996; Kimura et al., 1996; Faulkner-Jones et al., 1999). Cdh11 is expressed in the synapse and was shown to regulate long-term potentiation in the hippocampus (Manabe et al., 2000). Like Cdh2 (Letourneau et al., 1990; Jontes et al., 2004), Cdh11 is expressed in growth cones during axonal migration (Marthiens et al., 2005), and also like Cdh2 (Williams et al., 1994), migratory activity of Cdh11 is regulated by Fgf signaling (Boscher and Mege, 2008). Together, these findings illustrate that Cdh11 has diverse roles during embryo development.

Gene disruption of cdh11 in mouse was analyzed, showing somite and skeletal development defects (Horikawa et al., 1999; Kawaguchi et al., 2001). Neural function of Cdh11 was defined in the hippocampus, showing that Cdh11 is strongly expressed in the neuropil region and regulates long-term potentiation in the CA1 hippocampus region (Manabe et al., 2000). Interesting experiments using Xenopus revealed specific function of Cdh11 in neural crest migration (Borchers et al., 2001; McCusker et al., 2009). Using zebrafish, we previously showed that Cdh11 has an extracellular activity, controlling otolith growth in the inner ear (Clendenon et al., 2009). Analysis of Cdh11 function during development of the visual system was not previously reported.

Previous studies showed that Cdh11 is expressed in the ganglion cell layer and inner nuclear layer of the developing mouse retina (Honjo et al., 2000; Yamagata et al., 2006). Our immunofluorescence experiments using Cdh11 specific antibody to examine distribution in the developing zebrafish retina were consistent with the findings in mouse. Expression of cadherins, including Cdh11, suggests that differential adhesion control retina morphogenesis. However, Cdh11 distribution is different than other cadherins, displaying punctate staining. Cdh11 distribution in the cytoplasm and in the extracellular space may indicate an unusual activity for this cadherin, like that detected in the inner ear (Clendenon et al., 2009). In addition, Cdh11 was highly expressed in the epiderm of the eye, including the cornea and lens, which may exert an indirect effect on retinal development. Additional experiments will be needed to evaluate these possibilities.

Morpholino oligonucleotide knockdown experiments were used to examine the consequences of zebrafish Cdh11 loss-of-function. Knocking down Cdh11 expression produced a small eye phenotype. cdh11 knockdown did not prevent eye specification, but Cdh11 loss-of-function increased frequency of cell death in the retina. cdh11 knockdown also induced increased cell division, perhaps a compensatory response to the cell death. Thus, increased cell death is part of the mechanism for the small eye phenotype seen in cdh11 knockdown embryos. Cell death increases were detected in other cadherin loss-of-function studies. For example, Cdh4 knockdown in zebrafish produced a small eye phenotype that was due to increased apoptosis (Babb et al., 2005). Cadherin adhesion supplies a survival signal in many cellular contexts, which is critical for morphogenesis, including that of the retina. The lens and retina have reciprocal tissue interactions during development (Graw, 1996), but additional experiments will be needed to determine whether Cdh11 loss-of-function in the lens affects retina differentiation or vice versa.

Reduced differentiation also contributes to the Cdh11 loss-of-function retina phenotype. Blocking Cdh11 function prevented appropriate differentiation in the layers of the retina. Differentiation processes alter various cellular qualities; for example, in neural cells, differentiation produces various specialized membrane domains and structures, including axons, dendrites and, in photoreceptors, the outer segment. Reduced differentiation observed in cdh11 knockdown embryos may also contribute significantly to the small eye phenotype. Reduced differentiation was associated with other cadherin loss-of-function phenotypes, including zebrafish Cdh4 and Cdh6 (Babb et al., 2005; Liu et al., 2008). In contrast, zebrafish Cdh2 loss-of-function does not prevent differentiation of retinal cell types, but cell polarity defects result in the mixing of cell types from different retinal laminae (Erdmann et al., 2003; Malicki et al., 2003; Masai et al., 2003). In cdh11 knockdown embryos, some disorganization of retinal lamina was observed with retinal cells forming abnormal clusters or rosettes. However, differentiated cells were found in their appropriate laminae in cdh11 knockdown embryo retinas in our analysis.

Cadherin activity may affect the balance of progenitors and differentiated cell populations during retina and lens differentiation. Effects of cadherins on stem cell niche and differentiation mechanisms are incompletely understood (Song and Xie, 2002). Future studies could identify the primary effects of the various cadherins expressed in the retina. Recent findings that characterize progenitor populations in zebrafish will present many exciting possibilities for future research (Bernardos et al., 2007; Thummel et al., 2008, 2010; Cerveny et al., 2010).

Cdh11 was previously shown to regulate cell motility, including roles in neural crest cell migration and spinal cord neurite outgrowth (Simonneau and Thiery, 1998; Vallin et al., 1998; Borchers et al., 2001; Marthiens et al., 2005; Boscher and Mege, 2008; Chu et al., 2008; Nakajima et al., 2008; Huang et al., 2010). Retinal ganglion cells project axons to the visual centers of the brain through stereotypical trajectories (Burrill and Easter, 1994). Normally, zebrafish retinal ganglion cell axons fasciculate in the retina and exit the eye, sending projections to the contralateral brain visual targets. In cdh11 knockdown embryos, ganglion cells were reduced, but those ganglion cells that differentiated would send axons to the contralateral visual centers. Within the optic tectum, ganglion cell axons normally defasciculate and arborize within neuropil, where synapse formation occurs. Cdh11 loss-of-function does not prevent defasciculation and arborization of the axons. Of interest, retinotectal projections in cdh11 knockdown embryos showed ectopic neurite processes at the midline and neurite processes that overextend the neuropil, entering the cellular layer of the tectum. In Cdh4 knockdown embryos, neurites were also observed to overextend the neuropil and enter the cellular layer of the tectum (Babb et al., 2005). In Cdh2 loss-of-function embryos, large numbers of axons projected ipsilaterally (Masai et al., 2003), which is not observed in control embryos. cdh11 knockdown embryos did not send projections to the ipsilateral tectum, only producing small processes observed at the chiasm. Together, these findings indicate that different cadherin molecule subtypes control separate processes during retinotectal projection of ganglion cell axons, including pathfinding, fasciculation and target recognition.

In summary, consequences of Cdh11 loss-of-function for visual system development were characterized. Previous studies showed that Cdh11 loss-of-function affects gastrulation, somitogenesis, neurogenesis, and skeleton and inner ear development (Horikawa et al., 1999; Kawaguchi et al., 2001, Borchers et al., 2001; McCusker et al., 2009, Clendenon et al., 2009). A small eye phenotype was also noted in zebrafish cdh11 knockdown embryos. Here, we show that increased cell death in the retina and lens and reduced retinal differentiation contribute to the Cdh11 loss-of-function small eye phenotype, and analysis of retinotectal projections indicate that axon guidance in ganglion cells was defective, participating with other cadherin adhesion molecules to orchestrate retinotectal projection to central nervous system visual targets (Erdmann et al., 2003; Malicki et al., 2003; Masai et al., 2003; Babb et al., 2005; Raymond et al., 2006; Liu et al., 2008; Lewis et al., 2011). Additional studies are needed to characterize how Cdh11 activity interacts with other cadherin adhesion systems and with other regulatory molecules that control visual system development.

EXPERIMENTAL PROCEDURES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Zebrafish Husbandry and Morpholino Injection

Zebrafish (Danio rerio) maintenance and breeding was conducted under standard laboratory conditions (Westerfield, 2000) in accordance with Indiana University and University of Akron policies on animal care and use. cdh11 specific and control morpholino oligonucleotides (cdh11MOA: 5′-CTA AAG AAG GTA AAG TGT GTG AAT G-3′; cdh11MOB: 5′-CCC CAT CAG GTA GAG TCT GCT TCC T-3′; and cdh11MOC: 5′-TCA AGG ACA ACA GAG GTG CGT GAC A-3′; standard control; Gene Tools, Corvallis OR) were previously described (Clendenon et al., 2009). Morpholino oligonucleotides were microinjected into one- to four-cell stage embryos (Nasevicius and Ekker, 2000), and the injected embryos were kept in embryo medium (Westerfield, 2000) at 28.5°C. Phenylthiourea (PTU; 0.2 mM) was also added to embryo medium to prevent melanization.

BODIPY-ceramide Labeling and Transmitted Light Microscopy

BODIPY-ceramide-Fl-C5 (Molecular Probes) stock solution (5 mM in dimethyl sulfoxide) was diluted to 100 μM in embryos medium with 10 mM HEPES, and dechorionated living embryos were soaked in BODIPY-ceramide solution for 2 hr in the dark (Cooper et al., 1999). The embryos were washed, and confocal images were acquired using a Zeiss LSM 700 confocal microscope (Göttingen, Germany).

Differential interference contrast images of unstained eyes were collected using a Zeiss Observer Z1 with ×20 0.8 NA objective equipped with a Orca-AG CCD camera (Hamamatsu Photonics, Bridgewater, NJ).

Immunohistochemistry

Affinity purified, polyclonal, anti-peptide, Cdh11 specific rabbit antibody was previously described (Clendenon et al., 2009). Embryos were fixed in 4% paraformaldehyde in phosphate buffered saline (PBS) at 4°C overnight. Fixed embryos were dissected from their chorions and placed in blocking solution (PBS containing 0.5% Triton X-100, 5% goat serum, 0.2% bovine serum albumin, 0.05 M NH4 Cl and 0.02% sodium azide) overnight. Cdh11 specific antibody was diluted in blocking solution at 1:100 and incubated at least 2 days at room temperature. For some experiments, anti-acetylated tubulin monoclonal antibody (Sigma Chemical Co., St. Louis, MO) was diluted 1:200 in blocking solution with the Cdh11 specific antibody. Secondary antibody (TRITC-conjugated or Alexa-488-conjugated anti-rabbit for Cdh11 antibody or TRITC-conjugated anti-mouse for anti-acetylated tubulin monoclonal antibody; Jackson ImmunoResearch) was diluted in blocking solution at 1:50 and incubated overnight. TO-PRO-3 was incubated with embryos for 1–2 hr (1 μM TO-PRO-3 in PBS) to stain nuclei, and then rinsed twice with PBS before imaging. Two-photon microscope image volumes were acquired using a Zeiss LSM-510 Meta Confocal microscope System equipped with a tunable Titanium-Sapphire laser at the Indiana Center for Biological Microscopy. Confocal images were also acquired using a Zeiss LSM 700 system equipped with a Observer Z1. Projection images were produced from image volumes using Voxx software (Clendenon et al., 2002) or Volocity (Perkin Elmer, Walther, MA).

Immunolabeling of eye cryosections was performed as described previously (Babb et al., 2005) using the following primary antibodies: Pax6 antibody (AB5409, Chemicon International, Temecula, CA) at 1:50 dilution; phosphohistone 3 antibody (Upstate Biotechnology) at 1:200 dilution; Cdh11 affinity purified anti-peptide antibody (Clendenon et al., 2009) at 1:100 dilution; Zpr-1 (Zebrafish Rescource Center) at 1:200 dilution; and anti-HuC/D antibody (Molecular Probes, Eugene, OR,) at 1:500 dilution. Texas Red-conjugated anti-rabbit or anti-mouse secondary antibodies made in goat (Jackson Immunoresearch Labs, West Grove, PA) were used at 1:100. Confocal images were acquired using a Zeiss LSM 510 microscope system. Metamorph (Universal Imaging Corp, Downingtown, PA) and LSM Image Browser (Carl Zeiss, Inc.) were used to obtain single plane and projection images from confocal image stacks.

In Situ Hybridization

Whole-mount in situ hybridization of zebrafish embryos was performed as described (Liu et al., 1999b). Digoxigenin-labeled riboprobes for rx1 and shh (generously provided by Pamela Raymond and Stephen Ekker, respectively) were synthesized using the Genius System DIG RNA Labeling Kit (Roche, Indianapolis, IN).

Acridine Orange Staining

Apoptosis was detected in live embryos using acridine orange staining. Embryos were removed from their chorions and placed in embryo medium (Westerfield, 2000) containing 5 μg/ml of acridine orange (A-3568, Molecular Probes) for 2 min. Embryos were rinsed several in embryo medium and differential interference contrast transmitted light PMT and fluorescence images were collected using a Zeiss LSM 700 confocal microscopy system.

Mitosis and Cell Death Measurements and Statistical Methods

To examine microphthalmia cause, mitosis and cell death rate was measured in control and cdh11 knockdown embryos. Morphant embryos were separated into groups based on phenotypic severity (Slight, Moderate, Severe) using previously defined criteria (Clendenon et al., 2009). Mitotic (histone-H3 immunostaining) and apoptotic cells (acridine orange staining) in eyes of 48 hr postfertilization (hpf) embryos were counted. Numbers of eyes (individual) analyzed is given in the legend (Fig. 4). A ratio of number of H3-positive nuclei in the neural retina divided by the neural retina area (arbitrary units; measured using Image J software) was used for comparisions. analysis of variance (ANOVA) comparisons of control vs. pooled data from all groups and control vs. slight, moderate or severe groups was performed (see Fig. 4 legend). For apoptosis measurement, acridine orange stained nuclei were counted. Our findings underestimate the effect on apoptosis because we did not normalize the findings to area (neural retina is smaller in morphant embryos). ANOVA comparisons of control vs. pool data from all groups and control vs. slight, moderate or severe groups were performed (see Fig. 4 legend).

DiI Injections

DiI injections were performed as previously described (Babb et al., 2005). Briefly, zebrafish larvae (3 dpf) were fixed using 4% paraformaldehyde in PBS overnight at 4°C; rinsed with PBS; and embedded in 1% low melting temperature agarose. Eye were filled with dye by pressure injection of DiI (V-22888, Molecular Probes). Two-photon microscopy image volumes were acquired using a Bio-Rad MRC1024 laser scanning confocal system (Bio-Rad, Hercules, CA) equipped with a Titanium-Sapphire laser mounted on a Nikon inverted microscope, using an excitation wavelength of 760 nm. These image volumes were rendered using Voxx software (Clendenon et al., 2002).

Acknowledgements

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES

Drs. Pamela Raymond (University of Michigan) and Stephen Ekker (Mayo Clinic Cancer Center) generously provided plasmid probes to detect rx1 and shh mRNAs. Two-photon microscopy was performed at the Indiana Biological Microscopy Center. J.A.M. and Q.L. were funded by grants from the NIH.

REFERENCES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. RESULTS
  5. DISCUSSION
  6. EXPERIMENTAL PROCEDURES
  7. Acknowledgements
  8. REFERENCES