SEARCH

SEARCH BY CITATION

Keywords:

  • chick;
  • eye development;
  • LIM-homeobox;
  • neural retina;
  • optic vesicle

Abstract

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Concluding remarks
  8. Acknowledgements
  9. References

Elucidating the mechanisms underlying eye development is essential for advancing the medical treatment of eye-related disorders. The primordium of the eye is an optic vesicle (OV), which has a dual potential for generation of the developing neural retina and retinal pigment epithelium. However, the factors that regulate the differentiation of the retinal primordium remain unclear. We have previously shown that overexpression of Lhx1 and Lhx5, members of the LIM-homeobox genes, induced the formation of a second neural retina from the presumptive pigmented retina of the OV. However, the precise timing of Lhx1 expression required for neural retina differentiation has not been clarified. Moreover, RNA interference of Lhx5 has not been previously reported. Here, using a modified electroporation method, we show that, Lhx1 expression in the forebrain around stage 8 is required for neural retina formation. In addition, we have succeeded in the knockdown of Lhx5 expression, resulting in conversion of the neural retina region to a pigment vesicle-like tissue, which indicates that Lhx5 is also required for neural retina differentiation, which correlates temporally with the activity of Lhx1. These results suggest that Lhx1 and Lhx5 in the forebrain regulate neural retina differentiation by suppressing the development of the retinal pigment epithelium, before the formation of the OV.


Introduction

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Concluding remarks
  8. Acknowledgements
  9. References

Understanding the detailed mechanisms of retinal development is essential for the treatment of retinal diseases, including the formation of the retinal sheet used for retinal transplantation (Eiraku et al. 2011). The eye originates from a bilateral evagination of the forebrain, called the optic vesicle (OV). Subsequently, the distal part of the OV begins to invaginate to form the optic cup with two layers. These inner and outer layers develop into neural tissues, including the neural retina (NR) and the retinal pigmented epithelium (RPE) (Chow & Lang 2001).

Differentiation of the RPE depends on exposure to extracellular signals, such as transforming growth factor (TGF)-β molecules, including activin and bone morphogenetic proteins (BMPs), from periocular mesenchyme (Hyer et al. 1998; Fuhrmann et al. 2000; Müller et al. 2007). These signals are thought to promote RPE specification by inducing the expression of microphthalmia-associated transcription factor (Mitf), which is involved in the acquisition and maintenance of RPE identity (Martínez-Morales et al. 2004; Horsford et al. 2005). Simultaneously, Otx2, which is induced by Mitf, is also required for RPE specification (Crossley et al. 2001; Martinez-Morales et al. 2001; Martínez-Morales et al. 2003). In addition, the surface ectoderm, which abuts the OV, produces several fibroblast growth factor (FGF) signals that confer prospective NR characteristics at the distal region of the OV (Pittack et al. 1997; Hyer et al. 1998; Nguyen & Arnheiter 2000; Chow & Lang 2001; Kurose et al. 2004; Kurose et al. 2005). FGF signals then upregulate the expression of Chx10, which represses Mitf and induces the NR to undergo the normal specification of retinal cell types (McFarlane et al. 1998; Nguyen & Arnheiter 2000; Vogel-Höpker et al. 2000; Rowan et al. 2004; Horsford et al. 2005; Martinez-Morales et al. 2005). The interaction between Mitf and Chx10 confers the differences between the NR and RPE (Müller et al. 2007).

In the present study we focus on the LIM class homeodomain transcription factors, Lhx1 and Lhx5 (Fujii et al. 1994; Tsuchida et al. 1994; Sun et al. 2008) . We have previously shown that both Lhx1 and Lhx5 are expressed in the boundary region of the forebrain and optic vesicle, while overexpression studies have suggested that Lhx1 and Lhx5 facilitate NR development (Kawaue et al. 2012). However, the phenotypes of RNA interference (RNAi) against Lhx1 are inconsistent and the precise timing of Lhx1 expression required for neural retina differentiation has not been clarified. Moreover, RNAi of Lhx5 has not been previously reported. Therefore, the role for Lhx5 in early chicken retinal development has not yet been revealed by RNAi. In this study, we demonstrate that a new method of electroporation which can introduce more RNAi vector DNA into the embryo, thereby providing a consistent Lhx1/Lhx5 silencing effect than has been previously shown. We found that silencing Lhx5 expression, as early as Hamburger and Hamilton's (HH) stage 8, inhibits NR formation, resulting in the conversion of the NR region into a pigment vesicle-like tissue.

Materials and methods

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Concluding remarks
  8. Acknowledgements
  9. References

In ovo electroporation of RNAi vectors

Chicken eggs (Goto Co., Gifu, Japan) were used for Lhx1/Lhx5-RNAi experimentation. Embryos were grown in a humidified incubator at 38°C for 30 h. Thereafter, these were staged according to Hamburger and Hamilton (Hamburger & Hamilton 1992) and harvested after a specified period of time post-fertilization. Subsequently, the eggs were used for in ovo electroporation. The pRFPRNAiA vector (Das et al. 2006; Mende et al. 2008), which includes Lhx1 or Lhx5 short hairpin RNA target sequence (cLhx1-1/pRFPRNAi or cLhx5-1/pRFPRNAi; Kawaue et al. 2012), was electroporated into the right prospective forebrain region at HH stage 8. A sharpened tungsten needle (CUY614T; Unique Medical Imada, Miyagi, Japan) was used as a cathode and inserted into the center of the prospective forebrain region (Fig. 1). A platinum wire electrode (1 mm, CUY611P3-1; Unique Medical Imada), which acted as an anode, was placed diagonally to the cathode electrode. Placing the anode more posteriorly to the embryo resulted in efficient gene transfer. After the DNA solution (50 nL) with fast green (0.1%) was injected into the forebrain of the embryo, electric pulses (50 V, poration pulse, 5-msec duration, 1 pulse; 8 V, driving pulse, pulse width 30-msec duration, pulse interval 50-msec, four pulses) were applied by a pulse generator CUY21EX (BEX, Tokyo, Japan). Windows of these eggs were covered with parafilm and placed at 38°C until desired stages.

image

Figure 1. Electroporation targeting the presumptive forebrain region in the chicken embryo. (A) Old electroporation method. Electroporation of the stage 10 optic vesicle. (B) Electroporation of a stage 8 forebrain region. A tungsten needle electrode (cathode) and a platinum wire electrode (anode) were used to achieve region-specific gene knockdown. Placing the anode more posteriorly to the embryo enhances the effect of Lhx1- or Lhx5-RNA interference. Injected DNA solution is shown in green. (C) Antero-lateral view of the embryo with both electrodes. The cathode pierces the anterior neural fold while the anode is vertical to the vitelline membrane.

Download figure to PowerPoint

Immunostaining, hematoxylin-eosin (HE) staining, and terminal deoxynucleotidyl transferase-mediated dUTP nick end labeling (TUNEL) staining

Embryos were fixed in 4% paraformaldehyde/phosphate buffer saline (PBS) on ice for 1 h. They were washed in PBS. Embryos were then equilibrated in 15% sucrose in PBS until they sank, and stored overnight in 30% sucrose in PBS, at 4°C. Embryos were viewed under a stereomicroscope to determine the morphology of the retina. After observation, fixed tissues were frozen in optimal cutting temperature compound (Sakura, Tokyo, Japan), and sectioned at 14 μm thickness. Sections were used for immunostaining and HE staining. For immunostaining, sections were treated with PBS containing 0.2% Triton X-100 (PBST) for 3 min at room temperature three times. Subsequently, they were incubated overnight at 4°C in the appropriate primary antibody diluted in PBS containing 5% normal goat serum. The antibodies used were as follows: anti-Sox2 (Merck Millipore, Billerica, MA, USA: AB5603, rabbit immunoglobulin G [IgG], diluted at 1:1000), anti-HuC/D (Life Technologies, Carlsbad, CA, USA: 16A11, mouse IgG, 1:500), anti-Otx2 (Abcam, Cambridge, UK: ab21990, rabbit IgG, 1:1000). Sections were rinsed in PBST and incubated with Alexa Fluor 488-conjugated secondary antibody (Life Technologies, 1:250) at room temperature. After washing, the sections were mounted in Vectashield containing 4′,6-diamidino-2-phenylindole (DAPI: Vector Laboratories, Peterborough, UK). For histology, sections were washed three times in PBST, and HE staining was performed according to the standard procedure. TUNEL staining was performed using a Click-iT TUNEL Assay (Life Technologies) as previously described (Kawaue et al. 2012).

Results

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Concluding remarks
  8. Acknowledgements
  9. References

Electroporation of the RNAi vector against Lhx1 into the forebrain region

We previously reported that normal Lhx1 and Lhx5 expression is observed in the developing forebrain from HH stage 8 (Kawaue et al. 2012). We have also shown that overexpression of Lhx1 or Lhx5 induced the ectopic expression of NR differentiation markers (Kawaue et al. 2012). Therefore, it seemed likely that the activity of Lhx1 and Lhx5 in the forebrain might have a role in promoting NR identity. To address this possibility, a vector for RNAi against Lhx1 or Lhx5 was introduced into the forebrain region by in ovo electroporation at HH stage 8 of the chick embryo. We first performed Lhx1-RNAi by delivering the pRFPRNAi vector containing the target messenger RNA (mRNA) sequences into the chick embryos (n > 30) using a new pulse-pattern generator (see 'Materials and methods') and a modified electroporation method (Fig. 1). We found that placing the anode diagonally to the cathode, that is, more posteriorly to the embryo, suppressed the expression of Lhx1 or Lhx5 efficiently than previous methods. After 48 h of electroporation, we observed the embryos and found that over 90% of them exhibited severe small eye phenotypes (class III, > 61% reduction according to Kawaue et al. 2012).

Retinal development after Lhx1-RNAi

At 96 h incubation after electroporation, we examined morphology of the developing eye. In most cases, the operated eye was smaller and darker than that on the contralateral side (Fig. 2). Histological analysis revealed that the Lhx1-RNAi eye had one cell layer of presumptive pigmented epithelium in the NR region. Furthermore, there were some ectopic pigments in the induced presumptive pigmented epithelium (Fig. 3F, arrowheads). Thus, Lhx1-RNAi inhibited the formation of NR. No NR morphogenesis was detectable in the thinner retina of Lhx1-RNAi embryos (Fig. 3).

image

Figure 2. Lhx1-RNA interference (RNAi) inhibits normal eye formation. Lateral (A, B, D, E) and frontal (C) views of the chicken embryonic head at stage 29. (A, B) On the control (contralateral) side, normal eye morphology and pigmentation are observed. (C–E) Compared with the control (left) side, a small eye with high pigmentation is formed on the Lhx1-RNAi (operated) side.

Download figure to PowerPoint

image

Figure 3. Histology of the eye after Lhx1-RNA interference (RNAi). (A-F) Hematoxylin-eosin staining at stage 29 (96 h post-electroporation). (B, C) Magnified view of the posterior retina on the contralateral side (A). (C) Distinct pigmentation is observed on the basal side of the single-layered retinal pigmented epithelium (RPE) (black arrowheads), but not in the neural retina (NR). (E, F) Magnified view of the posterior retina on the Lhx1-RNAi side (D). (D, E) Small eye is observed on the Lhx1-RNAi side, in which a pigmented vesicle forms and normal lens development is perturbed. (F) Intense pigmentation is preserved on the basal side of the RPE (black arrowheads). Ectopic pigmentation is observed in the induced pigmented epithelium (white arrowheads). Scale bar 100 μm.

Download figure to PowerPoint

The effect of Lhx1-RNAi on NR formation

Considering that the retina after Lhx1-RNAi resulted in morphological abnormalities, that is, loss of retinal cyto-differentiation, we tested whether Lhx1-RNAi converted the identity of NR cells to the RPE fate. The antibodies against early neural differentiation markers Sox2 and HuC/D, and an RPE marker Otx2 were used. Strong expression of HuC/D (Fig. 4A,B) or Sox2 protein (Fig. 4E,F) was detected in the NR on the contralateral side, and Otx2 protein was detected in the RPE (Fig. 4I,J). In contrast, Lhx1-RNAi retinal cells exhibited intense staining for Otx2 protein in the NR region (Fig. 4K,L), whereas no staining of the NR marker HuC/D or Sox2 was detected in the same region (Fig. 4C,D,G,H).

image

Figure 4. Immunostaining for retinal differentiation markers of a stage 29 eye after Lhx1-RNA interference (RNAi). (A, B, E, F, I, J) Contralateral control eye. (C, D, G, H, K, L) After Lhx1-RNAi. An RNA-binding protein HuC/D (A-D) and Sox2 (E-H) are used as neural retina (NR) differentiation markers. (I-L) Otx2 is used as a retinal pigmented epithelium (RPE) differentiation marker. Signals are shown in green and nuclei are stained with 4′,6-diamidino-2-phenylindole (DAPI) (blue). Scale bar 100 μm.

Download figure to PowerPoint

Retinal development post Lhx5-RNAi

At 96 h after electroporation, distinct pigmentation was present in the normal embryonic eye. After Lhx5-RNAi, as observed after Lhx1-RNAi, the operated eye was small and highly pigmented, when compared with the contralateral eye (Fig. 5). Histological analysis revealed that the Lhx5-RNAi eye exhibited a thin optic cup with pigmentation (Fig. 6). The inner layer of the cup (i-RPE in Fig. 6D-F) was a pseudo-multi-layered retina with pigmentation (Fig. 6F; arrowhead). Thus, Lhx5-RNAi inhibited formation of NR as was seen in Lhx1-RNAi.

image

Figure 5. Lhx5-RNA interference (RNAi) causes small eye formation. Lateral (A, B, D, E) and frontal (C) views of the chicken embryonic head at stage 29. (A, B) On the control (contralateral) side, normal eye morphology and pigmentation are present. (C-E) A small eye with high pigmentation is present on the Lhx5-RNAi (operated) side.

Download figure to PowerPoint

image

Figure 6. Histology of the eye after Lhx5-RNA interference (RNAi). (A-F) Hematoxylin-eosin staining at stage 29 (96 h post-electroporation). (B, C) Magnified view of the posterior retina on the contralateral side (A). (C) Pigmentation is observed on the basal side of the single-layered RPE (black arrowheads), but not in the NR. (E, F) Magnified view of the posterior retina on the Lhx5-RNAi side (D). (D, E) Small eye is observed on the Lhx5-RNAi side, and the eye exhibits a narrow cup-shape. (F) Ectopic pigmentation is observed in the induced pigmented epithelium (white arrowheads). i-RPE, induced retinal pigment epithelium; LE, lens epithelium; NR, neural retina; RPE, retinal pigment epithelium. Scale bar100 μm.

Download figure to PowerPoint

Are the effects of Lhx5-RNAi the same as those of Lhx1-RNAi?

Lhx5-RNAi resulted in defects in the formation of NR and loss of cyto-differentiation. As such, we tested whether the identity of authentic NR cells shifted to the RPE cells after Lhx5-RNAi. As mentioned, the phenotype after Lhx5-RNAi exhibited a small eye on the operated side. We performed TUNEL staining to determine whether apoptosis was induced by the RNAi or electroporation, resulting in a small eye or disruption of eye development (Mende et al. 2008). We confirmed that no apoptosis was evident in the RNAi OV, while it was observed in the hindbrain region where apoptosis normally occurs (Fig. S1).

We next examined the embryos by comparing marker expression between the operated side and contralateral side. Sox2, HuC/D and Otx2 were used. On the operated side, retinal cells of the Lhx5-RNAi embryo exhibited intense staining for Otx2 in the peripheral NR (Fig. 7I–L). In contrast, no staining for Sox2 or HuC/D was observed in the same regions (Fig. 7A–H).

image

Figure 7. Immunostaining for retinal differentiation markers of a stage 29 eye after Lhx5-RNA interference (RNAi). Signals are shown in green and nuclei are stained with DAPI (blue). (A, B, E, F, I, J) Contralateral control eye. (C, D, G, H, K, L) After Lhx5-RNAi. HuC/D (A–D) and Sox2 (E–H) are used as NR differentiation markers. (I–L) Otx2 is used as a retinal pigment epithelium (RPE) differentiation marker. Scale bar 100 μm.

Download figure to PowerPoint

Discussion

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Concluding remarks
  8. Acknowledgements
  9. References

Lhx1 and Lhx5 are LIM-homeodomain (LIM-HD) transcription factors (Taira et al. 1994; Shawlot & Behringer 1995; Hobert & Westphal 2000; Poché et al. 2007; Suga et al. 2009). We have previously showed that the LIN-11 class LIM-HD factors, Lhx1 and Lhx5 can induce the expression of NR differentiation markers and induce nascent OV cells to form the NR (Kawaue et al. 2012). In this study, after Lhx1-RNAi or Lhx5-RNAi, small eye phenotypes can be observed on the operated side. In addition, these small eyes are highly pigmented when compared with the contralateral eye. Histological analysis revealed that the NR region after Lhx1-RNAi exhibited a single layer of a pigmented retina, which is different from the normal NR. In contrast, the NR region after Lhx5-RNAi still exhibited a pseudo-multi-layered retina, with pigmentation. We tested whether the Lhx1 or Lhx5 RNAi converted the identity of NR cells to the RPE fate, using the neuronal differentiation markers Sox2 and HuC/D (Marusich et al. 1994; Hyer et al. 1998; Taranova et al. 2006;Ishii et al. 2009), and an RPE differentiation marker Otx2 (Crossley et al. 2001; Martínez-Morales et al. 2003; Nishihara et al. 2012). In RNAi-treated eyes, neither Sox2 nor HuC/D protein was detected in the authentic NR region. Furthermore, Lhx1-RNAi induced the expression of Otx2. Compared with Lhx1-RNAi, minor Otx2 protein was observed in the NR region after Lhx5-RNAi.

The earlier manipulation of embryos, a new electroporator and effective electrode placement enhanced the efficiency of in ovo RNAi

Based on the onset and domains of Lhx1 and Lhx5 expression patterns (Kawaue et al. 2012), we determined that placing the anode diagonally to the cathode and performing electroporation as early as HH stage 8 was optimum for suppressing the expression of Lhx1 or Lhx5 (Fig. 1). In addition, a modern electroporator with multi-pulse patterns allowed the more effective introduction of DNA into the chicken embryonic cells. This shows that the expression of Lhx1 in the prospective forebrain region may control NR differentiation at HH stage 8. However, the expression of Lhx1 post HH stage 9 has less impact on NR differentiation.

Lhx5 patterns the NR before OV development

The LIM-homeobox gene Lhx5 play roles in specifying the forebrain architecture (Sheng et al. 1997) and can mediate neuronal differentiation (Zhao et al. 1999). Therefore, we wished to determine whether Lhx5 plays a role in NR differentiation. In previous reports, we described the expression pattern of Lhx5 (Kawaue et al. 2012). Based on this data, we performed Lhx5-RNAi by electroporation under the same conditions with Lhx1-RNAi. At 96 h after electroporation, the majority of the embryos exhibited small eyes with excessive pigmentation, as is found in Lhx1-RNAi; however, the morphology and characteristics of the small eye were different from that of Lhx1-RNAi embryos. In the NR region, Sox2 and HuC/D protein disappeared, but the retina did not become a single layer of pigmented epithelium; however, a pseudo-multi-layered retina was seen in the neuroepithelium. Moreover, ectopic Otx2 protein was only detected in the periphery of the retina. We think that this is because the expression domain of Lhx5 is larger than that of Lhx1. Therefore, Lhx5-RNAi may not confer a complete silencing effect as Lhx1-RNAi and it is likely that the partially reduced expression of Lhx5 may not have fully induced NR differentiation. Instead, the residual mRNA of Lhx5 may antagonize the action of the extraocular mesenchyme that patterns the RPE (Fuhrmann et al. 2000). Although Lhx5-RNAi can be further optimized, this is the first report describing how the normal expression level of Lhx5 in the forebrain is required for NR development of the optic primordium.

Concluding remarks

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Concluding remarks
  8. Acknowledgements
  9. References

In the present study, we have shown that Lhx1 and Lhx5 expression is required to specify the neural retina domain in the optic vesicle. Since Lhx1 is induced by Lhx5 (Kawaue et al. 2012), it is conceivable that the outcome of Lhx5-RNAi treatment is similar to that of Lhx1-RNAi. It is possible that Lhx5 and Lhx1 influence NR differentiation by inducing the expression of diffusible factors, such as FGF (Vogel-Höpker et al. 2000; Martinez-Morales et al. 2005; Okamoto et al. 2009) and Wnt (Holliday et al. 1997). Subsequently, these mediators may induce the expression of early neuronal transcription factors such as Sox2 and Chx10 (Pittack et al. 1997; Zhao et al. 2001; Taranova et al. 2006). Simultaneously, RPE-specific transcription factors, such as Otx2 and Mitf, are induced by TGF-β-like signals from the extraocular mesenchyme. In normal eye development, Chx10 and Mitf repress each other's expression (Fuhrmann et al. 2000). In this report, we suggest that Lhx5 is necessary for NR development and the expression level of Lhx5 required to induce NR differentiation is higher than that of inhibiting RPE differentiation. However, diffusible mediators that link between Lhx5/Lhx1 transcription factors and NR development are elusive, and these important issues should be examined in future studies.

Acknowledgements

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Concluding remarks
  8. Acknowledgements
  9. References

We thank Thomas Jessell for providing the chicken Lhx1 complementary DNA. We also thank Stuart A. Wilson for the generous gift of the pRFPRNAi vector. Monoclonal antibodies were obtained from the Developmental Studies Hybridoma Bank developed under the auspices of the NICHD and maintained by the Department of Biology, the University of Iowa. This work was supported by JSPS KAKENHI Grant Number 24500382.

References

  1. Top of page
  2. Abstract
  3. Introduction
  4. Materials and methods
  5. Results
  6. Discussion
  7. Concluding remarks
  8. Acknowledgements
  9. References