SEARCH

SEARCH BY CITATION

Keywords:

  • competence;
  • cornea;
  • lens regeneration;
  • pax6;
  • transdifferentiation;
  • Xenopus laevis

Abstract

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

After lentectomy, larval Xenopus laevis can regenerate a new lens by transdifferentiation of the outer cornea and pericorneal epidermis (lentogenic area). This process is promoted by retinal factor(s) accumulated into the vitreous chamber. To understand the molecular basis of the lens-regenerating competence (i.e. the capacity to respond to the retinal factor forming a new lens) in the outer cornea and epidermis, we analysed the expression of otx2, pax6, sox3, pitx3, prox1, βB1-cry (genes all involved in lens development) by Real-time RT-PCR in the cornea and epidermis fragments dissected from donor larvae. The same fragments were also implanted into the vitreous chamber of host larvae to ascertain their lens-regenerating competence using specific anti-lens antibodies. The results demonstrate that there is a tight correlation between lens-regenerating competence and pax6 expression. In fact, (1) pax6 is the only one of the aforesaid genes to be expressed in the lentogenic area; (2) pax6 expression is absent in head epidermis outside the lentogenic area and in flank epidermis, both incapable of transdifferentiating into lens after implantation into the vitreous chamber; (3) in larvae that have undergone eye transplantation under the head or flank epidermis, pax6 re-expression was observed only in the head epidermis covering the transplanted eye. This is consistent with the fact that only the head epidermis reacquires the lens-regenerating competence after eye transplantation, forming a lens following implantation into the vitreous chamber; and (4) in larvae that have undergone removal of the eye, the epidermis covering the orbit maintained pax6 expression. This is consistent with the fact that after the eye enucleation the lentogenic area maintains the lens-regenerating competence, giving rise to a lens after implantation into the vitreous chamber. Moreover, we observed that misexpression of pax6 is sufficient to promote the acquisition of the lens-regenerating competence in flank epidermis. In fact, flank epidermis fragments dissected from pax6 RNA injected embryos could form lenses when implanted into the vitreous chamber. The data indicate for the first time that pax6 is a pivotal factor of lens-regenerating competence in the outer cornea and epidermis of larval X. laevis.


Introduction

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

Many studies on Xenopus laevis embryo have defined steps in lens formation during eye development (Henry & Grainger, 1990; Servetnick & Grainger, 1991; Grainger, 1992; Grainger et al. 1997). The early phase begins at the mid-gastrula stage when the ectoderm becomes competent to respond to lens-inductive signals. During the late gastrula and neurula stages, a planar signal from the presumptive retinal area causes a lens-forming bias in the presumptive lens epithelium (PLE). At the neural tube stage, the optic vesicle provides the final signal for lens specification. Later, during the tail bud stages, the eye cup and lens induce the overlying ectoderm to become the transparent corneal epithelium (Lopashov & Stroeva, 1961; Hay, 1980).

Larval X. laevis is provided with the capacity to regenerate a new lens after lentectomy (Freeman, 1963). Lens regeneration occurs through transdifferentiation of the outer cornea into lens. This process is promoted by the inductive factor produced by the neural retina and accumulated in the vitreous chamber (Filoni et al. 1982; Bosco et al. 1993, 1997). In X. laevis, the lens-regenerating competence extends from the outer cornea to the pericorneal epidermis (Freeman, 1963; Filoni et al. 1980), but is completely absent in epidermal territories outside the lentogenic area (the area extending for twice the diameter of the eye cup, according to Freeman, 1963). Thus fragments of outer cornea or pericorneal epidermis implanted into the vitreous chamber transdifferentiate into lens (Reeve & Wild, 1978; Filoni et al. 1980; Bosco & Filoni, 1992), whereas implants of epidermis outside the lentogenic area maintain their epidermal phenotype (Bosco & Filoni, 1992; Cannata et al. 2003).

Embryological manipulations have demonstrated that in the embryo the lens-forming competence (i.e. the capacity to respond to embryonic inductive signals) in the ectoderm is controlled by a developmental timer which is independent of tissue interactions (Grainger, 1992; Ogino & Yasuda, 2000). However, in X. laevis larva, the lens-regenerating competence (i.e. the capacity to respond to the retinal factor present in the vitreous chamber forming a new lens) in the lentogenic area is the result of both early signals, from the presumptive neural retina, linked to the acquisition of lens-forming bias during eye development, and to late signals, from the eye, linked to corneal induction (Cannata et al. 2003; Arresta et al. 2005).

In the early embryo (gastrula to neural tube stages), the capacity to respond to the retinal inducer in the vitreous chamber of a host larva is present not only in the presumptive lentogenic area, but also in the head and flank ectoderm (Henry & Mittleman, 1995). During further development, the lens-regenerating competence in the head ectoderm outside the lentogenic area is lost at the early larval stages and is strongly reacquired by auto-transplantation of an eye under the head epidermis (Arresta et al. 2005). However, the lens-regenerating competence in the flank ectoderm is lost at the embryonic stage 30/31. Nor is it reacquired by transplantation of an eye beneath the flank ectoderm in embryos from stage 30/31 onward (Arresta et al. 2005).

Although it has been ascertained that the expression of several homeobox transcription factors is correlated with different phases of lens induction (see Henry et al. 2002 for a review), to date the molecular basis for the maintenance of the lens-regenerating competence in the lentogenic area of larval X. laevis is unknown, as is that for the reacquisition of this competence in the epidermis outside the lentogenic area.

In this study we used the Real-time RT-PCR technique to analyse the expression of genes involved in lens development, such as otx2, pax6, sox3, pitx3, prox1, βB1-cry, in the cornea and epidermis fragments of normal X. laevis larvae and experimental larvae submitted to eye enucleation or eye transplantation. The same fragments were also implanted into the host larvae vitreous chamber to ascertain their lens-regenerating competence. Among the studied genes, only pax6 turned out to be correlated to lens-regenerating competence. In fact, we observed that epidermal areas maintaining or reacquiring lens-regenerating competence were characterized by pax6 expression or pax6 re-expression, respectively. Epidermal areas, lacking lens-regenerating competence and unable to reacquire this competence following eye transplantation, never expressed pax6.

Moreover, we observed that misexpression of pax6 is sufficient to promote the acquisition of the lens-regenerating competence in flank epidermis. Thus our data indicate for the first time the involvement of pax6 in the lens-regenerating competence of the outer cornea and epidermis of larval X. laevis.

Materials and methods

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

Experimental analysis of lens-regenerating competence: implantation of outer cornea and epidermal fragments dissected from donor larvae into the vitreous chamber of host larvae

The X. laevis larvae were obtained by hormone-induced mating and were staged according to the normal table of Nieuwkoop & Faber (1956). In preparation for surgery, the larvae were anaesthetized in 1 : 3000 MS 222. During the operation, the animals were kept in full-strength Holtfreter's solution. After the operation, the larvae were gradually transferred to dechlorinated tap water, kept at a temperature of 23 ± 1 °C and fed on powder nettles.

In each of experiments I–VII, 45 fragments of outer cornea, pericorneal epidermis, epidermis covering the enucleated orbit, head epidermis, head epidermis covering a transplanted eye, flank epidermis and flank epidermis covering a transplanted eye were dissected from stage 53 larvae (Fig. 1). Fifteen fragments were implanted into the vitreous chamber of host larvae at stage 55, while the remaining 30 fragments were collected in three different pools (10 fragments each) for Real-time PCR analysis (see below). In experiment III, the donor larvae derived from embryos submitted to optic vesicle removal at stage 20 (according to Cannata et al. 2003), were allowed to attain stage 53. In experiments V and VII, the right eye of stage 44 larvae was transplanted beneath the epidermis covering the forebrain or beneath the flank epidermis of stage 45 host larvae, respectively (according to Arresta et al. 2005). The operated larvae were allowed to reach stage 53.

image

Figure 1. Experiments I–VII. Fragments of outer cornea or epidermis were dissected from stage 53 larvae and implanted into the vitreous chamber of stage 55 host tadpole or collected for analysis by Real-time RT-PCR (Q RT-PCR). I, Outer cornea. II, Pericorneal epidermis. III, Epidermis covering the enucleated orbit. IV, Head epidermis. V, Head epidermis covering a transplanted eye. VI, Flank epidermis. VII, Flank epidermis covering a transplanted eye. Corneal and epidermal fragments were implanted together with a Millipore filter disk (f).

Download figure to PowerPoint

Tissue fragment implantations were performed according to Arresta et al. (2005). Briefly, tissue fragments were pushed against a Millipore filter disk (0.6 mm in diameter) and the disk with the adhering tissue was implanted into the vitreous chamber of the lentectomized right eye of stage 55 host larvae. The filter disk was inserted into the pupil to keep the implant inside the vitreous chamber and prevent lens regeneration from the outer cornea of the host larva (according to Cioni et al. 1982).

Six days after implantation, the host larvae were fixed in Carnoy's liquid. Their right eyes were dehydrated, embedded in paraffin, and cut into 7-µm-thick serial sections. Dewaxed sections were hydrated and incubated with a rabbit anti-lens polyclonal antibody (pAbL, 1 : 50) (Filoni et al. 1995). Secondary antibody HRP conjugated goat anti-rabbit (1 : 200, BioRad) was then applied and the immunoreactions were developed with 3-amino-9-ethylcarbazole as substrate (AEC, Sigma). Sections were then counterstained with hematoxylin.

In further cases (15 cases in each of experiments I–VII), the donor larvae were repeatedly injected (intraperitoneal injection, 3 µL every 24 h) with 5-bromodeoxyuridine (5-BrdU, Amersham Int.), 3 days before the dissection of tissue fragments and implantation into the vitreous chamber of not injected host larvae. Dewaxed sections of the right eyes of the host larvae killed 6 days after implantation, were tested for detection of both lens proteins and incorporated BrdU, using pAbL and a mouse monoclonal antibody anti-BrdU (Amersham Int.). Briefly, after incubation with pAbL, HRP-conjugated goat anti-rabbit antibody (BioRad) was applied and immunoreactions were developed with AEC as previously described. After washing in phosphate-buffered saline (PBS), the same sections were then incubated with the anti-BrdU monoclonal antibody. The detection of the bound antibody was achieved using HRP-conjugated anti-mouse antibody (1 : 200, BioRad) and polymerizing diaminobenzidine in the presence of nickel and cobalt. Sections were then counterstained with hematoxylin.

In all experiments, implants were scored positive for lens transdifferentiation only when they reacted positively to pAbL.

pax6 misexpression

pax6-FLAG RNA injection

Xenopus laevis embryos were obtained by standard procedures for artificial fertilization and microinjected in one blastomere at two-cell stage with pax6-FLAG and GFP (Green Fluorescent Protein) capped RNAs according to Smith & Harland (1991). Production of synthetic capped mRNA was obtained using the mMESSAGE MACHINE kit (Ambion) according to the instruction manual. To obtain sense mRNA, xPax-6-FLAG∖CS2++ (Altmann et al. 1997) (gift from C. Altmann and A. Hemmati-Brivanlou) was linearized with HpaI and transcribed with Sp6 RNA polymerase according to Chow et al. (1999). The capped mRNA produced was quantified by gel electrophoresis and OD measurement. Capped RNA was re-suspended in NAM and injected with a Drummond nanoinjector.

In a preliminary experiment, performed to test the effect of different concentrations of synthetic pax6-FLAG capped RNA, some embryos were injected with a dose of 30–1200 pg of pax6-FLAG RNA, allowed to attain the embryonic stage 40–41 and fixed in Dent's fluid. Embryos were processed for whole mount immunolocalization of crystallins (according to Gargioli & Slack, 2004) using the polyclonal antibody pAbL; they were then embedded in paraffin and cut at 7 µm. Sections were stained with hematoxylin and eosin. As most specific eye defects were observed in specimens injected with 120 pg of pax6-FLAG RNA, subsequent experiments were carried out using this concentration.

Implantation of epidermal fragments misexpressing pax6 into the vitreous chamber or into the enucleated orbit of host larvae

Some embryos were injected with 120 pg of pax6-FLAG RNA plus 120 pg of GFP RNA (batch 1), whereas other embryos were injected with 120 pg of GFP RNA (batch 2, controls). All the injected embryos were allowed to attain stage 40–41. Sixty fragments of flank epidermis (30 excised from batch 1 and 30 from batch 2 were collected in six different pools (10 fragments each) for Real-time RT-PCR. In experiments VIII and IX (Fig. 3), 40 flank epidermis fragments dissected from batch 1 and batch 2 embryos, respectively, were implanted into the vitreous chamber of the lentectomized right eye of stage 53 host larvae. Moreover, to be certain that the transdifferentiated implants eventually observed in experiment VIII were not due to auto-differentiation into lens of ventral epidermis fragments expressing pax6, in experiment X fragments of ventral epidermis dissected from batch 1 embryos were implanted into the enucleated orbit of a stage 53 host larva (Fig. 3). In all experiments, epidermal fragments were implanted together with a Millipore filter disk.

image

Figure 3. Implants of flank epidermis fragments dissected from stage 40–41 embryos injected, at two-cell stage with capped RNA. VIII, Implant into the vitreous chamber of flank epidermis derived from pax6-FLAG plus GFP RNAs injected embryos. IX, Implant into the vitreous chamber of flank epidermis derived from GFP RNAs injected embryos. X, Implant into the enucleated orbit of flank epidermis derived from pax6-FLAG plus GFP RNAs injected embryos. Epidermal fragments were implanted together with a Millipore filter disk (f).

Download figure to PowerPoint

Six days after implantation, the host larvae were fixed in Carnoy's liquid. Their right eyes or enucleated orbits were cut into 7-µm-thick serial sections. To label donor tissues expressing GFP, a rabbit polyclonal antibody to GFP (Molecular Probes) was utilized. The localization of the transdifferentiated tissue was achieved using a mouse anti-lens monoclonal antibody (mAbH, whole culture supernatant, Cannata et al. 2003), rather than pAbL, to avoid cross-reaction in double immunostaining. Dewaxed sections were hydrated, incubated with mAbH, and then, after PBS washing, with the anti-GFP (1 : 200). Immunoreactions were developed by using a goat anti-mouse TRITC conjugated (1 : 200, Pierce) and a goat anti-rabbit FITC conjugated (1 : 200, Pierce), respectively. After PBS washing, sections were counterstained with DAPI (4’,6-diamidino-2-phenylindole).

Real-time RT-PCR

For total RNA extraction, tissue pools were processed according to the ‘SV Total RNA Isolation System’ (Promega) and reverse-transcribed using M-MLV RTase (Gibco BRL). Each retro-transcription (RT) reaction was performed on 1 µg of OD quantified total RNA.

The Real time PCR reactions were performed using a LightCycler (Roche). The mastermix of reaction components, the LightCycler experimental run protocol, the melting curve program and the mathematical model for determining the relative gene expression level were performed according to Pfaffl (2001). As positive controls, we performed Real-time RT-PCR on RNAs from extracts of stage 32 embryonic eyes for pax6, otx2 and sox3 (Zygar et al. 1998), prox1 (Schaefer et al. 1999), pitx3 (Khosrowshahian et al. 2005) and βB1-cry (Mizuno et al. 1999a). As negative controls, we performed Real-time PCR on non-retrotranscribed RNAs from embryos and fragments of larval cornea and epidermis. After 40 Real-time PCR cycles, all PCR products were run on agarose gel electrophoresis. Amplified products resulted in a single band with the expected bp number. LightCycler Melting Curve analyses resulted in a single product of specific melting temperatures. No primer dimers were generated during the applied 40 RT PCR amplification cycles.

The following primers were used:

otx2up GGATGGATTTGTTACATCCGTC down CACTCTCCGAGCTCACTTCCC; pax6up GCAACCTGGCGAGCGATAAGC down CCTGCCGTCTCTGGTTCCGTAGTT; pax6-FLAGup gtct accagccaatcccaca down catcgtcgtccttgtactgtaa; sox3up TGATGCAGGACCAGTTGGGC down TGAAGTGAAGGGTCGCTGGC; pitx3up AAGTCCGTTGTCATCACA down CTTCTGGAAAGTGGAGCA; prox1up ACACGAGGTATCCGAGCTC down CTCGTAACGTGATCTGAGCC; βB1-cryup TTTGAACAGGAGAATTTTCA down CAG GATAAACATCTCACC; h4up CGGGATAACATTCAGGGTATCACT down ATCCATGGCGGTAACTGTCTTCC; (pax6, 448 bp*; pax6-FLAG, 399 bp; otx2, 315 bp*; h4, 182 bp*, sox3, 352 bp †; pitx3, 559 bp ‡; prox1, 291 bp‡‡; βB1-cry, 158 bp **; *from Zuber et al. 2003; † from Penzel et al. 1997; ‡ from Khosrowshahian et al. 2005; ‡‡ from Schaefer et al. 1999; ** from Mizuno et al. 1999a).

Results

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

Lens-regenerating competence

Larval outer cornea or epidermis implants into the vitreous chamber

Larval X. laevis lens-regenerating competence can vary significantly during the larval life span and in different batches (Freeman, 1963; Waggoner & Reyer, 1975; Filoni et al. 1997). In experiments I–VII (Fig. 1), the implants into the vitreous chamber of outer cornea and epidermis fragments were performed to ascertain the actual lens-regenerating competence of the same tissue type dissected from stage 53 larvae and collected for Real time RT-PCR.

The immunostaining with pAbL (Table 1; Fig. 2) revealed that implants of outer cornea, pericorneal epidermis, epidermis covering the enucleated orbit and head epidermis covering a transplanted eye all transdifferentiated into lens in high percentages (about 70–80%), very similar to those observed in previous experiments (Bosco & Filoni, 1992; Cannata et al. 2003). By contrast, fragments of head epidermis, flank epidermis and flank epidermis covering a transplanted eye all maintained their epithelial phenotype, confirming that these tissues, in agreement with the findings of Arresta et al. (2005), have no competence to respond to the retinal factor.

Table 1.  Experimental analysis of lens-regenerating competence. Summary of the results obtained after implantation into the vitreous chamber of fragments of outer cornea and several epidermal areas dissected from stage 53 larvae. Six days after operation
Experiment number (Implanted tissue)No. of implantsNo. of implants examinedNo. of Ts* implants (pAbL + )Ts*%
  • *

    Ts, transdifferentiated implants (positive to anti-lens polyclonal antibody).

I (Outer cornea)30302480
II (Pericorneal epidermis)30282279
III (Epidermis covering the enucleated orbit)30282279
IV (Head epidermis)3030 0 0
V (Head epidermis covering a transplanted eye)30302273
VI (Flank epidermis)3030 0 0
VII (Flank epidermis covering a transplanted eye)3030 0 0
image

Figure 2. Experimental analysis of lens-regenerating competence. Cornea and different epidermal fragments dissected from stage 53 larvae were implanted into the vitreous chamber of stage 55 host larvae and fixed 6 days after implantation. Arrows mark implants. Left panels, A–F. Immunoreactions to pAbL anti-lens antibody. (A) Implant of outer cornea (experiment I). (B) Implant of pericorneal epidermis (experiment II). (C) Implant of epidermis covering the enucleated orbit (experiment III). (D) Implant of head epidermis (experiment IV). (E) Implant of head epidermis covering a transplanted eye (experiment V). (F) Implant of flank epidermis covering a transplanted eye (experiment VII). Note that in A–C and E the implants had transdifferentiated into pAbL positive lenses (red). In D and F, the implants had formed a pAbL negative epithelial vesicle. Right panels, G–L. Immunoreactions to both pAbL and to a mAb anti-BrdU. In these cases, the implants were dissected from larvae previously injected with BrdU. (G) Implant of outer cornea (experiment I). (H) Implant of pericorneal epidermis (experiment II). (I) Implant of epidermis covering the enucleated orbit (experiment III). (J) Implant of head epidermis (experiment IV). (K) Implant of head epidermis covering a transplanted eye (experiment V). (L) Implant of flank epidermis covering a transplanted eye (experiment VII). All the implants were BrdU positive (dark brown nuclei), indicating the donor origin of the analysed tissues (white arrowheads mark some labeled nuclei). Note that in G,H,I,K the implants had transdifferentiated into pAbL positive lens cells (red cytoplasm). In J,L implants had formed a pAbL negative epithelial vesicle. f, Millipore filter disk implanted with the tissue fragment; vc, vitreous chamber; nr, neural retina. Scale bars: A–F, 50 µm; G–I,K, 20 µm; J–L, 30 µm.

Download figure to PowerPoint

In the cases where the donor larvae were injected with BrdU, all the implants were BrdU-positive (dark brown nuclei) indicating the donor origin of the implanted tissue (Fig. 2G–L). All the transdifferentiated implants were both pAbL (red cytoplasm) and anti-BrdU (dark brown nuclei) positive (Fig. 2G,H,I,K), indicating that the lens tissue had always derived from the implant.

Gene expression in fragments of larval outer cornea or epidermis

The Real-time RT-PCR for pax6, otx2, sox3, pitx3, prox1 and βB1-cry on RNA extracts (Table 2) revealed that pax6 was expressed in ectodermal tissues with lens-regenerating competence, i.e. outer cornea, pericorneal epidermis, epidermis covering the enucleated orbit and head epidermis covering a transplanted eye. The same gene was not expressed, however, in ectodermal tissues without lens-regenerating competence, i.e. head epidermis, flank epidermis and flank epidermis covering a transplanted eye. An almost undetectable otx2 expression was observed only in the outer cornea, pericorneal epidermis and epidermis covering the enucleated orbit. No expression of sox3, pitx3, prox1 andβB1-cry was observed in any of the examined tissues. Thus a tight correlation between lens-regenerating competence and gene expression was evident only for pax6.

Table 2.  Relative gene expression level in the outer cornea and several epidermal areas dissected from stage 53 larvae. Real-time RT-PCR. The values were calculated by the comparison between the crossing points of the target genes and the reference gene h4. Each value represents the mean ± sd of three Real-time PCR determinations made on three different retrotranscriptions. Each RT derived from 10 samples
Analysed tissueR = 2−Δcp*
otx2pax6Sox3pitx3prox1βB1-cry
  • *

    R = relative gene expression; ΔCP = difference between the crossing point of the target gene and the crossing point of the housekeeping gene h4.

Embryonic eye at stage 31 (Positive control)0.03 ± 0.011.36 ± 0.410.51 ± 0.910.12 ± 0.020.03 ± 0.010.16 ± 0.01
Outer cornea (Exp. I)< 0.010.56 ± 0.080.000.000.000.00
Pericorneal epidermis (Exp. II)< 0.010.43 ± 0.070.000.000.000.00
Epidermis covering the enucleated orbit (Exp. III)< 0.010.49 ± 0.030.000.000.000.00
Head epidermis (Exp. IV)0.000.000.000.000.000.00
Head epidermis covering a transplanted eye (Exp. V)0.000.27 ± 0.040.000.000.000.00
Flank epidermis (Exp. VI)0.000.000.000.000.000.00
Flank epidermis covering a transplanted eye (Exp. VII)0.000.000.000.000.000.00

pax6 misexpression

The Real-time RT-PCR for pax6 and pax6-FLAG from fragments of flank epidermis dissected from stage 41 embryos injected at two-cell stage with pax6-FLAG plus GFP RNA, revealed that pax6 was expressed in flank epidermis (Relative gene expression, R = 0.19 ± 0.02), while pax6-FLAG was not expressed. In flank epidermis fragments, dissected from stage 41 embryos injected at two-cell stage with GFP RNA (controls), pax6 was not detectable. Considering that pax6 RNA autoregulates (Grocott et al. 2007), these data indicate that the injected pax6-FLAG RNA had activated endogenous pax6 expression in flank epidermis.

Implants into the vitreous chamber or the enucleated orbit of flank epidermis fragments dissected from embryos injected with pax6 and GFP RNAs

Six of 40 (15%) fragments of flank epidermis dissected from stage 40–41 pax6 plus GFP RNAs injected embryos and implanted into the vitreous chamber (experiment VIII) underwent transdifferentiation forming lens vesicles with mAbH positive lens fibers (Table 3; Fig. 4A–D).

Table 3. pax6 misexpression. Summary of the results obtained following implantation into the vitreous chamber or the enucleated orbit of flank epidermis fragments dissected from stage 40–41 embryos injected, at two-cell stage, with 120 pg of both pax6-FLAG and GFP RNAs/embryo or 120 pg of GFP RNA/embryo. Immunoreactions to mAbH (anti-lens monoclonal antibody) and to anti-GFP on implants fixed 6 days after operation
RNA injectedExperiment number (Implantation site)Implants examined (N)mAbH+ and anti-GFP+ implants (N)Anti-GFP implants (N)mAbH+ and anti-GFP+ implants (%)
pax6-FLAG + GFPVIII (vitreous chamber)4064015
GFPIX (vitreous chamber)40040 0
pax6-FLAG + GFPX (enucleated orbit)40040 0
image

Figure 4. pax6 misexpression. A–D. Implants of flank epidermis fragments dissected from stage 40–41 embryos injected, at two-cell stage, with 120 pg of pax6-FLAG and GFP RNAs and implanted into the vitreous chamber of stage 55 host larvae (experiment VIII). (E–H) Implants of flank epidermis fragments dissected from stage 40–41 embryos injected, at two-cell stage, with GFP RNAs and implanted into the vitreous chamber of stage 55 host larvae (experiment IX). Implants were fixed 6 days after operation. Double immunofluorescence with an anti-GFP polyclonal antibody and an anti-lens monoclonal antibody (mAbH); sections were counterstained with DAPI. Arrows mark the implants. (A–D) The implant had transdifferentiated into a lens vesicle with mAbH positive lens fibers. (E–H) The implant had formed a mAbH negative hollow epithelial vesicle. (A,E) Anti-GFP and FITC conjugated secondary antibody, green fluorescence; (B,F) mAbH and TRITC conjugated secondary antibody, red fluorescence; (C,G) TRITC/FITC merge; D, H. TRITC/FITC/DAPI merge. Bar: 50 µm.

Download figure to PowerPoint

None of the 40 flank epidermis implants dissected from stage 40–41 GFP RNA injected embryos and implanted into the vitreous chamber (experiment IX) underwent mAbH positive lens transformations. These implants had formed a mAbH negative hollow epithelial vesicle (Table 3, Fig. 4E–H).

All 40 flank epidermis fragments dissected from stage 40–41 embryos injected with pax6 plus GFP RNAs and implanted into the enucleated orbit (experiment X), had formed a mAbH negative epithelial vesicle (Table 3; Fig. 4E–H). This result demonstrates that the lens-forming transformations observed in experiment VIII were actually due to the acquisition of the competence to respond to the retinal inducer forming a lens and not to auto-differentiation into lens of epidermis misexpressing pax6. Thus, misexpression of pax6 can convert flank epidermis from incompetent to lens competent tissue.

Discussion

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

pax6 is assumed to be a ‘master regulator’ gene of eye development (Ashery-Padan & Gruss, 2001) because misexpression of pax6 (or eyeless, a pax6 homologue of Drosophila) results in ectopic eye formation in Drosophila and Xenopus (Halder et al. 1995; Altmann et al. 1997; Chow et al. 1999), indicating that pax6 can initiate the genetic cascade for eye formation in both invertebrates and vertebrates. Tissue recombination experiments and the time course of pax6 expression indicate the involvement of this gene in the induction of lens bias in the PLE of the early embryo (Zygar et al. 1998; Ogino & Yasuda, 2000; Henry et al. 2002). Furthermore, pax6 is continuously expressed throughout the later stages of lens development and is required for the activation of differentiation genes such as crystallins (Piatigorsky, 1998; Cui et al. 2004).

Via in situ hybridization, no expression of pax6 is detected in the outer cornea of the unoperated eye of larval X. laevis (Schaefer et al. 1999; Mizuno et al. 1999b; Cannata et al. 2003). However, after lentectomy of the larval eye or implantation into the vitreous chamber of outer cornea fragments, hybridization reveals pax6 re-expression in the outer cornea during the first few days after operation (Schaefer et al. 1999; Mizuno et al. 1999b; Cannata et al. 2003). Further, pax6 expression is maintained until the beginning of lens-fiber formation. Thus, the data from in situ hybridization indicated that pax6 re-expression was an early marker of the cornea to lens transdifferentiation process.

However, via PCR analysis, Henry et al. (2002), observed pax6 expression in the outer cornea of unoperated larva. Using Real-time RT-PCR, we firstly observed that pax6 transcripts were present not only in the outer cornea but also in the pericorneal epidermis of the normal larva. Since the outer cornea and pericorneal epidermis are the only ectodermal regions of the larva capable of responding to the retinal inducer regenerating a lens, our results suggested the possibility that a basal level of pax6 expression could be related to the lens-regenerating competence. The comparison between the results obtained in experiments I–VII on outer cornea and epidermal fragments implanted into the vitreous chamber and the results of Real-time RT-PCR analysis on the same fragment type demonstrate that a tight correlation between lens-regenerating competence and pax6 expression actually exists. In fact, pax6 expression was absent in head epidermis outside the lentogenic area and in flank epidermis, both incapable of transdifferentiating into lens after implantation into the vitreous chamber. Also, in larvae that have undergone eye transplantation beneath the epidermis of the head or flank, pax6 re-expression was observed only in the head epidermis covering the transplanted eye. This is consistent with the fact that only the head epidermis reacquires the lens-regenerating competence after eye transplantation, forming a lens when implanted into the vitreous chamber. Finally, in larvae that have undergone removal of the eye, the epidermis covering the orbit maintained pax6 expression. This is consistent with the fact that after eye enucleation this epidermis maintains the lens-regenerating competence, giving rise to a lens when implanted into the vitreous chamber.

None of the other genes examined, which are expressed both earlier (otx2) and later (sox3, pitx3, prox1 andβB1-crystallin) than pax6 during eye development, appeared to be correlated with the maintenance or reacquisition of the lens-regenerating competence in the outer cornea or epidermis. Indeed, otx2 has a very low expression in the outer cornea, pericorneal epidermis and epidermis covering the enucleated orbit, although it is not expressed at all in the head epidermis covering a transplanted eye. Aside from this, none of the competent lentogenic areas expresses sox3, pitx3, prox1 andβB1-cry.

In the embryo, at neurula stages, otx2 is one of the earliest genes to demarcate the PLE, during the lens-forming bias (Zygar et al. 1998). pax6 expression occurs in PLE immediately following the decrease in otx2 expression during the late phase of lens-forming bias, whereas sox3, prox1, βB1-cry and pitx3 are expressed during lens specification (Zygar et al. 1998; Ogino & Yasuda, 2000; Khosrowshahian et al. 2005).

Present results indicate that the pattern of gene expression in the larval areas with lens-regenerating competence is similar to that observed during the late phase of the lens-forming bias in PLE. Thus, lens-regenerating competence in the larva seems correspond to a condition similar to that of the biased head ectoderm. The maintaining or reacquisition of pax6 expression could be a pivotal factor for the maintaining and reacquisition of lens-regenerating competence in the ectodermal areas of the larva. Consistent with this view, the results of experiments VIII to X demonstrate that misexpression of pax6 can promote lens-regenerating competence. In fact, at stage 40–41, when flank epidermis of controls had completely lost this capacity after implantation into the vitreous chamber, implants of flank epidermis of stage 40–41 embryos misexpressing pax6 could transdifferentiate into a lens vesicle with mAbH positive lens fibers. As pax6 activates β-crystallin expression in X. laevis animal caps (Altmann et al. 1997), one could object that these results may not be due to a recovery in the competence to respond to the retinal inducer, but, rather, to β-crystallin expression in epidermal cells misexpressing pax6. The results of experiment X demonstrate that this is not the case, because fragments of flank epidermis misexpressing pax6 and implanted into the enucleated orbit, maintained their epithelial phenotype forming simple mAbH negative hollow vesicles.

In conclusion, the experiments I–X demonstrate for the first time that in larval X. laevis pax6 expression and lens-regenerating competence are closely related. In the newt, which is able to regenerate a new lens by transdifferentiation of the dorsal iris, such a correlation has not been observed. In fact, even though it has been established that pax6 controls cell proliferation and lens fiber differentiation in the dorsal iris of lentectomized eyes (Madhavan et al. 2006), in intact eyes pax6 expression has been found not only in the dorsal iris but also in the ventral iris, which is unable to regenerate a lens (Del Rio-Tsonins et al. 1995; Hayashi et al. 2004). This could indicate that the molecular bases of competence in the two regenerative model systems, namely lens regeneration from outer cornea in larval Xenopus laevis and from dorsal iris in the newt, are not the same.

Acknowledgements

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

We are very grateful to R. Altmann and A. Hemmati-Brivanlou for providing the Xenopus CS2-Xpax6-Flag. We also thank Prof. Giulio Cossu for his valuable advice. Grant sponsor: MIUR (Ministero dell’Università e della Ricerca Scientifica); grant number 2004057145_006.

References

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