SEARCH

SEARCH BY CITATION

Keywords:

  • Mitf;
  • Otx;
  • Chx10;
  • Pax6;
  • activin;
  • sonic hedgehog;
  • fibroblast growth factor;
  • neuroepithelial domain specification;
  • evolution

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Developmental origin of vertebrate RPE cells
  5. Signaling molecules in RPE development
  6. Integration of signaling pathways at the transcriptional level
  7. Sorting out neuroretina from RPE: the role of MITF and its cooperators and detractors
  8. The development of the iris and the ciliary body
  9. Sorting out optic stalk from RPE: the roles of Vax genes
  10. The downstream effectors of RPE specification – cell proliferation and differentiation
  11. Thoughts on the differences between melanocytes and RPE cells
  12. Conclusions
  13. Acknowledgements
  14. References

Vertebrate retinal pigment epithelium (RPE) cells are derived from the multipotent optic neuroepithelium, develop in close proximity to the retina, and are indispensible for eye organogenesis and vision. Recent advances in our understanding of RPE development provide evidence for how critical signaling factors operating in dorso-ventral and distal-proximal gradients interact with key transcription factors to specify three distinct domains in the budding optic neuroepithelium: the distal future retina, the proximal future optic stalk/optic nerve, and the dorsal future RPE. Concomitantly with domain specification, the eye primordium progresses from a vesicle to a cup, RPE pigmentation extends towards the ventral side, and the future ciliary body and iris form from the margin zone between RPE and retina. While much has been learned about the molecular networks controlling RPE cell specification, key questions concerning the cell proliferative parameters in RPE and the subsequent morphogenetic events still need to be addressed in greater detail.


Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Developmental origin of vertebrate RPE cells
  5. Signaling molecules in RPE development
  6. Integration of signaling pathways at the transcriptional level
  7. Sorting out neuroretina from RPE: the role of MITF and its cooperators and detractors
  8. The development of the iris and the ciliary body
  9. Sorting out optic stalk from RPE: the roles of Vax genes
  10. The downstream effectors of RPE specification – cell proliferation and differentiation
  11. Thoughts on the differences between melanocytes and RPE cells
  12. Conclusions
  13. Acknowledgements
  14. References

The light-absorbing properties of dark pigments lend themselves ideally to regulate light reception in biology. It is no accident of evolution, therefore, that photoreception in general and that of eyes in particular are commonly found associated with dark pigments such as melanins (for evolutionary considerations, see Box 1). In vertebrates, the photoreceptive retina is abutted on its back not only by a layer of choroidal melanocytes, but also by a specialized pigmented monolayer, the retinal pigment epithelium or RPE. RPE cells are cuboidal cells that on their apical side form multiple villi that make direct contact with the outer segments of the photoreceptor cells. On their lateral walls, they form tight, adherens and gap junctions, and on their basal side, they contact the underlying basal membrane, termed Bruch's membrane, that separates the RPE from the choroid (for review, see Clark, 1986).

Table Box 1..   Photoreception and dark pigment – evolutionary considerations
Charles Darwin, ever on the mark, noted in his famous book that ‘The simplest organ which can be called an eye consists of an optic nerve, surrounded by pigment cells and covered by translucent skin…’ (Darwin, 1886). Indeed, although light reception without associated pigment does exist in extraocular tissues, functional eyes, even the most primitive ones, always come with dark pigment. Typically, this pigment is found in cells that lie adjacent to photoreceptor cells, and in some organisms, as in some crustaceans such as Palaemonetes pugio, it is also found within the photoreceptor cells themselves (Doughtie and Rao, 1984). Even the subcellular light-responsive organelle of the single-celled dinoflagellate Erythropsis pavillardi contains a pigment cup – so suitable is the association of photoreception with dark pigment (Gehring, 2005).
The evolutionary ancient role of this pigment is to form a light screen, shielding photoreceptor cells against light from one direction but not from others. In this way, eyes can see the light and discern the angle where it comes from – a major feat if we think, for instance, of translucent animals such as jelly fish. The shielding role of pigment is preserved from the simplest eye designs in larval trematodes to more complex designs in compound eyes of arthropods or the optically perfected camera eyes of cephalopods and vertebrates. In these advanced designs, the pigment also lends itself to regulate the amount of light that reaches photoreceptors. In compound eyes, this is achieved by moving pigment granules up and down the ommatids, and in the pinhole eye of Nautilus or the lens-containing camera eyes, by the formation of a pupil whose diameter varies with light intensity. As an added benefit, a pupil aids in image resolution, increasing resolution by narrowing in bright light and striking a compromise between resolution and sensitivity by widening in dim light (for details, see Land and Nilsson, 2002).

While choroidal melanocytes are neural crest-derived cells that migrate towards the eye during development, RPE cells are generated directly from the optic neuroepithelium and remain embedded locally in their epithelial environment. Their formation is intimately linked with the development of the eye from its primordium and they continue to fulfill important roles in the adult. During embryogenesis, RPE cells participate in ciliary body and iris formation, control the closure of the optic fissure, influence retinal neurogenesis and ganglion cell projections, and are implicated in the regulation of the choroidal vasculature (for review, see Chow and Lang, 2001; Martinez-Morales et al., 2004). In the adult, they serve as life-long partners of photoreceptor cells by providing nutritional support, forming a blood/retinal barrier, replenishing 11-cis retinal after its photoisomerization to all-trans retinal, controlling ion flow and oxidative damage, and cleaning up the bits of membrane that accumulate at the apex of the photoreceptors’ outer segments. In fact, RPE and retina form a functional unit in which primary abnormalities in one tissue can lead to secondary degeneration of the other (for review, see Strauss, 2005). For all these reasons, RPE cells merit more than cursory mention in vertebrate eye development and function. We here intend to highlight recent insights into their developmental specification and differentiation.

Developmental origin of vertebrate RPE cells

  1. Top of page
  2. Summary
  3. Introduction
  4. Developmental origin of vertebrate RPE cells
  5. Signaling molecules in RPE development
  6. Integration of signaling pathways at the transcriptional level
  7. Sorting out neuroretina from RPE: the role of MITF and its cooperators and detractors
  8. The development of the iris and the ciliary body
  9. Sorting out optic stalk from RPE: the roles of Vax genes
  10. The downstream effectors of RPE specification – cell proliferation and differentiation
  11. Thoughts on the differences between melanocytes and RPE cells
  12. Conclusions
  13. Acknowledgements
  14. References

The vertebrate eye develops from an evagination of the neuroepithelium in the region of the ventral forebrain. As illustrated and described in more detail in Figure 1A for mice, soon after the formation of the evagination, the optic neuroepithelium becomes partitioned into a distal territory which gives rise to the future retina; a proximal territory which gives rise to the optic stalk; and a dorsal territory which gives rise to the RPE. The transition zone between the future retina and the RPE, the ciliary margin zone (CMZ), goes on to form the ciliary body and the iris, and the transition zone between RPE and optic stalk forms the exit point of the optic nerve.

image

Figure 1.  The development of the mouse eye and its territorialization. (A) Overview/fate map: (a–c) Eye development in the mouse starts at around embryonic day (E) 7.5 from an ‘eye field’ of cells in the midline of the neural plate at the level of the prospective diencephalon. This portion of the neuroepithelium evaginates laterally in a dorsal-distal direction, extending towards the surface ectoderm (which is omitted in the figure for clarity). The process generates proximally the optic stalk (which will become the optic nerve) and distally the optic vesicle (which will become neuroretina, RPE, ciliary body, and iris). (a) The optic vesicle is readily visible at E9.5. (b) When the vesicle contacts the surface ectoderm, its distal-most part becomes indented to form the optic cup. The indentation extends towards the ventral part of the optic stalk, generating a fissure along the ventral side of stalk and retina, called the optic fissure. The optic cup is now a highly polarized structure, both in the dorso-ventral and proximal-distal orientation. (c) The fissure progressively seals and is usually closed by E13.5. As a result of this process, the ventral optic stalk epithelium comes to lie inside the stalk where it will generate the astrocytes of the optic nerve. With the closure of the fissure, the cup's outer wall is entirely composed of RPE. This morphogenetic process is further schematically illustrated in the cross-sections through the optic stalk (a′–c′) and the coronal sections through the optic vesicle/optic cup (a′′–c′′). Not shown in these drawings is the process of lens formation from the overlying surface ectoderm. Color code for (A): yellow, future RPE; green, future retina; red: ventral optic stalk and derivatives; gray (in sections): dorsal optic stalk and derivatives. (B) Gene expression during territorialization: The budding optic vesicle expresses many factors that initially overlap (a), but eventually become responsible for distinct eye tissues (b). This requires sorting out their expression patterns into the different domains of the neuroepithelium, a process that is controlled by cell-extrinsic factors such as activins and FGFs that emanate from specific local points and locally impinge on gene expression in the nearby neuroepithelium (see text). For instance, the overlapping expression patterns of MITF, PAX2, and PAX6 at E9.5 (a) are sorted into the expression domains indicated in (b) (see also Bäumer et al., 2003). PAX2, for instance, becomes concentrated in the optic stalk as early as E10.5 and eventually is found in the astrocytes of the optic nerve along with Vax1 (not shown, see text). Equally early, MITF is downregulated in the future retina while PAX6 stays on. In contrast, in the future RPE, where MITF becomes prominent, PAX6 fades away, though only several days later. Hence, the drawing on the right (b) represents not a specific developmental time point but a conceptualization of a dynamic process. Color code for (B): blue-green: overlapping expression of MITF, PAX6, and PAX2 during early vesicle formation. Later stages: colors as indicated in the figure.

Download figure to PowerPoint

As precise as these patterning events are both temporally and spatially, they are not cell-autonomous default pathways but influenced by cell-extrinsic factors. Experimental manipulations in vivo or in explant cultures show that the early optic neuroepithelium is capable of giving rise, in principle, to any of its subdomains (retina, RPE, and optic stalk) from any of its anatomical locations (distal, proximal, dorsal, and ventral) (for review, see Moshiri et al., 2004). In mammals, the associated cell fate changes are limited to early developmental phases. In contrast, in urodeles such as newts and salamanders, and anurans such as frogs and toads, the capacity to switch cell fates in the eye is retained into adulthood. In some of these animals, damaged retinal tissue can be replenished, for instance, from the differentiated RPE or precursor cells present in the CMZ (Del Rio-Tsonis and Tsonis, 2003). Such cell fate switches often are referred to as ‘transdifferentiation’ events, but the term needs careful definition as the mechanisms operating in developing precursor cells are not necessarily equivalent to those operating in adult, terminally differentiated cells.

The patterning of the optic neuroepithelium results from exposure to a variety of extracellular ligands produced either by the neuroepithelium itself or by the surrounding tissues. Of particular relevance are members of the transforming growth factor-β (TGFβ) superfamily of proteins that include bone morphogenic proteins (BMPs); members of the hedgehog (HH) family of proteins; and fibroblast growth factors (FGFs). Each of these ligands modulates the transcription profiles in the developing optic neuroepithelium in a precise, domain-specific manner and in this way controls the correct assembly of an eye (Figure 1B, and see below). For detailed informations on these signaling pathways and their general role in eye development we refer the reader to other recent reviews (Amato et al., 2004; Yang, 2004). Here, we shall focus on the role of the signaling molecules and the associated transcriptional regulatory circuits that control the development of the RPE.

Signaling molecules in RPE development

  1. Top of page
  2. Summary
  3. Introduction
  4. Developmental origin of vertebrate RPE cells
  5. Signaling molecules in RPE development
  6. Integration of signaling pathways at the transcriptional level
  7. Sorting out neuroretina from RPE: the role of MITF and its cooperators and detractors
  8. The development of the iris and the ciliary body
  9. Sorting out optic stalk from RPE: the roles of Vax genes
  10. The downstream effectors of RPE specification – cell proliferation and differentiation
  11. Thoughts on the differences between melanocytes and RPE cells
  12. Conclusions
  13. Acknowledgements
  14. References

Activins promote RPE development

The notion that the mesenchyme surrounding the optic vesicle influences eye development goes back to the earlier part of the last century when it was recognized that a naked eye primordium does not develop properly (Holtfreter, 1939). Direct evidence for the importance of the extraocular mesenchyme for RPE development has first been obtained in chick (Fuhrmann et al., 2000). After removal of the mesenchyme, explanted chick optic vesicles show a downregulation of RPE markers and a concomitant upregulation of neuroretinal markers. Activin A, but not several other BMPs, prevents these changes and hence can substitute for the missing mesenchyme (Fuhrmann et al., 2000). It is unclear, however, which activins are expressed in the surrounding mesenchyme and which ones are genetically required.

The role of hedgehog proteins in RPE development

Hedgehog proteins, originally identified in Drosophila, are secreted proteins with multiple functions in eye development in both vertebrates and invertebrates. In mice, a targeted mutation in the gene encoding sonic hedgehog (Shh) leads to cyclopia characterized by a single midline eye that lacks an optic stalk (Chiang et al., 1996). This demonstrates that SHH plays an important early role in eye field separation and in the formation of proximal eye structures. Hedgehog proteins, however, continue to be expressed later in development, not only in the neuroretina, but also in the RPE. For instance, in zebrafish, shh and tiggywinkle hedgehog (twhh) are expressed in retinal ganglion cells and in the RPE (Neumann and Nüsslein-Volhard, 2000; Stenkamp et al., 2000). In Xenopus, X-shh is found in the RPE, although only transiently (Perron et al., 2003). RPE expression is also seen with banded hedgehog (X-bhh), a homolog of mouse indian hedgehog (Ihh), and cephalic hedgehog (X-chh), a homolog of mouse desert hedgehog (Dhh). These latter HH proteins, however, are absent more peripherally towards the CMZ (Perron et al., 2003; Takabatake et al., 1997). In the mouse, Ihh has been found in the RPE and proximally adjacent to the RPE (Dakubo et al., 2003; Wallace and Raff, 1999). Hedgehog receptors such as patched-1 or patched-2, and the signal transduction protein smoothened are expressed in areas complementary to where the ligands are expressed. For instance, in Xenopus, these molecules are found in the peripheral RPE at locations where HH ligands are missing (reviewed in Amato et al., 2004).

Whether HH signaling plays a role in RPE development has been addressed genetically, by chemical inhibition, by the use of antibodies, and by deliberate exposure of the RPE to ectopic sources of SHH. In Xenopus tadpoles, for instance, cyclopamine treatment, which inhibits HH signaling, leads to a reduction or loss in RPE markers and a reduction in pigmentation at the peripheral RPE (Perron et al., 2003). The cells do not, however, transdifferentiate into neuroretina. In the chick, in contrast, anti-SHH antibodies interfere with ventral RPE formation, concomitant with the loss of RPE markers such as Otx2 that may signal the initiation of a transdifferentiation event (Zhang and Yang, 2001). SHH is also important in the developmental regeneration of the chick retina after experimental retina ablation. Here, the retina is developmentally regenerated by two mechanisms, transdifferentiation of the RPE, and replenishment from the CMZ, both processes involving the action of FGFs. CMZ-directed regeneration can be stimulated by SHH and this stimulation is inhibited by blockers of FGF receptor signaling, indicating that SHH acts through FGF. Intriguingly, however, transdifferentiation of the RPE is inhibited by SHH and favored by the inhibition of HH signaling (Spence et al., 2004). Nevertheless, both mechanisms underscore the importance of the HH pathway in RPE development.

FGFs are negative regulators of RPE development

The recognition that FGFs are important mediators of eye development was originally based on in vitro experiments that were later extended to manipulations in intact embryos. In chick and mouse, FGF1 and FGF2 are expressed at high concentrations in the surface ectoderm overlying the eye primordium while the corresponding receptors are found in the future neuroretina (De Iongh and McAvoy, 1993; Pittack et al., 1997; Tcheng et al., 1994; Wanaka et al., 1991). In addition, the future retina itself expresses FGFs, notably FGF3, FGF8 and FGF15 (Martinez-Morales et al., 2005; McWhirter et al., 1997; Vogel-Höpker et al., 2000). Despite this retinal expression, however, removal of the surface ectoderm, and hence removal of a major source of FGFs, favors a transition of presumptive neuroretina into pigmented RPE-like cells both in the embryonic chick and optic vesicles harvested from mouse embryos; this neuroretina-to-RPE transition is preventable by renewed addition of FGF1 or FGF2 alone. Conversely, FGF exposure of the presumptive RPE favors its development as neuroretina, both in culture (as determined by the acquisition of neuroretinal markers) and in vivo in transgenic mice (Hyer et al., 1998; Mochii et al., 1998; Nguyen and Arnheiter, 2000; Pittack et al., 1997; Zhao et al., 2001). Nevertheless, FGFs show a great degree of redundancy in the eye, and so genetically, no single ligand/receptor pair seems critical. For instance, FGF1/FGF2 double knockout mice show normal eyes (Miller et al., 2000). Implicitly, this finding illustrates that conclusions based solely on embryo manipulations may be misleading. As discussed earlier, that FGF1/FGF2 induce neuroretina formation in experimental situations does not necessarily mean that they are the physiological ligands; they could just as well mimic the action of other factors that induce overlapping downstream signaling pathways (Nguyen and Arnheiter, 2000). Hence, additional genetic models targeting the downstream pathways may be required to gain deeper insights into the relative importance of the multiple ligands that control domain specification in the optic neuroepithelium.

Integration of signaling pathways at the transcriptional level

  1. Top of page
  2. Summary
  3. Introduction
  4. Developmental origin of vertebrate RPE cells
  5. Signaling molecules in RPE development
  6. Integration of signaling pathways at the transcriptional level
  7. Sorting out neuroretina from RPE: the role of MITF and its cooperators and detractors
  8. The development of the iris and the ciliary body
  9. Sorting out optic stalk from RPE: the roles of Vax genes
  10. The downstream effectors of RPE specification – cell proliferation and differentiation
  11. Thoughts on the differences between melanocytes and RPE cells
  12. Conclusions
  13. Acknowledgements
  14. References

The above description of the expression patterns of some of the relevant signaling molecules and their actions confront us with two vexing problems of how extracellular signals are integrated during development. First, although the different types of ligands usually have expression peaks in anatomically distinct locations, they form overlapping gradients. This renders the transition zones between distinct domains particularly interesting battle zones where cells have to interpret colliding signals. Secondly, the notion that at least initially, the optic vesicle displays a high degree of developmental plasticity indicates that each cell in the neuroepithelium can, in principle, respond to each of these ligands. Hence, the cells are initially equipotent. Therefore, mechanisms must be in place that interpret gradients precisely and re-inforce initial fate decisions so that potential temporal fluctuations in the signaling systems do not perturb eye formation too easily. Such developmental robustness or buffering against environmental or genetic perturbations, much the focus of recent studies in development (Kitano, 2004), is thought to result from a variety of mechanisms including feed-back loops, gene redundancies, the action of chaperones, and the multipoint connections in molecular networks. Transcription factors are key components of such networks as they regulate each other's genes as well as different genes and do so both transcriptionally and by direct protein/protein interactions. Here, we shall focus on the transcriptional networks that are involved in the demarcation of the RPE distally against the neuroretina and proximally against the optic stalk.

Sorting out neuroretina from RPE: the role of MITF and its cooperators and detractors

  1. Top of page
  2. Summary
  3. Introduction
  4. Developmental origin of vertebrate RPE cells
  5. Signaling molecules in RPE development
  6. Integration of signaling pathways at the transcriptional level
  7. Sorting out neuroretina from RPE: the role of MITF and its cooperators and detractors
  8. The development of the iris and the ciliary body
  9. Sorting out optic stalk from RPE: the roles of Vax genes
  10. The downstream effectors of RPE specification – cell proliferation and differentiation
  11. Thoughts on the differences between melanocytes and RPE cells
  12. Conclusions
  13. Acknowledgements
  14. References

Neuroepithelial domain specification, be it in the brain or in the eye, follows the common principle of sorting out the expression domains of transcription factors that initially are co-expressed throughout the neuroepithelial sheet. In the eye, for instance, the expression of the paired-domain transcription factor genes Pax2 and Pax6 and the basic helix–loop–helix–leucine zipper (bHLHZip) gene Mitf initially overlap in the optic vesicle (Bäumer et al., 2003). Later, under the influence of cell-extrinsic factors, they become concentrated in different domains: Pax6 in neuroretina, Pax2 in optic stalk, and Mitf in RPE (Figure 1B). Here, we focus on Mitf (Hodgkinson et al., 1993) chiefly because of its central role in RPE formation and because of the availability of a large number of Mitf mutations in species as diverse as fish and man which allow us to gain further insights into RPE development. For details on the genetics, biology and molecular biology of Mitf; however, we refer the reader to several other recent reviews (Arnheiter et al., 2002; Arnheiter et al., 2006; Steingrimsson et al., 2004).

Mitf– a critical eye transcription factor

In 1942, a mutant line of mice was described whose phenotype was characterized by pigmentary abnormalities and small eyes, and so the mutant locus was termed microphthalmia (mi) (Hertwig, 1942). Some 50 yr later, we and others cloned the corresponding gene from transgenic insertional mutations, found it to encode a transcription factor of the bHLHZip class, and named it Mitf (Hodgkinson et al., 1993; Hughes et al., 1993; Tachibana et al., 1994). Mitf is prominently expressed in the RPE (Hodgkinson et al., 1993; Nakayama et al., 1998), and when rendered non-functional by mutation, leads to a hyperproliferating, unpigmented RPE that in its dorsal part assumes a neuroretinal fate (Figure 2). In fact, this dorsal part develops as a correctly layered, though inverted, second retina (Bumsted and Barnstable, 2000; Müller, 1950; Nguyen and Arnheiter, 2000). The RPE abnormalities lead to defects in the closure of the optic fissure and in the expansion of the eye ball and hence to small eyes. Because the ordinary retina continues to proliferate normally in an otherwise too small eye, it develops multiple folds and ultimately degenerates. The result is a complicated colobomatous microphthalmia that severely compromises image formation (Scholtz and Chan, 1987). But how is Mitf regulated in the RPE and which genes does it normally regulate to prevent this outcome?

image

Figure 2.  The role of Mitf in RPE formation. MITF is a basic–helix–loop–helix–leucine zipper transcription factor expressed in the developing RPE. (A and B) Cross-section through the dorsal RPE of an Mitf+/+, albino mouse embryo at E14.5 and its counterpart carrying an Mitf allele encoding a protein that lacks part of the basic domain (Mitfmiew). Both embryo sections were stained with a rabbit anti-MITF antiserum. Brackets mark the RPE monolayer in (A) and the thickened RPE in (B). Note that in the mutant (B), the part that is going to become the ‘transdifferentiated’ second retina (arrow) no longer expresses MITF protein. This second retina soon degenerates after birth, leading to additional abnormalities in the adjacent original retina. (C–E) Sections through the posterior RPE and adjacent retina in 6-week-old wild type (pigmented mouse) and two Mitf mutant compound heterozygotes (for allele description, see Arnheiter et al., 2002; Steingrimsson et al., 2004). In the wild type (C), one normally sees a pigmented RPE sandwiched between the rod outer segments (ros) of the retina and the pigmented choroid. (D) In mice carrying a relatively mild Mitf allele, Mitfmivit, here in combination with a null allele (Mitfmivga−9), individual RPE cells are unpigmented (arrows), while others retain pigment (arrowheads). The choroid, however, is entirely unpigmented. In these mice, the retina progressively degenerates. (E) In mice carrying a more severe Mitf mutation (MitfMior/mivga−9), there is considerable hypercellularity in an RPE that remains totally unpigmented. The adjacent retina is abnormal and lacks lamination.

Download figure to PowerPoint

Negative regulation of Mitf

An initial clue to Mitf regulation in the eye came from studies of FGF signaling in cultured quail RPE when it was found that experimental addition of FGF, which interferes with RPE development, leads to MITF downregulation (Mochii et al., 1998). Similar observations were later made in explanted mouse optic vesicles next to which FGF soaked beads were implanted (Nguyen and Arnheiter, 2000). It is critical to note that in the mouse, before the optic vesicle makes contact with the surface ectoderm, MITF is found in the future retina and in the future RPE and only later is downregulated in the future retina by surface ectodermal signals such as FGFs (Figure 1B and Bora et al., 1998; Nguyen and Arnheiter, 2000). This fact immediately leads to two questions. (1) How do FGFs downregulate MITF – transcriptionally, or post-transcriptionally? (2) Are the same mechanisms responsible for the physiological downregulation of Mitf in the future neuroretina and for the experimental downregulation in the RPE?

Previous in vitro studies indicated that extracellular signal-regulated kinase-2 (ERK2), which is activated by a variety of receptor tyrosine kinases including FGF receptors, phosphorylates MITF at serine-73 and, through the regulation of p90RSK, also leads to phosphorylation at serine-409 (Wu et al., 2000). These phosphorylation events result in a transient increase in MITF's transcriptional activity, followed by a decrease in protein stability. This observation prompted experiments in which an activated MAP/ERK kinase, MEK1, was ectopically expressed in chicken RPE. Activated MEK1 leads to ERK phosphorylation and activation and, as expected, resulted in reduced MITF protein levels and perturbed RPE pigmentation. This finding was consistent with the observation that FGF2 also leads to phospho-ERK in avian RPE cells (Galy et al., 2002). Independent experiments in mouse showed, however, that ectopic FGF leads to Mitf downregulation at the mRNA level (Figure 3), suggesting a role for a transcriptional repressor. Because Mitf alleles encoding non-functional MITF proteins do not lead to an increase in Mitf mRNA levels (Nakayama et al., 1998), an autoregulatory loop is unlikely the explanation for this downregulation. A promising hint about the nature of an Mitf repressor in the eye, however, came from earlier genetic studies in which mice homozygous for a mutation in the paired-like homeodomain transcription factor CHX10 (Chx10or/or) were crossed with Mitfmi/mi mice (Konyukhov and Sazhina, 1966). Both types of mice have small eyes, though the underlying mechanisms causing their microphthalmia are different. The double homozygotes displayed more normal retina and RPE development although the adult eyes were still small and lacked bipolar cells (a particular type of retinal interneurons), similar to those of Chx10or/or single mutants, and the RPE remained unpigmented as in Mitfmi/mi mutants (see Box 2 for the history of these experiments and their interpretations). These findings suggested that CHX10, which is induced by FGF and is prominently expressed and needed for the development of the future neuroretina and the retina derived from the transdifferentiating RPE, might serve as a repressor of Mitf, and that Mitf has a negative activity on retinal development (Horsford et al., 2005; Rowan et al., 2004). This interpretation is supported by recent results which show that CHX10 lies downstream of FGF and upstream of MITF (FGF[RIGHTWARDS ARROW] CHX10 ⊣ MITF ⊣ retina) and which help explain why in the double mutants, both retina and RPE look more normal when compared with the single mutants (Horsford et al., 2005). How FGFs induce CHX10, however, still needs to be explored.

image

Figure 3.  FGF downregulates Mitf at the mRNA level. The frontal part of the heads of E9.5 mouse embryos were explanted. A polyacrylamide bead, coated with bovine serum albumin and labeled ‘control bead’, was implanted on one side near the optic vesicle. Another bead, coated with FGF2 and labeled ‘FGF2 bead’, was implanted close to the optic vesicle on the other side. After 72 h, the heads were fixed and processed for whole mount non-radioactive in situ hybridization using an Mitf RNA probe. Note that on the control side, Mitf mRNA is expressed as expected (dark brown stain) but no Mitf label is seen on the FGF2 treated side.

Download figure to PowerPoint

Table Box 2..   CHX10, MITF and the regulation of eye size – historical and modern concepts
Mice with mutations at the ocular retardation locus (for instance, Chx10or/or) display microphthalmia associated with a normal RPE but a hypoplastic neuroretina which lacks bipolar interneurons. Mice with mutations at Mitf (for instance, Mitfmi/mi) display microphthalmia associated with an abnormal hyperproliferation of the RPE but a normal proliferation of the neuroretina. These Mitf-associated abnormalities then lead to a failure in the closure of the optic fissure, the generation of retinal infoldings, and the formation of an inverted, RPE-derived second retina on the dorsal side. Intriguingly, as originally reported by Konyukhov and Sazhina in 1966 and schematically depicted below, if mice are bred to simultaneous homozygosity for mutations at both loci, the proliferative disturbances seen in the respective single-gene homozygotes are partially corrected in both neuroretina and RPE (Konyukhov and Sazhina, 1966). Konyukhov and Sazhina thought that the Chx10 mutant retina was growing poorly because of overproduction of some retina-derived, growth-inhibitory factor, and that Mitf mutant RPE was hyperproliferating because of overproduction of some RPE-derived, growth-stimulatory factor. They recognized, however, that in the respective single mutants, the hypothetical retinal factor did not reduce RPE proliferation and the RPE factor did not stimulate retinal proliferation. Nevertheless, they thought that once the two mutations were combined, the negative retinal factor would win out against the positive RPE factor in the RPE (although not in the retina itself, as the retina resumes more normal proliferation in the double mutants); and that the positive RPE factor would win out against the retinal factor in the retina (although not in the RPE, as the RPE stops hyperproliferating in the double mutants). Be that as it may, this interpretation implied that the Mitf and Chx10-associated phenotypes were each separately due to the addition, or increase in the level, of a secreted, cell-extrinsic factor.
Alternatively, however, one may also see the phenotypes as resulting from cell-intrinsic defects. The loss of a negatively acting factor in RPE cells, or of a positively acting factor in retinal cells, would likewise account for the observed proliferative abnormalities. Recent research shows that this second view is indeed the correct one. CHX10 is a paired-like homeodomain transcription factor that cell-autonomously suppresses Mitf expression in the future retinal cells (Horsford et al., 2005; Rowan et al., 2004). MITF is a basic helix–loop–helix leucine zipper transcription factor that inhibits, cell-autonomously, proliferation of at least some cell types (Bismuth et al., 2005; Carreira et al., 2005; Loercher et al., 2005). On the one hand, then, when Chx10 is non-functional, there is prolonged expression of the negatively acting Mitf in the retina, hence retinal hypoproliferation. When Mitf is equally non-functional, it can no longer inhibit cell proliferation, hence retinal hypoproliferation is corrected. This correction is only partial, though, as the regulation of Mitf is just one of the pathways by which Chx10 stimulates retinal growth. Because Chx10 is also needed for the formation of the second, RPE-derived retina in Mitf mutants (Horsford et al., 2005), the RPE of such mutants reverts to a more normal proliferative state when Chx10 is equally non-functional. Of course, in the end, the eyes of double mutants are still abnormal as the retinae lack bipolar cells (Chx10 has a later role in cell lineage determination) and the RPE cells remain unpigmented (Mitf regulates genes involved in pigment biosynthesis). Thus, even though the combination of two deleterious mutations can cancel each other out with respect to the regulation of cell proliferation early in development, the need to generate a fully functional eye still puts evolutionary pressure on both genes. This consideration would apply even if making eyes were the only tasks of Mitf or Chx10, but Mitf, for instance, is also critical for coat pigmentation and other cellular functions for which there is selective pressure.
inline image

It is unclear at present whether CHX10 affects Mitf transcription directly or indirectly through other gene products. A direct action is complicated by the observation that Mitf is transcribed from at least nine distinct promoters, of which at least eight are potentially expressed in the optic neuroepithelium (for review, see Arnheiter et al., 2006; Steingrimsson et al., 2004, K. Bharti, unpublished observations). None of these promoters, individually, has been reported to be critical for eye development or to respond to CHX10. Nevertheless, the individual promoters are conserved among different mammals and several of them contain potential CHX10 binding sites (K. Bharti, unpublished observation). Although it is not a priori clear that the FGF-mediated negative regulation of Mitf in neuroretina operates through the same mechanisms as those in RPE, it appears that Chx10 is involved in regulating Mitf in the future RPE as it is in the furture retina. In fact, in Chx10 mutant optic vesicles, implantation of FGF-coated beads can no longer downregulate MITF and inhibit pigmentation (Horsford et al., 2005). This is not to say that Chx10 is the only inhibitor of Mitf expression which is repressed not only in the presumptive retina, but also in the ventral optic stalk. For instance, Vax genes, which are prominently expressed ventrally, also appear to repress Mitf expression (see below).

Positive regulation of Mitf

While the transcriptional downregulation of Mitf helps explain the generation of a neuroretina and optic stalk, positive regulators of Mitf are needed to explain the making of an RPE. It is conceivable that distinct regulators are involved in the initial neuro-epithelial expression of Mitf and its subsequent pre-eminence in the dorsal RPE. An interesting study showed that both Pax6 and Pax2, critical eye transcription factors of the paired-domain class (for review, see Kozmik, 2005) with initial pan-vesicular expression patterns similar to Mitf, can activate at least one of the multiple Mitf promoters in vitro, the A-promoter, which is the 5′ most promoter of the 200 kb Mitf gene (Bäumer et al., 2003). In vivo, however, loss-of-function of either Pax6 or Pax2 alone does not reduce Mitf expression, at least as seen with antibodies recognizing all MITF isoforms, but when both genes are non-functional, Mitf is no longer expressed (Bäumer et al., 2003). A regulation of all Mitf isoforms through the A promoter is intriguing as it may suggest that this promoter acts as a ‘locus control region’. Nevertheless, A-Mitf and other isoforms are expressed as well in various non-ocular tissues and in cell lines that lack Pax6/Pax2 expression. Moreover, other Mitf promoters, conspicuously inactive in most extra-ocular tissues, contain promoter sequences that are compatible with PAX2/PAX6 binding (K. Bharti, unpublished observation). Furthermore, in the developing RPE, where Pax6 is soon downregulated and Pax2 is absent, Mitf eventually reaches its highest level of expression (Nakayama et al., 1998). Lastly, bacterial artificial chromosomes that contain the entire Mitf gene except for its A promoter and associated A exon, are still capable of rescuing eye development in Mitf null mutants (E. Steingrímsson, N. Jenkins and N. Copeland, unpublished observation). Taken together, these findings suggest that (1) Pax2/Pax6 may be involved in the initial onset of Mitf expression; (2) Chx10 wins against Pax2/Pax6 during the phase of rapid downregulation of Mitf in the future neuroretina; and (3) Pax2/Pax6 do not play major roles for the maintenance of Mitf expression in the RPE. Which Mitf promoter(s) these transcription factors target, however, remains to be elucidated.

The search for additional positive regulators has concentrated on other factors expressed in RPE. Among them are the Otx genes (Otx1 and Otx2) which are related to the Drosophila orthodenticle gene and which, like Mitf, are initially expressed in the entire mouse optic vesicle and then become most prominent in the RPE (Martinez-Morales et al., 2001, 2003). Otx2 overexpression in retinal cells results in pigmentation, and targeted Otx1−/−;Otx2/+ mice show an almost complete lack of Mitf expression in the presumptive RPE, a lack of pigmentation, and an RPE-to-retina transition similar to that observed following FGF-mediated Mitf downregulation. Nevertheless, in Mitf mutants, Otx2 is missing, and so it is not possible to simply place Otx genes above Mitf genetically. Rather, the two types of genes may be involved in a mutual regulation and may cooperate to activate downstream pigmentation target genes such as tyrosinase, tyrosinase-related protein-1 (Tyrp1) and dopachrome tautomerase (Dct) (Martinez-Morales et al., 2003, for review, see Martinez-Morales et al., 2004).

A potential role in Mitf regulation has also been suggested for TGFβ-induced transcription factors. TGFβ proteins signal through heterodimeric complexes of membrane spanning serine/threonine receptor kinases which become activated by phosphorylation and in turn phosphorylate specific SMAD transcription factors. Activins bind to activin receptors (termed ActRs or activin receptor-like kinases, Alks) and activate the receptor-activated SMADs (R-SMADs), SMAD2 and SMAD3, which associate with the common-partner SMAD (co-SMAD), SMAD4. R-SMAD/co-SMAD heterodimers then translocate to the nucleus to activate gene transcription (for review, see Massague et al., 2005). Since in the chick, activin A induces Mitf and promotes RPE development (Fuhrmann et al., 2000), it is likely that this activity involves SMAD2 and/or SMAD3. Recent studies in mice, however, show that at least Smad3 is not required for normal RPE development. Nevertheless, it is required for the epithelial-to-mesenchymal transition of RPE cells into α-smooth muscle actin-expressing cells following experimental detachment of the retina in the adult eye (Saika et al., 2004). This indicates that during development, Smad3 is not limiting in the regulation of Mitf but plays a role in the adult RPE, although not necessarily by regulating Mitf.

Chx10, Pax, Otx and perhaps some Smad genes likely represent but a subset of the genes involved in Mitf regulation in the RPE. Mutations in the transcription factor AP-2α, for instance, also result in dorsal RPE-to-neuroretina transitions (West-Mays et al., 1999), conceivably by regulating Mitf levels. Moreover, mutations in genes regulating cell proliferation, such as Gas1, lead to a gradual loss of Mitf expression and retinal transdifferentiation in the ventral RPE (Lee et al., 2001). Future studies using promoter-specific targeted mutations and in vitro analyses will be required to determine which are the relevant Mitf promoters, how the regulatory networks operate, and whether there are variations in these networks between species such as Xenopus, chick, and mouse, in which eye formation shows clear, if subtle, differences.

The development of the iris and the ciliary body

  1. Top of page
  2. Summary
  3. Introduction
  4. Developmental origin of vertebrate RPE cells
  5. Signaling molecules in RPE development
  6. Integration of signaling pathways at the transcriptional level
  7. Sorting out neuroretina from RPE: the role of MITF and its cooperators and detractors
  8. The development of the iris and the ciliary body
  9. Sorting out optic stalk from RPE: the roles of Vax genes
  10. The downstream effectors of RPE specification – cell proliferation and differentiation
  11. Thoughts on the differences between melanocytes and RPE cells
  12. Conclusions
  13. Acknowledgements
  14. References

The sorting out of neuroretina from RPE does not generate a sharp boundary between the two tissues but a transition zone, the CMZ. Interestingly, in fish and amphibians, the CMZ serves as a growth zone for retinal neurons throughout life, and in the post-hatch chick, the CMZ contributes to the generation of at least some types of retinal neurons. In mammals, such as mice and monkeys, however, retinal neurogenesis is completed relatively soon after birth and there is no evidence for a peripheral retinal growth zone (for review, see Moshiri et al., 2004). Nevertheless, in the mouse, even in the adult mouse, the ciliary body still contains stem cells capable of generating, in vitro, retinal neurons including photoreceptor cells (Ahmad et al., 2000; Tropepe et al., 2000). Intriguingly, the number of these stem cells is increased in mutants with fewer retinal or RPE progenitor cells, suggesting a negative feed back loop between progenitors and stem cells (Coles et al., 2006, Dhomen et al., 2006). The CMZ thus represents a distinct zone in the eye that requires special attention.

The CMZ is characterized anatomically by the expression of several marker proteins, including collagen XI, XVIII, laminin α1, β1, and γ1 (which together form the laminin1 trimer), nidogen, and connexin43 (Dong and Chung, 1991; Dong et al., 2002; Halfter et al., 2002; Hyer, 2004; Kubota et al., 2004). Because the CMZ can be molecularly characterized by such markers, one could test whether it is simply the relative levels of neuroretinal and RPE inducers that specify a CMZ, or whether additional, perhaps lens-derived factors are involved as well.

Once specified as CMZ, the corresponding, contiguous epithelium forms the ciliary body and the iris. As described in more detail in Figure 4, the RPE extends into the ciliary body and the back layer of the iris, and the retina into the unpigmented epithelium that covers the extension of the RPE. The front layer of the iris, however, is populated with neural crest-derived pigment cells. Interestingly, in none of these pigmented cells (nor in the remainder of the RPE and choroid) is Mitf expression maintained after birth as determined by non-radioactive in situ hybridization or antibody staining in the mouse (Nakayama et al., 1998). This suggests that maintenance transcription of the developmental Mitf target genes including pigmentation genes may require only very low MITF levels, or may entirely depend on other transcription factors.

image

Figure 4.  Ciliary body and iris formation. The ciliary body and the iris form as a special elongation from the rim of the optic cup, called the ciliary margin zone or CMZ, where the outer layer (RPE) turns around into the inner layer (retina). The picture on the left shows a section through an albino eye of an Mitf+/+ mouse embryo at E16.5. This section was labeled by non-radioactive in situ hybridization for a pigmentation gene, Dct. Note the presence of Dct-positive neural crest-derived cells at the surface and close to the CMZ. The two pictures on the right show sections through adult eyes from wild type and a compound heterozygous Mitf mutant, Mitfmibw/mivga−9. The CMZ has given rise to the ciliary processes which provide the excess tissue needed during closure of the pupil. The extension of the retina forms the unpigmented ciliary epithelium and the extension of the RPE the pigmented back of the iris. The front of the iris is populated with neural crest-derived melanocytes. The distinct origin of these two pigmented layers of the iris can be seen in the mutant eye in which the generation of neural crest-derived melanocytes but not of RPE cells is affected. The irises of such mutant mice have a pigmented layer on the back but lack pigmented cells in the front; here, only unpigmented stromal cells are found. Nevertheless, on visual inspection, the mutant eyes are black, in contrast to the red eyes of albino animals which lack melanin in both layers. Hence, pigmentation of the RPE and the back layer of the iris is sufficient for the visual appearance of a black eye.

Download figure to PowerPoint

Sorting out optic stalk from RPE: the roles of Vax genes

  1. Top of page
  2. Summary
  3. Introduction
  4. Developmental origin of vertebrate RPE cells
  5. Signaling molecules in RPE development
  6. Integration of signaling pathways at the transcriptional level
  7. Sorting out neuroretina from RPE: the role of MITF and its cooperators and detractors
  8. The development of the iris and the ciliary body
  9. Sorting out optic stalk from RPE: the roles of Vax genes
  10. The downstream effectors of RPE specification – cell proliferation and differentiation
  11. Thoughts on the differences between melanocytes and RPE cells
  12. Conclusions
  13. Acknowledgements
  14. References

The development of the RPE not only requires setting precise boundaries on the distal side where the neuroretina develops, but also on the proximal side where the optic stalk will form. In fact, we can distinguish between two boundaries, one between the dorsal RPE and the dorsal optic stalk, and one between the presumptive retina and the ventral optic stalk (see Figure 1A). As in other parts of the developing central nervous system, dorso-ventral polarity is determined by BMPs dorsally and SHH ventrally that repress each other's action (Yang, 2004). Two types of transcription factors seem to be instrumental in this mutual repression in the eye, the T-box protein TBX5 (Koshiba-Takeuchi et al., 2000), induced dorsally by BMP4, and the ventral homeodomain proteins VAX1 and VAX2, induced ventro-proximally by SHH as determined in fish and frog (Sasagawa et al., 2002; Take-Uchi et al., 2003). In the mouse, at around E9.5, Vax1 and Vax2 expression nearly overlap in the ventral optic stalk and optic vesicle. Soon, however, Vax1 fades in the ventral optic vesicle and remains only in the optic stalk while Vax2 becomes confined to the ventral neuroretina (Mui et al., 2002, 2005). Hence, the expression patterns of these genes become complementary to those of Mitf which is retained in the RPE but shows a sharp boundary towards the optic stalk.

These expression patterns are suggestive of mutually repressive interactions between Mitf and Vax. We have recently begun to explore this possibility using mice with targeted mutations in Vax1 and Vax2 and extant mutations in Mitf. Indeed, in compound Vax1/Vax2 mutants, Mitf expression is expanded into the dorsal optic stalk which can lead to its pigmentation (Mui et al., 2005, K. Bharti and S. Bertuzzi, unpublished observations). This result is reminiscent of the observation made in mice in which PAX6 is ectopically expressed under a Pax2 enhancer active in the optic stalk (Schwarz et al., 2000). Here, too, there is ectopic pigmentation in the dorsal optic stalk, suggesting a regulatory link between Pax6 and Vax genes. In fact, Vax loss-of-function mutations lead to an expansion of Pax6 expression into the optic stalk, concomitant with a vast expansion of the neuroretina which becomes gigantic yet well differentiated and laminated (Mui et al., 2005). It is still unclear, however, what additional signals are responsible for the observation that in Vax1/Vax2 compound mutants, the expansion of Mitf expression is restricted to the dorsal stalk where it limits the excessive cell proliferation that is seen in the ventral stalk (Mui et al., 2005). We have also obtained preliminary results suggesting that Mitf mutations may lead to an expansion of Vax1 expression into the ventral RPE (K. Bharti and S. Bertuzzi, unpublished observation). These analyses suggest, then, that Mitf and Vax are engaged in a mutual repression of their expression and that it is this mutual repression which leads to the formation of the sharp boundaries between optic stalk and RPE.

The downstream effectors of RPE specification – cell proliferation and differentiation

  1. Top of page
  2. Summary
  3. Introduction
  4. Developmental origin of vertebrate RPE cells
  5. Signaling molecules in RPE development
  6. Integration of signaling pathways at the transcriptional level
  7. Sorting out neuroretina from RPE: the role of MITF and its cooperators and detractors
  8. The development of the iris and the ciliary body
  9. Sorting out optic stalk from RPE: the roles of Vax genes
  10. The downstream effectors of RPE specification – cell proliferation and differentiation
  11. Thoughts on the differences between melanocytes and RPE cells
  12. Conclusions
  13. Acknowledgements
  14. References

Cell proliferation

Changes in gene expression patterns in the eye primordium give us intimations of the future eye, but it is the changes in anatomy that finally lead to a functioning visual organ. Surprisingly little information is available, however, on how signaling and transcriptional networks are ultimately translated into shape and function.

One of the earliest signs of the temporal progression in eye development is the differential thickening of the future RPE and neuroretina, soon followed by the formation of the optic cup. This difference in thickening reflects differential cell numbers and cell shape, but whether these are produced by differential cell cycle times, rates of cell death, or other mechanisms that differentially add or subtract cells from the distinct epithelial domains has not been widely studied. In zebrafish, the future RPE and NR start out from roughly equal cell numbers, but over a 12-h period (from 15 to 27 h post-fertilization), the RPE loses about two-thirds of its cells while the neuroretina gains about double (Li et al., 2000a,b). Differential cell death cannot account for this but the addition of cells from the RPE to the retina by involution through the ciliary margin can. Furthermore, the time of S-to-M phase, extracted from the percentage of BrdU-pulse-labeled cells that progress to mitosis, shows age-dependent changes (Li et al., 2000a). What drives these changes is unknown, however.

In the mouse, studies on cell proliferation in the developing eye are usually limited to measuring rates of DNA synthesis or mitosis, although a few studies have determined RPE clone sizes by analyzing experimental chimeras or mosaics. Mosaics are produced, for instance, by the spontaneous reversion of the pink-eyed-unstable (pun) mutation which is brought about by intrastrand homologous recombination of a sequence duplication (Gondo et al., 1993) and which generates pigmented clones within an unpigmented RPE (Bodenstein and Sidman, 1987b). From these studies, it appears that soon after domain specification, the rates of cells in S-phase or mitosis in the future RPE are about half of those in the future neuroretina (Bodenstein and Sidman, 1987a,b; Kong et al., 1992; Tang et al., 1996). Unlike in zebrafish, however, there is no contribution of cells from the RPE to the neuroretina in mice.

Several findings point to the importance of Mitf and Chx10 in regulating cell proliferation in the eye. First, mouse Mitf mutations lead to a thickening of the RPE and increased cell numbers as early as E10.5 (Nguyen and Arnheiter, 2000), and similar RPE thickening is seen in mutant quail (Mochii et al., 1998) and hamster (Arnheiter et al., 2002). Secondly, failed downregulation of Mitf in the future neuroretina of Chx10 mutants impairs neuroretinal growth (Horsford et al., 2005; Rowan et al., 2004). Thirdly, in cultured melanocytes or heterologous cells, MITF, at least its (+) ‘melanocyte isoform’ [(+) M-MITF] which is characterized by the insertion of six residues upstream of the basic domain, inhibits cell proliferation (Bismuth et al., 2005; Carreira et al., 2005; Loercher et al., 2005). This inhibition is thought to be mediated by stimulation of the transcription of two negative regulators of the cell cycle: the CIP/KIP family member p21, which like the related genes p27 and p57 encodes an inhibitor of cyclin-dependent kinase-2 (CDK2) and suppresses G1-to-S progression; and Ink4a, which encodes an inhibitor of CDK4 and CDK6 and promotes cell cycle exit. Nevertheless, depending on the conditions, MITF also stimulates Cdk2 and hence promotes G1-to-S progression (Du et al., 2004), and it stimulates Tbx2, leading to suppression of p21 (Vance et al., 2005). It is unknown, however, whether the MITF isoforms specifically found in the optic neuroepithelium are equally active on the promoters of the above cell cycle genes. Moreover, it is conceivable that MITF controls cell proliferation at least in part through the regulation of tyrosinase as tyrosinase catalyzes DOPA synthesis and DOPA negatively regulates cell proliferation (for a discussion of this effect, see Jeffery, 1997).

A possible connection between the in vitro findings in melanocytes and the regulation of cell proliferation in the RPE may come from studies of p27. In contrast to p21, which is not prominently expressed in RPE and is not essential for its development and maintenance (Zhang et al., 1998), p27 is expressed and is important in the RPE (Defoe and Levine, 2003; Yoshida et al., 2004; Zhang et al., 1998). This expression, however, does not start before E15 in the rat, and then only in the central parts and not peripherally where most dividing cells are found (Defoe and Levine, 2003). Only later is it found towards the marginal zone. Thus, it appears that p27 starts to be expressed just as RPE cells stop proliferating, mimicking the situation in the retina where p27 is found in most if not all cells exiting the cell cycle. Not surprisingly, lack of p27 leads to an increased proliferation of retinal progenitor cells (Dyer and Cepko, 2001; Levine et al., 2000) and to a doubling of the number of nuclei in the RPE (Yoshida et al., 2004). Moreover, p27 is found ectopically in progenitors of the hypoproliferating Chx10orJ/orJ mutant retina. Since p27 mRNA remains unchanged in the mutant retina, it is thought that ectopic p27 protein expression results from a post-transcriptional mechanism and not the potential transcriptional stimulation by Mitf which, as described above, is also upregulated in Chx10orJ/orJ mutants. Nevertheless, the retinal hypoproliferation and the morphological abnormalities are corrected in Chx10/p27 compound mutants (Green et al., 2003) as they are in Chx10/Mitf compound mutants, suggesting that Mitf and p27 might still be linked in a common pathway. Changes in Mitf expression, however, have yet to be shown in either single p27 mutants or their compound Chx10/p27 counterparts.

Cell differentiation

For many cell types in the nervous system as well as other organs, terminal differentiation is concomitant with exit from the cell cycle. In the mouse RPE, however, cell proliferation is compatible with pigmentation, itself a clear sign of differentiation. Melanin synthesis depends on a cascade of enzymatic steps of which the first, the synthesis of l-(3,4)dihydroxyphenylalanine from tyrosine, is catalyzed by the rate limiting enzyme tyrosinase. Not surprisingly, the all-purpose pigment cell transcription factor MITF is also critical in regulating pigment genes and hence pigmentation in the RPE (Nakayama et al., 1998). Other transcriptional regulators of these genes include the Otx genes (Martinez-Morales et al., 2003) and, potentially, the Mitf-relative Tfec which is expressed at relatively low levels (Rowan et al., 2004). Hence, in RPE cells, genes involved in regulating cell proliferation such as Mitf may also control cell differentiation, but the details of the mechanisms that link these two parameters still need to be analyzed.

Thoughts on the differences between melanocytes and RPE cells

  1. Top of page
  2. Summary
  3. Introduction
  4. Developmental origin of vertebrate RPE cells
  5. Signaling molecules in RPE development
  6. Integration of signaling pathways at the transcriptional level
  7. Sorting out neuroretina from RPE: the role of MITF and its cooperators and detractors
  8. The development of the iris and the ciliary body
  9. Sorting out optic stalk from RPE: the roles of Vax genes
  10. The downstream effectors of RPE specification – cell proliferation and differentiation
  11. Thoughts on the differences between melanocytes and RPE cells
  12. Conclusions
  13. Acknowledgements
  14. References

Much of our knowledge on RPE development and differentiation has been facilitated by studies of melanocytes. RPE cells share with melanocytes the transcription factor gene Mitf and at least a subset of its target genes, and yet, melanocytes and RPE cells are fundamentally different in origin, form, and function.

One of the intriguing manifestations of these differences is their differential susceptibility to perturbations in Mitf. In the mouse and quail, severe Mitf mutations lead to death of melanocytes but survival and hyperproliferation of RPE cells. One might argue, therefore, that for developing melanocytes, the pro-survival role of Mitf is more important, or manifested earlier, than its anti-proliferative role, while for developing RPE cells, its anti-proliferative role is critical and its pro-survival role negligible. Moreover, the over 30 distinct Mitf alleles in mice, some of which, when homozygous, can lead to extensive white spotting of the coat or even total whiteness, are generally less severe with respect to RPE development. Even in the human deafness/pigmentation syndromes, Waardenburg II and Tietz syndrome, where so far only heterozygosity for Mitf mutations has been found, skin and iris pigmentation are regularly affected but RPE alterations (which would lead to colobomas) are apparently not common (Read, 2000). These observations might suggest, then, that the balance between Mitf’s pro-survival and anti-proliferative roles in RPE and coat pigment cells is allele-independent. Conceptually, however, we think this to be unlikely given that the alleles differ widely in effects on protein level, subcellular distribution, splicing patterns, and coding sequence. Rather, we would predict that a careful analysis of extant and new Mitf alleles will eventually allow us to distinguish between effects on survival, cell proliferation, and differentiation, and hence genetically dissect the underlying mechanisms controlling these parameters.

One of the prominent differences between RPE cells and melanocytes is their embryological origin, but whether this is critical for the observed cell biological differences is not entirely clear. After all, the optic neuroepithelium shows a high degree of plasticity and, as discussed in this review, is capable of giving rise to a variety of different cell types. In fact, one recent study argues that neuroretinal tissue, forced to express MITF, can transdifferentiate into both RPE-like cells as well as melanocyte-like cells (Planque et al., 2004). This observation, though, was made under experimental in vitro conditions in the quail. To our knowledge, similar transitions from neuroretina to melanocytes have not been observed in vivo, neither in birds nor mice. Even the ‘natural’ neuroretinal Mitf overexpressor, the Chx10orJ/orJ mouse, seems to generate only RPE-like cells and not melanocytes from neuroretina (Horsford et al., 2005). Nevertheless, it is conceivable that both melanocytes and RPE cells, or the molecular gene expression ‘packages’ that characterize them, are evolutionarily derived from a common ancestral cell type, which was light-sensitive and pigmented (Arnheiter, 1998), and that ontogenetic differences are perhaps less decisive for the differences between RPE cells and melanocytes.

What remains, then, are differences in the tissue environment in which the two cell types are embedded. Interestingly, many of the signaling systems that are vital for melanocytes, such as KIT-ligand/KIT and EDN3/EDNRB, are of no importance for RPE cells. Tissue environment, however, not only means contacts of RPE cells with the neuroretina or the extraocular mesenchyme and choroid but also with their direct epithelial neighbors – other RPE cells – with which they maintain intimate contacts and communications. Developing melanocytes, in contrast, are individual, highly migratory cells that face constantly changing tissue environments. We suggest, therefore, that for melanocytes, staying the course of development requires a molecular program capable of firmly resisting the various tissue influences that otherwise could change their developmental fate. In contrast, RPE cells, facing a more constant environment, have hardly been selected for a similar degree of resistance to cell fate changes, and so it may become understandable why by comparison, RPE cells are apparently more ready to let down their guard when confronted with fate-changing insults (see also, Bharti and Arnheiter, 2005).

Conclusions

  1. Top of page
  2. Summary
  3. Introduction
  4. Developmental origin of vertebrate RPE cells
  5. Signaling molecules in RPE development
  6. Integration of signaling pathways at the transcriptional level
  7. Sorting out neuroretina from RPE: the role of MITF and its cooperators and detractors
  8. The development of the iris and the ciliary body
  9. Sorting out optic stalk from RPE: the roles of Vax genes
  10. The downstream effectors of RPE specification – cell proliferation and differentiation
  11. Thoughts on the differences between melanocytes and RPE cells
  12. Conclusions
  13. Acknowledgements
  14. References

It is clear that pigment cell biology, our love and fascination, touches on far more disciplines than just the regulation of coloration and malignant transformation. Genetic models show that without pigment cells, sensory organ physiology is disturbed. In mammals, for instance, hearing is seriously compromised in the absence of neural crest-derived melanocytes (for review, Tachibana, 1999), and vision is impaired without normal pigmentation in the eye (for review, Jeffery, 1997). RPE development presents an excellent model system for molecularly dissecting binary cell fate choices, the formation of a polarized epithelium with its intricate rules of intracellular trafficking (for review, Futter, 2006), and eventually for studying, and finding cures for, the retinal diseases that can result from alterations of the RPE in humans (see http://www.sph.uth.tmc.edu/retnet/). Understanding RPE development will not only help us treat diseases, however, but also gain insights into the principles that have shaped eye evolution over time and led to the tight association between photoreceptor and pigment cells. As speculated previously (Arnheiter, 1998), such ‘evo-devo’ studies, extended from the vertebrate subphylum to other phyla whose members benefit from vision, may eventually answer the intriguing question of what came first, pigmentation of the body, or the eyes to see it. RPE research may thus add to both the useful and the museful.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Developmental origin of vertebrate RPE cells
  5. Signaling molecules in RPE development
  6. Integration of signaling pathways at the transcriptional level
  7. Sorting out neuroretina from RPE: the role of MITF and its cooperators and detractors
  8. The development of the iris and the ciliary body
  9. Sorting out optic stalk from RPE: the roles of Vax genes
  10. The downstream effectors of RPE specification – cell proliferation and differentiation
  11. Thoughts on the differences between melanocytes and RPE cells
  12. Conclusions
  13. Acknowledgements
  14. References
  • Ahmad, I., Tang, L., and Pham, H. (2000). Identification of neural progenitors in the adult mammalian eye. Biochem. Biophys. Res. Commun. 270, 517521.
  • Amato, M.A., Boy, S., and Perron, M. (2004). Hedgehog signaling in vertebrate eye development: a growing puzzle. Cell Mol. Life Sci. 61, 899910.
  • Arnheiter, H. (1998). Evolutionary biology. Eyes viewed from the skin. Nature 391, 632633.
  • Arnheiter, H., Hou, L., Nguyen, M.T., Nakayama, A., Champagne, B., Hallsson, J.H., and Bismuth, K. (2002). The role of microphthalmia in pigment cell development. In Mechanisms of Suntanning, J.p.Ortonne, and R., Ballotti, eds (London: Martin Dunitz Publication), pp. 4963.
  • Arnheiter, H., Hou, L., Nguyen, M.T.T., Bismuth, K., Csermely, T., Murakami, H., Skuntz, S., Liu, W., and Bharti, K. (2006). Mitf—A matter of life and death for the developing melanocyte. In: From Melanocytes to Malignant Melanoma, V., Hearing, and S.P.L., Leong, eds (Totowa, NJ: Humana Press), pp. 2749.
  • Bäumer, N., Marquardt, T., Stoykova, A., Spieler, D., Treichel, D., Ashery-Padan, R., and Gruss, P. (2003). Retinal pigmented epithelium determination requires the redundant activities of Pax2 and Pax6. Development 130, 29032915.
  • Bharti, K., and Arnheiter, H. (2005). When pigment cells turn into neurons. J. Invest. Dermatol. 125, xxi.
  • Bismuth, K., Maric, D., and Arnheiter, H. (2005). MITF and cell proliferation: the role of alternative splice forms. Pigment Cell Res. 18, 349359.
  • Bodenstein, L., and Sidman, R.L. (1987a). Growth and development of the mouse retinal pigment epithelium. I. Cell and tissue morphometrics and topography of mitotic activity. Dev. Biol. 121, 192204.
  • Bodenstein, L., and Sidman, R.L. (1987b). Growth and development of the mouse retinal pigment epithelium. II. Cell patterning in experimental chimaeras and mosaics. Dev. Biol. 121, 205219.
  • Bora, N., Conway, S.J., Liang, H., and Smith, S.B. (1998). Transient overexpression of the Microphthalmia gene in the eyes of Microphthalmia vitiligo mutant mice. Dev. Dyn. 213, 283292.
  • Bumsted, K.M., and Barnstable, C.J. (2000). Dorsal retinal pigment epithelium differentiates as neural retina in the microphthalmia (mi/mi) mouse. Invest. Ophthalmol. Vis. Sci. 41, 903908.
  • Carreira, S., Goodall, J., Aksan, I., La Rocca, S.A., Galibert, M.D., Denat, L., Larue, L., and Goding, C.R. (2005). Mitf cooperates with Rb1 and activates p21Cip1 expression to regulate cell cycle progression. Nature 433, 764769.
  • Chiang, C., Litingtung, Y., Lee, E., Young, K.E., Corden, J.L., Westphal, H., and Beachy, P.A. (1996). Cyclopia and defective axial patterning in mice lacking Sonic hedgehog gene function. Nature 383, 407413.
  • Chow, R.L., and Lang, R.A. (2001). Early eye development in vertebrates. Annu. Rev. Cell Dev. Biol. 17, 255296.
  • Clark, V.M. (1986). The cell biology of the retinal pigment epithelium. In: The Retina: A Model for Cell Biology Studies. Part II, A., Ruben, and D., Farber, eds (Orlando, FL: Academic Press), pp. 129168.
  • Coles, B.L., Horsford, D.J., Mcinnes, R.R., and Van Der Kooy, D. (2006). Loss of retinal progenitor cells leads to an increase in the retinal stem cell population in vivo. Eur. J. Neurosci. 23, 7582.
  • Dakubo, G.D., Wang, Y.P., Mazerolle, C., Campsall, K., Mcmahon, A.P., and Wallace, V.A. (2003). Retinal ganglion cell-derived sonic hedgehog signaling is required for optic disc and stalk neuroepithelial cell development. Development 130, 29672980.
  • Darwin, C. (1886). The Origin of Species By Means of Natural Selection, or the Preservation of Favored Races in the Struggle for Life (New York: D. Appleton & Co).
  • De Iongh, R., and McAvoy, J.W. (1993). Spatio-temporal distribution of acidic and basic FGF indicates a role for FGF in rat lens morphogenesis. Dev Dyn. 198, 190202.
  • Defoe, D.M., and Levine, E.M. (2003). Expression of the cyclin-dependent kinase inhibitor p27Kip1 by developing retinal pigment epithelium. Gene Expr. Patterns 3, 615619.
  • Del Rio-Tsonis, K., and Tsonis, P.A. (2003). Eye regeneration at the molecular age. Dev. Dyn. 226, 211224.
  • Dhomen, N.S., Balaggan, K.S., Pearson, R.A., Bainbridge, J.W., Levine, E.M., Ali, R.R., and Sowden, J.C. (2006). Absence of chx10 causes neural progenitors to persist in the adult retina. Invest. Ophthalmol. Vis. Sci. 47, 386396.
  • Dong, L.J., and Chung, A.E. (1991). The expression of the genes for entactin, laminin A, laminin B1 and laminin B2 in murine lens morphogenesis and eye development. Differentiation 48, 157172.
  • Dong, S., Landfair, J., Balasubramani, M., Bier, M.E., Cole, G., and Halfter, W. (2002). Expression of basal lamina protein mRNAs in the early embryonic chick eye. J. Comp. Neurol. 447, 261273.
  • Doughtie, D., and Rao, K. (1984). Ultrastructure of the eyes of the grass shrimp, Palaemonetes pugio. General morphology, and light and dark adaption at noon. Cell Tissue Res. 238, 271288.
  • Du, J., Widlund, H.R., Horstmann, M.A., Ramaswamy, S., Ross, K., Huber, W.E., Nishimura, E.K., Golub, T.R., and Fisher, D.E. (2004). Critical role of CDK2 for melanoma growth linked to its melanocyte-specific transcriptional regulation by MITF. Cancer Cell 6, 565576.
  • Dyer, M.A., and Cepko, C.L. (2001). p27Kip1 and p57Kip2 regulate proliferation in distinct retinal progenitor cell populations. J. Neurosci. 21, 42594271.
  • Fuhrmann, S., Levine, E.M., and Reh, T.A. (2000). Extraocular mesenchyme patterns the optic vesicle during early eye development in the embryonic chick. Development 127, 45994609.
  • Futter, C.E. (2006). The molecular regulation of organelle transport in mammalian retinal pigment epithelial cells. Pigment Cell Res. 19, 104111.
  • Galy, A., Neron, B., Planque, N., Saule, S., and Eychene, A. (2002). Activated MAPK/ERK kinase (MEK-1) induces transdifferentiation of pigmented epithelium into neural retina. Dev. Biol. 248, 251264.
  • Gehring, W.J. (2005). New perspectives on eye development and the evolution of eyes and photoreceptors. J. Hered. 96, 171184.
  • Gondo, Y., Gardner, J.M., Nakatsu, Y., Durham-Pierre, D., Deveau, S.A., Kuper, C., and Brilliant, M.H. (1993). High-frequency genetic reversion mediated by a DNA duplication: the mouse pink-eyed unstable mutation. Proc. Natl. Acad. Sci. U. S. A. 90, 297301.
  • Green, E.S., Stubbs, J.L., and Levine, E.M. (2003). Genetic rescue of cell number in a mouse model of microphthalmia: interactions between Chx10 and G1-phase cell cycle regulators. Development 130, 539552.
  • Halfter, W., Dong, S., Yip, Y.P., Willem, M., and Mayer, U. (2002). A critical function of the pial basement membrane in cortical histogenesis. J. Neurosci. 22, 60296040.
  • Hertwig, P. (1942). Neue Mutationen und Kopplungsgruppen bei der Hausmaus. Z. Indukt. Abstammungs- u. Vererbungsl. 80, 220246.
  • Hodgkinson, C.A., Moore, K.J., Nakayama, A., Steingrimsson, E., Copeland, N.G., Jenkins, N.A., and Arnheiter, H. (1993). Mutations at the mouse microphthalmia locus are associated with defects in a gene encoding a novel basic–helix–loop–helix–zipper protein. Cell 74, 395404.
  • Holtfreter, J. (1939). Gewebeaffinität, ein Mittel der embryonalen Formbildung. Arch. Exp. Zellforsch. 23, 169209.
  • Horsford, D.J., Nguyen, M.T., Sellar, G.C., Kothary, R., Arnheiter, H., and Mcinnes, R.R. (2005). Chx10 repression of Mitf is required for the maintenance of mammalian neuroretinal identity. Development 132, 177187.
  • Hughes, M.J., Lingrel, J.B., Krakowsky, J.M., and Anderson, K.P. (1993). A helix–loop–helix transcription factor-like gene is located at the mi locus. J. Biol. Chem. 268, 2068720690.
  • Hyer, J. (2004). Looking at an oft-overlooked part of the eye: a new perspective on ciliary body development in chick. Dev. Neurosci. 26, 456465.
  • Hyer, J., Mima, T., and Mikawa, T. (1998). FGF1 patterns the optic vesicle by directing the placement of the neural retina domain. Development 125, 869877.
  • Jeffery, G. (1997). The albino retina: an abnormality that provides insight into normal retinal development. Trends Neurosci. 20, 165169.
  • Kitano, H. (2004). Biological robustness. Nat. Rev. Genet. 5, 826837.
  • Kong, Y., Usuda, N., and Nagata, T. (1992). Radioautographic study on DNA synthesis of the retina and retinal pigment epithelium of developing mouse embryos. Cell Mol. Biol. 38, 263272.
  • Konyukhov, B.V., and Sazhina, M.V. (1966). Interaction of the genes of ocular retardation and microphthalmia in mice. Folia Biol. (Praha) 12, 116123.
  • Koshiba-Takeuchi, K., Takeuchi, J.K., Matsumoto, K. et al. (2000). Tbx5 and the retinotectum projection. Science 287, 134137.
  • Kozmik, Z. (2005). Pax genes in eye development and evolution. Curr. Opin. Genet. Dev. 15, 430438.
  • Kubota, R., Mcguire, C., Dierks, B., and Reh, T.A. (2004). Identification of ciliary epithelial-specific genes using subtractive libraries and cDNA arrays in the avian eye. Dev. Dyn. 229, 529540.
  • Land, M., and Nilsson, D.-E. (2002). Animal Eyes (Oxford, NY: Oxford University Press).
  • Lee, C.S., May, N.R., and Fan, C.M. (2001). Transdifferentiation of the ventral retinal pigmented epithelium to neural retina in the growth arrest specific gene 1 mutant. Dev. Biol. 236, 1729.
  • Levine, E.M., Close, J., Fero, M., Ostrovsky, A., and Reh, T.A. (2000). p27(Kip1) regulates cell cycle withdrawal of late multipotent progenitor cells in the mammalian retina. Dev. Biol. 219, 299314.
  • Li, Z., Hu, M., Ochocinska, M.J., Joseph, N.M., and Easter, S.S., Jr (2000a). Modulation of cell proliferation in the embryonic retina of zebrafish (Danio rerio). Dev. Dyn. 219, 391401.
  • Li, Z., Joseph, N.M., and Easter, S.S., Jr (2000b). The morphogenesis of the zebrafish eye, including a fate map of the optic vesicle. Dev. Dyn. 218, 175188.
  • Loercher, A.E., Tank, E.M., Delston, R.B., and Harbour, J.W. (2005). MITF links differentiation with cell cycle arrest in melanocytes by transcriptional activation of INK4A. J. Cell Biol. 168, 3540.
  • Martinez-Morales, J.R., Signore, M., Acampora, D., Simeone, A., and Bovolenta, P. (2001). Otx genes are required for tissue specification in the developing eye. Development 128, 20192030.
  • Martinez-Morales, J.R., Dolez, V., Rodrigo, I., Zaccarini, R., Leconte, L., Bovolenta, P., and Saule, S. (2003). OTX2 activates the molecular network underlying retina pigment epithelium differentiation. J. Biol. Chem. 278, 2172121731.
  • Martinez-Morales, J.R., Rodrigo, I., and Bovolenta, P. (2004). Eye development: a view from the retina pigmented epithelium. Bioessays 26, 766777.
  • Martinez-Morales, J.R., Del Bene, F., Nica, G., Hammerschmidt, M., Bovolenta, P., and Wittbrodt, J. (2005). Differentiation of the vertebrate retina is coordinated by an FGF signaling center. Dev. Cell 8, 565574.
  • Massague, J., Seoane, J., and Wotton, D. (2005). Smad transcription factors. Genes Dev. 19, 27832810.
  • McWhirter, J.R., Goulding, M., Weiner, J.A., Chun, J., and Murre, C. (1997). A novel fibroblast growth factor gene expressed in the developing nervous system is a downstream target of the chimeric homeodomain oncoprotein E2A-Pbx1. Development 124, 32213232.
  • Miller, D.L., Ortega, S., Bashayan, O., Basch, R., and Basilico, C. (2000). Compensation by fibroblast growth factor 1 (FGF1) does not account for the mild phenotypic defects observed in FGF2 null mice. Mol. Cell Biol. 20, 22602268.
  • Mochii, M., Ono, T., Matsubara, Y., and Eguchi, G. (1998). Spontaneous transdifferentiation of quail pigmented epithelial cell is accompanied by a mutation in the Mitf gene. Dev. Biol. 196, 145159.
  • Moshiri, A., Close, J., and Reh, T.A. (2004). Retinal stem cells and regeneration. Int. J. Dev. Biol. 48, 10031014.
  • Mui, S.H., Hindges, R., O'leary, D.D., Lemke, G., and Bertuzzi, S. (2002). The homeodomain protein Vax2 patterns the dorsoventral and nasotemporal axes of the eye. Development 129, 797804.
  • Mui, S.H., Kim, J.W., Lemke, G., and Bertuzzi, S. (2005). Vax genes ventralize the embryonic eye. Genes Dev. 19, 12491259.
  • Müller, G. (1950). Eine entwicklungsgeschichtliche Untersuchung über das erbliche Kolobom mit Mikrophthalmus bei der Hausmaus. Z Mikrosk Anat Forsch 56, 520558.
  • Nakayama, A., Nguyen, M.T., Chen, C.C., Opdecamp, K., Hodgkinson, C.A., and Arnheiter, H. (1998). Mutations in microphthalmia, the mouse homolog of the human deafness gene MITF, affect neuroepithelial and neural crest-derived melanocytes differently. Mech. Dev. 70, 155166.
  • Neumann, C.J., and Nüsslein-Volhard, C. (2000). Patterning of the zebrafish retina by a wave of sonic hedgehog activity. Science 289, 21372139.
  • Nguyen, M., and Arnheiter, H. (2000). Signaling and transcriptional regulation in early mammalian eye development: a link between FGF and MITF. Development 127, 35813591.
  • Perron, M., Boy, S., Amato, M.A., Viczian, A., Koebernick, K., Pieler, T., and Harris, W.A. (2003). A novel function for Hedgehog signalling in retinal pigment epithelium differentiation. Development 130, 15651577.
  • Pittack, C., Grunwald, G.B., and Reh, T.A. (1997). Fibroblast growth factors are necessary for neural retina but not pigmented epithelium differentiation in chick embryos. Development 124, 805816.
  • Planque, N., Raposo, G., Leconte, L., Anezo, O., Martin, P., and Saule, S. (2004). Microphthalmia transcription factor induces both retinal pigmented epithelium and neural crest melanocytes from neuroretina cells. J. Biol. Chem. 279, 4191141917.
  • Read, A. (2000). Waardenburg syndrome. In: The Metabolic and Molecular Basis of Inherited Disease, 8th edn, C., Scriver, A., Beaudet, D., Valle, and W., Sly, eds (Toronto: McGraw-Hill), pp. 60976116.
  • Rowan, S., Chen, C.-M.A., Young, T.L., Fisher, D.E., and Cepko, C.L. (2004). Transdifferentiation of the retina into pigmented cells in ocular retardation mice defines a new function of the homeodomain gene Chx10. Development 131, 51395152.
  • Saika, S., Kono-Saika, S., Tanaka, T., et al. (2004). Smad3 is required for dedifferentiation of retinal pigment epithelium following retinal detachment in mice. Lab. Invest. 84, 12451258.
  • Sasagawa, S., Takabatake, T., Takabatake, Y., Muramatsu, T., and Takeshima, K. (2002). Axes establishment during eye morphogenesis in Xenopus by coordinate and antagonistic actions of BMP4, Shh, and RA. Genesis 33, 8696.
  • Scholtz, C.L., and Chan, K.K. (1987). Complicated colobomatous microphthalmia in the microphthalmic (mi/mi) mouse. Development 99, 501508.
  • Schwarz, M., Cecconi, F., Bernier, G., Andrejewski, N., Kammandel, B., Wagner, M., and Gruss, P. (2000). Spatial specification of mammalian eye territories by reciprocal transcriptional repression of Pax2 and Pax6. Development 127, 43254334.
  • Spence, J.R., Madhavan, M., Ewing, J.D., Jones, D.K., Lehman, B.M., and Del Rio-Tsonis, K. (2004). The hedgehog pathway is a modulator of retina regeneration. Development 131, 46074621.
  • Steingrimsson, E., Copeland, N.G., and Jenkins, N.A. (2004). Melanocytes and the microphthalmia transcription factor network. Annu. Rev. Genet. 38, 365411.
  • Stenkamp, D.L., Frey, R.A., Prabhudesai, S.N., and Raymond, P.A. (2000). Function for Hedgehog genes in zebrafish retinal development. Dev. Biol. 220, 238252.
  • Strauss, O. (2005). The retinal pigment epithelium in visual function. Physiol. Rev. 85, 845881.
  • Tachibana, M. (1999). Sound needs sound melanocytes to be heard. Pigment Cell Res. 12, 344354.
  • Tachibana, M., Perez-Jurado, L.A., Nakayama, A., Hodgkinson, C.A., Li, X., Schneider, M., Miki, T., Fex, J., Francke, U., and Arnheiter, H. (1994). Cloning of MITF, the human homolog of the mouse microphthalmia gene and assignment to chromosome 3p14.1-p12.3. Hum. Mol. Genet. 3, 553557.
  • Takabatake, T., Ogawa, M., Takahashi, T.C., Mizuno, M., Okamoto, M., and Takeshima, K. (1997). Hedgehog and patched gene expression in adult ocular tissues. FEBS Lett. 410, 485489.
  • Take-Uchi, M., Clarke, J.D., and Wilson, S.W. (2003). Hedgehog signalling maintains the optic stalk-retinal interface through the regulation of Vax gene activity. Development 130, 955968.
  • Tang, M., Ruiz, M., Kosaras, B., and Sidman, R.L. (1996). Increased cell genesis in retinal pigment epithelium of perinatal vitiligo mutant mice. Invest. Ophthalmol. Vis. Sci. 37, 11161124.
  • Tcheng, M., Fuhrmann, G., Hartmann, M.P., Courtois, Y., and Jeanny, J.C. (1994). Spatial and temporal expression patterns of FGF receptor genes type 1 and type 2 in the developing chick retina. Exp. Eye Res. 58, 351358.
  • Tropepe, V., Coles, B.L., Chiasson, B.J., Horsford, D.J., Elia, A.J., McInnes, R.R., and Van Der Kooy, D. (2000). Retinal stem cells in the adult mammalian eye. Science 287, 20322036.
  • Vance, K.W., Carreira, S., Brosch, G., and Goding, C.R. (2005). Tbx2 is overexpressed and plays an important role in maintaining proliferation and suppression of senescence in melanomas. Cancer Res. 65, 22602268.
  • Vogel-Höpker, A., Momose, T., Rohrer, H., Yasuda, K., Ishihara, L., and Rapaport, D.H. (2000). Multiple functions of fibroblast growth factor-8 (FGF-8) in chick eye development. Mech. Dev. 94, 2536.
  • Wallace, V.A., and Raff, M.C. (1999). A role for Sonic hedgehog in axon-to-astrocyte signalling in the rodent optic nerve. Development 126, 29012909.
  • Wanaka, A., Milbrandt, J., and Johnson, E.M. (1991). Expression of FGF receptor gene in rat development. Development 111, 455468.
  • West-Mays, J.A., Zhang, J., Nottoli, T., Hagopian-Donaldson, S., Libby, D., Strissel, K.J., and Williams, T. (1999). AP-2a transcription factor is required for early morphogenesis of the lens vesicle. Dev. Biol. 206, 4662.
  • Wu, M., Hemesath, T.J., Takemoto, C.M., Horstmann, M.A., Wells, A.G., Price, E.R., Fisher, D.Z., and Fisher, D.E. (2000). c-Kit triggers dual phosphorylations, which couple activation and degradation of the essential melanocyte factor Mi. Genes Dev. 14, 301312.
  • Yang, X.J. (2004). Roles of cell-extrinsic growth factors in vertebrate eye pattern formation and retinogenesis. Semin. Cell Dev. Biol. 15, 91103.
  • Yoshida, K., Kase, S., Nakayama, K. et al. (2004). Involvement of p27KIP1 in the proliferation of the developing corneal endothelium. Invest. Ophthalmol. Vis. Sci. 45, 21632167.
  • Zhang, X.M., and Yang, X.J. (2001). Temporal and spatial effects of Sonic hedgehog signaling in chick eye morphogenesis. Dev. Biol. 233, 271290.
  • Zhang, P., Wong, C., Depinho, R.A., Harper, J.W., and Elledge, S.J. (1998). Cooperation between the Cdk inhibitors p27(KIP1) and p57(KIP2) in the control of tissue growth and development. Genes Dev. 12, 31623167.
  • Zhao, S., Hung, F.C., Colvin, J.S., White, A., Dai, W., Lovicu, F.J., Ornitz, D.M., and Overbeek, P.A. (2001). Patterning the optic neuroepithelium by FGF signaling and Ras activation. Development 128, 50515060.