DEVELO

PMENT

3483RESEARCH ARTICLE

INTRODUCTIONThe vertebrate eye primordia are first visible as an outgrowth of theprosencephalic neuroepithelium (reviewed by Chow and Lang,2001; Martinez-Morales et al., 2004). Enlargement of the distalportion of the optic vesicle and dorsal expansion divides the opticvesicle into three territories (Hilfer, 1983): the narrow optic stalk(proximal), the neural retina (NR) and the retinal pigment epithelium(RPE). Formation of the lens vesicle from the surface ectoderminduces the distal region of the optic vesicle to invaginate and thisprocess results in the development of the bilayered optic cup. Theinner layer develops into the multilayered NR, whereas the outerlayer develops into the single-layered, pigmented RPE (reviewed byChow and Lang, 2001).

In vertebrates, optic vesicle cells initially co-express a number oftranscription factors (TFs) that become restricted to NR, RPE andoptic nerve later on, implicating that these cells are competent todevelop into these tissues (reviewed by Martinez-Morales et al.,2004). Extrinsic signals emanating from the surface ectoderm andocular mesenchyme appear to induce and repress specific TFs,which subsequently pattern the optic vesicle into NR and RPEdomains (for reviews, see Chow and Lang, 2001; Martinez-Moraleset al., 2004). For example, fibroblast growth factors (FGFs)expressed in the surface ectoderm and/or distal optic vesicle appearto be involved in NR induction and differentiation (Pittack et al.,1997; Hyer et al., 1998; Nguyen and Arnheiter, 2000; Vogel-Höpkeret al., 2000; Martinez-Morales et al., 2005). Embryonictransplantations and in ovo explant cultures of the chick optic vesicle

have shown that the dorsoventral polarity of the eye is alreadyspecified by stage 10 (Uemonsa et al., 2002; Kagiyama et al., 2005).At this time point, the dorsal half of the optic vesicle is fated todevelop mainly into RPE, whereas the ventral portion developsmainly into NR (Kagiyama et al., 2005).

Little is known about the molecular mechanisms that specify theRPE (reviewed by Martinez-Morales et al., 2004). The mesenchymeadjacent to the optic vesicle appears to be crucial for RPEdevelopment, but the molecular nature of the signal(s) is still unclear(reviewed by Chow and Lang, 2001; Martinez-Morales et al., 2004).Activin, a member of the transforming growth factor-� (TGF-�)superfamily, or a related growth factor appears to be released fromthe mesenchyme to induce RPE development (Fuhrmann et al.,2000). Cell-intrinsic TFs mediate the effect of mesenchymalsignalling molecules on RPE development (reviewed by Chow andLang, 2001). The best-studied example is the microphthalmia-associated transcription factor (MITF), a basic helix-loop-helixleucine zipper TF that is crucial for the acquisition and maintenanceof RPE cell identity (reviewed by Martinez-Morales et al., 2004).Ectopic Mitf expression in cultured avian neural retina cells resultsin the induction of pigmentation by initiating the expression of twomarkers of differentiated pigment cells: melanosomal matrix protein115 (MMP115) and tyrosinase (Mochii et al., 1998; Planque et al.,1999). By contrast, inhibition of Mitf by small interfering RNAs(siRNA) decreases MMP115 expression and promotes de-differentiation of the RPE (Iwakiri et al., 2005). In Mitf mutants, theRPE remains unpigmented and displays areas developing into asecond NR (Bumsted and Barnstable, 2000; Nguyen and Arnheiter,2000). Members of the orthodenticle-related family of TFs, Otx1and Otx2, are also required for RPE specification during vertebrateeye development (Martinez-Morales et al., 2001; Martinez-Moraleset al., 2003). In Otx1/Otx2 mutants, RPE development is disturbedand instead the outer layer of the optic cup develops NR-likefeatures. Similar to Mitf, Otx2 overexpression induces a pigmentedphenotype in cultured NR cells. Otx1 and Otx2 are initially

Bone morphogenetic proteins specify the retinal pigmentepithelium in the chick embryoFrank Müller*, Hermann Rohrer and Astrid Vogel-Höpker†

In vertebrates, the neuroepithelium of the optic vesicle is initially multipotential, co-expressing a number of transcription factorsthat are involved in retinal pigment epithelium (RPE) and neural retina (NR) development. Subsequently, extrinsic signalsemanating from the surrounding tissues induce the separation of the optic vesicle into three domains: the optic stalk/nerve, the NRand the RPE. Here, we show that bone morphogenetic proteins (BMPs) are sufficient and essential for RPE development in vivo.Bmp4 and Bmp7 are expressed in the surface ectoderm overlying the optic vesicle, the surrounding mesenchyme and/orpresumptive RPE during the initial stages of eye development. During the initial stages of chick eye development themicrophthalmia-associated transcription factor (Mitf), important for RPE development, is expressed in the optic primordium that iscovered by the BMP-expressing surface ectoderm. Following BMP application, the optic neuroepithelium, including thepresumptive optic stalk/nerve and NR domain, develop into RPE as assessed by the expression of Otx2, Mitf, Wnt2b and thepigmented cell marker MMP115. By contrast, interfering with BMP signalling prevents RPE development in the outer layer of theoptic cup and induces NR-specific gene expression (e.g. Chx10). Our results show that BMPs are sufficient and essential for RPEdevelopment during optic vesicle stages. We propose a model in which the BMP-expressing surface ectoderm initiates RPEspecification by inducing Mitf expression in the underlying neuroepithelium of the optic vesicle.

KEY WORDS: BMP, Eye development, Retinal pigment epithelium, RPE specification

Development 134, 3483-3493 (2007) doi:10.1242/dev.02884

Max-Planck-Institute for Brain Research, Department of Neurochemistry,Deutschordenstr. 46, 60528 Frankfurt/M., Germany.

*Present address: Institute for Physiological Chemistry, Martin-Luther-University,Hollystr. 1, D-06097 Halle, Germany†Author for correspondence (e-mail: veithopker@aol.com)

Accepted 9 July 2007I

DEVELO

PMENT

3484

expressed in the entire optic vesicle. Subsequently, Otx2 expressionis maintained in the presumptive RPE and expression persists in theadult RPE (reviewed by Martinez-Morales et al., 2004).

There appear to be differences in the Mitf expression patternbetween chick and mouse (Mochii et al., 1998; Fuhrmann et al.,2000; Nguyen and Arnheiter, 2000). In chick, Mitf expression seemsto be restricted to the dorsal region of the optic vesicle, thepresumptive RPE, and this region is covered by the surroundingmesenchyme. By contrast, the entire mouse optic vesicle is initiallycovered by a small amount of mesenchyme and here Mitf expressionis observed throughout the optic vesicle. Once the mesenchyme isdisplaced at the distal part of the optic vesicle at the time this regioncontacts the FGF-expressing surface ectoderm, Mitf expression isinhibited and instead NR induction occurs in the mouse (Bora et al.,1998; Nakayama et al., 1998; Nguyen and Arnheiter, 2000). Thepaired-like homeobox gene Chx10 is a specific marker of retinalprogenitor cells and functions to repress Mitf expression in the distaloptic vesicle (Rowan et al., 2004; Horsford et al., 2005). Moreover,overexpression of Chx10 in the chick RPE causes downregulationof Mitf expression and other pigment markers, leading to anonpigmented RPE (Rowan et al., 2004). Thus, the current model isthat the ocular mesenchyme is necessary to induce the RPE domainduring vertebrate eye development, whereas FGFs released from thesurface ectoderm ensure that the NR develops at the distal part of theoptic vesicle (reviewed by Chow and Lang, 2001).

Like activin, BMPs belong to the TGF-� superfamily and severalBMP ligands and their receptors are expressed in the developingchick and mouse eye and surrounding tissues (reviewed by Chowand Lang, 2001; Martinez-Morales et al., 2004). BMPs are involvedin several aspects of vertebrate eye development. For example, BMPsignalling is required for patterning the eye primordia duringblastula and gastrula stages in zebrafish (Hammerschmidt et al.,2003), whereas later on BMPs function in both dorsal and ventralpatterning of the vertebrate eye (Koshiba-Takeuchi et al., 2000;Sakuta et al., 2001; Adler and Belecky-Adams, 2002; Sasagawa etal., 2002; Murali et al., 2005). In addition, the generation of retina-specific BMP type 1 receptor mutant mice has shown that differentthreshold levels of BMP signalling regulate distinct developmentalprocesses such as dorsoventral patterning of the NR, as well as NRgrowth and differentiation (Murali et al., 2005). At present, however,the possible involvement of BMP signalling in RPE developmentduring optic vesicle stages has not been established (for a review, seeMartinez-Morales et al., 2004).

In this study, we show that BMP family members are expressed atthe right time and place to be involved in inducing Mitf expression inthe chick optic vesicle. Mitf expression is first observed at optic vesiclestages, being strongest in the distal optic vesicle that is covered by theBMP-expressing surface ectoderm. Gain-of-function experimentsshow that BMPs are sufficient to elicit RPE development in vivo.BMP treatment converts cells of the presumptive optic stalk and NRregion into RPE. By contrast, interfering with BMP signalling at opticvesicle stages inhibits RPE formation and induces NR-specific geneexpression in the outer optic cup. Thus, we provide evidence thatduring optic vesicle stages, BMPs are necessary and sufficient for RPEdevelopment in vivo.

MATERIALS AND METHODSAssaying gene expression in chick embryos by in situhybridisation (ISH)ISH was performed on whole embryos according to Wilkinson (Wilkinson,1993) and Henrique et al. (Henrique et al., 1995) and on cryostat sectionsusing the technique described by Reissmann et al. (Reissmann et al., 1996).

In some cases, we enhanced the signal by staining whole-mount embryostwo to three times or by leaving the colour reaction on sections overnight.Antisense RNA probes specific for chicken Bmp2, Bmp4 and Bmp7(Reissmann et al., 1996; Vogel-Höpker and Rohrer, 2002), Bmp5 (Oh et al.,1996), Bmpr1b (L. Niswander, Sloan-Kettering Institute, NY), Rx and Fgf8(T. Ogura, Tohoku University, Aoba, Japan), Mitf (Mochii et al., 1998),MMP115 (Rowan et al., 2004), Chx10 (D. Schulte, MPI Brain Research,Frankfurt/M, Germany), Wnt2b (H. Roelink, University of Washington,Seattle, WA) and Sox10 (M. Wegner, University of Erlangen, Germany)were used.

In vivo manipulations of the developing chick embryoGain-of-function experimentsA 2 �l drop of recombinant mouse BMP5 or BMP4 (0.7 mg/ml or 1 mg/ml;R&D Systems) was placed in a Petri dish and about eight drops (10 �l each)of distilled water were placed around it to keep it from evaporating. Ten tofifteen agarose beads (Affi-Gel blue beads, Biorad) were added to the BMPsolution, taking care to avoid transferring any fluid with the beads. Thesebeads were incubated in the BMP4 or BMP5 solution for a minimum of 1hour at room temperature.

Fertile white leghorn chicken eggs were incubated at 37.8°C until theyreached the desired stages (stages 8-12) according to Hamburger andHamilton (Hamburger and Hamilton, 1951). The embryonic membraneswere removed and a small incision was made either temporal (posterior) tothe optic vesicle/cup or into the midline of the forebrain. One BMP-soakedbead was transferred to the egg, inserted through the slit in the membranesand placed either temporal to the optic vesicle/cup into the mesenchyme orplaced into the forebrain/optic vesicle region. The embryos were left todevelop at 37.8°C until they reached the desired stages (stages 13-26). At thispoint, the embryos were fixed in 4% paraformaldehyde in PBS (PFA) at 4°Cfor 24-48 hours. Embryos to be used for whole-mount ISH were dehydratedand stored in 100% methanol. Those intended for ISH on sections werecryoprotected overnight in 15% sucrose in PBS at 4°C; consecutive 12-16�m sections were then cut and analysed by ISH. For control experiments,beads were soaked in PBS and implanted according to the same protocol.

Loss-of-function experimentsNoggin-expressing Chinese hamster ovary (CHO B3A4) cells were culturedand implanted as described (Vogel-Höpker and Rohrer, 2002). Briefly, forimplantation, a 90% confluent culture was harvested and centrifuged to forma pellet for implantation. The embryonic membranes of stage 8-12 chickembryos were removed and noggin-expressing CHO cells implanted/injected into the mesenchyme temporal to the optic vesicle or into the opticvesicle using fine glass micropipettes. For control experiments, CHO cellswere cultured, harvested and implanted according to the same protocol.After incubation for a further 1-6 days, the embryos were fixed and sectionedas described above.

Replication-competent RCAS (B) retroviruses engineered to express thedominant-negative BMPR1B (referred to here as dnBmpR1b) were kindlyprovided by L. Niswander. Retroviral stocks were prepared as describedpreviously (Vogel et al., 1995; Vogel et al., 1996). For the infection ofembryos with dnBmpr1b-RCAS (B), or with RCAS (B) as control, retroviralstock was injected either into the optic vesicle or into the mesenchymetemporal to the optic vesicle at stages 6-11, using fine glass micropipettes.The embryos were incubated for a further 3-8 days and analysed as describedabove.

RESULTSGene expression in the neuroepithelium of theoptic vesicle during the initial stages of chick eyedevelopmentAt the beginning of eye development, the entire optic vesicle co-expresses several genes known to be involved in RPE and NRdevelopment. To determine the time point when RPE and NRdevelopment is initiated in the chick, we analysed and compared thedistribution of transcripts of genes known to be involved in RPE andNR development in vertebrates.

RESEARCH ARTICLE Development 134 (19)

DEVELO

PMENT

In the chick, the separation of the optic vesicle into NR and RPEdomains is initiated in the distal region of the optic vesicle at stage10 (see below). Initially, Otx2 transcripts are detected throughout theoptic vesicle (data not shown) (Bovolenta et al., 1997). At stage 10,Otx2 expression weakens in the distal portion of the optic vesicle(Fig. 1A) and, by stage 13, Otx2 transcripts are abundant in thedorsal part of the optic vesicle, the cells that will give rise to the RPE(Fig. 1B). Otx2 expression is maintained in the RPE thereafter and,from about stage 23 onwards, Otx2 expression is also detected in NRcells (unoperated eye in Fig. 5B) (Bovolenta et al., 1997).

No Mitf transcripts were observed in the eye primordia of thechick at stage 8 (Fig. 2F). However, Mitf expression was observedin the optic vesicle at stage 9, where expression is strongest in thedistal region (Fig. 2G) that is covered by the overlying surfaceectoderm (Fig. 2, compare G with J). In the temporal part of the opticvesicle, downregulation of Mitf expression was observed in the distalportion at stage 10 (Fig. 1C; Fig. 2S,T), while expression is stillobserved in the distal optic vesicle more nasally (Fig. 2R).Subsequently, at around stage 12/13, Mitf transcripts were restrictedto the presumptive RPE (Fig. 1D). A marker of differentiatedpigment cells is melanosomal matrix glycoprotein 115 (MMP115),which is involved in melanin production. Unlike Otx2 and Mitf, wedid not observe MMP115 expression at the initial stages of chick eyedevelopment (Fig. 1E). The first MMP115 transcripts were detectedin the presumptive RPE from stage 13 onwards (Fig. 1F; Fig. 4C).This is about five stages earlier than previously reported (Mochii etal., 1988; Mochii et al., 1998). At stages 13-18, Wnt2b expression isdetected in the presumptive RPE and no transcripts are detectedwithin the NR (Fig. 3B) (Jasoni et al., 1999).

Next, we investigated the time point at which the NR domain isestablished during chick eye development. The retinal homeobox-containing gene Rx is initially expressed throughout the optic vesicle(data not shown) (Mathers et al., 1997). At stage 10, Rx expressionwas seen to be downregulated in the presumptive RPE (Fig. 1G) and,by stage 13, expression was restricted to cells in the distal portion ofthe optic vesicle (Fig. 1H). Chx10 is a NR-specific gene expressedin progenitor cells of the NR. At stage 10, Chx10 expression isdetected distally in the temporal region of the optic vesicle (Fig. 1I)(Fuhrmann et al., 2000), the region where Mitf transcripts are firstdownregulated (compare Fig. 1C or Fig. 2T with Fig. 1I).

A second NR-specific marker is FGF8, which appears to beinvolved in NR induction and differentiation (Vogel-Höpker et al.,2000; Martinez-Morales et al., 2005). At stage 10, Fgf8 transcriptswere not detected in the distal neuroepithelium of the chick opticvesicle (Fig. 1K). Fgf8 transcripts in the presumptive NR are firstobserved at stage 11/12 (13-16 somites) (Vogel-Höpker et al., 2000;Crossley et al., 2001) and expression persists in the central region ofthe chick NR at optic cup stages (Fig. 1L) (Vogel-Höpker et al.,2000).

Thus, in the chick, the subdivision of the optic vesicle into NR andRPE is observed at stage 10.

BMP expression during the initial stages of chickeye developmentA signal released from the mesenchyme is thought to be the primaryinducer of Mitf expression in the chick and mouse optic vesicle(Fuhrmann et al., 2000; Kagiyama et al., 2005). At stage 9, Mitfexpression was seen to be strongest in the distal part of the opticvesicle that is covered by the surface ectoderm (Fig. 2G). The firstmesenchymal cells that surround the dorsal region of the opticvesicle are of neural crest origin (Johnston et al., 1979; Hilfer, 1983),suggesting that initially a signal released from the surface ectoderm

induces Mitf expression within the optic vesicle, rather than a signalreleased from the adjacent mesenchyme. To document the presenceof neural crest-derived mesenchyme in more detail, we nextcompared the expression of the neural crest marker gene Sox10 withthe Mitf expression pattern during the initial stages of chick eyedevelopment. At stage 8+, the optic primordia are first visible. Atthis stage, Sox10 expression was detected in migrating neural crestcells in the dorsal neural folds (Fig. 2C). At stages 9 and 10, Sox10

3485RESEARCH ARTICLEBMPs in RPE development

Fig. 1. Division of the chick optic vesicle into a NR and RPEdomain at stage 10. Schematic (above) illustrates the location of thesections in the temporal part of the optic vesicle at stage 10/11.(A-F) RPE development. (A) At stage 10, Otx2 expression weakens inthe distal region of the optic vesicle, the presumptive NR (arrowheads).(B) At stage 13, Otx2 transcripts are abundant in the presumptive RPE(arrows). (C) At stage 10, Mitf expression is downregulated in the distalregion of the optic vesicle (arrowheads). (D) Mitf expression is observedin the presumptive RPE at stage 13 (arrows). (E) MMP115 expression isnot observed in the optic vesicle at stage 11 (arrows). (F) The firstMMP115 transcripts appear in the presumptive RPE at stage 13(arrows). (G-L) NR development. (G) Rx expression is downregulated inthe presumptive RPE at stage 10 (arrows). (H) Rx expression is restrictedto the distal region of the optic vesicle, the presumptive NR at stage 13(arrowheads). (I) Parallel section of the optic vesicle shown in C. Atstage 10, Chx10 expression is detected in the distal region of the opticvesicle (arrowheads), in the region where Mitf expression weakens.(J) Strong Chx10 expression is detected in the distal region, thepresumptive NR at stage 13 (arrowheads). (K) Fgf8 expression is notdetected in the optic vesicle at stage 10 (arrowheads). (L) At stage 13,Fgf8 expression is observed in the distal region of the optic vesicle, justbeneath the forming lens placode. Note that in all panels at stage10/11, the temporal region of the optic vesicle is shown. LP, lensplacode.

DEVELO

PMENT

3486

expression was restricted to migrating neural crest cells that overliethe dorso-temporal part of the optic vesicle and no transcripts wereobserved in the distal region (Fig. 2D,E) where Mitf expression isstrongest (Fig. 2G,H,R).

Two candidate genes that are detected in the presumptive lensectoderm in the mouse are Bmp4 and Bmp7 (Furuta et al., 1997;Furuta and Hogan, 1998). In the chick, Trousse et al. (Trousse et al.,2001) did not detect Bmp7 transcripts in the neuroepithelium of theoptic vesicle or overlying ectoderm until stage 13, whereas wepreviously observed Bmp7 transcripts in the presumptive RPE atstages 11-16 (Vogel-Höpker et al., 2000). Therefore, we investigatedthe expression pattern of BMP family members and their receptorsin comparison with the Mitf expression pattern during the initialstages of chick eye development (stages 8-14). By stage 8+ (6somites) the neural folds have closed and the neuroepithelium of theoptic primordia are first visible. At this time point, theneuroepithelium appeared to be close to the surface ectoderm (Fig.2) and transcripts of Bmp4, Bmp5 and Bmp7 were detected in thedorsal neural folds. Within the neural folds, Bmp7 expression wasdiffuse (Fig. 2I), whereas there appeared to be a regional restrictionof Bmp4 and Bmp5 transcripts (Fig. 2L,O). Both Bmp4 and Bmp7

transcripts, but not Bmp5 transcripts, were observed in the overlyingectoderm between stages 8 and 11 (Fig. 2B,I-Q). Both Bmp4 andBmp7 transcripts were also detected in the ectoderm overlying theocular mesenchyme at stage 10 (Fig. 2K,N). At optic cup stages(stages 13-16), Bmp4 transcripts are present in the dorsal NR,whereas Bmp5 and Bmp7 transcripts are detected in the presumptiveRPE and surrounding mesenchyme (data not shown) (Vogel-Höpkeret al., 2000; Trousse et al., 2001).

BMP receptor type 1a and type 1b are expressed in the eye fieldduring the initial stages of vertebrate eye development (Furuta andHogan, 1998; Trousse et al., 2001; Hyer et al., 2003). Consistent

RESEARCH ARTICLE Development 134 (19)

Fig. 2. Comparison of the Sox10, Mitf and Bmp expressionpatterns during the initial stages of chick eye development.(A) Schematic illustrating the location of the section in H, showing thenasal part of the optic vesicle at stage 10/11. (B) Bmp4 expression inthe surface ectoderm overlying the optic vesicle following whole-mountin situ hybridisation (13 somites). The optic primordium is shown atstage 8 (6 somites; C,F,I,L,O), at stage 9 (7-9 somites; D,G,J,M,P) andstage 10 (10-12 somites; E,H,K,N,Q). (C) At stage 8, Sox10-positiveneural crest cells are observed in the dorsal neural folds of theprosencephalon (arrows). (D) Sox10-expressing neural crest cells aredetected in the dorsal-most region of the prosencephalon (arrows) andno transcripts are detected distally. (E) Sox10 expression is detected inneural crest cells overlying the dorsal part of the optic vesicle (arrows).(F) At stage 8, Mitf expression is not observed in the optic primordium(arrow). (G) Mitf expression is strongest in the distal part of the opticvesicle at stage 9 (arrows). The arrowheads indicate neural crest cellsdorsally. (H) At stage 10, Mitf expression is observed in the presumptiveRPE (dorsal optic vesicle) and in the distal part (arrows). See A for thelocation of this section; see also R,S,T. (I) Bmp7 expression is observedin the ectoderm overlying the optic primordium at stage 8 (arrows), anddiffuse expression is detected in the neural folds (arrowheads).(J) Parallel section of the embryo shown in G. At stage 9, strong Bmp7expression is observed in the overlying ectoderm (arrows). Bmp7transcripts are also detected in the dorsal ectoderm that covers theneural crest cells (arrowheads). (K) Bmp7 expression in the ectodermoverlying the distal region of the optic vesicle at stage 10 (arrows).Transcripts are still observed in the ectoderm overlying the mesenchyme(arrowhead). (L) At stage 8, Bmp4 transcripts are detected in theoverlying ectoderm (arrow) and in the neural folds (arrowheads).(M) Strong Bmp4 expression is detected in the ectoderm overlying thedistal portion of the optic vesicle at stage 9 (arrows). Weak or noexpression is observed in the dorsal-most ectoderm overlying themesenchymal cells (arrowheads). (N) At stage 10, Bmp4 expression isstill strong in the ectoderm overlying the distal portion of the opticvesicle (arrows), whereas weak expression is observed in the ectodermoverlying the surrounding mesenchyme (arrowhead). Note that Bmp4expression appears to be stronger in the ectoderm overlying the dorsalportion of the optic vesicle. (O) At stage 8, Bmp5 expression is strong inthe dorsal midline, the neural folds (arrowheads). Transcripts appear tobe absent from the overlying ectoderm. (P) Bmp5 expression weakensin the dorsal midline at stage 9 (arrowhead). (Q) No Bmp5 transcriptsare detected in the neuroepithelium of the chick optic vesicle andsurrounding tissues (arrow) at stage 10. (R) Nasal region of the opticvesicle at stage 10. Mitf expression in the distal and dorsal part of theoptic vesicle (arrowheads). (S) Higher magnification of the Mitfexpression pattern in the more-temporal region of the optic vesicleshown in H. Mitf expression is detected in both the dorsal and distalregion of the optic vesicle (arrowheads), although expression weakensventrally (arrow). (T) In the most-temporal region of the optic vesicle,Mitf expression is downregulated in the disto-ventral region (arrow).

DEVELO

PMENT

with these observations, we detected Bmpr1b transcripts in theneuroepithelium of the optic vesicle, the overlying ectoderm andsurrounding mesenchyme at stages 8-10 (data not shown).

In summary, BMP family members are expressed at the right time(stage 8/9) and place (surface ectoderm) to be involved in RPEspecification by inducing Mitf expression in the neuroepithelium ofthe chick optic vesicle.

BMP application induces RPE development in thepresumptive optic stalk and NRIf BMP levels determine whether the cells of the neuroepithelium ofthe optic vesicle acquire a RPE instead of a NR phenotype,overexpression of BMPs should result in ectopic generation of RPEfrom cells of the optic vesicle (presumptive NR/optic stalk region).We implanted BMP-soaked beads into the head mesenchyme oroptic vesicle at stages 8-12 and analysed these embryos for changesin NR and RPE gene expression patterns. Vertebrate BMPs havebeen divided into two subgroups, suggesting that different ligandsmight have different functions during embryogenesis. In our

experiments, we implanted BMP4 and BMP5, which belong todifferent BMP subclasses. BMP4 and BMP2 belong to the Dppfamily, whereas both BMP5 and BMP7 belong to the 60A family(Zhao, 2002).

As described above, Otx2 and Mitf expression is downregulatedin cells that develop into NR, but maintained in cells that willdevelop into RPE during vertebrate eye development. By contrast,the RPE-specific marker MMP115, which is involved in melanin

3487RESEARCH ARTICLEBMPs in RPE development

Fig. 3. Effects of BMP and noggin application on the Wnt2bexpression pattern during early stages of eye development.(A) Schematic of a stage 10/11 chick embryo showing the implantationsite of the BMP5-soaked bead in E. (B) At stage 13, Wnt2b transcriptsare restricted to the presumptive RPE (arrow). The arrowhead indicatesWnt2b expression in the ectoderm. (C) Wnt2b transcripts are detectedin the RPE (arrows) and surface ectoderm (arrowhead) on thecontralateral side of the BMP5-treated eye shown in E. (D) Wnt2bexpression in the contralateral, untreated eye following implantation ofnoggin-expressing cells. Wnt2b transcripts are restricted to the RPE(arrowhead) and no transcripts are detected within the NR. The arrowshows Wnt2b expression within the ectoderm and anterior lens.(E) Following BMP5 application (asterisk), Wnt2b expression is alsodetected in the distal region of the optic vesicle, the presumptive NR(arrows). The arrowhead indicates Wnt2b expression in the ectoderm.(F) Parallel section of the noggin-treated eye shown in Fig. 5B,D. Wnt2bexpression is downregulated in the entire outer optic cup (arrowheads).In the surface ectoderm (arrow) and anterior lens, Wnt2b expression isstill detected.

Fig. 4. Effects of BMP5 application on the distribution of genesexpressed within the NR and RPE at optic cup stages. (A)Schematic illustrating the location of the BMP5-soaked bead followingimplantation at stage 10/11 as shown in E-I; B-D are PBS-soaked beadcontrols (B) In control embryos, Mitf expression is weakly detected inthe RPE at stage 15 (arrowheads). (C) The RPE-specific marker MMP115is restricted to the RPE at this stage (arrowheads). (D) Strong Rxexpression is detected in the NR at this stage (arrowhead). (E) Followingimplantation of a BMP5-soaked bead (asterisk), optic cup formation isnot observed and Mitf expression is detected in the distal optic vesicle(arrowheads). (F) Parallel section of the embryo shown in E and G.MMP115 expression is induced in the presumptive NR (arrowhead) andthe optic stalk region following BMP5 exposure. (G) Implantation of aBMP5-soaked bead (asterisk) leads to downregulation of Rx expressionin the presumptive NR. (H) BMP5 application downregulates Chx10expression in the distal optic vesicle/cup, the presumptive NR (right,arrowheads). By contrast, Chx10 expression is strongly observed withinthe presumptive NR of the contralateral eye (left, arrow). (I) Parallelsection of the embryo shown in H. MMP115 expression is induced byBMP5 in the presumptive NR (arrowheads), whereas in thecontralateral, unoperated eye, MMP115 transcripts are absent from theNR (left, arrow). L, Lens.

DEVELO

PMENT

3488

pigment production, is first detected in the presumptive RPE at stage13 (Fig. 1F). Similarly, Wnt2b expression is detected in thepresumptive RPE at early optic cup stages (Fig. 3B) (Jasoni et al.,1999; Fuhrmann et al., 2000). Application of BMP4 or BMP5 atstages 8-12 induced (MMP115, Wnt2b) and maintained (Otx2, Mitf)RPE genes in both the distal and proximal region of the optic vesiclein 43% of the embryos (n=19/44). In some cases in which the beadhad been placed close to the eye region, optic cup and nerveformation was not observed, so that the BMP-treated eyes had stilloptic vesicle-like morphology (Fig. 3E; Fig. 4E-G). The distal andproximal regions of these BMP-treated eyes had lost thecharacteristic morphology of the multilayered NR and opticstalk/nerve, respectively. Instead, these regions developed RPE-likefeatures, including the appearance of pigment granules (Fig. 5I,J).In these embryos, the expression of Mitf, Otx2, MMP115 and Wnt2bwas maintained or induced in the proximal region that normally

develops into the optic nerve (Fig. 4F; Fig. 5C,F), and/or in the distalregion that normally gives rise to the NR (Fig. 3E; Fig. 4E,F,I; Fig.5D,G,N,O). In BMP-treated embryos that developed RPE-likefeatures, expression of NR-specific genes such as Rx, Chx10 andFgf8 was downregulated or absent in the distal optic vesicle (Fig.4G,H; Fig. 5J,K). Three BMP-treated embryos that were left todevelop until stage 25/26 developed a single-layered pigmentedregion within the neuroepithelium of the forebrain, expressing bothOtx2 and MMP115 (data not shown).

BMP beads placed into the mesenchyme lying more temporal tothe optic vesicle did not prevent optic cup formation, and eyemorphology, including lens development, appeared normal (Fig.4H,I; Fig. 5L-O). In 44% of these cases, the RPE-specific markerMMP115 was expressed in single cells within the NR (n=12/27; Fig.4I; Fig. 5N,O). Application of PBS-soaked beads into the opticvesicle or into the mesenchyme temporal to the optic vesicle did not

RESEARCH ARTICLE Development 134 (19)

Fig. 5. Effects of BMP4 application at the opticvesicle stage (stage 8/9) on the distribution ofgenes expressed within the NR and RPE atstage 24. (A) Schematic illustrating theimplantation site of the BMP4-soaked beadfollowing the operations at stage 8/9. (B) FollowingBMP4 application into the optic vesicle, optic cupformation is not observed. Instead, a huge vesiclewith RPE-like morphology, expressing Otx2 in boththe proximal (C) and distal (D) regions developed.The arrowhead indicates Otx2 expression in the RPEof the contralateral, unoperated eye. Note that atthis stage (stage 24), Otx2 transcripts are alsoobserved in cells of the native NR (Bovolenta et al.,1997). This expression pattern is also seen in asmall ventral region of the BMP-treated eye thathas still NR-like morphology (arrow). (C) The entireoptic stalk region has a single-layered RPE-likemorphology expressing Otx2 (arrows) followingBMP4 exposure. (D) Following BMP4 application,the region that normally develops into the NR has aRPE-like morphology and expresses Otx2 (arrows).(E) BMP4 application into the optic vesicle inhibitsNR and optic stalk/nerve development and inducesMitf expression in the entire optic vesicle. Note thatthe entire eye has RPE-like morphology (arrow). Inthe unoperated, contralateral eye, Mitf expressionis restricted to the RPE (arrowhead) and notranscripts are observed in the multilayered NR.(F) Mitf expression (arrows) in the optic stalk/nerveregion that developed a single-layered morphologyfollowing BMP4 application. (G) Following BMPtreatment, Mitf expression is observed in the regionthat normally develops into the multilayered NR(arrows). (H) Parallel section of the embryo shownin B. Chx10 expression is only detected in a smallventral portion of the BMP-treated eye (arrow). Inthe contralateral eye, strong Chx10 expression isrestricted to cells of the NR (arrowhead). (I) Theoptic stalk/nerve region of the BMP-treated eye hasRPE-like morphology and pigment granules(arrows) are observed. (J) Following BMP4application, expression of the NR-specific markerChx10 is not observed in the region that normallydevelops into the NR. Instead, this region is pigmented (arrows) and has RPE-like morphology. (K) Fgf8 expression is not observed in the distalregion of the BMP-treated eye (arrows). (L) Following BMP4 application more temporal to the optic vesicle, eye morphology including lensdevelopment appears to be normal although the operated eye is slightly smaller. (M) MMP115 expression is restricted to the outer optic cup, thedeveloping RPE and no transcripts are observed within the NR (arrows). (N,O) BMP4 application induced MMP115 expression in single cells of theNR (arrows). The asterisk indicates the neurepithelium of the diencephalon. L, Lens.

DEVELO

PMENT

convert NR into RPE (n=12; data not shown) (Vogel-Höpker et al.,2000). These results show that BMPs are sufficient to induce RPEdevelopment in vivo.

BMP signalling is required for RPE developmentTo address the functional relevance of BMPs by loss-of-functionexperiments during the initial stages of eye development, weinterfered with BMP signalling using the protein noggin. Nogginspecifically inhibits BMP signalling by binding to BMP dimers,thereby preventing their interaction with cell surface receptors(reviewed by Balemans and Van Hul, 2002).

Noggin-expressing CHO cells were injected either into the headmesenchyme or into the ventricle of developing chick embryos atstages 8-11. At several time points after the injection, the embryoswere analysed for the expression of RPE- and NR-specific markers.In general, the noggin-treated eyes were smaller than thecontralateral unoperated eye and displayed aberrant development ofthe optic stalk/nerve (coloboma), NR, RPE and lens as previouslyreported by Adler and Belecky-Adams (Adler and Belecky-Adams,2002). We therefore implanted the cells slightly further away, at thelevel of the midbrain. In 32% of the embryos, parts of the outer opticcup no longer had a single-layered morphology and instead a regiondeveloped with NR-like morphology (Fig. 3F, arrowheads; Fig.6B,D, arrowheads; Fig. 6E-J, arrows). In these regions, the pigmentmarker MMP115, Wnt2b and Mitf expression was downregulated(n=7/22; Fig. 3F; Fig. 6B,D,J) and pigment granules were notobserved (Fig. 6E,F,J, arrow). Instead, we observed expression ofthe retinal markers Rx and Chx10 in these regions (Fig. 6G; data notshown). Pax6 is initially expressed throughout the optic vesicle, butexpression is lost from the proximal RPE at late optic cup stages(Fig. 6I, arrowhead) (reviewed by Martinez-Morales et al., 2004).

At stage 25, Pax6 expression is strong within the chick NR and noPax6 transcripts are observed within the single-layered RPE (Fig.6H,I). Overexpression of Pax6 in the chick RPE inducestransdifferentiation of the RPE into NR (Azuma et al., 2005).Following noggin treatment, we observed strong Pax6 expression ina small, multilayered region of the RPE (Fig. 6H,I), whereasexpression of the RPE-specific gene MMP115 was downregulatedand pigment granules were absent (Fig. 6J, arrow). However, weakinduction of Pax6 expression within the RPE (Fig. 6H, arrowhead)did not result in the downregulation of MMP115 (data not shown).In control experiments, CHO cells were grafted into themesenchyme temporal to the optic vesicle of stage 10-11 chickembryos. In these cases, eye morphology was normal and MMP115expression and pigment granules were restricted to the RPE (n=12;data not shown).

In a second set of experiments, we blocked BMP signalling withinthe RPE by viral overexpression of a dnBmpr1b construct. Injectionof dnBmpR1b-RCAS (B) into the eye field at stages 6-11 resultedin partial loss of RPE development in 21% (7/33) of cases. The mostdramatic effects were observed when the operation was carried outat stage 6/7. The outer layer of the optic cup was no longer single-layered and instead developed a NR-like morphology (3/4 cases).Expression of both Otx2 and MMP115 was downregulated in theouter optic cup (Fig. 7F,G,J). By contrast, the NR marker Chx10 wasnow detected in the outer layer of the optic cup (Fig. 7H).Thickening of the outer layer was not as prominent when theoperation was carried out at stages 8-11 (observed in 4/29 cases; Fig.7K,L). Injection of control RCAS (B) retrovirus at the same stagesof development did not result in any alterations in gene expression,and pigment granules were still observed in the outer layer of theoptic cup (n=8; Fig. 7A-D).

3489RESEARCH ARTICLEBMPs in RPE development

Fig. 6. Effects of interfering with BMP signallingon the distribution of transcripts known to beinvolved in NR and RPE development. (A) Mitfexpression in the contralateral, untreated eyefollowing implantation of noggin-expressing CHOcells. Mitf expression is restricted to the RPE(arrowhead). (B) Inhibition of BMP signallingdownregulates Mitf expression in the outer optic cup(arrowheads) and only a small single-layered regionstill expresses Mitf (arrows). (C) MMP115 expression inthe contralateral, untreated eye followingimplantation of noggin-expressing cells. MMP115transcripts are restricted to the RPE. (D) Parallelsection of the noggin-treated eye shown in B.MMP115 expression is downregulated in the outeroptic cup (arrowheads), although MMP115 expressionis still maintained in a small region that still has RPE-like morphology (arrow). (E) Following implantation ofnoggin-expressing CHO cells, the single-layered RPEsuddenly thickens in the proximal region of the eye(arrow). Pigmentation is still observed in theunaffected areas of the RPE (arrowheads). (F) Highermagnification of the eye shown in E. The RPE ispigmented (arrowhead), whereas the adjacent multilayered area is unpigmented (arrow). (G) Parallel section of the noggin-treated eye shown in Eand F. Rx expression is observed in the native NR and in the outer optic cup that has NR-like morphology (arrow). Expression is absent in the single-layered RPE (arrowhead). (H) Proximal region of a noggin-treated eye at stage 25. At this stage, Pax6 expression is strongly detected in the nativeNR, whereas transcripts are absent from the pigmented single-layered RPE (see also arrowhead in I). Following noggin treatment, strong Pax6expression is induced in a small, multilayered region of the outer layer of the optic cup (box). The arrowhead indicates weak Pax6 expression in theRPE. Note that Pax6 expression is also observed in the brain on the right side. (I) Higher magnification of the multilayered region in the outer opticcup shown in H. This region strongly expresses Pax6 (arrow), whereas expression is not observed in the single-layered, pigmented RPE (arrowhead).(J) Parallel section of the eye shown in H and I. MMP115 expression is downregulated in the Pax6-expressing region that is multilayered (arrow).

DEVELO

PMENT

3490

Taken together, the data suggest that BMP signalling is requiredduring the initial stages of chick eye development for properdevelopment of the RPE.

DISCUSSIONIn this study, we have determined the time point at which the opticvesicle is subdivided into NR and RPE domains. We presentevidence from gain- and loss-of-function studies that BMPs arenecessary and sufficient for RPE development during optic vesiclestages in the chick. In addition, the BMP expression pattern incomparison to the expression of RPE and/or NR marker genessuggests that the BMP-expressing surface ectoderm, rather than themesenchyme, is involved in RPE specification by inducing Mitfexpression in the underlying optic vesicle.

In vertebrates, the neuroepithelium of the optic vesicle initiallyco-expresses several TFs that are involved in RPE and NRdevelopment. For example, Mitf and Otx2 are initially expressed inthe entire optic vesicle, but expression is subsequently maintainedonly in the presumptive RPE. MITF and OTX2 are key signalsinvolved in initiating and maintaining pigmentation in the RPE ofvertebrates (for reviews, see Chow and Lang, 2001; Martinez-Morales et al., 2004). The retinal homeobox-containing gene Rx,which is also initially expressed throughout the optic vesicle,becomes downregulated in the presumptive RPE, whereasexpression is maintained in the presumptive NR (Mathers et al.,1997). In this study, we show that in the chick optic vesicle, RPEdevelopment is initiated first and that induction of NR development,marked by Chx10 expression, leads to the separation of the chickoptic vesicle into NR and RPE. Expression of Chx10, a marker ofretinal progenitor cells, is detected at stage 10 in the distal region ofthe chick optic vesicle (this study) (Fuhrmann et al., 2000) (for areview, see Chow and Lang, 2001) at the time when Mitf expressionis downregulated in this region (this study). Members of the FGFfamily – Fgf1, Fgf2 and Fgf19 – are expressed in the surfaceectoderm overlying the distal portion of the chick optic vesicle

(reviewed by Chow and Lang, 2001; Martinez-Morales et al., 2004;Kurose et al., 2004). The separation of the optic vesicle into NR andRPE domains is initiated through FGF-mediated induction ofChx10, which subsequently leads to the repression of Mitf (Horsfordet al., 2005) and possibly also of Otx2 in the presumptive NR. Anantagonistic interaction between Chx10 and Mitf regulates retinalcell identity. CHX10 negatively regulates Mitf expression bybinding to its promoter, thereby ensuring NR development in thedistal portion of the optic vesicle (Rowan et al., 2004; Horsford etal., 2005). Thus, it appears that, similar to the situation in mouse,RPE development is the fate of the neuroepithelium of the opticvesicle in the absence of NR-inducing signals. Removal of theectoderm after BMP-mediated RPE induction and before FGFproduction should thus lead to RPE development. Indeed, surfaceectoderm removal at stage 10 prevents the separation of the opticvesicle into NR and RPE, and instead a pigmented vesicle develops(Hyer et al., 1998; Nguyen and Arnheiter, 2000). At stage 10, Mitfexpression is mainly observed in the distal optic vesicle, whereas atthis time only a few cells express Chx10 (this study). Thus, in theabsence of FGF-induced Chx10 expression, the neuroepithelial cellswill mainly develop into RPE and only a few neuronal cells areobserved (Hyer et al., 1998). FGF application to the distal opticvesicle restores proper separation of the NR and RPE domains inthe absence of the surface ectoderm (reviewed by Martinez-Moraleset al., 2004). FGF family members are also expressed in thepresumptive NR (reviewed by Chow and Lang, 2001; Martinez-Morales et al., 2004; Kurose et al., 2004). For example, in the chick,Fgf8 and Fgf19 transcripts are observed in the distal optic vesicle atabout the time when Chx10 expression is first detected in this region(Vogel-Höpker et al., 2000; Crossley et al., 2001; Kurose et al.,2004). Indeed, FGF8 application into the chick ocular mesenchymeinhibits Mitf, Otx2 and Bmp7 expression in the presumptive RPEand Bmp7 expression in the surrounding mesenchyme, and thisallows NR development to occur in the outer optic cup (Vogel-Höpker et al., 2000; Martinez-Morales et al., 2005). On the other

RESEARCH ARTICLE Development 134 (19)

Fig. 7. Effects of interfering with BMPsignalling on the distribution of transcriptsknown to be involved in NR and RPEdevelopment. (A-D) Control experimentsoverexpressing a viral RCAS (B) (labelled RCAS-B)construct only (parallel sections). (A) Eyemorphology is normal following viral infectionwith RCAS (B). Expression of the viral reversetranscriptase gene (RT) indicates infected areas ofthe RPE (arrows). (B) In the eyes infected withRCAS (B), Otx2 expression is still restricted to theRPE (arrowheads) and no transcripts are observedin the NR at stage 20. (C) Following viral injectionof RCAS (B), MMP115 expression is unchangedand is restricted to the RPE (arrowheads).(D) Expression of the NR-specific marker Chx10 isnot observed in the outer layer of the optic cupfollowing injection of RCAS (B) (arrowheads).(E-H) The effects following injection of the viraldnBmpR1b-RCAS (B) construct at stage 6. (E) Following injection of dnBmpR1b, a large region of the outer layer of the optic cup is thickened(arrowheads). The infection of the RPE is shown by the expression of RT. (F) Otx2 expression weakens in the outer layer of the optic cup followinginjection of dnBmpR1b (arrowheads). (G) MMP115 expression is downregulated following inhibition of BMP signalling (arrowheads). (H) Expressionof the NR-specific gene Chx10 is induced in the outer layer of the optic cup following viral overexpression of the dnBmpR1b construct (arrowheads).(I) RT expression in the outer layer of the optic cup (arrowheads) following viral infection of the dnBmpR1b construct at stage 7. (J) MMP115expression is downregulated (arrowheads) in the infected region of the outer layer of the optic cup. Strong MMP115 expression is still observed inthe single-layered portion (arrow). (K) RT expression within the proximal RPE (arrowheads) following infection at stage 10. (L) The arrowheadsindicate the expression of the NR marker gene Chx10 in the RPE in the parallel section. L, lens; NR, neural retina.

DEVELO

PMENT

hand, BMP application leads to a downregulation of Fgf8expression within the NR, and this allows RPE development tooccur in the distal region of the neuroepithelium (this study, seebelow).

It has been suggested that RPE development is initiated by signalsreleased from the ocular mesenchyme (for a review, see Martinez-Morales et al., 2004; Kagiyama et al., 2005). Previous studiesconsidered the mesenchyme as a source of RPE-inducing signals forthree reasons. First, Mitf expression was first detected in thepresumptive chick RPE at stage 12/13 (Mochii et al., 1998;Fuhrmann et al., 2000), at the time the presumptive RPE issurrounded by mesenchyme. Second, in the mouse, the initialexpression of Mitf throughout the optic vesicle coincides with thetime when it is entirely covered by a small amount of mesenchyme(Bora et al., 1998; Nguyen and Arnheiter, 2000). Third, embryonictransplantations and explant studies supported the idea that themesenchyme induces RPE-specific gene expression within theneuroepithelium of the optic vesicle (Fuhrmann et al., 2000;Kagiyama et al., 2005). In the chick, Mitf expression is induced inthe distal optic vesicle before mesenchymal cells are present (thisstudy) (Hilfer, 1983; Johnston et al., 1979; Sullivan et al., 2004;Kagiyama et al., 2005), and here expression is strongest in the distaloptic vesicle that is covered by the Bmp4- and Bmp7-expressingsurface ectoderm. Our gain-of-function experiments show thatectopic BMP application at optic vesicle stages can induce thedevelopment of a single-layered RPE, including the appearance ofpigment granules, by inducing the pigment cell marker MMP115and/or maintaining the expression of Otx2 and Mitf in cells thatwould normally have developed into the optic nerve or NR. In vitrostudies have shown that optic vesicles isolated at stages 11-15 andcultured in the presence of mesenchyme express Mitf, Wnt2b andMMP115 (Fuhrmann et al., 2000). The results of these co-cultureexperiments may be explained by BMP-producing/-containingmesenchyme that has a maintenance function at this later stage (seebelow). Fuhrmann et al. (Fuhrmann et al., 2000) reported thatactivin, but not BMPs, can substitute for the mesenchyme to induceRPE development in optic vesicle explants. The discrepancy withthe present in vivo data might be best explained by the fact that

BMPs specify different cell fates in a concentration-dependentmanner and that the BMP concentrations used were either too lowor too high to elicit RPE induction (Wilson et al., 1997; Simeoni andGurdon, 2007). BMP beads applied close to the optic vesicle inducethe development of a single-layered RPE in cells that would havenormally developed into NR or optic stalk. By contrast, followingexposure to a lower BMP concentration owing to a different positionof the BMP-soaked bead, only single cells within the NR itselfexpress the RPE-specific marker MMP115.

What is the cellular mechanism that is responsible for thegeneration of RPE instead of a two-layered optic cup with NR?BMP treatment does not lead to increased apoptosis, excluding thepossibility of selective death of presumptive NR (Ohkubo et al.,2002). The significant defects in eye vesicle morphogenesis uponBMP overexpression raised the question of whether the effect ofBMPs is direct or, alternatively, is secondary to an invaginationdefect. Optic vesicle invagination fails when the NR domain has notbeen correctly specified (Uemonsa et al., 2002). The finding thatlower BMP levels do not interfere with optic cup formation and leadto RPE-specific gene expression in single cells within the NR arguesin favour of a direct BMP-induced differentiation process (e.g. Fig.5O).

In the chick and mouse, several BMP family members and relevantreceptors are expressed at the right time and place to play a role ininducing and maintaining RPE development (Lyons et al., 1995;Dudley and Robertson, 1997; Furuta et al., 1997; Furuta and Hogan,1998; Wawersik et al., 1999; Fuhrmann et al., 2000; Vogel-Höpker etal., 2000; Crossley et al., 2001; Trousse et al., 2001; Belecky-Adamset al., 2002; Müller and Rohrer, 2002; Hyer et al., 2003; Liu et al.,2003). BMPs mainly signal via complexes composed of type 1 andtype 2 transmembrane serine/threonine kinase receptors, which areboth required for signal transduction (Mishina, 2003). Activated type1 receptor kinases subsequently phosphorylate intracellular mediatorsknown as Smad proteins. The type 1 receptors, also known as activinreceptor-like kinases (ALKs), ALK1 (ACVRL1), ALK2 (ACTR1;ACVR1), ALK3 (BMPR1A) and ALK6 (BMPR1B) phosphorylateSMAD1, SMAD5 and SMAD8 (also known as SMAD9 in mouse),which transduce the extracellular signal to the nucleus. Activin

3491RESEARCH ARTICLEBMPs in RPE development

SURFACE ECTODERM OPTIC VESICLE SURFACE ECTODERMDISTAL

OPTIC VESICLE MESENCHYME RPE NR

BMPsBMPs BMPs

BMPs FGFs

Chx10MitfMitf

Chx10FGFs

Mitf

(B) NR SPECIFICATIONstage 10

(A) RPE SPECIFICATIONstage 8/9

(C) MAINTENANCEstage 11 - 17

Fig. 8. Proposed model for the regulation of the RPE and NR domain in the developing chick eye. (A) The RPE is specified first at stage 9.BMP genes expressed in the surface ectoderm at stage 8 (e.g. Bmp4 and Bmp7) induce Mitf expression in the underlying optic vesicle. Mitfexpression is strongest in the region of the optic vesicle that directly contacts the overlying ectoderm. (B) At stage 10, NR specification is initiated bysignals released from the FGF-expressing surface ectoderm. FGF-mediated (e.g. FGF1, 2 and 19) induction of Chx10 in the distal portion of the opticvesicle inhibits Mitf and Otx2 expression in this region, which results in the subdivision of the optic vesicle into a RPE and NR domain. (C) At earlyoptic cup stages, Mitf expression is still maintained in the presumptive RPE by BMPs expressed within the RPE itself and in the surroundingmesenchyme (e.g. BMP5, BMP7). At these stages, BMP family members expressed in the dorsal surface ectoderm and adjacent diencephalon couldalso be emanating into the underlying mesenchyme to maintain Mitf expression and hence RPE development. On the other hand, FGFs present inthe NR itself (e.g. FGF3, 8, 15 and 19) maintain Chx10 expression, which allows NR development in the adjacent inner layer of the optic cup. Theantagonistic interaction between BMPs/MITF within the RPE and FGFs/CHX10 within the NR ensures the development of the vertebrate eye at earlyoptic cup stages.

DEVELO

PMENT

3492

receptor type 2 mediates BMP signalling when bound to BMPR1A orBMPR1B (for reviews, see Balemans and Van Hul, 2002; Larsson andKarlsson, 2005). In the chick, Bmpr1a, Bmpr1b and activin type 2aand type 2b receptors are expressed in the neuroepithelium of the opticvesicle and/or surrounding tissues at optic vesicle stages (data notshown) (Stern et al., 1995; Fuhrmann et al., 2000; Hyer et al., 2003)and ACTR1 is present in the optic primordia of the developing mouseembryo (Yoshikawa et al., 2000). Interestingly, neither the �A nor �Bactivin subunit has been detected at optic vesicle stages in thedeveloping chick embryo (Fuhrmann et al., 2000), whereasphosphorylated SMAD1 was observed in both the neuroepithelium ofthe optic vesicle and in the surface ectoderm (Belecky-Adams et al.,2002; Faure et al., 2002; Sakai et al., 2005). We finally demonstratethe physiological importance of BMPs in RPE development byinterfering with BMP signalling at optic vesicle stages. Application ofthe BMP-inhibitor noggin or of the dnBmpR1b constructdownregulated MMP115, Mitf and Otx2 expression in the RPE andinstead induced the expression of the NR marker genes (e.g. Chx10,Rx). FGF8 application into the mesenchyme near to the opticvesicle/cup induces the development of a second NR in the outer opticcup (Vogel-Höpker et al., 2000; Martinez-Morales et al., 2005).However, during chick eye development, Fgf8 and Fgf19 areexpressed within the NR, but NR induction in the outer optic cup doesnot occur. If BMPs within the RPE and FGFs within the NR actantagonistically, the absence of BMPs within the RPE should allowthe development of NR-like features in the outer optic cup (see Figs 6and 7). BMP inhibition at optic cup stages results in the upregulationof Fgf8 expression within the NR itself (Adler and Belecky-Adams,2002).

BMPs have multiple functions during early and late stages ofvertebrate eye development (Koshiba-Takeuchi et al., 2000; Sakutaet al., 2001; Adler and Belecky-Adams, 2002; Sasagawa et al., 2002;Hammerschmidt et al., 2003; Murali et al., 2005). For example,deletion of the BMPR1A/B function specifically within the mouseretina leads to reduced growth of the NR and failure of retinalneurogenesis (Murali et al., 2005). We show that at optic vesiclestages, BMPs are involved in patterning the vertebrate eye byregulating RPE gene expression within the neuroepithelium of theoptic vesicle. On the basis of our results, we propose the followingmodel (Fig. 8). Within a short period of time, both RPE and NRspecification are induced by signals released from the overlyingectoderm. Initially, the BMP-expressing surface ectoderm isinvolved in inducing and maintaining Mitf expression in theneuroepithelium of the chick optic vesicle. At this time, the opticvesicle is in direct contact with the surface ectoderm (this study)(Johnston et al., 1979; Hilfer, 1983; Sullivan et al., 2004; Kagiyamaet al., 2005). The subdivision of the optic vesicle into NR and RPEdomains is initiated by FGFs (e.g. FGF1, 2 and/or 19) released fromthe surface ectoderm a few hours later at stage 9/10. FGF-mediatedinduction of Chx10 expression in the distal portion of the opticvesicle downregulates genes involved in RPE development (e.g.Mitf). Subsequently, during early optic cup stages, BMPs (e.g.BMP5 and BMP7) in the presumptive RPE itself, the mesenchymeand/or released from the surrounding tissues (dorsal surfaceectoderm, diencephalon) into the mesenchyme, are involved instabilising the RPE domain in the outer optic cup. FGF familymembers (e.g. FGF3, 8, 15 and 19), being now expressed in the NRitself, maintain Chx10 expression and allow NR development tooccur adjacent to the RPE. Thus, at the early optic cup stages whenthe NR and RPE are in close contact, BMPs/MITF within the RPEand FGFs/CHX10 within the NR, act antagonistically to ensurevertebrate eye development.

BMP ligands are expressed in overlapping domains and geneticstudies strongly argue that BMP family members are functionallyredundant in vivo (Solloway et al., 1998; Solloway and Robertson,1999; Kim et al., 2001). It is possible that cooperative signalling ofdifferent BMP family members, which may also involve BMPheterodimers (Butler and Dodd, 2003), might be involved inregulating RPE development at optic vesicle stages. However, whichspecific BMP family members are involved in RPE specification,differentiation and maintenance remains to be elucidated.

We thank Sabine Richter for excellent technical assistance; R. Harland, B.Houston, R. Johnson, M. Mochii, L. Niswander, H. Roelink, T. Ogura forproviding reagents; and Sabine Fuhrmann, Juan Ramon Martinez-Morales,Dorothea Schulte and Jochen Wittbrodt for critical reading of the manuscript.F.M. and H.R. were supported from the SFB 269. A.V.-H. was supported by theDeutsche Forschungsgemeinschaft.

ReferencesAdler, R. and Belecky-Adams, T. L. (2002). The role of bone morphogenetic

proteins in the differentiation of the ventral optic cup. Development 129, 3161-3171.

Azuma, N., Tadokoro, K., Asaka, A., Yamada, M., Yamaguchi, M., Handa, H.,Matsushima, S., Watanabe, T., Kida, Y., Ogura, T. et al. (2005).Transdifferentiation of the retinal pigment epithelium to the neural retina bytransfer of the Pax6 transcriptional factor. Hum. Mol. Genet. 14, 1059-1068.

Balemans, W. and Van Hul, W. (2002). Extracellular regulation of BMP signalingin vertebrates: a cocktail of modulators. Dev. Biol. 250, 231-250.

Belecky-Adams, T. L., Adler, R. and Beebe, D. C. (2002). Bone morphogeneticprotein signaling and the initiation of lens fiber cell differentiation. Development129, 3795-3802.

Bora, N., Conway, S. J., Liang, H. and Smith, S. B. (1998). Transientoverexpression of the Microphthalmia gene in the eyes of Microphthalmiavitiligo mutant mice. Dev. Dyn. 213, 283-292.

Bovolenta, P., Mallamaci, A., Briata, P., Corte, G. and Boncinelli, E. (1997).Implication of OTX2 in pigment epithelium determination and neural retinadifferentiation. J. Neurosci. 17, 4243-4252.

Bumsted, K. M. and Barnstable, C. J. (2000). Dorsal retinal pigment epitheliumdifferentiates as neural retina in the microthalmia (mi/mi) mouse. Invest.Ophthalmol. Vis. Sci. 41, 903-908.

Butler, S. J. and Dodd, J. (2003). A role for BMP heterodimers in roof plate-mediated repulsion of commissural axons. Neuron 38, 389-401.

Chow, R. L. and Lang, R. A. (2001). Early eye development in vertebrates. Annu.Rev. Cell Dev. Biol. 17, 255-296.

Crossley, P. H., Martinez, S., Ohkubo, Y. and Rubenstein, J. L. (2001).Coordinate expression of Fgf8, Otx2, Bmp4, and Shh in the rostralprosencephalon during development of the telencephalic and optic vesicles.Neuroscience 108, 183-206.

Dudley, A. T. and Robertson, E. J. (1997). Overlapping expression domains ofbone morphogenetic protein family members potentially account for limitedtissue defects in BMP7 deficient mice. Dev. Dyn. 208, 349-362.

Faure, S., de Santa Barbara, P., Roberts, D. J. and Whitman, M. (2002).Endogenous patterns of BMP signaling during early chick development. Dev.Biol. 244, 44-65.

Fuhrmann, S., Levine, E. M. and Reh, T. (2000). Extraocular mesenchymepatterns the optic vesicle during early eye development in the embryonic chick.Development 127, 4599-4609.

Furuta, Y. and Hogan, B. L. (1998). BMP4 is essential for lens induction in themouse embryo. Genes Dev. 12, 3764-3775.

Furuta, Y., Piston, D. W. and Hogan, B. L. M. (1997). Bone morphogeneticproteins (BMPs) as regulators of dorsal forebrain development. Development124, 2203-2212.

Hamburger, V. and Hamilton, H. (1951). A series of normal stages in thedevelopment of the chick embryo. J. Morphol. 88, 49-92.

Hammerschmidt, M., Kramer, C., Nowak, M., Herzog, W. and Wittbrodt, J.(2003). Loss of maternal Smad5 in zebrafish embryos affects patterning andmorphogenesis of optic primordia. Dev. Dyn. 227, 128-133.

Henrique, D., Adam, J., Myat, A., Chitnis, A., Lewis, J. and Ish-Horowicz, D.(1995). Expression of Delta homologue in prospective neurons in chick. Nature375, 787-790.

Hilfer, S. R. (1983). Development of the eye of the chick embryo. Scan. ElectronMicrosc. III, 1353-1369.

Horsford, C. J., Nguyen, M. T., Sellar, G. C., Kothary, R., Arnheiter, H. andMcInnes, R. R. (2005). Chx10 repression of Mitf is required for the maintenanceof mammalian neuroretinal identity. Development 132, 177-187.

Hyer, J., Mima, T. and Mikawa, T. (1998). FGF-1 patterns the optic vesicle bydirecting the placement of the neural retina domain. Development 125, 869-877.

RESEARCH ARTICLE Development 134 (19)

DEVELO

PMENT

Hyer, J., Kuhlman, J., Afif, E. and Mikawa, T. (2003). Optic cup morphogenesisrequires pre-lens ectoderm but not lens differentiation. Dev. Biol. 259, 351-363.

Iwakiri, R., Kobayashi, K., Okinami, S. and Kobayashi, H. (2005). Suppressionof Mitf by small interfering RNA induces transdifferentiation of chick embryonicretinal pigment epithelium. Exp. Eye Res. 81, 15-21.

Jasoni, C., Hendrickson, A. and Roelink, H. (1999). Analysis of chicken Wnt-13expression demonstrates coincidence with cell division in the developing eye andis consistent with a role in induction. Dev. Dyn. 215, 215-224.

Johnston, M. C., Noden, D. M., Hazelton, R. D., Coulombre, J. L. andCoulombre, A. J. (1979). Origins of avian ocular and periocular tissues. Exp. EyeRes. 29, 27-43.

Kagiyama, Y., Gotouda, N., Sakagami, K., Yasuda, K., Mochii, M. and Araki,M. (2005). Extraocular dorsal signal affects the developmental fate of the opticvesicle and patterns the optic neuroepithelium. Dev. Growth Differ. 47, 523-536.

Kim, R. Y., Robertson, E. J. and Solloway, M. J. (2001). Bmp6 and Bmp7 arerequired for cushion formation and septation in the developing mouse heart.Dev. Biol. 235, 449-466.

Koshiba-Takeuchi, K., Takeuchi, J. K., Matsumoto, K., Momose, T., Uno, K.,Hoepker, V., Ogura, K., Takahashi, N., Nakamura, H., Yasuda, K. et al.(2000). Tbx5 and the retinotectum projection. Science 287, 134-137.

Kurose, H., Bito, T., Adachi, T., Shimizu, M., Noji, S. and Ohuchi, H. (2004).Expression of Fibroblast growth factor 19 (Fgf19) during chicken embryogenesisand eye development, compared with Fgf15 expression in the mouse. GeneExpr. Patterns 4, 687-693.

Larsson, J. and Karlsson, S. (2005). The role of Smad signaling in hematopoiesis.Oncogene 24, 5676-5692.

Liu, J., Wilson, S. and Reh, T. (2003). BMP receptor 1b is required for axonguidance and cell survival in the developing retina. Dev. Biol. 256, 34-48.

Lyons, K. M., Hogan, B. L. and Robertson, E. J. (1995). Colocalization of BMP 7and BMP 2 RNAs suggests that these factors cooperatively mediate tissueinteractions during murine development. Mech. Dev. 50, 71-83.

Martinez-Morales, J. R., Signore, M., Acampora, D., Simeone, A. andBovolenta, P. (2001). Otx genes are required for tissue specification in thedeveloping eye. Development 128, 2019-2030.

Martinez-Morales, J. R., Dolez, V., Rodrigo, I., Zaccarini, R., Leconte, L.,Bovolenta, P. and Saule, S. (2003). OTX2 activates the molecular networkunderlying retina pigment epithelium differentiation. J. Biol. Chem. 278, 21721-21731.

Martinez-Morales, J. R., Rodrigo, I. and Bovolenta, P. (2004). Eyedevelopment: a view from the retina pigmented epithelium. BioEssays 26, 766-777.

Martinez-Morales, J. R., Del Bene, F., Nica, G., Hammerschmidt, M.,Bovolenta, P. and Wittbrodt, J. (2005). Differentiation of the vertebrate retinais coordinated by an FGF signaling center. Dev. Cell 8, 565-574.

Mathers, P. H., Grinberg, A., Mahon, K. A. and Jamrich, M. (1997). The Rxhomeobox gene is essential for vertebrate eye development. Nature 387, 603-607.

Mishina, Y. (2003). Function of bone morphogenetic protein signaling duringmouse development. Front. Biosci. 8, d855-d869.

Mochii, M., Agata, K., Kobayashi, H., Yamamoto, T. S. and Eguchi, G. (1988).Expression of gene coding for a melanosomal matrix protein transcriptionallyregulated in the transdifferentiation of chick embryo pigmented epithelial cells.Cell Differ. 24, 67-74.

Mochii, M., Mazaki, Y., Mizuno, N., Hayashi, H. and Eguchi, G. (1998). Role ofMitf in differentiation and transdifferentiation of chicken pigmented epithelialcell. Dev. Biol. 193, 47-62.

Müller, F. and Rohrer, H. (2002). Molecular control of ciliary neurondevelopment: BMPs and downstream transcriptional control in theparasympathetic lineage. Development 129, 5707-5717.

Murali, D., Yoshikawa, S., Corrigan, R. R., Plas, D. J., Crair, M. C., Oliver, G.,Lyons, K. M., Mishina, Y. and Furuta, Y. (2005). Distinct developmentalprograms require different levels of Bmp signaling during mouse retinaldevelopment. Development 132, 913-923.

Nakayama, A., Nguyen, M. T., Chen, C. C., Opdecamp, K., Hodgkinson, C. A.and Arnheiter, H. (1998). Mutations in microphthalmia, the mouse homolog ofthe human deafness gene MITF, affect neuroepithelial and neural crest-derivedmelanocytes differently. Mech. Dev. 70, 155-166.

Nguyen, M.-T. and Arnheiter, H. (2000). Signaling and transcriptional regulationin early mammalian eye development: a link between FGF and MITF.Development 127, 3581-3591.

Oh, S. H., Johnson, R. and Wu, D. K. (1996). Differential expression of bonemorphogenetic proteins in the developing vestibular and auditory sensoryorgans. J. Neurosci. 16, 6463-6475.

Ohkubo, Y., Chiang, C. and Rubenstein, J. L. R. (2002). Coordinate regulation

and synergistic actions of BMP4, SHH and FGF8 in the rostral prosencephalonregulate morphogenesis of the telencephalic and optic vesicles. Neuroscience111, 1-17.

Pittack, C., Grunwald, G. B. and Reh, T. A. (1997). Fibroblast growth factors arenecessary for neural retina but not pigmented epithelium differentiation in chickembryos. Development 124, 805-816.

Planque, N., Turque, N., Opdecamp, K., Bailly, M., Martin, P. and Saule, S.(1999). Expression of the Microthalmia-associated basic helix-loop-helix leucinezipper transcription factor Mi in avian neuroretina cells induces a pigmentedphenotype. Cell Growth Differ. 10, 525-536.

Reissmann, E., Ernsberger, U., Francis-West, P. H., Rueger, D., Brickell, P. M.and Rohrer, H. (1996). Involvement of bone morphogenetic protein-4 andbone morphogenetic protein-7 in the differentiation of the adrenergicphenotype in developing sympathetic neurons. Development 122, 2079-2088.

Rowan, S., Chen, A., Young, T. L., Fisher, D. E. and Cepko, C. L. (2004).Transdifferentiation of the retina into pigmented cells in ocular retardation micedefines a new function of the homeodomain gene Chx10. Development 131,5139-5152.

Sakai, D., Tanaka, Y., Endo, Y., Osumi, N., Okamoto, H. and Wakamatsu, Y.(2005). Regulation of Slug transcription in embryonic ectoderm by ß-catenin-Lef/Tcf and BMP-Smad signaling. Dev. Growth Differ. 47, 471-482.

Sakuta, H., Suzuki, R., Takahashi, H., Kato, A., Shintani, T., Iemura, S.,Yamamoto, T. S., Ueno, N. and Noda, M. (2001). Ventroptin: a BMP-4antagonist expressed in a double-gradient pattern in the retina. Science 293,111-115.

Sasagawa, S., Takabatake, T., Takabatake, Y., Muramatsu, T. and Takeshima,K. (2002). Axes establishment during eye morphogenesis in Xenopus bycoordinate and antagonistic actions of BMP4, Shh, and RA. Genesis, 33, 86-96.

Simeoni, I. and Gurdon, J. B. (2007). Interpretation of BMP signaling in earlyXenopus development. Dev. Biol. 308, 82-92.

Solloway, M. J. and Robertson, E. J. (1999). Early embryonic lethality inBmp5;Bmp7 double mutant mice suggests functional redundancy within the60A subgroup. Development 126, 1753-1768.

Solloway, M. J., Dudley, A. T., Bikoff, E. K., Lyons, K. M., Hogan, B. L. andRobertson, E. J. (1998). Mice lacking Bmp6 function. Dev. Genet. 22, 321-339.

Stern, C. D., Yu, R. T., Kakizuka, A., Kintner, C. R., Methews, L. S., Vale, W.W., Evans, R. M. and Umesono, K. (1995). Activin and its receptors duringgastrulation and the later phases of mesoderm development in the chickembryo. Dev. Biol. 172, 192-205.

Sullivan, C. H., Braunstein, L., Hazard-Leonards, R. M., Holen, A. L., Samaha,F., Stephens, L. and Grainger, R. M. (2004). A re-examination of lensinduction in chicken embryos: in vitro studies of early tissue interactions. Int. J.Dev. Biol. 48, 771-782.

Trousse, F., Esteve, P. and Bovolenta, P. (2001). Bmp4 mediates apoptotic celldeath in the developing chick eye. J. Neurosci. 21, 1292-1301.

Uemonsa, T., Sakagami, K., Yasuda, K. and Araki, M. (2002). Development ofdorsal-ventral polarity in the optic vesicle and its presumptive role in eyemorphogenesis as shown by embryonic transplantation and in ovo explantculturing. Dev. Biol. 248, 319-330.

Vogel, A., Rodriguez, C., Warnken, W. and Izpisua-Belmonte, J. C. (1995).Dorsal cell fate specified by chick Lmx-1 during vertebrate limb development.Nature 378, 716-720.

Vogel, A., Rodriguez, C. and Izpisua-Belmonte, J.-C. (1996). Involvement ofFGF-8 in initiation, outgrowth and patterning of the vertebrate limb.Development 122, 1737-1750.

Vogel-Höpker, A. and Rohrer, H. (2002). The specification of noradrenergic locuscoeruleus (LC) neurones depends on bone morphogenetic proteins (BMPs).Development 129, 983-991.

Vogel-Höpker, A., Momose, T., Rohrer, H., Yasuda, K., Ishihara, L. andRapaport, D. H. (2000). Multiple functions of fibroblast growth factor-8 (FGF-8)in chick eye development. Mech. Dev. 94, 25-36.

Wawersik, S., Purcell, P., Rauchman, M., Dudley, A. T., Robertson, E. J. andMaas, R. (1999). BMP7 acts in murine lens placode development. Dev. Biol.207, 176-188.

Wilkinson, D. G. (1993). Whole mount in situ hybridisation of vertebrateembryos. In In Situ Hybridisation (ed. D. G. Wilkinson), pp. 75-83. Oxford:Oxford University Press.

Wilson, P. A., Lagne, G., Suzuki, A. and Hemmati-Brivanlou, A. (1997).Concentration-dependent patterning of the Xenopus ectoderm by BMP4 and itssignal transducer Smad1. Development 124, 3177-3184.

Yoshikawa, S., Aota, S., Shirayoshi, Y. and Okazaki, K. (2000). The ActR-Iactivin receptor protein is expressed in notochord, lens placode and pituitaryprimordium cells in the mouse embryo. Mech. Dev. 91, 439-444.

Zhao, G.-Q. (2002). Consequences of knocking out BMP signaling in the mouse.Genesis 35, 43-56.

3493RESEARCH ARTICLEBMPs in RPE development