fate maps of neural crest and mesoderm in the mammalian...

Fate Maps of Neural Crest and Mesoderm in theMammalian Eye

Philip J. Gage,1,2 William Rhoades,1 Sandra K. Prucka,3 and Tord Hjalt4

PURPOSE. Structures derived from periocular mesenchyme ariseby complex interactions between neural crest and mesoderm.Defects in development or function of structures derived fromperiocular mesenchyme result in debilitating vision loss, in-cluding glaucoma. The determination of long-term fates forneural crest and mesoderm in mammals has been inhibited bythe lack of suitable marking systems. In the present study, thefirst long-term fate maps are presented for neural crest andmesoderm in a mammalian eye.

METHODS. Complementary binary genetic approaches wereused to mark indelibly the neural crest and mesoderm in thedeveloping eye. Component one is a transgene expressing Crerecombinase under the control of an appropriate tissue-spe-cific promoter. The second component is the conditional Crereporter R26R, which is activated by the Cre recombinaseexpressed from the transgene. Lineage-marked cells werecounterstained for expression of key transcription factors.

RESULTS. The results established that fates of neural crest andmesoderm in mice were similar to but not identical with thosein birds. They also showed that five early transcription factorgenes are expressed in unique patterns in fate-marked neuralcrest and mesoderm during early ocular development.

CONCLUSIONS. The data provide essential new information to-ward understanding the complex interactions required for nor-mal development and function of the mammalian eye. Theresults also underscore the importance of confirming neuralcrest and mesoderm fates in a model mammalian system. Thecomplementary systems used in this study should be useful forstudying the respective cell fates in other organ systems. (In-vest Ophthalmol Vis Sci. 2005;46:4200–4208) DOI:10.1167/iovs.05-0691

The vertebrate eye is constructed during development fromthree general embryonic precursor sources: neural ecto-

derm, surface ectoderm, and a loose array of cells termed theperiocular mesenchyme. A primary function of the periocular

mesenchyme is to provide multiple mature cell lineages thatare necessary for normal ocular development and vision, in-cluding the corneal endothelium and stroma, trabecular mesh-work, Schlemm’s canal, sclera, ciliary body muscles, irisstroma, extraocular muscles, and ocular blood vessels. A sec-ond essential function of the periocular mesenchyme duringocular development is to provide essential signals for pattern-ing of ocular ectoderm primordia, including induction of lac-rimal glands from the surface ectoderm, specification of retinalpigmented epithelium from the optic cup, and differentiationof the optic stalk from the neural ectoderm.1–3 Genetic oracquired defects in development or function of the periocularmesenchyme result in ocular disease and vision loss, includingparticularly glaucoma.4,5

Historically, the periocular mesenchyme was thought toarise developmentally from the mesoderm. However, fatemaps developed in birds by using quail chick chimeras, vitaldye labeling, or neural crest-specific antibodies have demon-strated that periocular mesenchyme actually receives initialcontributions from both neural crest and mesoderm.6–10 Thesestudies further established that the corneal endothelium andstroma, trabecular meshwork, most of the sclera, and theciliary muscles in birds are derived solely from the neural crest.In contrast, all vascular endothelium, the caudal region of thesclera, Schlemm’s canal, and extraocular muscles are derivedfrom the mesoderm. These results have generally served wellas the model for all vertebrates, including mammals. However,the documentation of important developmental differencesbetween birds and mammals has become increasingly com-mon.11,12 In the neural crest, these include differences in thetiming of neural crest migration, the migratory pathways taken,and their ultimate fates,11 making it essential to uncover anyspecies differences that may exist in mammals to accountaccurately for the embryonic origins of each mature structure.Several key regulators of periocular mesenchyme developmenthave recently been identified.2,13–17 Determining the neuralcrest and mesoderm expression patterns, as well as the lineage-specific effects of genetic lesions in these genes, would signif-icantly enhance our understanding of the normal mechanismsby which these genes function during ocular development.The major impediment to addressing any of these questionshas been the lack of a reliable strategy for quantitative, long-term labeling of neural crest and mesoderm in a mammalianeye. Fortunately, the use of two complementary Cre-lox basedapproaches has recently provided the necessary breakthrough.

A binary transgenic system has been developed that allowsfor indelible, permanent labeling of early presumptive neuralcrest and all subsequent progeny cells.18 The first componentof this system is a Wnt1-Cre transgene, which expresses Crerecombinase under the control of the promoter for the Wnt1gene. Like Wnt1 itself, Cre expression from the transgene istransient and limited to a 24-hour period in early presumptiveneural crest cells before their emigration from the dorsal neuraltube.18 The second component of the system is the conditionalCre reporter allele, R26R.19 Cre-mediated activation of �-galac-tosidase (�-gal) expression from the R26R reporter allele byDNA recombination provides for indelible, long-term labelingof all cells derived from the neural crest.19 This system has

From the Departments of 1Ophthalmology and Visual Sciences,2Cell and Developmental Biology, and 3Human Genetics, University ofMichigan Medical School, Ann Arbor, Michigan; and the 4Departmentof Cell and Molecular Biology, Lund University, Lund, Sweden.

Supported by a University of Michigan UROP Summer BiomedicalResearch Fellowship (WR), the Glaucoma Research Foundation (PJG),National Eye Institute Grants EY014126 and EY07003 (PJG), Researchto Prevent Blindness (PJG), the Swedish Science Council (TH), andNational Institute of Child Health and Development Grant 34283 toSally Camper.

Submitted for publication June 2, 2005; revised July 1, 2005;accepted August 30, 2005.

Disclosure: P.J. Gage, None; W. Rhoades, None; S.K. Prucka,None; T. Hjalt, None

The publication costs of this article were defrayed in part by pagecharge payment. This article must therefore be marked “advertise-ment” in accordance with 18 U.S.C. §1734 solely to indicate this fact.

Corresponding author: Philip J. Gage, Department of Ophthalmol-ogy and Visual Sciences, University of Michigan Medical School, 350Kellogg Eye Center, 1000 Wall Street, Ann Arbor, MI 48105;[email protected].

Investigative Ophthalmology & Visual Science, November 2005, Vol. 46, No. 114200 Copyright © Association for Research in Vision and Ophthalmology

been used to map neural crest fates in cardiovascular, cranio-facial, and skull vault development.20–22

The mouse glycoprotein hormone �-subunit gene (�GSU) isexpressed initially in Rathke’s pouch, the oral-ectoderm–de-rived primordium of the anterior pituitary gland, and subse-quently in multiple lineages of the mature anterior pituitarygland.23 �GSU expression has also been reported in mesodermof the early ocular primordium and in the olfactory epithe-lium.24 The promoter regulatory elements required to recapit-ulate this expression pattern have been identified and used toexpress Cre recombinase in transgenic mice.25 Mice doublytransgenic for �GSU-Cre and a Cre-responsive reporter cassetteexhibit indelible expression of �-gal in a pattern consistentwith expression of endogenous �GSU itself.25 �GSU-Cre hasalso been used successfully to generate a lineage-specific geneknockout of the transcription factor gene Dax1 in the anteriorpituitary gland.26,27

In the current study, we used the complementary Wnt1-Cre/R26R and �GSU-Cre/R26R labeling systems to establishfor the first time the long-term fates of neural crest and meso-derm, respectively, in a model mammalian eye. The fates weresimilar to but not identical with those previously reported inbirds, with the most significant differences occurring in theanterior segment. We also demonstrated that five known orpotential transcriptional regulators of periocular mesenchymehave unique expression patterns in the neural crest and meso-derm during early ocular development. Finally, the data alsoimply that the mechanism of activation for the homeobox genePitx2 in the neural crest and mesoderm is likely to be distinct.These findings have important implications for ocular develop-ment and function and may provide a model for understandinginteractions between the neural crest and mesoderm in otherorgan systems.

MATERIALS AND METHODS

Animals and Isolation of Lineage-Marked Mice

Wnt1-Cre mice transmit a transgene consisting of a cDNA cassetteexpressing Cre protein placed under the transcriptional control of themurine Wnt1 promoter.18 �Gsu-Cre transgenic mice (TgN(Cga-cre)S3SAC, cat no. 004426; Jackson Laboratories, Bar Harbor, ME)transmit a construct including the cDNA for a nuclear-localized Creprotein under the transcriptional control of the pituitary glycoproteinsubunit � promoter.25 R26R mice were purchased from Jackson Lab-oratories and transmit a �-galactosidase reporter allele that is geneti-cally activated in vivo by Cre recombinase activity.19 All proceduresusing mice were approved by the University of Michigan Committeeon Use and Care of Animals. All experiments were conducted inaccordance with the principles and procedures outlined in the NIHGuidelines for the Care and Use of Experimental Animals and in theARVO Statement for the Use of Animals in Ophthalmic and VisionResearch.

Timed pregnancies were produced by mating homozygous R26Rfemales either with Wnt1-Cre transgenic males to produce progenywith marked neural crest or with �Gsu-Cre transgenic males to pro-duce progeny with marked mesoderm. The morning a plug was de-tected was designated as embryonic day (E)0.5. Embryos were col-lected by cesarian section after the mother was euthanatized. Allgenetic loci were genotyped by using PCR-based methods with prim-ers for Cre25 and R26R.19

Detection of �-Galactosidase Activity

Enucleated whole eyes from 6-week-old mice were fixed for 1 to 2hours at 4°C with 4% paraformaldehyde in phosphate-buffered saline(PBS) and rinsed three times in wash buffer (0.1 M sodium phosphate[pH 8.0], 2 mM MgCl2, 2% NP-40). Fixed eyes were stained overnightat 37°C in standard staining solution (5 mM potassium ferricyanide, 5

mM potassium ferrocyanide, 0.1% X-gal in wash buffer), rinsed in washbuffer, embedded in JB-4 resin, and sectioned at 3 �m. Mountedsections were counterstained with nuclear fast red (Sigma-Aldrich, St.Louis, MO).

Immunohistochemistry

Embryos harvested for immunostaining were fixed for 2 to 4 hourswith 4% paraformaldehyde in PBS, washed and dehydrated, and em-bedded in paraffin. Mounted sections were deparaffinized and treatedfor antigen retrieval by boiling for 10 minutes in citrate buffer (pH 6.0).Immunostaining was performed according to standard methods.Briefly, sections were incubated with antibodies directed against �-ga-lactosidase (Eppendorf-5Prime, Boulder, CO), FOXC1 or -2 (Abcam,Inc., Cambridge, MA), myogenin (MYOG; clone F5D, developed byWoodring Wright and obtained from NICHD/Developmental Hybrid-oma Studies Bank, Iowa City, IA), PITX1 (a gift from Jacques Drouin,Montreal, Canada28) or PITX229 followed by biotinylated species-spe-cific secondary antibodies (Jackson ImmunoResearch, West Grove,PA). Signals were detected using tyramide signal-amplification kits(PerkinElmer, Boston, MA).

RESULTS

Contributions of Cranial Neural Crest andAnterior Paraxial Mesoderm to the EarlyOcular Primordia

We crossed Wnt1-Cre males with R26R females to generateWnt1-Cre;R26R progeny (neural-crest–marked) with indeliblylabeled neural crest by virtue of Cre activation of �-gal reporterexpression. In parallel experiments, we crossed �Gsu-Cre andR26R mice to generate analogous �Gsu-Cre;R26R mice (me-soderm-marked) with �-gal-labeled ocular mesoderm. In eithersystem, the presence and location of lineage-marked �-gal�

cells was detected by incubation with X-gal, a chromogenicsubstrate of �-gal, or by immunohistochemistry with an anti-�-gal antibody.

X-gal staining of whole embryos from each class at E10.5demonstrated the distinct labeling specificity of each binarylabeling system. The heads of neural-crest–marked embryoswere heavily invested with �-gal� cells by this time point,particularly in the branchial arches and trigeminal ganglia (Fig.1A). There was also heavy labeling of the midbrain neuralectoderm, as has been reported.18 In the trunk, there wassubstantial staining of the cardiac outflow tract and the dorsalroot ganglia along either side of the neural tube in neural-crest–marked embryos (Fig. 1A). In contrast, head staining in �Gsu-Cre;R26R embryos at E10.5 was limited to a wedge of mesen-chyme located ventrally and caudally to the optic cup and theoral ectoderm. These positions are analogous to those reportedin �Gsu-lacZ transgenic mice and in endogenous �Gsu expres-sion (Fig. 1B).24 �Gsu-Cre;R26R embryos also showed heavystaining of the somites, atria and ventricles of the heart, and gutmesentery (Fig. 1B). In histologic sections, classic neural-crest–derived structures such as the trigeminal ganglia, meninges,and presumptive calvarial vault were heavily stained in neural-crest–marked embryos (Figs. 1C, 1E) but were unstained inmesoderm-marked embryos (Figs. 1D, 1F). The dorsal neuraltube, which was labeled in neural-crest–marked embryos (Fig.1E), was unstained in mesoderm-marked embryos (Fig. 1F).Collectively, these data are consistent with the Wnt1-Cre;R26Rand �Gsu-Cre;R26R marking systems labeling mesenchymederived from neural crest and mesoderm, respectively.

We next examined each binary system for �-gal labeling inthe developing eye at the cellular level. Labeled neural crestcells in Wnt1-Cre;R26R embryos surrounded the optic cup andstalk at E10.5 (Fig. 1G). The neural crest also occupied the

IOVS, November 2005, Vol. 46, No. 11 Ocular Mesenchyme Fates in Mammals 4201

presumptive cornea, beginning immediately after separation ofthe lens vesicle from the surface ectoderm, and a few cellsappeared within the hyaloid space posterior to the lens. Mes-enchymal cells located in a morphologically distinct wedge ofcells located ventrally and caudally to the optic cup and stalkdid not label as neural crest in Wnt1-Cre;R26R embryos, sug-gesting that these cells were of mesoderm origin (Fig. 1G).Indeed, the position of these cells corresponds well with thatof the previously reported rostral limit of cranial paraxial me-soderm.30,31 Cells in this morphologically distinct wedge areprominently labeled in �GsuCre;R26R embryos (Fig. 1H). The

unique morphology of these cells relative to surrounding mes-enchyme, the previous mapping of cranial paraxial mesodermto this location at this time point, and the absence of labelingin Wnt1-Cre;R26R embryos all argue that the labeled cells inthe ocular primordia of �GsuCre;R26R embryos represent cra-nial paraxial mesoderm. Thus, the two lineage-specific �-gallabeling systems resulted in unique staining patterns in theearly ocular primordia that were consistent with the previouslyreported arrangement of neural crest and mesoderm precur-sors in the chick.6–10 Occasional labeled mesoderm cells werealso present within the presumptive cornea, in the hyaloidspace, and adjacent to the optic cup and stalk in �GsuCre;R26R embryos at this early time point (Fig. 1H and data notshown), an observation that is consistent with previous reportsthat mesoderm cells have already begun to intermingle withneural crest by this time point.30,31 By E12.5, neural crest andmesoderm were extensively comingled in multiple ocularstructures (data not shown and see Figs. 3, 4, and 5).

Transcription Factor Expression in MarkedNeural Crest and MesodermMolecular markers that are specifically expressed in neuralcrest or mesoderm precursors have not been described for themammalian ocular primordia. Therefore, we examined theprotein expression patterns of several transcription factorspreviously implicated in ocular development to determinetheir expression profiles between the two embryonic precur-sor pools. The homeodomain transcription factor PITX2 isassociated with ocular defects in humans and is essential forocular development in mice.2,13,32,33 PITX2 expression labeledthe neural crest, beginning initially in the presumptive anteriorsegment and quickly extending to the periphery of the opticcup in Wnt1Cre;R26R embryos by E11.5 (Fig. 2A). In contrast,PITX2 expression had marked all �-gal-labeled mesoderm in�GsuCre;R26R embryos at E11.5, before these cells enteredthe ocular field (Fig. 2B). By E12.5, PITX2 expression hadspread to all ocular neural crest (data not shown). The fork-head transcription factor FOXC1 is also essential for POMdevelopment in mice and humans.16,34 FOXC1 expressionmarked the neural-crest–derived cells in the anterior segmentand surrounding the optic cup and stalk at E11.5 (Fig. 2C).Colabeling was also identifiable in a small percentage of meso-derm cells at E11.5 (Fig. 2D). Therefore, PITX2 and FOXC1were each expressed in both the neural crest and mesoderm,albeit to very different degrees in the mesoderm. The transcrip-tion factor FOXC2 is necessary in periocular mesenchyme fornormal anterior segment development in mice, but not inhumans.17,35 In contrast to FOXC1, expression of FOXC2 wasdetectable only in neural-crest–labeled cells and was notpresent in mesoderm-labeled cells of �Gsu-Cre;R26R embryosat E11.5 (Figs. 2E, 2F). The homeodomain protein PITX1 isrelated to PITX2, and the two genes are often coexpressed andare functionally redundant in other tissues.36,37 However, mes-enchymal PITX1 expression in the developing eye was limitedto a subset of labeled mesoderm cells and was never observedin neural-crest–labeled cells at E11.5 (Figs. 2G, 2H). Consistentwith a previous report, PITX1 expression was also detected inthe early lens vesicle (Fig. 2H).38 Finally, we examined MYOGas a second marker for mesoderm-derived muscle precursors.MYOG expression was limited to a subset of mesoderm-labeledcells in �GsuCre;R26R embryos and was not expressed inneural-crest–labeled cells at E11.5 (Figs. 2I, 2J). Based on thesecollective results, particularly the complete absence of FOXC2colabeling in mesoderm-marked embryos and the specificity ofPITX1 and MYOG for subsets of mesoderm cells, we concludethat, in the developing eye, the Wnt1Cre;R26R and �GsuCre;R26R systems specifically mark the neural crest and meso-derm, respectively.

FIGURE 1. Lineage-specific �-galactosidase expression in neural crestand mesoderm of Wnt1-Cre;R26R and �Gsu-Cre;R26R embryos. X-galstaining of E10.5 neural crest (A) and mesoderm (B) labeled embryosreveals distinct �-gal expression patterns generated by two binarylabeling systems. Strong �-gal expression in the trigeminal ganglion ofneural crest (C) but not mesoderm (D) labeled embryos stained withX-gal. The mesoderm-derived endothelial lining of the primary headvein is labeled in mesoderm embryos (D, inset). Strong �-gal expres-sion in the dorsal root ganglion of neural crest (E) but not mesoderm(F) marked embryos stained with X-gal. The dorsal neural tube (��) islabeled in neural crest (E) but not mesoderm (F) marked embryos.Complementary, nonoverlapping �-gal expression patterns detectedby immunohistochemistry in E10.5 ocular primordia of neural crest (G)and mesoderm (H) labeled embryos. X-gal-stained sections are coun-terstained with nuclear fast red to display all nuclei. TG, trigeminalganglion; BA, branchial arches; DRG, dorsal root ganglia; C, cornea;CM, cephalic mesoderm; G, gut; H, heart; S, somites.

4202 Gage et al. IOVS, November 2005, Vol. 46, No. 11

Fates in the Extraocular Muscles

The early primordia of extraocular muscles were apparent asmesenchymal condensations lateral to the optic cup andstalk by E12.5 (Figs. 3A, 3B). A small number of neural crestand a significant number of mesoderm precursor cells arealready present within the presumptive muscle condensa-tions at this early time point (Figs. 3A, 3B). Subsequently,

the fate of each embryonic precursor pool was determinedbased on cellular morphology. Connective fascia cells inperinatal and adult muscles were derived exclusively fromthe neural crest (Figs. 3C, 3E). In contrast, myotubules in thedeveloping muscle (Fig. 3D) and myofibers in the maturemuscle (Fig. 3F) are derived exclusively from mesoderm. Noectopic labeling by either labeling system, such as neuralcrest labeling of muscle cells or mesoderm labeling of con-nective fascia cells, was ever observed in multiple speci-mens examined at any time point.

Fates in the Ocular Vasculature

The hyaloid vasculature forms posterior to the lens as a tran-sient, embryonic blood system that maintains the posteriorlens and developing retina. Mesenchymal precursor cells werepresent within the hyaloid space by E11.5 and primitive bloodvessels were present by E12.5 (Figs. 4A, 4B). Both neural crestand mesoderm-derived cells contributed to the early hyaloidvessels (Figs. 4A, 4B). Subsequently, endothelial cells lining themature hyaloid blood vessels were derived solely from themesoderm (Figs. 4D, 4F). Pericytes attaching themselves to theunstained endothelial tube were derived exclusively from theneural crest (Figs. 4C, 4E). Endothelial and smooth muscle cellsof the hyaloid artery were derived from the mesoderm andneural crest, respectively (data not shown).

FIGURE 3. Detection of �-gal by immunohistochemistry in developingand mature extraocular muscles. �-gal expression (green) in presump-tive extraocular muscles (PM) of E12.5 neural crest (A) and mesoderm(B) marked embryos. �-gal expression labels connective fascia cells(arrowheads) in developing (C) and mature (E) extraocular muscles ofneural-crest–marked mice. �-gal expression labels myotubules andmyofibers (arrows) in developing (D) and mature (F) extraocularmuscles of mesoderm marked mice. Adult micrographs (E, F) arepresented as the merge of red and green channels to enhance visual-ization of the myofibers.

FIGURE 2. Coexpression of transcription factors and �-gal in fate-marked embryos. E11.5 Wnt1Cre;R26R and �Gsu-Cre;R26R embryoswere immunostained for expression of �-gal (green) to display neuralcrest (A, C, E, G, I) and cephalic mesoderm (B, D, F, H, J), respec-tively. Sections were counterstained (red) to detect the expression ofPITX2 (A, B), FOXC1 (C, D), FOXC2 (E, F), PITX1 (G, H), and MYOG(I, J). Labeling in (A) provides orientation for all panels. NC, neuralcrest; CM, cephalic mesoderm; C, cornea; L, lens; OC, optic cup; OS,optic stalk.


The choriocapillary bed comprises a network of microves-sels that immediately overlie the retinal pigmented epithelialcells and provide vascular functions to the RPE and underlyingphotoreceptors of the retina. Fates of the embryonic precursorpools in the choroid were identical with those in the hyaloid.Pericytes and other associated cells were derived from theneural crest (Fig. 4G), whereas endothelial cells were derivedfrom the mesoderm (Figs. 4H). Endothelial cells of blood ves-sels within the ciliary body were similarly derived from themesoderm (Fig. 5H).

Fates in Anterior Segment Structures

Multiple cell lineages within the anterior segment of the eyewere derived from mesenchyme, including, in approximate

order of appearance, the corneal endothelium and stroma,ciliary muscles, ciliary blood vessels, anterior iris, and trabec-ular meshwork and Schlemm’s canal within the iridocornealangle. Neural crest cells migrated into the presumptive anteriorsegment between the anterior face of the newly formed lensvesicle and the surface ectoderm by E10.5 (Fig. 1G). Mesodermcells followed 12 to 24 hours later (data not shown). By E14.5,neural-crest–derived cells comprised most of the cells withinthe endothelium and stroma of the cornea, and the presump-tive iridocorneal angle (Fig. 5A). Although significantly fewerin number, mesoderm-derived cells were also consistentlypresent in the same structures (Fig. 5B). The presence andrelative ratios of the two embryonic cell lineages remain un-changed as these structures proceeded through differentiation(Figs. 5C, 5D).

To detect �-gal� cells while preserving the integrity ofmature anterior segment structures, we stained enucleatedadult eyes in X-gal and then processed them for plastic-embed-ded sections. The relative contributions of neural crest andmesoderm after this processing was unchanged from that ob-served with fluorescent immunostaining on samples derivedfrom earlier time points, indicating efficient penetration of thestain. Most cells in the endothelium and stroma layers of themature cornea expressed �-gal in neural-crest–marked animals,establishing a neural crest origin (Fig. 5E). However, cells thatdid not express �-gal were also consistently present in bothlayers of neural-crest–marked embryos (Fig. 5E), indicating thatnot all cells in these layers derived from neural crest in mice. Acomparable number of �-gal� cells were consistently presentin the mature corneal endothelium and stroma of mesoderm-marked mice (Fig. 5F). The number of �-gal� cells in meso-derm-marked mice parallels the number of unlabeled cells inneural-crest–marked mice. Collectively, these data indicate thatmost of the corneal endothelial and stromal cells are derivedfrom neural crest but additional cells in both layers are derivedfrom mesoderm. This finding was unexpected, because onlycells of neural crest origin have been reported to be present inthe corneal endothelium and stroma of birds.6–10

Contributions of the neural crest and mesoderm to thelimbal region and iridocorneal angle of the mature eye werecomplex and highly variable. Ciliary muscles were derivedfrom the neural crest but not the mesoderm (Figs. 5I–L). Theendothelial lining of Schlemm’s canal was derived from themesoderm (Figs. 5K, 5L). Most cells within the trabecularmeshwork labeled as neural crest, but a measurable numberdid not (Figs. 5I, 5J). Similar to observations in the cornea,labeling of some trabecular meshwork cells in mesoderm-la-beled mice indicates that these cells derived from the meso-derm (Figs. 5K, 5L). We observed little to no labeling for neuralcrest in the iris stroma (Figs. 5M, 5O). In contrast, iris stromaregularly labels as mesoderm (Figs. 5N, 5P).

DISCUSSION

We used binary transgenic systems based on Cre/loxP technol-ogy to label specifically the neural crest or mesoderm in eyes ofmouse as a model vertebrate. Early indelible labeling of therespective embryonic precursor cell lineages by the two com-plementary systems allowed their fates to be followed through-out development and in mature mouse eyes. The data demon-strate that neural crest and mesoderm fates in mammalian eyesare similar to but not identical with those previously reportedin birds.6–10 We also report the neural crest and mesodermexpression patterns of several key transcription factors duringearly ocular development. Collectively, these results have im-portant mechanistic implications for understanding the devel-opment and function of structures derived from periocular

FIGURE 4. Neural crest and mesoderm fates in ocular blood vessels.Immunohistochemical detection of �-gal expression (green) in eyesfrom neural crest–and mesoderm-marked animals. Neural crest (A) andmesoderm (B) precursor cells are present within the hyaloid space byE12.5. Pericytes (C, E, arrowheads) in mature hyaloid vessels arederived from neural crest, whereas the endothelial lining (D, F, ar-rows) of mature hyaloid is derived from mesoderm. Neural crest (G,arrowheads) and mesoderm (H, arrows) have analogous fates in themature choroid vasculature. All micrographs presented as the merge ofred and green channels to facilitate visualization of the blood vesselsand the location of red blood cells, which are autofluorescent (red/orange). Hy, hyaloid blood vessels; L, lens; R, retina; PE, pigmentedepithelium; S, sclera; RBC, red blood cells.


mesenchyme, and the roles mediated by specific transcriptionfactors in these processes.

Characterization of the Cell-Labeling Systems

The ability to accurately track cell fates throughout develop-ment and into adulthood depends on the fidelity and perma-nence of the labeling system used.39 Neural crest and meso-derm fates have been extensively characterized in birds byusing quail chick chimeras, vital dye labeling, and antibodystaining for neural-crest–specific markers.6–10,40,41 However,these labeling systems are not effective for determining partic-ularly the long-term fates of these lineages in mammals.39

Activation of the R26R reporter construct by Cre-mediatedDNA recombination in the two systems used in the studyensured indelible �-gal labeling of the precursor cells in whichCre was originally expressed, as well as in all subsequentprogeny cells. The Wnt1-Cre;R26R labeling system was initiallyshown to generate specific, quantitative labeling of the mid-brain neural ectoderm and all premigratory neural crest.18

Subsequent experiments examining cardiovascular, craniofa-

cial, and skull vault development further established that this isa robust approach for quantitative, specific, and long-term fatemapping of neural crest in mice.20–22 Our results confirm andextend these conclusions.

Our data also identify the �Gsu-Cre;R26R combination as asimilarly robust and specific system for labeling mesenchymepopulations derived from mesoderm. The cells marked in�Gsu-Cre;R26R embryos are distinct from those in comparableneural crest-labeled Wnt1-Cre;R26R embryos during early oc-ular development, when the two embryonic precursor popu-lations have not yet mixed. The location of cells labeled in�Gsu-Cre;R26R embryos at E11.5 corresponds precisely withthe previously established position of cephalic paraxial meso-derm.30,31 The expression patterns of the transcription factorsFOXC2, PITX1, and MYOG also distinguish between the �-gal-marked populations in Wnt1-Cre;R26R versus �Gsu-Cre;R26Rembryos. It is unlikely that there was a previously unrecog-nized population of neural crest cells that was not marked by�-gal activation in Wnt1-Cre;R26R embryos. Therefore, theseresults are consistent with the �-gal-marked cells in �Gsu-Cre;

FIGURE 5. Complex neural crest and mesoderm fates in the ocular anterior segment. Immunohistochemistry (green) and X-gal (blue) were usedto detect �-gal expression in neural crest (NC) and mesoderm (M). Neural crest–and mesoderm-derived cells were both present in the developinganterior segment of E14.5 (A, B) and P1 (C, D) eyes. Neural crest and mesoderm contributions to the mature corneal endothelium and stroma (E,F), ciliary body (G, H), iridocorneal angle (I–L), and iris (M–P) of adult eyes. Note �-gal� cells (∧ ) in corneal stroma and endothelium of neuralcrest-marked animals (E). Limbus (�), Schlemm’s canal (open arrows), trabecular meshwork (closed arrows), iris stroma (arrowheads). X-galstained eyes are counterstained with nuclear fast red to display all nuclei. S, corneal stroma; E, corneal endothelium; ICA, iridocorneal angle.


R26R being of mesoderm origin. Future identification of addi-tional molecular markers is likely to refine our understandingof these cells further.

The absence of staining in well-established neural-crest–derived tissues, such as the trigeminal ganglia in the head anddorsal root ganglia in the trunk, provides further evidence thatthe �Gsu-Cre;R26R system marks mesoderm and not neuralcrest elsewhere as well, suggesting that this combination mayhave further utility in other mesoderm-derived tissues else-where in the body. However, further analysis of lineage-spe-cific markers within individual organ systems would be neces-sary to confirm this preliminary interpretation. Although thesource of rare �-gal� cells in lens and surface ectoderm is notclear, PITX1 is capable of transactivating the same �GSU pro-moter in cell culture and PITX1 expression has been reportedin the lens placode and early lens vesicle (Fig. 2H and Ref. 38).Therefore, it may be that expression of PITX1 ectopicallyactivates the �Gsu-Cre transgene at a low frequency in lens orthe surface ectoderm from which it derives.

Ocular Fate of Neural Crest and Mesoderm inBirds Versus Mammals

Our results indicate that there is general conservation of neuralcrest and mesoderm fates in the eye between birds and mam-mals and that there is frequently a division of labor withinindividual structures where the neural crest and mesodermcontribute different mature lineages (Fig. 6).38 For example, inextraocular muscles the muscle fibers themselves are derivedfrom the mesoderm, whereas connective fascia cells arise fromthe neural crest. Similarly, the endothelial lining of ocularblood vessels arises from the mesoderm, whereas associatedsmooth muscle and pericytes derive from the neural crest.Differentiation of ciliary muscles from the neural crest is alsoconserved between birds and mammals.

Despite the general conservation of neural crest and meso-derm fates in the eye between birds and mammals, severalimportant differences are apparent (Fig. 6). Of note, these alloccur within anterior segment structures. The endothelial andstromal layers of the cornea, as well as cells of the trabecularmeshwork are reported to derive solely from the neural crest inbirds.6,7,9,10 Endothelial cells lining Schlemm’s canal are thesole anterior segment lineage reported to derive from themesoderm in birds.40 Our results indicate that most cellswithin the corneal endothelial and stroma layers and the tra-becular meshwork are also derived from the neural crest in

mammals. However, a small population of cells within each ofthese structures consistently did not express �-gal in neural-crest–labeled animals, strongly suggesting that certain cells inthese structures have a different origin. These cells werepresent by E12.5 and persisted into adulthood. In contrast, acomparable number of cells in each of these structures consis-tently labeled positively for �-gal in mesoderm-marked animalsthroughout the same time period, consistent with a mesodermorigin of these cells. Although the possibility of ectopic expres-sion is always a concern when using transgenic mice, theabsence of ectopic labeling in any other neural-crest–derivedstructures or the dorsal neural tube, from which the neuralcrest arises, strongly argues that these cells were not neuralcrest. This, together with the demonstration that not all cells inthe affected structures labeled as neural crest, leads us toconclude that a minority population of cells within the cornealendothelium and stroma, the limbus, and the trabecular mesh-work are derived from the mesoderm. The identification ofadditional lineage-specific markers that are expressed at latertime points is essential for further confirmation of this inter-pretation and to determine the mature identity of these cells.

The underlying reason(s) for the surprising presence ofmesoderm cells in multiple anterior segment structures inmice, where they had not been reported in birds, remainsunknown. One possibility is that these cells are also present inthe chick but are not recognized because they represent arelatively small minority of cells in the relevant structures. Thisseems unlikely, given that multiple laboratories have reportedequivalent findings using different labeling approaches.6–10

Therefore, we propose a model wherein the mature cornealendothelial and stromal layers, the mesenchymal layer of thelimbus, and the trabecular meshwork in mammals each consistof two mature cell lineages, one of which is derived from theneural crest and the other from the mesoderm. The lineagesderived from the mesoderm may not be present in homologousstructures in birds. This addition of new cells would representan evolutionary advance from birds to mammals. Attractivecandidates for the identity of at least some of the mesoderm-derived cells are a complex mix of antigen-presenting cells thatserve as immune sentinels and are present in mammalian eyes,including murine eyes. In the limbus and peripheral cornea,these include dendritic cells and macrophages.42–44 A similarpopulation of cell lineages has recently been demonstrated inthe central cornea of mice as well, but the cells in this locationare generally immature, lacking the typical dendritic cell mor-

FIGURE 6. Neural crest and meso-derm fates in bird and mammalianeyes. Cross-sectional diagrams summa-rizing similarities and differences incontributions of neural crest (red) andmesoderm (blue) to adult chicken andmouse eyes. The major differences oc-cur in the anterior segment. Chickenfates are synthesized from previouslypublished reports.6–10


phology and do not express major histocompatibility complex(MHC) class II antigen in normal, uninflamed eyes.45 Thesecells probably arise from bone marrow, which is consistentwith a mesoderm origin. It is intriguing that there have been nopublished reports of dendritic or macrophage cells’ havingbeen identified in chick eyes, and highly related Langerhanscells are reportedly absent from chick eyes.46 Taken together,these observations suggest that cells within the dendritic andmacrophage lineages may account for at least some of themesoderm-derived cells that we identified in the anterior seg-ment of murine eyes. Recently, lymphohematopoietic lineageshave been shown to modulate pathogenicity in a murine modelof pigmentary glaucoma.47 Although the significance of thesefindings remains to be established for human disease, immunesurveillance cells in the form of dendritic and Langerhans cellsare also found in the anterior segment of human eyes. There-fore, the presence of mesoderm-derived dendritic cells maycontribute to the underlying etiology of glaucoma.

Implications of Transcription FactorExpression Patterns

The ability to identify lineage-specific expression patterns ofkey regulatory and functional genes in the developing eye is animportant advance, because it provides useful insights intopotential mechanisms of gene function and interpretation ofmutant phenotypes. Demonstration that the five transcriptionfactors we examined were expressed in overlapping but dis-tinct patterns in the neural crest and mesoderm at E11.5 issignificant because it suggests that initial molecular patterningof the mesenchymal precursor cells is probably already in placeat this early time point, even though morphologic differenceswere not yet discernible. It is particularly striking that FOXC1and PITX2 had the most similar expression patterns, becauseheterozygous mutations in the human FOXC1 or PITX2 genesboth result in Axenfeld-Rieger syndrome, an autosomal domi-nant condition including anterior segment defects and a highrisk for glaucoma.33,34 This supports the hypothesis that thetwo factors probably regulate common downstream targets incells in which they are coexpressed. Our current demonstra-tion of wider Pitx2 expression in the developing eye is con-sistent with the demonstration that complete loss of Pitx2function in mice results in a more severe ocular phenotypethan loss of Foxc1. Foxc2 heterozygous mice exhibit ocularphenotypes very similar to the Foxc1 heterozygous pheno-type.17 This is consistent with the two genes’ being coex-pressed in neural crest.

Pitx2 is essential to specify correctly the multiple lineagesderived from the periocular mesenchyme, including the cor-neal endothelium and stroma and the extraocular mus-cles.2,13,15,32,48 Our current demonstration that these struc-tures are chimeric, receiving contributions from both theneural crest and mesoderm, raises the intriguing mechanisticquestion of whether the requirement for Pitx2 function ineach affected structure lies in the neural crest or the meso-derm. For example, specification of extraocular muscles mayrequire Pitx2 for an intrinsic function in mesoderm precursorsfor them to initiate their myogenic program. Alternatively,Pitx2 function may be necessary in the neural crest for expres-sion of an upstream signaling molecule(s) that acts extrinsicallyon the mesoderm precursor cells. Neural crest- and mesoderm-specific knockouts will ultimately be necessary to address thisquestion in each relevant tissue.

The distinct timing and location of initial Pitx2 expressionin neural crest versus mesoderm prompts intriguing new in-sights into potential gene function in the developing corneaand provides strong evidence that the mechanism of geneactivation is likely to be different between the two embryonic

precursor cell populations. Anterior epithelial cells of the lensvesicle act as a signaling center that is needed in multiple stepsduring development of the anterior segment, including induc-tion of the corneal endothelium and stroma layers.49 Pitx2expression is activated within neural crest cells immediatelyafter their arrival within the nascent anterior segment. To-gether with the lack of corneal endothelium and stroma spec-ification in Pitx2-deficient mice, this suggests a model in whichPitx2 itself is an essential early target of the cornea-inducingsignal(s) expressed by the anterior lens epithelia. Activation ofPitx2 in wild-type mice, presumably together with additionalessential genes, results in the initiation of a genetic cascadeleading to specification and differentiation of corneal endothe-lium and stroma. Pitx2-deficient animals are genetically unableto express PITX2 protein, effectively resulting in uncoupling ofthe required genetic cascade and agenesis of the cornea. Fu-ture extension of this pathway by identification of the up-stream inducing signal(s), additional genes coinduced withPitx2 in the neural crest, and the downstream effector geneswill provide essential insights into our understanding of cor-neal development. In contrast to the neural crest, activation ofPitx2 in the mesoderm occurs before interaction of these cellswith the ocular primordia and therefore, presumably, by adistinct mechanism.

We have illustrated the use of complementary systems forlong-term monitoring of neural crest and mesoderm cell fatesin the mammalian eye. The subtle but important differences incell fates between mature chick and mouse eyes underscorethe importance of determining whether such variations existelsewhere in the body where the neural crest and mesoderminteract during development. Differential expression of tran-scription factors within the two embryonic lineages in wild-type mice provides important new insights into the potentialmechanisms of gene function in normal and abnormal oculardevelopment. Determining the lineage-specific expression pro-files of additional important genes that are expressed in peri-ocular mesenchyme should provide similarly useful insights.Mutant mice are now available for the genes encoding each ofthese transcription factors14,16,17 and introduction of the twolabeling systems onto each mutant genetic background willallow identification of potential lineage-specific changes in thebehavior of embryonic cell types and/or their derivatives. Fi-nally, it will be possible, by using the Wnt1-Cre and �Gsu-Cretransgenes, to generate neural-crest–and mesoderm-specificknockouts of relevant genes in the mammalian eye. For geneshaving complex expression patterns and mutant phenotypes,this will allow parsing of specific features of the phenotype toa requirement for gene function either in neural crest or me-soderm, and to distinguish intrinsic versus extrinsic effects.

Acknowledgments

The authors thank Andy McMahon and Sally Camper for the gifts ofWnt1-Cre and aGsu-Cre transgenic mice, respectively; Mitchell Gillettfor expert technical assistance; David Reed for graphic illustration;Amanda Evans, Sally Camper, Peter Hitchcock, and Anand Swaroop forcritical readings of the manuscript; and Sally Camper for supportduring the early stages of the work.

References

1. Fuhrmann S, Levine EM, Reh TA. Extraocular mesenchyme pat-terns the optic vesicle during early eye development in the em-bryonic chick. Development. 2000;127:4599–4609.

2. Gage, PJ, Suh H, Camper SA. Dosage requirement of Pitx2 fordevelopment of multiple organs. Development. 1999;126:4643–4651.

3. Govindarajan V, Ito M, Makarenkova HP, Lang RA, Overbeek PA.Endogenous and ectopic gland induction by FGF-10. Dev Biol.2000;225:188–200.


4. Alward WL. Axenfeld-Rieger syndrome in the age of moleculargenetics. Am J Ophthalmol. 2000;130:107–115.

5. Lines MA, Kozlowski K, Walter MA. Molecular genetics of Axen-feld-Rieger malformations. Hum Mol Genet. 2002;11:1177–1184.

6. Johnston MC, Noden DM, Hazelton RD, Coulombre JL, CoulombreAJ. Origins of avian ocular and periocular tissues. Exp Eye Res.1979;29:27–43.

7. Noden DM. Periocular mesenchyme: neural crest and mesodermalinteractions. In: Jacobiec FA, ed. Ocular Anatomy, Embryology,and Teratology. Hagerstown, MD: Harper & Row; 1982:97–119.

8. Le Douarin NM. The Neural Crest. Cambridge, UK: CambridgeUniversity Press; 1982.

9. Le Douarin NM. Migration and differentiation of neural crest cells.In: Moscona A, Monroy A, eds. Current Topics in Development.New York: Academic Press; 1980.

10. Le Lievre CS, Le Douarin NM. Mesenchymal derivatives of theneural crest: analysis of chimaeric quail and chick embryos. JEmbryol Exp Morphol. 1975;34:125–154.

11. Kulesa P, Ellies DL, Trainor PA. Comparative analysis of neuralcrest cell death, migration, and function during vertebrate embry-ogenesis. Dev Dyn. 2004;229:14–29.

12. Barald K, Kelly M. From placode to polarization: new tunes ininner ear development. Development. In press.

13. Kitamura K, Miura H, Miyagawa-Tomita S, et al. Mouse Pitx2deficiency leads to anomalies of the ventral body wall, heart, extra-and periocular mesoderm and right pulmonary isomerism. Devel-opment. 1999;126:5749–5758.

14. Pressman CL, Chen H, Johnson RL. LMX1B, a LIM homeodomainclass transcription factor, is necessary for normal development ofmultiple tissues in the anterior segment of the murine eye. Gene-sis. 2000;26:15–25.

15. Lin CR, Kioussi C, O’Connell S, et al. Pitx2 regulates lung asym-metry, cardiac positioning and pituitary and tooth morphogenesis.Nature. 1999;401:279–282.

16. Kidson SH, Kume T, Deng K, Winfrey V, Hogan BL. The forkhead/winged-helix gene, Mf1, is necessary for the normal developmentof the cornea and formation of the anterior chamber in the mouseeye. Dev Biol. 1999;211:306–322.

17. Smith RS, Zabaleta A, Kume T, et al. Haploinsufficiency of thetranscription factors FOXC1 and FOXC2 results in aberrant oculardevelopment. Hum Mol Genet. 2000;9:1021–1032.

18. Danielian PS, Muccino D, Rowitch DH, Michael SK, McMahon AP.Modification of gene activity in mouse embryos in utero by atamoxifen-inducible form of Cre recombinase. Curr Biol. 1998.8:1323–1326.

19. Soriano, P. Generalized lacZ expression with the ROSA26 Crereporter strain. Nat Genet. 1999;21:70–71.

20. Chai Y, Jiang X, Ito Y, et al. Fate of the mammalian cranial neuralcrest during tooth and mandibular morphogenesis. Development.2000;127:1671–1679.

21. Jiang X, Rowitch DH, Soriano P, McMahon AP, Sucov HM. Fate ofthe mammalian cardiac neural crest. Development. 2000;127:1607–1616.

22. Jiang X, Iseki S, Maxson RE, Sucov HM, Morriss-Kay GM. Tissueorigins and interactions in the mammalian skull vault. Dev Biol2002;241:106–116.

23. Japon, MA, M Rubinstein, MJ Low, In situ hybridization analysis ofanterior pituitary hormone gene expression during fetal mousedevelopment. J Histochem Cytochem. 1994;42:1117–1125.

24. Kendall SK, Gordon DF, Birkmeier TS, et al. Enhancer-mediatedhigh level expression of mouse pituitary glycoprotein hormonealpha-subunit transgene in thyrotropes, gonadotropes, and devel-oping pituitary gland. Mol Endocrinol. 1994;8:1420–1433.

25. Cushman LJ, Burrows HL, Seasholtz AF, Lewandoski M, MuzyczkaN, Camper SA. Cre-mediated recombination in the pituitary gland.Genesis. 2000;28:167–174.

26. Zhao L, Bakke M, Krimkevich Y, et al. Hypomorphic phenotype inmice with pituitary-specific knockout of steroidogenic factor 1.Genesis. 2001;30:65–69.

27. Zhao L, Bakke M, Krimkevich Y, et al. Steroidogenic factor 1 (SF1)is essential for pituitary gonadotrope function. Development 2001;128:147–154.

28. Lanctot, C, Gauthier Y, Drouin J. Pituitary homeobox 1 (Ptx1) isdifferentially expressed during pituitary development. Endocrinol-ogy. 1999;140:1416–1422.

29. Hjalt TA, Semina EV, Amendt BA, Murray JC. The Pitx2 protein inmouse development. Dev Dyn. 2000;218:195–200.

30. Trainor PA, PP Tam, Cranial paraxial mesoderm and neural crestcells of the mouse embryo: co-distribution in the craniofacialmesenchyme but distinct segregation in branchial arches. Devel-opment 1995;121:2569–2582.

31. Trainor PA, Tan SS, Tam PP. Cranial paraxial mesoderm: regional-isation of cell fate and impact on craniofacial development inmouse embryos. Development. 1994. 120:2397–2408.

32. Lu MF, Pressman C, Dyer R, Johnson RL, Martin JF. Function ofRieger syndrome gene in left-right asymmetry and craniofacialdevelopment. Nature. 1999;401:276–278.

33. Semina EV, Reiter R, Leysens NJ, et al. Cloning and characteriza-tion of a novel bicoid-related homeobox transcription factor gene,RIEG, involved in Rieger syndrome. Nat Genet. 1996. 14:392–399.

34. Mears AJ, Jordan T, Mirzayans F, et al. Mutations of the forkhead/winged-helix gene, FKHL7, in patients with Axenfeld-Rieger anom-aly. Am J Hum Genet. 1998;63:1316–1328.

35. Finegold DN, Kimak MA, Lawrence EC, et al. Truncating mutationsin FOXC2 cause multiple lymphedema syndromes. Hum MolGenet. 2001;10:1185–1189.

36. Charles MA, Suh H, Hjalt TA, Drouin J, Camper SA, Gage PJ. PITXgenes are required for cell survival and Lhx3 activation. MolEndocrinol. In press.

37. Marcil A, Dumontier E, Chamberland M, Camper SA, Drouin J.Pitx1 and Pitx2 are required for development of hindlimb buds.Development. 2003;130:45–55.

38. Chang W, KhosrowShahian F, Chang R, Crawford MJ. xPitx1 playsa role in specifying cement gland and head during early Xenopusdevelopment. Genesis. 2001;29:78–90.

39. Stern CD, Fraser SE. Tracing the lineage of tracing cell lineages.Nat Cell Biol. 2001;3:E216–E218.

40. Couly GF, Coltey PM, Le Douarin NM. The developmental fate ofthe cephalic mesoderm in quail-chick chimeras. Development.1992;114:1–15.

41. Tucker GC, Aoyama H, Lipinski M, Tursz T, Thiery JP. Identicalreactivity of monoclonal antibodies HNK-1 and NC-1: conservationin vertebrates on cells derived from the neural primordium and onsome leukocytes. Cell Differ. 1984;14:223–230.

42. Dana MR. Corneal antigen-presenting cells: diversity, plasticity,and disguise: the Cogan lecture. Invest Ophthalmol Vis Sci. 2004;45:722–727.

43. Hamrah P, Zhang Q, Liu Y, Dana MR. Novel characterization ofMHC class II-negative population of resident corneal Langerhanscell-type dendritic cells. Invest Ophthalmol Vis Sci. 2002;43:639–646.

44. Hamrah P, Huq SO, Liu Y, Zhang Q, Dana MR. Corneal immunityis mediated by heterogeneous population of antigen-presentingcells. J Leukoc Biol. 2003;74:172–178.

45. Hamrah P, Liu Y, Zhang Q, Dana MR. The corneal stroma isendowed with a significant number of resident dendritic cells.Invest Ophthalmol Vis Sci. 2003;44:581–589.

46. Rodrigues MM, Rowden G, Hackett J, Bakos I. Langerhans cells inthe normal conjunctiva and peripheral cornea of selected species.Invest Ophthalmol Vis Sci. 1981;21:759–765.

47. Mo JS, Anderson MG, Gregory M, et al. By altering ocular immuneprivilege, bone marrow-derived cells pathogenically contribute toDBA/2J pigmentary glaucoma. J Exp Med. 2003;197:1335–1344.

48. Suh H, Gage PJ, Drouin J, Camper SA. Pitx2 is required at multiplestages of pituitary organogenesis: pituitary primordium formationand cell specification. Development 2002;129:329–337.

49. Beebe DC, Coats JM. The lens organizes the anterior segment:specification of neural crest cell differentiation in the avian eye.Dev Biol. 2000;220:424–431.


fate maps of neural crest and mesoderm in the mammalian...

Documents