Transcriptional profiling of endogenous germ layerprecursor cells identifies dusp4 as an essentialgene in zebrafish endoderm specificationJamie L. Brown*, Mirit Snir*, Houtan Noushmehr*†, Martha Kirby‡, Sung-Kook Hong§, Abdel G. Elkahloun¶,and Benjamin Feldman*�

*Medical Genetics Branch, ‡Genetics and Molecular Biology Branch, and ¶Genome Technology Branch, National Human Genome Research Institute;and §Program on Genomics of Differentiation, Eunice Kennedy Shriver National Institute of Child Health and Human Development, NationalInstitutes of Health, Bethesda, MD 20892

Communicated by Francis S. Collins, National Institutes of Health, Bethesda, MD, June 10, 2008 (received for review February 21, 2008)

A major goal for developmental biologists is to define the behav-iors and molecular contents of differentiating cells. We havedevised a strategy for isolating cells from diverse embryonicregions and stages in the zebrafish, using computer-guided laserphotoconversion of injected Kaede protein and flow cytometry.This strategy enabled us to perform a genome-wide transcriptomecomparison of germ layer precursor cells. Mesendoderm and ec-toderm precursors cells isolated by this method differentiatedappropriately in transplantation assays. Microarray analysis ofthese cells reidentified known genes at least as efficiently aspreviously reported strategies that relied on artificial mesend-oderm activation or inhibition. We also identified a large set ofuncharacterized mesendoderm-enriched genes as well as ecto-derm-enriched genes. Loss-of-function studies revealed that one ofthese genes, the MAP kinase inhibitor dusp4, is essential for earlydevelopment. Embryos injected with antisense morpholino oligo-nucleotides that targeted Dusp4 displayed necrosis of head tissues.Marker analysis during late gastrulation revealed a specific loss ofsox17, but not of other endoderm markers, and analysis at laterstages revealed a loss of foregut and pancreatic endoderm. Thisspecific loss of sox17 establishes a new class of endoderm speci-fication defect.

microarray � zgc:55423 � gastrulation � mkp2 � sox17

The establishment of germ layers and their coordinatedrearrangement during gastrulation is a core developmental

process (1). In organisms ranging from humans to flatworms,three primary germ layers arise: the mesoderm forms deeptissues such as blood and muscle, the endoderm produces visceralike the digestive tract and liver, and the ectoderm forms theepidermis and central nervous system. Cell lineage tracingstudies have mapped the precursor cells that are fated to formthese germ layers, and there are key commonalities to the fatemaps of different vertebrates (2). Germ layer precursor cells aretypically distributed in extended fields, making their quantitativeisolation by manual microsurgery particularly challenging.

Fluorescent proteins have opened the door to new microdis-section approaches in developmental biology. Placing differentfluorescent proteins behind different promoters, distinct colorshave been expressed in cell types comprising the nascent skin orlate gastrula endoderm of developing mice (3, 4). In thesestudies, f luorescently labeled cells from either newborn orembryonic mice were dissociated, separated by FACS, andmRNA was harvested to identify tissue-specific genes by mi-croarray analysis. More recently, photoconvertible fluorescentproteins such as Kaede (5) have become available, allowingresearchers to interactively convert expressing cells from onecolor to another. Several promising uses of photoconvertibleproteins in embryology have been described; however, they havenot yet been used to facilitate tissue-specific microdissection.

A reproducible map of embryonic cell fates in zebrafish isestablished �5 h post-fertilization (hpf), at the late blastula stage(6). At this time, two to four embryonic cell layers are arrangedin a dome that surrounds one hemisphere of a single large yolkcell (Fig. 1A). The endoderm precursor field lies within thelowest four tiers of this dome (Fig. 1 A and B) (7). The mesodermprecursor field overlaps the endoderm precursor field andextends to higher tiers (6). The cells that comprise these over-lapping fields are collectively termed mesendoderm precursors.Above the sixth tier, there is a staggered transition from meso-derm fates to ectoderm fates, with ectoderm precursors con-tinuing to the top of the embryo (6).

Traditional methods for isolating early mesendoderm andectoderm cells are technically challenging (excising the tops ofembryos with an eyebrow hair or explantation of individual cellswith a micropipet) and only a few studies have been publishedthat examined the fate changes they undergo when transplantedto host embryos (8, 9) or when explanted and cultured (10).Transgenic zebrafish lines with mesoderm-, endoderm-, or ec-toderm-restricted fluorescent protein expression have been re-ported (11, 12), but at the late-blastula stage, none of these arebright enough for FACS-based purification (J.L.B. and B.F.,unpublished observations).

To further our understanding of germ-layer differentiation,we have developed FACS-assisted microdissection of photola-beled (FAM-P) cells, a strategy for isolating zebrafish embryoniccells. FAM-P uses recombinant photoconvertible Kaede protein,which overcomes the spatiotemporal limitations of transgeni-cally expressed fluorescent proteins or injected mRNA andallows the quantitative isolation of early embryonic zebrafishcells from user-defined regions of interest. We isolated endog-enous mesendoderm and ectoderm precursors by FAM-P andtransplanted them into host embryos, where they contributed toa similar range of tissues as mesendoderm and ectoderm pre-cursors that were traditionally explanted by micropipet. DNAmicroarray analysis of FAM-P-isolated mesendoderm and ecto-

Author contributions: J.L.B., M.S., M.K., A.G.E., and B.F. designed research; J.L.B., M.S., M.K.,S.-K.H., and A.G.E. performed research; J.L.B., M.S., H.N., S.-K.H., and B.F. analyzed data;and B.F. wrote the paper.

The authors declare no conflict of interest.

Data deposition: The data reported in this paper have been deposited in the GeneExpression Omnibus (GEO) database, www.ncbi.nlm.nih.gov/geo (accession no. GSE8654).

†Present address: Eli and Edythe Broad Center for Regenerative Medicine and Stem CellResearch, Keck School of Medicine, University of Southern California, Los Angeles, CA90089.

�To whom correspondence should be addressed at: Vertebrate Embryology Section, Med-ical Genetics Branch, National Human Genome Research Institute, National Institutes ofHealth, Building 35 Room 1B205, 35 Convent Drive, MSC 3717, Bethesda, MD 20892-3717.E-mail: bfeldman@mail.nih.gov.

This article contains supporting information online at www.pnas.org/cgi/content/full/0805589105/DCSupplemental.

© 2008 by The National Academy of Sciences of the USA

www.pnas.org�cgi�doi�10.1073�pnas.0805589105 PNAS � August 26, 2008 � vol. 105 � no. 34 � 12337–12342

DEV

ELO

PMEN

TAL

BIO

LOG

Y

Dow

nloa

ded

by g

uest

on

Oct

ober

29,

202

0

derm precursors revealed they continue to express molecularmarkers specific to their regions of origin. Microarray analysisalso identified dozens of new mesendoderm precursor-enrichedgenes. Morpholino knockdown of one these genes, dusp4, pro-duced a loss-of-function phenotype, characterized by the down-regulation of a single endoderm marker during gastrulation,sox17, although other endoderm markers were expressed nor-mally, and a loss of foregut and pancreatic endoderm at laterstages. Our studies illustrate just some of the ways that FAM-Pcan be applied to experimental embryology. Given the spatialand temporal f lexibility that FAM-P offers, it should be usefulfor many other experiments in zebrafish and possibly in othermodel organisms.

ResultsDifferential Labeling of Germ Layer Precursors with the Kaede Pro-tein. The mesendoderm precursor field of zebrafish late blastulastage (40–50% epiboly, �5 hpf) embryos consists of a narrowband of cells that encircle a delicate yolk cell (Fig. 1 A and B).Because an efficient isolation of these cells by using traditionalmicrosurgical tools is untenable, we set out to develop a methodfor fluorescently labeling cells of interest and isolating them byFACS. For this purpose, we selected the photoconvertiblefluorescent reagent Kaede, a stony coral protein that fluorescesgreen (518 nm) in its full-length form but red (582–627 nm) afterphotocleavage by near-UV light (405 nm) (5). We discoveredthat nascent fluorescent protein maturation is slow relative tozebrafish development, such that late blastula-stage fluores-cence from injected kaede mRNA [see supporting information(SI) Fig. S1] and zygotically expressed transgenic fluorescentproteins (J.L.B. and B.F., unpublished observations) were notbright enough for FACS. To circumvent this difficulty, we

generated recombinant Kaede protein (Fig. 1C). Injection ofpurified Kaede protein was nontoxic and brightly labeled em-bryos from the time of injection (Fig. 1D) through the lateblastula stage (Fig. 1 E–G) and persisted until 24 hpf (Fig. 1H),after which time the signal gradually faded (data not shown),presumably because of cellular catabolism.

To specifically label mesendoderm precursors, we microin-jected Kaede protein into fertilized eggs and at the late blastulastage these embryos were mounted on a scanning confocalmicroscope, with the top of the embryo nearest to the lens (Fig.1E). Orientation of multiple specimens was facilitated by mount-ing embryos, their transparent chorions intact, in arrayed conicalmolds (13, 14). A circular swath was defined comprising theoutermost, and therefore lowest, tiers of cells, and these mes-endoderm precursors were photoconverted from green to red bycomputer-controlled scanning with a near-UV laser (Fig. 1F).Using this method, we can label the mesendoderm precursors of8–12 embryos in a single experiment.

Traditional fate mapping requires the lineage tracing of singlecells in dozens of embryos. Labeling the entire margin visualizedkey features of the fate map (Fig. 1B) in a single embryo. Imageanalysis on day 2 shows tissues normally derived from lower tierswith more red pixels, tissues normally from higher tiers withmore green pixels, and tissues normally derived from near themargin/nonmargin boundary with similar numbers of red andgreen pixels, presumably representing a mixture of red, green,and half-labeled yellow cells (Fig. 1H). This analysis demon-strates that we could correctly target the mesendoderm precur-sor field and indicates that Kaede protein labeling does not altercellular destinies. We have similarly used Kaede protein to labeland trace outcomes of other embryonic regions (Fig. S1). Suchlineage tracing experiments are facilitated by the ability of cells

Fig. 1. Use of Kaede protein to label embryonic cells and trace their lineages. (A) Schematic of a late blastula stage (40% epiboly) embryo. The embryonic layer(EL) is color coded according to germ layer precursor distribution, with dark red representing endoderm precursor-enriched, red representing mesodermprecursors, and green representing ectoderm precursors. Also indicated are the yolk cell (YC, tan), the enveloping layer (EVL, purple), and the yolk syncytial layer(YSL, light blue). For clarity, the yolk cell is represented as separated from the overlying tissues. (B) Fate map of a late-blastula stage embryo after Kimmel et al.(6) using the same color scheme as A and including precursor domains for the indicated outcomes in H. (C) Coomassie-blue stained polyacrylamide gel, showinga 34-kDa monomer of bacterially expressed Kaede after column purification. (D) Four-cell-stage (1 hpf) embryo shortly after injection with 3.5 ng of purifiedKaede protein, lateral view. (E) 40% epiboly-stage (5 hpf) embryo just before photolabeling along the margin, animal pole view. (F) Same embryo as in E, justafter photolabeling along the margin, animal pole view. (G) Close-up of a Kaede protein-injected embryo, directly after photolabeling of the margin, lateralview. Only optical sections from near the surface are shown. Brackets and numbers indicate the lowest cellular tiers. (H) Same embryo as in G, but now atpharyngula stage (24 hpf), lateral view, anterior to left. Representative tissues are indicated in H and their relative pixel values (% red of total red � green) matchexpected outcomes as follows: 1 (22%), 2 (25%), 3 (52%), 4 (52%), 5 (77%), and 6 (78%).

12338 � www.pnas.org�cgi�doi�10.1073�pnas.0805589105 Brown et al.

Dow

nloa

ded

by g

uest

on

Oct

ober

29,

202

0

labeled with Kaede protein to hold their color over time. Whenkaede mRNA is injected, photolabeled lineages become signif-icantly yellowed because of newly synthesized green Kaedeprotein (see Fig. S1).

Isolation of Mesendoderm and Ectoderm Precursors. To quantita-tively purify mesendoderm and ectoderm precursors, we disso-ciated labeled embryos into single cell suspensions and sortedcells by FACS (Fig. S2). To minimize losses of labeled cellsduring processing (e.g., pipetting, centrifugation, and FACSsteps), we mixed them with unlabeled cells from uninjectedsibling embryos at the point of trypsinization. Forward and sidescatter values for single embryonic cells were determined, andevents within those gates were further sorted by their relative redand green fluorescence. Using these parameters, we could purifyhomogenous populations of green and red cells, with a recoveryrate of input embryonic cells close to 75% and over 99% viability(data not shown).

A critical test of FAM-P’s utility is to determine whethersorted cells retain their developmental potentials and biases. Ithas previously been shown that ectoderm precursors trans-planted to the animal pole are reliably incorporated into hostectoderm (8, 9). We find the same is true for FAM-P-isolatedectoderm precursors, which contributed to ectoderm fates, suchas retinal neurons and forebrain cells (Fig. S2 and Table S1). Itwas previously shown that early-stage mesendoderm precursors(4.7–5 hpf) are readily reprogrammed to ectoderm (8, 9), butshortly thereafter [beyond 50% epiboly (5.3 hpf)] become com-mitted to mesendoderm fates (8). Our FAM-P purified mesen-doderm cells come from an intermediate stage [40–50% epiboly(5.0–5.3 hpf)] and consistent with this, they contributed both toectoderm- and mesendoderm-derived tissues, as did marginalcells of the same stage that we traditionally transplanted directlyfrom one embryo to another (Fig. S2 and Table S1). Weconclude that FAM-P-purified mesendoderm and ectodermprecursors maintain their endogenous commitment status anddevelopmental potential.

Identification of Previously Characterized and Uncharacterized GermLayer-Specific Genes. To study the transcriptomes of both themesendoderm and ectoderm precursors, we harvested and am-plified the RNA of purified precursor cells and cohybridizedthem onto oligo chips for microarray analysis (Fig. 2A). As acontrol against photolabeling effects, we also prepared andcohybridized amplified RNA probes from green and red controlcells, yielding no substantially enriched genes (Fig. 2B). Thus,the expression profile of embryonic cells is not substantiallyaltered by photolabeling per se. We identified 188 full-lengthcDNAs that were significantly enriched (P � 0.05) in themesendoderm precursor pool and 106 similarly enriched cDNAsin the ectoderm precursor pool.

We compiled cohorts of the top 60 unique genes from eachcomparison for which a minimum of annotated information wasavailable (Table S2) and examined them for over-representedgene ontology terms (Fig. 2C and Table S2). A striking percent-age (45%) of the mesendoderm precursor genes encode tran-scription factors (Fig. 2C). The frequency of transcription factorson the entire microarray, by contrast, is below 10%, and a similarfrequency is seen in the control cohort (Fig. 2C). We also notedan excess of ligands, agonists, and receptors and a paucity ofenzymes among the mesendoderm precursors compared withthe control cohort (Fig. 2C). An elevated number of transcrip-tion factors (32%) was also seen among ectoderm precursors(Fig. 2C), but no other molecular class seems to be over orunderrepresented.

We also scored the mesendoderm, ectoderm, and controlcohorts for membership in four signaling pathways—Nodal,WNT, FGF, and retinoic acid—which are known to be active

among mesendoderm precursors (15) and for known ectoderm-specific genes. We found that a full 38.3% of the mesendodermprecursor cohort genes fall into at least one of these fourpathways, whereas ectoderm and control cohort incidences were8.3% and 1.7%, respectively (Fig. 2D). Although we identifiedfar fewer highly enriched ectoderm genes, a slightly increasedfalse discovery filter of 5.5% reveals a substantial enrichment ofseveral well characterized neurectoderm genes, such as sox3 (16),sox 19a (17), lhx5 (18) and otx2 (19), as well as the nonneuralectoderm gene foxi1 (20) (Fig. 2C and Table S2).

A number of the mesendoderm-enriched genes we identifiedhave no reported expression in the late blastula stage. Wecharacterized 21 such genes from the mesendoderm precursorpool by cloning and performing whole mount in situ hybridiza-

Fig. 2. Transcriptome analysis of mesendoderm and neurectoderm precur-sor cells. (A and B) Volcano plots showing 31,356 (A) and 25,837 (B) oligos each.Each oligo is mapped along the y axis according to its relative enrichment ina direct comparison of the two source tissues (average normalized Log2 of thehybridization signal ratio) and along the x axis according to its significance(negative Log2 scale of uncorrected P value). (A) FAM-P isolated mesendodermprecursor cells (� 0 on Y axis) vs. cognate FAM-P isolated ectoderm precursorcells (� 0 on Y axis). (B) Kaede-injected and photoconverted embryo cells (� 0on Y axis) vs. cells from Kaede-injected embryos that were not photoconverted(� 0 on Y axis). (C and D) Bar graphs showing the representation of selectedmolecular functions (C) and signaling pathways (D) among the 60 most-enriched genes (Table S2). Mes, mesendoderm precursor enriched; Ect, ecto-derm precursor enriched; Con, red control cell enriched; E, Enzyme; LAR,ligand (agonist), antagonistic ligand or receptor; TF, transcription factor; N,Nodal pathway; W, Wnt pathway; F, FGF pathway; RA, retinoic acid pathway.Genes acting in multiple pathways were arbitrarily assigned to one pathwaywith the following priority: RA � F � W � N.

Brown et al. PNAS � August 26, 2008 � vol. 105 � no. 34 � 12339

DEV

ELO

PMEN

TAL

BIO

LOG

Y

Dow

nloa

ded

by g

uest

on

Oct

ober

29,

202

0

tions on late blastula-stage embryos. Ten of these genes showedmargin-specific staining in late blastula embryos, validating theirenrichment among mesendoderm precursors (Fig. 3 A–H anddata not shown). We also validated the ectoderm-specific ex-pression of several previously uncharacterized ectoderm precur-sor genes (data not shown).

We used antisense morpholino oligonucleotides (MOs) toblock the translation of several validated mesendoderm precur-sor genes and found one to associate with a distinct andreproducible phenotype: zgc:55423, which has 86% homology(Clustal alignment) to human dual-specificity phosphatase 4(DUSP4, alias: MKP2) and which we have accordingly termeddusp4. Dusps attenuate mitogen-activated protein kinase(MAPK) signaling via de-phosphorylation of two key residues,with human DUSP4 reported to target the MAPKs ERK andJNK (21). At the pharyngula stage, Dusp4-depleted embryosdisplayed developmental delays and small and necrotic heads,suggesting an essential role for Dusp4 in anterior development(Fig. 3J). Given dusp4’s expression in the margin, we investigated

the expression of mesoderm and endoderm markers in Dusp4morphants. We observed a drastic reduction in the expression ofsox17 in gastrula-stage endoderm cells as well as in a specializedgroup of cells termed the dorsal forerunners (Fig. 3 L and M)(22). Intriguingly, at this stage, other endoderm markers such asfoxa2 (23) (Fig. 3 N and O) were expressed normally in Dusp4morphants. These phenotypes are specific as we were able tophenocopy the morphological and molecular defects with asecond, nonoverlapping MO against Dusp4 (data not shown)and we could rescue them by coinjection of dusp4 mRNA (Fig.3K). Analysis of more differentiated endoderm with probes forfoxa2 (24) and insulin (24) transcripts revealed a loss of foregut(Fig. 3 P and Q) and pancreatic (Fig. 3 R and S) endoderm.

DiscussionThis paper introduces a method for viably isolating embryoniccells from user-defined regions and stages of interest. InjectedKaede protein can brightly label cells much earlier than exoge-nous mRNA, and there is no problem with de novo synthesis overthe course of lineage tracing experiments. Caged fluorophoresmight be suitable alternatives to Kaede protein; however, thesereagents have been commercially unavailable for several years.Another possible alternative is recombinant photoconvertibleEosFP protein, which has been used to achieve early cell labelingin frog embryos (25).

For the studies described here, we microdissected embryoswithin 1 h of labeling; however, labeled cells can be allowed tofurther differentiate before isolation. For instance, we easilyisolated early-labeled mesendoderm precursors at mid- andlate-gastrula stages (data not shown). Thus, a diverse range ofembryonic cell types can be labeled and purified by FAM-P. Onetopological limitation should be noted. The cells we labeled wereconveniently arranged in a shallow half dome. Differentiallabeling of deep and superficial cells is not feasible with thesingle-photon apparatus we used but might be achievable withmultiphoton systems, as has been reported for photoconversionof KikGR (26).

To our knowledge, our microarray analysis is the first suchstudy on endogenous germ layer precursors in zebrafish; how-ever, three papers have reported on the transcriptomes ofzebrafish embryos forced to over- or underproduce mesend-oderm via alterations in Nodal signaling (27–29). Each of thesestudies yielded important and complementary data, but it isuseful to perform a detailed comparison to assess the endoge-nous approach. Our method was best for identifying non-Nodalmesendoderm pathway members (i.e., FGF, WNT, and retinoicacid), with 12 of our top 60 genes being in these categories,compared with 9 of 98 (28) and 6 of 75 (27) in the other studiesthat presented their nonvalidated hits (Table S2). Our approachwas also successful for identifying Nodal pathway genes. Thehighest number of validated Nodal-regulated genes emergingfrom the four studies was 34 of a curated list of 72 (29). Our 188mesendoderm genes included 26 genes from this list and we havedozens more, such as those in Fig. 3, pending validation (TableS3). Thus, screening endogenous cells yielded an overall greaterdiversity of pathway-specific genes. Our proposed explanationfor this difference is that hyperactivation of Nodal signalingextinguishes the expression of Nodal and non-Nodal pathwaygenes that are uniquely expressed at lower endogenous Nodalsignaling levels.

Many of the mesendoderm precursor-enriched genes we iden-tified were not previously shown to be expressed by mesend-oderm precursors, and half of those we tested showed margin-enriched expression in situ, including dusp4. The loss ofexpression of sox17, but not foxa2, that we observed in Dusp4morphants during gastrulation is unusual and may define a newclass of endoderm defect; although reduction of sox17 has beenseen in zebrafish sqt, cyc, oep, bon, fau, and cas mutants, foxa2

Fig. 3. Expression and function of new mesendoderm genes. (A–H) Wholemount in situ hybridizations (WISH) on late-blastula-stage (5 hpf) embryos,validating the mesendoderm-specific expression of eight mesendoderm pre-cursor-enriched genes, as indicated. Animal pole views are shown, and C–Fhave their dorsal sides to the right (black arrows), as confirmed by doublestains with the chordin gene (data not shown). (I–K) Embryos injected with acontrol morpholino (Ctrl MO), Dusp4 morpholino (Dusp4 MO), or a rescuingcombination of Dusp4MO and dusp4 mRNA, shown at the pharyngula stage(28 hpf). (L–O) WISH on late-gastrula-stage (80% epiboly) control and Dusp4MO-injected embryos, visualizing expression of the endoderm markers sox17(L and M, dorsal views) and foxa2 (N and O, lateral views). Black arrowheadsindicate sample endoderm cells, and open arrows point to stained dorsalforerunner cells. (P–S) WISH on 18 somite-stage control (P), pharyngula-stage(26 hpf) Dusp4 MO-injected (Q), and pharyngula-stage (30 hpf) control andmorphant (R and S) embryos, visualizing foxa2 (P and Q) and insulin (R and S).Embryos in P and Q were stage matched rather than time matched to accountfor developmental delays of morphants. Black arrows in P and Q point toposterior endoderm, including foregut region. Black arrows and magnifica-tions in R and S highlight pancreatic stains.

12340 � www.pnas.org�cgi�doi�10.1073�pnas.0805589105 Brown et al.

Dow

nloa

ded

by g

uest

on

Oct

ober

29,

202

0

has been seen to be coordinately down-regulated where exam-ined (30). We also report a loss of foregut and pancreaticendoderm in Dusp4 morphants. This outcome is consistent withsox17 decreases, as a mouse knockout of Sox17 has beenreported to lack gut endoderm (31).

In summary, the ability to microdissect a wide range of earlycells in zebrafish opens up numerous experimental possibilitiesfor developmental biologists. Future studies might include, forinstance, comparisons of dorsal vs. ventral or younger vs. olderembryonic cell populations. We have shown that FAM-P-purified zebrafish cells can be used in transplantation assays andmicroarray studies, but many other applications can be envi-sioned, such as direct transcriptome sequencing, proteome anal-ysis, or the establishment of new embryonic cell lines.

Materials and MethodsMicroinjection and Kaede Photoconversion. Embryos were obtained from WTadult zebrafish (AB, WIK, or EK) via natural mating. pKaede-S1 vector (MBLInternational) was subcloned into pRSET B (Invitrogen), expressed in E. coli,and purified using Ni-NTA agarose affinity, sizing, and hydroxyapatite col-umns (expression and purification by ProteinOne Inc.). The protein was storedin 5 mM Hepes (pH 7) and 0.2 M KCl buffer at �80°C. For kaede and dusp4mRNA, NotI-linearized Kaede-pCS2� (32) or dusp4-pCS2 was used as a tem-plate for mMessage mMachine synthesis using SP6 RNA polymerase (Ambion).Kaede protein injections of 3.5 ng/embryo and mRNA injections of 300 pg/embryo were carried out at the 1–2 cell stage and embryos were subsequentlyincubated at 28°C. Embryos at the stage of interest were mounted in theirchorions in an agarose array of conical imprints (13, 14). By using a laserscanning microscope (Carl Zeiss LSM 510 META), Kaede was photoconvertedin regions of interest, such as the margin, with a 405-nm laser and imaged via488-nm and 543-nm excitation for green and red Kaede, respectively. Emittedlight was respectively filtered through 505–530-nm BP and 560-nm LP filters.For loss-of-function studies, 1–3 ng of an antisense splicing MO (Gene Tools),targeting zygotic dusp4 translation (5�-TAGCCCCCTGTAAAAGTGGAAAAGG-3�), was injected into one cell-stage embryos. A similar, nonoverlapping MO(5�-TGCTGCTTGTGTATTTACCTGGTCG-3�) yielded indistinguishable results.For control embryos, 5 ng of a standard control MO (5�-TAGTTAAGC-CTAGCTCTCATAAACT-3�) was injected. For rescue, 50-pg full-length dusp4mRNA was coinjected.

Embryonic Cell Sorting and Transplantation. Embryos were dechorionated withPronase (P8811, Sigma) and transferred to siliconized glass vials. Uninjectedembryos were added to samples with �20 labeled embryos to bring the totalnumber to 20–30. For dissociation, embryos were incubated for 6 min in 300�l of trypsin solution (27) and triturated once with a transfer pipette midin-cubation. After dissociation with trypsin, 30 �l of heat-inactivated FBS (HI-FBS)(Gibco) was added and the vial was then flooded with L-15 Medium (Sigma)containing 0.1% BSA, 0.8 mM CaCl2, 2 mM L-glutamine, 100 U/ml penicillin,and 0.1 mg/ml streptomycin. The cells were centrifuged at 100 � g for 4 minat 4°C and all but �700 �l of media was removed. The cells were transferredto a Titertube Micro Tube (Bio-Rad) inside of a 5-ml polystyrene tube, and 70�l of HI-FBS was added. The cells were transported on ice to a flow cytometer(FACSDiVa Vantage, BD Biosciences) where they were sorted based on theirforward scatter vs. side scatter and green vs. red fluorescence by using a100-�m tip at 8–10 psi into either TRIzol (Invitrogen) for RNA isolation or cell

media for cell viability assays or transplants. Methods for transplantationexperiments can be found in the legend of Fig. S2.

RNA Isolation, Amplification, and Cy Labeling. Cells were sorted into TRIzol,vortexed, frozen on dry ice, stored at �80°C, and extracted according to themanufacturer’s protocol with slight modifications. Approximately 300 ng to 1�g RNA was linearly amplified by using the Amino Allyl MessageAmp II aRNAAmplification kit (Ambion) with yields ranging from 12 to 30 �g of aRNA.aRNA samples were split and labeled, half with Cy3 mono NHS ester and halfwith Cy5 mono NHS ester (CyDyes from GE Healthcare; post-labeling reagentsfrom the MessageAmp II kit). Labeled aRNAs were fragmented by incubationat 94°C for 15 min in fragmentation buffer (Affymetrix). Paired probes wereprepared for four independent experiments microdissecting mesendodermand ectoderm precursors and three independent experiments using controlred and control green cells prepared as follows. A number of embryos wereinjected with Kaede and photoconverted to red in toto, and an equivalentnumber of Kaede-injected embryos were left in their green-fluorescent state.These red-labeled and green-labeled embryos were then pooled togetherwith unlabeled stage-matched embryos, disaggregated, and the entire cellpopulation was sorted to isolate red and green control cells.

Microarray Hybridization, Transcriptomes Analysis, and Validation. In-housemicroarray chips were used (13). Four biological replicates of mesendodermand ectoderm precursor probes were cohybridized, each with one dye swap,for a total of eight hybridizations. Similarly, three biological replicates ofred-cell and green-cell probes with one dye swap were cohybridized for a totalof six additional hybridizations. Hybridizations were performed overnight at45°C in Maui Mixer FL hybridization chambers (BioMicro Systems). Post-hybridization processing and scanning were as previously described (13). Datapoints with quality values of �0.85 were eliminated, and the dataset wasnormalized by median shift following a log2 transformation by using theAvadis 4.0 software program. P values and average log ratios were deter-mined for each oligo of the array as described in ref. 13 and predicted genes,ESTs, and duplicates were disregarded. Representation of molecular catego-ries among the most enriched genes in mesendoderm precursor vs. ectodermprecursor comparisons as well as red vs. green control-cell comparisons waspreliminarily assessed with GeneSifter software (VizX Labs), looking at ap-proximately the top 100 genes with available RefSeq IDs. This assessment wasfollowed by a hand-curated analysis of the 60 most-enriched genes from eachcategory (Table S2).

For validation studies, primers were designed to amplify the coding se-quence as well as the 5� and 3� UTRs of genes selected for study and cloned byPCR into TOPO TA vectors (Invitrogen). DIG-labeled RNA probes were pre-pared by using PCR-generated DNA templates and a DIG RNA labeling kit witheither SP6 or T7 polymerase (Roche), and in situ hybridizations were carriedout essentially as described in ref. 33, with post-hybridization washes per-formed by an automated liquid exchanger (Biolane HTI, Holle & Huttner).

All zebrafish work was approved by the National Human Genome ResearchInstitute’s Animal Care and Use Committee.

ACKNOWLEDGMENTS. We thank Darryl Leja for artwork and expert assis-tance with graphics; Alexandra Joyner, Paul Liu, David Bodine, Igor Dawid,and Mary LaMarca for helpful comments on the manuscript; Erich Roessler andStacie Anderson for valuable discussions; and Atsushi Miyawaki for sharingthe Kaede-pCS2� vector. This research was supported by the IntramuralResearch Program of the National Human Genome Research Institute, Na-tional Institutes of Health.

1. Haeckel E (1874) Die gastraea-theorie: Die phylogenetische classification des thier-reichs und die homologie der keimblatter. Jenaische Zeitschrift fur Naturwissen-schaften 8:1–55.

2. Hatada Y, Stern CD (1994) A fate map of the epiblast of the early chick embryo.Development 120:2879–2889.

3. Rendl M, Lewis L, Fuchs E (2005) Molecular dissection of mesenchymal-epithelialinteractions in the hair follicle. PLoS Biol 3:e331.

4. Sherwood RI, et al. (2007) Prospective isolation and global gene expression analysis ofdefinitive and visceral endoderm. Dev Biol 304:541–555.

5. Ando R, Hama H, Yamamoto-Hino M, Mizuno H, Miyawaki A (2002) An optical markerbased on the UV-induced green-to-red photoconversion of a fluorescent protein. ProcNatl Acad Sci USA 99:12651–12656.

6. Kimmel CB, Warga RM, SchillingTF (1990) Origin and organization of the zebrafish fatemap. Development 108:581–594.

7. Warga RM, Nusslein-Volhard C (1999) Origin and development of the zebrafishendoderm. Development 126:827–838.

8. David NB, Rosa FM (2001) Cell autonomous commitment to an endodermal fate andbehaviour by activation of Nodal signaling. Development 128:3937–3947.

9. Ho RK, Kimmel CB (1993) Commitment of cell fate in the early zebrafish embryo.Science 261:109–111.

10. Sagerstrom CG, Gammill LS, Veale R, Sive H (2005) Specification of the envelopinglayer and lack of autoneuralization in zebrafish embryonic explants. Dev Dyn232:85–97.

11. Ellingsen S, et al. (2005) Large-scale enhancer detection in the zebrafish genome.Development 132:3799–3811.

12. Fan X, et al. (2007) Nodal signals mediate interactions between the extra-embryonicand embryonic tissues in zebrafish. Dev Biol 310:363–378.

13. Pei W, et al. (2007) An early requirement for maternal FoxH1 during zebrafish gastru-lation. Dev Biol 310:10–22.

14. Schroter C, et al. (2008) Dynamics of zebrafish somitogenesis. Dev Dyn 237:545–553.

15. Schier AF, Talbot WS (2005) Molecular genetics of axis formation in zebrafish.Ann Rev Gen 39:561– 613.

16. Kudoh T, Concha ML, Houart C, Dawid IB, Wilson SW (2004) Combinatorial Fgf and Bmpsignalling patterns the gastrula ectoderm into prospective neural and epidermaldomains. Development 131:3581–3592.

Brown et al. PNAS � August 26, 2008 � vol. 105 � no. 34 � 12341

DEV

ELO

PMEN

TAL

BIO

LOG

Y

Dow

nloa

ded

by g

uest

on

Oct

ober

29,

202

0

17. Fong SH, Emelyanov A, Teh C, Korzh V (2005) Wnt signalling mediated by Tbx2bregulates cell migration during formation of the neural plate. Development 132:3587–3596.

18. Toyama R, et al. (1995) The LIM class homeobox gene lim5: Implied role in CNSpatterning in Xenopus and zebrafish. Dev Biol 170:583–593.

19. Nguyen VH, et al. (1998) Ventral and lateral regions of the zebrafish gastrula, includingthe neural crest progenitors, are established by a bmp2b/swirl pathway of genes. DevBiol 199:93–110.

20. Nissen RM, Yan J, Amsterdam A, Hopkins N, Burgess SM (2003) Zebrafish foxi onemodulates cellular responses to Fgf signaling required for the integrity of ear and jawpatterning. Development 130:2543–2554.

21. Teng CH, Huang WN, Meng TC (2007) Several dual specificity phosphatases coordinateto control the magnitude and duration of JNK activation in signaling response tooxidative stress. J Biol Chem 282:28395–28407.

22. Kikuchi Y, et al. (2004) Notch signaling can regulate endoderm formation in zebrafish.Dev Dyn 229:756–762.

23. Strahle U, Blader P, Henrique D, Ingham PW (1993) Axial, a zebrafish gene expressedalong the developing body axis, shows altered expression in cyclops mutant embryos.Genes Dev 7:1436–1446.

24. Kim HJ, et al. (2006) Genetic analysis of early endocrine pancreas formation in zebrafishMolecular endocrinology. Mol Endocrinol 20:194–203.

25. Wacker SA, Oswald F, Wiedenmann J, Knochel W (2007) A green to red photoconvert-ible protein as an analyzing tool for early vertebrate development. Dev Dyn 236:473–480.

26. Hatta K, Tsujii H, Omura T (2006) Cell tracking using a photoconvertible fluorescentprotein. Nature protocols 1:960–967.

27. Dickmeis T, et al. (2001) Identification of nodal signaling targets by array analysis ofinduced complex probes. Dev Dyn 222:571–580.

28. Link V, et al. (2006) Identification of regulators of germ layer morphogenesis usingproteomics in zebrafish. J Cell Sci 119:2073–2083.

29. Bennett JT, et al. (2007) Nodal signaling activates differentiation genes during ze-brafish gastrulation. Dev Biol 304:525–540.

30. Fukuda K, Kikuchi Y (2005) Endoderm development in vertebrates: Fate mapping,induction and regional specification. Dev Growth Differ 47:343–355.

31. Kanai-Azuma M, et al. (2002) Depletion of definitive gut endoderm in Sox17-nullmutant mice. Development 129:2367–2379.

32. Mutoh T, Miyata T, Kashiwagi S, Miyawaki A, Ogawa M (2006) Dynamic behavior ofindividual cells in developing organotypic brain slices revealed by the photoconvert-able protein Kaede. Exp Neurol 200:430–437.

33. Thisse C, Thisse B (1998) High resolution whole-mount in situ hybridization. ZebrafishSci Monitor 5.

12342 � www.pnas.org�cgi�doi�10.1073�pnas.0805589105 Brown et al.

Dow

nloa

ded

by g

uest

on

Oct

ober

29,

202

0