DEVELOPMENTAL DYNAMICS 215:352–370...

Zebrafish stat3 Is Expressed in Restricted Tissues DuringEmbryogenesis and stat1 Rescues Cytokine Signaling in aSTAT1-Deficient Human Cell LineANDREW C. OATES,1,2* PATRIK WOLLBERG,1 STEPHEN J. PRATT,3 BARRY H. PAW,3 STEPHEN L. JOHNSON,4ROBERT K. HO,2 JOHN H. POSTLETHWAIT,5 LEONARD I. ZON,3 AND ANDREW F. WILKS1

1Ludwig Institute for Cancer Research, Melbourne Tumour Biology Branch, Royal Melbourne Hospital, Victoria, Australia2Department of Molecular Biology, Princeton University, Princeton, New Jersey3Howard Hughes Medical Institute, Children’s Hospital, Boston, Massachusetts4Department of Genetics, Washington University Medical School, St. Louis, Missouri5Institute of Neurosciences and Institute of Molecular Biology, University of Oregon, Eugene, Oregon

ABSTRACT Transcription factors of theSTAT family are required for cellular responsesto multiple signaling molecules.After ligand bind-ing-induced activation of cognate receptors, STATproteins are phosphorylated, hetero- or ho-modimerize, and translocate to the nucleus. Sub-sequent STAT binding to specific DNA elementsin the promoters of signal-responsive genes al-ters the transcriptional activity of these loci.STAT function has been implicated in the trans-duction of signals for growth, reproduction, viraldefense, and immune regulation. We have iso-lated and characterized two STAT homologs fromthe zebrafish Danio rerio. The stat3 gene is ex-pressed in a tissue-restricted manner during em-bryogenesis, and larval development with high-est levels of transcript are detected in the anteriorhypoblast, eyes, cranial sensory ganglia, gut, pha-ryngeal arches, cranial motor nuclei, and lateralline system. In contrast, the stat1 gene is notexpressed during early development. The stat3gene maps to a chromosomal position syntenicwith the mouse and human STAT3 homologs,whereas the stat1 gene does not. Despite a higherrate of evolutionary change in stat1 relative tostat3, the stat1 protein rescues interferon-signal-ing functions in a STAT1-deficient human cellline, indicating that cytokine-signaling mechanismsare likely to be conserved between fish and tetra-pods. Dev Dyn 1999;215:352–370. r 1999 Wiley-Liss, Inc.

Key words: zebrafish; stat1; stat3; STAT; JAK; cy-tokine; interferon-a; interferon-g;IL-6; sIL-6; signal transduction; phos-photyrosine; transcription factor; de-velopment; embryogenesis; blastula;prechordal plate; cranial neuralcrest; cranial motor nuclei; sensoryganglia; lateral line; retinal ganglioncell; chromosome evolution; synteny;in situ hybridization; degenerate oli-gonucleotide PCR; gene mapping

INTRODUCTION

Signal tranducer and activator of transcription (STAT)proteins are activated in response to the binding of awide range of signaling molecules to their specificreceptors. The best understood system in which STATsare required is the JAK/STAT pathway downstream ofcytokine receptors (for review, see Wilks and Harpur,1996; Darnell, 1997). On ligation of a cytokine receptor,constitutively associated JAK kinases become acti-vated, phosphorylating themselves, the receptor cyto-plasmic domain, and associated or recruited proteins.STAT family proteins are recruited to an activecytokine receptor complex by a C-terminal-locatedSH2 domain binding to phosphotyrosine residues inthe receptor (Greenlund et al., 1995). Phosphorylationof the STAT activation tyrosine, immediately C-terminal to the SH2 domain (Shuai et al., 1993), allowsSTAT proteins to dimerize and translocate to thenucleus (Shuai et al., 1994) where they recognizespecific DNA sequences or STAT-binding elements(SBEs) with high affinity (Horvath et al., 1995; Schind-ler et al., 1995) and alter transcriptional events (Zhanget al., 1996; Horvai et al., 1997; Kurokawa et al., 1998;Korzus et al., 1998). The closely related STAT1 andSTAT3 proteins are activated by signaling moleculesthat use cytokine receptors such as interferons (IFNs)and interleukin-6 (IL-6) (Fu et al., 1992; Akira et al.,1994), growth factors that signal through receptortyrosine kinases (RTKs) such as epidermal growthfactor (EGF) (Zhong et al., 1994a), and hormones that

Grant sponsors: Anti Cancer Council of Victoria PostgraduateResearch Scholarship; Swedish Natural Science Research Council;Swedish Institute; Wenner-Gren Center Foundation for ScientificResearch; National Institutes of Health; National Science Foundation;Rita Allen Foundation.

Dr. Patrik Wollberg’s present address is The Ludwig Institute forCancer Research, Biomedical Center, S-751 24 Uppsala, Sweden.

Dr. Andrew Wilks’ present address is Mutation Research Centre, St.Vincent’s Hospital, 41 Victoria Parade, VIC 3065, Australia.

*Correspondence to: Andrew C. Oates, Department of MolecularBiology, Princeton University, Washington Road, Princeton, NJ 08544.E-mail: aoatesrmolbio.princeton.edu

Received 24 November 1998; Accepted 21 May 1999

DEVELOPMENTAL DYNAMICS 215:352–370 (1999)

r 1999 WILEY-LISS, INC.

bind and activate serpentine receptors such as angioten-sin II and serotonin (Marrero et al., 1995; Guillet-Deniau et al., 1997).

The developmental role of members of the STATfamily has been analyzed in Dictyostelium, Drosophila,and mice. During Dictyostelium mound formation, Dd-

Figure 1. (Continued on following page)

353ANALYSIS OF stat1 AND stat3 GENES FROM ZEBRAFISH

STAT is required for stalk cell differentiation (Kawataet al., 1997) and is activated in this capacity by theserpentine cAMP receptor (Araki et al., 1998). A defi-ciency of STAT92E in the developing fruit fly embryoresults in segmental defects and larval lethality (Hou etal., 1996; Yan et al., 1996), an effect that is phenocopiedby loss of the Drosophila JAK homolog, hopscotch(Perrimon and Mahowald, 1986; Binari and Perrimon,1994). Further evidence for the combined regulatoryactivity of the JAK and STAT genes in Drosophilacomes from the suppression of the dominant leukemichopTuml phenotype by a loss of stat92E function (Yan etal., 1996). Stat3 is activated and competent to bindDNA in extracts from mouse embryos, suggesting thatSTAT-mediated signaling is taking place in early verte-brate development (Duncan et al., 1997). Consistentwith this, null mutants in the Stat3 gene of mice areembryonic lethal, causing developmental arrest imme-

diately before gastrulation (Takeda et al., 1997). Theearly mouse embryo is relatively intractable to embryo-logical manipulation, however, and neither the tissueaffected by the absence of Stat3 nor the signal trans-duced is understood. Furthermore, the requirement forSTAT3 in other, later developmental processes is un-known. In contrast, although STAT1 is activated by avariety of signaling molecules, a Stat1-deficient mouseis viable and fertile,but displays no innate response toviral or bacterial infection, corresponding to defectsspecific to the transduction of IFN-a/b and IFN-gsignals (Durbin et al., 1996; Meraz et al., 1996). Thissurprising specificity in biological effect suggests thatmost of STAT1 activity in vertebrates is dispensable fornormal development and adult life. However, loss of

Figure 1. (Continued.)

Fig. 1. Structure and evolutionary analysis of zebrafish stat genes a:Alignment of zebrafish and mammalian STAT1 and STAT3 proteins withDrosophila STAT92E. The amino acid sequences of zebrafish (Daniorerio, Dr) stat1 and stat3 were aligned with the human (Homo sapiens, Hs)STAT1 and STAT3 protein sequences (Schindler et al., 1992; Akira et al.,1994) and the D. melanogaster (Dm) STAT92E sequence (Yan et al.,1996) as an outgroup by the CLUSTAL alignment algorithm (Higgins et al.,1996). The names of the sequences are displayed to the left of thealignment, and the amino acid position is indicated on the right. Residuesthat are conserved in most of the sequences (3 of 5) are highlighted byblack squares for identity and colored squares for similarity, to indicateregions of high evolutionary constraint. Structural features of the STATprotein family as determined by functional studies and crystal structures(Becker et al., 1998; Chen et al., 1998; Vinkemeier et al., 1998) areindicated on the alignment. The position of the N-terminal proteininteraction, coiled-coil, DNA binding, linker and SH2 domains are indi-cated by square, arrowed brackets. Amino acid residues required for DNAbinding within the DNA binding domain are denoted with asterisks abovethe sequence (Horvath et al., 1995, Fukada et al., 1996), and the tyrosineand serine phosphorylation motifs in the C-terminus are indicated witharrowheads (Shuai et al., 1993; Kaptein et al., 1996; Wen et al., 1995;Wen and Darnell, 1997). The binding site for PI3-kinase in Stat3 isindicated with an arrowhead (Pfeffer et al., 1997). The position of thealternate exon 23 splicing sites are indicate by a ‘‘3’’ and a ‘‘1’’ above thelast residue of exon 22 of STAT3 and STAT1, respectively. The sequenceof the alternate C-terminus of Dr stat3 or Hs STAT3 is not shown. b:Phylogeny of the mammalian STAT gene family. The amino acid se-quences of all known STAT homologs from a single mammalian species(human, Homo sapiens, Hs) were aligned with the amino acid sequenceof the D. melanogaster (Dm) STAT homolog, Dm STAT92E as above. Thisalignment was used to construct a maximum parsimony dendrogram toinfer the likely genealogy of the mammalian STAT family. The names ofthe sequences are indicated to the right of the dendrogram, and thesequences are taken from the following sources: STAT1 (Schindler et al.,1992), STAT2 (Fu et al., 1992), STAT3 (Akira et al., 1994), STAT4 (Xu etal., 1996), STAT5A (Hou et al., 1995), STAT5B (Silva et al., 1996), STAT6(Hou et al., 1994), and DSTAT92E (Yan et al., 1996). c: Phylogeny of theSTAT1/3/4 gene subfamily. The amino acid sequences of all knownmembers of the STAT1/3/4 clade of STAT homologs from human (Homosapiens, Hs) mouse, (Mus musculus; Mm), rat (Rattus norvegicus, Rn) andrainbow trout (Oncorhynchus mykiss, Om) were aligned with the zebrafish(Danio rerio, Dr) stat1 and stat3 protein sequences and the D. melanogaster(Dm) STAT92E sequence as an outgroup as above (a). This alignment wasused to construct a dendrogram as above (b) to infer the likely genealogy of themammalian STAT1/3/4 gene subfamily. The names of the sequences areindicated to the right of the dendrogram, and the sequences are taken from thefollowing sources: Om Stat1a/b (Johnson et al., unpublished 1996, accessionnumbers: U60331, U60332), Hs STAT1 (Schindler et al., 1992), Mm Stat1(Zhong et al., 1994b), Hs STAT4 (Xu et al., 1996), Mm Stat4 (Zhong et al.,1994b), Hs STAT3 (Akira et al., 1994), Mm Stat3 (Zhong et al., 1994a), RnStat3 (Ripperger et al., 1995), Om Stat3 (Johnson et al., unpublished 1996,accession number: U60333), and STAT92E (Yan et al., 1996).

354 OATES ET AL.

function studies do not reveal the full range of STAT-mediated biological effects. The deregulated activationof STAT proteins is a common occurrence in viralinfection and neoplasia and may contribute to malignanttransformation (reviewed in Garcia and Jove, 1998).

To better understand the role of the STAT genes indevelopment and disease and to establish the evolution-ary history of the STAT gene family, we have isolatedSTAT1 and STAT3 homologs from the teleost Daniorerio. In this report we examine the conservation ofstructure in vertebrate STAT1 and STAT3 proteins and

test conservation of function in a cytokine-signalingassay in human fibroblasts. Furthermore, we map thestat genes to the linkage groups of the zebrafish andexamine the expression of stat3 and stat1 in the develop-ing embryo and larva.

RESULTSIsolation of STAT Transcripts From the Zebrafish

To isolate STAT homologs from the zebrafish, degen-erate oligonucleotide-primed polymerase chain reac-tion (PCR) was used with primers designed to amplify

b

Fig. 2. Complementation of cytokine signaling in the STAT1-deficientU3A human cell line with zebrafish stat1 cDNA. a: Structure of stat1mutants used in the rescue of U3A cells. The N-terminus of the full-lengthstat1 is depicted in the upper portion of the diagram with the conservedarginine (R) and glutamic acid (E) residues indicated. Below this, the stat1DPIC mutant illustrating the in-frame removal of the core of the N-terminaldomain is depicted. Last, the truncated stat1 TGA is shown, whichcontains an in-frame stop codon after 28 amino acids of the nativesequence. b: Survival in response to IFN-a of stat1-expressing U3A cellsunder HAT selection. Photographs of a representative experiment show-

ing the survival after 2 weeks of parental cell line 2fTGH, the STAT1-deficient derivative cell line U3A, and U3A cells stably transfected withvarious mouse Stat1 or zebrafish stat1 cDNA constructs as indicated inthe presence of IFN-a (600 IU/mL) and HAT medium. c: Growthretardation of stat1-expressing U3A cells in response to IFNa and IFNg.Stably transfected STAT1-deficient U3A cell lines expressing variousmouse Stat1 or zebrafish stat1 cDNA constructs were grown in IFN-a (600IU/mL) or IFN-g (15 ng/mL) and cell number counted each day for 6–7days. Results are the mean and standard deviation of three experiments.(Contined on following page.)


the DNA binding and SH2 domains of STAT genes.Sequencing of subcloned PCR products amplified fromembryonic complementary DNA (cDNA) revealed thepresence of two distinct PCR products with high similar-ity to mammalian STAT genes. These products wereused to screen several cDNA libraries, resulting in theisolation of full-length cDNA clones corresponding toeach of the PCR products. On the basis of databasesearching with these sequences, the cDNAs representtranscripts from zebrafish stat1 and stat3 genes.

One full-length cDNA from the stat1 gene was recov-ered, comprising 2,520 nucleotides (accession number:AJ005692). Examination of the sequence of the longestopen reading frame (ORF) present in this clone indi-cates a predicted protein of 749 amino acids. Fourindependent full-length stat3 cDNAs were isolated,possessing both start and in frame stop codon as well asa poly (A) tract. The longest of these was 3,071 nucleo-tides in length (accession number: AJ005693), and thepredicted size of the stat3 protein is 806 amino acids.From 14 independent stat3 cDNAs, two clones con-tained identical short internal deletions of 56 bp,consistent with the omission of an alternately splicedexon. The alternate cDNA is predicted to yield 2 more

amino acids from the deletion site before reaching astop codon, giving rise to a C-terminal-truncated pro-tein of 720 amino acids. The longer stat3 transcript istermed stat3a, and the shorter, less abundant tran-script is named stat3b, consistent with the nomencla-ture for alternately spliced forms of the mammalianSTAT1 and STAT3 genes (Darnell et al., 1994; Schaeferet al., 1995).

Structure of the Zebrafish stat Transcripts

Alignment of the zebrafish stat1 and stat3 proteinswith their mammalian STAT homologs and the Dro-sophila STAT92E protein indicates that the modularstructural features of STAT proteins are conservedbetween fish and mammals; sequence motifs character-istic of functional STAT proteins are annotated inFigure 1a. Residues in the N-terminus required forcooperative binding to DNA (Vinkemeier et al., 1996,Xu et al., 1996) and residues critical for DNA bindingare highly conserved. In mouse Stat3, mutation of–E417E– to –A417A– or of –V444VV– to –A444AA– abro-gates DNA-binding activity (Horvath et al., 1995);however, in stat1 there was a conservative amino acidchange at the position corresponding to V444: an isoleu-

Figure 2. (Continued.)

356 OATES ET AL.

cine was found here instead. These valine residues areconserved in all other STAT1 and STAT3 proteins andin STAT92E, suggesting that they are essential forDNA binding and raising the possibility that zebrafishstat1 may exhibit altered DNA-binding characteristics.The SH2 domains of both zebrafish proteins displayed aconservation of the essential arginine required forphosphotyrosine binding (Shuai et al., 1994; Improta etal., 1994), and the PI3K binding site found in the mouseStat3 was conserved in stat3 at –Y658KIM–. Examina-tion of the activation tyrosine immediately C-terminalto the SH2 domain indicates that the –YxxT– motifbound by the SH2 domain of other STAT proteins wasconserved in all teleost and mammalian sequences.Zebrafish stat3 (–Y716LKT–) was identical to mouse,whereas zebrafish stat1 (–Y698IST–) varied from mouse(–Y701IKT–), suggesting that stat1 may possess alteredbinding characteristics. The most C-terminal of thesequence motifs is the site of serine phosphorylation inthe activation domain (Wen et al., 1995; Wen andDarnell, 1997); in both zebrafish proteins the consensus–PMSP– was found. The position of the deletion in thevariant stat3 cDNA was identical to that in the mouseStat3b homolog, which results from the omission ofexon 23 due to alternate mRNA splicing (Schaefer et al.,1995; Shi et al., 1996). The putative 58-splice siteoccurred in the codon of –P718– and the 38-splice sitewas in the codon of –P737–. The result is a shift in thereading frame of the equivalent of mouse exon 24,resulting in the addition of two amino acids beforereaching a stop codon thus: –WD*. Consequently, thesite of serine phosphorylation in stat3 (–PMS751P– islost. Further comparison of the sequences is detailed inthe legend to Figure 1a.

There are seven known mammalian STAT genes.Alignment of the mammalian STAT proteins using theClustal algorithm and subsequent phylogenetic recon-struction, using the D. melanogaster STAT92E as anoutgroup, produces the dendrogram shown in Figure1b. From this analysis of sequence structure, it appearsthat STAT1, STAT3, and STAT4 form a natural clade,indicating that they share a common ancestor after thedivergence of the other STAT genes. With an overallidentity of 63.9% to mouse Stat1, the zebrafish stat1protein is significantly more diverged than the paralo-gous stat3, which displays an overall identity to mouseStat3 of 86.5%. By comparison, the jak2 proteins ofzebrafish are approximately 65% identical to theirmouse homolog (Oates et al., 1999). Phylogenetic analy-sis of all known STAT1 and STAT3 genes indicates thatthe STAT3 gene appears to be evolving 2.5 times slowerthan STAT1 in both fish and mammalian lineages (Fig.2c).

Rescue of a Human Cell Line Deficient for STAT1With Zebrafish stat1 cDNA

The high degree of sequence divergence between themammalian STAT1 and fish stat1 genes prompted theexamination of functional conservation between the

corresponding protein products. Use was made of ahuman fibroblast cell line that is STAT1 deficient (U3A)to test whether zebrafish stat1 could substitute forhuman STAT1 in signaling responses to IFN-a/b, IFN-g,and IL-6 (Pellegrini et al., 1989; McKendry et al., 1991;Muller et al., 1993). In U3A cells, expression of the E.coli guanine phosphoribosyltransferase (gpt) gene istightly controlled by 1.8 kb of the upstream region ofthe human IFN-a/binducible gene 6–16 (Porter et al.,1988) and can be used to select for functional IFNsignaling on the basis of survival in antibiotic medium.

stat1 cDNA was expressed in U3A cells and thesurvival of these cells under selection in the presence ofIFN-a/b, IFN-g or IL-6/sIL-6R was compared with U3Acells expressing mouse Stat1 and parental cells with afunctional STAT1 gene (2fTGH). As controls, two vari-ants of the zebrafish stat1 cDNA were also expressed(Fig. 2a): stat1 DPIC contains an in-frame deletion of 22amino acids in the conserved N-terminal oligomeriza-tion domain (Vinkemeier et al., 1996), and stat1 TGAexpresses a truncated stat1 product by virtue of anin-frame stop codon after 28 amino acids of the nativeprotein. As shown previously (Muller et al., 1993),mouse Stat1 was capable of rescuing U3A survival inresponse to IFN-a selection in a manner indistinguish-able from the parental 2fTGH cells, whereas transfec-tion of the expression vector alone led to cell death (Fig.2b). Expression of the zebrafish stat1 gene allowed cellsurvival under IFN-a stimulation, in contrast to expres-sion of either stat1 DPIC or stat1 TGA (Fig. 2b). stat1expression also conferred survival to U3A cells inresponse to IFN-g and IL-6 signaling (data not shown).This indicates that the zebrafish stat1 protein is ca-pable of mediating transcriptional activation from the6–16 promoter in response to IFNs and IL-6.

One of the main biological functions of the IFNs inmany cell types is growth inhibition. Consistent withthis, the human 2fTGH cell line arrests growth inresponse to either IFN-a or IFN-g; this response is lostin U3A cells, indicating that it is STAT1 dependent(Bromberg et al., 1996). To determine whether thecomplementation of human STAT1 function by ze-brafish stat1 was restricted to the activation of the 6–16promoter, the ability of the zebrafish stat1 protein tointegrate into the mammalian cell’s growth inhibitionpathway was assayed by measuring cell number underIFN stimulation in U3A cell lines expressing stat1. Asillustrated in Figure 2c, the growth inhibition conferredto U3A cells by expression of zebrafish stat1 in responseto either IFN-a or IFN-g was indistinguishable to thatconferred by mouse Stat1. In contrast, IFN stimulationdid not alter the growth of the U3A line expressingeither stat1 mutant (Fig. 2c). These results indicate afull rescue of IFN-mediated growth inhibition, suggest-ing that zebrafish stat1 is broadly active in the mamma-lian IFN-dependent mechanism of growth suppression.Combined, these results indicate a high degree offunctional conservation in the STAT1 protein betweenfish and mammals.


Genetic Mapping of the Zebrafish stat Genes

To enable genetic mapping of the stat1 and stat3genes, sequence from physically linked genomic DNA(data not shown) was used to design PCR primers thatamplified polymorphic DNA markers segregating in aC32xSJD hybrid mapping cross (Johnson et al., 1996).stat1 was mapped by using SSCP analysis of the93.T7–1/93.T7–3 sequence polymorphism in 78 meioses(Fig. 3a). According to the marker’s segregation, stat1lies on LG22, 13.4 cM proximal to the 14Y.1280.s RAPDmarker (Johnson et al., 1996) and 4.3 cM distal to9AB.590.c (Fig. 3b). stat3 was mapped by using SSCP

analysis to follow the segregation of the 138.SP6–1/138.SP6–2 sequence polymorphism in 53 meiotic events(Fig. 3c). This analysis placed the stat3 gene distal tothe hoxba cluster on LG3. Analysis of an additional 137meioses with breakpoints in this region (data notshown) resolved the placement of stat3 to 2.8 cMproximal to simple sequence repeat marker SSR14(Goff et al., 1992) and 3.0 cM distal to SSLP z5033(Knapik et al., 1996) (Fig. 3d).

The linkage of the 138.SP6–1/138.SP6–2 mappingpolymorphism to the stat3 gene was confirmed byanalysis of genomic DNA from embryos derived from

Fig. 3. Genetic mapping of the stat1 and stat3 genes.a: Segregation of the 93.T7–1/93.T7–2 stat1 SSCPpolymorphism in the C32xSJD mapping cross. Eightrepresentative haploid embryos of the C32xSJD map-ping cross typed for a polymorphic marker derived fromgenomic DNAassociated with the stat1 gene by nondena-turing polyacrylamide gel electrophoresis are shown. Thepolymorphism presented as a presence or absence ofproduct, possibly due to a sequence variant in one of theprimer sites. The presence or absence of the productswas assigned to the maternal (M) or paternal (P) genomeat random, and segregation in the panel was scored. b:Genetic map position of the stat1 gene. Analysis of thesegregation of a stat1-associated marker in 71 meioticevents of the C32xSJD panel places the zebrafish stat1locus on LG22. c: Segregation of the 138.SP6–1/138.SP6–2 stat3 SSCP polymorphism in the C32xSJDmapping cross. Eight representative haploid embryos ofthe C32xSJD mapping cross typed for a polymorphicmarker derived from genomic DNA associated with thestat1 gene by nondenaturing polyacrylamide gel electro-phoresis are shown. The size variants of the productswere assigned to the maternal (M) or paternal (P)genome at random, and segregation in the panel wasscored. d: Genetic map position of the stat3 gene.Analysis of the segregation of a stat3-associated markerin 53 meiotic events places the zebrafish stat3 locus onLG3. e: Synteny of STAT3 loci between the zebrafish,mouse and human. A schematic diagram of the syntenicchromosome segment containing STAT3 homologs fromzebrafish (LG3), human (Hsa 17), and mouse (Mmu 11)genomes. Note that genes have been illustrated in thesame relative positions on syntenic chromosomes; how-ever, in situ, local gene order varies between chromo-somes. The diagram is not to scale.

358 OATES ET AL.

the g-ray-induced deletion mutant b245, which re-moves LG3 distal to, and including the hoxba cluster(Fritz et al., 1996). On the basis of the position of138.SP6–1/138.SP6–2 marker, the stat3 locus should beremoved by this deletion. To confirm that the mappingmarker was linked to the stat3 locus in situ, embryoswere scored for the b245 phenotype and typed for thepresence of LG3 markers (data not shown). The distal,hoxba linked marker cdc27 (B. Paw, unpublished) wasabsent from mutant b245 DNA but present in thewild-type, whereas the centromeric marker a-globinwas unaffected by b245, consistent with a distal loss ofLG3. Both the 138 C20 mapping marker and a productderived from within the 38-UTR of the stat3 cDNA werelost only in the mutant b245 embryos, similar to thecdc27 marker. These results demonstrate the linkage ofthe 138.SP6–1/138.SP6–2 mapping marker to the stat3gene and independently confirm the assignment ofstat3 to distal LG3. The arrangement of loci on syntenicchromosome segments containing the STAT3 gene fromzebrafish, human, and mouse is indicated in diagram-matic form in Figure 3e.

Expression of Zebrafish stat Genesin Development

The expression of stat1 was examined in the develop-ing embryo and larva but could not be detected by insitu hybridization at any stage before 6 dpf (data notshown). In contrast, strong, dynamic expression of stat3was observed. stat3 expression could not be detected incleavage stage embryos (data not shown), indicatingthat there was no maternal contribution of stat3 to theembryo. After midblastula transition (MBT), when thezygotic genome becomes transcriptionally active, stat3was expressed at high level (Fig. 4a). This expressionpersisted through the onset of epiboly (Fig. 4b) and wasfound only in the deep cells of the embryo, beingexcluded from the enveloping layer (Fig. 4c). At theonset of gastrulation, stat3 transcript levels droppedrapidly so that most of gastrulation takes place in theabsence of stat3 message (data not shown). Toward theend of gastrulation, at 90% epiboly, stat3 expressionwas detected in a new domain in the anteriormostregion of the axial mesendoderm (Fig. 4d). As the yolkplug was closed by the completion of epiboly, theanterior domain of expression was enlarged and inten-sified (Fig. 4e). The stat3-expressing cells formed acircular patch that lay just posterior to the edge of theneural plate (Fig. 4f) and anterior to the rostral limits ofshh and hlx1 expression (data not shown). At bud stage,the position of the stat3-expressing cells in the anterior-most prechordal plate mesoderm, or polster, was clear(Fig. 4g), and the first expression of stat3 in theoverlying neurectoderm was detected in a limited num-ber of cells (Fig. 4h). Elevated expression of stat3 in theanterior mesendoderm suggests the existence of asignal passing into this tissue.

With the beginning of somite formation, expression ofstat3 was widely upregulated along the embryonic axiswith the exception of the adaxial cells, an epitheliallayer lying laterally to the notochord (Fig. 5a). At the5-somite stage, elevated expression was detected in thepresumptive hindbrain and in the paraxial mesodermof the newly forming somites (Fig. 5b). By midsegmen-tation stages, expression in the hindbrain and trunkdecreased, whereas expression in the forebrain andmidbrain and the growing optic primordia remainedhigh (Fig. 5c). stat3 levels increased dramatically in thetail bud as it began extension at the 13-somite stage; bythis stage also, a rostrocaudal gradient of stat3 expres-sion was evident along the notochord. At the 16-somitestage, stat3 message was concentrated in the forebrainand eyes and in a paired cluster of cells lateral to theneural tube at an axial level rostral to the otic primor-dia (Fig. 5d). Shortly after this first pair of cell clustersexpressed stat3, two more were detected: one immedi-ately caudal to the midbrain hindbrain junction and theother immediately caudal to the otic primordia. Thesecell clusters correspond in location and morphology tothe primordia of the cranial sensory ganglia as markedby expression of isl1 (Inoue et al., 1994). In the posteriormesoderm, stat3 message was found at higher level inthe six most newly formed somites and the presegmen-tal mesoderm corresponding to the somite about toform, as well as the everting tail bud. Expression in thenewly formed somites persisted, but by the 19-somitestage became localized to the ventral portion of thesomites (Fig. 5e). Expression of stat3 in the primordia ofthe cranial sensory ganglia increased rapidly so that by24 hpf the maturing structures of the trigeminal,anterior lateral line, acoustic and posterior lateral lineganglia (Metcalfe et al., 1990; Holder and Hill, 1991)expressed high levels of transcript (Fig. 5f, g). Thedorsal midbrain and a stripe of cells in the mid-diencephalon between the eyes were also expressingstat3 at this stage. Expression within the eyes wasrestricted by 24 hpf, with highest accumulation of stat3transcript found in the cells of the neuroepitheliumclosest to the lens (Fig. 5g).

The expression of stat3 was followed throughout theremainder of embryonic life and into the larval stages.At 2 days postfertilization (dpf), expression of stat3 wasconfined to the cranial sensory ganglia (Fig. 6a). Asubset of these structures, the trigeminal and theposterior lateral line ganglia, also expresses the cyto-plasmic kinase jak1 (Oates et al., 1999) at this stage(Fig. 6b). Over the next 2 days of development, expres-sion of stat3 expanded to include nuclei of the brain, theendoderm by 3 dpf (Fig. 6c), and neuromasts of thelateral line by 4 dpf (Fig. 6d). Close examination of thebrain at 3 dpf indicated that dorsally located cranialmotor nuclei of the midbrain (occulomotor) and hind-brain (trigeminal, abducens/facial, glossopharyngeal,and vagus) were expressing stat3 (Fig. 6e) as were twopaired nuclei of unknown identity in the forebrain (Fig.


6h). At a more ventral location in the brain, stat3expression persisted in the sensory cranial ganglia andcould be detected in the peripheral margin of the retina(Fig. 6f). The endoderm of 3 dpf zebrafish showed stat3expression, with the strongest signal from the liver andweaker expression in the gut (Fig. 6g). Also at this time,a diffuse signal was observed in the pharyngeal arches.By 4 dpf, stat3 expression was detected in the dorsalcells of the ventricular walls of the midbrain andhindbrain, forming a cross-shaped arrangement whenviewed dorsally (Fig. 6i). This pattern was also ob-served for the expression of the bHLH transcriptionfactor gene scl (Liao et al., 1998) (Fig. 6j). The retinal

ganglion cells of the eye expressed stat3 in patchymanner (Fig. 6k). Expression of stat3 in the elements ofthe lateral line was strong by 4 dpf (Fig. 6d, l),appearing as a ring of positive neuromast support cellssurrounding a nonexpressing core of differentiatinghair cells (Metcalfe et al., 1985). These support cellsalso express jak1 (data not shown). Expression of stat3in the pharyngeal arches increased and was main-tained in the liver and gut at 4 dpf (Fig. 6l). By 5 days,an epithelial ring of cells around the opening mouth ofthe larva was strongly positive for stat3 (Fig. 6m).

Thus, stat3 is expressed in a variety of tissues in thedeveloping zebrafish, most prominently in the anterior

Fig. 4. stat3 expression in the early zebrafish embryo. Zebrafishembryos collected from natural spawning and staged according toKimmel et al. (1995) were hybridized with a riboprobe to stat3 to localizethe sites of endogenous stat3 expression during blastula and gastrulastages (3–11 hpf). Embryos are shown in lateral view with animal pole tothe top and dorsal to the right of the panel unless otherwise noted. Scalebars are 250 µm for panels a,b,d–f and 100 µm for panels c, g, and h. a:1K stage showing ubiquitous zygotic stat3 transcript; arrowheads indicateblastoderm margin. b: 40% epiboly, ubiquitous expression, arrowheadsindicate blastoderm margin. c: 40% epiboly, higher magnification viewdemonstrating stat3 exclusion from enveloping layer (white arrowheads),black arrowhead indicates blastoderm margin. d: 90% epiboly. stat3

expression is restricted to the anterior-most mesendoerm of the pre-chordal plate (white arrowheads). Note open yolk plug; black arrowheadsindicate the gastrula margin. e: 100% epiboly, viewed as in (d) showingintensified signal from prechordal plate (white arrowheads) and closedyolk plug (black arrowhead). f: 100% epiboly, animal view, dorsal to thetop of panel, showing the stat3 expression domain. The anterior edge ofthe neural plate is marked with white arrowheads, and a black arrowheaddenotes the position of the dorsal axis. g: Bud stage, lateral view at highmagnification showing restriction of stat3 message to the anterior-mostmesendoderm. h: Bud stage, anterior view at high magnification showinglateral extent of stat3 expression in prechordal plate (white arrowheads)and in isolated cells of anterior neurectoderm (black arrowhead).

360 OATES ET AL.

Figure 5. (Legend on following page.)


hypoderm in late gastrulation, the sensory cranialganglia during the pharyngula stages, and the retina,lateral line, liver, and cranial motor nuclei duringlarval stages, suggesting that it may be involved in thetransduction of signals important for the developmentof these structures.

DISCUSSION

We have isolated cDNAs encoding the zebrafishhomologs of the transcription factors STAT1 and STAT3,genes required for cytokine signaling in mammals. Thehigh similarity of the sequences of these cDNAs to thoseof their mammalian homologs, combined with the abil-ity of zebrafish stat1 to rescue cytokine signaling in aSTAT1-deficient human cell line, indicates that theprocesses of cytokine signal transduction are likelyconserved between mammals and teleosts.

Structural and Functional RelationshipsBetween the STAT Genes

Functionally characterized regions with high iden-tity between zebrafish and mammalian STAT proteinsinclude the N-terminal 120 amino acid oligomerizationdomain, the DNA-binding domain, and the SH2 do-main. The otherwise highly variable C-terminal do-main has two islands of high identity positioned aroundthe activation tyrosine and the serine phosphorylationsite. Notably, the domain immediately N-terminal to

the SH2 domain and the region N-terminal to the DNAbinding domain are also highly conserved between fishand mammals. Recent crystallographic data indicatethat both regions play important roles in the relativeorientation and spacing of the N-terminal oligomeriza-tion, DNA binding, and SH2 domains (Chen et al., 1998;Becker et al., 1998). The variant stat3 cDNA possessinga short internal deletion is likely to represent analternately spliced mRNA. On the basis of experimentswith the mouse Stat3b, this alteration would be pre-dicted to abolish regulatory changes based on serinephosphorylation (Wen et al., 1995; Wen and Darnell,1997). Furthermore, it would be predicted to causeconstitutive activation in the absence of cytokines at somepromoters (Sasse et al., 1997; Schaefer et al., 1997) andact as a dominant negative repressor at others (Calden-hoven et al., 1996). The deletion of the C-terminus inthis manner would also confer cooperative transcrip-tional activity with c-Jun (Shaefer et al., 1995). Thus,primary sequence analysis suggests that the zebrafishstat cDNAs isolated in this study represent transcriptsfrom intact, functional STAT gene family members.

This conclusion is supported by the demonstrationthat the zebrafish stat1 cDNA will rescue the cytokine-signaling defect in human fibroblasts caused by the lossof endogenous STAT1 activity. The reconstitution ofIFN-g, IFN-a, and IL-6 signaling to the 6–16 promoterand the integration of the zebrafish stat1 protein intointerferon-mediated growth suppression processes inmammalian cells suggests that the variation in DNAbinding and activation tyrosine motifs found in thezebrafish stat1 protein are without functional conse-quence. Thus by inference, the zebrafish protein iscompetent to bind to human cytokine receptor cytoplas-mic domains, act as a substrate for human JAK pro-teins, and, after translocation to the nucleus, bind toSBEs in the human genome and mediate transcrip-tional activation. This high degree of conservationsuggests that the zebrafish may provide a reliablemodel for the developmental genetic analysis of cyto-kine signal transduction.

Synteny and Evolution of the STATFamily of Genes

In the human genome, each STAT gene is tightlylinked to a HOX cluster in a locus that contains two orthree tandemly arranged STAT genes (Copeland et al.,1995; Spring, 1997). The chromosomal position of ze-brafish stat3 on LG3 places this gene in a chromosomesegment that shares extended gene order with regionsof human chromosome 17 and mouse chromosome 11(Fig. 3e). In addition to stat3/STAT3/Stat3, the LG3 locihoxb5/HOXB5/Hoxb5, dlx8/DLX3, pyy/PYY, rara2b/RARA/Rara, and hbaa1/Hba appear to be orthologs inzebrafish, human, and mouse, respectively (Njolstad etal., 1988, 1990; Molven et al., 1993; Joore et al., 1994;Ellies et al., 1997). The inclusion of stat3 in this syntenygroup extends the region conserved between human,mouse, and zebrafish a further 15.7 cM distal from the

Fig. 5. (Figure on preceding page.) stat3 expression in the segmenta-tion and early pharyngula stage zebrafish embryo. Zebrafish embryoscollected from natural spawning and staged according to Kimmel et al.(1995) were hybridized with a riboprobe to stat3 to localize the sites ofendogenous stat3 expression during segmentation and early pharyngulastages (12–24 hpf). Embryos are shown dissected from their yolk, vieweddorsally in flat mount with anterior to the left of the panel, unless indicatedotherwise. Scale bars are 250 µm for panels a–d, and f, and 100 µm forpanels e and g. a: 3-somite stage, showing stat3 expression in theanterior neural plate, ventral tail bud (arrow), and notochord. Noteexclusion of stat3 from the adaxial cells, visible as a clear stripe adjacentto the notochord (arrowhead). b: 5-somite stage, stat3 expressionpersists as in (a), with elevated expression in the anterior neural keelcorresponding to the presumptive hindbrain (star) and in the newlyforming somites (bracket). c: 13-somite stage, showing elevated stat3transcript in forebrain and midbrain, including the optic primordia (thinarrow), whereas expression in the hindbrain has decreased. Stainingbecomes stronger in the anterior notochord (arrowhead) relative to theposterior, and expression in the tail bud (thick arrow) increases strongly.d: 16-somite stage, stat3 expression is present at high levels in the eyesand midbrain, tailbud (arrow), presomitic mesoderm and newly formedsomites (bracket). Small arrowheads mark the position of the mostrecently formed somite boundary. stat3 transcript is seen for the first timein cranial ganglia primordia (large arrowheads). e: 19-somite stage,lateral view of tail showing persistent expression of stat3 in ventralsomites (white arrowhead). The position of the most recently formedsomite boundary is indicated with arrowheads. f: 24 hpf, lateral view ofwhole mount embryo. stat3 expression is seen in the cranial ganglia(white arrowheads), dorsal midbrain (black arrowheads), eyes, and in arestricted cluster of neurons in the diencephalon between the eyes(arrow). g: 24 hpf, stat3 expression is detected in the midbrain (out of theplane of focus) and in eyes at elevated levels in the retina (whitearrowheads), diencephalon (arrow), and the cranial ganglia (from rostralto caudal: trigeminal, anterior lateral line/acoustic and posterior lateralline, arrowheads). The position of the otic vesicle is marked with a star.

362 OATES ET AL.

hoxba cluster. The high-sequence similarity and syn-tenic position of stat3 leads us to conclude that itrepresents a genuine ortholog of the known mamma-lian STAT3 genes. Furthermore, it indicates that thissyntenic relationship has persisted for more than 420million years, the estimated time since the divergenceof lobe-fin and ray-fin fishes (Ahlberg and Milner, 1994).In contrast, stat1 maps to LG22, where the only markerpresent that enables syntenic analysis is the morecentromeric scl gene (Liao et al., 1998). The presumedorthologs of these genes are not syntenic in mouse orhuman: Stat1 is located on Mmu1 and Tal1 on Mmu4 at49.5, and STAT1 is found at Hsa2q32-q34, whereasTAL1 is at Hsa1p32, suggesting that their chromo-somal linkage differs in zebrafish. We are currentlyinvestigating the possibility of additional stat1 ho-mologs in the zebrafish.

Comparison of early stat1 gene expression pat-terns in the Zebrafish and in the mammal. In situhybridization with Stat1 in early postimplantationmouse development demonstrates maternal expressionin endothelial cells of decidual vasculature, and de-cidual cells apposed to the myometrium (Duncan et al.,1997). Extraembryonic parietal endoderm and the ecto-placental cone express Stat1 message, whereas at E 9.5,the embryo proper has not initiated expression. Simi-larly, zebrafish stat1 cannot be detected in the embryoproper at the stages examined, but unlike the mammal,zebrafish stat1 is not seen in the extraembryonic tissueof the zebrafish, the yolk syncitial layer (YSL, seebelow). Stat1 expression has been detected in thepostnatal brain of mice (Planas et al., 1997a) and can bedetected in multiple adult tissues (Zhong et al., 1994b;Herrada and Wolgemuth, 1997; Illa et al. 1997), raisingthe possibility that expression of the zebrafish stat1gene may be seen at much later stages in zebrafish.

Comparison of early stat3 gene expression pat-terns in the zebrafish and in the mammal. Differ-ences are seen in STAT3 expression patterns betweenmouse and fish at early stages of development. Anessential functional role for Stat3 in the mouse isindicated by the failure of Stat3 mice to initiate gastru-lation (Takeda et al., 1997). Stat3-deficient mouseembryos produce an egg cylinder at E 6.0, which issmaller than wild-type siblings, but appears otherwisenormal. However, by E 6.5, the Stat3-deficient embryobegins to degenerate and does not produce any meso-derm at E 7.0, in contrast to wild-type siblings. Stat3transcripts are detected by in situ hybridization in themouse conceptus at E 6.0 in extraembryonic columnarvisceral endoderm (which forms the visceral wall of theyolk sac) (Takeda et al., 1997). Duncan and co-workers(1997) confirm this localization at E 7.5 of developmentwhen Stat3 is expressed in multiple extraembryonictissues. This expression data suggest that functionalStat3 is required in the extraembryonic tissues forsurvival of the conceptus but does not rule out anearlier role within the embryo proper. The primitiveendoderm of mammals may be homologous to the YSL

of the teleost; induction and dorsoventral specificationof mesoderm from the zebrafish blastoderm is a prop-erty of the YSL (Mizuno et al., 1996; Yamanka et al.,1998; Koos and Ho, 1998). However, we have not beenable to detect stat3 in the YSL, suggesting that thefunction of Stat3 in the extraembryonic tissues of themouse may be a mammalian specific adaptation.

A conserved role in early embryogenesis for STAT3may be found in the preservation of the potential forself-renewal of the earliest cells of the blastula. Inembryonic stem (ES) cell lines, derived from the innercell mass of the early mouse blastocyst, a functionalStat3 gene is required for the maintenance of pluripo-tency (Raz et al., 1999). In the absence of Stat3, or onoverexpression of dominant negative versions of thegene, an enhanced and premature differentiation ofthis cell type is seen in the absence of changes inproliferation (Boeuf et al., 1997; Niwa et al., 1998; Razet al., 1999). Early stat3 expression in the zebrafishmay play an analogous role; it is abundant in the cells ofthe blastula as they proliferate and determine theirposition within the axial coordinates of the embryo butis down-regulated at the onset of gastrulation when themesoderm is created and overt differentiation begins.In any case, a prediction of a potential functionalconservation for STAT3 in vertebrates is that introduc-tion of dominant negative forms of stat3 (Kaptein et al.,1996; Minami et al., 1996; Nakajima et al., 1996) or aloss of function stat3 mutation would perturb gastrula-tion in zebrafish. It will be interesting to compare themap positions of zebrafish zygotic mutants that do notgastrulate with that of stat3.

It is interesting to note the absence of both zebrafishstat genes from the extraembryonic tissues (YSL) of thefish. Because the role of mammalian Stat genes inextraembryonic tissue function is unknown, we cannotdetermine if this function is provided in fish by the YSL.If this were the case, the differences detected by thisstudy may reflect evolutionary modification of the rolesplayed by STAT genes in different vertebrates. Alterna-tively, the early role of mammalian Stat genes mayconcern a novel function of the extraembryonic tissuesrelated to the processes of implantation and maternalinteraction, and no stat expression in the YSL of fishwould be expected.

At E 9.5 of mouse development, embryonic expres-sion of Stat3 is found in the endothelia of the vascula-ture (dorsal aorta and umbilical vein), nucleated bloodcells, myocardium, and endocardium of the atria andventricles of the heart (Duncan et al., 1997). However,zebrafish stat3 is not detected in these tissues at anequivalent stage of development (24 hpf) or thereafterat stages we have examined, suggesting that a role forStat3 in blood and vascular development is a specificfunction of the mammalian gene. Alternatively, theremay exist additional stat3 homologs in the zebrafishthat have retained a blood-specific expression pattern.In this regard, it is noted that there are two JAK2homologs in the zebrafish, only one of which is ex-


Figure 6.

364 OATES ET AL.

pressed in erythroblasts (Oates et al., 1999). stat3expression in head mesenchyme of the pharyngealarches is not evident in the zebrafish till 3 dpc; thiscontrasts with an early (E 9.5) Stat3 expression at highlevel in head mesenchyme in the mouse embryo; how-ever, in the mouse, the expression does not appear to berestricted to the pharyngeal arches (Duncan et al.,1997). In contrast to the differences noted above, theexpression of stat3 in the PNS and CNS of zebrafish is afeature shared by mammals; activated stat3 proteincan be detected in the developing mammalian brain bygel shift assay and Western blot (Planas et al., 1997b;De-Fraja et al., 1998; Potts et al., 1998), and functionsfor Stat proteins in mammalian neurons are welldocumented (see below).

Potential Functions of JAK and STAT Genesin the Peripheral and Central Nervous Systemof Vertebrates

The late expression of stat3 along with jak1 in cells ofthe sensory cranial ganglia relative to markers ofneural crest identity (dlx2, Akimenko et al., 1994),neural commitment (isl1, Inoue et al., 1994) and differ-entiation (L1, Tongiorgi et al., 1995) raises a questionas to the nature of the signal that may be transduced,because it argues against those that impart specifica-tion or differentiation. Rather, signals that direct axon

guidance or neuronal survival may be better candi-dates. Members of the Trk receptor family transducedistinct survival signals in response to the binding ofmember of the nerve growth factor (NGF) family ofgrowth factors in the cranial ganglia of mice (Klein etal., 1993; Crowley et al., 1994; Jones et al., 1994;Smeyne et al., 1994; Pinon et al., 1996), and thezebrafish trkC1 gene is expressed in the cranial gangliawith a similar developmental timing to the stat genesreported here (Martin et al., 1998). NGF signaling inPC12 cells causes the rapid serine phosphorylation ofStat3, and in combination with IL-6, the formation ofDNA-binding complexes containing Stat1 and Stat3(Wu and Bradshaw, 1996). We note that the NGF-related signaling molecule BDNF is expressed in ze-brafish in the central hair cells of the lateral lineneuromasts (Hashimoto and Heinrich, 1997), which aresurrounded by stat3-expressing support cells. This sug-gests that BDNF released from the hair cells may bind tothe support cells and that stat3 in these cells may trans-duce this signal. Given the requirement for STAT signalingin mediating the effects of signaling from other RTKs, suchas the HGFR (Boccaccio et al., 1998), it is tempting tosuggest that STAT1 and/or STAT3 function may be con-nected to the effectiveness of Trk- mediated signals in thecranial ganglia and lateral line of vertebrates.

Examination of Stat3 expression and activation inthe developing cerebellum of the postnatal rat (Planaset al., 1997b) indicates that neuronal expression in theCNS is a feature of both fish and mammalian STAT3genes. One role for STAT1 and STAT3 in the develop-ment of the CNS is found in the transduction of theastrocyte differentiation signal CNTF, which signalsvia a cytokine-specific subunit, the CNTFR-a, and theJAK/STAT-activating gp130 and LIFR-b signalingchains. CNTFR-a-deficient mice display severe motorneuron defects (DeChiara et al., 1995). In rat corticaland spinal neuron precursor cells, Stat1 and Stat3 arerapidly phosphorylated in response to CNTF and bindto a functionally critical SBE in the promoter of theastrocyte specific GFAP gene; furthermore, a dominantnegative Stat3 isoform suppresses the CNTF differen-tiation activity (Bonni et al., 1997). In the zebrafish, theJAK/STAT pathway may play an analogous role in CNSdevelopment, and the expression of stat1 and stat3 (aswell as jak1 and jak2a, Oates et al., 1999) in regions ofthe developing zebrafish CNS is consistent with thisnotion. In this study, we demonstrate the expression ofstat3 in cranial motor nuclei in the midbrain andhindbrain from day 3 onward. The formation of thecranial motor nuclei is marked by isl1 (Inoue et al.,1994; Chandrasekhar et al., 1997) expression throughthe first and second day of development, but to date,there has been no marker of later stages of developmentof these structures. Again, the timing of stat3 expres-sion is consistent with a role in survival or maintenanceof the differentiated state of these cells. Coexpression ofstat3 with four distinct Trk receptors (Martin et al.,1995) in the retinal ganglion cell layer of the developing

Fig. 6. stat3 expression in the developing zebrafish larva. Zebrafishembryos collected from natural spawning and staged according toKimmel et al. (1995) were hybridized with a riboprobe to stat3 to localizethe sites of endogenous stat3 expression during larval stages (2–5 dpf).Embryos are shown dissected from their yolk, viewed dorsally in flatmount with anterior to the top of the panel, except c, d which are lateralviews and i, which is viewed ventrally. Scale bars are 100 µm for panels a,b, and e–m, and 250 µm for panels c and d. a: 2 dpf head, showing stat3expression in sensory cranial ganglia: tr, trigeminal; all, anterior lateralline; ac, acoustic; pll, posterior lateral line. The position of the ear ismarked with a star. b: 2 dpf head, showing jak1 expression in trigeminaland posterior lateral line sensory cranial ganglia (arrows) and ear (star). c:3 dpf, lateral view of head. Dashed lines indicate location of stat3expressing features displayed at high magnification in e–h, below. d: 4dpf, lateral view of head. Dashed lines indicate location of stat3 express-ing features displayed at high magnification in i–k, below. Neuromasts ofthe lateral line system are indicated with arrowheads. e: 3 dpf head, DICoptical section of dorsal brain showing stat3 expression in the cranialmotorneuron nuclei. III, occulomotor; V, trigeminal; ab/VII, abducens andfacial; IX, glossopharyngeal; X, vagus. Dashed outline indicates relativeposition of (h). f: 3 dpf head, DIC optical section, ventral to (e), showingstat3 expression in the sensory cranial ganglia denoted as in (a).Expression is also seen in the marginal zone of the retina (arrowheads).g: 3 dpf, DIC optical section through ventral head showing expression inthe liver (arrowheads) gut (arrow) and pharyngeal arches (small arrows).h: 3 dpf, DIC optical section intermediate between (e) and (f) in headshowing additional neuronal nuclei of unknown identity in forebrain. i: 4dpf head, dorsal optical section showing stat3 expression in cells liningthe ventricles (arrowheads). j: 3 dpf, same section of head as (i), showingscl expression in the walls of the ventricles. k: 4 dpf, DIC optical sectionthrough center of eye, showing stat3 expression in the retinal ganglion celllayer (arrowheads). l: 4 dpf, ventral view of head, showing stat3 expres-sion in the liver (arrowheads), pharyngeal arches (small arrows) andneuromasts of the anterior lateral line (larger arrows). m: 5 dpf, DICoptical section at the level of the opening of the mouth, showing stat3expression in a ring of epithelial cells around the opening.


retina suggests the interaction of these signaling compo-nents within the CNS, in addition to their proposed rolein the PNS described above.

Thus, comparison of early embryonic expression pat-terns between the mouse and zebrafish STAT genesindicates that the temporal and spatial regulation oftranscription is not highly conserved. STAT1 is ex-pressed in neither animal in the embryo proper duringearly development, and stat3 expression appears to beconserved only in the CNS of both animals. However,these comparisons suffer from a paucity of data fromthe mouse beyond E 13.5; further similarities maybecome known as the expression patterns of mamma-lian STAT1 and STAT3 genes are better characterized.

In summary, this work has demonstrated the exis-tence and high structural and functional conservationof members of the STAT gene family, essential compo-nents of cytokine signal transduction pathways, inteleost fish. The results presented here strongly sug-gest that the biochemical mechanism of cytokine signal-ing is conserved across vertebrata. Furthermore, thedata indicate that genes of these families are deployedduring embryogenesis in a dynamic and tissue-restricted manner, suggesting that they may partici-pate in multiple distinct ontogenetic signaling events.This sharp spatiotemporal restriction of expression ofwhat were thought to be ubiquitous intracellular signal-ing components suggests that differential expression ofsignal transduction componentry is a mechanismwhereby cytokine signaling may be regulated duringdevelopment. The genes of the stat family have beengenetically mapped, demonstrating that a member ofthe stat family appears to be situated in a paralogouschromosomal segment, the long-range structure of whichhas been conserved for more than 400 million years. Inaddition, mapping allows rapid assessment of whetherone or more of the hundreds of mutant zebrafish linesmay harbor a mutation in a stat gene. It is importantbecauses this study has opened the possibility foranalysis of cytokine signaling in the zebrafish, a modelorganism that allows genetic and cell biological ap-proaches to be married in the investigation of complexgenetic and physiological mechanisms.

EXPERIMENTAL PROCEDURESIsolation of Zebrafish stat Homologs

STAT-directed degenerate oligonucleotide primerswere designed after Wilks (1989), based around regionsof sequence similarity defined by aligning H. sapiensSTAT1 and STAT2 proteins (Fu et al., 1992). For theamplification of STAT sequences, a nested protocolusing mixed stage cDNA as template was used. Thefirst reaction used a primer targeting the amino acidmotif QPCMP, found at position 307 in STAT1, incombination with the reverse complement primer tar-geting the motif LWNDG, found at position 561. Theproduct of this reaction was used as template in anested PCR with a primer targeted to the motif GFRKF,found at position 367, in combination with the LWNDG

primer described above. The sequence of the primers isdescribed below: QPCMP, GGGAATTCCA(A/G)CC(G/A/T/C)TG(T/C)ATGCC; GFRKF, GGGAATTCT(T/C)(A/C)G(G/A/T/C)AA(A/G)TT(T/C)AA; LWNDG, CCGAATTC(G/A/T/C)CC(A/G)TC(A/G)TTCCA(G/A/T/C)A. cDNAlibrarieswereplated and screened at high stringency according tostandard methods (Sambrook et al., 1989) with PCRproducts generated by STAT-directed degenerate oligo-nucleotide PCR as above and specific for the zebrafishstat1 and stat3 genes. Mixed developmental stage D.rerio cDNA libraries in Lambda Zap (Stratagene, LaJolla, CA), both random and poly-A primed, were a giftof J. Campos-Ortega (University of Koln). Lambda ZapcDNAlibraries generated from staged embryonic mRNApopulations were made by Bob Riggleman and KathrynHelde and were a kind gift of D. Grunwald (EcclesInstitute, University of Utah). The ‘‘Contig Manager’’application of the DNAstar suite of programs was usedto create and monitor contigs from the primary se-quence data. cDNA sequences presented herein corre-sponding to each of the gene transcripts under studywere sequenced in both directions to a minimum oftwofold coverage.

Zebrafish Care and Strains

Zebrafish were raised and maintained as described(Westerfield, 1995) by using wild-type fish purchasedfrom St. Kilda Aquarium (Melbourne, Australia). Theb245 gamma ray deletion allele was described in Fritzet al. (1996).

Sequence Analysis and EvolutionaryComparison

Electronic database searches were made by submit-ting nucleic acid sequence and putative amino acidsequence to the public search facility at the BaylorCollege of Medicine (http://kiwi.imgen.bcm.tmc.edu) us-ing ‘‘WU-BLAST’’ (Altschul and Gish, 1996). To studythe evolutionary relationships between the identifiedSTATs, the deduced amino acid sequences of the genesin question were aligned with the ‘‘CLUSTAL’’ proteinalignment program (Higgins et al., 1996) of theMegAlign application (DNAstar suite) and refined byhand using information from crystal structures whereavailable (Vinkemeier et al., 1998; Chen et al., 1998;Becker et al., 1998), and functional studies (Wen et al.,1995). These alignments were used to create maximumparsimony phylogenetic trees and distance matricesusing the options of that program. The topology of thephylogenetic trees shown was insensitive to the order ofsequence addition.

Wild-Type and Mutated stat1 Variants

The stat1 cDNA was excised from a full-length plas-mid (3705pA-16.1) with EcoRI and ligated into pcDNA3(Invitrogen, CA). stat1 DPIC and stat1 TGA weregenerated from 3705pA-16.1 by restricting with TthIand NcoI, blunting the overhangs with T4 DNApolymer-ase and religating with T4 DNA ligase. Sequencing

366 OATES ET AL.

confirmed that stat1 DPIC has a 22 amino acid in-framedeletion between residue 23 and 45, and stat1 TGAresulted from an incomplete blunting reaction thatintroduces an in-frame stop codon after 28 amino acidsof the native stat1 sequence. Mouse Stat1 in pcDNA3was a kind gift of Dr. S. E. Nicholson (Walter and ElizaHall Institute, Melbourne).

Cell Culture, DNA Transfection, and SelectionRegimens

Parental 2fTGH and mutant U3A cells lacking STAT1(Pellegrini et al., 1989; McKendry et al., 1991; Muller etal., 1993) were maintained in Dulbecco’s modified Ea-gle’s medium (DMEM) with 10% (v/v) FCS and 5 µML-glutamine. Hygromycin (Sigma, St. Louis, MO) wasincluded at 250 µg/mL for the stabilization of the gptvector in the U3A cells. DNA transfection was per-formed by using either the calcium phosphate method(Velazquez et al., 1992) with 106 cells per 10-cm dishand 20 µg of DNA per dish (4-µg expression vector DNAand 16-µg carrier vector pEF BOS) or by using theDOTAP Liposomal transfection kit (Boehringer Mann-heim) as described by the manufacturer with 5 3 105

cells per 10-cm dish, and 10-µg expression vector DNA.Following transfection, cells were grown for 3 days,split 1:10 and selected in 500 µg/mL G418-containingmedia (Gibco BRL) for another 7–14 days.

Stably transfected cell lines were assayed for respon-siveness to cytokine signaling by plating in negativeselection media (HAT: 20 µg hypoxanthine/mL, 0.2 µgaminopterin/mL, 20 µg thymidine/mL in DMEM) orpositive selection media (6TG 30 µg 6-thioguanine) inthe presence of IFN-a? 600 IU/mL Refron A (Roche Pty.Ltd. New South Wales, Australia), or IFN-g? 15 ng/mLImukin (Boehringer Ingelheim Pty. Ltd., New SouthWales, Australia), or 100 nM IL-6 plus 200 pM solubleIL-6R (kind gift of Dr. R. Simpson, Ludwig Institute forCancer Research, Mebourne, Australia). Representa-tive cell fields were photographed under phase contrastusing a Nikon-inverted microscope after 2 weeks ofselection. Growth curves were measured by plating 3 3104 cells per 3-cm dish and counting viable cell numberdaily (in triplicate) for the following 8 days. DNAsynthesis was measured by plating 1 3 105 cells perwell and growing cells to 20% confluency. The cells wereserum starved for 20 hr and then selected in thepresence of cytokines at various doses as describedabove. After 12 hr, cells were pulsed with 3H-thymidine(1 µCi/mL) for 2 hr and harvested by using a CellHarvester 96 (Tomtec). For each cell line the number ofcounts were normalized to the maximal incorporationfor the cell line and displayed as percent incorporation.

Generation of DNA Polymorphisms and GeneticMapping

A bacterial artificial chromosome (BAC) library (Ge-nome Systems, St. Louis, MO) containing large insertzebrafish genomic DNA was screened by hybridizationto oligonucleotide probes derived from stat1 and stat3

cDNA. The presence of the target genes was confirmedwith PCR by using independent coding region-derivedprimers. Clones corresponding to the genomic loci wereobtained for stat1 (46J1, 43J3, and 93N19) and stat3(82H8, 138C20, and 155 K12). Sequence informationfrom the ends of each of these genomic clones wasdetermined (data not shown) and used to design PCRprimer pairs that amplified products from genomic DNA,which segregated in a C32xSJD mapping cross (Johnson etal., 1996). The primers for stat1 (93 T7–1 CTCTCATGTCT-GCCATTATG, 93 T7–3 CGCACTTTCAAACATTGGGCTG,derived from BAC 93 N19) give a recessive sequencepolymorphism (Orita et al., 1989); the stat3 codominantsize polymorphism is from the primer pair 138 SP6–1TTTTGTCCCTCGTCTTTCAG, 138 SP6–2 AGACT-GATCTGGCTCCATAG, derived from BAC 138 C20.Linkage was determined by comparison of a givenmarker on a C32xSJD haploid panel consisting of 96individuals that were genotyped against approximately600 markers (Goff et al., 1992; Johnson et al., 1996;Knapik et al., 1996) for close correlation of segregationpatterns using the program ‘‘MapMaker’’ (Lander et al.,1987) and the program ‘‘mapmanager’’ (Manly et al.,1996). Linkage between the stat3 gene and the b245deletion (Fritz et al., 1996) was determined by PCR(Knapik et al., 1996) using the following primer pairs:stat3 gene, s3.10 TATAGTGAGCTTAGGTGC, and s3.14TGGGCGGATTCTGATCAC; cdc27 gene, cdc27F1 GATTT-GCCAATGACTCCACCGTC and cdc27F2 GGCTAATT-GACTTACTTAACTCC, cdc27R1 TTAGCAGATGAATCT-TATGCC and cdc27R2 TAACAGATGGCAATCTGAAAG;a-globin locus: a-globinF CTGCCAAAGACAAAGCTGC-CGTCA and a-globinR AGCATGGAGCTCACTGAG-GTTCAG.

Whole-Mount In Situ Hybridization

Embryos were staged according to Kimmel and co-workers (1995). Embryos raised to time points beyond24 hr postfertilization (hpf) were transferred to E3embryo medium with 0.003% phenylthiourea (PTU,Sigma) to prevent melanization. Riboprobe synthesisand in situ hybridization were conducted essentially asdescribed (Shulte-Merker et al., 1992) with modifica-tions as in Oates et al. (1999). Furthermore, an in-creased stringency was afforded by hybridization andwashes with 0.053 SSC at 70°C. The stat1 riboprobewas generated from the full-length cDNA by restrictingwith Not I and then transcribing with T7 RNA polymer-ase in vitro. The stat3 riboprobe was generated byrestricting the full-length stat3 cDNA with Acc I andreligating, to remove the two internal Acc I fragments.The resulting plasmid, containing the 58 most 1.5 kbencoding the ATPID and coiled-coil domain, was re-stricted with Bam HI and transcribed in vitro with T7RNA polymerase. The color reaction was stopped at thesame time for all stages up to and including 1 dpf, after6–8 hr of development. Larvae up to 5 dpf weresimilarly treated, being developed for approximately2days. Sense riboprobes from stat1 and stat3 plasmids


did not produce any signal over these times (data notshown).

ACKNOWLEDGMENTS

The authors thank Jana Strickland for her invalu-able help with the figures. We also thank Cuong Do andthe members of the Growth Regulation and CytokineBiology Laboratories for many discussions;Ashley Bruceand This Ernst who made critical comments on themanuscript; and members of the Zon lab fish collectivefor their support and encouragement. A.C.O. was sup-ported by an Anti Cancer Council of Victoria Postgradu-ate Research Scholarship. P.W. was supported by fellow-ships from the Swedish Natural Science ResearchCouncil, the Swedish Institute, and the Wenner-GrenCenter Foundation for Scientific Research. B.H.P. wassupported in part by a H.H.M.I. Postdoctoral Fellow-ship for Physicians. J.H.P. was supported by grantsRO1RR10715 and PHS PO1HD22486. R.K.H. is sup-ported by NIH R01 HD34499, NSF IBN 9808351, andRita Allen Foundation grants.

REFERENCESAhlberg PE, Milner AR. 1994. The origin and diversification of

tetrapods. Nature 368:507–511.Akimenko MA, Ekker M, Wegner J, Lin W, Westerfield M. 1994.

Combinatorial expression of three zebrafish genes related to distal-less: part of a homeobox gene code for the head. J Neurosci14:3475–3486.

Akira S, Nishio Y, Inoue M, Wang XJ, Wei S, Matsusaka T, Yoshida K,Sudo T, Naruto M, Kishimoto T. 1994. Molecular cloning of APRF, anovel IFN-stimulated gene factor 3 p91-related transcription factorinvolved in the gp130-mediated signaling pathway. Cell 77:63–71.

Altschul SF, Gish W. 1996. Local alignment statistics. MethodsEnzymol 266:460–471.

Araki T, Gamper M, Early A, Fukuzawa M, Abe T, Kawata T, Kim E,Firtel RA, Williams JG. 1998. Developmentally and spatially regu-lated activation of a Dictyostelium STAT protein by a serpentinereceptor. EMBO J 17:4018–4028.

Becker S, Groner B, Muller CW. 1998. Three-dimensional structure ofthe Stat3beta homodimer bound to DNA. Nature 394:145–151.

Binari R, Perrimon N. 1994. Stripe-specific regulation of pair-rulegenes by hopscotch, a putative Jak family tyrosine kinase inDrosophila. Genes Dev 8:300–312.

Boccaccio C, Ando M, Tamagnone L, Bardelli A, Michieli P, BattistiniC, Comoglio PM. 1998. Induction of epithelial tubules by growthfactor HGF depends on the STAT pathway. Nature 391:285–288.

Boeuf H, Hauss C, Graeve FD, Baran N, Kedinger C. 1997. Leukemiainhibitory factor-dependent transcriptional activation in embryonicstem cells. J Cell Biol 138:1207–1217.

Bonni A, Sun Y, Nadal-Vicens M, Bhatt A, Frank DA, Rozovsky I, StahlN, Yancopoulos GD, Greenberg ME. 1997. Regulation of gliogenesisin the central nervous system by the JAK-STAT signaling pathway.Science 278:477–483.

Bromberg JF, Horvath CM, Wen Z, Schreiber RD, Darnell JE Jr. 1996.Transcriptionally active Stat1 is required for the antiproliferativeeffects of both interferon alpha and interferon gamma. Proc NatlAcad Sci USA 93:7673–7678.

Caldenhoven E, van Dijk TB, Solari R, Armstrong J, RaaijmakersJAM, Lammers JWJ, Koenderman L, de Groot RP. 1996. STAT3beta,a splice variant of transcription factor STAT3, is a dominantnegative regulator of transcription. J Biol Chem 271:13221–13227.

Chandrasekhar A, Moens CB, Warren JT Jr, Kimmel CB, Kuwada JY.1997. Development of branchiomotor neurons in zebrafish. Develop-ment 124:2633–2644.

Chen X, Vinkemeier U, Zhao Y, Jeruzalmi D, Darnell JE Jr, Kuriyan J.

1998. Crystal structure of a tyrosine phosphorylated STAT-1 dimerbound to DNA. Cell 93:827–839.

Copeland NG, Gilbert DJ, Schindler C, Zhong Z, Wen Z, Darnell JE Jr,Mui AL, Miyajima A, Quelle FW, Ihle JN, et al. 1995. Distribution ofthe mammalian Stat gene family in mouse chromosomes. Genomics29:225–228.

Crowley C, Spencer SD, Nishimura MC, Chen KS, Pitts-Meek S,Armanini MP, Ling LH, MacMahon SB, Shelton DL, Levinson AD,et al. 1994. Mice lacking nerve growth factor display perinatal loss ofsensory and sympathetic neurons yet develop basal forebrain cholin-ergic neurons. Cell 76:1001–1011.

Darnell JE Jr. 1997. STATs and gene regulation. Science 277:1630–1635.

Darnell JE Jr, Kerr IM, Stark GR. 1994. Jak-STAT pathways andtranscriptional activation in response to IFNs and other extracellu-lar signaling proteins. Science 264:1415–1421.

DeChiara TM, Vejsada R, Poueymirou WT, Acheson A, Suri C, ConoverJC, Friedman B, McClain J, Pan L, Stahl N, et al. 1995. Mice lackingthe CNTF receptor, unlike mice lacking CNTF, exhibit profoundmotor neuron deficits at birth. Cell 83:313–322.

De-Fraja C, Conti L, Magrassi L, Govoni S, Cattaneo E. 1998.Members of the JAK/STAT proteins are expressed and regulatedduring development in the mammalian forebrain. J Neurosci Res54:320–330.

Duncan SA, Zhong Z, Wen Z, Darnell JE Jr. 1997. STAT signaling isactive during early mammalian development. Dev Dyn 208:190–198.

Durbin JE, Hackenmiller R, Simon MC, Levy DE. 1996. Targeteddisruption of the mouse Stat1 gene results in compromised innateimmunity to viral disease. Cell 84:443–450.

Ellies DL, Stock DW, Hatch G, Giroux G, Weiss KM, Ekker M. 1997.Relationship between the genomic organization and the overlappingembryonic expression patterns of the zebrafish dlx genes. Genomics45:580–590.

Fritz A, Rozowski M, Walker C, Westerfield M. 1996. Identification ofselected gamma-ray induced deficiencies in zebrafish using multi-plex polymerase chain reaction. Genetics 144:1735–1745.

Fu XY, Schindler C, Improta T, Aebersold R, Darnell JE Jr. 1992. Theproteins of ISGF-3, the interferon alpha-induced transcriptionalactivator, define a gene family involved in signal transduction. ProcNatl Acad Sci USA 89:7840–7843.

Fukada T, Hibi M, Yamanaka Y, Takahashi-Tezuka M, Fujitani Y,Yamaguchi T, Nakajima K, Hirano T. 1996. Two signals are neces-sary for cell proliferation induced by a cytokine receptor gp130:involvement of STAT3 in anti-apoptosis. Immunity 5:449–460.

Garcia R, Jove R. 1998. Activation of STAT transcription factors inoncogenic tyrosine kinase signaling. J Biomed Sci 5:79–85.

Goff DJ, Galvin K, Katz H, Westerfield M, Lander ES, Tabin CJ. 1992.Identification of polymorphic simple sequence repeats in the genomeof the zebrafish. Genomics 14:200–202.

GreenlundAC, Morales MO, Viviano BL, Yan H, Krolewski J, SchreiberRD. 1995. Stat recruitment by tyrosine-phosphorylated cytokinereceptors: an ordered reversible affinity-driven process. Immunity2:677–687.

Guillet-Deniau I, Burnol AF, Girard J. 1997. Identification andlocalization of a skeletal muscle secrotonin 5-HT2A receptor coupledto the Jak/STAT pathway. J Biol Chem 272:14825–14829.

Hashimoto M, Heinrich G. 1997. Brain-derived neurotrophic factorgene expression in the developing zebrafish. Int J Dev Neurosci15:983–997.

Herrada G, Wolgemuth DJ. 1997. The mouse transcription factorStat4 is expressed in haploid male germ cells and is present in theperinuclear theca of spermatozoa. J Cell Sci 110:1543–1553.

Higgins DG, Thompson JD, Gibson TJ. 1996. Using CLUSTAL formultiple sequence alignments. Methods Enzymol 266:383–402.

Holder N, Hill J. 1991. Retinoic acid modifies development of themidbrain-hindbrain border and affects cranial ganglion formation inzebrafish embryos. Development 113:1159–1170.

Horvai AE, Xu L, Korzus E, Brard G, Kalafus D, Mullen TM, Rose DW,Rosenfeld MG, Glass CK. 1997. Nuclear integration of JAK/STATand Ras/AP-1 signaling by CBP and p300. Proc Natl Acad Sci USA94:1074–1079.

368 OATES ET AL.

Horvath CM, Wen Z, Darnell JE Jr. 1995. A STAT protein domain thatdetermines DNA sequence recognition suggests a novel DNA-binding domain. Genes Dev 9:984–994.

Hou J, Schindler U, Henzel WJ, Ho TC, Brasseur M, McKnight SL.1994. An interleukin-4-induced transcription factor: IL-4 Stat.Science 265:1701–1706.

Hou J, Schindler U, Henzel WJ, Wong SC, McKnight SL. 1995.Identification and purification of human Stat proteins activated inresponse to interleukin-2. Immunity 2:321–329.

Hou XS, Melnick MB, Perrimon N. 1996. Marelle acts downstream ofthe Drosophila HOP/JAK kinase and encodes a protein similar tothe mammalian STATs. Cell 84:411–419.

Illa I, Gallardo E, Gimeno R, Serrano C, Ferrer I, Juarez C. 1997.Signal transducer and activator of transcription 1 in human muscle:implications in inflammatory myopathies. Am J Pathol 151:81–88.

Improta T, Schindler C, Horvath CM, Kerr IM, Stark GR, Darnell JEJr. 1994. Transcription factor ISGF-3 formation requires phosphory-lated Stat91 protein, but Stat113 protein is phosphorylated indepen-dently of Stat91 protein. Proc Natl Acad Sci USA 91:4776–4780.

Inoue A, Takahashi M, Hatta K, Hotta Y, Okamoto H. 1994. Develop-mental regulation of islet-1 mRNA expression during neuronaldifferentiation in embryonic zebrafish. Dev Dyn 199:1–11.

Johnson SL, Gates MA, Johnson M, Talbot WS, Horne S, Baik K, RudeS, Wong JR, Postlethwait JH. 1996. Centromere-linkage analysisand consolidation of the zebrafish geneticmap. Genetics 142:1277–1288.

Jones KR, Farinas I, Backus C, Reichardt LF. 1994. Targeted disrup-tion of the BDNF gene perturbs brain and sensory neuron develop-ment but not motor neuron development. Cell 76:989–999.

Joore J, van der Lans GB, Lanser PH, Vervaart JM, Zivkovic D,Speksnijder JE, Kruijer W. 1994. Effects of retinoic acid on theexpression of retinoic acid receptors during zebrafish embryogen-esis. Mech Dev 46:137–150.

Kaptein A, Paillard V, Saunders M. 1996. Dominant negative stat3mutant inhibits interleukin-6-induced Jak-STAT signal transduc-tion. J Biol Chem 271:5961–5964.

Kawata T, Shevchenko A, Fukuzawa M, Jermyn KA, Totty NF,Zhukovskaya NV, Sterling AE, Mann M, Williams JG. 1997. SH2signaling in a lower eukaryote: a STAT protein that regulates stalkcell differentiation in dictyostelium. Cell 89:909–916.

Kimmel CB, Ballard WW, Kimmel SR, Ullmann B, Schilling TF. 1995.Stages of embryonic development of the zebrafish. Dev Dyn 203:253–310.

Klein R, Smeyne RJ, Wurst W, Long LK, Auerbach BA, Joyner AL,Barbacid M. 1993. Targeted disruption of the trkB neurotrophinreceptor gene results in nervous system lesions and neonatal death.Cell 75:113–122.

Knapik EW, Goodman A, Atkinson OS, Roberts CT, Shiozawa M, SimCU, Weksler-Zangen S, Trolliet MR, Futrell C, Innes BA, et al. Areference cross DNA panel for zebrafish (Danio rerio) anchored withsimple sequence length polymorphisms. Development 123:451–460.

Koos DS, Ho RK. 1998. The nieuwkoid gene characterizes andmediates a Nieuwkoop-center-like activity in the zebrafish. CurrBiol 8:1199–1206.

Korzus E, Torchia J, Rose DW, Xu L, Kurokawa R, McInerney EM,Mullen TM, Glass CK, Rosenfeld MG. 1998. Transcription factor-specific requirements for coactivators and their acetyltransferasefunctions. Science 279:703–707.

Kurokawa R, Kalafus D, Ogliastro MH, Kioussi C, Xu L, Torchia J,Rosenfeld MG, Glass CK. 1998.Differential use of CREB bindingprotein-coactivator complexes. Science 279:700–703.

Lander ES, Green P, Abrahamson J, Barlow A, Daly MJ, Lincoln SE,Newburg L. 1987. MAPMAKER: an interactive computer packagefor constructing primary genetic linkage maps of experimental andnatural populations. Genomics 1:174–178.

Liao EC, Paw BH, Oates AC, Pratt SJ, Postlethwait JH, Zon LI. 1998.SCL/Tal-1 transcription factor acts downstream of cloche to specifyhematopoietic and vascular progenitors in zebrafish. Genes Dev12:621–626.

Manly K, Cudmore R, Kohler G. at mcbio.med.buffalo.edu/map-mgr.html, Roswell Park Cancer Institute, 1996.

Marrero MB, Schieffer B, Paxton WG, Heerdt L, Berk BC, Delafon-

taine P, Bernstein KE. 1995. Direct stimulation of Jak/STAT path-way by the angiotensin II AT1 receptor. Nature 375:247–250.

Martin SC, Marazzi G, Sandell JH, Heinrich G. 1995. Five Trkreceptors in the zebrafish. Dev Biol 169:745–758.

Martin SC, Sandell JH, Heinrich G. 1998. Zebrafish TrkC1 and TrkC2receptors define two different cell populations in the nervous systemduring the period of axonogenesis. Dev Biol 15;195:114–130.

McKendry R, John J, Flavell D, Muller M, Kerr IM, Stark GR. 1991.High-frequency mutagenesis of human cells and characterization ofa mutant unresponsive to both a and g interferons. Proc Natl AcadSci USA 88:11455–11459.

Metcalfe WK, Kimmel CB, Schabtach E. 1985. Anatomy of theposterior lateral line system in young larvae of the zebrafish. JComp Neurol 223:377–389.

Metcalfe WK, Myers PZ, Trevarrow B, Bass MB, Kimmel CB. 1990.Primary neurons that express the L2/HNK-1 carbohydrate duringearly development in the zebrafish. Development 110:491–504.

Meraz MA, White JM, Sheehan KC, Bach EA, Rodig SJ, Dighe AS,Kaplan DH, Riley JK, Greenlund AC, Campbell D, et al. Targeteddisruption of the Stat1 gene in mice reveals unexpected physiologicspecificity in the JAK-STAT signaling pathway. Cell 84:431–442.

Minami M, Inoue M, Wei S, Takeda K, Matsumoto M, Kishimoto T,Akira S. 1996. STAT3 activation is a critical step in gp130-mediatedterminal differentiation and growth arrest of a myeloid cell line.Proc Natl Acad Sci USA 93:3963–3966.

Mizuno T, Yamaha E, Wakahara M, Kuroiwa A, Takeda H. 1996.Mesoderm induction in zebrafish. Nature 383:131–132.

Molven A, Hordvik I, Njolstad PR. 1993. Sequence analysis of thezebrafish hox-B5/B6 region. Biochim Biophys Acta Apr 29;1173(1):102–106.

Muller M, Laxton C, Briscoe J, Schindler C, Improta T, Darnell JE Jr,Stark GR, Kerr IM. 1993. Complementation of a mutant cell line:central role of the 91 kDa polypeptide of ISGF3 in the interferon-alpha and -gamma signal transduction pathways. EMBO J 12:4221–4228.

Nakajima K, Yamanaka Y, Nakae K, Kojima H, Ichiba M, Kiuchi N,Kitaoka T, Fukada T, Hibi M, Hirano T. 1996. A central role for Stat3in IL-6-induced regulation of growth and differentiation in M1leukemia cells. EMBO J 15:3651–3658.

Niwa H, Burdon T, Chambers I, Smith A. 1998. Self-renewal ofpluripotent embryonic stem cells is mediated via activation ofSTAT3. Genes Dev 12:2048–2060.

Njolstad PR, Molven A, Hordvik I, Apold J, Fjose A. 1988. Primarystructure, developmentally regulated expression and potential dupli-cation of the zebrafish homeobox gene ZF-21. Nucl Acids Res16:9097–10111.

Njolstad PR, Molven A, Apold J, Fjose A. 1990. The zebrafish ho-meobox gene hox-2.2: transcription unit, potential regulatory re-gions and in situ localization of transcripts. EMBO J 9:515–24.

Oates AC, Brownlie A, Pratt SJ, Irvine DV, Liao EC, Paw BH, DorianKJ, Johnson SL, Postlethwait JH, Zon LI, Wilks AF. 1999. Geneduplication of zebrafish JAK2 homologs is accompanied by diver-gent embryonic expression patterns; only jak2a is expressed duringerythropoiesis. In press, Blood, 1999.

Orita M, Iwahana H, Kanazawa H, Hayashi K, Sekiya T. 1989.Detection of polymorphisms of human DNA by gel electrophoresis assingle-strand conformation polymorphisms. Proc Natl Acad Sci USA86:2766–2771.

Pellegrini S, John J, Shearer M, Kerr IM, Stark GR. 1989. Use of aselectable marker regulated by alpha interferon to obtain mutationsin the signaling pathway. Mol Cell Biol 9:4605–4612.

Perrimon N, Mahowald AP. 1986. l(1)hopscotch, a larval-pupal zygoticlethal with a specific maternal effect on segmentation in Drosophila.Dev Biol 118:28–41.

Pfeffer LM, Mullersman JE, Pfeffer SR, Murti A, Shi W, Yang CH.1997. STAT3 as an adapter to couple phosphatidylinositol 3-kinaseto the IFNAR1 chain of the type I interferon receptor. Science276:1418–1420.

Pinon LG, Minichiello L, Klein R, Davies AM. 1996. Timing ofneuronal death in trkA, trkB and trkC mutant embryos revealsdevelopmental changes in sensory neuron dependence on Trk signal-ling. Development 122:3255–3261.


Planas AM, Justicia C, Ferrer I. 1997a. Stat1 in developing and adultrat brain. Induction after transient focal ischemia. Neuroreport8:1359–1362.

Planas AM, Berruezo M, Justicia C, Barron S, Ferrer I. 1997b. Stat3 ispresent in the developing and adult rat cerebellum and participatesin the formation of transcription complexes binding DNA at thesis-inducible element. J Neurochem 68:1345–1351.

Porter AC, Chernajovsky Y, Dale TC, Gilbert CS, Stark GR, Kerr IM.1988. Interferon response element of the human gene 6–16. EMBO J7:85–92.

Potts JD, Kornacker S, Beebe DC. 1998. Activation of the Jak-STAT-signaling pathway in embryonic lens cells. Dev Biol 204:277–292.

Raz R, Lee CK, Cannizzaro LA, d’Eustachio P, Levy DE. 1999.Essential role of STAT3 for embryonic stem cell pluripotency. ProcNatl Acad Sci USA 96:2846–2851.

Ripperger JA, Fritz S, Richter K, Hocke GM, Lottspeich F, Fey GH.1995. Transcription factors Stat3 and Stat5b are present in rat livernuclei late in an acute phase response and bind interleukin-6response elements. J Biol Chem 270:29998–30006.

Sasse J, Hemmann U, Schwartz C, Schniertshauer U, Heesel B,Landgraf C, Schneider-Mergener J, Heinrich PC, Horn F. 1997.Mutational analysis of acute-phase response factor/Stat3 activationand dimerization. Mol Cell Biol 17:4677–4686.

Sambrook J, Fritsch EF, Maniatis T. 1989. Molecular cloning, alaboratory manual, 2nd ed. Cold Spring Harbor: Cold Spring HarborLaboratory Press.

Schaefer TS, Sanders LK, Nathans D. 1995. Cooperative transcrip-tional activity of Jun and Stat3 beta, a short form of Stat3. Proc NatlAcad Sci USA 92:9097–10101.

Schaefer TS, Sanders LK, Park OK, Nathans D. 1997. Functionaldifferences between Stat3alpha and Stat3beta. Mol Cell Biol 17:5307–5316.

Shi W, Inoue M, Minami M, Takeda K, Matsumoto M, Matsuda Y,Kishimoto T, Akira S. 1996. The genomic structure and chromo-somal localization of the mouse STAT3 gene. Int Immunol 8:1205–1211.

Schindler C, Shuai K, Prezioso VR, Darnell JE Jr. 1992. Interferon-dependent tyrosine phosphorylation of a latent cytoplasmic transcrip-tion factor. Science 257:809–813.

Schindler U, Wu P, Rothe M, Brasseur M, McKnight SL. 1995.Components of a Stat recognition code: evidence for two layers ofmolecular selectivity. Immunity 2:689–697.

Schulte-Merker S, Ho RK, Herrmann BG, Nusslein-Volhard C. 1992.The protein product of the zebrafish homologue of the mouse T geneis expressed in nuclei of the germ ring and the notochord of the earlyembryo. Development 116:1021–1032.

Shuai K, Stark GR, Kerr IM, Darnell JE Jr. 1993. A single phosphoty-rosine residue of Stat91 required for gene activation by interferon-gamma. Science 261:1744–1746.

Shuai K, Horvath CM, Huang LH, Qureshi SA, Cowburn D, DarnellJE Jr. 1994. Interferon activation of the transcription factor Stat91involves dimerization through SH2-phosphotyrosyl peptide interac-tions. Cell 76:821–828.

Silva CM, Lu H, Day RN. 1996. Characterization and cloning of STAT5from IM-9 cells and its activation by growth hormone. Mol Endocri-nol 10:508–518.

Smeyne RJ, Klein R, Schnapp A, Long LK, Bryant S, Lewin A, Lira SABarbacid M. 1994. Severe sensory and sympathetic neuropathies inmice carrying a disrupted Trk/NGF receptor gene. Nature 368:246–249.

Spring,J. 1997. Vertebrate evolution by interspecific hybridization—are we polyploid? FEBS Lett 400:2–8.

Takeda K, Noguchi K, Shi W, Tanaka T, Matsumoto M, Yoshida N,Kishimoto T, Akira S. 1997. Targeted disruption of the mouse Stat3gene leads to early embryonic lethality. Proc Natl Acad Sci USA94:3801–3804.

Tongiorgi E, Bernhardt RR, Schachner M. 1995. Zebrafish neuronsexpress two L1-related molecules during early axonogenesis. JNeurosci Res 42:547–561.

Velazquez L, Fellous M, Stark GR, Pellegrini S. 1992. A proteintyrosine kinase in the interferon alpha/beta signaling pathway. Cell70:313–22.

Vinkemeier U, Cohen SL, Moarefi I, Chait BT, Kuriyan J, Darnell JEJr. 1996. DNA binding of in vitro activated Stat1 alpha, Stat1 betaand truncated Stat1: interaction between NH2-terminal domainsstabilizes binding of two dimers to tandem DNA sites. EMBO J15:5616–5626.

Vinkemeier U, Moarefi I, Darnell JE Jr, Kuriyan J. 1998. Structure ofthe amino-terminal protein interaction domain of STAT-4. Science279:1048–1052.

Wen Z, Darnell JE Jr. 1997. Mapping of Stat3 serine phosphorylationto a single residue (727) and evidence that serine phosphorylationhas no influence on DNA binding of Stat1 and Stat3. Nucl Acids Res25:2062–2067.

Wen Z, Zhong Z, Darnell JE Jr. 1995. Maximal activation of transcrip-tion by Stat1 and Stat3 requires both tyrosine and serine phosphor-ylation. Cell 82:241–250.

Westerfield, M. 1995. The zebrafish book, 3rd ed. Eugene, Oregon:University of Oregon Press, Eugene.

Wilks AF. 1989. Two putative protein-tyrosine kinases identified byapplication of the polymerase chain reaction. Proc Natl Acad SciUSA 86:1603–1607.

Wilks AF, Harpur AG. 1996. Intracellular signal transduction: theJAK-STAT pathway. Austin, Texas: RG Landes.

Wu YY, Bradshaw RA. 1996. Induction of neurite outgrowth byinterleukin-6 is accompanied by activation of Stat3 signaling path-way in a variant PC12 cell (E2) line. J Biol Chem 271:13023–13032.

Xu X, Sun YL, Hoey T. 1996. Cooperative DNA binding and sequence-selective recognition conferred by the STAT amino-terminal domain.Science 273:794–797.

Yamanaka Y, Mizuno T, Sasai Y, Kishi M, Takeda H, Kim CH, Hibi M,Hirano T. 1998. A novel homeobox gene, dharma, can induce theorganizer in a non-cell-autonomous manner. Genes Dev 12:2345–2353.

Yan R, Small S, Desplan C, Dearolf CR, Darnell JE Jr. 1996.Identification of a Stat gene that functions in Drosophila develop-ment. Cell 84:421–430.

Zhang JJ, Vinkemeier U, Gu W, Chakravarti D, Horvath CM, DarnellJE Jr. 1996. Two contact regions between Stat1 and CBP/p300 ininterferon gamma signaling. Proc Natl Acad Sci USA 93:15092–15096.

Zhong Z, Wen Z, Darnell JE Jr. 1994a. Stat3: a STAT family memberactivated by tyrosine phosphorylation in response to epidermalgrowth factor and interleukin-6. Science 264:95–98.

Zhong Z, Wen Z, Darnell JE Jr. 1994b. Stat3 and Stat4: members of thefamily of signal transducers and activators of transcription. ProcNatl Acad Sci USA 91:4806–4810.

370 OATES ET AL.

DEVELOPMENTAL DYNAMICS 215:352–370...

Documents

Transcript of DEVELOPMENTAL DYNAMICS 215:352–370...