Evolution of developmental regulation in the vertebrate ...uoneuro.uoregon.edu › ~cristian ›...

Evolution of Developmental Regulationin the Vertebrate FgfD Subfamily

RICHARD JOVELIN1, YI-LIN YAN2, XINJUN HE2, JULIAN CATCHEN2,3,ANGEL AMORES2, CRISTIAN CANESTRO2, HAYATO YOKOI2, AND

JOHN H. POSTLETHWAIT2�1Center for Ecology and Evolutionary Biology, University of Oregon, Eugene,Oregon2Institute of Neuroscience, University of Oregon, Eugene, Oregon3Department of Computer and Information Science, University of Oregon,Eugene, Oregon

ABSTRACT Fibroblast growth factors (Fgfs) encode small signaling proteins that help regulateembryo patterning. Fgfs fall into seven families, including FgfD. Nonvertebrate chordates have asingle FgfD gene; mammals have three (Fgf8, Fgf17, and Fgf18); and teleosts have six (fgf8a, fgf8b,fgf17, fgf18a, fgf18b, and fgf24). What are the evolutionary processes that led to the structuralduplication and functional diversification of FgfD genes during vertebrate phylogeny? To study thisquestion, we investigated conserved syntenies, patterns of gene expression, and the distribution ofconserved noncoding elements (CNEs) in FgfD genes of stickleback and zebrafish, and comparedthem with data from cephalochordates, urochordates, and mammals. Genomic analysis suggests thatFgf8, Fgf17, Fgf18, and Fgf24 arose in two rounds of whole genome duplication at the base of thevertebrate radiation; that fgf8 and fgf18 duplications occurred at the base of the teleost radiation;and that Fgf24 is an ohnolog that was lost in the mammalian lineage. Expression analysis suggeststhat ancestral subfunctions partitioned between gene duplicates and points to the evolution of novelexpression domains. Analysis of CNEs, at least some of which are candidate regulatory elements,suggests that ancestral CNEs partitioned between gene duplicates. These results help explain theevolutionary pathways by which the developmentally important family of FgfD molecules arose andthe deduced principles that guided FgfD evolution are likely applicable to the evolution ofdevelopmental regulation in many vertebrate multigene families. J. Exp. Zool. (Mol. Dev. Evol.)312B, 2009. r 2009 Wiley-Liss, Inc.

How to cite this article: Jovelin R, Yan Y-L, He X, Catchen J, Amores A, Canestro C, Yokoi H,Postlethwait JH. 2009. Evolution of developmental regulation in the vertebrate FgfDsubfamily. J. Exp. Zool. (Mol. Dev. Evol.) 312B:[page range].

The development of body form and the differ-entiation of cells depend on a variety of processes,including cell adhesion accomplished by celladhesion proteins (e.g. cadherins, integrins, selec-tins), cell communication performed by extracel-lular signaling proteins (e.g. Notch, Hedgehog,Wnt, Tgf-b, Fgf), and differential gene expressioncontrolled by transcription factors (e.g. homeobox,bHLH, C2H2-zinc finger, bZIP). Families ofstructurally related genes encode these proteinsand constitute an ancestral toolkit for the con-struction of animal embryos. Recent genomicinvestigations reveal a surprisingly complete gene

set already present in the last common ancestor ofall metazoans (Technau et al., 2005; Guder et al.,2006; Nichols et al., 2006; Kasbauer et al., 2007;Matus et al., 2007), but this ancestral toolkit

Published online in Wiley InterScience (www.interscience.wiley.com). DOI: 10.1002/jez.b.21307

Received 11 February 2009; Revised 4 May 2009; Accepted 27 May2009

Richard Jovelin and Yi-Lin Yan contributed equally to the work.Grant sponsor: The National Center for Research Resources; Grant

number: 5R01RR020833; Grant sponsor: National Institutes ofHealth; Grant number: P01 HD22486.�Correspondence to: John H. Postlethwait, Institute of

Neuroscience, 1254 University of Oregon, Eugene, OR 97403.E-mail: [email protected]

r 2009 WILEY-LISS, INC.

JOURNAL OF EXPERIMENTAL ZOOLOGY (MOL DEV EVOL) 312B (2009)

diversified by gene duplication and gene deletionin the great lineages of animal life.

Stem chordates appear to have inherited asubset of toolkit genes present in stem metazoans,and these genes amplified into several members ofeach subfamily, many in two rounds of wholegenome duplication (WGD) (R1 and R2) thatappear to have occurred at the base of thevertebrate radiation before the divergence ofchondrichthyans and osteichthyans about 530million years ago (Kumar and Hedges, ’98;Kortschak et al., 2001; Ornitz and Itoh, 2001;Lundin et al., 2003; Dehal and Boore, 2005;Garcia-Fernandez, 2005; Bourlat et al., 2006;Delsuc et al., 2006; Guder et al., 2006; Jacob andLum, 2007; Kitisin et al., 2007; Sundstrom et al.,2008; Yu et al., 2008; Lynch and Wagner, 2009).Genes derived from a WGD event are called‘‘ohnologs’’ to emphasize their special characterswith respect to genes duplicated by other mechan-isms, such as tandem gene duplication, unequalcrossing-over, or retrotransposition (Wolfe, 2000;Postlethwait, 2007). An example is the family ofHox clusters, which is initially amplified bytandem duplication into a gene cluster that thenreplicated in vertebrates into four clusters thatappear to be related to each other in the relation-ship ((A,B)(C,D)), as would be expected by the 2Rmodel (Meulemans and Bronner-Fraser, 2007;Amemiya et al., 2008).

The widespread occurrence of multigene fa-milies and subfamilies raises two major questionsregarding the evolution of developmental mechan-isms: What are the processes by which individualmembers of gene families evolve their specializedroles in development? And, to what extent doesthe origin of individual family members by geneduplication contribute to morphological innova-tion and lineage divergence?

A classic multigene family essential for animaldevelopment, the fibroblast growth factor (Fgf)gene family, encodes small signaling proteinsdeployed in many aspects of embryo patterning(Ornitz and Itoh, 2001). Human and mousegenomes have 22 Fgf genes grouped into sevensubfamilies, each containing two to four members(FgfA (1/2), FgfB (3/7/10/22), FgfC (4/5/6), FgfD(8/17/18), FgfE (9/16/20), FgfF (11/12/13/14), andFgfG (19/21/23)) (Itoh and Ornitz, 2004; Popoviciet al., 2005). The Fgf gene complement of theurochordate Ciona intestinalis, whose lineagediverged from the vertebrate lineage before theR2 events (Oda et al., 2002; Delsuc et al., 2006;Vienne and Pontarotti, 2006; Wada et al., 2006;

Delsuc et al., 2008), helps elucidate the pattern ofdiversification of mammalian Fgf subfamilies(Satou et al., 2002). Based on strong evidencefrom sequence similarities, two of the sixC. intestinalis Fgf genes are orthologs of FgfDand FgfF subfamilies, respectively, and moderatesupport, based on conserved protein signaturemotifs and conserved genomic syntenies, suggeststhat three other genes are orthologs of the FgfB,FgfC, and FgfE subfamilies (Satou et al., 2002;Popovici et al., 2005). It is not clear whetherCi-FgfL is orthologous to the FgfA subfamily orwhether it is unique to C. intestinalis.

Consider the origin of the FgfB, FgfC, and FgfGsubfamilies. A representative of FgfG appears tobe missing from Ciona and thus either arose bygene duplication in vertebrates or was secondarilylost in the urochordate lineage. Because FGF3,FGF4, and FGF19 of the FgfB, C, and Gsubfamilies are contiguous in a cluster in humanchromosome 11q13.3, these three genes likelyarose by tandem duplication. Those tandemduplications must have occurred before R1 be-cause FGF6 and FGF23 (FgfC and FgfG subfami-lies) are nearest neighbors in chromosome12p13.32 (the other FgfB member was apparentlylost from this chromosome), and FGF21 andFGF22 (FgfB and FgfG) are linked on humanchromosome 19 (Hsa19) (the FgfC member wasapparently deleted from this chromosome). Thisgenomic pattern suggests the model that anancient FgfB/C/G gene tandemly duplicated toFgfB and FgfC/G before the divergence ofurochordates and vertebrates and that subse-quently, FgfC/G duplicated to FgfC and FgfGbefore R1; after R2, chromosome rearrangementsand gene loss led to the current human genomicarrangement.

Unexpectedly, the cnidarian Nematostella vec-tensis has at least 13 Fgf genes, twice as many asCiona and five or six times more than Drosophilamelanogaster and Caenorhabditis elegans (Matuset al., 2007). Phylogenetic analysis suggests thatnine Nv-Fgfs may be species-specific with no clearrelationship to vertebrate Fgfs, whereas fourNv-Fgfs fall within the FgfD subfamily (Matuset al., 2007).

In mammals, the FgfD subfamily includes Fgf8,Fgf17, and Fgf18 (Ornitz and Itoh, 2001) and inzebrafish, fgf24 (Itoh and Ornitz, 2004; Popoviciet al., 2005). The vertebrate FgfD phylogenysuggests either the hypothesis that Fgf24 waslost in the tetrapod lineage (Jovelin et al., 2007), orthe hypothesis that fgf24 was an innovation in

R. JOVELIN ET AL.2

J. Exp. Zool. (Mol. Dev. Evol.)

teleosts. FgfD members play critical roles inmesoderm and neuronal induction across largephylogenetic distances. For instance Fgf8, Fgf17,and Fgf18 are expressed at the midbrain–hind-brain border (MHB) (Sato et al., 2004). FGF18 isrequired for skeletal development (Ohuchi et al.,2000; Liu et al., 2002; Ohbayashi et al., 2002), andFgf8 plays an important role in MHB and limbbud development (Lewandoski et al., 2000; Moonand Capecchi, 2000; Sun et al., 2002; Boulet et al.,2004). Remarkably, FGF signaling in cnidariansmay be important during gastrulation and neuralinduction as well (Matus et al., 2007).

Until recently, the comparison of Fgf genecontent between mammals and zebrafish did notreflect the R3 hypothesis, a WGD occurring in theteleost lineage after the teleosts diverged fromother ray-fin fish (Sidow, ’96; Amores et al., ’98;Postlethwait et al., ’98; Wittbrodt et al., ’98; Meyerand Schartl, ’99; Christoffels et al., 2004; Hoegget al., 2004; Naruse et al., 2004; Meyer and Van dePeer, 2005; Crow et al., 2006). Our previousphylogenetic analysis of the vertebrate FgfDsubfamily showed that fgf8 and fgf18 wereduplicated early in the teleost lineage and pro-vided evidence for a diversification compatiblewith the ray-fin fish genome duplication (Jovelinet al., 2007). Duplicates of human fgf6, fgf10, andfgf20 genes have also recently been identified inzebrafish (Itoh and Konishi, 2007).

Because of the central role of FGF signaling indevelopment, mutations in Fgf genes and theirreceptors cause hereditary human diseases (Itoh,2007). An understanding of the precise evolution-ary relationships among Fgf genes and insightsinto the mechanisms of their functional diversifi-cation are required to fully connect teleostresearch to human biology. For instance, inmouse, Fgf8 is required for heart development(Abu-Issa et al., 2002; Frank et al., 2002; Ilaganet al., 2006). In zebrafish, fgf8a is required for theexpression of cardiac genes (Reifers et al., 2000b),but fgf8b is not expressed in the heart; conversely,in stickleback, fgf8b but not fgf8a is expressed inthe heart (Jovelin et al., 2007). After the ray-finfish genome duplication, independent evolution ofregulatory elements in lineages leading to zebra-fish and stickleback led to the differential parti-tioning of fgf8 subfunctions in heart developmentbetween paralogs so that orthologous genes in thetwo species have come to express different func-tions (Jovelin et al., 2007).

In this study, we investigated how the develop-mental roles of FgfD subfamily members evolved

as these genes originated from R1 to R3. First, weanalyzed the conservation of syntenies amongchordate FgfD genes to discern evolutionaryrelationships ambiguous in phylogenies (Satouet al., 2002; Itoh and Ornitz, 2004; Popoviciet al., 2005; Jovelin et al., 2007). Second, we usedin situ hybridization in stickleback and zebrafish,whose lineages diverged early in teleost phylo-geny, about 300 million years ago (Hedges, 2002),to examine how functions of duplicate genesevolve in separate lineages. Third, we analyzedthe conservation of noncoding regions in thevicinity of FgfD genes in the context of genephylogeny and gene expression to learn howdivergence in regulatory elements in the face ofdifferential gene loss could affect functionaldivergence of surviving ohnologs.

Results confirmed the presence of an Fgf24 geneat the base of gnathostome vertebrates, thepartitioning of functions between Fgf24 andFgf8, and revealed the origin of teleost fgfDparalogs. Expression analyses identified sharedancestral functions and newly evolved functions invarious members of the FgfD group, and explora-tion of conserved noncoding elements (CNEs) gavestrong support for a pattern of functional diver-gence through subfunctionalization. The generalpattern of gene duplication and functional diver-gence that this study reveals is likely to apply tomany other vertebrate multiple gene families.

MATERIALS AND METHODS

Conserved synteny

Genome sequences were investigated usingthe Ensembl databases (http://www.ensembl.org/index.html) for zebrafish Zv7, human NCBI 35,mouse NCBI m36, and C. intestinalis version 2.0.Orthologies were determined by best reciprocalBLAST hit analysis (Wall et al., 2003) automatedin a genome-wide scale in our Synteny Databasegenomic analysis software (Catchen et al., 2009).From the elephant shark genomic database weisolated the following FgfD gene fragments:Fgf8: exon2 AAVX01442442; Fgf17: no fragmentsidentified; Fgf18: exon2 AAVX01015792 andAAVX01460739, exon3 AAVX01088747, and exon4AAVX01419960; Fgf24: exon2 AAVX01560519 andAAVX01583584, exon3 AAVX01466056 andAAVX01521786, and exon4 AAVX01108229. Theabsence of Fgf17 is not significant owing to the low(1.4�) covera ge of the shark genome (Venkateshet al., 2007).

FgfD GENES IN EVOLUTION AND DEVELOPMENT 3


Gene expression

In situ hybridization probes were made by invitro transcription of linearized clones and labeledwith digoxigenin-UTP. Stickleback probes weresynthesized from TOPO/Not-1 linearized clonesusing T3 RNA polymerase. Zebrafish fgf8a, fgf8b,fgf17, and fgf24 and stickleback fgf8a and fgf8bprobes were as described (Reifers et al., 2000a;Draper et al., 2003; Cao et al., 2004; Jovelin et al.,2007). Additional probes were: zebrafish fgf18,Fgf18173 GTGTTTGGGGTGGACGGTGTGAATand Fgf18-578 TTTTTGCTCCGCTTGCTGACT-GTAG (NM_001013264) covering the codingregion; zebrafish fgf18l, Fgf18L11 ATGCGGT-CCGTCCTGTGGTCT and Fgf18L-620 ATGG-CAGTGGCGGAGGAGAGG (NM_001012379)covering the coding region; stickleback fgf18,Gacfgf18195 TCAGCGTGCACGTGGAGAACCand Gacfgf18-585 GCTTGCCCCGCTTGCTGACG(ENSGACT00000023719) covering most of thecoding region but missing some of the 30 portionof the gene; stickleback fgf17ex3F CAACGGCCG-GAGCAGGGATT and fgf17-3’R TTCTTTACTC-TACTTCAGT (ENSGACG00000003432) covering442 bp of the last exon and 30 UTR; sticklebackfgf24-ex1F TTACATCGAGAACCACAC andfgf24-30R GCTGTCCCGACGTTCCGT (ENSGA-CG00000016697) covering 552 bp of codingsequence. Embryos were fixed in 4% paraformal-dehyde at 41C for at least 2 days and thendechorionated by hand. In situ hybridizationswere performed as described (Yan et al., 2002).Stickleback embryos were staged relative to thezebrafish staging series (Kimmel et al., ’95), whichshowed that at 201C, stickleback develop about 2.5times slower than zebrafish does at 28.51C (Creskoet al., 2003). The University of Oregon IACUCapproved protocols for this study.

Analysis of conserved noncoding elements

Genomic sequences including the immediateneighbors of each Fgf gene were obtained fromEnsembl release 47 (Hubbard et al., 2007) forhuman and teleosts (zebrafish, and the fourpercomorph fish: stickleback, medaka, fugu, andgreen-spotted pufferfish). Some neighbors flank-ing medaka fgf24 and fugu fgf17 were unavailableowing to short scaffolds, and a flanking neighborof zebrafish fgf18 was too distant to be included inthe analysis. Sequences were aligned usingthe program LAGAN (Brudno et al., 2003) andCNEs were identified using the VISTA server(Mayor et al., 2000, Frazer et al., 2004). Repeated

elements in the reference sequence were maskedusing human and fugu masks.

RESULTS

Vertebrate FgfD genes originated in tworounds of duplication

To learn the origins of vertebrate FgfD genes,we investigated conserved syntenies amongchordate genomes, beginning with nonvertebratechordates.

The human genome has four segments corre-sponding to the FgfD region of nonvertebratechordates:

Sequence similarities between the C. intestinalisFgfD gene Ci-Fgf8/17/18 (NP_001027648, ENS-CING00000009302) and vertebrate FgfD mem-bers suggest that the FgfD subfamily originatedfrom a single gene already present in stemolfactores, the last common ancestor of urochor-dates and vertebrates (Satou et al., 2002; Popoviciet al., 2005). The R2 hypothesis predicts that fourco-orthologs of the prototypic FgfD gene should bepresent in four paralogous chromosome segmentsin vertebrate genomes in the absence of chromo-somal rearrangements or gene losses. To test thisprediction, we examined the conservation ofsyntenies to compare the genomic neighborhoodof the single C. intestinalis Fgf8/17/18 gene withgenomic regions surrounding human FgfD genes(Fig. 1). Results showed that many genes flankingCi-Fgf8/17/18 have human orthologs near FGF8,FGF17, or FGF18. Immediately to the left of Fgf8/17/18 lies ENSCINT00000018903, the ortholog ofFBXW4, and immediately to the right of Fgf8/17/18 is XP_002122436, which is the ortholog ofhuman NPM1/2/3, and all three genes are tran-scribed in the same orientation (Fig. 1A). Humanorthologs of these three genes are also contiguousand transcribed in the same orientation onHsa10 (FBXW4, FGF8, NPM3) (Fig. 1A). Thissituation represents a fully conserved chromosomesegment since the divergence of urochordate andvertebrate lineages, estimated to be 800 millionyears ago from molecular clock data (Petersonet al., 2004; Blair and Hedges, 2005). The orienta-tion and nearest-neighbor relationship of FGF18/NPM1 and FGF17/NPM2 have also been pre-served, although the FBXW4 paralog is missingfrom these paralogons (Fig. 1A). Of 24 loci withclear human orthologs in the portion ofC. intestinalis chromosome 5q shown in Figure 1A,15 have orthologs in Hsa2, Hsa5, Hsa8, or Hsa10,

R. JOVELIN ET AL.4


suggesting that these may be paralogons from theR1 and R2 genome duplication events.

To determine whether the human orthologs ofgenes neighboring C. intestinalis Fgf8/17/18 wereclustered on human chromosomes or widelydistributed, we used circle plots. Figure 1B showsthat human orthologs of the neighbors of theC. intestinalis Fgf8/17/18 gene cluster around thehuman Fgf8, Fgf17, and Fgf18 genes rather thanbeing splattered over the full extent of eachhuman chromosome. We conclude that syntenieshave tended to be conserved in this region of thegenome, which would be expected if local inver-sions were more frequent than translocationsduring chordate evolution. Furthermore, thisresult is as predicted from the hypothesis thatstem olfactores (urochordates1vertebrates) had achromosome segment that duplicated twice, givingrise to paralogons on Hsa5, 8, and 10 containing

FGF18, FGF17, and FGF8, and another paralogoncontaining parts of Hsa4 or Hsa2p that today lacksan FgfD gene.

The human genome has four FgfD-relatedparalogons

The hypothesis that FGFD genes arose in tworounds of chromosome duplication predicts that adot plot that localizes paralogs of genes surround-ing FGF8 on Hsa10 should identify four para-logons, including parts of Hsa5 and Hsa8, and oneother chromosome segment. To test this predic-tion, we investigated the chromosome locations ofthe paralogs of all genes in the 10 Mb intervalsurrounding FGF8, from 98 to 108 Mb on Hsa10.Figure 1C shows dots for genes from the region ofHsa10 surrounding FGF8 and their paralogs onother chromosomes displayed by red pluses in a

Cin5q

Hsa2

Hsa5

Hsa8

Hsa10

FGF18

FGF17

FGF8

(FGF24)

B

Human Chromosomes

108Mb

Hsa10Paralogs

C

98Mb22212019181716151413121110987654321YX

FG

F8

FG

F17

FG

F18

MVP Hsa16

RIOK3 Hsa18

FOLR1 Hsa11 B3GAT1 Hsa11SCP2 Hsa1

KIAA0368 Hsa9

HD Hsa4FMOD Hsa1; ECM2 Hsa9; DCN Hsa12

3.50 Mb 3.55 Mb 3.60 Mb 3.65 Mb 3.70 Mb 3.75 Mb 3.80 Mb 3.85 Mb 3.90 Mb

TSGA14 Hsa7

POLE4 Hsa2

HMGCS1 Hsa5TMEM20 Hsa10

SLC7A2 Hsa8

FBXW4 Hsa10FGF8 Hsa10; FGF17 Hsa8; FGF18 Hsa5

PPAPDC1B Hsa8; PPAPDC1A Hsa10 FBXW11 Hsa5; BTRC Hsa10

CYP2U1 Hsa4; CYP2E1 Hsa10; CYP2C18 Hsa10

ADD2 Hsa2; ADD1 Hsa4; ADD3 Hsa10 FBXW11 Hsa5

MCCC2 Hsa5

RNPEPL1 Hsa2; RNPEP Hsa1

TMEM20 Hsa10

Cin5q

NPM3 Hsa10; NPM2 Hsa8; NPM1 Hsa5

A

Fig. 1. The chromosome region containing Fgf8/17/18 in the urochordate Ciona intestinalis shares conserved synteny withFGFD-containing regions of the human genome. (A) A portion of C. intestinalis chromosome 5q (rectangle) is enlarged to showthe positions of predicted C. intestinalis genes listed with the symbols for their human orthologs, often several, and theirlocations in the human genome. Black font represents Ciona genes with human orthologs near FGFD genes and gray fontrepresents other locations in the human genome. The black rectangle marks a trio of genes preserved in order and orientationfor 800 million years since the divergence of urochordate and vertebrate lineages. (B) A circle plot of the Fgf8/17/18 region of theC. intestinalis genome with arcs linking to human orthologs on chromosomes around the circumference of the circledemonstrating the limited position of the orthologs on each human chromosome. (C) A dot plot of paralogs for genes in a 10 Mbwindow surrounding FGF8 on Hsa10, with the location of FGF8, FGF17, and FGF18 circled. Paralogs to Hsa10 genes are shownon their chromosome directly above the location of the gene on Hsa8. Paralogs lie mainly on Hsa2, 4, 5, 8, and 10.



column above or below the Hsa10 gene. Resultsshowed that chromosomes 2, 4, 5, and 8 have 13,13, 8, and 9 paralogs of the genes indicated by graycircles on Hsa10, but all other chromosomes havesix or fewer paralogs, with an average of 1.5paralogs per chromosome. These results supportthe conclusion from the analysis of conservedsyntenies shared between C. intestinalis andhuman genomes that parts of Hsa2p and/or Hsa4are members of the FgfD-containing paralogon.

Teleost FgfD genes reflect R3 and anohnolog gone missing

Having identified paralogous chromosome seg-ments in the human genome that have propertiesexpected for ‘‘ohnologons’’ arising from the R1and R2 WGD events, we turned to understandingthe relationship of the six teleost FgfD genes tothe one Ciona and three human FgfD genes.

Is fgf24 an ohnolog gone missing from thehuman genome?

Recognizing that the primary fate of duplicategenes is nonfunctionalization (pseudogenation) of

one of the two paralogs (Bershtein and Tawfik,2008), we explored the hypothesis that thepredicted fourth R2-derived co-ortholog of theprototypical FgfD gene is Fgf24 and that thisgene was preserved in the teleost lineage but waslost in the human lineage, as suggested byphylogenetic analysis (Jovelin et al., 2007). Analternative hypothesis to be ruled out is that fgf24did not exist in the last common ancestor ofteleosts and humans but was a teleost innovation.

The R2 hypothesis makes three predictions.First, the teleost fgf24 gene should be embeddedin a chromosome segment that is conserved withthe human genome; second, the conserved regionshould have paralogs near other human FGFDgenes; and third, this segment should showconserved synteny with the Ciona Fgf8/17/18 geneneighborhood shown in Figure 1. Here we testthese predictions by examining conserved synte-nies around zebrafish fgf24, starting at the firstgene to its right (Fig. 2).

The gene lying immediately to the right of fgf24in the zebrafish and stickleback genomes (Q566X7and ENSGACT00000022099, respectively) encodeproteins that form a well-supported sister clade to

Dre14

Hsa2

74.56Mb74.64Mb

Hsa2

AUP1PCGF1

TLX2LBX2 LOXL3

bM57.1bM22.1

Dre14

aup1

tlx2npm4fgf24

loxl3lbx2

3.12Mb 3.57Mb

lbx2

tpm4fgf24

pcgf1

pcgf1

GacIV

A

B

C GacIV

tlx2

Fig. 2. Conserved syntenies for fgf24. (A) A portion of Hsa2p with orthologs to neighbors of teleost fgf24 genes. (B) A portionof zebrafish (Danio rerio) linkage group 14 (Dre14) containing fgf24. (C) A portion of stickleback (Gasterosteus aculeatus)linkage group IV (GacIV) containing fgf24. This evidence is as expected if the human orthologs of fgf24 and npm4 were deletedfrom the human genome. Lines connect orthologs. Positions on chromosomes given in megabases (Mb).

R. JOVELIN ET AL.6


tetrapod NPM1 and the zebrafish gene npm1; wewill call this gene npm4 for now (Fig. 2B andSuppl. Fig. S1). Thus, npm4 is either an npm1duplicate (e.g. npm1b), which would make fgf24 aduplicate of fgf18, or npm4 is a gene, like itsneighbor fgf24, without a human ortholog. Thehuman genome has three NPM genes: NPM1(5q35), which is adjacent to FGF18 (8p21.3);NPM2, which is the nearest neighbor of FGF17,and NPM3, which is the nearest neighbor of FGF8(10q24) (Figs. 3B, 4C, 5A). Thus, an FgfD gene andan Npm gene are neighbors for all extant humanFgfD genes, as would be expected if this was theancestral condition before R1, and neither npm4nor its neighbor fgf24 have unique humanorthologs.

The next full gene to the right of fgf24(ENSDARG00000011273) is annotated as tlx3abut is actually an ortholog of TLX2 (Wotton et al.,2008). Human and zebrafish genomes both have

three TLX family genes (Andermann andWeinberg, 2001; Langenau et al., 2002): TLX1 isfive genes distant from FGF8 and tlx1 is threegenes distant from fgf8a. TLX3 is the second genefrom FGF18, whereas tlx3b is 2.8 Mb from fgf18l.TLX2 is in Hsa2p13.1-p12 without a nearbyFGFD-family gene, but tlx2 (NP_705937) is thesecond gene from fgf24, which favors the modelthat fgf24 is an ohnolog gone missing from thehuman genome.

The next gene to the right, ENSDARG-00000063176 from Ensembl release 45, is notannotated in Zv7 despite the continued presenceof its nucleotide sequence in the position indicatedin Figure 2B. ENSDARG00000063176 has asa best human BLAST hit KAZALD1 locatedat 10q24.31 seven genes from FGF8, but in aphylogenetic tree, it forms, with its ortholog inthe pufferfish Tetraodon nigroviridis, an out-group to the tetrapod1zebrafish KAZALD1 clade

0.52Mb0.62Mb

Dre14

fgf18a

169.68Mb 171.58Mb

Hsa5

3XLT1PINCK FGF18

FBXW11STK10

16.71Mb

Dre10

ENSDARG00000062737

zgc:63728

zgc:101531npm1

fgf18btlx3b

NPM1

fbxw11

GacIVA

B

C

D

bM04.01bM07.9

fgf18LOC567566

777600001COL1hdcp

fbxw11etf1pcdh1

pcdh1pcdh1

pcdh1pcdh1

pcdh1

LOC567566etf1 pcdh1

pcdh1

Dre10

Hsa5

Dre14

GacIVE.

PCDH1FGF18

18.04Mb

Fig. 3. Conserved syntenies for Fgf18. (A) A portion of stickleback chromosome GacIV containing fgf18. (B) A portion ofzebrafish chromosome Dre10 containing fgf18b. (C) A portion of human chromosome Hsa5 containing FGF18. (D) A portion ofDre14 containing fgf18a. (E) Locations of each region shown related to the whole chromosome. Black genes have orthologs inthese sections, gray genes do not.



(Suppl. Fig. S2). This phylogenetic groupingsuggests that, like npm4 and fgf24, EN-SDARG00000063176 is a paralog without a humanortholog and should be called kazald2 (Wottonet al., 2008).

The two neighbors to the immediate left ofzebrafish fgf24 (lbx2 and pcgf1, Fig. 2B) haveorthologs adjacent on Hsa2p (Fig. 2A). In human,LBX2, PCGF1, and TLX2 are adjacent, but in thezebrafish genome fgf24 and npm4 lie betweenpcgf1 and tlx2 (Fig. 2A and B). Moving to the rightof TLX2, two of the next four genes in human,AUP1 and LOXL3, have orthologs also to the rightof fgf24 in zebrafish (Fig. 2A and B). Thus, thezebrafish and human share most of the genes in

this region of Hsa2p/Dre14, but orthologs of fgf24and npm4 are specifically excised from the humangenome. Stickleback linkage group IV (GacIV),which has five genes in a row orthologous to thezebrafish segment in Dre14 just discussed, and anortholog of aup1 syntenic on GacIV, but about halfa chromosome away (Fig. 2C). These resultssuggest that the five genes (lbx2, pcgf1, fgf24,npm4, tlx2) were neighbors before the divergenceof zebrafish and stickleback lineages and thatinversions in the stickleback lineage removedaup1 from its former neighbors.

The most parsimonious explanation of these datais that fgf24 and npm4 were adjacent in the FgfDparalogon in the last common ancestor of zebrafish

zgc:136955

29.31Mb

Dre1

poll

lbx1bzgc:153412

bM82.4bM27.3

GacVI

TLX1C6orf57

OGFRL1

ENSGACG00000003828fgf8a

2TAG3B4WXBF

102.95Mb 105.13Mb

Hsa10

LOC100007451NP_001082838.1

nt5c2

cyp17a1 si:ch211-103o12.1

fgf8afbxw4

zgc:114169

A2ATX7ldb4

C10orf26AS3MT

CNNM2

INA

LBX1FGF8

FBXW4LDB1 CYP17A1 NT5C2

PCGF6

RP11-529I10.4

POLL

Dre13

NP_739571.1si:busm1-265n4.4

ogflr11

slc2a5 21.33Mbzgc:91925zgc:103568A2ATX5

HPS6

fgf8b

zgc:111946

zgc:136933 gefiltin

zgc:15365129.69Mb

bM49.01bM27.01

GacIX

ENSGACG00000018247NP_056263.1

LBX1CENPK

INAfgf8bPOLL

C10orf26AS3MT

cenpk

NPM3

A

B

C

D

E

20.255Mb

Fig. 4. Conserved syntenies for Fgf8. (A) A portion of stickleback chromosome GacVI containing fgf8a. (B) A portion ofzebrafish chromosome Dre13 containing fgf8b. (C) A portion of human chromosome Hsa10 containing FGF8. (D) A portion ofDre1 containing fgf8b. (E) A portion of GacIX containing fgf8b. Black genes have orthologs in these sections, gray genes do not.

R. JOVELIN ET AL.8


and human, and that they were lost in the humanlineage. Note that the orientation of TLX2/tlx2relative to LBX2/lbx2 and PCGF1/pcgf1 is oppositein human and zebrafish, suggesting that a localinversion stirred the sequences, either in thehuman or zebrafish lineage. The human FGF18region can serve as an outgroup to order the vectorof evolutionary change: TLX3, NPM1, and FGF18are all in the same orientation, suggesting that thedeletion of the tetrapod Fgf24 and Npm4 ohnologswas associated with a small local inversion.

When did the duplication event that producedFgf24 occur? BLAST searches of the trace files ofthe genome sequencing project of the elephantshark Callorhinchus milli revealed that chon-drichthyes have at least an Fgf8 gene (exon 2AAVX01442442) and both an Fgf18 gene (exon 3AAVX01088747) and an Fgf24 gene (exon 3AAVX01521786), as shown by exclusive aminoacid positions in an alignment of exon-3 of FgfDgenes from various vertebrates (Suppl. Fig. S3A).Because the shark genome, which diverged fromthe bony vertebrates before the divergence ofteleost and tetrapod lineages, has an Fgf24 gene,this gene must have already existed in the lastcommon ancestor of all jawed vertebrates. Theinclusion of shark sequences for Fgf18 and Fgf24in a phylogenetic tree of exon 3 results in thetopology ((Fgf8,Fgf17)(Fgf18,Fgf24)) as expectedfrom their origin in the R1 and R2 genomeduplications (Suppl. Fig. S3B).

Altogether, the best interpretation of thesefindings is that the vertebrate FgfD subfamilyoriginated from a single gene present in theancestor of vertebrates and urochordates by tworounds of genome duplication. These R1 and R2events lead to four Fgf genes in the last commonancestor of teleosts and tetrapods but the orthologof fgf24 was subsequently lost in the tetrapodlineage.

fgf18a and fgf18b are duplicates from theteleost genome duplication

Phylogenetic analysis shows that zebrafish hastwo genes (fgf18a (NM_001013264) and fgf18b(also called fgf18l; NM_001012379)) in a cladewith FGF18 as the most closely related humangene. Zebrafish Fgf18b is sister-taxon to a teleostFgf18a clade (Itoh and Konishi, 2007; Jovelinet al., 2007; Kikuta et al., 2007) as would beexpected from gene duplication after the diver-gence of ray-fin and lobe-fin fish (Itoh and Konishi,2007; Jovelin et al., 2007; Kikuta et al., 2007).Four other sequenced teleost genomes have anortholog of fgf18a but lack an annotated fgf18band BLAST searches failed to identify an orthologof fgf18b in genome databases for stickleback,medaka, Tetraodon, or fugu, all of which arepercomorph fish. This situation could have arisenif fgf18a and fgf18b arose by the R3 duplication instem teleosts and fgf18b was lost in the lineage of

21.93Mb 25.40Mb

Hsa8

bM31.74bM46.54

Dre8

FGF17CHMP7

ENTPD4

KCTD9

zgc:92668

zgc:113115lgi3chmp7

LGI3

nr5a1a

zgc:123194

LOC571805

LOC100004958cyr61

fgf17b

0.30Mb 0.78Mb

fgf17ENSGACG00000003433

ENSGACG00000003446

ENSGACG00000003456

1A5RN17430000000GCAGSNE

GacXIII

NPM2

A

B

C

Fig. 5. Conserved syntenies for Fgf17. (A) A portion of Hsa8 containing FGF17. (B) A portion of Dre8 containing fgf17.(C) A portion of GacXIII containing fgf17. Black genes have orthologs in these sections, gray genes do not.



the Percomorpha but was retained in the lineageof the Ostariophysi (including zebrafish). Analternative hypothesis is that fgf18b arose bytandem duplication in the zebrafish lineage afterit diverged from the percomorphs.

Conserved syntenies can help distinguish the R3and tandem duplication hypotheses for the originof fgf18 paralogs. The fgf18a gene of zebrafish ison Dre14 near fgf24 along with fbxw11, theortholog of which is separated from FGF18 by asingle hypothetical gene (Fig. 3C and D), thussupporting the orthology of zebrafish fgf18aand human FGF18. Note that the C. intestinalisgenome has two orthologs of FBXW11, onelocated 10 genes distant and one 13 genes awayfrom Fgf8/17/18 (NP_001027626 and ENSCINT-00000018903) (Fig. 1A); this is another example ofan ancient FgfD-region synteny conserved for 800million years. A cluster of pcdh1 genes flankingfgf18 (not shown) are, as a group, co-orthologousto PCDH1 protocadherin genes in Hsa5q31, abit distant but syntenic to FGF18 in Hsa5q35(Fig. 3C–E). These gene clusters were apparentlytandemly duplicated at least partly independentlyin the zebrafish and human lineages. Together,these conserved syntenies support orthology offgf18a and FGF18.

The orthologs of several genes located nearzebrafish fgf18b on zebrafish LG10 are adjacentto or near FGF18 on Hsa5q (Fig. 3B and C),consistent with the orthology of fgf18b and FGF18as shown in the phylogeny. The nearest neighborto fgf18b (Fig. 3B) is an ortholog of FBXW11,which is separated from FGF18 by a single smallhypothetical gene (Fig. 3C). Like fgf18a andFGF18 neighborhoods, the fgf18b region of zebra-fish and stickleback both have a series of tandemlyduplicated pcdh genes. The region surroundingfgf18 in stickleback has at least five neighborswith orthologies to genes in the region surround-ing fgf18b in zebrafish, a result expected if thetwo regions are orthologous (Fig. 3A and B). Thisis in conflict with the phylogeny, which shows thatFgf18 in stickleback (ENSGACT00000023719) ismore closely related to Fgf18a than it is to Fgf18b(Jovelin et al., 2007). If the conserved synteniescorrectly suggest orthologies, then the zebrafishFgf18b sequence is evolving more rapidly thanthe Fgf18a sequence. Alternatively, gene lossesmay account for the conflict. If orthologs neigh-boring stickleback fgf18 were lost near fgf18abut not near fgf18b, then the reciprocal bestblast method for identifying orthologies couldfind only genes near fgf18b and call them as

orthologs given the loss of these genes fromthe neighborhood of fgf18a. Both explanations,however, are consistent with the origin offgf18a and fgf18b in the R3 event in teleostevolution.

fgf8 duplicates arose in R3

Zebrafish has one uncontested ortholog of FGF8(Brand et al., ’96; Reifers et al., ’98), and a secondgene (fgf17a) that was initially assigned as a co-ortholog of tetrapod Fgf17 (Reifers et al., 2000a;Cao et al., 2004), but that was later shown to be afgf8 duplicate and is now called fgf8b (Itoh andKonishi, 2007; Jovelin et al., 2007; Kikuta et al.,2007). A detailed analysis of conserved synteniesin human, zebrafish, and stickleback supports thisconclusion. Orthologs of six genes surroundinghuman FGF8 in a 2.16 Mb region on chromosomeHsa10q (Fig. 4C) are located in a 0.4 Mb region onzebrafish chromosome Dre1 containing fgf8b (Fig.4B). The orthologous portion of stickleback link-age group GacIX contains fgf8b and eight genes inthe corresponding region of the human genome(Fig. 4A). The orthologs of a somewhat differentset of human genes flanking FGF8 lie nearfgf8a in zebrafish Dre13 and stickleback LG VI(Fig. 4D and E). We conclude that zebrafishfgf8 (NM_131281) and the so-called fgf17a(NM_182856) are co-orthologs of human FGF8and support the conclusion that they should benamed fgf8a and fgf8b (Itoh and Konishi, 2007;Jovelin et al., 2007; Kikuta et al., 2007). Figure 4shows that in addition to FGF8 co-orthologs,zebrafish and stickleback have co-orthologs ofLBX1 (Q5CZV7 and si:busm1-265n4.1) andC10orf26 (zgc:111946 and Q1ECU4) in thesegenomic regions, indicating that that theseduplicated loci also originated in the R3 genomeduplication.

Teleosts have a single copy of fgf17

Sequenced teleost genomes have a single ortho-log of FGF17. Several genes in the immediateneighborhood the gene originally called fgf17b onzebrafish Dre8 (Cao et al., 2004) are orthologs ofgenes that are near FGF17 in Hsa8p21 (Fig. 5Aand B). These two segments are largely conservedbetween zebrafish and stickleback (Fig. 5B and C).We conclude that fgf17b should be called simplyfgf17 (NM_182856) and it is the zebrafish orthologof FGF17.

R. JOVELIN ET AL.10


Expression patterns of FgfD family genes

With this understanding of the evolutionaryrelationships of teleost and tetrapod FgfD genes,we wondered how paralog functions evolved asjudged by gene expression patterns.

Mid-segmentation stages

fgf8 paralogs: At mid-segmentation stages(40 hpf stickleback and 16 hpf zebrafish) (Kimmelet al., ’95), the expression of fgf8a and fgf8b aresimilar in zebrafish and stickleback (Reifers et al.,’98, 2000a; Draper et al., 2003; Jovelin et al.,2007), with strong expression of both genes inboth species in MBH and somites (Fig. 6A–H). In

addition, fgf8a is expressed in the dorsal dience-phalon and the tailbud in both species. In theheart, stickleback and zebrafish express differentfgf8 paralogs (Jovelin et al., 2007).

fgf17: At mid-segmentation in stickleback andzebrafish, fgf17 is expressed weakly in the centralnervous system (CNS), and more strongly andspecifically around the tail bud and in thesegmental plate and segmented somites (Fig. 6I–L)(Cao et al., 2004; Hamade et al., 2006). The majordifference between sfgf17 (stickleback fgf17)and zfgf17 (zebrafish fgf17) is that the zebrafishgene is expressed in the dorsal diencephalon(Fig. 6K). For both species, fgf8a and fgf8bare both expressed in the MHB, but fgf17 is not,

Gac 40 hpf Dre 16 hpfA B C D

E F G H

I J K L

M N O P

Q R

S T U V

a8fgfa8fgf

b8fgfb8fgf

71fgf71fgf

a81fgf81fgf

fgf18b

42fgf42fgf

mhb mhbdd dd

r r

ss

tbtb

bhmbhm

mhb

mhb

mhb

s s

ss

s

s

tb tb

umum

ov

ovov

dd

dddd

rr

r

rr

r r

h h

Fig. 6. Expression of stickleback and zebrafish FgfD genes at midsegmentation stages. A, B, E, F, I, J, M, N, S, T, 40 hpfstickleback embryos. C, D, G, H, K, L, O, P, Q, R, U, V, 16 hpf zebrafish embryos. A–D, fgf8a. E–H, fgf8b. I–L, fgf17. M–P,fgf18a. Q, R, fgf18b. S–V, fgf24. dd, dorsal diencephalon; h, heart; mhb, midbrain–hindbrain boundary; ov, otic vesicle; r, retina;s, somite; tb, tail bud; um, unsegmented mesenchyme.



and all three genes are expressed in the somites.In the tailbud, fgf8a in both species is expressed inthe tailbud itself; fgf8b is not expressed in thetailbud region; and fgf17 is expressed in thesegmental mesoderm that surrounds the tailbudin a pattern complementary to that of fgf8a. Bothzfgf17 and zfgf8a are expressed in the dorsaldiencephalon.

fgf18: The expression of zfgf18a and zfgf18bhave not been previously described. At mid-segmentation, the single copy of fgf18 in stickle-back is expressed weakly in the MHB and in theheart primordium, but not in the somites (Fig. 6Mand N). In zebrafish, zfgf18b, but not zfgf18a, isexpressed in the MHB like zfgf8a, and recipro-cally, zfgf18a, but not zfgf18b, is expressed in theheart like zfgf8b (Fig. 6O–R). Although sfgf18 isexpressed only weakly if at all in the somites(Fig. 6N), zfgf18a and zfgf18b are both expressedin the somites, but zfgf18a is expressed broadly insomites like zfgf8b (Fig. 6H and P), but zfgf18b isexpressed in just a part of each somite like zfgf8a(Fig. 6D and R).

fgf24: In mid-segmentation stage sticklebackembryos, sfgf24 is expressed in the dorsal dience-phalon, MBH, and tailbud, but not in the somites,like sfgf18 (Fig. 6S and T). In zebrafish as instickleback, fgf24 is expressed in the dorsaldiencephalon like fgf8a (Draper et al., 2003). Incontrast to sfgf24, zfgf24 is not expressed in theMHB (Fig. 6U and V). In both stickleback andzebrafish, fgf24 is expressed in the unsegmentedpresomitic mesoderm.

Long pec stage

fgf8: At the ‘‘long pec’’ stage (120 hpf sticklebackand 48 hpf zebrafish) (Kimmel et al., ’95), theexpression patterns of both fgf8 genes in bothspecies occupy several of the earlier domains, butin addition, in both species, new domainsappear for fgf8a (but not fgf8b) in the telencepha-lon, olfactory epithelium, pharyngeal arches,(Meulemans and Bronner-Fraser, 2007), andapical epidermal ridge (AER) of the pectoral finbud (Fig. 7A–D).

fgf17: At the long pec stage, stickleback sfgf17 isexpressed in the retina (Fig. 7E) in a pattern thatoverlaps in time and space expression of sfgf8a inthe eye, (Fig. 7A) and is stronger than sfgf8bexpression in the eye at the same stage (Fig. 7C).In addition, sfgf17 and zfgf17 are expressed in theotic vesicle (Fig. 7E and F). sfgf17 was only weaklyexpressed in the pharyngeal arches compared with

zebrafish (Fig. 7E and F). No expression of sfgf17was apparent in the MHB or in the pectoral finbud. In zebrafish, zfgf17 is also expressed in theeye, but broadly in several layers. In mouse, theexpression patterns of Fgf17 and Fgf8 largelyoverlap, but Fgf17 is expressed in a broader regionof the frontal cortex and cerebellum of the brainand is required for dorsal frontal cortex andcerebellar development (Xu et al., 2000; Cholfinand Rubenstein, 2007; Dominguez and Rakic,2008). The regenerating adult zebrafish heart alsoexpresses fgf17 (Lepilina et al., 2006).

fgf18: To the expression domains present at mid-segmentation stages (Fig. 6M–R), at the long pecstage, stickleback sfgf18 and zebrafish zfgf18a addexpression in the retina, olfactory epithelium, oticvesicle, and pharyngeal arches 3–7. In contrast,

Gac 5 dpf Dre 2 dpf

A B

C D

E F

G H

I

J K

fgf8a fgf8a

fgf8b fgf8b

fgf17 fgf17

fgf18a fgf18a

fgf18b

fgf24 fgf24

dd

mhb

pa

ov

rt

aeraer

mhbdd

r

ov

pat

os

mhbmhb

ovov

ov ov

pa

pa

rr

pa

pa pa

ovov

rr

rh

mhb

ov

pa

rn

mhb

ov ov

n

mhb

r

n

pa

mhb

naeraer

nn

Fig. 7. Expression of stickleback and zebrafish FgfD genesat long pec stage. A, C, E, G, J, 5 dpf stickleback embryos. B,D, F, H, I, K, 2 dpf zebrafish embryos. A, B, fgf8a. C, D, fgf8b.E, F, fgf17. G, H, fgf18a. I, fgf18b. J, K, fgf24. aer, apicalectodermal ridge of pectoral fin; dd, dorsal diencephalon; h,heart; mhb, midbrain-hindbrain boundary; n, nose; ov, oticvesicle; os, optic stalk; pa, pharyngeal arches; r, retina;t, telencephalon.

R. JOVELIN ET AL.12


zebrafish zfgf18b only adds expression in theretina and pharyngeal arches (Fig. 7G–I).

fgf24: At the long pec stage, stickleback sfgf24 isexpressed in the MHB, olfactory epithelium, mostcells of the otic vesicle, endoderm of the phar-yngeal pouches, and apical epidermal ridge of thepectoral fin (Fig. 7J). Expression of sfgf24 in theeye and MHB is broader than the expression ofsfgf8a and sfgf18 in these domains (Fig. 7A–G).Here we extend previous descriptions of zfgf24expression (Draper et al., 2003; Fischer et al.,2003); zfgf24 is expressed in a small group of cellsin the dorsal MHB, in the olfactory epithelium,endoderm of the pharyngeal pouches, and the oticvesicle. In the AER of the pectoral fin in bothstickleback and zebrafish, fgf24 is expressed broad-

er and stronger than fgf8a, with the fgf8 domainentirely within the fgf24 domain (Fig. 7A, B, J, K).

Eyes and ears

Several FgfD genes are expressed in the eyesand ears, but often in different domains. Theganglion cell layer of the eye expresses sfgf8a,zfgf8a, sfgf17, and sfgf24 (Fig. 8A, C, I, S)(Pickeret al., 2005). In the retina, zfgf24 is expressed inisolated cells scattered in various layers (Fig. 8S,U). Except for the retina and MHB, zfgf24 andsfgf24 are expressed similarly.

The developing ear expresses all FgfD genes atlong pec stages in both stickleback and zebrafish.Some genes are expressed generally throughout

A B C D

E F G H

I J K L

M N O P

Q R

S T U V

fgf18fgf18

fgf8a fgf8a

fgf8b fgf8b

fgf17 fgf17

fgf18l fgf18l

fgf24 fgf24

Gac 120 hpf Dre 48 hpf

ac

am

am

cpac

ac

pc

cp

lc

am

am

am

am

ac

am

pc

ac pc

cp

cp

cp

cp

pcam

cp

pp

acpc

gl gl

gl

vp

il

Fig. 8. Expression of stickleback and zebrafish FgfD genes in the ear. A, B, E, F, I, J, M, N, S, T, 5 dpf stickleback embryos.C, D, G, H, K, L, O, P, Q, R, U, V, 2 dpf zebrafish embryos. A–D, fgf8a. E–H, fgf8b. I–L, fgf17. M–P, fgf18a. Q, R, fgf18b. S–V,fgf24. ac, anterior crista; am, anterior macula; cp, canal projection; fp, fusion plate; gl, gangloin layer; il, inner nuclear layer; lc,lateral crista; pc, posterior crista; pp, pharyngeal pouches; vp, ventral patch.



the ear, but others are only expressed in specificregions. The anterior sensory patch shows strongexpression of sfgf8b in stickleback and its paralogzfgf8a in zebrafish, whereas expressing zfgf8b onlyweakly (Fig. 8D, F, H). In addition, the anteriorsensory patch expresses zfgf18b strongly andzfgf17 and zfgf18a only weakly (Fig. 8L, P, R).The anterior and posterior cristae express zfgf24(Fig. 8V).

In both stickleback and zebrafish during theearly pharyngula period, fgf18 and fgf18a areexpressed in specific spinal cord neurons (Fig. 9Aand B). None of the other FgfD genes areexpressed in spinal cord neurons at this stage inzebrafish, and none of the FgfD genes in mouseare expressed in spinal cord neurons at thisdevelopmental stage, suggesting that this expres-sion domain is either an innovation in stemteleosts, or that it was an ancestral function lostin the mammalian lineage.

Of the three FgfD genes in mouse, only Fgf18 isexpressed in the pancreas (Dichmann et al., 2003).In stickleback and zebrafish long pec stage, sfgf18,zfgf18a, sfgf24, and zfgf24 are also expressed inthe pancreas (Fig. 9C–F), suggesting that apancreas regulatory element was present beforethe gene duplication event that produced fgf18and fgf24.

Analysis of conserved noncoding elements

Expression patterns of the FgfD gene family justdescribed show many examples of conservationacross species and across paralogs. These con-served patterns are likely to be controlled byregulatory elements, the functions of which havebeen conserved for hundreds of millions of years.

Because some regulatory elements conserved infunction are also conserved in sequence, wesearched teleost and human FgfD genomic regionsfor CNEs.

CNE sharing among FgfD paralogs withinspecies

Human: To discover ancient CNEs derived fromthe R1 and R2 genome duplication events, weconducted a Vista plot analysis of human FGFDgenes. The analysis allowed several conclusions(Fig. 10A). First, CNEs shared between FGF8 andFGF18 were more frequent and longer than thoseshared by FGF8 and FGF17. Second, CNEsembedded in the genes flanking FGF8 (FBXW4and NPM3) are conserved with those flankingFGF18, which is located adjacent to NPM1 andnear FBXW11. NPM2 is adjacent to FGF17(Fig. 5A), but CNEs embedded in NPM3 (nearFGF8) and NPM1 (near Fgf18) are apparently notconserved with NPM2 (near FGF17). Third, aCNE (labeled ‘‘a’’) shared by FGF8 and FGF18 isfound within intron four of FBXW4—this issignificant because an enhancer trap elementinserted in a similar position in zebrafish confersan fgf8 expression pattern (Kikuta et al., 2007).Finally, three highly conserved CNEs located 3’ toFGF8 (labeled d, e, and f) are shared by all threehuman FGFD genes. We conclude that at leastthree noncoding regions in the genomic neighbor-hood of FGFD genes have been highly conservedin sequence for at least 500 million years (Kumarand Hedges, ’98; Peterson et al., 2004) since theR1 and R2 genome duplication events. Our resultsraise the hypothesis that these conserved se-quences are regulatory elements that drive theexpression of FGFD genes in domains common toall three genes.

Stickleback: To test if CNEs near FgfD genes arealso conserved in teleosts, we constructed Vistaplots for stickleback. Fig. 10B shows conservationfor the exons of fgf8a/fgf8b and scl2a5a/scl2a5b atgreater than threshold, but no CNEs shared by thetwo paralogs. This result is a bit surprising giventhe similarities of expression patterns between thetwo sfgf8 ohnologs. In contrast, several CNEs wereapparent in the comparison of sfgf8a and sfgf17,although most exons failed to rise above threshold(Minimum Y value on the VISTA plot 5 50%,Minimum conservation identity 5 70%, Minimumlength for a CNE 5 100).

Three of the stickleback CNEs (a, b, e) were inintrons of fgfD neighbor genes, and two were found

Gac Drefgf18

fgf18

fgf18a

fgf18a

fgf24 fgf24

A B

C D

E F

2.5 dpf 1dpf

5 dpf

5 dpf

2 dpf

2 dpf

spnspn

pan pan

panpan

Fig. 9. Expression of stickleback and zebrafish FgfD genesin the trunk. A, C, E, stickleback embryos. B, D, F, zebrafishembryos. A, B, fgf18 and fgf18a. C, D, fgf18 and fgf18a. E, F,fgf24. pan, pancreas; spn, spinal cord neurons.

R. JOVELIN ET AL.14


in intergenic region flanking the fgfD gene. Notethat CNE ‘‘a’’ in Fig. 10B is in the same intron ofstickleback fbxw4 as CNE ‘‘a’’ of human FBXW4although it is shared between the fgf8a and fgf17regions for stickleback and FGF8 and FGF18regions for human. Neither the exons nor anyCNEs were detected that were shared betweensfgf8a and either sfgf18 or sfgf24, suggesting rapidevolution, translocations, or problems with genomeassembly. The sparse pattern of CNEs in stickle-back fgfD genes contrasts with the rather broadCNEs found for human FGFD genes.

Zebrafish: For zebrafish (Fig. 10C), a strong CNE(labeled ‘‘d’’ in Fig. 10B and C and ‘‘a’’ in Fig. 10A)was detected in the same intron of fbxw4 asdetected in stickleback and in human, and it wasshared by all zebrafish fgfD regions except fgf18b.An enhancer trap insertion nearby (CLGY667)confers reporter expression in telencephalon, opticstalk, midbrain–hindbrain boundary, somites,heart, olfactory pits, and tail bud in a patternsimilar to that of fgf8 genes (Kikuta et al., 2007).This CNE, conserved since before R1 and main-tained by fgf8a, fgf8b, fgf17, fgf18a, and fgf24, may

Fig. 10. Distribution of CNEs among FgfD paralogs in human (A), stickleback (B) and zebrafish (C). The higher density ofCNEs among human paralogs is probably the result of the R3 genome duplication in teleosts with subsequent relaxed selectionand partitioning of regulatory elements (Force et al., 1999). The reference sequence is indicated at the top. The position of exonsis indicated along the toop. For each conservation plot, the top line represents 100% conservation with the reference sequence,and the bottom line indicates 50% conservation. Gray arrows show gene orientation. Lower case letters indicate peaksmentioned in the text.



play a role in defining expression domains sharedby vertebrate FgfD genes.

Despite their similarity in expression patterns(Fig. 8A–H, 9A–D), fgf8a shares fewer CNEs withfgf8b than it does with fgf17, fgf18a, and fgf24(Fig. 10C). Using fgf8a as base, one CNE (d) ispresent near five zebrafish fgfD genes, three (b, n,and r) are present near three fgfD genes, eight(f, i, k, l, m, q, w, and y) are near two fgfD genes,and 14 (a, c, e, g, h, j, o, p, s, t, u, v, x, and z) areshared by fgf8a and only one or another fgfD gene.fgf8a shares most CNEs (17) with fgf24 and aboutan equal number with fgf17 and fgf18a (11 and 10,respectively).

Enhancer trap CLG508 is located in betweenfgf8a and fbxw4 and confers expression in theapical ectodermal ridge of the pectoral fin bud(Kikuta et al., 2007), but Vista detected no CNEshared by fgf8a and other fgfD genes in thisintergenic region despite the expression of bothfgf8 and fgf24 in the AER. Enhancer trapCLGY1030 is inserted near CNE ‘‘u’’ in zebrafishand matches the expression pattern of fgf8a onlyin the tail bud (Kikuta et al., 2007).

CNE sharing among FgfD orthologs

CNEs displayed when comparing FgfD geneswithin a species show elements conserved fromthe R2 event, whereas CNEs revealed by compar-ing a single FgfD gene among species identifieselements conserved since species divergence.

fgf17: Few CNEs appeared in comparisons of thestickleback fgf17 gene to its orthologs in humanand several teleosts, although exons were readilyapparent (Fig. 11). At least six CNEs were

apparent when comparing stickleback and zebra-fish fgf17 genes, although the flanking genes werenot conserved (Fig. 11). Given that stickleback andpufferfish are more closely related than stickle-back and zebrafish, it was surprising to find moresfgf17 CNEs shared with zebrafish than puffer-fish. The lack of CNEs near fgf17 in the twopufferfish species could be due in part to incom-plete sequencing in this genomic region: fgf17in fugu (Takifugu rubripes) lies in a smallscaffold without predicted neighbors, and numer-ous sequencing gaps surround fgf17 in thegreen spotted pufferfish T. nigroviridis. Amongthe species examined, only stickleback andT. nigroviridis share a neighboring gene(ENSGACG00000003433) (Fig. 11). Close inspec-tion of chromosome segments carrying fgf17 inzebrafish and stickleback shows that global syn-teny is conserved but gene orders are not,indicating local genomic micro-rearrangements(Fig. 5). Comparison with human reveals rearran-gements over still larger genomic distances(Fig. 5). Chromosome breaks near fgf17 may havecarried CNEs along with fgf17 neighbors disrupt-ing synteny of coding and noncoding sequences inthese regions.

fgf18: BLAST searches identified a single fgf18gene in stickleback, medaka and both pufferfishand the topology of the fgf18 clade suggests thatfgf18b may have been lost in the lineage leading tothese species (Jovelin et al., 2007). In contrast tothe four percomorph species, zebrafish has twofgf18 genes. A number of sfgf18 CNEs are foundassociated only with zebrafish fgf18a (b, c. d, e, g,h, i, j, l, n, o, q), a few only with zfgf18b (a, f ), andseveral with both (k, m, r) (Fig. 12). This pattern

Fig. 11. Distribution of CNEs among Fgf17 orthologs. The low number of CNEs is likely owing to micro-rearrangementsaround the Fgf17 locus disrupting synteny of coding and noncoding sequences.

R. JOVELIN ET AL.16


of reciprocal distribution of CNEs is what wouldbe expected if the single stickleback fgf18 generepresented much of the ancestral distribution ofCNEs and they had become distributed as pre-dicted by subfunctionalization in the zebrafishlineage after gene duplication.

The high density of CNEs surrounding fgf18suggests that fgf18b may have been lost shortlyafter the separation of the lineages leading tozebrafish and stickleback and that CNEs may havebeen maintained by selection in lineages in which

only one fgf18 gene subsisted. However, in zebra-fish, which retained both fgf18 duplicates, mostCNEs found around fgf18 of species retaining justone fgf18 gene have been partitioned betweenfgf18a and fgf18b (Fig. 12).

fgf24: The high density of CNEs in proximity tofgf24 orthologs in all tested species (Fig. 13) isconsistent with our inference that one fgf24duplicate was lost before the divergence of thelineages leading to zebrafish and stickleback(Jovelin et al., 2007). Two particularly strong

Fig. 12. Most of the CNEs conserved among stickleback, medaka and pufferfish fgf18 are partitioned between fgf18a andfgf18b in zebrafish, consistent with the DDC model (Force et al., ’99). The high density of CNEs in teleost species having a singlefgf18 gene suggests that the second fgf18 duplicate was lost shortly after the R3 duplication. The region lacking CNEs intetraodon is owing to a gap in the sequence alignment.

Fig. 13. The high density of CNEs among teleost fgf24 orthologs suggests that the second fgf24 duplicate was lost shortlyafter the R3 duplication before the partitioning of fgf24 subfunctions. Medaka fgf24 is located on a short scaffold, hence theapparent lack of conservation further away from fgf24.



CNEs (labeled a and d) lie 30 and 50 respectively tofgf24 and two others (b and c) are within introns ofthe fgf24 gene. We checked to see if Fgfd-relatedCNEs could be detected in the location predictedfor the missing human FGF24, but found none.

fgf8: In stickleback and zebrafish, fgf8 dupli-cates exhibit common and distinct expressionpatterns, and comparisons between fgf8 orthologsshow that partitioning of most ancestral functionsoccurred before the divergence of these species(Jovelin et al., 2007). Thus, we should expectregulatory elements to be largely shared betweenfgf8 orthologs in stickleback and zebrafish ratherthan being species-specific.

For comparisons based on stickleback fgf8a, atleast four CNEs were shared with the humanFGF8 region (Fig. 14A). These included elementsin introns of fbxw4 (CNEs a, b, and d in Fig. 14A)

and an element in an intron of fgf8a itself(element h). Among the five stickleback fgfDgenes, the fgf8 comparisons were the only onesthat revealed any CNEs shared with human. All ofthe elements sfgf8a shared with human were alsoshared with zebrafish. In addition, zebrafish fgf8ashared with stickleback fgf8a several elements inthe intergenic spaces 30 and 50 to fgf8a. As notedfor other fgfD genes, stickleback fgf8a had moreand longer CNEs shared with percomorphs thanwith zebrafish. With few exceptions (see elementd, Fig. 14A), the percomorphs had very similarCNE patterns.

The comparison of the stickleback fgf8b regionto that of human revealed a CNE 30 to thestickleback fgf8b gene (element b, Fig. 14B) sharedwith other teleosts except T. nigroviridis. A CNEin an intron of fgf8b (element c in Fig. 14B) was

Fig. 14. Distribution of CNEs among Fgf8 orthologs. CNEs are partitioned in unique sets between teleost fgf8a (A) andfgf8b (B), reflecting the partitioning of expression domains between fgf8 duplicates (this work and (Jovelin et al., 2007)). Fewerfgf8b CNEs are shared with human FGF8.

R. JOVELIN ET AL.18


shared with zebrafish and fugu. A large number ofCNEs located 50 to fgf8b (e.g. element g) were heldin common by several teleosts that were not seenin the comparison with zebrafish.

Therefore, the overall pattern of conservation ofnoncoding sequences in the proximity of fgf8correlates with the evolutionary history of fgf8duplicates and their subsequent functional diver-sification: fgf8a and fgf8b orthologs have uniquesets of CNEs, some of which are shared withhuman as it would be expected if subfunctionshave been partitioned among paralogs.

DISCUSSION

Vertebrate FgfD genes originated asohnologs in R1 and R2

To infer the evolutionary history of expressiondomains and conserved noncoding sequences, it isessential to understand the phylogenetic relation-ships of the four FgfD genes (Suppl. Fig. S4). Thephylogenetic tree of these genes shows a history of(((Fgf8, Fgf17) Fgf18) Fgf24) (Jovelin et al., 2007).Several other data sets, however, support theconclusion that the four FgfD genes arose in theR1 and R2 WGD events. Evidence supporting thisconclusion comes from four independent sources:(1) The chromosome region surrounding theCiona FgfD gene contains genes with humanorthologs mainly on human chromosome seg-ments containing FGF8, FGF17, and FGF18, aswell as part of Hsa2p. (2) An automated search forparalogous chromosome segments around FGF8revealed chromosome segments surroundingFGF17, FGF18, and a portion of Hsa2p. (3) Theelephant shark has a copy of Fgf24, demonstratingthe origin of this gene before the divergence ofteleost and tetrapod genomes. (4) The immediategenomic neighborhood of the teleost fgf24 gene ispreserved in the human genome with the ratherprecise excision of an ortholog to fgf24 and itsnearest neighbor npm4. These data support theconclusion that R1 and R2 produced four ohnologsrelated as ((Fgf8, Fgf17)(Fgf18, Fgf24)) and thatthe tetropod Fgf24 gene went missing fromtetrapod genomes.

Comparative analysis of expressionpatterns

After the R2 genome duplication event, the fourFgfD genes would have been preserved either bysubfunctionalization (the duplicate preservationby reciprocal partitioning of ancestral gene sub-

functions) or neofunctionalization (the origin ofnovel domains) (Force et al., ’99; Hughes, ’99;Stoltzfus, ’99; Postlethwait et al., 2004; Chainet al., 2008; Conant and Wolfe, 2008), andsubsequent to preservation, additional expressiondomains or other functions could have been lostor ‘‘invented’’. To analyze these events, weinvestigated expression patterns of FgfD genes intwo teleosts and compared them with publisheddata available on their orthologs in mouse(Suppl. Fig. S4).

The single FgfD gene present in stem olfactores(urochordates1vertebrates) was an ancestorof Fgf8/17/18 in the ascidian urochordateC. intestinalis and its expression pattern serves asan outgroup to the vertebrate FgfD genes. InCiona, Fgf8/17/18 is expressed in the embryonicCNS coincident with En and Pax2/5/8 (Ikuta andSaiga, 2007). Correspondingly, the vertebrateorthologs of these three Ciona genes (Fgf8, Fgf17,Fgf18, En1, En2, Pax2, Pax5, and Pax8) are co-expressed in the vertebrate CNS at the midbrain–-hindbrain boundary (MHB) (Joyner, ’96). Althoughurochordates may lack an MHB organizer secon-darily (Canestro et al., 2005; Ikuta and Saiga,2007), this gene set appears to constitute an ancientregulatory cassette in chordate CNS development.The expression of all four vertebrate FgfD ohnologsin the MHB region in at least one vertebratespecies (Suppl. Fig. S4) suggests that this was anancient expression domain existing before thedivergence of urochordates and vertebrates andthat this expression domain was retained by allfour genes after the R1 and R2 WGD events at thebase of the vertebrate radiation.

Besides its expression in the CNS, Fgf8/17/18 isalso expressed in C. intestinalis in mesodermalcells lateral and anterior to the CNS expressiondomain, as well as cells at the tip of the tail (Imaiet al., 2002; Ikuta and Saiga, 2007). The relation-ship of these two urochordate domains to similardomains in the vertebrate somitic and unsegmen-ted mesodermal expression and tail bud domain isuncertain. In the cephalochordate amphioxus,Fgf8/17/18 is expressed transiently in the rostralCNS and in the pharyngeal endoderm, whichare likely conserved domains with vertebrates(Muelemans and Bonner-Fraser, 2007).

A substantial number of morphological innova-tions or elaborations occurred after the divergenceof urochordates and vertebrates but before thedivergence of bony fishes into ray-fin and lobe-finclades. These include the elaboration of neural crestand placodes, the invention of bone, and the origin



of paired appendages (Gans and Northcutt, ’83;Bassham and Postlethwait, 2005; Jeffery, 2006;Sauka-Spengler and Bronner-Fraser, 2008). Alsooccurring at about this time period were two roundsof WGD, R1 and R2 (Kortschak et al., 2001; Ornitzand Itoh, 2001; Lundin et al., 2003; Dehal andBoore, 2005; Garcia-Fernandez, 2005; Bourlat et al.,2006; Delsuc et al., 2006; Guder et al., 2006; Jacoband Lum, 2007; Kitisin et al., 2007; Sundstromet al., 2008; Yu et al., 2008) that we show here arehighly likely to have produced the four FgfDohnologs (Fgf8, Fgf17, Fgf18, and Fgf24). Anyexpression domain that appears in the orthologs ofthree or four of these four genes in at least one of thethree species stickleback, zebrafish, and mouse, ishighly likely to have been a site of expression in thelast common ancestor of all bony fishes.

In addition to the MHB and somite/unsegmen-ted mesoderm, at least one of the three investi-gated vertebrates (stickleback, zebrafish, mouse)expresses each FgfD gene in the retina, pharyngealarches, somites, and otic vesicle (Suppl. Fig. S4). Weconclude that these expression domains wereprobably present in the four orthologous FgfDgenes in the last common ancestor of ray-fin andlobe-fin fish and were differentially lost fromvarious genes in different lineages. Because thesedomains were in common in all four FgfD genes,we infer that they were also in common in theunduplicated FgfD gene in the lineage leading tovertebrates before R1 and R2. An alternativehypothesis is that these shared expression do-mains evolved independently but convergentlymultiple times, a possibility that seems unlikelybased on parsimony. If the hypothesis is true thatexpression domains shared by all four FgfD geneswere present in the last pre-R1-duplication FgfDgene, then that vertebrate ancestor would havehad expression domains appropriate for eyes, ears,pharyngeal arches, somites, and MHB. Availabledata for FgfD expression patterns in cephalochor-dates and urochordates do not contradict thisexplanation (Imai et al., 2002; Ikuta and Saiga,2007); Muelemans and Bonner-Fraser, 2007).

Expression domains in the dorsal diencephalonand the apical epidermal ridge (AER) of thepectoral fin/limb bud are shared by Fgf8 andFgf24 (Figs. 6, 7). Furthermore, the examinationof mutant phenotypes shows that Fgf8 is requiredfor paired pectoral appendage development intetrapods, which lack an ortholog of fgf24, andreciprocally, fgf24 is required for paired pectoralappendage development in zebrafish but fgf8a isnot (Reifers et al., ’98; Lewandoski et al., 2000;

Draper et al., 2003) (fgf8b is not expressed in theAER of fish pectoral appendages (Jovelin et al.,2007)). If the R1, R2 evolutionary history of thefour FgfD genes is correct, then the pre-duplica-tion FgfD gene should have had an expressiondomain corresponding to the dorsal diencephalonand AER domains in modern vertebrates. Accord-ing to this hypothesis, the last ancestor before thevertebrate genome duplications would thus havehad an expression domain for the AER, eventhough, paradoxically, it is unlikely that thatorganism had paired appendages (Coates, ’94).Thus, the ancestral function of this expressiondomain would have been co-opted into appendagedevelopment after paired appendages began toevolve in early vertebrates, and it may have beenused in the development of unpaired midline fins(Freitas et al., 2006). Alternatively, one or both ofthese expression domains could have evolvedconvergently as neofunctionalizations by the twogenes Fgf8 and Fgf24 well after the R1 and R2events during the evolution of paired fins.

Some additional expression domains are displayedby just two of the four FgfD genes (Suppl. Fig. 4).Expression in tailbud is shared by Fgf8 and Fgf17 andin unsegmented mesoderm and pancreas by Fgf18and Fgf24. This distribution might be predicted bythe (Fgf8, Fgf17)(Fgf18, Fgf24) lineage hypothesis ifthese domains had been a part of the repertoire of theFgf8/17 and Fgf18/24 genes after the R1 WGD event.As mentioned above, the tailbud expression domainmay be related to the expression domain at the tip ofthe tail in Ciona and inherited from stem olfactores(Imai et al., 2002; Ikuta and Saiga, 2007). In addition,both Fgf8 and Fgf18 share expression in the heartand Fgf17 is expressed in zebrafish heart regenera-tion, consistent with either inheritance from the FgfDgene of stem olfactores followed by loss in the Fgf24clade or, less parsimoniously, independent acquisitionby convergent evolution.

Several expression domains appear only on singleclades in the tree. These include segmental plate onlyin teleost fgf17 genes, telencephalon only in Fgf8 ofteleosts and tetrapods, olfactory epithelium only inteleost fgf8 genes, and spinal neurons only in fgf18aof teleosts. The most parsimonious explanation ofthese domains is that they represent neofunctiona-lization events in the corresponding clades.

Correlation of CNEs and expressiondomains

Because many CNEs are shared by fgf8a andother zebrafish FgfD genes (Fig. 10C), we can

R. JOVELIN ET AL.20


attempt to correlate individual CNEs with expres-sion domains in embryonic zebrafish that aresummarized in Supplementary Fig. 4. The firstconclusion is that none of the CNEs identified wasunique for a single tissue. CNE q is shared only byfgf8a, fgf18a, and fgf24, and only these genesexpress in the olfactory organ, making this acandidate regulatory element for contributing toolfactory expression, although genes possessingCNE q are expressed in tissues in addition to theolfactory epithelium. The dorsal diencephalonexpresses only fgf8a, fgf17, and fgf24, and thesegenes share CNEs labeled b, d, i, k, l, n, and w.(Fig. 10C); of these shared elements, i, k, l, and ware absent from paralogs that do not showexpression in the dorsal diencephalon, so theseare candidate CNEs for positive regulatory ele-ments contributing to expression in the dorsaldiencephalon of zebrafish embryos. Elements e, g,h, p, t, x, and z are found in both fgf8a and fgf24and in none of the other fgfD genes, and only thesetwo genes are expressed in the AER at 2 dpf,making these elements candidates for controllingexpression in the apical ectodermal ridge of thepectoral fin bud. For the retina at 16 hpf, onlyfgf18b lacks expression, and element d is the onlyelement possessed by all zebrafish fgfD genesexcept fgf18b, and so element d is a candidate forthe early expression in the retina. Other factorsmay turn on fgf18b in the retina at pharyngulastages. Genes expressed in the heart and thespinal cord neurons have in common the same setof elements (b, c, d, m, n, q, r, s, u, and v). Theunsegmented mesenchyme, segmental plate, andpancreas each have unique CNE combinations notshared with other tissues. Because vertebrates canconserve regulatory function without conservingsequence similarity (Fisher et al., 2006), thesignificance of these conserved noncoding regionsmust be tested by experiment to draw firmconclusions.

CNEs among paralogs reflect theevolutionary history of the FgfD subfamily

Gene fates after duplication include the loss ofone of the two paralogs by nonfunctionalization,or the divergence of gene function by the acquisi-tion of new function(s) (neofunctionalization)(Ohno, 1970) or the partitioning of ancestralfunctions (subfunctionalization) (Force et al., ’99;Hughes, ’99; He and Zhang, 2005). Subfunctionpartitioning occurs by the complementary fixationof otherwise deleterious mutations in ancestral

control elements of gene duplicates. This modelessentially follows neutral evolution, andtherefore one might expect that subfunctionpartitioning should increase with the time ofcoexistence between duplicates within a lineage.Consequently, the Duplication–Degeneration–Complementation (DDC) model predicts that moreancestral functions should have been partitionedbetween paralogs that have coexisted over longperiods of evolutionary time than in cases whereduplicates have existed for a shorter time period orwhere duplicates rapidly became reduced to singlecopy after the duplication event.

Examples of subfunction partitioning can befound for a variety of gene duplicates (e.g. Listeret al., 2001; Cresko et al., 2003; Liu et al., 2005).The FgfD subfamily in teleosts provides never-theless a rare opportunity to test predictions of theDDC model because its members share an originthrough the R1, R2, and R3 genome duplicationevents, and because since R3, different FgfDsubfamily members have spent different amountsof time as duplicate copies in various lineages. Thefgf24 and fgf17 genes apparently became singlecopy before the divergence of zebrafish andstickleback lineages; the fgf18a and fgf18b para-logs co-existed from R3 to the present in zebrafish,but became single copy in the stickleback lineageafter it diverged from the zebrafish lineage; andthe fgf8a and fgf8b paralogs have co-existed fromR3 to the present in several lineages. In addition,CNEs identified using phylogenetic footprinting inthe proximity of the Fgf8 locus have been shown tohave regulatory activity in mouse (Beermannet al., 2006) and in zebrafish (Inoue et al., 2006).Moreover, regulatory modules of fgf8 functionextend to intronic sequences of fgf8 neighbors(Kikuta et al., 2007). It is therefore likely thatCNEs identified around other Fgf genes have afunctional role as well.

The large number of long CNEs detected amongthe three human FGFD genes contrasts with thefewer shorter CNEs found in zebrafish andstickleback. What could be responsible for thisdifference? It is not mere antiquity because thelast common ancestor of the zebrafish fgf8 andfgf18a genes was in fact the same gene as the lastcommon ancestor of the human FGF8 and FGF18genes. Thus, either the human lineage preservedmore broad CNEs than teleost lineages, perhapsamplifying them locally by tandem duplication,or the zebrafish lineage lost more CNEs afterthe divergence of human and zebrafish lineages.The pattern of just a few, short, but nevertheless



well-conserved CNEs within stickleback and zeb-rafish genomes may relate to the additional WGDin the teleost lineage not suffered by the tetrapodlineage (R3). Post-R3 fgfD paralogs may haveexperienced greater relaxation of selective con-straints than post-R2 FgfD genes. After bothevents, we would expect the partitioning ofexpression domains by fixation of complementarydeleterious mutations in regulatory elements(Force et al., ’99; Lynch a nd Force, 2000).

After R3, one fgf18 duplicate was lost after thedivergence of the lineages leading to zebrafish andstickleback resulting in one fgf18 gene in stickle-back, medaka, and pufferfish whereas zebrafishretained two fgf18 duplicates. Analysis of CNEs inthe proximity of the stickleback fgf18 locus showsextensive conservation between stickleback andother percomorph fish (Fig. 12), suggesting thatthe integrity of regulatory elements in thisgenomic region is important for fgf18 functionand has been maintained by selection in specieshaving a single fgf18 gene. In contrast, fewerCNEs are shared between zebrafish and thepercomorphs, and most CNEs have been parti-tioned between fgf18a and fgf18b in zebrafish.This pattern of shared and partitioned CNEsmakes two predictions. First, the amount andlevel of conservation among CNEs surroundingfgf18 in percomorphs suggest that little subfunc-tion partitioning may have occurred before thesecond fgf18 duplicate was lost in this lineage andthus fgf18 in these species may have retained mostof the ancestral pre-R3 fgf18 functions. Second,the pattern of shared CNEs between fgf18a andfgf18b in zebrafish, where both duplicates havebeen maintained over longer evolutionary time,suggests the partitioning of ancestral pre-R3 fgf18functions between paralogs. Supporting thesepredictions, stickleback fgf18 is expressed in theheart field precursor and in the MHB at mid-segmentation, whereas in zebrafish, these expres-sion domains are partitioned between fgf18a andfgf18b, respectively (Fig. 6). At the long pec stage,expression of stickleback fgf18 in the olfactoryepithelium and in the MHB is partitioned betweenzebrafish fgf18a and fgf18b (Fig. 7). Similarly, atthe early pharyngula stage, expression in thespinal chord neurons partitioned to zfgf18a(Suppl. Fig. S4). Specifically, one or more ofelements ‘‘a’’, ‘‘f’’, and ‘‘r’’ in Fig. 12 arecandidates for expression of fgf18b in the MHB,and elements ‘‘b–e’’, ‘‘g–j’’, and ‘‘o–q’’ are candi-dates for expression of fgf18a in the nose, spinalneurons, and pancreas, assuming all regulatory

elements act positively, although, of course somemay act to diminish expression in certain tissues.

Sequenced teleost genomes have a single fgf24and fgf17 gene, indicating that one duplicate ofeach locus was lost after R3, before the divergenceof the lineages leading to zebrafish and other fishspecies. The extensive number and length of CNEsaround all fgf24 orthologs also supports thishypothesis and suggests that regulatory controlelements of fgf24 have been maintained byselection (Fig. 13). By contrast, the lower numberof CNEs surrounding Fgf17 genes may be me-chanistically related to the multiple local chromo-some rearrangements disrupting syntenies of bothCNEs and Fgf17 neighboring genes betweenteleosts and human and within teleosts (Figs. 5,11). CNEs detached from their target genes bychromosomal rearrangements are expected to belost by genetic drift (Kikuta et al., 2007). Forinstance, it is noteworthy that Fgf17 expression isthe least conserved between mouse and teleostsamong FgfD members (Suppl. Fig. S4). Theexpression patterns of fgf17 orthologs in stickle-back and zebrafish, however, are well conserved,suggesting that important ancestral regulatoryelements may have been retained in both lineages,and CNEs ‘‘a–e’’ in Figure 11 are candidates forsuch elements.

Because one fgf24 duplicate was lost before thedivergence of zebrafish and percomorph lineagesand one fgf18 duplicate was lost after theselineages diverged, fgf24 duplicates originatingfrom R3 have coexisted over shorter evolutionarytime than fgf18 duplicates originating from thesame WGD before one copy of each gene was lost.Thus, we would expect the functions of fgf24orthologs to be more conserved than those of fgf18orthologs. Indeed, not only does zfgf24 sharemore and longer CNEs with its orthologs thanzfgf18a and zfgf18b share with their orthologs(Figs. 12 and 13) but zfgf24 and sfgf24 expressiondomains are remarkably conserved with onlyminor differences in the retina and MHB(Figs. 6S–V, 7J–K, 8S–V), whereaas fgf18 ortho-logs show expression differences in somites, retinaand ear (Figs. 6N, P, R, 8M–R).

In summary, this comparison of CNEs andexpression domains verifies predictions made bythe molecular mechanisms of subfunction parti-tioning and the DDC model (Force et al., ’99) forall teleost fgfD members for which synteny has notbeen disrupted by local chromosomal rearrange-ments. Additionally, these results identify a clearrelationship, predicted by the subfunctionalization

R. JOVELIN ET AL.22


model, between the length of time that paralogscoexist and their distribution of both CNEs andexpression domains. Importantly, this work iden-tifies a number of CNEs that are candidates forregulatory elements for specific expression do-mains in development. The testing of thesedomains for developmental functions is anongoing project.

ACKNOWLEDGMENTS

We thank Dong Liu for discussions on earmorphology, Amanda Rapp, Tim Mason and theUniversity of Oregon Zebrafish Facility for provid-ing animals and excellent fish care, and PohKheng Loi and Amber Selix of the HistologyFacility for sectioning. This work was supportedby the National Center for Research Resources(5R01RR020833) and National Institutes ofHealth (P01 HD22486); the contents of this studyare solely the responsibility of the authors and donot necessarily represent the official views ofNCRR or NIH. RJ is an Associate of the IGERTTraining Program for Development, Evolution,and Genomics, NSF DGE-9972830.

LITERATURE CITED

Abu-Issa R, Smyth G, Smoak I, Yamamura K, Meyers EN.2002. Fgf8 is required for pharyngeal arch and cardiovas-cular development in the mouse. Development 129:4613–4625.

Amemiya CT, Prohaska SJ, Hill-Force A, Cook A,Wasserscheid J, Ferrier DE, Pascual-Anaya J, Garcia-Fernandez J, Dewar K, Stadler PF. 2008. The amphioxusHox cluster: characterization, comparative genomics, andevolution. J Exp Zool B Mol Dev Evol 310:465–477.

Amores A, Force A, Yan YL, Joly L, Amemiya C, Fritz A,Ho RK, Langeland J, Prince V, Wang YL, Westerfield M,Ekker M, Postlethwait JH. 1998. Zebrafish hox clusters andvertebrate genome evolution. Science 282:1711–1714.

Andermann P, Weinberg ES. 2001. Expression of zTlxA,a Hox11-like gene, in early differentiating embryonicneurons and cranial sensory ganglia of the zebrafishembryo. Dev Dyn 222:595–610.

Bassham S, Postlethwait JH. 2005. The evolutionary historyof placodes: a molecular genetic investigation of thelarvacean urochordate Oikopleura dioica. Development132:4259–4272.

Beermann F, Kaloulis K, Hofmann D, Murisier F, Bucher P,Trumpp A. 2006. Identification of evolutionarily conservedregulatory elements in the mouse Fgf8 locus. Genesis44:1–6.

Bershtein S, Tawfik DS. 2008. Ohno’s model revisited:measuring the frequency of potentially adaptive mutationsunder various mutational drifts. Mol Biol Evol 25:2311–2318.

Blair JE, Hedges SB. 2005. Molecular phylogeny and diver-gence times of deuterostome animals. Mol Biol Evol22:2275–2284.

Boulet AM, Moon AM, Arenkiel BR, Capecchi MR. 2004. Theroles of Fgf4 and Fgf8 in limb bud initiation and outgrowth.Dev Biol 273:361–372.

Bourlat SJ, Juliusdottir T, Lowe CJ, Freeman R, Aronowicz J,Kirschner M, Lander ES, Thorndyke M, Nakano H,Kohn AB, Heyland A, Moroz LL, Copley RR, Telford MJ.2006. Deuterostome phylogeny reveals monophyletic chor-dates and the new phylum Xenoturbellida. Nature 444:85–88.

Brand M, Heisenberg CP, Jiang YJ, Beuchle D, Lun K, Furutani-Seiki M, Granato M, Haffter P, Hammerschmidt M, Kane DA,Kelsh RN, Mullins MC, Odenthal J, van Eeden FJ, Nusslein-Volhard C. 1996. Mutations in zebrafish genes affecting theformation of the boundary between midbrain and hindbrain.Development 123:179–190.

Brudno M, Do CB, Cooper GM, Kim MF, Davydov E,Green ED, Sidow A, Batzoglou S. 2003. LAGAN andMulti-LAGAN: efficient tools for large-scale multiple align-ment of genomic DNA. Genome Res 13:721–731.

Canestro C, Bassham S, Postlethwait J. 2005. Development ofthe central nervous system in the larvacean Oikopleuradioica and the evolution of the chordate brain. Dev Biol285:298–315.

Cao Y, Zhao J, Sun Z, Zhao Z, Postlethwait J, Meng A. 2004.fgf17b, a novel member of Fgf family, helps patterningzebrafish embryos. Dev Biol 271:130–143.

Catchen JM, Conery JS, Postlethwait JH. 2009. Automatedidentification of conserved synteny after whole genomeduplication. Genome Res. [Epub ahead of print].

Chain FJ, Ilieva D, Evans BJ. 2008. Duplicate gene evolutionand expression in the wake of vertebrate allopolyploidiza-tion. BMC Evol Biol 8:43.

Cholfin JA, Rubenstein JL. 2007. Patterning of frontalcortex subdivisions by Fgf17. Proc Natl Acad Sci USA 104:7652–7657.

Christoffels A, Koh EG, Chia JM, Brenner S, Aparicio S,Venkatesh B. 2004. Fugu genome analysis provides evidencefor a whole-genome duplication early during the evolution ofray-finned fishes. Mol Biol Evol 21:1146–1151.

Coates MI. 1994. The origin of vertebrate limbs. Dev Suppl169–180.

Conant GC, Wolfe KH. 2008. Turning a hobby into a job: howduplicated genes find new functions. Nat Rev Genet9:938–950.

Cresko WA, Yan YL, Baltrus DA, Amores A, Singer A,Rodriguez-Mari A, Postlethwait JH. 2003. Genome duplica-tion, subfunction partitioning, and lineage divergence: Sox9in stickleback and zebrafish. Dev Dyn 228:480–489.

Crow KD, Stadler PF, Lynch VJ, Amemiya C, Wagner GP.2006. The ‘‘fish-specific’’ Hox cluster duplication is coin-cident with the origin of teleosts. Mol Biol Evol 23:121–136.

Dehal P, Boore JL. 2005. Two rounds of whole genomeduplication in the ancestral vertebrate. PLoS Biol 3:e314.

Delsuc F, Brinkmann H, Chourrout D, Philippe H. 2006.Tunicates and not cephalochordates are the closest livingrelatives of vertebrates. Nature 439:965–968.

Delsuc F, Tsagkogeorga G, Lartillot N, Philippe H. 2008.Additional molecular support for the new chordate phylo-geny. Genesis 46:592–604.

Dichmann DS, Miller CP, Jensen J, Scott Heller R, Serup P.2003. Expression and misexpression of members of the FGF



and TGFbeta families of growth factors in the developingmouse pancreas. Dev Dyn 226:663–674.

Dominguez MH, Rakic P. 2008. Neuroanatomy of the FGFsystem. J Comp Neurol 509:141–143.

Draper BW, Stock DW, Kimmel CB. 2003. Zebrafish fgf24functions with fgf8 to promote posterior mesodermaldevelopment. Development 130:4639–4654.

Fischer S, Draper BW, Neumann CJ. 2003. The zebrafishfgf24 mutant identifies an additional level of Fgf signalinginvolved in vertebrate forelimb initiation. Development130:3515–3524.

Fisher S, Grice EA, Vinton RM, Bessling SL, McCallion AS.2006. Conservation of RET regulatory function from humanto zebrafish without sequence similarity. Science 312:276–279.

Force A, Lynch M, Pickett FB, Amores A, Yan YL,Postlethwait J. 1999. Preservation of duplicate genes bycomplementary, degenerative mutations. Genetics 151:1531–1545.

Frank DU, Fotheringham LK, Brewer JA, Muglia LJ,Tristani-Firouzi M, Capecchi MR, Moon AM. 2002. AnFgf8 mouse mutant phenocopies human 22q11 deletionsyndrome. Development 129:4591–4603.

Frazer KA, Pachter L, Poliakov A, Rubin EM, Dubchak I.2004. VISTA: computational tools for comparative geno-mics. Nucleic Acids Res 32:W273–W279.

Freitas R, Zhang G, Cohn MJ. 2006. Evidence that mechan-isms of fin development evolved in the midline of earlyvertebrates. Nature 442:1033–1037.

Gans C, Northcutt RG. 1983. Neural crest and the origin ofvertebrates: a new head. Science 220:268–274.

Garcia-Fernandez J. 2005. The genesis and evolution ofhomeobox gene clusters. Nat Rev Genet 6:881–892.

Guder C, Philipp I, Lengfeld T, Watanabe H, Hobmayer B,Holstein TW. 2006. The Wnt code: cnidarians signal theway. Oncogene 25:7450–7460.

Hamade A, Deries M, Begemann G, Bally-Cuif L, Genet C,Sabatier F, Bonnieu A, Cousin X. 2006. Retinoic acidactivates myogenesis in vivo through Fgf8 signalling. DevBiol 289:127–140.

He X, Zhang J. 2005. Rapid subfunctionalization accompaniedby prolonged and substantial neofunctionalization in dupli-cate gene evolution. Genetics 169:1157–1164.

Hedges SB. 2002. The origin and evolution of model organ-isms. Nat Rev Genet 3:838–849.

Hoegg S, Brinkmann H, Taylor JS, Meyer A. 2004. Phyloge-netic timing of the fish-specific genome duplication corre-lates with the diversification of teleost fish. J Mol Evol59:190–203.

Hubbard TJ, Aken BL, Beal K, Ballester B, Caccamo M,Chen Y, Clarke L, Coates G, Cunningham F, Cutts T,Down T, Dyer SC, Fitzgerald S, Fernandez-Banet J, Graf S,Haider S, Hammond M, Herrero J, Holland R, Howe K,Johnson N, Kahari A, Keefe D, Kokocinski F, Kulesha E,Lawson D, Longden I, Melsopp C, Megy K, Meidl P,Ouverdin B, Parker A, Prlic A, Rice S, Rios D,Schuster M, Sealy I, Severin J, Slater G, Smedley D,Spudich G, Trevanion S, Vilella A, Vogel J, White S,Wood M, Cox T, Curwen V, Durbin R, Fernandez-Suarez XM, Flicek P, Kasprzyk A, Proctor G, Searle S,Smith J, Ureta-Vidal A, Birney E. 2007. Ensembl 2007.Nucleic Acids Res 35:D610–D617.

Hughes AL. 1999. Adaptive evolution of genes and genomes.New York: Oxford University Press.

Ikuta T, Saiga H. 2007. Dynamic change in the expressionof developmental genes in the ascidian centralnervous system: revisit to the tripartite model and theorigin of the midbrain-hindbrain boundary region. Dev Biol312:631–643.

Ilagan R, Abu-Issa R, Brown D, Yang YP, Jiao K, Schwartz RJ,Klingensmith J, Meyers EN. 2006. Fgf8 is required foranterior heart field development. Development 133:2435–2445.

Imai KS, Satoh N, Satou Y. 2002. Region specific geneexpressions in the central nervous system of the ascidianembryo. Mech Dev 119:S275–S277.

Inoue F, Nagayoshi S, Ota S, Islam ME, Tonou-Fujimori N,Odaira Y, Kawakami K, Yamasu K. 2006. Genomic organi-zation, alternative splicing, and multiple regulatory regionsof the zebrafish fgf8 gene. Dev Growth Differ 48:447–462.

Itoh N. 2007. The Fgf families in humans, mice, and zebrafish:their evolutional processes and roles in development,metabolism, and disease. Biol Pharm Bull 30:1819–1825.

Itoh N, Konishi M. 2007. The zebrafish fgf family. Zebrafish4:179–186.

Itoh N, Ornitz DM. 2004. Evolution of the Fgf and Fgfr genefamilies. Trends Genet 20:563–569.

Jacob L, Lum L. 2007. Hedgehog signaling pathway. Sci STKE2007:cm6.

Jeffery WR. 2006. Ascidian neural crest-like cells:phylogenetic distribution, relationship to larval complexity,and pigment cell fate. J Exp Zoolog B Mol Dev Evol306:470–480.

Jovelin R, He X, Amores A, Yan YL, Shi R, Qin B, Roe B,Cresko WA, Postlethwait JH. 2007. Duplication and diver-gence of fgf8 functions in teleost development and evolution.J Exp Zool B Mol Dev Evol 308B:730–743.

Joyner AL. 1996. Engrailed, Wnt and Pax genes regulatemidbrain–hindbrain development. Trends Genet 12:15–20.

Kasbauer T, Towb P, Alexandrova O, David CN, Dall’armi E,Staudigl A, Stiening B, Bottger A. 2007. The Notch signalingpathway in the cnidarian Hydra. Dev Biol 303:376–390.

Kikuta H, Laplante M, Navratilova P, Komisarczuk AZ,Engstrom PG, Fredman D, Akalin A, Caccamo M, Sealy I,Howe K, Ghislain J, Pezeron G, Mourrain P, Ellingsen S,Oates AC, Thisse C, Thisse B, Foucher I, Adolf B, Geling A,Lenhard B, Becker TS. 2007. Genomic regulatory blocksencompass multiple neighboring genes and maintain con-served synteny in vertebrates. Genome Res 17:545–555.

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

Kitisin K, Saha T, Blake T, Golestaneh N, Deng M, Kim C,Tang Y, Shetty K, Mishra B, Mishra L. 2007. Tgf-Betasignaling in development. Sci STKE 2007:cm1.

Kortschak RD, Tamme R, Lardelli M. 2001. Evolutionaryanalysis of vertebrate Notch genes. Dev Genes Evol 211:350–354.

Kumar S, Hedges SB. 1998. A molecular timescale forvertebrate evolution. Nature 392:917–920.

Langenau DM, Palomero T, Kanki JP, Ferrando AA, Zhou Y,Zon LI, Look AT. 2002. Molecular cloning and develop-mental expression of Tlx (Hox11) genes in zebrafish (Daniorerio). Mech Dev 117:243–248.

Lepilina A, Coon AN, Kikuchi K, Holdway JE, Roberts RW,Burns CG, Poss KD. 2006. A dynamic epicardial injuryresponse supports progenitor cell activity during zebrafishheart regeneration. Cell 127:607–619.

R. JOVELIN ET AL.24


Lewandoski M, Sun X, Martin GR. 2000. Fgf8 signalling fromthe AER is essential for normal limb development. NatGenet 26:460–463.

Lister JA, Close J, Raible DW. 2001. Duplicate mitf genes inzebrafish: complementary expression and conservation ofmelanogenic potential. Dev Biol 237:333–344.

Liu Z, Xu J, Colvin JS, Ornitz DM. 2002. Coordination ofchondrogenesis and osteogenesis by fibroblast growth factor18. Genes Dev 16:859–869.

Liu RZ, Sharma MK, Sun Q, Thisse C, Thisse B, Denovan-Wright EM, Wright JM. 2005. Retention of the duplicatedcellular retinoic acid-binding protein 1 genes (crabp1a andcrabp1b) in the zebrafish genome by subfunctionalization oftissue-specific expression. FEBS J 272:3561–3571.

Lundin LG, Larhammar D, Hallbook F. 2003. Numerousgroups of chromosomal regional paralogies strongly indicatetwo genome doublings at the root of the vertebrates.J Struct Funct Genomics 3:53–63.

Lynch M, Force A. 2000. The probability of duplicategene preservation by subfunctionalization. Genetics 154:459–473.

Lynch VJ, Wagner GP. 2009. Multiple chromosomal rearran-gements structured the ancestral vertebrate Hox-bearingprotochromosomes. PLoS Genet 5:e1000349.

Matus DQ, Thomsen GH, Martindale MQ. 2007. FGFsignaling in gastrulation and neural development inNematostella vectensis, an anthozoan cnidarian. Dev GenesEvol 217:137–148.

Mayor C, Brudno M, Schwartz JR, Poliakov A, Rubin EM,Frazer KA, Pachter LS, Dubchak I. 2000. VISTA: visualiz-ing global DNA sequence alignments of arbitrary length.Bioinformatics 16:1046–1047.

Meulemans D, Bronner-Fraser M. 2007. Insights fromamphioxus into the evolution of vertebrate cartilage. PLoSONE 2:e787.

Meyer A, Schartl M. 1999. Gene and genome duplications invertebrates: the one-to-four (-to-eight in fish) rule and theevolution of novel gene functions. Curr Opin Cell Biol11:699–704.

Meyer A, Van de Peer Y. 2005. From 2R to 3R: evidence for afish-specific genome duplication (FSGD). Bioessays 27:937–945.

Moon AM, Capecchi MR. 2000. Fgf8 is required for outgrowthand patterning of the limbs. Nat Genet 26:455–459.

Naruse K, Tanaka M, Mita K, Shima A, Postlethwait J,Mitani H. 2004. A medaka gene map: the trace of ancestralvertebrate proto-chromosomes revealed by comparativegene mapping. Genome Res 14:820–828.

Nichols SA, Dirks W, Pearse JS, King N. 2006. Early evolutionof animal cell signaling and adhesion genes. Proc Natl AcadSci USA 103:12451–12456.

Oda H, Wada H, Tagawa K, Akiyama-Oda Y, Satoh N,Humphreys T, Zhang S, Tsukita S. 2002. A novel amphioxuscadherin that localizes to epithelial adherens junctions hasan unusual domain organization with implications forchordate phylogeny. Evol Dev 4:426–434.

Ohbayashi N, Shibayama M, Kurotaki Y, Imanishi M,Fujimori T, Itoh N, Takada S. 2002. FGF18 is required fornormal cell proliferation and differentiation during osteo-genesis and chondrogenesis. Genes Dev 16:870–879.

Ohno S. 1970. Evolution by gene duplication. New York:Springer.

Ohuchi H, Kimura S, Watamoto M, Itoh N. 2000. Involvementof fibroblast growth factor (FGF)18-FGF8 signaling in

specification of left-right asymmetry and brain and limbdevelopment of the chick embryo. Mech Dev 95:55–66.

Ornitz DM, Itoh N. 2001. Fibroblast growth factors. GenomeBiol 2:REVIEWS3005.

Peterson KJ, Lyons JB, Nowak KS, Takacs CM, Wargo MJ,McPeek MA. 2004. Estimating metazoan divergence timeswith a molecular clock. Proc Natl Acad Sci USA 101:6536–6541.

Picker A, Scholpp S, Bohli H, Takeda H, Brand M. 2002. Anovel positive transcriptional feedback loop in midbrain-hindbrain boundary development is revealed throughanalysis of the zebrafish pax2.1 promoter in transgeniclines. Development 129:3227–3239.

Popovici C, Roubin R, Coulier F, Birnbaum D. 2005. Anevolutionary history of the FGF superfamily. Bioessays27:849–857.

Postlethwait JH. 2007. The zebrafish genome in context:ohnologs gone missing. J Exp Zool B Mol Dev Evol308:563–577.

Postlethwait JH, Yan YL, Gates MA, Horne S, Amores A,Brownlie A, Donovan A, Egan ES, Force A, Gong Z,Goutel C, Fritz A, Kelsh R, Knapik E, Liao E, Paw B,Ransom D, Singer A, Thomson M, Abduljabbar TS, Yelick P,Beier D, Joly JS, Larhammar D, Rosa F, Westerfield M,Zon LI, Johnson SL, Talbot WS. 1998. Vertebrate genomeevolution and the zebrafish gene map. Nat Genet 18:345–349.

Postlethwait J, Amores A, Cresko W, Singer A, Yan YL. 2004.Subfunction partitioning, the teleost radiation and theannotation of the human genome. Trends Genet 20:481–490.

Reifers F, Bohli H, Walsh EC, Crossley PH, Stainier DY,Brand M. 1998. Fgf8 is mutated in zebrafish acerebellar(ace) mutants and is required for maintenance of mid-brain–hindbrain boundary development and somitogenesis.Development 125:2381–2395.

Reifers F, Adams J, Mason IJ, Schulte-Merker S, Brand M.2000a. Overlapping and distinct functions provided by fgf17,a new zebrafish member of the Fgf8/17/18 subgroup of Fgfs.Mech Dev 99:39–49.

Reifers F, Walsh EC, Leger S, Stainier DY, Brand M. 2000b.Induction and differentiation of the zebrafish heart requiresfibroblast growth factor 8 (fgf8/acerebellar). Development127:225–235.

Sato T, Joyner AL, Nakamura H. 2004. How doesFgf signaling from the isthmic organizer induce midbrainand cerebellum development? Dev Growth Differ46:487–494.

Satou Y, Imai KS, Satoh N. 2002. Fgf genes in the basalchordate Ciona intestinalis. Dev Genes Evol 212:432–438.

Sauka-Spengler T, Bronner-Fraser M. 2008. Evolution of theneural crest viewed from a gene regulatory perspective.Genesis 46:673–682.

Sidow A. 1996. Gen(om)e duplications in the evolution of earlyvertebrates. Curr Opin Genet Dev 6:715–722.

Stoltzfus A. 1999. On the possibility of constructive neutralevolution. J Mol Evol 49:169–181.

Sun X, Mariani FV, Martin GR. 2002. Functions of FGFsignalling from the apical ectodermal ridge in limb devel-opment. Nature 418:501–508.

Sundstrom G, Larsson TA, Larhammar D. 2008. Phylogeneticand chromosomal analyses of multiple genefamilies syntenic with vertebrate Hox clusters. BMC EvolBiol 8:254.



Technau U, Rudd S, Maxwell P, Gordon PM, Saina M,Grasso LC, Hayward DC, Sensen CW, Saint R, Holstein TW,Ball EE, Miller DJ. 2005. Maintenance of ancestralcomplexity and non-metazoan genes in two basal cnidarians.Trends Genet 21:633–639.

Venkatesh B, Kirkness EF, Loh YH, Halpern AL, Lee AP,Johnson J, Dandona N, Viswanathan LD, Tay A, Venter JC,Strausberg RL, Brenner S. 2007. Survey sequencing andcomparative analysis of the elephant shark (Callorhinchusmilii) genome. PLoS Biol 5:e101.

Vienne A, Pontarotti P. 2006. Metaphylogeny of 82 genefamilies sheds a new light on chordate evolution. Int J BiolSci 2:32–37.

Wada H, Okuyama M, Satoh N, Zhang S. 2006. Molecularevolution of fibrillar collagen in chordates, with implicationsfor the evolution of vertebrate skeletons and chordatephylogeny. Evol Dev 8:370–377.

Wall DP, Fraser HB, Hirsh AE. 2003. Detecting putativeorthologs. Bioinformatics 19:1710–1711.

Wittbrodt J, Meyer A, Schartl M. 1998. More genes in fish?Bioessays 20:511–515.

Wolfe K. 2000. Robustness—it’s not where you think it is. NatGenet 25:3–4.

Wotton KR, Weierud FK, Dietrich S, Lewis KE. 2008.Comparative genomics of Lbx loci reveals conservation ofidentical Lbx ohnologs in bony vertebrates. BMC Evol Biol8:171.

Xu J, Liu Z, Ornitz DM. 2000. Temporal and spatial gradientsof Fgf8 and Fgf17 regulate proliferation and differentiationof midline cerebellar structures. Development 127:1833–1843.

Yan YL, Miller CT, Nissen R, Singer A, Liu D, Kirn A, DraperB, Willoughby J, Morcos PA, Amsterdam A, Chung BC,Westerfield M, Haffter P, Hopkins N, Kimmel C, Post-lethwait JH. 2002. A zebrafish sox9 gene required forcartilage morphogenesis. Development 129:5065–5079.

Yu WP, Rajasegaran V, Yew K, Loh WL, Tay BH,Amemiya CT, Brenner S, Venkatesh B. 2008. Elephantshark sequence reveals unique insights into the evolution-ary history of vertebrate genes: A comparative analysis ofthe protocadherin cluster. Proc Natl Acad Sci USA105:3819–3824.

R. JOVELIN ET AL.26


Evolution of developmental regulation in the vertebrate ...uoneuro.uoregon.edu › ~cristian ›...

Documents

Transcript of Evolution of developmental regulation in the vertebrate ...uoneuro.uoregon.edu › ~cristian ›...