Loss of fish actinotrichia proteins and the fin-to-limbtransitionJing Zhang1, Purva Wagh1, Danielle Guay1, Luis Sanchez-Pulido2, Bhaja K. Padhi3, Vladimir Korzh4,Miguel A. Andrade-Navarro5 & Marie-Andree Akimenko1
The early development of teleost paired fins is strikingly similar tothat of tetrapod limb buds and is controlled by similar mechan-isms1,2. One early morphological divergence between pectoral finsand limbs is in the fate of the apical ectodermal ridge (AER), thedistal epidermis that rims the bud. Whereas the AER of tetrapodsregresses after specification of the skeletal progenitors3, the AER ofteleost fishes forms a fold that elongates4,5. Formation of the fin foldis accompanied by the synthesis of two rows of rigid, unmineralizedfibrils called actinotrichia, which keep the fold straight6,7 and guidethe migration of mesenchymal cells within the fold5,8. The actino-trichia are made of elastoidin, the components of which, apart fromcollagen, are unknown. Here we show that two zebrafish proteins,which we name actinodin 1 and 2 (And1 and And2), are essentialstructural components of elastoidin. The presence of actinodinsequences in several teleost fishes and in the elephant shark(Callorhinchus milii, which occupies a basal phylogenetic position),but not in tetrapods, suggests that these genes have been lost duringtetrapod species evolution. Double gene knockdown of and1 andand2 in zebrafish embryos results in the absence of actinotrichiaand impaired fin folds. Gene expression profiles in embryos lackingand1 and and2 function are consistent with pectoral fin truncationand may offer a potential explanation for the polydactyly observedin early tetrapod fossils. We propose that the loss of both actinodinsand actinotrichia during evolution may have led to the loss oflepidotrichia and may have contributed to the fin-to-limb transition.
The structure and composition of the actinotrichia have intriguedscientists since the end of the nineteenth century because of theproperties of the component elastoidin9,10, a term coined to designatethe ensemble of proteins found in shark fins that has propertiesintermediate between those of collagen and elastin10. Elastoidin iscomposed of collagen and of a non-collagenous protein that hasnot yet been characterized11–13.
In a screen for genes differentially expressed during zebrafish finregeneration, we identified two genes, 2-F11 (LOC794315, EntrezGeneID 794315) and 2-H06 (zgc:175106, Entrez GeneID 566702),encoding proteins with unknown functions that are expressed specifi-cally in fins14. Here we show that these genes code for structural proteinsthat belong to a family of elastoidin components. We therefore namedthis family actinodin (and), for actinotrichia plus elastoidin.
The and1 (2-F11) and and2 (2-H06) genes encode proteins (Sup-plementary Fig. 1 and Fig. 1) which contain a signal peptide suggest-ing that they are secreted. They also contain two potential cleavagesites (RXRR; X, any amino acid) for convertases, which are enzymesthat catalyse the maturation of proprotein substrates in the secretorypathway15. Both And1 and And2 are rich in tyrosine (8.25% and5.71%, respectively), one of the criteria previously used to distinguish
elastoidin from collagen11. And1 and And2 respectively com-prise eight and ten repeats of the nine-amino-acid motif C(N/D)PXXDPXC (Fig. 1 and Supplementary Fig. 1). The presence oftwo proline and two cysteine residues suggests that each repeatmay form a loop. We found, in the zebrafish genome database, twoadditional and genes, and3 (si:ch211-237l4.2, Entrez GeneID561047) and and4 (zgc:172167, Entrez GeneID 100192207), whichare predicted to encode proteins with fewer repeats than And1 andAnd2 (Fig. 1 and Supplementary Fig. 1). Given the large and variablenumber of repeats in actinodin proteins, we postulate that repeatsmay stack on each other to form an elongated domain16.
Expression of the four zebrafish and genes correlates spatially andtemporally with formation of the actinotrichia in both the median finfold that develops around the posterior part of the embryo, starting at24 h post-fertilization (h.p.f.), and in the pectoral fin buds, starting at36 h.p.f. (Fig. 2 and Supplementary Figs 2 and 3). Transcripts are firstfound in the ectodermal cells of the fold, with the notable exceptionsof the cleft cells and the AER of the median and pectoral fin folds,respectively (Fig. 2 and Supplementary Figs 2 and 3). They are alsofound in mesenchymal cells invading the fold (Fig. 2 and Supplemen-tary Figs 2 and 3). Actinotrichia become visible shortly after the onsetof and expression (Fig. 2 and Supplementary Figs 2 and 3). Analysisof the expression of and genes in regenerating fins of adults furthersupports their having a role in actinotrichia formation (Supplemen-tary Fig. 4). We confirmed the presence of and gene products inactinotrichia of fin regenerates using an antibody that was raisedagainst the DVECLQYHLRAAYGYR peptide, which is highly con-served between And1 and And2 (Fig. 2).
1CAREG, Department of Biology, University of Ottawa, Ottawa, Ontario K1N 6N5, Canada. 2Functional Genomics Unit, Department of Physiology, Anatomy and Genetics, University ofOxford, Oxford OX1 3QX, UK. 3Environmental Health Science and Research Bureau, Health Canada, Ottawa, Ontario K1A 0L2, Canada. 4Institute of Molecular and Cell Biology, 138673Singapore. 5Max Delbruck Center for Molecular Medicine, 13125 Berlin, Germany.
Zebrafish
Elephant shark
And1
And2
And3
And4
Signal peptideConserved region
Repeat
Sequence unknown
And
Figure 1 | Schematic representations of the zebrafish and elephant sharkactinodin proteins. The representation of the zebrafish actinodin proteinswas deduced from sequences shown in Supplementary Fig. 1. The partialpredicted sequence for elephant shark actinodin was obtained from genomicclone AAVX01004882.1 of the genome sequencing survey database19.
Actinotrichia are found in the fins of developing and adultActinopterygii (ray-finned fishes including the teleostei) and in thedeveloping fins of the lobe-finned fishes. They are homologous to theceratotrichia of Chondrichthyes (cartilaginous fishes)17. They do notexist in tetrapod limbs. Database searches revealed the existence ofgenes closely related to and1–4 in other teleosts (ref. 18 and Sup-plementary Fig. 5) but not in any tetrapod species. The elephantshark, a cartilaginous fish of the holocephalian lineage, representsan interesting outgroup with which to study the evolution of the
teleosts and tetrapods19. We searched the survey sequence19 of theelephant shark genome and identified one incomplete and-likesequence (Fig. 1 and Supplementary Fig. 5). The elephant sharkdatabase is relatively small and, therefore, its genome may containadditional and-related genes. The existence of an elephant shark andgene indicates that and is an ancient vertebrate gene family that hasbeen lost or has highly diverged in tetrapods.
The role of and1 and and2 during fin development was investi-gated using morpholino-mediated gene knockdown. Major findefects were not observed in morphants for either and1 or and2(Fig. 3a–c and Supplementary Fig. 6). However, co-injection of both
MFF
NTNT
MFF
PFF
and1, 72 h.p.f.
72 h.p.f.
72 h.p.f.
and1, 72 h.p.f.
a
ab
Mergede
e
a
ab
Anti-And1
e
e
Anti-And1 Anti-And1 + And1 peptide
---50 kDa
a b
c d
e
f g
h i
j
k l
Figure 2 | Expression of and1 correlates with the growth of actinotrichia.a–e, In situ hybridization of and1 riboprobe in the median fin fold (MFF)(a, b) and in pectoral fin buds (c–e) of 72-h.p.f. embryos. Note the absence ofexpression in the cleft cells of the MFF and the AER of the pectoral fin buds(arrows in a and c). Transverse sections close to the posterior end of theembryo (b) or in the distal pectoral fin fold (PFF) (d) show and1 expressionin epidermal cells (red arrow) and in mesenchymal cells (blue arrow). e, Amore proximal section of a pectoral fin shows and1 expression only inepidermal cells. f–i, At 72 h.p.f., actinotrichia occupy the entire length of theMFF (f) and the PFF (h), and form a tight array of fibrils. NT, notochord.g, i, Magnified views of the boxed areas in f (g) and h (i). Red lines in f andi indicate actinotrichia orientation. j, Western blot analysis of total proteinsof larvae at 3 d post-fertilization (d.p.f.) with an anti-And1 antibody (leftlane) reveals a band of 50 kDa, which is slightly larger than the expectedmolecular mass of the predicted mature peptide for And1 (46.2 kDa) andAnd2 (45.6 kDa). The 50-kDa band is lost in the presence of excess And1peptide (right lane), demonstrating antibody specificity.k, Immunofluorescence of a sectioned fin regenerate at 6 d post-amputation(d.p.a.), showing a strong anti-And1 antibody staining of the twosymmetrical rows of actinotrichia (white arrows). a, actinotrichia; b,blastema; e, epidermis. l, Merged image of k and bright-field image. Scalebars, 40 mm (a); 20 mm (b–f, h, k); 5mm (g, i).
e MFF
NT
3 d.p.f., control
f
NT
MFF
3 d.p.f., co-MO
3 d.p.f., control 3 d.p.f., co-MO
g
PFF
h
PFF
and1 MO Co-MOand2 MO
---50 kDa
i
6 d.p.a., control
e
e
a
a
b
j
c
Control
NT
MFFd
Co-MO
NT
MFF
b
and2 MO
NT
MFFa
and1 MO
NT
MFF
6 d.p.a., co-MO
ke
e
a
a
b
Figure 3 | Double gene knockdown of and1 and and2 leads to an absence ofactinotrichia and reduced or absent fin folds. a–d, Picrosirius-red stainingof the caudal part of 4-d.p.f. larvae: and1 morphant (6 ng morpholino (MO);a), and2 morphant (3 ng MO; b), control larva (c), and1 1 and2 doublemorphants (co-MO; d). In a–c, actinotrichia are still visible; the black linesindicate their orientations and lengths. In d, actinotrichia do not form andthe fin fold is reduced, granular and curled. e–h, The ET37 enhancer-trapline, expressing green fluorescent protein in mesenchymal cells of the finfolds30, was used to examine cell migration following and knockdown in theMFF (e, f) and in pectoral fins (g, h). e, g, In 3-d.p.f. control larvae, migratingcells are aligned and show filopodia orientated in the direction of migration.f, h, In double morphants, mesenchymal cells are disorganized and do notshow oriented filopodia in the MFF (f) or are absent in pectoral fins(h). i, Western blot analysis of 3-d.p.f. morphants shows the 50-kDaactinodin band in and1 morphants and and2 morphants but not in doublemorphants. j, k, Picrosirius-red staining of transverse sections of a 6-d.p.a.fin ray regenerate. The control MO (j) or co-MO (k) was injected at 3 d.p.a.Compare the symmetrical arrays of actinotrichia in control rays (n 5 15)and the disorganized actinotrichia in co-MO-injected rays (n 5 15). Blackarrows indicate individual actinotrichia. Scale bars, 20mm.
and1 and and2 morpholinos resulted in the absence of actinotrichiain the developing fin folds (Fig. 3d and Supplementary Fig. 6). Themedian fin fold of morphants at 3 d post-fertilization (d.p.f.) was lessdeveloped and curled onto itself, demonstrating the supporting func-tion of the actinotrichia. Similarly, pectoral fin buds were lessdeveloped, with reduced or no fold formation (Fig. 3g, h andSupplementary Fig. 6). This result suggests that actinodins are notonly necessary for actinotrichia synthesis but are also essential forproper growth of the fold. Inactivation of and1 and and2 also per-turbed or, in the case of the pectoral fin buds, prevented the migra-tion of mesenchymal cells into the fold (Fig. 3e–h). This is significantbecause these cells include the progenitors of cells that will formlepidotrichia, the major dermal bones of the ray exoskeleton.Staining of cartilage with Alcian blue confirmed growth defects(Supplementary Fig. 8). Western blot analysis of protein extractsfrom single and double morphants confirmed the loss, in doublemorphants, of the 50-kDa band that corresponds to both proteins(Fig. 3i). The absence of actinotrichia was rescued in 85% and 92% ofdouble morphants by injection of 4 ng of and1 or and2 messengerRNA (n 5 250), respectively (data not shown). Transcripts of and1and and2 were downregulated in double morphants until about4 d.p.f. but, as morpholinos are effective for 4–5 d, transcripts pro-gressively reappeared afterwards and this correlated with a progress-ive appearance of actinotrichia (Supplementary Fig. 9). Knockdownof and1 and and2 in adult fin regenerates further confirmed theimportance of actinodins in actinotrichia formation because mor-phants showed an irregular distribution of actinotrichia 3 d aftermorpholino electroporation (n 5 15; Fig. 3j, k).
Bud morphology and gene expression patterns of double morphantswere identical to controls before the onset of and expression (Sup-plementary Fig. 10 and data not shown). However, at 48 h.p.f. and56 h.p.f. the pectoral fin buds of morphants morphologicallyresembled the buds of 40-h.p.f. and 42–45-h.p.f. controls, respectively.We therefore compared gene expression in morphology-matched andstage-matched morphant and control buds. Fibroblast growth factor(FGF) signalling is required for limb/fin outgrowth20 and, consistentwith the growth defects observed in morphants, fgf8a expression wasreduced in the AER of morphants relative to controls (in 9 of 12 buds;Fig. 4a–c). Expression of erm (also known as etv5b) and pea3, twodownstream effectors of FGF signalling, was also reduced in themesenchymal cells of morphant fin buds (Fig. 4d–f and data notshown).
Expression of shha is restricted to the posterior mesenchyme cor-responding to the zone of polarizing activity in normal limb/fin buds.In tetrapods, the graded propagation of the Shha molecule across thebud specifies digit identity. We observed an anterior expansion of theshha expression domain in pectoral fin buds of double morphants(Fig. 4j–l and Supplementary Fig. 11). Expression of hoxd13a isnormally restricted to the posterior margin of the fin bud up to36 h.p.f., and later occupies most of the posterior half of thebud21,22. In zebrafish, hoxd13a expression becomes dependent onshha expression starting around 36 h.p.f. (ref. 21), a time correspond-ing to the onset of and gene expression. In morphants, we observedan expansion of hoxd13a expression in cells of the anterior half of thefins that correlates with the expanded shha expression domain(Fig. 4g–i and Supplementary Fig. 11). The expansion of shha andhoxd13a expression may be due to the reduced expression of pea3 anderm (Fig. 4d–f and data not shown), which have recently been shownto be necessary to restrict shha expression to the posterior region ofthe fin/limb mesenchyme23,24. Expansion of Shh and Hoxd13 expres-sion in mouse limb buds is also observed in Gli3 mutants25. Weobserved a significant reduction or complete loss of gli3 expressionin morphants fin buds (in 6 of 9 fin buds; Fig. 4m–o). It has beenshown that the absence of Gli3 expression, with or without anteriorexpansion of posterior Hoxd expression, results in an anterior shift inFgf10 expression in mouse limb mesenchyme26. In zebrafish, fgf10a isexpressed along the distal mesenchyme of pectoral fin buds at 40 and
48 h.p.f. (Fig. 4p, r). In double and morphants, we observed ananterior shift in fgf10a expression consistent with the misregulationof gli3 and hoxd13a (Fig. 4q).
The altered expression of genes involved in anteroposterior pattern-ing of the limb following impaired function of and1 and and2 andactinotrichia formation allows us to propose hypotheses regardingappendage evolution. The reduced growth of morphant pectoral finsand their defective fold suggest that the resulting adult fins would beshorter and distally truncated. Notably, formation of the lepidotrichia,the dermal bony rays, seems to be compromised, as suggested by theabsence of migration of mesenchymal cells (which include progenitorsof lepidotrichia-forming cells) in the distal portion of the morphantpectoral fins (Fig. 3g, h). If our observations in zebrafish apply to anancestral fish, it is tempting to propose that, during evolution, the lossof the actinotrichia in an ancestral fish may have contributed to theconcurrent loss of the lepidotrichia. However, this consequence mayapply uniquely to the pectoral fins as the effects of the loss of function ofand1 and and2 in zebrafish seem to be less drastic in the tail, andlepidotrichia may develop normally (Fig. 3). This hypothesis fits with
WT, 40 h.p.f.Co-MO, 48 h.p.f.WT, 48 h.p.f.
shha
gli3
fgf10a
fgf8a
erm
hoxd13a
a b c
d e f
g h i
j k l
m n o
p q r
Figure 4 | Gene expression analysis in pectoral fin buds of double andmorphants. In situ hybridization of pectoral fin buds of double morphantsat 48 h.p.f. (b, e, h, k, n, q) compared with wild-type (WT) embryos at48 h.p.f. (same developmental stage; a, d, g, j, m, p) and at 40 h.p.f.(morphologically similar fin buds; c, f, i, l, o, r). The expression domain offgf8a is reduced in length in the AER of and1 1 and2 morphants (9 of 12buds; a–c), the expression of erm is reduced in the mesenchyme of morphantfin buds (d–f), and hoxd13a (g–i) and shha (j–l) show an expansion of theirexpression domains towards the anterior margin of morphant fin buds (9 of15 buds and 7 of 10 buds, respectively). Expression of gli3 is significantlyreduced in morphants (6 of 9 buds; m–o) and fgf10a transcripts are detectedin the distal mesenchyme of the control fin buds (p, r) but expression isshifted anteriorly in morphants (q). No change was observed for genes, suchas msxb, that are not known to be associated with hoxd function (data notshown). Red arrows in a–c and in p–r indicate the limits of fgf8a expressionin the AER and of fgf10a expression in the mesenchyme, respectively. Redcurved and straight dashed lines in g–i delineate the fin buds and mark theirmidlines, respectively.
fossil records showing that the earliest primitive aquatic tetrapods ofthe late-Devonian period had lost lepidotrichia in paired appendagesbut still retained a ‘fish tail’ with lepidotrichia27.
The reduced fin growth resulting from an impaired AER notwith-standing, the gene expression profile of and1 1 and2 morphants ishighly reminiscent of that described in chick and mouse mutantswith loss of GLI3/Gli3 expression. Polydactyly and the similarity ofdigits are common characteristics of Gli3 mouse mutants and ofhumans with GLI3 mutations28. Again assuming that gene expressionprofiles of and1 1 and2 morphants apply to an ancestral fish, it isplausible that, following recovery of a functional AER necessary forgrowth of the paired appendages, this pattern of gene expressionwould have rendered appendages susceptible to displaying multipledistal skeletal elements that all look similar. This is also in agreementwith the fossil record, which indicates that the earliest primitiveaquatic tetrapods of the late Devonian were polydactylous27.Acanthostega, the earliest aquatic tetrapod known so far, had eightsymmetrical digits. The number of digits decreased to six inTulerpeton, a primitive ‘true tetrapod’, and then to five (pentadac-tyly), which is the largest number of digits in extant tetrapods.
The loss of formation of actinotrichia during evolution may haveinduced profound changes in the morphology of the adult pectoralfins that perhaps led to short appendages without lepidotrichia andto gene expression profiles conducive to polydactyly in the earliesttetrapod species. Thus, the loss of actinotrichia may have contributedto the evolutionary transition from fin to limb.
METHODS SUMMARYA polyclonal antibody, anti-And1, was made against the peptide DVECLQYH
LRAAYGYR (Open Biosystems) and was used for immunostaining and western
blot analysis. We used the and1 morpholino (59-GGAAGATCCTCTCAAATG
AGCCATG-39), the and2 morpholino (59-GCCATTTTTCTCGATTAGCTG
AG-39) and the standard control morpholino (59-CCTCTTACCTCAGTTACAATTTATA-39) for gene knockdown experiments in embryos and fin regenerates.
In situ hybridization and picrosirius-red staining were performed as described
previously29.
Full Methods and any associated references are available in the online version ofthe paper at www.nature.com/nature.
Received 18 January; accepted 30 April 2010.Published online 23 June 2010.
1. Grandel, H. & Schulte-Merker, S. The development of the paired fins in thezebrafish (Danio rerio). Mech. Dev. 79, 99–120 (1998).
2. Mercader, N. Early steps of paired fin development in zebrafish compared withtetrapod limb development. Dev. Growth Differ. 49, 421–437 (2007).
3. Pizette, S. & Niswander, L. BMPs negatively regulate structure and function of thelimb apical ectodermal ridge. Development 126, 883–894 (1999).
4. Dane, P. J. & Tucker, J. B. Modulation of epidermal cell shaping and extracellularmatrix during caudal fin morphogenesis in the zebrafish Brachydanio rerio. J.Embryol. Exp. Morphol. 87, 145–161 (1985).
5. Wood, A. Early pectoral fin development and morphogenesis of the apicalectodermal ridge in the killifish, Aphysosemion scheeli. Anat. Rec. 204, 349–356(1982).
6. Bouvet, J. Differenciation et ultrastructure du squelette distal de la nageoirepectorale chez la truite indigene (Salmo Trutta Fario L.). I. Differenciation etultrastructure des actinotriches. Arch. Anat. Microsc. Morphol. Exp. 63, 79–96(1974).
7. Geraudie, J. Initiation of the actinotrichial development in the early fin bud of thefish, Salmo. J. Morphol. 151, 353–361 (1977).
8. Wood, A. & Thorogood, P. An analysis of in vitro cell migration during teleost finmorphogenesis. J. Cell Sci. 66, 205–222 (1984).
9. Garrault, H. Developpement des fibres d’elastoidine (actinotrichia) chez lessalmonides. Arch. Anat. Microsc. 130, 105–137 (1936).
10. Krukenberg, C. F. Ueber die chemische Beschaffenheit der sog. Hornfaden vonMustelus und uber die Zusammensetzung der keratinosen Hullen um den Eiernvon Scyllium stellate. Mitt. Zool. Stat. Neapel 6, 286–296 (1885).
11. Damodaran, M., Sivaraman, C. & Dhavalikar, R. S. Amino acid composition ofelastoidin. Biochem. J. 62, 621–625 (1956).
12. Kimura, S. & Kubota, M. Studies on elastoidin I. Some chemical and physicalproperties of elastoidin and its components. J. Biochem. 60, 615–621 (1966).
13. Kimura, S., Uematsu, Y. & Miyauchi, Y. Shark (Prionace glauca) elastoidin:characterization of its collagen as [alpha 1(E)]3 homotrimers. Comp. Biochem.Physiol. B 84, 305–308 (1986).
14. Padhi, B. K. et al. Screen for genes differentially expressed during regeneration ofthe zebrafish caudal fin. Dev. Dyn. 231, 527–541 (2004).
15. Seidah, N. G. & Chretien, M. Proprotein and prohormone convertases: a family ofsubtilases generating diverse bioactive polypeptides. Brain Res. 848, 45–62(1999).
16. Andrade, M. A., Perez-Iratxeta, C. & Ponting, C. P. Protein repeats: structures,functions, and evolution. J. Struct. Biol. 134, 117–131 (2001).
17. Geraudie, J. in Biology of Invertebrate and Lower Vertebrate Collagens (eds Bairati, A.& Garrone, R.) 451–455 (Plenum, 1985).
18. Nishidate, M., Nakatani, Y., Kudo, A. & Kawakami, A. Identification of novelmarkers expressed during fin regeneration by microarray analysis in medaka fish.Dev. Dynamics 236, 2685–2693 (2007).
19. Venkatesh, B. et al. Survey sequencing and comparative analysis of the elephantshark (Callorhinchus milii). PLoS Biol. 5, e101 (2007).
20. Martin, G. R. The roles of FGFs in the early development of vertebrate limbs. GenesDev. 12, 1571–1586 (1998).
21. Ahn, D. & Ho, R. K. Tri-phasic expression of posterior Hox genes duringdevelopment of pectoral fins in zebrafish: implications for the evolution ofvertebrate paired appendages. Dev. Biol. 322, 220–233 (2008).
22. Sordino, P., van der Hoeven, F. & Duboule, D. Hox gene expression in teleost finsand the origin of vertebrate digits. Nature 375, 678–681 (1995).
23. Mao, J., McGlinn, E., Huang, P., Tabin, C. J. & McMahon, A. P. Fgf-dependentEtv4/5 activity is required for posterior restriction of Sonic hedgehog andpromoting outgrowth of the vertebrate limb. Dev. Cell 16, 600–606 (2009).
24. Zhang, Z., Verheyden, J. M., Hassel, J. A. & Sun, X. FGF-regulated Etv genes areessential for repressing Shh expression in mouse limb buds. Dev. Cell 16, 607–613(2009).
25. Buscher, D., Bosse, B., Heymer, J. & Ruther, U. Evidence for genetic control of Sonichedgehog by Gli3 in mouse limb development. Mech. Dev. 62, 175–182 (1997).
26. Zakany, J., Zacchetti, G. & Duboule, D. Interactions between HOXD and Gli3 genescontrol the limb apical ectodermal ridge via Fgf10. Dev. Biol. 306, 883–893(2007).
27. Long, J. A. & Gordon, M. S. The greatest step in vertebrate history: apaleobiological review of the fish-tetrapod transition. Physiol. Biochem. Zool. 77,700–719 (2004).
28. Hui, C. C. & Joyner, A. L. A mouse model of Greig cephalopolysyndactylysyndrome: the extra-toesJ mutation contains an intragenic deletion of the Gli3gene. Nature Genet. 3, 241–246 (1993).
29. Smith, A. et al. Gene expression analysis on sections of zebrafish regenerating finsreveals limitations in the whole-mount in situ hybridization method. Dev. Dyn.237, 417–425 (2008).
30. Choo, B. G. et al. Zebrafish transgenic enhancer TRAP line database (ZETRAP).BMC Dev. Biol. 6, 5 (2006).
Supplementary Information is linked to the online version of the paper atwww.nature.com/nature.
Acknowledgements The authors would like to thank I. Duran for technical advice,A. Maurya for assistance in the early stages of this work and M. Ekker, F. Avaronand M. Debiais for discussions and critical reading of the manuscript. L.S.-P. issupported by a European Modular Biology Organization Long-Term Fellowship(ALT 325-2008). This work was supported by grants to M.-A.A. from the NaturalScience Engineering and Research Council of Canada and the Canadian Institutesof Health Research. M.A.A.-N. received funding from the Canada Research Chairsprogramme and from the Helmholtz Alliance on Systems Biology. V.K. receivedfunding from the Agency for Science, Technology and Research of Singapore.
Author Contributions J.Z., P.W., D.G. and B.K.P. performed the experiments; J.Z.collected and analysed the data; V.K. produced and provided the enhancer-traptransgenic line; M.A.A.-N. and L.S.-P. performed the analysis of the actinodinsequences; and M.-A.A. designed experiments, analysed data and wrote themanuscript.
Author Information Reprints and permissions information is available atwww.nature.com/reprints. The authors declare no competing financial interests.Readers are welcome to comment on the online version of this article atwww.nature.com/nature. Correspondence and requests for materials should beaddressed to M.-A.A. ([email protected]).
ref. 29), gli3 (2 kb; modified from a clone kindly provided by R. O. Karlstrom32
and fgf10a (1.6 kb; kindly provided by I. Belmonte).
Antibody production, immunostaining and western blot analysis. A polyclo-
nal antibody, anti-And1, was made against the peptide DVECLQYHLRAAYGYR
(Open Biosystems) and was used for immunostaining and western blot analysis.For immunostaining with anti-And1, fin regenerates were fixed in a gradient of
methanol in PBS and stored in 100% methanol at 220 uC before cryosectioning.
Binding of anti-And1 (dilution, 1:20) was revealed with an anti-rabbit fluores-
cein antibody (dilution, 1:300; Invitrogen, catalogue no. A-11008). For western
blot analysis, embryos and fin regenerates were homogenized in 32 sample
buffer33 using 80 ml sample buffer with 40 embryos and 100ml sample buffer with
three fin regenerates at 3 d.p.a. SDS–polyacrylamide gel electrophoresis analysis
of 10 ml of the homogenized samples and immunoblotting using anti-And1
(dilution, 1:10,000) were performed as described in ref. 33.
Morpholino and mRNA microinjections and electroporation. We used the
and1 morpholino (59-GGAAGATCCTCTCAAATGAGCCATG-39), the and2
morpholino (59-GCCATTTTTCTCGATTAGCTGAG-39) and the standard con-
trol morpholino (59-CCTCTTACCTCAGTTACAATTTATA-39) for gene
knockdown experiments in embryos and fin regenerates. The and1 morpholino,
the and2 morpholino and the standard control morpholino (Gene Tools) were
injected at various concentrations into embryos at the one-cell stage to deter-
mine the lowest morpholino amount that leads to a consistent and reproducible
phenotype. The best results were obtained by co-injection of 6 ng of and1 mor-
pholino and 3 ng of and2 morpholino. For rescue experiments, 4 ng of in vitro-
synthesized full-length and1 and and2 mRNAs (mMESSAGE mMACHINE Kit,
Ambion) were injected. For electroporation in fin regenerates, 40 nl each of and1
and and2 morpholinos at 0.6 mM were injected into the blastema of each ventralray of a 3-d.p.a. fin regenerate and the same amount of the standard control
morpholino was injected into the dorsal rays. Electroporation using a Grass
Technologies SD9 square-wave stimulator and Tweezertrodes (Harvard
Apparatus) was performed immediately following injection as described in ref.
34. Picrosirius-red staining was performed as described in ref. 29.
Alcian-blue staining. Alcian-blue staining was performed as previously
described35.
31. Thisse, C. & Thisse, B. High-resolution in situ hybridization to whole-mountzebrafish embryos. Nature Protocols 3, 59–69 (2008).
32. Tyurina, O. V. et al. Zebrafish Gli3 functions as both an activator and a repressor inHedgehog signaling. Dev. Biol. 277, 537–556 (2005).
33. Lang, C. et al. Molecular characterization and developmentally regulatedexpression of Xenopus lamina-associated polypeptide 2 (XLAP2). J. Cell Sci. 112,749–759 (1999).
34. Thummel, R. et al. Inhibition of zebrafish fin regeneration using in vivoelectroporation of morpholinos against fgfr1 and msxb. Dev. Dyn. 235, 336–346(2006).
35. Avaron, F., Hoffman, L., Guay, D. & Akimenko, M. A. Characterization of two newzebrafish members of the hedgehog family: atypical expression of a zebrafishindian hedgehog gene in skeletal elements of both endochondral and dermalorigins. Dev. Dyn. 235, 478–489 (2006).
