Trunk neural crest origin of dermal denticles in acartilaginous fishJ. Andrew Gillisa,b,1, Els C. Alsemaa,c, and Katharine E. Criswella,b

aDepartment of Zoology, University of Cambridge, Cambridge CB2 3EJ, United Kingdom; bWhitman Center, Marine Biological Laboratory, Woods Hole, MA02543; and cKarolinska Institutet, SE-171 77 Stockholm, Sweden

Edited by Marianne Bronner, California Institute of Technology, Pasadena, CA, and approved October 26, 2017 (received for review August 4, 2017)

Cartilaginous fishes (e.g., sharks and skates) possess a postcranialdermal skeleton consisting of tooth-like “denticles” embeddedwithintheir skin. As with teeth, the principal skeletal tissue of dermal den-ticles is dentine. In the head, cranial neural crest cells give rise to thedentine-producing cells (odontoblasts) of teeth. However, trunk neu-ral crest cells are generally regarded as nonskeletogenic, and so theembryonic origin of trunk denticle odontoblasts remains unresolved.Here, we use expression of FoxD3 to pinpoint the specification andemigration of trunk neural crest cells in embryos of a cartilaginousfish, the little skate (Leucoraja erinacea). Using cell lineage tracing,we further demonstrate that trunk neural crest cells do, in fact, giverise to odontoblasts of trunk dermal denticles. These findings expandthe repertoire of vertebrate trunk neural crest cell fates during nor-mal development, highlight the likely primitive skeletogenic poten-tial of this cell population, and point to a neural crest origin ofdentine throughout the ancestral vertebrate dermal skeleton.

skate | neural crest | denticle | evolution | skeleton

Jawed vertebrates (gnathostomes) primitively possessed anextensive postcranial dermal skeleton, consisting of both a

basal osteogenic component (composed of acellular bone and/orlamellar bone-like tissues) and a superficial odontogenic com-ponent (composed of dentine overlain by enamel or enameloid)(1). Elaboration of this dermal skeleton is closely tied to thediversification of gnathostomes, but the developmental andevolutionary origins of this skeletal system remain unresolved. Apostcranial dermal skeleton is present in several extant gna-thostome lineages—e.g., in the form of scales, denticles, orscutes of fishes (2)—although these elements exhibit consider-able histological variation and are often much reduced relative tothe dermal armor of stem gnathostomes. For example, the vastmajority of teleost fishes possess lightly mineralized bony scalesthat lack typical dentine or enamel and that likely derive fromthe osteogenic component of the primitive gnathostome dermalskeleton. Conversely, cartilaginous fishes (sharks, skates, andrays) possess well-mineralized dermal denticles (Fig. 1) that arecomposed of typical dentine and enameloid, but that lack basalbone of attachment (3).Bone and dentine are both mesenchymal in origin and are

secreted by osteoblasts and odontoblasts, respectively, whileenamel matrix is secreted by epithelial ameloblasts (with“enameloid” resulting from the mineralization of mixed dentineand enamel matrices; ref. 4). Based on comparison with thecranial dermal skeleton, in which much of the dermal bone (i.e.,of the skull vault) and all dentine (i.e., of oral teeth) is neuralcrest-derived (5–7), it has long been assumed that these tissuesshare a similar neural crest origin in the trunk (8, 9). However,recent fate-mapping studies in teleosts have called this assump-tion into question: cell lineage-tracing experiments in zebrafish(10, 11) and medaka (12) have demonstrated a paraxial meso-dermal origin of scale-producing osteoblasts, leading to sugges-tions that trunk neural crest cells lack skeletogenic potential infishes (10, 12, 13). However, the embryonic origin of the denti-nous component of the postcranial dermal skeleton remainsunresolved. While it has been suggested that teleost scales are

odontogenic in nature (14, 15), they do not exhibit key histo-logical features of dentine (e.g., dentine tubules), and their modeof development and growth is more similar to that of acellularbone (16, 17). We therefore sought to test the odontogenic po-tential of trunk neural crest cells in a taxon that has retainedtypical dentine in their postcranial dermal skeleton—a cartilag-inous fish, the little skate, Leucoraja erinacea.

ResultsThe neural crest is a multipotent cell population that undergoesepithelial-to-mesenchymal transition from the developing centralnervous system of vertebrate embryos and that migrates exten-sively to give rise to a diversity of skeletal and nonskeletal celltypes (18). Histological studies of shark (Scyliorhinus torazame)embryos have revealed that neural crest cells begin to emigratefrom the cranial neural tube at developmental stage (S)18 andfrom the trunk neural tube at approximately stage S19 (ref. 19,with staging according to ref. 20). To verify the timing of trunkneural crest cell emigration in skate embryos, we conducted aseries of mRNA in situ hybridization experiments to characterizeexpression of the transcription factor, FoxD3—a neural crestspecifier that is expressed in nascent and early migratory neuralcrest cells (21). At S18 (Fig. 2A), we observed FoxD3-expressingcells (i.e., premigratory neural crest cells) in the dorsal trunkneural tube. We observed that, as in the shark, trunk neural crestcells begin emigrating in skate embryos at S19, based on theappearance FoxD3-expressing cells undergoing epithelial-to-mesenchymal transition from the dorsal trunk neural tube (Fig. 2B).

Significance

The earliest mineralized skeleton of vertebrates was the der-mal skeleton: superficial armor of tooth-like skeletal unitscomposed of dentine and basal bone of attachment. Remnantsof this dentinous armor have been retained as teeth in thehead of all jawed vertebrates and as dermal denticles in theskin of cartilaginous fishes (sharks and skates). Cranial neuralcrest cells (NCCs) give rise to dentine-secreting odontoblasts ofteeth. However, trunk NCCs are regarded as nonskeletogenic,raising questions about the embryonic origin of postcranialdenticles in cartilaginous fishes. Here, we show that trunkNCCs give rise to trunk denticle odontoblasts in the skate,Leucoraja erinacea. This finding expands the repertoire oftrunk NCC fates, highlighting the primitive skeletogenic po-tential of this cell population.

Author contributions: J.A.G. designed research; J.A.G. and K.E.C. performed research;J.A.G., E.C.A., and K.E.C. analyzed data; and J.A.G. wrote the paper.

The authors declare no conflict of interest.

This article is a PNAS Direct Submission.

Published under the PNAS license.

Data deposition: The sequence reported in this paper has been deposited in the GenBankdatabase (accession no. MF281542).1To whom correspondence should be addressed. Email: jag93@cam.ac.uk.

This article contains supporting information online at www.pnas.org/lookup/suppl/doi:10.1073/pnas.1713827114/-/DCSupplemental.

13200–13205 | PNAS | December 12, 2017 | vol. 114 | no. 50 www.pnas.org/cgi/doi/10.1073/pnas.1713827114

Dow

nloa

ded

by g

uest

on

Dec

embe

r 13

, 202

0

This migration continues through S22 (Fig. 2C), at which pointwe also begin to observe rostral differentiation of trunk neuralcrest derivatives (e.g., dorsal root ganglia).To label premigratory trunk neural crest cells in skate embryos,

we microinjected the lipophilic dye CM-DiI into the lumen of thetrunk neural tube at S18 (Fig. 3A). This strategy ensures focal la-beling of the neural tube (including premigratory neural crest cells)and has been used extensively in other systems to test neural crestcell fates (22–26). Five days postinjection, labeled neural crest cellscould be observed emigrating from the trunk neural tube of CM-DiI–injected skate embryos (Fig. 3B). Embryos with CM-DiI–la-beled neural tubes were left to develop for 4–5 mo (to S33, whendermal denticles have differentiated within the trunk), at whichpoint, histological examination revealed CM-DiI–labeled cells inexpected trunk neural crest derivatives, such as dorsal root ganglia(Fig. 3C) and melanocytes (Fig. 3D), as well as in the odontoblastlayer of trunk dermal denticles.In cartilaginous fish dermal denticles, the odontoblast layer is

morphologically distinct within the pulp cavity and consists of

polarized secretory cells with characteristic cell processes lying ad-jacent to newly secreted dentine matrix (3) (Fig. 1D). In total, werecovered 15 CM-DiI–labeled trunk denticle odontoblasts in em-bryos that were examined histologically (n = 8; Fig. 3 E and F—additional examples provided in Fig. S1), indicating their neuralcrest origin. Seven out of eight examined embryos possessed CM-DiI–labeled odontoblasts, with each positive embryo possessingbetween one and four labeled cells (Table S1). CM-DiI–labeledodontoblasts always occurred singly, with their rarity likely due todilution of the label through substantial proliferation betweenlabeling and analysis.Some embryos (n = 4) possessed CM-DiI–labeled cells within

tissue of the denticle pulp cavity, and some (n = 4) also possessedoccasional CM-DiI–positive ameloblasts (Fig. S2 and Table S1).While tooth ameloblasts are generally regarded as an exclusivelyectodermally derived cell type (6), one genetic lineage-tracingstudy using P0-Cre/R26R mice previously suggested that a smallnumber of tooth ameloblasts might, in fact, derive from neuralcrest cells (27). Our observation of CM-DiI–positive ameloblasts

Fig. 1. Dermal denticles of the little skate. (A) A skate hatchling and (B) skeletal preparation exhibit prominent dorsal median (arrow) and paired dorso-lateral rows (arrowheads) of dermal denticles extending the length of the trunk. (C) These denticles (*) are recurved, with a tooth-like morphology. (D) Asagittal section through a differentiated dermal denticle (preeruption) at embryonic stage 33 reveals a pulp cavity filled with mesenchyme (Pulp in Di), amorphologically distinct layer of dentine-secreting odontoblasts (Odont. in Di) lining the wall of the pulp cavity, and a layer of cuboidal enamel-secretingameloblasts (Amel. in Di). (Scale bars: A and B, 1 cm; C, 750 μm; D, 20 μm; Di, 10 μm.)

Gillis et al. PNAS | December 12, 2017 | vol. 114 | no. 50 | 13201

DEV

ELOPM

ENTA

LBIOLO

GY

Dow

nloa

ded

by g

uest

on

Dec

embe

r 13

, 202

0

in chondrichthyan trunk dermal denticles could be indicative of aneural crest contribution to the ameloblast layer of dental organswithin the vertebrate dermal skeleton. However, we do observesigns of ectodermal contamination in embryos with CM-DiI–labeled neural tubes (a result of colabeling of some surface ec-toderm at the site of trunk neural tube injection), and wetherefore cannot rule out ectodermal contamination as the basisof these labeled ameloblasts.When labeling the trunk neural tube with CM-DiI, we would

sometimes contaminate adjacent paraxial mesodermal cells (asevidenced by the presence of some CM-DiI–positive chon-drocytes within the axial skeleton of S33 embryos). Given theparaxial mesodermal origin of teleost scales (10–12), we soughtto specifically test for mesodermal contributions to skate trunkdermal denticles. In three S18 skate embryos, we focally labeledthe presomitic paraxial mesoderm of the trunk (Fig. S3A), whichsits ventrolateral to the neural tube (Fig. S3B) and again leftembryos to develop until S33. Upon examination, no CM-DiI–labeled odontoblasts or pulp cavity cells were observed in thedenticles of any of these embryos, despite extensive labeling ofother paraxial mesodermal derivatives, such as chondrocytes ofthe axial skeleton (Fig. S3C) and pectoral girdle (Fig. S3D), andtrunk musculature (Fig. S3E). These experiments thereforeprovide no evidence of a paraxial mesodermal contribution tothe trunk denticles of the little skate.

DiscussionIt has been demonstrated that the trunk neural crest cells ofmouse (28), chick (29), and axolotl (30) may be pushed toward askeletogenic fate through transplantation or recombination withcranial epithelia in vitro, and it has been suggested that this re-flects the vestigial skeletogenic potential of trunk neural crestcells that is suppressed in tetrapods. Our demonstration of aneural crest origin of trunk dermal denticles during normal de-velopment in a cartilaginous fish is consistent with this view andpoints to the likely primitive skeletogenic potential of trunkneural crest cells in vertebrates. Given the mesodermal origin ofteleost scales (a likely derivative of the basal-most bony layerof the ostracoderm dermal skeleton), we hypothesize that themultilayered dermal armor of stem gnathostomes may haveconsisted of a basal, mesodermally derived osteogenic compo-nent and a superficial, neural crest-derived odontogenic com-ponent (Fig. 4), with each of these layers being variouslyretained, reduced, or lost in different vertebrate lineages. Thescales of some nonteleost actinopterygian fishes, such as Poly-pterus, appear to have retained both the osteogenic and odon-togenic components of the postcranial dermal skeleton (15),rendering them an important model system with which to furthertest this hypothesis.It has recently been demonstrated that several well-established

neural crest derivatives in both the head (e.g., tooth odontoblasts—ref. 31) and the trunk (e.g., melanocytes, parasympathetic ganglia,

Fig. 2. Trunk neural crest cell migration in the little skate. (A) Wholemount mRNA in situ hybridization for the neural crest specifier FoxD3 at embryonicstage 18 reveals neural tube expression in the cranial and rostral trunk dorsal midline, as well as in the pharyngeal and gut endoderm. At this stage, cross-sections through the rostral trunk (Ai and Aii) reveal no emigrating neural crest cells, although premigratory neural crest cells may be recognized by theirexpression of FoxD3 within the dorsal neural tube. (B) At stage 19, FoxD3 expression persists in the developing central nervous system and endoderm, andcross-sections through the rostral trunk (Bi and Bii) now reveal FoxD3-expressing neural crest cells emigrating from the dorsal neural tube. (C) By stage 22,FoxD3 expression in the cranial region localizes to the mandibular (m), hyoid (h), and branchial (b) streams of cranial neural crest cells, while expressionpersists in the dorsal midline of the trunk. A row of trunk neural crest-derived dorsal root ganglia (DRG, *) may now be recognized in the rostral trunk, andcross-sections through the trunk just beyond the front of DRG differentiation (Ci and Cii) reveal continued emigration of FoxD3-expressing trunk neural crestcells from the dorsal neural tube. (Scale bars: A–C, 375 μm; Ai, Aii, Bi, Bii, Ci, and Cii, 10 μm.)

13202 | www.pnas.org/cgi/doi/10.1073/pnas.1713827114 Gillis et al.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 13

, 202

0

and enteric neurons—refs. 26 and 32−34) arise from neuralcrest-derived Schwann cell precursors associated with nerve fi-bers. If trunk denticle odontoblasts share a similar Schwanncell precursor origin, such a cell lineage bottleneck (along withdilution of CM-DiI through extensive proliferation) could ac-count for the small number of CM-DiI–positive odontoblastsobserved here. While further cell lineage-tracing experimentsusing indelible labels (e.g., genomically integrated fluorescentreporters) will eventually allow us to more accurately assess theprecise nature and extent of the neural crest contribution to the

postcranial skeleton of cartilaginous fishes, our current findingssupport the notion that the odontoblast is a potentially uniquelyneural crest-derived skeletal cell type (35) and highlights the neuralcrest as the source of the dentine throughout the primitive verte-brate dermal skeleton.

Materials and MethodsEmbryo Collection and Fate Mapping. Skate (Leucoraja erinacea) eggswere obtained at the Marine Biological Laboratory and maintained in a flow-through seawater system at ∼16 °C. CM-DiI was prepared for microinjection

Fig. 3. A trunk neural crest origin of skate dermal denticles. (A) Premigratory trunk neural crest cells were labeled by microinjection of CM-DiI into the lumenof the trunk neural tube at stage 18. (B) This approach effectively labeled trunk neural crest cells, and at 5 d postinjection, CM-DiI–labeled neural crest cellswere observed emigrating from the trunk neural tube. By stage 33, histological analysis revealed CM-DiI–labeled cells within anticipated trunk neural crestderivatives, such as (C) dorsal root ganglia (Ci is the same section as C, subsequently stained with Masson’s trichrome) and (D) melanocytes (Di is the samesection as D, subsequently imaged with light). Following labeling of trunk neural crest cells, (E and F) CM-DiI–labeled odontoblasts were also recovered withintrunk dermal denticles, indicating their neural crest origin. Ei and Eii and Fi and Fii are images of the same sections, imaged initially with fluorescence (CM-DiIwith DAPI counterstain) and subsequently with Masson’s trichrome. CM-DiI–labeled odontoblasts are indicated with white arrows and dashed outline. (Scalebars: A, 375 μm; B, 20 μm; C and Ci, 20 μm; D and Di, 10 μm; E, 60 μm; Ei and Eii, 20 μm; F, 75 μm; Fi and Fii, 15 μm.)

Gillis et al. PNAS | December 12, 2017 | vol. 114 | no. 50 | 13203

DEV

ELOPM

ENTA

LBIOLO

GY

Dow

nloa

ded

by g

uest

on

Dec

embe

r 13

, 202

0

as described (36). For injection of S18 skate embryos, eggs were removedfrom the water, briefly dried, and a small window was cut in the topsurface of the egg case with a razor blade. CM-DiI was microinjectedthrough the jelly layer into the lumen of the neural tube or the paraxialmesoderm using a Picospritzer pressure injector, and the window was thensealed completely using a piece of donor egg shell adhered with a cya-noacrylate adhesive (“Krazy Glue”). All injected embryos were left to

develop in a flow-through seawater system until S33 (∼4–5 mo post-injection), at which point embryos were euthanized with an overdose ofMS-222 (1 g/L), fixed in 4% paraformaldehyde in phosphate-buffered sa-line (PBS) overnight at 4 °C, rinsed three times in PBS, and stored in PBS +0.02% sodium azide at 4 °C. All animal work complied with protocolsapproved by the Institutional Animal Care and Use Committee at theMarine Biological Laboratory (MBL).

Fig. 4. Hypothesis of embryonic origin and evolution of the vertebrate dermal skeleton. The primitive vertebrate dermal skeleton, as seen in heterostracans, wascomposed of a mesodermally derived osteogenic layer (Ost.) and a neural crest-derived odontogenic layer (Odont.). This dermal skeleton has undergone re-ductions and modifications through vertebrate phylogeny, including the loss of the osteogenic layer in cartilaginous fishes (e.g., skate) and the loss of theodontogenic layer in teleosts (e.g., zebrafish). Certain nonteleost bony fishes, such as the bichir, have retained both the osteogenic and odontogenic layers in theirscales, and we predict that the former would be mesodermal in origin, while the latter would be neural crest-derived.

13204 | www.pnas.org/cgi/doi/10.1073/pnas.1713827114 Gillis et al.

Dow

nloa

ded

by g

uest

on

Dec

embe

r 13

, 202

0

Skeletal Preparation, Histology, and mRNA in Situ Hybridization.Whole-mountskeletal preparations and paraffin sections were produced as previouslydescribed (37, 38), and histochemical staining of paraffin sections was per-formed according to the Masson’s trichrome protocol of Witten and Hall (39)or according to a standard hematoxylin and eosin (H&E) staining protocol.Sections of CM-DiI–injected embryos were coverslipped with FluoromountG + DAPI, imaged with fluorescence, and then decoverslipped (by soakingovernight in water) and stained with Masson’s trichrome. Wholemount andparaffin section mRNA in situ hybridization for FoxD3 (GenBank AccessionNumber MF281542) was performed as previously described (36).

ACKNOWLEDGMENTS. We acknowledge Dr. Kate Rawlinson, Prof. RichardBehringer, Prof. Alejandro Sánchez Alvarado, Prof. Jonathan Henry, the MBLembryology community and the staff of the MBL Marine Resources Centerfor facilitating this research, and Dr. Clare Baker for helpful comments onthe manuscript. We also acknowledge technical support and microscopeloans from Leica Microsystems at MBL. This research was supported by RoyalSociety University Research Fellowship UF130182 (to J.A.G.) and by the Uni-versity of Cambridge Isaac Newton Trust Grant 14.23z (to J.A.G.). K.E.C. wassupported by Royal Society Shooter International Postdoctoral FellowshipNF160762.

1. Keating JN, Marquart CL, Donoghue PCJ (2015) Histology of the heterostracan dermalskeleton: Insight into the origin of the vertebrate mineralised skeleton. J Morphol276:657–680.

2. Sire JY, Huysseune A (2003) Formation of dermal skeletal and dental tissues in fish: Acomparative and evolutionary approach. Biol Rev Camb Philos Soc 78:219–249.

3. Miyake T, Vaglia JL, Taylor LH, Hall BK (1999) Development of dermal denticles inskates (Chondrichthyes, Batoidea): Patterning and cellular differentiation. J Morphol241:61–81.

4. Gillis JA, Donoghue PCJ (2007) The homology and phylogeny of chondrichthyan toothenameloid. J Morphol 268:33–49.

5. Couly GF, Coltey PM, Le Douarin NM (1993) The triple origin of skull in higher ver-tebrates: A study in quail-chick chimeras. Development 117:409–429.

6. Chai Y, et al. (2000) Fate of the mammalian cranial neural crest during tooth andmandibular morphogenesis. Development 127:1671–1679.

7. Jiang X, Iseki S, Maxson RE, Sucov HM, Morriss-Kay GM (2002) Tissue origins and in-teractions in the mammalian skull vault. Dev Biol 241:106–116.

8. SmithMM, Hall BK (1990) Development and evolutionary origins of vertebrate skeletogenicand odontogenic tissues. Biol Rev Camb Philos Soc 65:277–373.

9. Smith MM, Hall BK (1993) A developmental model for evolution of the vertebrateexoskeleton and teeth: The role of cranial and trunk neural crest. Evol Biol 27:387–448.

10. Lee RT, Thiery JP, Carney TJ (2013) Dermal fin rays and scales derive from mesoderm,not neural crest. Curr Biol 23:R336–R337.

11. Mongera A, Nüsslein-Volhard C (2013) Scales of fish arise from mesoderm. Curr Biol23:R338–R339.

12. Shimada A, et al. (2013) Trunk exoskeleton in teleosts is mesodermal in origin. NatCommun 4:1639.

13. Lee RT, Knapik EW, Thiery JP, Carney TJ (2013) An exclusively mesodermal origin offin mesenchyme demonstrates that zebrafish trunk neural crest does not generateectomesenchyme. Development 140:2923–2932.

14. Sire JY, Akimenko MA (2004) Scale development in fish: A review, with description ofsonic hedgehog (shh) expression in the zebrafish (Danio rerio). Int J Dev Biol 48:233–247.

15. Sire JY, Donoghue PCJ, Vickaryous MK (2009) Origin and evolution of the in-tegumentary skeleton in non-tetrapod vertebrates. J Anat 214:409–440.

16. Meunier FJ (1984) Spatial organization and mineralization of the basal plate ofelasmoid scales in osteichthyans. Am Zool 24:953–964.

17. Meunier FJ, Huysseune A (1991) The concept of bone tissue in osteichthyes. Neth JZool 42:445–458.

18. Green SA, Simoes-Costa M, Bronner ME (2015) Evolution of vertebrates as viewedfrom the crest. Nature 520:474–482.

19. Adachi N, Kuratani S (2012) Development of head and trunk mesoderm in the dog-fish, Scyliorhinus torazame: I. Embryology and morphology of the head cavities andrelated structures. Evol Dev 14:234–256.

20. Ballard WW, Mellinger J, Lechenault H (1993) A series of normal stages for devel-opment of Scyliorhinus canicula, the lesser spotted dogfish (Chondrichthyes: Scylio-rhinidae). J Exp Zool 267:318–336.

21. Sasai N, Mizuseki K, Sasai Y (2001) Requirement of FoxD3-class signaling for neural

crest determination in Xenopus. Development 128:2525–2536.22. Serbedzija GN, Fraser SE, Bronner-Fraser M (1990) Pathways of trunk neural crest cell

migration in the mouse embryo as revealed by vital dye labelling. Development 108:

605–612.23. Serbedzija GN, Burgan S, Fraser SE, Bronner-Fraser M (1991) Vital dye labelling

demonstrates a sacral neural crest contribution to the enteric nervous system of chick

and mouse embryos. Development 111:857–866.24. Reyes M, Zandberg K, Desmawati I, de Bellard ME (2010) Emergence and migration of

trunk neural crest cells in a snake, the California Kingsnake (Lampropeltis getula

californiae). BMC Dev Biol 10:52.25. Juarez M, et al. (2013) Characterization of the trunk neural crest in the bamboo shark,

Chiloscyllium punctatum. J Comp Neurol 521:3303–3320.26. Green SA, Uy BR, Bronner ME (2017) Ancient evolutionary origin of vertebrate enteric

neurons from trunk-derived neural crest. Nature 544:88–91.27. Wang S-K, Komatsu Y, Mishina Y (2011) Potential contribution of neural crest cells to

dental enamel formation. Biochem Biophys Res Commun 415:114–119.28. Lumsden AGS (1988) Spatial organization of the epithelium and the role of neural

crest cells in the initiation of the mammalian tooth germ. Development 103:155–169.29. McGonnell IM, Graham A (2002) Trunk neural crest has skeletogenic potential. Curr

Biol 12:767–771.30. Graveson AC, Smith MM, Hall BK (1997) Neural crest potential for tooth development

in a urodele amphibian: Developmental and evolutionary significance. Dev Biol 188:

34–42.31. Kaukua N, et al. (2014) Glial origin of mesenchymal stem cells in a tooth model sys-

tem. Nature 513:551–554.32. Espinosa-Medina I, et al. (2014) Neurodevelopment. Parasympathetic ganglia derive

from Schwann cell precursors. Science 345:87–90.33. Dyachuk V, et al. (2014) Neurodevelopment. Parasympathetic neurons originate from

nerve-associated peripheral glial progenitors. Science 345:82–87.34. Adameyko I, et al. (2009) Schwann cell precursors from nerve innervation are a cel-

lular origin of melanocytes in skin. Cell 139:366–379.35. Hall BK, Gillis JA (2013) Incremental evolution of the neural crest, neural crest cells

and neural crest-derived skeletal tissues. J Anat 222:19–31.36. Gillis JA, et al. (2012) Electrosensory ampullary organs are derived from lateral line

placodes in cartilaginous fishes. Development 139:3142–3146.37. O’Neill P, McCole RB, Baker CVH (2007) A molecular analysis of neurogenic placode

and cranial sensory ganglion development in the shark, Scyliorhinus canicula. Dev Biol

304:156–181.38. Gillis JA, Dahn RD, Shubin NH (2009) Chondrogenesis and homology of the visceral

skeleton in the little skate, Leucoraja erinacea (Chondrichthyes: Batoidea). J Morphol

270:628–643.39. Witten PE, Hall BK (2003) Seasonal changes in the lower jaw skeleton in male Atlantic

salmon (Salmo salar L.): Remodelling and regression of the kype after spawning.

J Anat 203:435–450.

Gillis et al. PNAS | December 12, 2017 | vol. 114 | no. 50 | 13205

DEV

ELOPM

ENTA

LBIOLO

GY

Dow

nloa

ded

by g

uest

on

Dec

embe

r 13

, 202

0