Evolution, development, and regeneration of tooth-like epithelial appendages in sharks.

Sharks and their relatives are typically covered in highly specialized epithelial appendages embedded in the skin called dermal denticles; ancient tooth-like units (odontodes) composed of dentine and enamel-like tissues. These 'skin teeth' are remarkably similar to oral teeth of vertebrates and share comparable morphological and genetic signatures. Here we review the histological and morphological data from embryonic sharks to uncover characters that unite all tooth-like elements (odontodes), including teeth and skin denticles in sharks. In addition, we review the differences between the skin and oral odontodes that reflect their varied capacity for renewal. Our observations have begun to decipher the developmental and genetic shifts that separate these seemingly similar dental units, including elements of the regenerative nature in both oral teeth and the emerging skin denticles from the small-spotted catshark (Scyliorhinus canicula) and other chondrichthyan models. Ultimately, we ask what defines a tooth at both the molecular and morphological level. These insights aim to help us understand how nature makes, replaces and evolves a vast array of odontodes.

Development of branchial ionocytes in embryonic and larval stages of cloudy catshark, Scyliorhinus torazame.

In teleost fish, branchial ionocytes are important sites for osmoregulation and acid-base regulation by maintaining ionic balance in the body fluid. During the early developmental stages before the formation of the gills, teleost ionocytes are localized in the yolk-sac membrane and body skin. By comparing with teleost fish, much less is known about ionocytes in developing embryos of elasmobranch fish. The present study investigated the development of ionocytes in the embryo and larva of cloudy catshark, Scyliorhinus torazame. We first observed ionocyte distribution by immunohistochemical staining with anti-Na+/K+-ATPase (NKA) and anti-vacuolar-type H+-ATPase (V-ATPase) antibodies. The NKA- and V-ATPase-rich ionocytes appeared as single cells in the gill filaments from stage 31, the stage of pre-hatching, while the ionocytes on the body skin and yolk-sac membrane were also observed. From stage 32, in addition to single ionocytes on the gill filaments, some outstanding follicular structures of NKA-immunoreactive cells were developed to fill the inter-filament region of the gill septa. The follicular ionocytes possess NKA in the basolateral membrane and Na+/H+ exchanger 3 in the apical membrane, indicating that they are involved in acid-base regulation like single NKA-rich ionocytes. Three-dimensional analysis and whole-mount immunohistochemistry revealed that the distribution of follicular ionocytes was limited to the rostral side of gill septum. The rostral sides of gill septum might be exposed to faster water flow than caudal side because the gills of sharks gently curved backward. This dissymmetric distribution of follicular ionocytes is considered to facilitate efficient body-fluid homeostasis of catshark embryo.

Artificial insemination and parthenogenesis in the whitespotted bamboo shark Chiloscyllium plagiosum.

Non-lethal methods for semen collection from elasmobranchs to better understand species reproduction has accompanied the development of artificial insemination. Ejaculates (n = 82) collected from whitespotted bamboo sharks Chiloscyllium plagiosum (n = 19) were assessed and cold-stored raw or extended at 4 °C. Females (n = 20) were inseminated with fresh or 24-48 h cold-stored raw or extended semen and paternity of offspring determined with microsatellite markers. Insemination of females with fresh semen (n = 10) resulted in 80 hatchlings and 27.6% fertility. Insemination of females with semen cold-stored 24 h (n = 4) and 48 h (n = 1) semen resulted in 17 hatchlings and fertilization rates of 28.1% and 7.1% respectively. Two females inseminated with fresh or cold-stored semen laid eggs that hatched from fertilization and parthenogenesis within the same clutch. Parthenogenesis rate for inseminated females was 0.71%. Results demonstrate artificial insemination with cold-stored semen can provide a strategy for transport of male genetics nationally and internationally, precluding the need to transport sharks. Production of parthenotes in the same clutch as sexually fertilized eggs highlights the prevalence of parthenogenesis in whitespotted bamboo sharks and poses important considerations for population management.

Cephalochordates: A window into vertebrate origins.

How vertebrates evolved from their invertebrate ancestors has long been a central topic of discussion in biology. Evolutionary developmental biology (evodevo) has provided a new tool-using gene expression patterns as phenotypic characters to infer homologies between body parts in distantly related organisms-to address this question. Combined with micro-anatomy and genomics, evodevo has provided convincing evidence that vertebrates evolved from an ancestral invertebrate chordate, in many respects resembling a modern amphioxus. The present review focuses on the role of evodevo in addressing two major questions of chordate evolution: (1) how the vertebrate brain evolved from the much simpler central nervous system (CNS) in of this ancestral chordate and (2) whether or not the head mesoderm of this ancestor was segmented.

Developmental hourglass and heterochronic shifts in fin and limb development.

How genetic changes are linked to morphological novelties and developmental constraints remains elusive. Here, we investigate genetic apparatuses that distinguish fish fins from tetrapod limbs by analyzing transcriptomes and open-chromatin regions (OCRs). Specifically, we compared mouse forelimb buds with the pectoral fin buds of an elasmobranch, the brown-banded bamboo shark (Chiloscyllium punctatum). A transcriptomic comparison with an accurate orthology map revealed both a mass heterochrony and hourglass-shaped conservation of gene expression between fins and limbs. Furthermore, open-chromatin analysis suggested that access to conserved regulatory sequences is transiently increased during mid-stage limb development. During this stage, stage-specific and tissue-specific OCRs were also enriched. Together, early and late stages of fin/limb development are more permissive to mutations than middle stages, which may have contributed to major morphological changes during the fin-to-limb evolution. We hypothesize that the middle stages are constrained by regulatory complexity that results from dynamic and tissue-specific transcriptional controls.

Animals come in all shapes and sizes. This diversity arose through genetic mutations during evolution, but it is unclear exactly how these variations led to the formation of new shapes. There is increasing evidence to suggest that not all shapes are possible and that variability between animals is limited by a phenomenon known as "developmental constraint". These limitations direct parts of the body towards a specific shape as they develop in the embryo. Therefore, understanding the mechanisms underlying these developmental constraints could help explain how different body shapes evolved. The limbs of humans and other mammals evolved from the fins of fish, and this transition is often used to study the role developmental constraints play in evolution. This is an ideal model as there is already a detailed fossil record mapping this evolutionary event, and data pinpointing some of the genes involved in the development of limbs and fins. But this data is incomplete, and a full comparison between the genes activated in the fin and the limb during embryonic development had not been achieved. This is because most fish used for research have undergone recent genetic changes, making it hard to spot which genetic differences are linked to the evolution of the limb. To overcome this barrier, Onimaru et al. compared genetic data from the developing limbs of mice to the developing fins of the brown-banded bamboo shark, which evolves much slower than other fish. This revealed that although many genes commonly played a role in the development of the fin and the limb in the embryo, the activity of these shared genes was not the same. For example, genes that switched on in the late stages of limb development, switched off in the late stages of fin development. But in the middle of development, those differences were relatively small and both species activated very similar sets of genes. Many of these genes were pleiotropic, which means they have important roles in other tissues and therefore mutate less often. This suggests that the mid-stage of limb development is under the strongest level of constraint. Darwin's theory of natural selection explains that mutations drive evolution. But the theory cannot predict what kinds of new body shapes new mutations will produce. Understanding how the activity levels of different genes affect development could help to fill this knowledge gap. This has potential medical applications, for example, understanding why some genetic changes cause more serious problems than others. This work suggests that mutations in genes that are active during the mid-stage of limb development may have the most serious impact.

Future thermal regimes for epaulette sharks (Hemiscyllium ocellatum): growth and metabolic performance cease to be optimal.

Climate change is affecting thermal regimes globally, and organisms relying on their environment to regulate biological processes face unknown consequences. In ectotherms, temperature affects development rates, body condition, and performance. Embryonic stages may be the most vulnerable life history stages, especially for oviparous species already living at the warm edge of their distribution, as embryos cannot relocate during this developmental window. We reared 27 epaulette shark (Hemiscyllium ocellatum) embryos under average summer conditions (27 °C) or temperatures predicted for the middle and end of the twenty-first century with climate change (i.e., 29 and 31 °C) and tracked growth, development, and metabolic costs both in ovo and upon hatch. Rearing sharks at 31 °C impacted embryonic growth, yolk consumption, and metabolic rates. Upon hatch, 31 °C-reared sharks weighed significantly less than their 27 °C-reared counterparts and exhibited reduced metabolic performance. Many important growth and development traits in this species may peak after 27 °C and start to become negatively impacted nearing 31 °C. We hypothesize that 31 °C approximates the pejus temperature (i.e., temperatures at which performance of a trait begin to decline) for this species, which is alarming, given that this temperature range is well within ocean warming scenarios predicted for this species' distribution over the next century.

Isotopic enrichment in Rhizoprionodon porosus (Poey, 1861) and Carcharhinus porosus (Ranzani, 1840) embryos.

Isotopic values of two Caribbean sharpnose shark Rhizoprionodon porosus litters (Poey, 1861) with two and three embryos and one litter of 11 smalltail shark Carcharhinus porosus embryos showed enriched 15 N and 13 C compared to their mothers. In R. porosus, embryonic isotope values were 3.06 ± 0.07‰ and 0.69 ± 0.15‰ greater than their mothers' for δ15 N and δ13 C, respectively, whereas in C. porosus, δ15 N and δ13 C were 1.79 ± 0.09‰ and 1.31 ± 0.17‰ greater in embryos than their mothers.

Evolutionary developmental genetics of teeth and odontodes in jawed vertebrates: a perspective from the study of elasmobranchs.

Most extant vertebrates display a high variety of tooth and tooth-like organs (odontodes) that vary in shape, position over the body and nature of composing tissues. The development of these structures is known to involve similar genetic cascades and teeth and odontodes are believed to share a common evolutionary history. Gene expression patterns have previously been compared between mammalian and teleost tooth development but we highlight how the comparative framework was not always properly defined to deal with different tooth types or tooth developmental stages. Larger-scale comparative analyses also included cartilaginous fishes: sharks display oral teeth and dermal scales for which the gene expression during development started to be investigated in the small-spotted catshark Scyliorhinus canicula during the past decade. We report several descriptive approaches to analyse the embryonic tooth and caudal scale gene expressions in S. canicula. We compare these expressions wih the ones reported in mouse molars and teleost oral and pharyngeal teeth and highlight contributions and biases that arise from these interspecific comparisons. We finally discuss the evolutionary processes that can explain the observed intra and interspecific similarities and divergences in the genetic cascades involved in tooth and odontode development in jawed vertebrates.

The denticle surface of thresher shark tails: Three-dimensional structure and comparison to other pelagic species.

Shark skin denticles (scales) are diverse in morphology both among species and across the body of single individuals, although the function of this diversity is poorly understood. The extremely elongate and highly flexible tail of thresher sharks provides an opportunity to characterize gradients in denticle surface characteristics along the length of the tail and assess correlations between denticle morphology and tail kinematics. We measured denticle morphology on the caudal fin of three mature and two embryo common thresher sharks (Alopias vulpinus), and we compared thresher tail denticles to those of eleven other shark species. Using surface profilometry, we quantified 3D-denticle patterning and texture along the tail of threshers (27 regions in adults, and 16 regions in embryos). We report that tails of thresher embryos have a membrane that covers the denticles and reduces surface roughness. In mature thresher tails, surfaces have an average roughness of 5.6 µm which is smoother than some other pelagic shark species, but similar in roughness to blacktip, porbeagle, and bonnethead shark tails. There is no gradient down the tail in roughness for the middle or trailing edge regions and hence no correlation with kinematic amplitude or inferred magnitude of flow separation along the tail during locomotion. Along the length of the tail there is a leading-to-trailing-edge gradient with larger leading edge denticles that lack ridges (average roughness = 9.6 µm), and smaller trailing edge denticles with 5 ridges (average roughness = 5.7 µm). Thresher shark tails have many missing denticles visible as gaps in the surface, and we present evidence that these denticles are being replaced by new denticles that emerge from the skin below.

Report on fetal mummification in the scalloped hammerhead shark Sphyrna lewini.

This study describes the fetal mummification process in two embryos of a 310 cm total length scalloped hammerhead shark Sphyrna lewini caught in southeastern Brazil, in December 2017. Fourteen embryos were observed in total, in which two males in the left uterus presented different stages of mummification. Both mummified embryos were covered by an exudate (i.e., a mucous substance), indicating a hematic mummification process. All embryos were at the placentotrophic stage of development, indicating that they were close to parturition. An intrinsic characteristic is suggested as possible etiology for this condition, such as umbilical torsion, because both embryos were at different sizes and, therefore, at different development stages. In addition, the sample size did not allow the authors to presume any pollution effect once only one female was observed. Finally, fetal mummification and other embryonic development disorders might have populational impacts due to reduction in embryo survival and, consequently, recruitment. For this reason and considering that S. lewini is categorized as a "critically endangered species," this study's results have conservational relevance.

Ocean warming and hypoxia affect embryonic growth, fitness and survival of small-spotted catsharks, Scyliorhinus canicula.

Elasmobranchs are key to a healthy marine ecosystem but are under threat from human activities, such as destructive fisheries and shark finning. Embryos of oviparous elasmobranchs may be further challenged during development by rising temperatures and falling dissolved oxygen concentrations in their intertidal environment. However, the impact of climate change on survival and growth of oviparous elasmobranchs is still poorly understood. Here, we investigate the effects of temperature and hypoxia on the growth and survival of small-spotted catshark (Scyliorhinus canicula) embryos by incubating eggs in normoxia 15°C, normoxia 20°C, hypoxia 15°C, or hypoxia 20°C. Incubation under the elevated temperature increased the embryonic growth rate, yolk consumption rate and Fulton's condition factor at hatching, whilst decreasing the total length and body mass of newly hatched sharks. Under low oxygen conditions (50% air saturation) the survival rate of S. canicula embryos dropped significantly and the temperature-induced increase in Fulton's condition factor was reversed. Together, these data demonstrate both the individual and compound effects of elevated temperature and hypoxia on the survival and growth during early ontogeny of a ubiquitous, coastal elasmobranch, S. canicula.

Growth trajectories of prenatal embryos of the deep-sea shark Chlamydoselachus anguineus (Chondrichthyes).

Chlamydoselachus anguineus, Garman 1884, commonly called the frilled shark, is a deep-sea shark species occurring up to depths of 1300 m. It is assumed to represent an ancient morphotype of sharks (e.g., terminal mouth opening, more than five gill slits) and thus is often considered to represent plesiomorphic traits for sharks. Therefore, its early ontogenetic developmental traits are important for understanding the evolution of its particular phenotype. Here, we established six stages for prenatal embryos and used linear measurements and geometric morphometrics to analyse changes in shape and size as well as their timing during different embryonic stages. Our results show a change in head shape and a relocation of the mouth opening at a late stage of development. We also detected a negative allometric growth of the head and especially the eye compared to the rest of the body and a sexual dimorphism in total body length, which differs from the known data for adults. A multivariate analysis of covariance shows a significant interaction of shape related to the logarithm of centroid size and developmental stage. Geometric morphometrics results indicate that the head shape changes as a covariate of body size while not accounting for differences between sexes. The growth pattern of stages 32 and 33 indicates a shift in head shape, thus highlighting the moment in development when the jaws start to elongate anteriorly to finally achieve the adult condition of terminal mouth opening rather than retaining the early embryonic subterminal position as is typical for sharks. Thus, the antero-terminal mouth opening of the frilled shark has to be considered a derived feature.

Effects of elevated carbon dioxide on the hematological parameters of a temperate catshark.

Atmospheric CO2 levels have been rising due to an increase in anthropic activities and its implications over marine ecosystems are unprecedented. The present study focused on the effects of ocean acidification (OA) on key hematological parameters of the juvenile small-spotted catsharks (Scyliorhinus canicula). Eggs were reared throughout the entire embryogenesis (~4 months) plus 5 additional months, in two experimental treatments (control: pCO2 ~ 400 µatm; and high CO2 : pCO2 ~ 900 µatm, Δ -0.3 pH units). After blood collection, the following hematological parameters were evaluated: (a) normal blood cells count (erythrocytes, leukocytes, and thrombocytes), (b) presence of erythrocytes with nuclear abnormalities, and (c) erythrocyte nucleus to cytoplasmic ratio. Concomitantly, to determine the cardiac and hematopoietic conditions, the spleen and heart to body ratios were also assessed. The present findings indicate that the measured variables may not be affected by elevated pCO2 in this temperate species, as no significant differences were observed between treatments across all the endpoints tested. Nonetheless, it is worth mentioning a decreasing trend observed in a number of thrombocytes associated with OA, which should foster further investigation, regarding other aspects of their coagulation response. Along with OA, other stressors are expected to impact marine life, such as warming and hypoxia. Thus, future research should aim to investigate the cumulative effect of these stressors on hematological parameters in sharks.

Comparison of Maternal and Embryonic Trace Element Concentrations in Common Thresher Shark (Alopias vulpinus) Muscle Tissue.

This study compared the concentration of essential (Co, Cr, Cu, Fe, Mn, Ni, Se, Zn) and nonessential (Ag, As, Cd, Hg, Pb) trace elements in the muscle tissue of a pregnant common thresher shark (Alopias vulpinus) to the concentration in the three embryos. With the exception of Ag, Cd, Cr, and Ni which were below the detection limit, all other elements accumulated in the embryo muscle tissue. The Se:Hg molar ratios in the embryos averaged 9.8, indicating that Se may have a protective role against Hg toxicity during this early life stage. Maternal transfer as a source of trace elements in sharks should not be overlooked and future studies need to focus on how reproductive strategy influences this process.

Molecular ontogeny of the stomach in the catshark Scyliorhinus canicula.

The origin of extracellular digestion in metazoans was accompanied by structural and physiological alterations of the gut. These adaptations culminated in the differentiation of a novel digestive structure in jawed vertebrates, the stomach. Specific endoderm/mesenchyme signalling is required for stomach differentiation, involving the growth and transcription factors: 1) Shh and Bmp4, required for stomach outgrowth; 2) Barx1, Sfrps and Sox2, required for gastric epithelium development and 3) Cdx1 and Cdx2, involved in intestinal versus gastric identity. Thus, modulation of endoderm/mesenchyme signalling emerges as a plausible mechanism linked to the origin of the stomach. In order to gain insight into the ancient mechanisms capable of generating this structure in jawed vertebrates, we characterised the development of the gut in the catshark Scyliorhinus canicula. As chondrichthyans, these animals retained plesiomorphic features of jawed vertebrates, including a well-differentiated stomach. We identified a clear molecular regionalization of their embryonic gut, characterised by the expression of barx1 and sox2 in the prospective stomach region and expression of cdx1 and cdx2 in the prospective intestine. Furthermore, we show that gastric gland development occurs close to hatching, accompanied by the onset of gastric proton pump activity. Our findings favour a scenario in which the developmental mechanisms involved in the origin of the stomach were present in the common ancestor of chondrichthyans and osteichthyans.

Involvement of HGF/MET signaling in appendicular muscle development in cartilaginous fish.

In amniotes, limb muscle precursors de-epithelialize from the ventral dermomyotome and individually migrate into limb buds. In catsharks, Scyliorhinus, fin muscle precursors are also derived from the ventral dermomyotome, but shortly after de-epithelialization, they reaggregate within the pectoral fin bud and differentiate into fin muscles. Delamination of muscle precursors has been suggested to be controlled by hepatocyte growth factor (HGF) and its tyrosine kinase receptor (MET) in amniotes. Here, we explore the possibility that HGF/MET signaling regulates the delamination of appendicular muscle precursors in embryos of the catshark Scyliorhinus canicula. Our analysis reveals that Hgf is expressed in pectoral fin buds, whereas c-Met expression in fin muscle precursors is rapidly downregulated. We propose that alteration of the duration of c-Met expression in appendicular muscle precursors might underlie the evolution of individually migrating muscle precursors, which leads to the emergence of complex appendicular muscular systems in amniotes.

Expression of radial glial markers (GFAP, BLBP and GS) during telencephalic development in the catshark (Scyliorhinus canicula).

Radial glial cells (RGCs) are the first cell populations of glial nature to appear during brain ontogeny. They act as primary progenitor (stem) cells as well as a scaffold for neuronal migration. The proliferative capacity of these cells, both in development and in adulthood, has been subject of interest during past decades. In contrast with mammals where RGCs are restricted to specific ventricular areas in the adult brain, RGCs are the predominant glial element in fishes. However, developmental studies on the RGCs of cartilaginous fishes are scant. We have studied the expression patterns of RGCs markers including glial fibrillary acidic protein (GFAP), brain lipid binding protein (BLBP), and glutamine synthase (GS) in the telencephalic hemispheres of catshark (Scyliorhinus canicula) from early embryos to post-hatch juveniles. GFAP, BLBP and GS are first detected, respectively, in early, intermediate and late embryos. Expression of these glial markers was observed in cells with radial glia morphology lining the telencephalic ventricles, as well as in their radial processes and endfeet at the pial surface and their expression continue in ependymal cells (or tanycytes) in early juveniles. In addition, BLBP- and GS-immunoreactive cells morphologically resembling oligodendrocytes were observed. In late embryos, most of the GFAP- and BLBP-positive RGCs also coexpress GS and show proliferative activity. Our results indicate the existence of different proliferating subpopulations of RGCs in the embryonic ventricular zone of catshark. Further investigations are needed to determine whether these proliferative RGCs could act as neurogenic and/or gliogenic precursors.

Inner ear development in cyclostomes and evolution of the vertebrate semicircular canals.

Jawed vertebrates have inner ears with three semicircular canals, the presence of which has been used as a key to understanding evolutionary relationships. Ostracoderms, the jawless stem gnathostomes, had only two canals and lacked the lateral canal1-3. Lampreys, which are modern cyclostomes, are generally thought to possess two semicircular canals whereas the hagfishes-which are also cyclostomes-have only a single canal, which used to be regarded as a more primitive trait1,4. However, recent molecular and developmental analyses have strongly supported the monophyly of cyclostomes5-7, which has left the evolutionary trajectory of the vertebrate inner ear unclear8. Here we show the differentiation of the otic vesicle of the lamprey Lethenteron camtschaticum and inshore hagfish Eptatretus burgeri. This is the first time, to our knowledge, that the development of the hagfish inner ear is reported. We found that canal development in the lamprey starts with two depressions-which is reminiscent of the early developmental pattern of the inner ear in modern gnathostomes. These cyclostome otic vesicles show a pattern of expression of regulatory genes, including OTX genes, that is comparable to that of gnathosomes. Although two depressions appear in the lamprey vesicle, they subsequently fuse to form a single canal that is similar to that of hagfishes. Complete separation of the depressions results in anterior and posterior canals in gnathostomes. The single depression of the vesicle in hagfishes thus appears to be a secondarily derived trait. Furthermore, the lateral canal in crown gnathostomes was acquired secondarily-not by de novo acquisition of an OTX expression domain, but by the evolution of a developmental program downstream of the OTX genes.

Oviparous elasmobranch development inside the egg case in 7 key stages.

Embryological stages of oviparous elasmobranch during development can be difficult to identify, requiring magnification and/or fixation of an anaesthetized embryo. These restrictions are poorly suited for monitoring the development of living elasmobranchs inside their egg cases. There are two major aims of this study. The first was to observe elasmobranch embryonic development non-invasively and produce a non-invasive developmental key for identifying the life stages for an elasmobranch inside the egg case. To this end, 7 key developmental stages were identified for the greater spotted catshark, Scyliorhinus stellaris, and are provided here with diagrams from multiple perspectives to demonstrate the key features of each stage. The physiological and ecological relevance of each stage are discussed in terms of structure and function for embryonic survival in the harsh intertidal zone. Also discussed is the importance of the egg case membrane and the protective embryonic jelly. The second aim of the study was to understand the applicability of the 7 developmental stages from S. stellaris to other oviparous elasmobranchs. Thus, changes in embryonic body size and egg yolk volume at each stage were measured and compared with those of the closely related, lesser spotted catshark, Scyliorhinus canicula. We find nearly identical growth patterns and yolk consumption patterns in both species across the 7 developmental stages. Thus, although the 7 developmental stages have been constructed in reference to the greater spotted catshark, we suggest that it can be applied to other oviparous elasmobranch species with only minor modification.

An ancient Turing-like patterning mechanism regulates skin denticle development in sharks.

Vertebrates have a vast array of epithelial appendages, including scales, feathers, and hair. The developmental patterning of these diverse structures can be theoretically explained by Alan Turing's reaction-diffusion system. However, the role of this system in epithelial appendage patterning of early diverging lineages (compared to tetrapods), such as the cartilaginous fishes, is poorly understood. We investigate patterning of the unique tooth-like skin denticles of sharks, which closely relates to their hydrodynamic and protective functions. We demonstrate through simulation models that a Turing-like mechanism can explain shark denticle patterning and verify this system using gene expression analysis and gene pathway inhibition experiments. This mechanism bears remarkable similarity to avian feather patterning, suggesting deep homology of the system. We propose that a diverse range of vertebrate appendages, from shark denticles to avian feathers and mammalian hair, use this ancient and conserved system, with slight genetic modulation accounting for broad variations in patterning.