Research ArticleEVOLUTIONARY BIOLOGY

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

See allHide authors and affiliations

Science Advances  07 Nov 2018:
Vol. 4, no. 11, eaau5484
DOI: 10.1126/sciadv.aau5484
  • Fig. 1 RD modeling can explain catshark denticle patterning.

    (A) Catsharks display two rows of dorsal denticle placodes (DP) at developmental stage 32 (~80 dpf). (B to E and G to J) These placodes undergo morphogenesis and mineralize to become dorsal denticles (DD). (C, D, F, and G to J) Their emergence precedes subsequent eruption of parallel, adjacent rows of body denticles (BD). Dorsal denticles also begin to mineralize (H) before body denticle development (I). Dorsal denticles are longer and broader than body denticles (E, F, and J). RD modeling suggests that diffusion and interaction of an activator and inhibitor from an initiator row representing dorsal denticles (K) can explain the patterning of surrounding body denticles (L and M). (A) to (C) are computed tomography (CT) scans, (D) to (F) are scanning electron microscopy (SEM) images, and (G) to (J) show alizarin red–stained samples. See Materials and Methods for details of RD modeling. Scale bars, 250 μm (D), 200 μm (E), 100 μm (F), 10 mm (G), and 400 μm (H to J).

  • Fig. 2 Conserved initiator rows may trigger surrounding epithelial placodes in the shark and chick.

    Whole-mount ISH for β-cat was undertaken throughout epithelial appendage patterning of shark denticles (A, B, E, F, I, and J) and chick feathers (C, D, G, H, K, and L). At E6, the chick displays a continuous stripe of β-cat expression (C and D), which then becomes compartmentalized into feather placodes (G and H). This initiator row triggers the emergence of surrounding feather placodes, following an RD system (17). (A and B) At stage 31 (~70 dpf), shark denticle placodes are not visible, although patterning of the lateral line sensory system is demarked by β-cat. (E and F) By stage 32 (~80 dpf), two dorsolateral rows of denticle placodes are visible. (I and J) Later in stage 32 (~100 dpf), surrounding rows of body denticle placodes also express β-cat. The shark dorsal denticle rows may be triggering body denticle emergence following a Turing-like system comparable to feather patterning. LL, lateral line; BP, body placode; P, placode. Scale bars, 2000 μm (A, E, and I), 1000 μm (B, C, G, J, and K), 500 μm (D, F, and H), and 750 μm (L).

  • Fig. 3 Conserved markers of RD are expressed during shark denticle patterning.

    The expression of genes thought to control RD patterning of chick feathers was charted during shark denticle patterning (17). (A to C) At stage 32 (~80 dpf), shark dorsal denticle placodes express fgf4 and shh, which are considered activators of feather patterning, and bmp4, which is considered an inhibitor (17). (D and E) Dorsal rows also express fgf3, a dermal marker of feather bud development, and runx2, which is associated with FGF signaling during mammalian tooth development (44, 45). (F to O) Later in stage 32 (~100 dpf), these genes are expressed during patterning of adjacent, parallel rows of body denticle placodes. (P to R and T) Section ISH of body denticles revealed epithelial expression of shh and mesenchymal expression of fgf4, bmp4, and runx2. (S) Expression of fgf3 was observed in the epithelium and mesenchyme. White dashed lines separate columnar cells of the basal epithelium and the underlying mesenchyme. Scale bars, 500 μm (A to E), 2000 μm (F to J), 1000 μm (K to O), and 50 μm (P to T).

  • Fig. 4 Bead inhibition experiments reveal functional conservation of RD-associated genes.

    (A) Beads loaded with the FGFR inhibitor SU5402 were implanted beneath the epithelium of shark embryos at 75 dpf. (C to N) First, we analyzed gene expression at 5 dpt. We propose that breaking a conserved activator-inhibitor feedback system between fgf4, shh, and bmp4 (B) led to localized down-regulation of both shh and bmp4, resulting in stunted growth of dorsal denticle primordia, highlighted by black and white arrowheads (C to J) (17). (K to N) Expression of spry2, a transcriptional readout of FGF signaling, was also reduced (50). We observed localized inhibition of gene expression at 5 dpt in all SU5402 beaded samples (n = 5/5) and no DMSO control samples (n = 5/5). (O) Expression of fgf4 at 25 dpt showed that this inhibition resulted in a gap in the dorsal denticle row, which became occupied by smaller body denticles (n = 2/2). (P) No gap was observed in DMSO control samples (n = 2/2). Alizarin red staining revealed that this gap was maintained in 75% of SU5402-treated dorsal rows at 50 dpt (n = 6/8), whereas no gap was observed in rows treated with DMSO control bead (n = 7/7) (fig. S5). (Q) This pattern was maintained in SU5402 beaded dorsal rows at 75 dpt, once body denticles had begun to mineralize (n = 7/8). (R) DMSO control samples did not show a gap (n = 9/9). The output of RD simulation including a gap in the initiator row (S) was consistent with the experimental patterning observed; smaller units occupied the gap in the row (T and U). Dashed black lines show the location of vibratome sections from whole-mount ISH (E, F, I, J, M, and N). Scale bars, 200 μm (C, D, G, H, K, and L), 50 μm (E, F, I, J, M, and N), 300 μm (O and P), and 400 μm (Q and R).

  • Fig. 5 Alterations to RD parameter values can explain denticle patterning diversity.

    (A to F) Denticle diversity varies between elasmobranchs, with patterning becoming decreasingly dense from the catshark (S. canicula) to the thornback skate (R. clavata) and the little skate (L. erinacea). (G) Parameters of the RD model were initially set to result in catshark-like patterning. (H) Decreasing the inhibitor’s constitutive degradation rate (dv) and maximum net production rate (Gmax) while increasing its diffusion coefficient (Dv) resulted in a less dense thornback skate–like pattern. (E) Initiator spots were made larger and placed further apart to reflect the skate’s dorsal row. (I) Decreasing the activator’s constitutive production rate (cu) further reduced coverage density, resulting in a little skate–like pattern. See Materials and Methods for details of RD modeling and table S1 for specific parameter values. Scale bars, 400 μm (D) and 1000 μm (E).

  • Table 1 Summary of the number of replicates for bead inhibition experiments (shown in Fig. 4 and figs. S4 and S5).
    Stage fixed
    (dpf)
    Analysis
    type
    SU5402 bead
    (number of
    dorsal rows
    affected/total)
    DMSO control
    bead (number
    of dorsal rows
    unaffected/total)
    80ISH5/55/5
    100ISH2/22/2
    125Alizarin red6/87/7
    150Alizarin red7/89/9
    Total20/23 = 87%23/23 = 100%

Supplementary Materials

  • Supplementary material for this article is available at http://advances.sciencemag.org/cgi/content/full/4/11/eaau5484/DC1

    Fig. S1. Phylogenetic gene trees reconstructed from protein coding sequences extracted from www.ensembl.org.

    Fig. S2. Dorsal denticle placodes are not visible at stage 31 (~70 dpf).

    Fig. S3. Individual vibratome section images comprising false-colored ISH composite images.

    Fig. S4. Replicates of beaded shark embryos after whole-mount ISH.

    Fig. S5. Replicates of clear and stained shark embryos showing RD response to SU5402 beading.

    Fig. S6. SEM images of shark embryo 75 days after beading.

    Table S1. Activator and inhibitor values for RD model.

    Python script for RD simulations

  • Supplementary Materials

    The PDF file includes:

    • Fig. S1. Phylogenetic gene trees reconstructed from protein coding sequences extracted from www.ensembl.org.
    • Fig. S2. Dorsal denticle placodes are not visible at stage 31 (~70 dpf).
    • Fig. S3. Individual vibratome section images comprising false-colored ISH composite images.
    • Fig. S4. Replicates of beaded shark embryos after whole-mount ISH.
    • Fig. S5. Replicates of clear and stained shark embryos showing RD response to SU5402 beading.
    • Fig. S6. SEM images of shark embryo 75 days after beading.
    • Table S1. Activator and inhibitor values for RD model.

    Download PDF

    Other Supplementary Material for this manuscript includes the following:

    Files in this Data Supplement:

Navigate This Article