Dermoskeleton morphogenesis in zebrafish fins

Abstract

Zebrafish fins have a proximal skeleton of endochondral bones and a distal skeleton of dermal bones. Recent experimental and genetic studies are discovering mechanisms to control fin skeleton morphogenesis. Whereas the endochondral skeleton has been extensively studied, the formation of the dermal skeleton requires further revision. The shape of the dermal skeleton of the fin is generated in its distal growing margin and along a proximal growing domain. In these positions, dermoskeletal fin morphogenesis can be explained by intertissue interactions and the function of several genetic pathways. These pathways regulate patterning, size, and cell differentiation along three axes. Finally, a common genetic control of late development, regeneration, and tissue homeostasis of the fin dermoskeleton is currently being analyzed. These pathways may be responsible for the similar shape obtained after each morphogenetic process. This provides an interesting conceptual framework for future studies on this topic.

Fig. 1

Fin development in zebrafish. A: Tail fin of a 29.5-mm standard length (SL, from mouth to tail fin base) individual. Continuous and discontinuous lines, respectively, delimit ray and interray regions. Black arrows show ray branching-points. White arrows show ray joints. B: Scheme showing formation of pectoral apical fold (arrow). Left and right, respectively, are 31- and 46-hpf fin buds. Asterisk shows apical thickening. C, D: In situ hybridization of shh mRNA in early (48 hpf, C) and late (72 hpf, D) pectoral fin buds (according to Hoffman et al., 2002). Stained regions at Po are ZPA. Double-arrows label AER (C) and fin fold (D). E–H: Tail fins of a 4.5- (E), 5.5- (F), 8.5- (G) and 26.5-mm (H) SL zebrafish. Arrows show ACFP (E) and notochord dorsal bending (F). Small arrows indicate axis renaming (F). Discontinuous and continuous arrows are morphological proximodistal and anteroposterior axes (G). Length and distance between double-line are inter-joint distance and inter-ray width (G,H). I: Ray cross section. H, HI, and E, respectively, show hemi-ray, hemi-inter-ray, and epidermis. Contralateral divergent arrows show ray thickness. Double arrow shows ray width. Opposed arrows show lepidotrichia thickness. Do, V, Po, A, P, and D, dorsal, ventral, posterior, anterior, proximal, and distal, respectively. A, C–H: Nomarski optics. Scale bar = 10 (I), 40 (C), 80 (D), and 500 μm (A, E–H).

Fig. 2

Distal growing margin hypothesis. A: Scheme of the microscopic anatomy of the distal growing margin (DGM). Ep, distal epidermis; Me, distal mesenchyme. The fiber bundle is the distal actinotrichia. Lepidotrichia are contralateral hemirays. B: Inter-tissue interactions may occur as a 3D orthogonal system at the DGM of each ray and interray (pinnamere) during development and regeneration. Po, posterior; A, anterior; R, ray; I, inter-ray. a–d arrows are the candidate intertissue/genetic interactions discussed in text. a occurs at the distal organizer. b and d occur across the ray-interray boundary organizer. b and c interactions show uncertainties on the tissue/s in which they are exerted (epidermis and/or mesenchyme). e has not been consistently related to any intertissue/genetic interaction, a genes are fgfs, fgfr1, rarγ; or wnt5b. b genes may be rarγ, eve1, or evx2. c genes may be Shh pathway, cx43, or the gene mutated in another long fin. d genes are those of ActβA or Shh pathway. e genes may be evx1, hoxa13b, or cx43.

Fig. 3

Experiments on a specification map and outgrowth direction. A: Distal ablation of early pectoral fin bud and grown morphology in Salmo trutta fario (according to Bouvet, 1971). Grey, present; white, absent. B: Scheme showing proximal intercalary growth (distance increase between P, M [Medial] and D fates) and distalization (new fates distal to D) during early fin development. 1 and 2 are cut positions explained in text. Grey, an earlier stage. C: Perpendicular regenerate following oblique cut in Fundulus tail fin. C is reproduced from Nabrit (1929). D: Scheme of potential wound epidermal (We)-mesenchymal (Me) interactions that generate the results shown in C. Curved arrows, We-Me interactions; thin arrow, original proximodistal ray polarity and direction (discontinuous in regenerated regions shown in grey); thick arrow, grown proximodistal polarity and direction; 90° and small right angle, outgrowth angle with respect to the cut plane. Position and axis symbols are as in Figure 1. Grey circles in A and C, respectively, show the fin regions drawn in B and D.

Fig. 4

Experiments on local morphogenesis control. A: Proximal hole with distal oblique cut in R1 (discontinuous rectangle and inset) and regenerated ectopic R1 (eR1). eR1 outgrows outside the fin and is joined to R1 by an ectopic interray (eI). The larger arrow is a subsequent distal, transversal cut. B: Following the operation in A, the experimental ray1 regenerate (Ray 1) may show branching. Ray branching may depend on ectopic interactions from neighboring tissues (small, thick arrows). Ray/interray symbols are as in A. C: A recombinant H1H9 ray (discontinuous rectangle) is obtained substituting by grafting (curved arrow) a hemiray 9 (H9) by a hemiray fragment (small rectangle) from ray 1 (H1). H1H9 regenerate is obtained following fin cut (small thick arrow). D: Registered joint positioning at distal H1H9 regenerate. Continuous and discontinuous transversal lines show joints in contralateral hemirays. A, C: Reproduced from Murciano et al. (2002, , respectively) with permission of the publisher. A and C have been clockwise rotated 90°. Definitions are as in Figure 3.

Fig. 5

Interactions among axis genes in the early pectoral fin bud. Black, blue, and green arrows are interactions between proximodistal, anteroposterior, and dorsoventral (contralateral) regulatory genes, respectively. Aldh1a2 initiates these interactions from the somites. Suggested interactions are subsequent gene transcription activation. Light blue arrow, ligand-receptor interaction; red arrow, enzyme-product relationship. The arrows from hand2 and shh genes show initiation control. Dotted arrows from these genes show maintenance control. Broken arrows, repression; white arrows, interactions among proximodistal, anteroposterior, and dorsoventral genes; other discontinuous arrow, uncertainty. Orange genes are expressed in the AER. Gene colors and definitions are as in Figures 1, 2, and 7. Several-colored genes show various expression domains at different stages. 1, Gibert et al. (2006); 2, Mercader et al. (2006); 3, Fischer et al. (2003); 4, Harvey and Logan (2006); 5, Lee and Roy (2006); 6, Nomura et al. (2006); 7, Grandel et al. (2000); 8, Norton et al. (2005); 9, Yelon et al. (2000); 10, Neumann et al. (1999); 11, Hatta et al. (1991).

Fig. 6

Proposed interactions among proximodistal and ray, interray and joint differentiation genes in the fin blastema. The scheme shows a lateral view of two ray blastema and an intermediate interray. To the left, the ray DGM skeleton is absent. Fgfs (e.g., Wfgf or Fgf20a; see 1,2, 4; Smith et al., 2008; Whitehead et al., 2005) may be regulated by Wnt10a/Wnt5b (Stoick-Cooper et al., 2007). CP, cell proliferation at blastema mesenchyme (Me). Fgfr1/ERK and Wnt signaling pathways regulate raldh2-dependent retinoic acid-synthesis at distalmost blastema (3). Gene colors are as in Figure 7E and H. PD, RD, IRD, and JD genes, respectively, regulate the proximodistal axis, ray, interray, and joint differentiation (small box) during blastema formation. Symbols and colors are as in Figure 5. Blue/white broken arrows, PD-IRD interaction along the anteroposterior axis. 1, Lee et al. (2005); 2, Poss et al. (2000); 3, Mathew et al. (2009); 4, Yin and Poss (2008); 5, Lee et al. (2009); 6, White et al. (1994); Murciano et al. (personal communication); 7, Laforest et al. (1998); 8, Jaźwińska et al. (2007); 9, Quint et al. (2002); 10, Sims et al. (2009); 11, Poss et al. (2002); 12, Brulfert et al. (1998).

Fig. 7

Gene expression domains during fin morphogenesis. A,B: Schemes of transversal sections of zebrafish embryos at 18 hpf (A, according to Mercader, 2007) and 40 hpf (B, according to Harvey and Logan, 2006). Blue, green, and dark orange, respectively, are somites, intermediate, and lateral plate mesoderms. Arrows, suggested genetic interactions. C,D: Expression domains of hox genes in Figure 5 (similar grey hue). C: Darker to lighter grey, respectively, illustrates hoxa9/hoxa13, hoxa9/hoxa10/hoxa11, hoxa9/hoxa10 and hoxa9 domains. D: Except for hoxc6, expression domains of 5′ hoxd genes are posteriorly overlapped. E: Gene expression domains in Figure 6 (similar color). F: Adult tail fin of a long fin mutant. G: Proximoanterior region of an alf^ty86d tail fin. Arrows in same color show out-of-register joints. H: Gene expression domains (brackets and circles in same color) is ray blastema (according to Poss et al., 2000, 2002; Yoshinari et al., 2009; Brown et al., 2009). Overlapping regions, co-expression domains; PId and JPd, potential positional identity (PI) and joint positioning (JP) domains; triangles, lepidotrichia; R, right; L, left. Zns5 is an antibody. I: PI independently regulate (arrows) size (right) and joint positioning (left) genes. CP/AP is a balance between cell proliferation and apoptosis. Broken arrow, repression. Lettering as in Figures 1, 3–6. Scale bar = 500 (G) and 2,000 μm (F).

Fig. 8

Phenotypes of pattern and differentiation gene perturbations. A: Heat-shocked regenerate of hsp70:dn:fgfr1 tail fin. Reproduced from Lee et al. (2005) with permission of the publisher. The line and the arrow show where the cut took place. The discontinuous curved line shows the expected size of the outgrowing lobe. The circles show ray branching. B: Fusion phenotype (asterisk) obtained by injection of 0.2–0.6 nl of 100 ng/μl shh expression-plasmid in a branching ray blastema (according to Quint et al., 2002). Arrow shows cut plane. C: Cross-section of a simplet morphant ray. Arrowheads show ectopic lepidotrichia. Reproduced from Kizil et al. (2009) with permission of the publisher. D: Serrate phenotype (double arrow) obtained after 12-hr treatment with 5 μM Alk4/5/7-inhibitor SB431542 during wound healing (according to Jazwinska et al., 2007). Scale bar = 100 (C), 125 (B), and 250 (D) μm.

Fig. 9

Positional model and experiments showing PGD. A: Step-like joints (arrows) in zebrafish tail fin. B: Proximoanterior regions of an alf^ty86d tail fin regenerated for 22 days post-amputation (dpa). Inset shows the same joints regenerated for 14 dpa. Arrowheads indicate joint erasing (according to Murciano et al., 2007). C: Heat shock pulses in an adult tail fin of hsp70:dn:fgfr1 fish for 2 months. Grey arrows show hypertrophic joints. Reproduced from Wills et al. (2008) with permission of the publisher. D: Growing domains in the wild type tail fin. DGM and PGD are as in text. E: Spatial positional model. Dotted curved lines are PI gradients. Up-pointing arrow shows PI increase during development. After cut (small vertical line), PI gradient regenerates (curved arrow). Bullet profiles are regenerating (top) or developing (bottom) DGM. 1 and 2 are proximodistal positions in F. Discontinuous arrow is experimental reduction of PI transduction (Wills et al., 2008). F: PI control of joint differentiation. Discontinuous oblique lines, PI slope at DGM; discontinuous horizontal lines, PI activity up-regulating periodic scleroblast repression by JP (circles). 1 and 2 are as in E. Rectangles show ray segment size generated at previous positions. PD is proximodistal. Scale bar = 300 (A) and 500 μm (B, D). Original figure (C) does not show bars.