Heterochronic shift in Hox-mediated activation of sonic hedgehog leads to morphological changes during fin development

Abstract

We explored the molecular mechanisms of morphological transformations of vertebrate paired fin/limb evolution by comparative gene expression profiling and functional analyses. In this study, we focused on the temporal differences of the onset of Sonic hedgehog (Shh) expression in paired appendages among different vertebrates. In limb buds of chick and mouse, Shh expression is activated as soon as there is a morphological bud, concomitant with Hoxd10 expression. In dogfish (Scyliorhinus canicula), however, we found that Shh was transcribed late in fin development, concomitant with Hoxd13 expression. We utilized zebrafish as a model to determine whether quantitative changes in hox expression alter the timing of shh expression in pectoral fins of zebrafish embryos. We found that the temporal shift of Shh activity altered the size of endoskeletal elements in paired fins of zebrafish and dogfish. Thus, a threshold level of hox expression determines the onset of shh expression, and the subsequent heterochronic shift of Shh activity can affect the size of the fin endoskeleton. This process may have facilitated major morphological changes in paired appendages during vertebrate limb evolution.

Figure 1. Shh expression commences late in S. canicula (Sc) fin development, concomitant with Hoxd13 expression.

(A–G) Pectoral fin buds. Anterior is to the left. (A) ScShh expression at stages 27, 29 and early stage 32. Transcripts were present in the posterior region (arrowheads) at stage 29 but absent at stages 27 and early stage 32. (B, C) Expression of ScHoxa11 (B) and ScHoxa13 (C). ScHoxa11 transcripts were first detected in the posterior region and in the muscle buds. By early stage 32, transcripts were restricted to the posterior-distal region. ScHoxa13 transcripts were restricted to the distal part of the fin buds throughout fin development. Arrowheads indicate limits of ScHoxa expression. (D–G) Expression of ScHoxd10 (D), ScHoxd11 (E), ScHoxd12 (F) and ScHoxd13 (G). The ScHoxd genes were expressed collinearly at early stages. ScHoxd10–d12 transcripts were apparent at stage 27, whereas ScHoxd13 transcripts were first observed in the posterior mesenchyme at stage 29. Arrowheads indicate the anterior limits of ScHoxd expression. (H) Quantitative PCR analysis to determine the expression levels of ScHoxd10–13 in the pectoral fins of stage 26, 27 and 29 dogfish embryos. Relative expression was normalized against ScGAPDH transcripts. Note that levels of ScHoxd10–13 transcript expression increased at stage 29. Expression of ScHoxd10–d13 in stage 26 pectoral fins, or expression of ScHoxd13 in stage 27 pectoral fins, was not detectable. (I) Schematic representation of temporal Hoxd expression and Shh expression during pectoral fin development in S. canicula. Shh was expressed concomitantly with Hoxd13.

Figure 2. Timing of shh expression in zebrafish embryo fin primordia depends on hox transcript accumulation.

(A) Schematic representation of temporal hox and shh expression in the pectoral fin primordia of zebrafish embryos. shh was expressed at 24 hpf concomitantly with hoxd10a expression. (B) RT-PCR analysis to determine the efficiency of the hoxd10a or hoxd13a splice-blocking morpholino (MO). In the schematics, arrows represent forward (F) and reverse (R) primers, and the short red bars represent the hoxd10a MO and hoxd13a MO. Lower panel, analysis of RT-PCR products by agarose gel electrophoresis. Products of 618 bp and 1333 bp represent spliced and unspliced hoxd10a mRNA, respectively. The 316-bp RT-PCR product represents spliced hoxd13a mRNA. Amplification of eif4a cDNA was used as a control. (C) Whole-mount in situ hybridization to detect shh expression in the pectoral fin primordia of D. rerio embryos injected with 5 ng control MO (top panels), 5 ng hoxd10a MO (middle panels) or 5 ng hoxd13a MO (bottom panels) at the indicated hpf. Red ovals highlight the pectoral fin primordia. Note that shh expression was first observed at 24 hpf in the fin primordia of embryos injected with control (top) or hoxd13a MO (bottom), whereas shh transcripts became detectable at 25.5 hpf in the primordia of most embryos injected with hoxd10a MO (middle). (D) Percentages of embryos with detectable or undetectable levels of shh expression observed at 22.5, 24, and 25.5 hpf following injection of control MO, hoxd10a MO or hoxd13a MO (see also Figure S4). A representative image depicting the detectable or undetectable levels of shh expression in the pectoral fin primordia is shown at the left. Insets show high magnification views of pectoral fin primordia. (E) Semi-quantitative RT-PCR analysis to determine the expression levels of 5′ hoxd when shh is transcribed in pectoral fin buds. The relative levels of hoxd10a and hoxd11a transcripts in the lateral plate mesoderm of morphants were quantified. Relative expression was normalized against gapdh transcripts.

Figure 3. hox transcript accumulation is critical for the onset of shh expression in fin development.

(A) Expression of shh in pectoral fin primordia of D. rerio embryos injected with 5 ng control MO, 20 pg hoxd10a mRNA, 20 pg hoxd13a mRNA or 20 pg hoxa13a mRNA at 22.5 hpf. Red ovals highlight the pectoral fin primordia. Note that transcripts of shh became detectable at 22.5 hpf in the fin primordia of embryos injected with hoxd10a, hoxd13a or hoxa13a mRNA. (B) The percentage of embryos with the indicated level of shh expression at 22.5 hpf following injection of control MO, hoxd10a mRNA, hoxd13a mRNA or hoxa13a mRNA is shown in the bottom panel (see also Figure S4). A representative image depicting the detectable or undetectable levels of shh expression in the pectoral fin primordia is shown at the left.

Figure 4. Schematic representation of temporal hox and shh expression in pectoral fin primordia of zebrafish embryos.

Expression of shh was observed at 24 hpf and was concomitant with hoxd10a expression. The onset of shh expression in hoxd10a morphants was concomitant with the onset of hoxd11a expression, whereas shh expression was not delayed in hoxd13a morphants. In embryos injected with hoxd10a, hoxd13a or hoxa13a mRNA, shh expression was observed at 22.5 hpf.

Figure 5. Temporal shift of Shh activity leads to changes in pectoral fin morphology.

(A) shh expression appears at 25.5 hpf in pectoral fin primordia of D. rerio embryos injected with 5 ng of hoxd10a MO. (B) At 5 dpf, pectoral fins of embryos injected with control MO or with hoxd10a MO were stained with Alcian Blue (left). Cleithrum (cl), scapulocoracoid (sc), postcoracoid process (pop), endoskeletal disc (ed) and actinotrichs (ac) are indicated. Scale bars: 100 µm. The relative lengths of the endoskeletal disc are presented in the graph (right). *P<0.001, as assessed by Student's t-test. (C) shh expression appears at 24 hpf, concomitantly with hoxd10a, in pectoral fin primordia of D. rerio. Hedgehog signaling was blocked by treatment with 60 µM cyclopamine from 23 to 27 hpf, resulting in ablation of ptc1 expression until at least 27 hpf. ptc1 expression was recovered by 30 hpf in pectoral fin primordia of cyclopamine-treated embryos. (D) ptc1 and shh expression were examined in control or cyclopamine-treated embryos at the indicated stages (left). At 5 dpf, pectoral fins of control or cyclopamine-treated embryos were stained with Alcian Blue (middle). Scale bars: 200 µm. The relative lengths of the endoskeletal disc are represented in a graph (right). *P<0.05, as assessed by Student's t-test with Welch's correction. (E) Shh and Ptc2 expression disappeared before stage 31 in pectoral fin buds of S. canicula embryos. Hedgehog signaling was extended by treatment with SAG for 6 days from stage 30 to 31, resulting in extension of Ptc2 expression until at least stage 31. (F) Ptc2 expression was examined in control or SAG-treated embryos at stage 31 (5 days after the initial treatment). (G) Pectoral fins of control or SAG-treated embryos were stained with Alcian Blue. Anterior is to the left. Proximal is to the top. Insets show magnified views of the pectoral fin metapterygium. Note that the width of the metapterygium (arrows) of SAG-treated embryos was significantly increased. Scale bars: 1 mm. (H) Comparison of the size of the pectoral fin endoskeleton between control and SAG-treated S. canicula embryos. The table shows the total body length (TL), metapterygium length (ML), metapterygium width (MW), width across the base of pectoral fin endoskeleton (WPF), and length of pectoral fin endoskeleton (LPF) of control and SAG-treated embryos. The metapterygium lengths are represented in the bar graph. *P<0.05, as assessed by Student's t-test.

Figure 6. Diagram representing the effect of Shh expression heterochrony on vertebrate paired appendage evolution.

A model suggesting that the early fin buds may have acquired low levels of Hox expression by co-option of collinear Hox expression in the main body axis . Changes in accumulated Hox could have led to altered onset of Shh expression, resulting in enlargement of the endoskeletal elements during fin evolution.