The larval pharynx of the cephalochordate Branchiostoma (amphioxus) is asymmetrical. The mouth is on the left, and endostyle and gill slits are on the right. At the neurula, Nodal and Hedgehog (Hh) expression becomes restricted to the left. To dissect their respective roles in gill slit formation, we inhibited each pathway separately for 20 min at intervals during the neurula stage, before gill slits penetrate, and monitored the effects on morphology and expression of pharyngeal markers. The results pinpoint the short interval spanning the gastrula/neurula transition as the critical period for specification and positioning of future gill slits. Thus, reduced Nodal signaling shifts the gill slits ventrally, skews the pharyngeal domains of Hh, Pax1/9, Pax2/5/8, Six1/2 and IrxC towards the left, and reduces Hh and Tbx1/10 expression in endoderm and mesoderm, respectively. Nodal auto-regulates. Decreased Hh signaling does not affect gill slit positions or Hh or Nodal expression, but it does reduce the domain of Gli, the Hh target, in the pharyngeal endoderm. Thus, during the neurula stage, Nodal and Hh cooperate in gill slit development – Hh mediates gill slit formation and Nodal establishes their left-right position.

In bilateral animals, some organs are often arrayed asymmetrically about the midline. For example, in vertebrates, the viscera are asymmetrical. In the invertebrate chordate amphioxus, the larval pharynx is highly asymmetrical, with gill slits, the endostyle (homologous to the vertebrate thyroid gland) and a larval secretory organ, the club-shaped gland, forming on the right, and the mouth developing on the left. In many organisms, the specification of left-right asymmetry is mediated by the secreted signaling protein Nodal on the left (Boorman and Shimeld, 2002). In amphioxus, Nodal is initially expressed symmetrically but, by the early neurula, expression becomes restricted to the left side (Yu et al., 2002, 2007). Inhibition of Nodal signaling during the neurula stage, when pharyngeal structures are being specified, causes a duplicate endostyle, which is normally on the right side, to develop on the left, and also prevents formation of the first gill slit and mouth (Soukup et al., 2015). As in other organisms, Nodal, Cerberus, Lefty and Pitx cooperate (Li et al., 2017). However, Nodal is not the only signaling protein that is asymmetrically expressed during development. In both vertebrates and amphioxus, Hedgehog (Hh) genes also become asymmetrically expressed on the left side during early development (Shimeld, 1999; Dyer and Kirby, 2009; Tsiairis and McMahon, 2009; Tsikolia et al., 2012; Otto et al., 2014), and the Hh target Gli is expressed in the developing gill slit primordia (characterized by a thickening of the endoderm where each gill slit will form) (Shimeld, 2007). In the mouse, knockout of one Hh gene, sonic hedgehog (Shh), results in the first pharyngeal arches being fused in the midline (Moore-Scott and Manley, 2005; Yamagishi et al., 2006; Swartz et al., 2012) and in the failure of the neural tube to fuse (holoprosencephaly) (Xavier et al., 2016). Similarly, in amphioxus, knockout of Hh eliminates the mouth; the gill slit primordia are either absent, or aberrant and ventralized (Wang et al., 2015; Hu et al., 2017). Thus, the effects of inhibiting either Nodal or Hh signaling on gill slit formation in amphioxus embryos are much the same, raising the issue of whether Nodal and Hh act together or independently in patterning the amphioxus pharynx.

Signaling by both Nodal and Hh proteins is complex. In the canonical pathway, Nodal, which is in the TGFβ family, binds to Activin receptors 1 and 2, leading to phosphorylation of Smad2 and Smad3 that then complex with Smad4. This Smad complex then translocates to the nucleus where it participates in activation of Nodal target genes. In canonical Hh signaling, Hh binds to the transmembrane protein Patched (Ptch), relieving repression by a second transmembrane protein Smoothened, thereby converting Gli from a repressor to an activator of target sequences (Jenkins, 2009). In addition, non-canonical signaling by Hh and by TGFβ proteins, as well as context dependency, have been described (Bertrand and Dahmane, 2006; Jenkins, 2009; Massagué, 2012; Wang et al., 2016; Szczepny et al., 2017).

The focus on possible interactions of Nodal and Hh signaling in vertebrates has been on the CNS and the lateral plate mesoderm. In some contexts, Nodal and Hh appear to act in parallel, whereas in other contexts one may regulate the other. For example, Nodal and Shh signaling are both involved in the development of the vertebrate neural tube (Luo, 2017). Separate or simultaneous inhibition of Shh and Nodal causes holoprosencephaly, indicating that the two proteins probably act synergistically (Monuki, 2007; Mercier et al., 2013). Moreover, they may be co-regulated: for example, knockdown of the gene coding for the transcription factor Zic1 reduces signaling by both Nodal and Hh (Maurus and Harris, 2009). Specification of ventral domains in the vertebrate CNS also requires signaling by both Nodal and Hh; in this instance, Nodal is upstream of Hh. In addition, non-canonical Nodal signaling in the prechordal mesoderm acts to maintain expression of Shh (Ellis et al., 2015), whereas canonical Hh signaling in the lateral plate mesoderm is indirectly responsible for initiation and propagation of Nodal signaling (Tsiairis and McMahon, 2009). Consequently, mutations in either Hh or Nodal can also result in laterality defects (Roessler and Muenke, 2001). In animals other than vertebrates, Hh may also regulate Nodal. For example, in sea urchins, Hh regulates the late asymmetric expression of Nodal (Warner et al., 2016). This is reminiscent of the late role of Hh in activating Nodal signaling in the lateral plate mesoderm of vertebrates (Tsiairis and McMahon, 2009).

It is unclear whether Hh and Nodal interact in pharyngeal patterning. In vertebrates, Hh is expressed in the pharyngeal endoderm, where it is involved in patterning the pharyngeal arches. It is upstream of, and positively regulates, Tbx1 (Garg et al., 2001), but negatively regulates Fgf8 (Haworth et al., 2007; Billmyre and Klingensmith, 2015), Bmp4 and Pax1 (Moore-Scott and Manley, 2005). Nodal signaling, upstream of Pax2, Nkx2.1 and Hex, is essential for development of the thyroid gland, a derivative of the pharyngeal endoderm (Elsalini et al., 2003; Porazzi et al., 2009). This raises the issue of whether Nodal and Hh act independently to pattern the pharynx or whether they interact and, if so, to what extent.

To investigate possible relationships between Nodal and Hh in pharyngeal patterning, we used amphioxus as a simple chordate model. Amphioxus has little genetic redundancy, making it highly suitable for understanding the role of specific genes in embryonic patterning. A recent study has reported that expression of the Nodal antagonist Cerberus is absent in amphioxus Hh-null mutants and, therefore, concluded that Hh signaling regulates Nodal indirectly (Hu et al., 2017). However, this interpretation is equivocal, as overexpression of Hh had no effect on either Cerberus expression or on left-right asymmetry of amphioxus embryos (Hu et al., 2017). To determine whether Nodal and Hh act together or independently in pharyngeal patterning in amphioxus, we inhibited each pathway separately for a limited time span at intervals from the late gastrula through to the neurula stage and determined the effect on pharyngeal morphology and on the expression of a variety of genes expressed in the pharyngeal endoderm and/or developing gill slits. Our results reveal a narrow window at the late gastrula/early neurula stage in which Nodal signaling establishes the left-right position of gill slits, whereas Hh functions to mediate gill slit formation. Although Nodal auto-regulates, and to some extent may also regulate Hh, the two pathways act largely in parallel in gill-slit patterning and formation in amphioxus.

In normal embryos of the Florida amphioxus (Branchiostoma floridae), the larval gill slits form in the center of the gill slit primordia (thickened regions of the endoderm overlain by a thin layer of ectoderm). After ∼30-34 h of development at 25°C, the ectoderm and endoderm fuse to form the first gill slit on the right side. Within the next 2 days, two more gill slits are added posterior to the first (Fig. 1). Genes associated with gill slit formation begin to be expressed from the early neurula, at ∼10 h post fertilization (hpf), well before there is any morphological indication of a gill slit or even a gill slit primordium. For example, initially Pax1/9 is broadly expressed in the pharyngeal endoderm at the early neurula, at ∼9 hpf, and becomes downregulated at ∼10-11 hpf in the primordium of the first gill slit (Holland et al., 1995). Therefore, to determine the roles of Nodal and Hh signaling in gill slit positioning and penetration, we focused our experiments on the late gastrula through neurula stages.

Fig. 1.

The Nodal inhibitor SB505124 (15 μM) or the Hh inhibitor cyclopamine (1.75 μM) applied for 20 min at the late gastrula stage of Branchiostoma floridae results in a range of pharyngeal phenotypes at the early larval stage. (A) Diagrams with the gill slit primordia in red. Anterior to the right. From top to bottom: control with two gill slits (the third gill slit has not yet penetrated); SB505124-treated, medium gill slits are ventralized, small gill slits are ventralized, no gill slits and very small gill slit primordia; cyclopamine treated, development is somewhat retarded, two medium gill slits, one small gill slit, no gill slits with two small gill slit primordia, no gill slits and no gill slit primordia. (B) Photographs of living larvae. Anterior to the right. SB505124 treatment shifts gill slit primordia and gill slits to the left. Cyclopamine does not affect the lateral position of the gill slits. 1, first gill slit; 2, second gill slit; m, mouth; cg, club-shaped gland; e, endostyle; hald, Hatsheck's anterior left diverticulum, destined to form part of the adenohypophyseal homolog. Scale bar: 50 μm.

Fig. 1.

The Nodal inhibitor SB505124 (15 μM) or the Hh inhibitor cyclopamine (1.75 μM) applied for 20 min at the late gastrula stage of Branchiostoma floridae results in a range of pharyngeal phenotypes at the early larval stage. (A) Diagrams with the gill slit primordia in red. Anterior to the right. From top to bottom: control with two gill slits (the third gill slit has not yet penetrated); SB505124-treated, medium gill slits are ventralized, small gill slits are ventralized, no gill slits and very small gill slit primordia; cyclopamine treated, development is somewhat retarded, two medium gill slits, one small gill slit, no gill slits with two small gill slit primordia, no gill slits and no gill slit primordia. (B) Photographs of living larvae. Anterior to the right. SB505124 treatment shifts gill slit primordia and gill slits to the left. Cyclopamine does not affect the lateral position of the gill slits. 1, first gill slit; 2, second gill slit; m, mouth; cg, club-shaped gland; e, endostyle; hald, Hatsheck's anterior left diverticulum, destined to form part of the adenohypophyseal homolog. Scale bar: 50 μm.

To inhibit Nodal and Hh signaling, we used the chemical inhibitors SB505124 {2-[5-benzo(1,3)dioxol-5-yl-2-tert-butyl-3H-imidazol-4-yl]-6-methylpyridine hydrochloride}, which inhibits the Alk4, Alk5 and Alk7 Activin/Nodal receptors, and cyclopamine, which specifically binds to Smoothened and prevents transduction of the Hh signal (Chen et al., 2002a). When embryos are continuously exposed to either 15 μM SB505124 or 1.5 μM cyclopamine from the early gastrula, the resulting larvae are severely deformed and essentially lack a pharynx: gill slits and gill slit primordia are absent, although there may be some small domains of expression of pharyngeal markers at the anteriormost region of the larva (Fig. S1). To confirm that cyclopamine blocks Hh signaling, we treated embryos at the early gastrula with 1.75 µM cyclopamine and fixed aliquots for in situ hybridization at the late gastrula/very early neurula and at the early to mid-neurula. In situ hybridization for Ptch, a target of Hh signaling in amphioxus (Hu et al., 2017), as in other organisms (Chen et al., 2002a,b), showed that cyclopamine inhibits Ptch, and thus blocks Hh signaling in amphioxus (Fig. S3). At high concentrations of cyclopamine, long exposures cause the ectoderm to dissociate into individual cells. Therefore, to determine the optimum conditions for affecting pharyngeal patterning without completely eliminating the gill slit primordia or killing the larvae, we first did pilot experiments, with each inhibitor added at the following concentrations: 1.5 μM, 1.75 μM, 2.0 µM and 4.0 μm cyclopamine for 20 min at 30 min intervals (the newly hatched late gastrula, the early neurula, the early-mid neurula and the mid-neurula; see Fig. 2), and 1.0 μM, 10 μM and 50 μM SB505124 for 20 min added at the same times plus mid/late neurula and late neurula (Table S1). The optimal concentrations of inhibitors for additional experiments were determined as those resulting in a lateral shift in the position of the gill slit primordia and/or the failure of the gill slits to open (Table S1). The results of these experiments showed that 10 μM SB505124 added at late gastrula or early neurula had a very mild effect with slightly smaller gill slits, whereas 50 μm had a severe effect, with most embryos either lacking gill slits or dying. We, therefore, chose 15 μM SB505124 for subsequent experiments. Cyclopamine also eliminated gill slits at the highest concentration we used (4 μM). Therefore, for subsequent experiments, we used 1.5 μM or 1.75 μM, which produced smaller gill slits.

Fig. 2.

Effects on gill slit formation of Nodal or Hh inhibition at different stages of development. (A,B) Left hand panels show effects on gill slit formation of amphioxus (B. floridae) embryos with treatment of 15 µM SB505124 (A) or 1.75 µM cyclopamine (B) applied for 20 min at the late gastrula (9 h), very early neurula (10 h) and early neurula (11 h) stages. Control larvae were treated with DMSO (for SB505124) or ethanol (for cyclopamine). Gill slit formation was assayed at the early larval stage (43-45 hpf). At this stage, control larvae have formed two or three gill slits. For SB505124, sample numbers ranged from 54 to 154, with an average of 87 for controls and 105 for each experiment. For cyclopamine, sample numbers averaged 100 for controls and 108 for each experiment. Data are mean±s.d. Right hand panels show expression patterns of Nodal (A) and Hh (B) in normal larvae at the late gastrula, very early neurula and early neurula stages. Dorsal views are on the left and cross-sections are on the right. g.s., gill slit. Scale bars: 50 μm.

Fig. 2.

Effects on gill slit formation of Nodal or Hh inhibition at different stages of development. (A,B) Left hand panels show effects on gill slit formation of amphioxus (B. floridae) embryos with treatment of 15 µM SB505124 (A) or 1.75 µM cyclopamine (B) applied for 20 min at the late gastrula (9 h), very early neurula (10 h) and early neurula (11 h) stages. Control larvae were treated with DMSO (for SB505124) or ethanol (for cyclopamine). Gill slit formation was assayed at the early larval stage (43-45 hpf). At this stage, control larvae have formed two or three gill slits. For SB505124, sample numbers ranged from 54 to 154, with an average of 87 for controls and 105 for each experiment. For cyclopamine, sample numbers averaged 100 for controls and 108 for each experiment. Data are mean±s.d. Right hand panels show expression patterns of Nodal (A) and Hh (B) in normal larvae at the late gastrula, very early neurula and early neurula stages. Dorsal views are on the left and cross-sections are on the right. g.s., gill slit. Scale bars: 50 μm.

As is typical when either inhibitors or activators of signaling pathways are added to amphioxus embryos in DMSO or ethanol, at a given concentration, there is a range of severe to mild phenotypes because of slightly uneven mixing. Fig. 1 shows the range of phenotypes resulting from treatment with 15 μM SB505124 or 1.75 μM cyclopamine for 20 min at the late gastrula stage. The effects on gill slit size of adding these inhibitors at three stages (late gastrula, early neurula and early-mid neurula) are shown in Fig. 2. In controls, at 46 hpf, the first two gill slits have opened on the right (Fig. 1, top). In the slightly milder phenotype resulting from addition of 15 μM SB505124 at the late gastrula stage, the gill slit primordia are present, but smaller than normal and shifted somewhat ventrally; in the most severe phenotype, gill slits are imperforate and the gill slit primordia are small (Fig. 1B). With 1.75 μM cyclopamine added at the late gastrula stage, the first two gill slit primordia appear to be on the right, but are much smaller than normal; the first gill slit has opened (Fig. 1B). In the more severe phenotype, neither the gill slits nor the gill slit primordia are detectable (Fig. 1). Fig. 2 shows that, for 15 μM SB505124, small differences in the time of treatment had major effects on the size of gill slits, with most gill slits failing to penetrate with the earliest treatment, and most being about half the normal size at the latest treatment (Fig. 2A). In contrast, most of the larvae treated with 1.75 μM cyclopamine had very small gill slits regardless of the time of treatment (Figs 1 and 2B). Thus, the effect of altered Nodal signaling on gill slit size is restricted to a narrow time window, whereas Hh regulates gill slit size through to at least mid-neurula.

Pharyngeal gene expression in embryos and larvae treated with SB505124 and cyclopamine

To determine the respective roles of Nodal and Hh in regulating expression of pharyngeal markers during gill slit formation in amphioxus, we performed in situ hybridization for seven genes expressed in the developing pharynx (Nodal, Hh, Gli, Pax1/9, Pax2/5/8, IrxC and Tbx1/10). As a high percentage of larvae treated with 15 μM SB505124 at the early-mid neurula stage had nearly normal gill slits, and the effects of adding 1.75 μM cyclopamine for 20 min at the early neurula and early-mid neurula stages were the same (Fig. 1), we restricted these experiments to treatments at the late gastrula (9 hpf) and early neurula (10 hpf) stages. Fig. 3 shows that addition of 15 μM SB505124 for 20 min at either late gastrula or early neurula severely reduces Nodal expression (Fig. 3A-L,A′-L′). In addition, when SB505124 is applied at the late gastrula stage, downregulation of Nodal in the dorsal-lateral endoderm on the right, which normally occurs at the early/mid-neurula stage, is retarded (compare Fig. 3D with J). In line with a lesser effect on morphology if Nodal inhibition is delayed until the early neurula stage, the reduction in Nodal expression is not as pronounced in embryos treated at the early neurula stage compared with those treated at the late gastrula stage (Fig. 3A′-L′). In contrast, cyclopamine, when added at either the late gastrula stage (Fig. 3M-X) or the early neurula stage (Fig. 3M′-X′), has little effect on Nodal expression, except that when it is applied at late gastrula stage, the ectodermal domain on the left side is reduced or eliminated (arrows Fig. 3P,R and V,X). These results show that, whereas Nodal likely auto-regulates, Hh does not regulate Nodal, except perhaps to a slight extent in the ectoderm.

Fig. 3.

Transient inhibition of Nodal, but not Hh, reduces Nodal expression in amphioxus embryos. Expression of Nodal in control embryos (A-F,M-R,A′-F′,M′-R′) and in those treated for 20 min with the Nodal inhibitor SB505124 (15 μM) at the late gastrula (G-L) and very early neurula (G′-L′) stages or with the Hh inhibitor cyclopamine (1.75 μM) at the late gastrula (S-X) and very early neurula (S′-X′) stages. Control larvae were treated with DMSO (for SB505124) or ethanol (ETOH) (for cyclopamine). Embryos were fixed for in situ hybridization at the stages indicated. Anterior to the left in wholemounts. Cross-sections (pink), viewed from the posterior, are at the levels indicated with the lowercase letters in the preceding wholemounts (blue). Arrow in J indicates persistent expression of Nodal in SB505124-treated embryos in dorsal/lateral endoderm on the right at the mid-neurula stage. Arrowheads in P,R,V and X indicate the ectodermal domain on the left. Scale bars: 50 μm.

Fig. 3.

Transient inhibition of Nodal, but not Hh, reduces Nodal expression in amphioxus embryos. Expression of Nodal in control embryos (A-F,M-R,A′-F′,M′-R′) and in those treated for 20 min with the Nodal inhibitor SB505124 (15 μM) at the late gastrula (G-L) and very early neurula (G′-L′) stages or with the Hh inhibitor cyclopamine (1.75 μM) at the late gastrula (S-X) and very early neurula (S′-X′) stages. Control larvae were treated with DMSO (for SB505124) or ethanol (ETOH) (for cyclopamine). Embryos were fixed for in situ hybridization at the stages indicated. Anterior to the left in wholemounts. Cross-sections (pink), viewed from the posterior, are at the levels indicated with the lowercase letters in the preceding wholemounts (blue). Arrow in J indicates persistent expression of Nodal in SB505124-treated embryos in dorsal/lateral endoderm on the right at the mid-neurula stage. Arrowheads in P,R,V and X indicate the ectodermal domain on the left. Scale bars: 50 μm.

To investigate whether Hh expression is also auto-regulated or is regulated by Nodal signaling (Fig. 4), we assayed for Hh expression in embryos treated with either SB505124 or cyclopamine at the late gastrula and early neurula stages (Fig. 4). The major effects of SB505124 on Hh expression are a reduction of the domain in the ventral endoderm and a shift of the endodermal domain to the left. This is in line with the first gill slit primordium being shifted ventrally (compare Fig. 4G,H, and J,K with R,S and U,V) and suggests that Nodal may regulate the ventral endodermal domain of Hh. In addition, in embryos treated with SB505124 at the late gastrula stage, downregulation of Hh on the right side (arrows, Fig. 4O) compared with the control (Fig. 4D) was retarded, and upregulation of Hh ventrally and on the left side was reduced (compare Fig. 4G,J with R,U). Hh expression was not reduced on the left when Nodal inhibition was delayed until early neurula (Fig. 4A″-V″). Inhibition of Hh signaling at either the late gastrula or early neurula stage had little effect on Hh expression (Fig. 4W-R′,W″-R‴). Therefore, it is unlikely that there is a feedback loop whereby Hh regulates itself.

Fig. 4.

Transient inhibition of Nodal, but not Hh, affects Hh expression in amphioxus embryos. Expression of Hh in control embryos treated with DMSO for SB505124 (A-K, W-G′) or ethanol (ETOH) for cyclopamine (A″-K″, W″-G‴) and in those treated for 20 min with the Nodal inhibitor SB505124 (15 μM) at the late gastrula (L-V) and very early neurula (L″-V″) stages or with the Hh inhibitor cyclopamine (1.75 μM) at the late gastrula (H′-R′) and very early neurula (H‴-R‴) stages. Embryos were fixed for in situ hybridization at the stages indicated. Anterior to the left in wholemounts. Cross-sections (pink), viewed from the posterior, are at the levels indicated with the lowercase letters in the preceding wholemounts (blue). Arrows indicate the limits of the endodermal expression domain. Scale bars: 50 μm.

Fig. 4.

Transient inhibition of Nodal, but not Hh, affects Hh expression in amphioxus embryos. Expression of Hh in control embryos treated with DMSO for SB505124 (A-K, W-G′) or ethanol (ETOH) for cyclopamine (A″-K″, W″-G‴) and in those treated for 20 min with the Nodal inhibitor SB505124 (15 μM) at the late gastrula (L-V) and very early neurula (L″-V″) stages or with the Hh inhibitor cyclopamine (1.75 μM) at the late gastrula (H′-R′) and very early neurula (H‴-R‴) stages. Embryos were fixed for in situ hybridization at the stages indicated. Anterior to the left in wholemounts. Cross-sections (pink), viewed from the posterior, are at the levels indicated with the lowercase letters in the preceding wholemounts (blue). Arrows indicate the limits of the endodermal expression domain. Scale bars: 50 μm.

Gli is a zinc-finger transcription factor that transduces the Hh signal. In early amphioxus embryos, the two genes are expressed in largely complementary patterns, with Hh expressed in the notochord and dorsal-lateral endoderm and Gli in the developing somites and anterior neural plate (Figs 4A,B and 5A,B). At the mid-neurula stage, both genes also become expressed in the ventral endoderm, although the domain of Hh is skewed to the left, whereas that of Gli is more medial (Figs 4D,G and 5D,F,T,V). Inhibition of Nodal signaling for 20 min at the late gastrula stage has little effect on Gli expression, except that by the early larval stage the ventral endodermal domain in the second gill slit is greatly reduced in severely affected larvae with imperforate gill slits (Fig. 5O), whereas Gli expression persists in the gill slit primordia of control embryos (Fig. 5G,H). Inhibition of Nodal signaling for 20 min at the early neurula has no clear effect on Gli expression (Fig. 5A″-P″). In contrast, when Hh signaling is inhibited at the late gastrula stage, the ventral endodermal domain of Gli is reduced at the mid and late neurula stages (Fig. 5S-V,A′-E′). Whether this is because of slightly retarded development or inhibition of Gli per se is not clear. As seen in normal embryos, the ventral domain of Gli only appears between the early and mid-neurula stages (compare Fig. 5Q,R with S,T). There is no apparent effect on Gli expression when the addition of cyclopamine is delayed to the early neurula stage (Fig. 5Q″-F‴). These results indicate that although Nodal signaling has little, if any, effect on Gli expression, at the late gastrula stage Hh signaling regulates the ventral endodermal domain of Gli.

Fig. 5.

Transient inhibition of Hh, but not Nodal, reduces Gli expression in amphioxus embryos. Expression of Gli in control embryos treated with DMSO for SB505124 (A-H, A″-H″) and ethanol (ETOH) for cyclopamine (Q-X, Q″-X″) and in those treated for 20 min with the Nodal inhibitor SB505124 (15 μM) at the late gastrula (I-P) and very early neurula (I″-P″) stages or with the Hh inhibitor cyclopamine (1.75 μM) at the late gastrula (Y-F′) and very early neurula (Y″-F‴) stages. Embryos were fixed for in situ hybridization at the stages indicated. Anterior to the left in wholemounts. Cross-sections (pink), viewed from the posterior, are at the levels indicated with the lowercase letters in preceding wholemounts (blue). Arrows indicate the extent of the ventral endodermal domain in controls (T,V,T″,V″) and in embryos treated with cyclopamine at the late gastrula (B′,D′) and early neurula (B‴,D‴) stages. Scale bars: 50 μm.

Fig. 5.

Transient inhibition of Hh, but not Nodal, reduces Gli expression in amphioxus embryos. Expression of Gli in control embryos treated with DMSO for SB505124 (A-H, A″-H″) and ethanol (ETOH) for cyclopamine (Q-X, Q″-X″) and in those treated for 20 min with the Nodal inhibitor SB505124 (15 μM) at the late gastrula (I-P) and very early neurula (I″-P″) stages or with the Hh inhibitor cyclopamine (1.75 μM) at the late gastrula (Y-F′) and very early neurula (Y″-F‴) stages. Embryos were fixed for in situ hybridization at the stages indicated. Anterior to the left in wholemounts. Cross-sections (pink), viewed from the posterior, are at the levels indicated with the lowercase letters in preceding wholemounts (blue). Arrows indicate the extent of the ventral endodermal domain in controls (T,V,T″,V″) and in embryos treated with cyclopamine at the late gastrula (B′,D′) and early neurula (B‴,D‴) stages. Scale bars: 50 μm.

In addition to determining the effects of Nodal and Hh inhibition on these genes themselves, and on the direct Hh target Gli, we examined the effects of their inhibition on five other genes expressed in the amphioxus pharynx at the neurula and early larval stages. These include Pax1/9, normally expressed throughout the pharyngeal endoderm at the early neurula and subsequently downregulated in the gill slit primordia; Pax2/5/8, expressed where the gill slits will penetrate; Six1/2, expressed in the ventral and lateral endoderm; IrxC, co-expressed in the pharyngeal endoderm with Six1/2; and Tbx1/10, expressed in the endoderm and mesoderm of the developing branchial bars. The effect of Nodal inhibition on the domain of Pax1/9 is in proportion to the effect on the size and position of the gill slit primordia (Fig. 6A-P,A″-P″). Just as Nodal inhibition shifts the gill slits ventrally, it shifts the endodermal domain of Pax1/9 to the left so that the region in which Pax1/9 is downregulated continues to coincide with the region in which the gill slit primordia will develop. The degree of shifting to the left is similar whether SB505124 is added at the late gastrula or early neurula stages. These results suggest that Pax1/9 is downstream of Nodal and functions to maintain the undifferentiated state of the pharyngeal endoderm outside the gill slit primordia. In contrast, the primary effect of cyclopamine, as it does not change the position of the gill slits, is to delay development such that Pax1/9 is downregulated in the endoderm somewhat later than in the controls (compare Fig. 6S,T with A′,B′, and S″,T″ with A‴,B‴).

Fig. 6.

Transient inhibition of Nodal, but not Hh, shifts expression of Pax1/9 in amphioxus embryos. Expression of Pax1/9 in control embryos treated with DMSO for SB505124 (A-H, A″-H″) or ethanol (ETOH) for cyclopamine (Q-X, Q″-X″) and in those treated for 20 min with the Nodal inhibitor SB505124 (15 μM) at the late gastrula (I-P) and early neurula (I″-P″) stages or with the Hh inhibitor cyclopamine (1.75 μM) at the late gastrula (Y-F′) and early neurula (Y″-F‴) stages. Embryos were fixed for in situ hybridization at the stages indicated. Anterior to the left in wholemounts. Cross-sections (pink), viewed from the posterior, are at the levels indicated with the lowercase letters in preceding wholemounts (blue). Arrows indicate boundaries of Pax1/9 expression in the pharyngeal endoderm. Scale bars: 50 μm.

Fig. 6.

Transient inhibition of Nodal, but not Hh, shifts expression of Pax1/9 in amphioxus embryos. Expression of Pax1/9 in control embryos treated with DMSO for SB505124 (A-H, A″-H″) or ethanol (ETOH) for cyclopamine (Q-X, Q″-X″) and in those treated for 20 min with the Nodal inhibitor SB505124 (15 μM) at the late gastrula (I-P) and early neurula (I″-P″) stages or with the Hh inhibitor cyclopamine (1.75 μM) at the late gastrula (Y-F′) and early neurula (Y″-F‴) stages. Embryos were fixed for in situ hybridization at the stages indicated. Anterior to the left in wholemounts. Cross-sections (pink), viewed from the posterior, are at the levels indicated with the lowercase letters in preceding wholemounts (blue). Arrows indicate boundaries of Pax1/9 expression in the pharyngeal endoderm. Scale bars: 50 μm.

Pax2/5/8 is first expressed in the pharyngeal endoderm at mid-neurula in a pattern complementary to that of Pax1/9 (Figs 6C,D and 7C,D). In embryos treated with SB505124 at either the late gastrula or early neurula stages, the pharyngeal domain of Pax2/5/8 expression is reduced (Fig. 7A-P,A″-P″), although the reduction is more pronounced when SB505124 is added at the earlier time. For example, when the inhibitor is added at the late gastrula stage, the domain of Pax2/5/8 associated with the mouth (Fig. 7G,H,O,P) is reduced and those associated with the first two gill slits are eliminated (arrows in Fig. 7G,H,O,P). When the Nodal inhibitor is added at the early neurula stage, there is only a slight reduction of the Pax2/5/8 domain in the pharynx (Fig. 7A″-P″). The addition of cyclopamine at either time has no clear effect on the Pax2/5/8 domain, except that, as development is delayed, initial expression of Pax2/5/8 in the pharynx is also delayed. These results indicate that expression of Pax2/5/8 in the pharynx probably requires Nodal signaling, but that Hh may not regulate Pax2/5/8.

Fig. 7.

Transient inhibition of Nodal, but not Hh, reduces Pax2/5/8 expression in amphioxus embryos. Expression of Pax2/5/8 in control embryos treated with DMSO for SB505124 (A-H, A″-H″) and ethanol (ETOH) for cyclopamine (Q-X, Q″-X″) and in those treated for 20 min with the Nodal inhibitor SB505124 (15 μM) at the late gastrula (I-P) and early neurula (I″-P″) stages or with the Hh inhibitor cyclopamine (1.75 μM) at the late gastrula (Y-F′) and early neurula (Y″-F‴) stages. Embryos were fixed for in situ hybridization at the stages indicated. Anterior to the left in wholemounts. Cross-sections (pink), viewed from the posterior, are at the levels indicated with the lowercase letters in preceding wholemounts (blue). Arrows indicate the extent of the ventral endodermal domain that marks the future edges of the gill slits and is reduced in SB505124-treated larvae. Scale bars: 50 μm.

Fig. 7.

Transient inhibition of Nodal, but not Hh, reduces Pax2/5/8 expression in amphioxus embryos. Expression of Pax2/5/8 in control embryos treated with DMSO for SB505124 (A-H, A″-H″) and ethanol (ETOH) for cyclopamine (Q-X, Q″-X″) and in those treated for 20 min with the Nodal inhibitor SB505124 (15 μM) at the late gastrula (I-P) and early neurula (I″-P″) stages or with the Hh inhibitor cyclopamine (1.75 μM) at the late gastrula (Y-F′) and early neurula (Y″-F‴) stages. Embryos were fixed for in situ hybridization at the stages indicated. Anterior to the left in wholemounts. Cross-sections (pink), viewed from the posterior, are at the levels indicated with the lowercase letters in preceding wholemounts (blue). Arrows indicate the extent of the ventral endodermal domain that marks the future edges of the gill slits and is reduced in SB505124-treated larvae. Scale bars: 50 μm.

The endodermal domains of Six1/2 and IrxC in normal neurulae largely overlap (compare Figs 8A-P,Q″-F‴ with 9A-P,Q″-F‴). By the early larva stage, however, the pharyngeal domains of both genes become restricted to the mouth and gill slit primordia (Figs 8G,H and 9G,H). In fact, by mid to late neurula, the pharyngeal domain of IrxC is congruent with the area in which Pax1/9 becomes downregulated around the gill slits, although that of Six1/2 is somewhat broader than this region. The pharyngeal domains of both genes are shifted to the left when SB505124 is added at the late gastrula stage (Figs 8A-P and 9A-P). However, there is no clear effect on expression of either gene when SB505124 is added at early neurula (Figs 8A″-P″ and 9A″-P″). There is also no clear effect of cyclopamine when it is added at late gastrula or early neurula (Figs 8Q-F′,Q″-F‴ and 9Q-F′,Q″-F‴). Thus, both Six1/2 and IrxC appear to be downstream of Nodal signaling, but are not downstream of Hh.

Fig. 8.

Transient inhibition of Nodal, but not Hh, shifts expression of Six1/2 in amphioxus embryos. Expression of Six1/2 in control embryos treated with DMSO for SB505124 (A-H, A″-H″) and ethanol (ETOH) for cyclopamine (Q-X, Q″-X″) and in those treated for 20 min with the Nodal inhibitor SB505124 (15 μM) at the late gastrula (I-P) and very early neurula (I″-P″) stages or with the Hh inhibitor cyclopamine (1.75 μM) at the late gastrula (Y-F′) and early neurula (Y″-F‴) stages. Embryos were fixed for in situ hybridization at the stages indicated. Anterior to the left in wholemounts. Cross-sections (pink), viewed from the posterior, are at the levels indicated with the lowercase letters in preceding wholemounts (blue). Arrows indicate the borders of the ventral endodermal domain, which is skewed to the right in normal larvae, but shifted to the left in embryos treated at the early gastrula stage with the SB505124. Scale bars: 50 μm.

Fig. 8.

Transient inhibition of Nodal, but not Hh, shifts expression of Six1/2 in amphioxus embryos. Expression of Six1/2 in control embryos treated with DMSO for SB505124 (A-H, A″-H″) and ethanol (ETOH) for cyclopamine (Q-X, Q″-X″) and in those treated for 20 min with the Nodal inhibitor SB505124 (15 μM) at the late gastrula (I-P) and very early neurula (I″-P″) stages or with the Hh inhibitor cyclopamine (1.75 μM) at the late gastrula (Y-F′) and early neurula (Y″-F‴) stages. Embryos were fixed for in situ hybridization at the stages indicated. Anterior to the left in wholemounts. Cross-sections (pink), viewed from the posterior, are at the levels indicated with the lowercase letters in preceding wholemounts (blue). Arrows indicate the borders of the ventral endodermal domain, which is skewed to the right in normal larvae, but shifted to the left in embryos treated at the early gastrula stage with the SB505124. Scale bars: 50 μm.

Fig. 9.

Transient inhibition of Nodal, but not Hh, shifts expression of IrxC in amphioxus embryos. Expression of IrxC in control embryos treated with DMSO for SB505124 (A-H, A″-H″) and ethanol (ETOH) for cyclopamine (Q-X, Q″-X″) and in those treated for 20 min with the Nodal inhibitor SB505124 (15 μM) at the late gastrula (I-P) and early neurula (I″-P″) stages or with the Hh inhibitor cyclopamine (1.75 μM) at the late gastrula (Y-F′) and early neurula (Y″-F‴) stages. Embryos were fixed for in situ hybridization at the stages indicated. Anterior to the left in wholemounts. Cross-sections (pink), viewed from the posterior, are at the levels indicated with the lowercase letters in preceding wholemounts (blue). Arrows indicate boundaries of endodermal domain of IrxC, which is shifted to the left by treatment with SB505124 at the early gastrula stage but not at the early neurula stage or by treatment with cyclopamine. Scale bars: 50 μm.

Fig. 9.

Transient inhibition of Nodal, but not Hh, shifts expression of IrxC in amphioxus embryos. Expression of IrxC in control embryos treated with DMSO for SB505124 (A-H, A″-H″) and ethanol (ETOH) for cyclopamine (Q-X, Q″-X″) and in those treated for 20 min with the Nodal inhibitor SB505124 (15 μM) at the late gastrula (I-P) and early neurula (I″-P″) stages or with the Hh inhibitor cyclopamine (1.75 μM) at the late gastrula (Y-F′) and early neurula (Y″-F‴) stages. Embryos were fixed for in situ hybridization at the stages indicated. Anterior to the left in wholemounts. Cross-sections (pink), viewed from the posterior, are at the levels indicated with the lowercase letters in preceding wholemounts (blue). Arrows indicate boundaries of endodermal domain of IrxC, which is shifted to the left by treatment with SB505124 at the early gastrula stage but not at the early neurula stage or by treatment with cyclopamine. Scale bars: 50 μm.

The final gene we examined was Tbx1/10, normally expressed in the ventral endoderm and lateral mesoderm on the left side that will give rise to the mesoderm of the branchial bars. The only effect of Nodal inhibition at the late gastrula stage is the disappearance of the domain of Tbx1/10 in the lateral mesoderm (Fig. 10D-L,P-X). There is no effect on Tbx1/10 expression when SB505124 addition is delayed until the early neurula stage (Fig. 10A″-X″) and no clear effect on Tbx1/10 expression of cyclopamine added at either late gastrula or early neurula (Fig. 10Y-V′,Y″-V‴).

Fig. 10.

Transient inhibition of Nodal, but not Hh, affects mesodermal expression of Tbx1/10. Expression of Tbx1/10 in control embryos treated with DMSO for SB505124 (A-L, A″-L″) and ethanol (ETOH) for cyclopamine (Y-J′, Y″-J‴) and in those treated for 20 min with the Nodal inhibitor SB505124 (15 μM) at the late gastrula (M-X) and early neurula (M″-X″) stages or with the Hh inhibitor cyclopamine (1.75 μM) at the late gastrula (K′-V′) and early neurula (K‴-V‴) stages. Embryos were fixed for in situ hybridization at the stages indicated. Anterior to the left in wholemounts. Cross-sections (pink), viewed from the posterior, are at the levels indicated with the lowercase letters in preceding wholemounts (blue). Arrows indicate boundaries of endodermal domain of Tbx1/10, which is shifted to the left by treatment with SB505124 at the early gastrula stage but not at the early neurula stage or by treatment with cyclopamine. Arrowheads indicate the lateral mesoderm on the left. Scale bars: 50 μm.

Fig. 10.

Transient inhibition of Nodal, but not Hh, affects mesodermal expression of Tbx1/10. Expression of Tbx1/10 in control embryos treated with DMSO for SB505124 (A-L, A″-L″) and ethanol (ETOH) for cyclopamine (Y-J′, Y″-J‴) and in those treated for 20 min with the Nodal inhibitor SB505124 (15 μM) at the late gastrula (M-X) and early neurula (M″-X″) stages or with the Hh inhibitor cyclopamine (1.75 μM) at the late gastrula (K′-V′) and early neurula (K‴-V‴) stages. Embryos were fixed for in situ hybridization at the stages indicated. Anterior to the left in wholemounts. Cross-sections (pink), viewed from the posterior, are at the levels indicated with the lowercase letters in preceding wholemounts (blue). Arrows indicate boundaries of endodermal domain of Tbx1/10, which is shifted to the left by treatment with SB505124 at the early gastrula stage but not at the early neurula stage or by treatment with cyclopamine. Arrowheads indicate the lateral mesoderm on the left. Scale bars: 50 μm.

Taken together, these results show that the precise left-right position of the gill slits is largely mediated by levels of Nodal signaling at the late gastrula/early neurula stages, with higher levels of Nodal shifting gill slits to the right and lower levels shifting them to the left. A Nodal feedback loop reinforces Nodal signaling. The expression domains of genes that are either downregulated in the future gill slit primordia (Pax1/9) or expressed in the gill slit primordia (Six1/2, IrxC, Pax2/5/8 and, at the larval stage, Gli) are shifted to the left as Nodal signaling is reduced. Surprisingly, a reduction in Nodal signaling at the late gastrula stage strongly reduces Tbx1/10 expression in the branchial arches. Thus, all of these genes are downstream of Nodal signaling (Fig. 11). In addition, Nodal signaling is required for gill slit penetration and may act by positively regulating Pax2/5/8 where the ectoderm and endoderm will fuse to form gill slits, and/or by regulating Hh, which is also required for gill slit penetration.

Fig. 11.

Diagram of gene interactions at the neurula stage in development of the amphioxus gill slits. The interactions shown may be either indirect or direct. Nodal expression is restricted to the left in all three germ layers. Nodal auto-regulates. Secreted Nodal protein diffuses towards the right. A high level of Nodal activates Hh expression in endoderm on the left. Very low levels of Nodal protein on the right are required for expression of Pax2/5/8, IrxC and Six1/2 in the gill slit primordia. Expression of these genes is inhibited on the left by high levels of Nodal protein. Hh protein secreted on the left diffuses toward the right. Hh in turn regulates the size of the Gli domain in the pharyngeal endoderm. Tbx1/10 expression in mesoderm on the left requires Nodal signaling. Nodal signaling maintains expression of Pax1/9 in the pharyngeal endoderm, except where gill slits will form.

Fig. 11.

Diagram of gene interactions at the neurula stage in development of the amphioxus gill slits. The interactions shown may be either indirect or direct. Nodal expression is restricted to the left in all three germ layers. Nodal auto-regulates. Secreted Nodal protein diffuses towards the right. A high level of Nodal activates Hh expression in endoderm on the left. Very low levels of Nodal protein on the right are required for expression of Pax2/5/8, IrxC and Six1/2 in the gill slit primordia. Expression of these genes is inhibited on the left by high levels of Nodal protein. Hh protein secreted on the left diffuses toward the right. Hh in turn regulates the size of the Gli domain in the pharyngeal endoderm. Tbx1/10 expression in mesoderm on the left requires Nodal signaling. Nodal signaling maintains expression of Pax1/9 in the pharyngeal endoderm, except where gill slits will form.

In contrast, although inhibition of Hh signaling at either the late gastrula or early neurula strongly inhibits gill slit formation, the effects on expression of pharyngeal markers (for example Pax1/9) were very mild and could be because of a general developmental delay. The ventral endodermal domain of Gli is somewhat reduced. However, even in the most severe phenotype, with imperforate gill slits, generated by cyclopamine treatment, reducing Hh signaling has no effect on the position of the gill slits. Even when gill slits completely fail to penetrate, the expression of pharyngeal markers in the gill slit primordia is essentially unchanged (Fig. S2, Fig. 10). Thus, at the late gastrula and early neurula stages, Hh appears to play little, if any, role in establishing the position of the gill slits and gill slit primordia, but does affect gill slit penetration.

Stage-specific signaling during embryogenesis

The above results on the roles of Nodal and Hh signaling in early embryonic patterning of the amphioxus pharynx emphasize the rapidly changing roles of signaling pathways in early development. For example, during cleavage stages through to the mid-gastrula stage of amphioxus, Nodal is expressed dorsally and acts in opposition to BMPs to specify neuroectoderm and dorsal-anterior identity (Onai et al., 2010). Thus, inhibition of Nodal with SB505124 from cleavage stages onward results in loss of dorsal and anterior identity (Onai et al., 2010). Not surprisingly, Nodal has been identified as a neural inducer in amphioxus (Le Petillon et al., 2017). However, as represented in Fig. 11, at the late gastrula/early neurula stage, the role of Nodal changes; Nodal expression is downregulated on the right. From then into the neurula stage, Nodal signaling regulates left-right asymmetry (Soukup et al., 2015; Li et al., 2017). Correspondingly, in very early embryos, overexpression of the Nodal inhibitor Cerberus, normally expressed on the right, causes the loss of dorsal and anterior structures (Onai et al., 2010). A mutation of Cerberus results in a milder phenotype, with the right side of amphioxus embryos adopting a left-side identity and the gill slits shifted ventrally (Li et al., 2017). Null mutants of Hh, similar to the embryos in which Nodal is inhibited during the neurula stage, have ventralized gill slits and lack Cerberus expression, suggesting that Hh may indirectly regulate Nodal via Cerberus (Hu et al., 2017). However, our experiments, carried out at the late gastrula through to the neurula stage, after Cerberus ceases to be expressed on the right, indicate that if Nodal does regulate left-right asymmetry, in part through Hh, it must do so very early in development. By the early neurula stage, when inhibiting Nodal affects left-right asymmetry (Soukup et al., 2015), inhibiting Hh signaling for 20 min has little or no effect on either Nodal expression or on the left-right position of the gill slits (Fig. 3). Instead, it is the opposite: inhibiting Nodal signaling for 20 min at late gastrula/very early neurula inhibits expression of Hh in the pharyngeal endoderm (Fig. 4).

Not only does Nodal signaling at the late gastrula/early neurula stage control the lateral position of the gill slits, it also regulates the size of the gill slits. In severe phenotypes caused by inhibition of Nodal signaling, gill slits do not penetrate; gill slit primordia can be fused or be completely absent. These effects could be mediated by reduced Hh expression in the pharyngeal endoderm and/or by reduced Tbx1/10 expression in the pharyngeal mesoderm. Reduced Nodal signaling at the late gastrula stage reduces the endodermal domain of Hh and shifts it to the left (Fig. 4), and a reduction in Hh signaling can completely eliminate the gill slit primordia (Fig. 2D). Inhibition of Nodal signaling at the late gastrula stage also eliminates mesodermal expression of Tbx1/10 (Fig. 10F,I,R,U), and knockdown of Tbx1/10 eliminates the gill slits and branchial bars (Koop et al., 2014).

Surprisingly, the Hh-null mutants, as in the milder phenotype of cyclopamine-treated larvae, have been reported to develop gill slits (Hu et al., 2017). Although we would have expected the null mutants to lack gill slits, as in the more severe phenotype resulting from cyclopamine treatment at the late gastrula/very early neurula stages (Fig. S2), it could be that the most severely affected Hh mutants failed to develop to the larval stage. Alternatively, although cyclopamine has been shown to be a very specific inhibitor of Hh signaling (Chen et al., 2002a,b), off-target effects cannot be completely ruled out (Meyers-Needham et al., 2012).

In addition to Nodal and Hh, other signaling pathways also function in early embryonic patterning. For example, during the gastrula stage, when Nodal and BMP act in opposition to establish dorsoventral identity in amphioxus embryos, retinoic acid (RA) signaling establishes position along the anterior-posterior axis (Koop et al., 2010), and Wnt signaling concomitantly specifies the posterior end of the amphioxus embryo. RA signaling also has a late role in pharyngeal patterning; for gill slits to penetrate, RA levels are kept low in the gill slit primordia by the RA antagonist Tr2/4 (Escriva et al., 2002; Koop et al., 2014). Fgf and Wnt genes are also expressed in the developing amphioxus pharynx, but their roles in pharyngeal patterning have not yet been reported (Bertrand et al., 2011, 2015; Schubert et al., 2000, 2001).

Nodal and Hh function in pharyngeal patterning in vertebrates

Nodal and Hh genes are both involved in patterning vertebrate embryos, but whether they interact in pharyngeal patterning is not clear. In vertebrates, Nodal signaling is required for endoderm specification and development of the thyroid, a derivative of the pharyngeal endoderm (Elsalini et al., 2003; Grapin-Botton and Constam, 2007; Porazzi et al., 2009). Similar to amphioxus, in which expression of Hh in the pharynx is highest immediately anterior to the first gill slit primordium (Fig. 4), in vertebrates, Shh is strongly expressed in the endoderm and ectoderm of the first pharyngeal arch, where it is essential for development of mandibular arch derivatives (Brito et al., 2006, 2008; Haworth et al., 2007; Gillis et al., 2009; Swartz et al., 2012; Billmyre and Klingensmith, 2015; Dworkin et al., 2016). Shh signaling is low in the posterior pharyngeal pouches, allowing the parathyroid marker Gcm2 to be expressed (Grevellec et al., 2011).

Shh interacts with a variety of genes in pharyngeal patterning (Yamagishi et al., 2006; Billmyre and Klingensmith, 2015; Xavier et al., 2016). For example, ectopic expression of Shh in the endoderm of the pharyngeal pouches of the mouse induces expression of Tbx1 (Yamagishi et al., 2003; Bain et al., 2016). In Shh−/− mice, the pharynx is initially patterned correctly, but the first arch atrophies and the first pharyngeal pouch does not form. The second, third and fourth pharyngeal arches are small (Moore-Scott and Manley, 2005). Like Nodal, Shh signaling is also required for normal development of the thyroid (Fagman et al., 2004; Moore-Scott and Manley, 2005; Bain et al., 2016; Figueiredo et al., 2016). Nodal and Hh cooperate in patterning the mesoderm, but whether they act together or in parallel in thyroid development has not been addressed. In some models, initial Nodal asymmetry was proposed to be downstream of Shh. For example, in the chick, Shh inhibition in early development represses Nodal in lateral plate and paraxial mesoderm (Otto et al., 2014). In contrast, proNodal, but not mature Nodal, acts indirectly to maintain Shh expression in the prechordal mesoderm (Ellis et al., 2015). However, it has also been suggested that other factors may regulate asymmetry of both genes in the paraxial mesoderm (Tsikolia et al., 2012).

Interactions of Nodal and Hh in development of the neural tube have been well documented in vertebrates. Mutations or chemical inhibition of either Nodal or Hh genes results in holoprosencephaly (Monuki, 2007; Mercier et al., 2013). In patterning the ventral neural tube, Nodal is upstream of Shh, which contains a Nodal-responsive enhancer (Muller et al., 2000; Ito et al., 2001; Rohr et al., 2001; Lupo et al., 2006). In addition, Nodal and Shh both function in the development of the hypothalamus (Mathieu et al., 2002), and both Nodal and Hh signaling in the ventral diencephalon of the zebrafish are regulated by Zic1 (Maurus and Harris, 2009).

Nodal and Hh in other invertebrate deuterostomes

In invertebrate deuterostomes other than amphioxus, Nodal and Hh sometimes interact. In sea urchin embryos, Nodal signaling establishes left-right asymmetry, and Hh signaling helps to maintain Nodal expression once asymmetry is established (Materna et al., 2013; Warner et al., 2016). Whether Nodal and Hh interact in patterning developing hemichordates is unknown. In larvae of the indirectly developing hemichordate Ptychodera flava, Nodal inhibition disrupts mesoderm formation and dorsoventral fates (Rottinger et al., 2015). Hh is expressed in the anterior region and pharyngeal endoderm of P. flava larvae (Arimoto and Tagawa, 2015). In another indirectly developing hemichordate, Balanoglossus simodensis, Hh is expressed in the dorsal endoderm and the stomochord (an anterior projection from the gut) (Miyamoto and Wada, 2013). In the direct developer Saccoglossus kowalevskii, Hh is expressed in the anterior ectoderm and anterior gut (Lowe et al., 2006). Expression of these genes in the developing adult pharynx of hemichordates has not been reported.

The emerging picture is that a role for Nodal in patterning the left-right axis existed at the base of the Bilateria and has been preserved in both spiralians and chordates (Grande and Patel, 2009; Martín-Durán et al., 2016). Subsequently, in deuterostomes, additional functions for these genes in patterning the dorsoventral and anterior-posterior axes arose. With multiple signaling pathways, including Hh and RA signaling working at the same time, and the same signaling pathways doing different things at different times, the developing picture of the genetic mechanisms that pattern the pharynx and establish the gill slits is increasingly complex. However, in the absence of whole-genome duplications, amphioxus is optimally positioned for understanding the fundamentals of pharyngeal patterning in chordates. Comparisons between amphioxus and vertebrates have shown that the ancestral chordate had a pharynx with gill slits and an endostyle that was patterned by signaling pathways including Nodal, Hh, RA and Wnt. Importantly, the roles of these pathways in embryogenesis change with time. First, they act simultaneously to specify the anterior-posterior and dorsoventral axes, then during the gastrula stage they divide the embryo into regions along these axes, and finally, during the neurula stage, they specify discrete tissues and organs. An important step towards understanding how the amphioxus genome generates a phenotype would be to further dissect the gene networks downstream of these pathways.

Embryos and pharmacological treatments

Breeding cultures of the Florida amphioxus (Branchiostoma floridae) are maintained in L.Z.H.’s laboratory. Larvae and adults are fed on brown phytoplankton (Isochrysis, Pavlova and Tisochrysis) originally obtained from NCMA at the Bigelow Laboratory for Ocean Sciences (East Boothbay, ME, USA). Animals are raised at 25°C. Adults are kept on a 10 h dark/14 h light cycle. Ripe animals are then maintained at 17°C for at least 2 weeks. When returned to 25°C, 20-80% of the animals usually spawned 24-28 h later when the light was turned off. Nodal was inhibited with SB505124 (Medchem Express). A stock solution of 50 mM was made in DMSO and used at final concentrations of 10-50 μM. Hh signaling was inhibited with cyclopamine (LC Laboratories). A stock solution of 50 mM was made in 100% ethanol (ETOH) and added to embryos at concentrations of 1.5-4.0 μM. Embryos were treated for 20 min with each inhibitor at the late gastrula (9 hpf) and early neurula stages (10 hpf). Controls were treated with DMSO or ETOH alone. To confirm that cyclopamine inhibits Hh signaling in amphioxus, we treated with 1.75 µM cyclopamine at the very early gastrula stage and fixed for in situ hybridization at the late gastrula/early neurula, early to mid-neurula and early larval stages. The experiment was repeated three times. Aliquots of embryos were fixed with 4% paraformaldehyde, 0.1 M MOPS, 1 mM EGTA, 2 mM MgSO4 and 0.5 M NaCl, and stored in 70% ETOH at −20°C as previously described (Holland et al., 1996). Pilot experiments to determine the optimal concentrations of SB505124 and cyclopamine, and the times of treatment, were carried out in duplicate. Experiments for determining the effects of these inhibitors on gene expression were carried out in triplicate. For each replicate, control and experimental samples were treated in parallel on embryos from the same egg batch.

In situ hybridization

Methods for in situ hybridization were as previously described (Yu and Holland, 2009) with the following modification: 50 µM EDTA was added to the PBST solution. Antisense riboprobes of AmphiNodal (AY083838), AmphiHedgehog (Y13858), AmphiPax1/9 (U20167), AmphiIrxC (EU754750), AmphiGli (CAB96572), AmphiPax2/5/8 (AF053762), AmphiTbx (AF262562) and AmphiSix1/2 (EF195742) were synthesized. Genbank accession numbers are in brackets. Templates for riboprobe synthesis of AmphiNodal, AmphiIrxCi, and AmphiPtch were amplified from EST clones (bfga04g04, CAXF10761 and GA035K03, respectively) in the pDONR222 vector. Vector-specific primers contained sites for T7 and Sp6 RNA polymerase (pDONR222-T7-reverse: 5′-TAATACGACTCACTATAGGGAGGGGATATCAGCTGGATG-3′; pDONR222-Sp6-forward: 5′-ATTTAGGTGACACTATAGAAGACGGCAGTCTTAAGCTC-3′). Controls and experimental samples were hybridized in parallel with the same riboprobes at the same concentrations, and color development was for the same period of time. For each in situ hybridization, 10 embryos of each stage were used. After in situ hybridization, the embryos were photographed in wholemount, and were stained with Ponceau-S, embedded in Spurr's resin and 3 μm sections cut with a glass knife as previously described (Holland et al., 1996). For each sample, three embryos of each stage were sectioned.

We thank Daniel Hartsook, Zaren Youngblood, Taylor Crabtree, Esmeralda Alcantar and Joel Lashomb for animal husbandry.

Author contributions

Conceptualization: L.Z.H.; Formal analysis: L.Z.H.; Investigation: H.O., D.K., L.Z.H.; Resources: L.Z.H.; Writing - original draft: L.Z.H.; Writing - review & editing: D.K.; Supervision: L.Z.H.; Project administration: L.Z.H.; Funding acquisition: L.Z.H.

Funding

This study was supported by a National Science Foundation grant (IOS 1353688 to L.Z.H.)

Arimoto
,
A.
and
Tagawa
,
K.
(
2015
).
Hedgehog expression during development and regeneration in the hemichordate, Ptychodera flava
.
Zool. Sci.
32
,
33
-
37
.
Bain
,
V. E.
,
Gordon
,
J.
,
O'Neil
,
J. D.
,
Ramos
,
I.
,
Richie
,
E. R.
and
Manley
,
N. R.
(
2016
).
Tissue-specific roles for sonic hedgehog signaling in establishing thymus and parathyroid organ fate
.
Development
143
,
4027
-
4037
.
Bertrand
,
N.
and
Dahmane
,
N.
(
2006
).
Sonic hedgehog signaling in forebrain development and its interactions with pathways that modify its effects
.
Trends Cell Biol.
16
,
597
-
605
.
Bertrand
,
S.
,
Camasses
,
A.
,
Somorjai
,
I.
,
Belgacem
,
M. R.
,
Chabrol
,
O.
,
Escande
,
M.-L.
,
Pontarotti
,
P.
and
Escriva
,
H.
(
2011
).
Amphioxus FGF signaling predicts the acquisition of vertebrate morphological traits
.
Proc. Natl. Acad. Sci. USA
108
,
9160
-
9165
.
Bertrand
,
S.
,
Aldea
,
D.
,
Oulion
,
S.
,
Subirana
,
L.
,
de Lera
,
A. R.
,
Somorjai
,
I.
and
Escriva
,
H.
(
2015
).
Evolution of the role of RA and FGF signals in the control of somitogenesis in chordates
.
PLoS ONE
10
,
e0136587
.
Billmyre
,
K. K.
and
Klingensmith
,
J.
(
2015
).
Sonic hedgehog from pharyngeal arch 1 epithelium is necessary for early mandibular arch cell survival and later cartilage condensation differentiation
.
Dev. Dyn.
244
,
564
-
576
.
Boorman
,
C. J.
and
Shimeld
,
S. M.
(
2002
).
The evolution of left–right asymmetry in chordates
.
BioEssays
24
,
1004
-
1011
.
Brito
,
J. M.
,
Teillet
,
M.-A.
and
Le Douarin
,
N. M.
(
2006
).
An early role for sonic hedgehog from foregut endoderm in jaw development: ensuring neural crest cell survival
.
Proc. Natl. Acad. Sci. USA
103
,
11607
-
11612
.
Brito
,
J. M.
,
Teillet
,
M.-A.
and
Le Douarin
,
N. M.
(
2008
).
Induction of mirror-image supernumerary jaws in chicken mandibular mesenchyme by Sonic Hedgehog-producing cells
.
Development
135
,
2311
-
2319
.
Chen
,
J. K.
,
Taipale
,
J.
,
Cooper
,
M. K.
and
Beachy
,
P. A.
(
2002a
).
Inhibition of Hedgehog signaling by direct binding of cyclopamine to Smoothened
.
Genes Dev.
16
,
2743
-
2748
.
Chen
,
J. K.
,
Taipale
,
J.
,
Young
,
K. E.
,
Maiti
,
T.
and
Beachy
,
P. A.
(
2002b
).
Small molecule modulation of Smoothened activity
.
Proc. Natl. Acad. Sci. U.S.A.
99
,
14071
-
14076
.
Dworkin
,
S.
,
Boglev
,
Y.
,
Owens
,
H.
and
Goldie
,
S. J.
(
2016
).
The role of sonic hedgehog in craniofacial patterning, morphogenesis and cranial neural crest survival
.
J. Dev. Biol.
4
,
24
.
Dyer
,
L. A.
and
Kirby
,
M. L.
(
2009
).
Sonic hedgehog maintains proliferation in secondary heart field progenitors and is required for normal arterial pole formation
.
Dev. Biol.
330
,
305
-
317
.
Ellis
,
P. S.
,
Burbridge
,
S.
,
Soubes
,
S.
,
Ohyama
,
K.
,
Ben-Haim
,
N.
,
Chen
,
C.
,
Dale
,
K.
,
Shen
,
M. M.
,
Constam
,
D.
and
Placzek
,
M.
(
2015
).
ProNodal acts via FGFR3 to govern duration of Shh expression in the prechordal mesoderm
.
Development
142
,
3821
-
3832
.
Elsalini
,
O. A.
,
Gartzen
,
J. V.
,
Cramer
,
M.
and
Rohr
,
K. B.
(
2003
).
Zebrafish hhex, nk2.1a, and pax2.1 regulate thyroid growth and differentiation downstream of Nodal-dependent transcription factors
.
Dev. Biol.
263
,
67
-
80
.
Escriva
,
H.
,
Holland
,
N. D.
,
Gronemeyer
,
H.
,
Laudet
,
V.
and
Holland
,
L. Z.
(
2002
).
The retinoic acid signaling pathway regulates anterior/posterior patterning in the nerve cord and pharynx of amphioxus, a chordate lacking neural crest
.
Development
129
,
2905
-
2916
.
Fagman
,
H.
,
Grände
,
M.
,
Gritli-Linde
,
A.
and
Nilsson
,
M.
(
2004
).
Genetic deletion of Sonic Hedgehog causes hemiagenesis and ectopic development of the thyroid in mouse
.
Am. J. Pathol.
164
,
1865
-
1872
.
Figueiredo
,
M.
,
Silva
,
J. C.
,
Santos
,
A. S.
,
Proa
,
V.
,
Alcobia
,
I.
,
Zilhão
,
R.
,
Cidadão
,
A.
and
Neves
,
H.
(
2016
).
Notch and Hedgehog in the thymus/parathyroid common primordium: crosstalk in organ formation
.
Dev. Biol.
418
,
268
-
282
.
Garg
,
V.
,
Yamagishi
,
C.
,
Hu
,
T.
,
Kathiriya
,
I. S.
,
Yaagishi
,
H.
and
Srivastava
,
D.
(
2001
).
Tbx1, a DiGeorge syndrome candidate gene, is regulated by sonic hedgehog during pharyngeal arch development
.
Dev. Biol.
235
,
62
-
73
.
Gillis
,
J. A.
,
Dahn
,
R. D.
and
Shubin
,
N. H.
(
2009
).
Shared developmental mechanisms pattern the vertebrate gill arch and paired fin skeletons
.
Proc. Natl. Acad. Sci. USA
106
,
5720
-
5724
.
Grande
,
C.
and
Patel
,
N. H.
(
2009
).
Nodal signalling is involved in left-right asymmetry in snails
.
Nature
457
,
1007
-
1011
.
Grapin-Botton
,
A.
and
Constam
,
D.
(
2007
).
Evolution of the mechanisms and molecular control of endoderm formation
.
Mech. Dev.
124
,
253
-
278
.
Grevellec
,
A.
,
Graham
,
A.
and
Tucker
,
A. S.
(
2011
).
Shh signalling restricts the expression of Gcm2 and controls the position of the developing parathyroids
.
Dev. Biol.
353
,
194
-
205
.
Haworth
,
K. E.
,
Wilson
,
J. M.
,
Grevellec
,
A.
,
Cobourne
,
M. T.
,
Healy
,
C.
,
Helms
,
J. A.
,
Sharpe
,
P. T.
and
Tucker
,
A. S.
(
2007
).
Sonic hedgehog in the pharyngeal endoderm controls arch pattern via regulation of Fgf8 in head ectoderm
.
Dev. Biol.
303
,
244
-
258
.
Holland
,
N. D.
,
Holland
,
L. Z.
and
Kozmik
,
Z.
(
1995
).
An amphioxus Pax gene, AmphiPax-1, expressed in embryonic endoderm, but not in mesoderm: implications for evolution of classI paired box genes
.
Mol. Mar. Biol. Biotechnol.
35
,
206
-
214
.
Holland
,
L. Z.
,
Holland
,
P. W. H.
and
Holland
,
N. D.
(
1996
).
Revealing homologies between body parts of distantly related animals by in situ hybridization to developmental genes: amphioxus versus vertebrates
. In
Molecular Approaches to Zoology and Evolution
(ed.
S.
Palumbi
and
J. D.
Ferraris
), pp.
267
-
282
.
New York
:
John Wiley
;
473-483
.
Hu
,
G.
,
Li
,
G.
,
Wang
,
H.
and
Wang
,
Y.
(
2017
).
Hedgehog participates in the establishment of left-right asymmetry during amphioxus development by controlling Cerberus expression
.
Development
144
,
4694
-
4703
.
Ito
,
Y.
,
Kuhara
,
S.
and
Tashiro
,
K.
(
2001
).
In synergy with Noggin and Follistatin, Xenopus Nodal-Related gene induces Sonic Hedgehog on notochord and floor plate
.
Biochem. Biophys. Res. Commun.
281
,
714
-
719
.
Jenkins
,
D.
(
2009
).
Hedgehog signalling: Emerging evidence for non-canonical pathways
.
Cell. Signal.
21
,
1023
-
1034
.
Koop
,
D.
,
Holland
,
N. D.
,
Sémon
,
M.
,
Alvarez
,
S.
,
de Lera
,
A. R.
,
Laudet
,
V.
,
Holland
,
L. Z.
and
Schubert
,
M.
(
2010
).
Retinoic acid signaling targets Hox genes during the amphioxus gastrula stage: Insights into early anterior-posterior patterning of the chordate body plan
.
Dev. Biol.
338
,
98
-
106
.
Koop
,
D.
,
Chen
,
J.
,
Theodosiou
,
M.
,
Carvalho
,
J. E.
,
Alvarez
,
S.
,
de Lera
,
A. R.
,
Holland
,
L. Z.
and
Schubert
,
M.
(
2014
).
Roles of retinoic acid and Tbx1/10 in pharyngeal segmentation: amphioxus and the ancestral chordate condition
.
Evodevo
5
,
1
-
16
.
Le Petillon
,
Y.
,
Luxardi
,
G.
,
Scerbo
,
P.
,
Cibois
,
M.
,
Leon
,
A.
,
Subirana
,
L.
,
Irimia
,
M.
,
Kodjabachian
,
L.
,
Escriva
,
H.
and
Bertrand
,
S.
(
2017
).
Nodal-activin pathway is a conserved neural induction signal in chordates
.
Nat. Ecol. Evol.
1
,
1192
-
1200
.
Li
,
G.
,
Liu
,
X.
,
Xing
,
C.
,
Zhang
,
H.
,
Shimeld
,
S. M.
and
Wang
,
Y.
(
2017
).
Cerberus–Nodal–Lefty–Pitx signaling cascade controls left–right asymmetry in amphioxus
.
Proc. Natl. Acad. Sci. USA
114
,
3684
-
3689
.
Lowe
,
C. J.
,
Terasaki
,
M.
,
Wu
,
M.
,
Freeman
,
R. M.
, Jr
,
Runft
,
L.
,
Kwan
,
K.
,
Haigo
,
S.
,
Aronowicz
,
J.
,
Lander
,
E.
,
Gruber
,
C.
, et al. 
(
2006
).
Dorsoventral patterning in hemichordates: Insights into early chordate evolution
.
PLoS Biol.
4
,
e291
.
Luo
,
K.
(
2017
).
Signaling cross talk between TGF-β/Smad and other signaling pathways
.
Cold Spring Harb. Perspect. Biol.
9
,
a022137
.
Lupo
,
G.
,
Harris
,
W. A.
and
Lewis
,
K. E.
(
2006
).
Mechanisms of ventral patterning in the vertebrate nervous system
.
Nat. Rev. Neurosci.
7
,
103
-
114
.
Martín-Durán
,
J. M.
,
Vellutini
,
B. C.
and
Hejnol
,
A.
(
2016
).
Embryonic chirality and the evolution of spiralian left–right asymmetries
.
Philos. Trans. R. Soc. B Biol. Sci.
371
,
20150411
.
Massagué
,
J.
(
2012
).
TGF-β signalling in context
.
Nat. Rev. Mol. Cell Biol.
13
,
616
-
630
.
Materna
,
S. C.
,
Swartz
,
S. Z.
and
Smith
,
J.
(
2013
).
Notch and Nodal control forkhead factor expression in the specification of multipotent progenitors in sea urchin
.
Development
140
,
1796
-
1806
.
Mathieu
,
J.
,
Barth
,
A.
,
Rosa
,
F. M.
,
Wilson
,
S. W.
and
Peyriéras
,
N.
(
2002
).
Distinct and cooperative roles for Nodal and Hedgehog signals during hypothalamic development
.
Development
129
,
3055
-
3065
.
Maurus
,
D.
and
Harris
,
W. A.
(
2009
).
Zic-associated holoprosencephaly: zebrafish Zic1 controls midline formation and forebrain patterning by regulating nodal, Hedgehog and retinoic acid signaling
.
Genes Dev.
23
,
1461
-
1473
.
Mercier
,
S.
,
David
,
V.
,
Ratié
,
L.
,
Gicquel
,
I.
,
Odent
,
S.
and
Dupé
,
V.
(
2013
).
NODAL and SHH dose-dependent double inhibition promotes an HPE-like phenotype in chick embryos
.
Dis. Model. Mech.
6
,
537
-
543
.
Meyers-Needham
,
M.
,
Lewis
,
J. A.
,
Gencer
,
S.
,
Sentelle
,
R. D.
,
Saddoughi
,
S. A.
,
Clarke
,
C. J.
,
Hannun
,
Y. A.
,
Norell
,
H.
,
da Palma
,
T. M.
,
Nishimura
,
M.
, et al. 
(
2012
).
Off-target function of the sonic hedgehog inhibitor cyclopamine in mediating apoptosis via nitric oxide–dependent neutral sphingomyelinase 2/ceramide induction
.
Mol. Cancer Therap.
11
,
1092
-
1102
.
Miyamoto
,
N.
and
Wada
,
H.
(
2013
).
Hemichordate neurulation and the origin of the neural tube
.
Nat. Commun.
4
,
2713
.
Monuki
,
E. S.
(
2007
).
The morphogen signaling network in forebrain development and holoprosencephaly
.
J. Neuropathol. Exp. Neurol.
66
,
566
-
575
.
Moore-Scott
,
B. A.
and
Manley
,
N. R.
(
2005
).
Differential expression of Sonic hedgehog along the anterior-posterior axis regulates patterning of pharyngeal pouch endoderm and pharyngeal endoderm-derived organs
.
Dev. Biol.
278
,
323
-
335
.
Muller
,
F.
,
Albert
,
S.
,
Blader
,
P.
,
Fischer
,
N.
,
Hallonet
,
M.
and
Strahle
,
U.
(
2000
).
Direct action of the nodal-related signal cyclops in induction of sonic hedgehog in the ventral midline of the CNS
.
Development
127
,
3889
-
3897
.
Onai
,
T.
,
Yu
,
J.-K.
,
Blitz
,
I. L.
,
Cho
,
K. W. Y.
and
Holland
,
L. Z.
(
2010
).
Opposing Nodal/Vg1 and BMP signals mediate axial patterning in embryos of the basal chordate amphioxus
.
Dev. Biol.
344
,
377
-
389
.
Otto
,
A.
,
Pieper
,
T.
,
Viebahn
,
C.
and
Tsikolia
,
N.
(
2014
).
Early left-right asymmetries during axial morphogenesis in the chick embryo
.
Genesis
52
,
614
-
625
.
Porazzi
,
P.
,
Calebiro
,
D.
,
Benato
,
F.
,
Tiso
,
N.
and
Persani
,
L.
(
2009
).
Thyroid gland development and function in the zebrafish model
.
Mol. Cell. Endocrinol.
312
,
14
-
23
.
Roessler
,
E.
and
Muenke
,
M.
(
2001
).
Midline and laterality defects: left and right meet in the middle
.
BioEssays
23
,
888
-
900
.
Rohr
,
K. B.
,
Barth
,
K. A.
,
Varga
,
Z. M.
and
Wilson
,
S. W.
(
2001
).
The Nodal pathway acts upstream of Hedgehog signaling to specify ventral telencephalic identity
.
Neuron
29
,
341
-
351
.
Rottinger
,
E.
,
DuBuc
,
T. Q.
,
Amiel
,
A. R.
and
Martindale
,
M. Q.
(
2015
).
Nodal signaling is required for mesodermal and ventral but not for dorsal fates in the indirect developing hemichordate, Ptychodera flava
.
Biol. Open
4
,
830
-
842
.
Schubert
,
M.
,
Holland
,
L. Z.
and
Holland
,
N. D.
(
2000
).
Characterization of two amphioxus Wnt genes (AmphiWnt4 and AmphiWnt7b) with early expression in the developing central nervous system
.
Dev. Dyn.
217
,
205
-
215
.
Schubert
,
M.
,
Holland
,
L. Z.
,
Stokes
,
M. D.
and
Holland
,
N. D.
(
2001
).
Three amphioxus Wnt Genes (AmphiWnt3, AmphiWnt5, and AmphiWnt6) associated with the tail bud: the evolution of somitogenesis in chordates
.
Dev. Biol.
240
,
262
-
273
.
Shimeld
,
S. M.
(
1999
).
The evolution of the hedgehog gene family in chordates: Insights from amphioxus hedgehog
.
Dev. Genes Evol.
209
,
40
-
47
.
Shimeld
,
S. M.
(
2007
).
An amphioxus Gli gene reveals conservation of midline patterning and the evolution of hedgehog signalling diversity in chordates
.
PLoS ONE
2
,
e864
.
Soukup
,
V.
,
Yong
,
L. W.
,
Lu
,
T.-M.
,
Huang
,
S.-W.
,
Kozmik
,
Z.
and
Yu
,
J.-K.
(
2015
).
The Nodal signaling pathway controls left-right asymmetric development in amphioxus
.
Evodevo
6
,
1
-
23
.
Swartz
,
M. E.
,
Nguyen
,
V.
,
McCarthy
,
N. Q.
and
Eberhart
,
J. K.
(
2012
).
Hh signaling regulates patterning and morphogenesis of the pharyngeal arch-derived skeleton
.
Dev. Biol.
369
,
65
-
75
.
Szczepny
,
A.
,
Rogers
,
S.
,
Jayasekara
,
W. S. N.
,
Park
,
K.
,
McCloy
,
R. A.
,
Cochrane
,
C. R.
,
Ganju
,
V.
,
Cooper
,
W. A.
,
Sage
,
J.
,
Peacock
,
C. D.
, et al. 
(
2017
).
The role of canonical and non-canonical Hedgehog signaling in tumor progression in a mouse model of small cell lung cancer
.
Oncogene
36
,
5544
-
5550
.
Tsiairis
,
C. D.
and
McMahon
,
A. P.
(
2009
).
An Hh-dependent pathway in lateral plate mesoderm enables the generation of left/right asymmetry
.
Curr. Biol.
19
,
1912
-
1917
.
Tsikolia
,
N.
,
Schröder
,
S.
,
Schwartz
,
P.
and
Viebahn
,
C.
(
2012
).
Paraxial left-sided nodal expression and the start of left–right patterning in the early chick embryo
.
Differentiation
84
,
380
-
391
.
Wang
,
H.
,
Li
,
G.
and
Wang
,
Y.
(
2015
).
Generating amphioxus Hegehog knockout mutants and phenotypic analysis
.
Heriditas (Beijing)
37
,
1036
-
1042
.
Wang
,
Y.
,
Jin
,
G.
,
Li
,
Q.
,
Wang
,
Z.
,
Hu
,
W.
,
Li
,
P.
,
Li
,
S.
,
Wu
,
H.
,
Kong
,
X.
,
Gao
,
J.
 et al.  (
2016
).
Hedgehog signaling non-canonical activated by pro-inflammatory cytokines in pancreatic ductal adenocarcinoma
.
J.Cancer
7
,
2067
-
2076
.
Warner
,
J. F.
,
Miranda
,
E. L.
and
McClay
,
D. R.
(
2016
).
Contribution of hedgehog signaling to the establishment of left–right asymmetry in the sea urchin
.
Dev. Biol.
411
,
314
-
324
.
Xavier
,
G. M.
,
Seppala
,
M.
,
Barrell
,
W.
,
Birjandi
,
A. A.
,
Geoghegan
,
F.
and
Cobourne
,
M. T.
(
2016
).
Hedgehog receptor function during craniofacial development
.
Dev. Biol.
415
,
198
-
215
.
Yamagishi
,
H.
,
Maeda
,
J.
,
Hu
,
T.
,
McAnally
,
J.
,
Conway
,
S. J.
,
Kume
,
T.
,
Meyers
,
E. N.
,
Yamagishi
,
C.
and
Srivastava
,
D.
(
2003
).
Tbx1 is regulated by tissue-specific forkhead proteins through a common Sonic hedgehog-responsive enhancer
.
Genes Dev.
17
,
269
-
281
.
Yamagishi
,
C.
,
Yamagishi
,
H.
,
Maeda
,
J.
,
Tsuchihashi
,
T.
,
Ivey
,
K.
,
Hu
,
T.
and
Srivastava
,
D.
(
2006
).
Sonic hedgehog is essential for first pharyngeal arch development
.
Pediatr. Res.
59
,
349
-
354
.
Yu
,
J. K.
and
Holland
,
L. Z.
(
2009
).
Amphioxus whole-mount in situ hybridization
.
Cold Spring Harb. Protoc.
2009
,
prot5286
.
Yu
,
J.-K.
,
Holland
,
L. Z.
and
Holland
,
N. D.
(
2002
).
An amphioxus nodal gene (AmphiNodal) with early symmetrical expression in the organizer and mesoderm and later asymmetrical expression associated with left–right axis formation
.
Evol. Dev.
4
,
418
-
425
.
Yu
,
J.-K.
,
Satou
,
Y.
,
Holland
,
N. D.
,
Shin-I
,
T.
,
Kohara
,
Y.
,
Satoh
,
N.
,
Bronner-Fraser
,
M.
and
Holland
,
L. Z.
(
2007
).
Axial patterning in cephalochordates and the evolution of the organizer
.
Nature
445
,
613
-
617
.

Competing interests

The authors declare no competing or financial interests.

Supplementary information