XB-ART-41269Dev Biol May 15, 2010; 341 (2): 375-88.
Neural crest migration requires the activity of the extracellular sulphatases XtSulf1 and XtSulf2.
In vertebrates, there are two related genes, Sulf1 and Sulf2 that code for extracellular heparan sulphate 6-0-endosulphatases. These enzymes act to post-synthetically remodel heparan sulphate chains, generating structural diversity of cell surface HSPGs; this activity provides an important mechanism to modulate developmental cell signalling. Here we describe the expression and activity of Xenopus tropicalis Sulf2 (XtSulf2), which like XtSulf1, can act extracellularly to inhibit BMP4 and FGF4 signalling. Consistent with its discrete expression in regions of the anterior developing nervous system, we found that overexpression of XtSulf2 disrupts the expression of a set of neural markers and inhibits the migration of the neural crest. Using a combination of grafting experiments and antisense morpholino based knockdown studies in Xenopus embryos, we demonstrate that endogenous XtSulf1 and XtSulf2 play an important role during cranial neural crest cell migration in vivo.
PubMed ID: 20206618
Article link: Dev Biol
Genes referenced: actc1 actl6a bmp4 bmpr1a egr2 en2 fgf3 fgf4 fgf8 fgfr1 hoxb3 isl1 lhx1 myc myod1 ncam1 neurod1 nucb1 otx2 pax6 snai2 sox3 sox8 sulf1 sulf2 twist1
Morpholinos: sulf1 MO1 sulf1 MO2 sulf1 MO3 sulf2 MO1
Article Images: [+] show captions
|Fig. 1. Zygotic expression of XtSulf2. Zygotic transcripts for XtSulf2 are detected by whole mount in situ hybridisation in X. tropicalis embryos. (A–B) At stage 18, zygotic transcripts for XtSulf2 are detected in the anterior neural tube most promimently at the mid-hindbrain junction (asterisk in B). (C–D) At stage 20, XtSulf2 is expressed in the ventral part of anterior neural tube, in the ventral midbrain and hindbrain and at the midbrain–hindbrain junction (asterisk), and in the mandibular mesendoderm underlying the presumptive cement gland (arrow in D and E). (E) At stage 25, XtSulf2 is expressed in the pineal gland (arrowhead), the mandibular mesendoderm subjacent to the cement gland (arrow), the MHJ (asterisk) and the pineal gland (arrowhead). Strong expression continues in the ventral midbrain and hindbrain. (F–J) At stage 28, XtSulf2 is expressed in the mandibular arch, pineal gland and neural retina. Expression is detected in the ventricular zone (VZ) of the midbrain and ventrally in the neural tube. (K) At stage 35, XtSulf2 is expressed in the MHJ (asterisk), pineal gland (arrowhead) olfactory placodes (arrows), and neural retina. (G–J, L–O) 50 µm vibratome transverse sections through respectively (F) and (K). (G, I, L, N) 10× magnification, (H, J, M, O) 40× magnification. R is retina, VZ is ventricular zone, NT is neural tube, and Nc is notochord.|
|Fig. 2. Localisation and activity of XtSulf2. RNAs coding for tagged proteins were injected into the animal hemisphere of both blastomeres of a two-cell X. laevis embryo and processed for immunohistochemistry at NF stage 10. (A) Embryos injected with 1 ng of Nuc-GFP. (B and C) Embryos injected with 2 ng of XtSulf2–GFP. (D–F) Embryos co-injected with RNA coding for XtSulf1-Myc (D) and XtSulf2–GFP. (E) The yellow regions in (F) show where XtSulf1 and XtSulf2 expression overlap. Embryos were co-injected with mRNA coding for FGFR1-Myc (G) and XtSulf2–GFP (H) proteins. While both proteins appear associated with the membrane, they do not completely overlap (I). Western analysis shows in (J) that XtSulf2 while injection of mRNA coding for XtSulf2 inhibits activation of dpERK by FGF4 protein, it has no effect on dpERK activation by the intracellularly drug induced iFGFR1. (K) BMP4 activity was assayed by Western blot analysis using antibodies against phospho-SMAD 1/5/8. The phosphorylation of SMAD 1/5/8 stimulated in animal caps by injection of mRNA coding for BMP4 is down-regulated in animal caps co-expressing XtSulf2. However, animal caps expressing a constitutively active BMP4 receptor, Alk3 continue to express phospho-SMAD 1/5/8 in the presence of XtSulf2, indicating that XtSulf2 acts upstream of the BMP4 receptor. (L) Neuralisation of animal caps was assayed using RT-PCR to detect NCAM expression. Whole embryos (WE) express NCAM at NF stage 18, while control animal caps do not. Injection of mRNA coding for XtSulf2 results in the activation of NCAM expression. Samples with water and without AMV Reverse Transcriptase (RT) were used as negative controls; L8 was used as a loading control.|
|Fig. 3. Overexpression of XtSulf2 affects the expression of a number of genes expressed in anterior neural tissue and the cranial neural crest. One blastomere of a two-cell Xenopus tropicalis embryo was injected with 0.5 ng of XtSulf2 mRNA, embryos were allowed to develop to NF stage 28 and fixed for in situ hybridization using En2 (A–B), Isl1 (C–D), Krox20 (E–G), Lim1 (H–I), NeuroD (J–L), Otx2 (M–N), Pax6 (O–Q), Slug (R–S), Sox3 (T–U), and Sox8 (V–W). En2 expression is shifted posteriorly on the injected side (A–B). Isl1 expression in the profundal and trigeminal placodes shifts posteriorly (black arrowheads) as does the expression in the heart (C–D). Krox20 expression in the migrating neural crest shifts posteriorly (E–F white arrowheads) as does the expression in r5 (G black lines). Lim1 expression in the pronephros is restricted anteriorly (H–I black lines). NeuroD expression in the profundal and trigeminal placodes is disrupted (J–K red arrowheads), and expression in the olfactory placodes fuses with expression in the eye placodes (L yellow arrowheads). Otx2 expression is restricted in the midbrain (M–N black lines). Pax6 expression is ectopically expressed in the posterior and dorsal head ectoderm (O–P–Q red asterisk). Slug expression at the leading edge of neural crest migration remains closer to the neural tube than in controls (R–S black lines). Sox3 expression is decreased in neural crest but expands laterally (T–U black lines). Sox8 expression in the anterior and posterior branchial arches is decreased and the leading edge of the migrating neural crest remains closer to the neural tube than in controls (V–W green arrowhead). (A–B, G, Q) are dorsal views of embryos, anterior at the top. (C–F, H–K, M–P, R–W) are lateral views of embryos, anterior to the left. (L) is an anterior view of embryo in (J–K), dorsal at the top. Asterisks indicate injected side. CS control side, IS injected side.|
|Fig. 4. XtSulf2 overexpression disrupts cranial neural crest migration. One blastomere of a two-cell X. tropicalis embryo was injected with 0.5 ng of XtSulf2 mRNA, embryos were allowed to develop to NF stage 22 and 30 and fixed for in situ hybridisation using HoxB3 (A–E), NeuroD (F–I), Slug (J–M), Sox8 (N–Q) and Twist (R–U). (A–B, F–G, J–K, N–O, R–S) Lateral views of control and injected sides of NF stage 22 embryos. (E) Dorsal view of embryo in (A–B), black asterisk indicated the injected side. (C–D, H–I, L–M, P–Q, T–U) Lateral views of control and injected sides of NF stage 30 embryos. At stage 22, on the control side 6–12% of embryos show disrupted migration, while on the injected side 58–71% of migration is found to be inhibited (n = 55–84) as assayed by the expression of CNC markers shown. At stage 30, on the control side 4–5% of embryos show disrupted migration, while on the injected side 40–53% of migration is found to be inhibited (n = 34–81).|
|Fig. 5. XtSulf1/2 double knockdown embryos show impaired early migration of the cranial neural crest. Xenopus tropicalis embryos were injected at two and four-cell stage into the animal hemisphere. Sibling embryos were cultured until NF stage 15 (top panels), stage 22 (middle panels) and NF stage 30 (bottom panels) and fixed for in situ hybridisation. This figure shows control uninjected embryos (E, J, O, T, Y, D′); embryos injected with 30 ng of control morpholino oligo (CMO) (A, C, F, K, P, U, Z, E′); embryos injected with 10 ng of antisense morpholino oligo (AMO) targeted against XtSulf1 (G, L, Q, V, A′, F′); embryos injected with 20 ng of AMO targeted against XtSulf2 (H, M, R, W, B′, G′); and embryos co-injected with 10 ng of AMO targeted against XtSulf1and 20 ng of AMO targeted against XtSulf1(B, D, I, N, S, X, C′, H′). Top panels (A–D) are dorsal views, the others are lateral views with anterior to the left. At stage 15, the expression of Slug marks the prospective neural crest at the boundary of the neural plate and the non-neural ectoderm (A, B). MyoD marks the dorsal, paraxial mesoderm (C, D). The expression of these genes is not effected when embryos are injected into the animal hemisphere with AMOs targeting XtSulf1 and XtSulf2 (B, D). At stage 22, NeuroD is expressed in the profundal, trigeminal and antero-dorsal lateral line placodes and lens (E,F). In single XtSulf1 knockdown, NeuroD expression in these placodes is reduced and disorganised (G), while in double Sulf knockdown, NeuroD expression is further reduced and restricted to a very small dorsal region (L). Sox8 and Twist are expressed in the mandibular (around the eye), hyoid and branchial arches (J, K, O, P; three arrow heads) at NF stage 22. In single XtSulf1 and XtSulf2 knockdown embryos the expression of these markers in the mandibular arch appears normal, but the expression in the hyoid and branchial arches does not extend as far ventrally as controls (L, M, R, Q; arrowhead and line). The anterior and posterior branchial arches are no longer visible as two distinct entities. In double Sulf1/2 knockdowns, Sox8 and Twist expression in the mandibular arch is normal, but expression in the hyoid and branchial arches is further restricted dorsally and the anterior and posterior branchial arches appear as one fused domain (N and S, arrowhead and line). At NF stage 30, NeuroD is expressed normally in the neural placodes and pineal gland (T,U). In XtSulf1 single knockout, expression in the antero-dorsal lateral line and trigeminal placodes migrate abnormally, all other expression is absent (V). In XtSulf2 knockouts, NeuroD expression is absent from the posterior lateral line placode, and expression in the trigeminal and antero-dorsal lateral line placode appear closer together (W). In double Sulf knockouts, NeuroD is expressed abnormally in the antero-dorsal lateral line and trigeminal placodes (X). Sox8 and Twist are expressed in the mandibular, hyoid and branchial arches (Y, Z, D′, E′). In single knockouts, the expression in the mandibular arch appears normal, but the expression in the hyoid and branchial arches is restricted dorsally. Furthermore, the anterior and posterior branchial arches appear fused together (A′, B′, F′, G′). In double Sulf knockouts, Sox8 and Twist expression domains are restricted dorsally in the hyoid and branchial arches, and the expression is fused in the anterior and posterior branchial arches (C′, H′). The effects of XtSulf 1/2 knockdown on CNC migration as assayed by the expression of the markers shown in this figure are summarised in Table 1.|
|Fig. 6. Targeting XtSulf1/2 knockdown to neural tissue disrupts cranial neural crest migration and does not affect mesoderm differentiation. X. tropicalis embryos were injected into the two dorsal blastomeres in the animal hemisphere at the 8-cell stage with 30 ng of control morpholino (CMO) or with 10 ng of S1 AMO together with 20 ng of S2 AMO (S1/S2 AMO). At stage 16, CMO injected (A, B) and S1/S2 AMO injected (C, D) embryos show normal expression of slug (CMO = 78% normal, n = 23; S1/S2 AMO = 59% normal, n = 17) and MyoD (CMO = 90% normal, n = 20, S1/S2 AMO = 65% normal; n = 23). At stage 22, twist expression is disrupted in knockdown embryos (F, 66% disrupted twist expression, n = 15) and does not extend as far ventrally as seen in those injected with control morpholino (E). The expression of myoD and α-cardiac actin in CMO injected embryos at stage 22 are shown in G and I. There is normal expression of myoD (66%, n = 18) and α-cardiac actin (64%, n = 14) in S1/S2 knockdown embryos (H and J). (K) The expression of twist in embryos injected with control morpholino at stage 30. (L) Twist expression is disrupted in knockdown embryos (69%, n = 16). The expression of myoD and α-cardiac actin in CMO injected embryos at stage 30 are shown in M and O. There is normal somite expression of myoD (83%, n = 12) and α-cardiac actin (80%, n = 10) in stage 30 S1/S2 knockdown embryos (N and P).|
|Fig. 7. Delay in cranial neural crest migration in XtSulf1/2 double knockdown embryos is rescued by XtSulf2 mRNA injections. The CNC migrates into the mandibular, hyoid and anterior and posterior branchial arches in control and CMO embryos (A–D). In Sulf 1/2 double knockdown embryos, the CNC migration is blocked (E–F), migration is rescued with XtSulf2 mRNA (G–H). Xenopus tropicalis embryos were injected at the two- and four-cell stage into the animal hemisphere with either 30 ng of CMO (C, D), 10 ng of S1 AMO and 20 ng of S2 AMO (E, F) or 10 ng of S1 AMO, 20 ng of S2 AMO and 1.5 ng of XtSulf-2 mRNA (G, H). Neural crest migration was assayed by analysing Twist expression by in situ hybridisation. All panels show lateral views with anterior to the left. A summary of the results shown in this figure are reported in Table 2.|
|Fig. 8. Grafting experiments reveal a requirement for XtSulf1/2 in the environment through which cranial neural crest cells migrate. On the left, cartoons depict the design of the neural crest grafting experiments, while the panels to the right show the resulting embryos at tailbud stages. The top set of experiments show X. laevis embryos injected at the two-cell stage with 0.5 ng of GFP mRNA, ± 3 ng of mRNA coding for XtSulf2 and from which are dissected CNC explants at NF stage 14. Explants are transplanted into either a control embryo or an embryo injected with 3 ng of XtSulf2 mRNA. (A–C) Control host embryos transplanted at NF stage 14 with GFP expressing neural crest explants. (D–F) Host embryos injected with 3 ng of XtSulf2 at the two-cell stage into the animal pole and grafted with GFP expressing CNC explants at NF stage 14. (G–I) Control X. laevis host embryos grafted at NF stage 14 with GFP + XtSulf2 expressing CNC explants. All pictures were taken at 10X magnification; the eye, cement gland, and labelled cells are outlined. The bottom set of experiment show X. tropicalis embryos injected with mRNA coding for GFP ± AMOs targeted against XtSulf1 and XtSulf2. (J–L) At NF stage 15, CNC explants were taken from control Xenopus tropicalis embryos expressing GFP and transplanted into uninjected control embryos at the same stage. The GFP labelled cells migrate into the branchial arches by NF stage 30. (M–O) GFP mRNA together with AMOs targeting both XtSulf1 and XtSulf2 were injected into the animal hemisphere at the 2-cell stage and at NF stage 15 CNC explants were transplanted into control host embryos at the same stage. Labelled S1/S2 knockdown cells migrate less extensively than the control cells, but have direction. (P–R) At NF stage 15, CNC explants were taken from control Xenopus tropicalis embryos expressing GFP and transplanted into host embryos in which XtSulf1 and XtSulf2 have been knocked down. At stage 30, labelled cells are seen to be widely dispersed, suggesting that these wildtype cells lack direction in the S1/S2 mutant background. The eye, cement gland, and labelled cells are outlined.|
|Supplementary Figure 3. Neural crest induction is not inhibited by overexpressing XtSulf2 Control embryos (top panel) show the normal expression of Slug at neural plate stage 15. Embryos overexpressing XtSulf2 show reduced Slug expression in some embryos (13/24) and in some of these there is a loss of a sharp lateral boundary (arrows) of Slug expression when XtSulf2 is overexpressed. However there is no loss of MyoD expression (not shown) in these embryos as injections were done into the animal hemisphere.|
|Supplementary Figure 4: XtSulf-1 and XtSulf-2 are expressed in the migrating CNC and are flanked by FGF3 and FGF8 expression. Zygotic transcripts for XtSulf-1, XtSulf-2 , FGF3 and FGF8 are detected by whole mount in-situ hybridisation using DIG-AP labelled RNA in X. tropicalis embryos at NF stage 25 30 35 and 40, lateral views of embryos, anterior to the left. XtSulf-1 is expressed in the mandibular arch and somites throughout all tailbud stages, in the hyoid and branchial arch and pronephros from NF stage 25 till 40 in the primary heart field from NF stage 30 till stage 40. XtSulf-2 is expressed in the anterior mesendoderm, just ventral to the mandibular arch at NF stage 20 to 30, expression is then restricted to the gill at NF stage 40. Frontal vibratome sections show XtSulf-1 expression in the pharyngeal mesoderm of the anterior and posterior branchial, hyoid and mandibular arches, XtSulf-2 expression in the pharyngeal endoderm of the mandibular arch and anterior ectodermal cleft of the hyoid arch, FGF3 and FGF8 expression in the anterior endodermal pharyngeal pouch of the posterior and anterior branchial, hyoid and mandibular arches. Xenbase Curator notes: 1) bottom 2 panels, left side are labelled sox8 as are curated as such. 2) NF staging mis-identified in figure legend, and are curated as stage 22/23, 26, 28/29 and 40 respectively (left to right) - not 2,5 30, 35, and 40 as stated above)|
|Supplementary Figure 6. The structure of somites is normal in Sulf1/2 knock-down embryos. Embryos were injected at the 8-cell stage with 30ng of control morpholino oligo (CMO, top panel), or with 20ng of antisense morpholino oligs directed against XtSulf2 together with 10ng of antisense morpholino oligs directed against XtSulf1 (S1/S2 AMO, bottom panel). At stage 30, the expression of a-cardiac actin was analysed by in situ hybridisation and subsequently sagital sections were taken with a vibratome. Anterior is to the left. Normal somite structure is evident.|
|Supplementary Figure 7: CNC explants. In-situ hybridisation using the CNC marker Slug. (A) Dorsal view of a control NF stage 18 X. laevis embryo in-situ hybridised with Slug (B) Dorsal view of an NF stage 18 X. laevis embryo from which CNC explants were dissected in-situ hybridised with Slug. The black arrowheads indicate the regions were the CNC has been removed. (C) CNC explants were fixed immediately after dissection and processed for in-situ hybridisation with Slug. (D) Dorsal view of a control NF stage 18 X. laevis embryo injected with 3ng of XtSulf-2 mRNA and in-situ hybridised with Slug.|
|fgf3 (fibroblast growth factor 3) gene expression in Xenopus tropicalis embryo, assayed via in situ hybridization, NF stage 22, lateral view, anterior left, dorsal up.|
|fgf3 (fibroblast growth factor 3) gene expression in Xenopus tropicalis embryo, assayed via in situ hybridization, NF stage 26, lateral view, anterior left, dorsal up.|
|fgf3 (fibroblast growth factor 3) gene expression in Xenopus tropicalis embryo, assayed via in situ hybridization, NF stage 28, lateral view, anterior left, dorsal up.|