XB-ART-39736Dev Biol. July 15, 2009; 331 (2): 281-91.
Flow on the right side of the gastrocoel roof plate is dispensable for symmetry breakage in the frog Xenopus laevis.
Leftward flow of extracellular fluid breaks the bilateral symmetry of most vertebrate embryos, manifested by the ensuing asymmetric induction of Nodal signaling in the left lateral plate mesoderm (LPM). Flow is generated by rotational beating of polarized monocilia at the posterior notochord (PNC; mammals), Kupffer''s vesicle (KV; teleost fish) and the gastrocoel roof plate (GRP; amphibians). To manipulate flow in a defined way we cloned dynein heavy chain genes dnah5, 9 and 11 in Xenopus. dnah9 expression was closely related to motile cilia from neurulation onwards. Morphant tadpoles showed impaired epidermal ciliary beating. Leftward flow at the GRP was absent, resulting in embryos with loss of asymmetric marker gene expression. Remarkably, unilateral knockdown on the right side of the GRP did not affect laterality, while left-sided ablation of flow abolished marker gene expression. Thus, flow was required exclusively on the left side of the GRP to break symmetry in the frog. Our data suggest that the substrate of flow is generated within the GRP and not at its margin, disqualifying Nodal as a candidate morphogen.
PubMed ID: 19450574
Article link: Dev Biol.
Genes referenced: dnah11 dnah5 dnah9 fes nodal pitx2 tuba4b
Antibodies referenced: Tuba4b Ab4
Morpholinos referenced: dnah5 MO1 dnah9 MO2 dnah9 MO3
Article Images: [+] show captions
|Fig. 1. Maternal and early zygotic expression of dnah9. Whole-mount in situ hybridization of staged embryos with a dnah9-specific antisense probe. (A, B) Maternal mRNA localizes to the animal pole of the zygote (hemi section in A), and 4-cell embryo in top (B) and side view (B’). (C–F) Zygotic expression in the dorsal lip at stage 11 (C) and at stage 13 (arrow in D). (E) Early neurula (stage 16) in frontal view (E) and transverse section (E’). Note expression in the floor plate (arrowheads; section at a level anterior to the gastrocoel roof plate). (F) Lateral view of a stage 21 neurula with floor plate (F’) and epidermal staining (anterior to the left). (G) Localization of dnah9 mRNA in multi-ciliated cells of the epidermis at stage 32, demonstrated by sequential in situ hybridization (G’) and immunohistochemistry using an antibody against acetylated tubulin (G”). no, notochord; s, somite. Scale bar represents 15 μm.|
|Fig. 2. Comparison of expression patterns of dnah5, dnah9 and dnah11 in ciliated embryonic tissues. (A–C) Immunohistochemistry using an antibody against acetylated alpha tubulin to highlight cilia. (A) Ciliated cells at the GRP, nephrostomes (B’), epidermis (B”) and tailbud (B’”) of the 2-day tadpoles, and (C) in stomach and small intestine. (D–L) Whole-mount in situ hybridization of staged embryos with dnah5 (D–F), dnah9 (G–I) and dnah11 (J–L) specific antisense probes. GRPs (A, D, G, J), outlined in dorsal explants by dotted lines, are shown in ventral view, anterior to the top. (B, E, H, K) 2-day tadpoles are oriented anterior to the left and dorsal up; blow-ups are indicated by dashed lines. Insets show frontal view of proctodeum (arrowhead). (C, F, I, L) Ventral views of stage 43/44 tadpoles. Proctodeum marked by arrowheads. Note that dnah5 and dnah9 are co-expressed in ciliated cells, with dnah9 showing markedly stronger signals, whereas dnah11 is only expressed in the proctodeum, tailbud and gastro-intestinal tract. Please note also, that dnah5 and dnah9 are restricted to the neural tube of the tailbud, whereas dnah11 is additionally expressed in the posterior wall. cnh, chordoneural hinge; fp, floor plate; nc, neuroenteric canal; no, notochord; s, stomach; si, small intestine.|
|Fig. 3. Knockdown of dnah9 in the epidermis inhibits ciliary motility. (A) Ventral injection of lineage tracer mRFP into the marginal regions of vegetal and animal right blastomeres resulted in specific labeling of right but not left epidermis at stage 32. (B) Tadpole motion assay. Stage 32 tadpoles were placed individually into six Petri dishes and monitored by standard (CCD) digital cameras at 25 fps. Movements were recorded via connected computers (PC) and calculated using a custom-made software. (C) Unilateral dnah9-MO-injection resulted in a very highly significant decrease of tadpole velocity on the injected side. (D–G) Ciliation of epidermal cells was unaffected in dnah9-MO-injected tadpoles. (D, E) Immunohistochemistry of stage 32 tadpoles using an antibody against acetylated tubulin (green) and red fluorescence from co-injected lineage tracer mRFP (red). Overlays reveal equal ciliation in specimens injected with control MO (D) and in dnah9 morphants (E). (F, G) Scanning electron micrographs of control MO (F) and dnah9-SB-MO-injected tadpoles. Targeted regions of skin are highlighted in (F) and (G) by overlay with injected lineage tracer (GFP). Higher power magnifications of targeted regions in (F’, F”, G’, G”) demonstrate equal ciliation in control-injected and morphant embryos. an, animal; d, dorsal; l, left; r, right; v, ventral; veg, vegetal. Scale bars represent 10 μm.|
|Fig. 4. No induction of Pitx2c in dnah9 and dnah5 morphants. Injection into the marginal region of left and right dorsal blastomeres (A) specifically targets the floor plate (B; external view) and GRP (B’; ventral view of dorsal explant). (C–E) Laterality defects in morphants. Wildtype left-asymmetric expression of Pitx2c in Co-MO-injected specimen (C) and absence of lateral plate expression in dnah9-MO-injected tadpole (D). (E) Quantification of results. Note that bilateral and right-asymmetric expression patterns were very rarely encountered, while absence of Nodal cascade induction was found in up to 75% of cases. Please note that all three MOs resulted in similar effects as well as dose-dependency of dnah9-SB-MO. an, animal; d, dorsal; l, left; r, right; v, ventral; veg, vegetal.|
|Fig. 5. No leftward flow at the GRP of dnah9-SB and dnah5-SB morphant embryos. Flow was analyzed by addition of fluorescent microbeads to dorsal explants and video microscopy. Representative examples of stage 17/18 dorsal explants of Co-MO (A), dnah9-SB-MO (B) and dnah5-SB-MO (C) injected embryos. Targeted areas indicated by red lines represent the limits of lineage tracer. Particle movements displayed as gradient-time trails (GTTs), representing 25 s from green to red (cf. bar in C). (A’–C’) Quantitative analysis of GTT directionality over the area of the GRP demonstrating strong leftward flow in Co-MO and absence of directed bead transport in dnah9-SB- and dnah5-SB-MO-injected specimen. (D, E) Collapsed movies of in vivo imaged cilia movements by fluorescence microscopy using a PACRGeGFP fusion construct. Co-MO injection resulted in wildtype rotational pattern (D), whereas dnah9-SB-MO-injected GRPs displayed variant phenotypes, namely irregular circular movements (E), wiggling (E’) or arrested motion (E”; cf. Suppl. Movie 5). Scale bar represents 50 μm. a, anterior; l, left; p, posterior; r, right.|
|Fig. 6. Flow is dispensable on the right side of the GRP. (A, B) Injection scheme to specifically target left and right half of floor plate (B, external view) and GRP (B’, ventral view of dorsal explant). (C–E) Flow analysis of Co-MO left (C), and dnah9-SB-MO right (D) and left (E) injected dorsal explants (ventral views, anterior up). Targeted areas indicated by red lines represent the limits of lineage tracer. Particle movements displayed as gradient-time trails (GTTs), representing 25 s from green to red (cf. bar in panel E). Note that flow was absent on dnah9-SB-targeted sides. (C’–E”) Quantitative analysis of GTT directionality over the respective left and right sides of the GRP. (F) Pitx2c expression analysis. Wildtype left-asymmetric expression in Co-MO and right dnah9-SB-MO-injected tadpoles, and absence of signals in left dnah9-SB-MO, dnah9-AUG-MO and dnah5-SB-MO-injected embryos. a, anterior; an, animal; d, dorsal; l, left; p, posterior; r, right; v, ventral; veg, vegetal. Scale bar in panel C represents 50 μm and applies to panels C–E.|
|Dnah11 (dynein, axonemal, heavy chain 11) gene expression in Xenopus laevis embryo via in situ hybridization, NF stage 43, ventral view, anterior up.|
|dnah9 (dynein, axonemal, heavy chain 9 ) gene expression in Xenopus laevis embryo via in situ hybridization, NF stage 43, ventral view, anterior up.|
|Suppl. Fig. 6. Lineage-specific knock-down of dnah9. Dorsal–-marginal injections (A) predominantly targeted GRP and floor plate (A’–-A’”’’’) and resulted in > 75% of altered Pitx2c gene expression (D). Dorsoal–-lateral marginal injections (B) targeted lateral tissue (neural plate and somites) in addition to GRP and floor plate, which were hit less frequently (B’–-B’”’’’). Efficiency of laterality defects consequently dropped to about 45% (D). Ventral marginal injections (C) resulted in lateral plate and skin targeting (C’–-C’”’’’) and did not affect Pitx2c gene expression on either side (D). Please note that although identical blastomeres were injected in (A) and (B) effects varied very highly significantly (p < 0.001), suggesting limited diffusion.|
|Suppl. Fig. 3. Specificity of early maternal expression patterns of dnah9. (A, B) Whole-mount in situ hybridization of zygotes using antisense (A) and sense (B) probes, shown in whole-mounts (A, B), hemi-sections (A’, B’) and 30 μm vibratome sections (A”’’, B”’’). (C) Semi-quantitative RT-PCR from 2-cell to tadpole stages, and of isolated GRP and SM tissue (top). Elongation factor 1 alpha (bottom) served as control. GRP, gastrocoel roof plate; SM, superficial mesoderm. Note the strong maternal expression of dnah9.|
|Suppl. Fig. 4. (A) Inhibition of intron2 splicing by dnah9-SB-MO. RT-PCR analysis of MO-injected embryos. Splicing was analyzed using forward primer 9e2 (located in exon 2) or forward primer 9i2 (located in intron 2), and reverse primer 9e3 (located in exon 3), schematically indicated by green (exons) and orange (intron) bars, and by blue arrows (primers). PCR of genomic DNA resulted in 1.5 kb (9e2–-9e3) and 1.4 kb (9i2–-9e3) bands. RT-PCR of Co-MO inMO-injected specimens yielded no band with 9i2–-9e3, and a strong band of 225 bp with 9e2–-9e3, corresponding to the joined exons. In dnah9-MO-injected embryos, bands representing unspliced RNAs were observed with both 9e2–-9e3 and 9i2–-9e3, while the spliced band observed with 9e2–-9e3 was much reduced compared to Co-MO-injected samples (cf. EF1 alpha loading control). No products were amplified without prior RT reaction (− -RT). (B) Inhibition of intron2 splicing by dnah5-SB-MO. RT-PCR analysis of MO-injected embryos. Splicing was analyzed using forward primer 5e2 (located in exon 2) and reverse primer 5i2 (located in intron 2), schematically indicated by green (exons) and orange (intron) bars, and by blue arrows (primers). PCR of genomic DNA resulted in 600 bp (5e2–-5i2) band. RT-PCR of Co-MO inMO-injected specimens yielded no band with 5e2–-5i2. In dnah5-SB-MO-injected embryos, a band representing unspliced RNA was observed with 5i2–-5e3 (cf. EF1 alpha loading control). No products were amplified without prior RT reaction (− -RT).|
|Suppl. Fig. 5. No induction of Xnr1 in dnah9 morphants. Experimental set-up was as described in Fig. 4. Xnr1 was not induced in the LPM in about 60% % (dnah9-SB-MO) and 40% % (dnah-AUG-MO) of morphants.|