XB-ART-14740Development July 1, 1998; 125 (14): 2687-700.
Frizzled-8 is expressed in the Spemann organizer and plays a role in early morphogenesis.
Wnts are secreted signaling molecules implicated in a large number of developmental processes. Frizzled proteins have been identified as likely receptors for Wnt ligands in vertebrates and invertebrates, but a functional role for vertebrate frizzleds has not yet been defined. To assess the endogenous role of frizzled proteins during vertebrate development, we have identified and characterized a Xenopus frizzled gene (xfz8). It is highly expressed in the deep cells of the Spemann organizer prior to dorsal lip formation and in the early involuting marginal zone. Ectopic expression of xfz8 in ventral cells leads to complete secondary axis formation and can synergize with Xwnt-8 while an inhibitory form of xfz8 (Nxfz8) blocks axis duplication by Xwnt-8, consistent with a role for xfz8 in Wnt signal transduction. Expression of Nxfz8 in dorsal cells has profound effects on morphogenesis during gastrulation and neurulation that result in dramatic shortening of the anterior-posterior axis. Our results suggest a role for xfz8 in morphogenesis during the gastrula stage of embryogenesis.
PubMed ID: 9636083
Article link: Development
Species referenced: Xenopus
Genes referenced: chrd.1 clstn2 frzb fzd8 gsc myod1 nodal3.1 nodal3.2 nog sia1 wnt5a wnt8a
Article Images: [+] show captions
|Fig. 1. Sequence and structure of Xenopus frizzled-8 (xfz8). Deduced amino acid sequences from xfz8, mouse frizzled-8 (mfz8) and Drosophila frizzled (Dfz1) were aligned using MacVector 6.0 ClustalW alignment. Identical amino acids are in black boxes and similar amino acids are shaded. Signal sequences are boxed, ten highly conserved cysteines of the putative extracellular ligand-binding domain are indicated by asterisks (*), transmembrane regions are underlined by black bars and putative PDZ-binding sites are indicated by an arrowhead. Xfz8 sequence has been deposited into the GenBank database under accession number AF033110. Illustrated in B is the proposed structure of frizzleds; the transmembrane domains, the putative Wnt-interacting cysteine-rich domain (CRD), and the structure of a putative inhibitory form of frizzled (Nfz).|
|Fig. 2. Expression of xfz8. Whole-mount in situ hybridization was performed on albino embryos from stage 9.5 (A), stage 10+ (B,F), stage 12 (C,G), stage 21 (D,H) and stage 30 (E). After visualization, embryos were sectioned (F-H). The dorsal blastopore lip is indicated by an arrow in B,C,F and G. Xfz8 expression at stage 10 (F) is in the anterior endomesodermal cells deep to Brachet’s cleft (b) (Keller, 1991), at stage 12 (G) in both anterior mesoderm and ectoderm and at stage 21 (H) in the stomatodeal-hypophyseal anlage (s) as well as surrounding tissues. At stage 30 (E), xfz8 is expressed in a variety of developing organ systems including the stomatodeal anlage (s), the heart anlage (h), the pronephric duct (p) and the branchial arches (b). Xfz8 expression in embryos is abolished by UV ventralization and radialized by lithium chloride treatment (I) similarly to chordin expression (J).|
|Fig. 3. Xfz8 induces ectopic axis formation and rescues UV ventralization. Ventral injection of 100 pg to 1000 ng of xfz8 RNA into a 4-cell blastomere can induce a single (A) ectopic axis complete with eyes and cement gland and frequently multiple ectopic axes (B, arrows in all panels represent cement glands of primary and ectopic axes). A dose-response curve for axis formation is shown in C. The number of ectopic axes per embryo is plotted with respect to the dose of xfz8 injected, with the solid black portion of the bar representing complete ectopic axes containing a well-formed cement gland and at least one eye and the open portion representing partial axes. The total number of embryos injected from three independent experiments is indicated at the top of each bar. The dashed line indicates an average frequency of one ectopic axis per embryo. UV ventralization (D) is rescued by injection of a low dose (100 pg) of xfz8 (E). This dose results in 81% of injected embryos (n=51) rescued to this extent. Note that at this and higher doses, cement gland is well formed, but there is a paucity of head structures and an open neural fold. Ventral injection of xfz8 induces organizer markers (F). Embryos were injected ventrally with 1 ng xfz8 RNA at the 8-cell stage. At stage 10.25, dorsal marginal zones (DMZs) and ventral marginal zones (VMZs) were dissected and assayed by RT-PCR for organizer marker gene expression. Siamois, Xnr3, goosecoid and noggin are upregulated in xfz8-injected VMZs.|
|Fig. 4. Xfz8 synergizes with Xwnt-8 to induce complete axes and Xwnt-5A to induce partial axes. HA-tagged Xwnt-8 (Xwnt-8HA, 2-5 pg) or Xwnt-5A (40 pg) were injected ventrally into 4-cell embryos either with or without xfz8 (50-100 pg). (A) Late neurula embryos oriented with the anterior toward the top. The anterior end of each ectopic axis is indicated by an arrow. (B) A histogram showing compilations of three independent experiments indicates percent of embryos with ectopic axes for the injections indicated. (C) Histogram illustrates that the activity of an axis-inducing dose of Xwnt-8HA (50 pg) is inhibited by coinjection with varying doses of Nxfz8 RNA (100 pg and 1 ng), but not b- galactosidase RNA (1 ng). (D) In a cell nonautonomous assay, Xwnt-8HA (50 pg) and Nxfz8 (1ng) or GFP RNA (1 ng) were injected into adjacent blastomeres of an 8-cell embryo. In all histograms, the bar represents percent of embryos with axis duplication, with the solid portion indicating complete axes as judged in Fig. 3. The total number of embryos injected is indicated above each bar.|
|Fig. 5. Nxfz8 enhances dorsoanterior structures and disrupts trunk structures. Injection of 1 ng of Nxfz8 or Nmfz8 RNA into a dorsal animal blastomere of 4-cell embryos (B,D,F,H,J,L,N) leads to a delay in closure of the blastopore (A versus B), expanded cement gland and failure to close the neural folds (C versus D), and tadpoles with enlarged heads, enlarged cement glands and shortened body axis (E versus F). Longitudinal (K,L) and cross-sectional (M,N) histological analysis of control (K,M) and Nxfz8-injected (L,N) embryos reveals the presence of expected tissues for this stage of development. These include notochord (n), somite (s), forebrain (f), cement gland (c) and heart (h) as indicated. In K-L, all illustrated at the same scale, it is evident that many structures, notochord being the most prevalent, are shortened in the anterior-posterior axis (K versus L) and expanded laterally (M versus N). This effect of Nfz8 injections is evident by neurula stage in embryos (G-J). Longitudinal (G,H) and cross sectional (I,J) histological analysis of control (G,I) and Nxfz8-injected (H,J) neurulae indicate A-P compression. Here, developing optic cup (o) is postitioned much closer to the blastopore (wide arrow) while the notochord (n) and the neural plate (edges marked by small arrows) is expanded laterally. Embryos in E-H,K,L) are oriented with the anterior to the left. Embryos in panels (A-D,I,J,M,N) are oriented with the dorsal to the top.|
|Fig. 6. The phenotype caused by Nxfz8 RNA injection is rescued by full-length xfz8 RNA. Embryos were injected with Nxfz8 (250 pg; B,D-F) and 125 pg (D), 250 pg (E) or 500 pg (F) of xfz8 RNA into a single dorsal animal B1 blastomere of the 32-cell embryo (Nakamura and Kishiyama, 1971). Uninjected stage 35 sibling embryos are shown for comparison (A). All embryos are oriented with anterior toward the left. A histogram summary (C) of these experiments indicates the percent of embryos that are scored as normal using the criteria outlined in the text. The head of a duplicated axis, complete with eye and cement gland is indicated with an arrow (F).|
|Fig. 7. In situ hybridization analysis with dorsal and lateral markers reveals that Nxfz8 inhibits dorsal convergence and embryonic elongation. Embryos, either uninjected (A,C,E,G) or injected with 1 ng of Nmfz8 RNA into an animal blastomere (B,D,F,H) were analyzed by in situ hybridization for chordin (A-D) or myoD (E-H). Stages shown are mid-gastrula (A,B,E,F) or late gastrula/early neurula (C,D,G,H). Embryos are oriented with dorsal (A,B,E,F) or anterior (C,D,G,H) toward the top of the picture. RT-PCR analysis indicates that Nxfz8s effect is not via a general delay in gastrulation (I). Embryos were injected with 1 ng of Nxfz8 RNA into a dorsal animal blastomere at the 8-cell stage. Whole embryos were collected beginning just after MBT (Stage 8.5) through stage 10.5, when the Nxfz8 phenotype is visibly apparent. Whole embryos were subjected to RT-PCR analysis for a number of markers to determine both temporal expression with respect to xfz8 and potential delays in gene expression caused by Nxfz8. Stages are indicated at the top, with a (+) indicating Nxfz8 RNA injection. Vertical lines have been superimposed to aid in comparison of uninjected and injected embryos for each time point. A negative control without reverse transcriptase is indicated in the right lane (-RT). No markers analyzed had appreciably different expression levels at any time point analyzed. Comparison of upstream Wnt inhibitors on myoD expression (J-Q). At the four cell stage, RNAs encoding the upstream wnt inhibitors Nxfz8 (K,O), Frzb-1 (L,P) and dnXwnt-8 (M,Q) were injected into a single dorsal animal blastomere of a pigmented embryo (J-M) or half the dose into the marginal zone of each blastomere (i.e. twice the total dose) of albino embryos (O-K). All injections resulted in a strong phenotype. Doses for dorsal injections were: Nxfz8, 500 pg; Frzb-1, 500 pg; dnXwnt-8, 2 ng.|
|Fig. 8. Nxfz8 inhibits elongation but not specification of activininduced animal caps. Embryos were injected at the 1-cell stage with 1 ng Nxfz8. At stage 8, animal explants were dissected and cultured in medium containing activin protein at a concentration previously titered to induce elongation of animal caps. Animal caps from uninjected embryos (A-D) and Nxfz8-injected embryos (E-H) are shown. Caps treated with activin (C,G) exhibit dramatic differences in the amount of elongation at the late neurulae stage (stage 17). Histological analysis of tadpole stage (stage 34) activin-treated caps (D,H) indicates notochord (n) in both uninjected (D) in and Nxfz8- injected caps (H). Untreated caps (B,F) exhibit only atypical epidermis. RT-PCR molecular analysis of these caps is shown in (I). Animal caps are shown in the left four lanes, and embryos in the right three. Activin treatment and injection of Nxfz8 RNA (1 ng) are indicated by a (+), as well as a negative control without reverse transcriptase (-RT).|
|Fig. 9. Activity of Nxfz8 expressed after MBT. Embryos were injected at the 8-cell stage into a single dorsal marginal blastomere. Xwnt-8/CS2 (30 pg) inhibited head formation in 100% of embryos (B; n=38) and Nxfz8/CS2 (200 pg) caused an expansion of cement gland and shortened axes in 94% of embryos (C; n=30). Nxfz8/CS2 was able to inhibit Xwnt-8/CS2 and rescue head structures in 93% of embryos (D; n=28).|