XB-ART-22588Cell May 21, 1993; 73 (4): 659-71.
Expression of an extracellular deletion of Xotch diverts cell fate in Xenopus embryos.
Xotch is a Xenopus homolog of Notch, a receptor involved in cell fate decisions in Drosophila. Using an extracellular deletion construct, Xotch delta E, we show that Xotch has a similar role in Xenopus embryos. Broad expression causes the loss of dorsal structures and the expansion and disorganization of the brain. Single blastomere injections of Xotch delta E induce autonomous neural and mesodermal hypertrophy, even in the absence of cell division. Xotch delta E inhibits the early expression of epidermal and neural crest markers yet enhances and extends the response of animal caps to mesodermal and neural induction. Our data suggest a mechanism for the function of Notch homologs in which they delay differentiation and leave undetermined cells competent to respond to later inductive signals.
PubMed ID: 8500162
Article link: Cell
Genes referenced: cdh2 elavl1 inhba krt12.4 ncam1 notch1 sia2 tbx2 twist1
Article Images: [+] show captions
|Figure 1. Schematic Diagrams of Xotch, XotchAE, and T6 The structure of the intact Xotch molecule is represented at the top (Coffman et al., 1990). To construct XotchAE, the carboxyl half of Xotch upstream of the transmembrane domain was fused to sequences encoding the signal peptide from N-cadherin (Kintner, 1992). The resulting chimeric molecule retains the transmembrane (TM) domain from Xotch but lacks two conserved cysteine residues in the extracellular domain of Xotch that have been implicated in dimerization (Greenwald and Seydoux, 1990; Kidd et al., 1969). The structure of XotchAE is similar to that of TAN-I, an altered form of human Notch (TAN-1 arrow; see Ellisen et al., 1991). T6, an extracellular and intracellular deleted form of N-cadherin, was used as a negative control (Kintner,‘1992).|
|Figure 2. Expression of XotchdE in Embryos Leads to Dorsoanterior Defects Embryos were injected into all blastomeres at the 4-cell stage with T6 control or XotchdE RNA and allowed to develop to stage 26 (Nieuwkoop and Faber, 1967). (A) An example of an embryo injected with a control RNA. Note the well-formed eye vesicles (E) and cement gland (CG), a heavily pigmented structure found at the anterior end of the animal. (B) An embryo injected with XotchdE RNA. Note that the cement gland is missing (open arrow). (C and D) Sections through animals from this same experiment taken at the level indicated by the arrowheads in (A) and (B); (C) shows a section through a control embryo, and (D) shows a section through a XotchdE embryo. Note that the ventricle of the diencephalon (V) remains, but the eye vesicle is completely missing in the XotchdE-injected embryo in (D). Instead, the walls of the brain are thickened (arrowheads in [D]), and the dorsal region is broader and flatter than in controls. In a serial reconstruction of more posterior sections, neural tissue was found to be expanded in XotchdE-injected animals by an average of 1.6-fold (range 1.3- to 1 .Q-fold). Bar, 100 pm (applies to both [C] and [D]).|
|Figure 3. Extra Neural and Muscle Tissue in Embryos Injected with XotchdE RNA (A-D) Embryos were injected once at the 2-to &cellstage witha mixtureof XotchdEand B-gal RNA. At stage 23, the embryos were fixed, sectioned, and then stained with various antibodies by immunoftuorescence (see Experimental Procedures). A section from an injected embryo stained with an anti-g-gal antibody is shown in (A), while (B) shows a neighboring section from the same animal stained with a neural-specific antibody, XAN-1. Note that the neural tube hypertrophy seen on the right side of the embryo in (B) greatly distorts the dorsal side of the embryo. The line labeled M in (A), (B), and (E) denotes the midline of the neural tube as defined by the orientation of the ventricular cell layer. In(C), a section from an injected embryo stained with an anti-p-gal antibody is shown, while (D) shows a neighboring section stained with a muscle-specific antibody, 12/ 101. Note the increase in muscle mass in regions injected with XotchdE RNA. The extra tissue between the somite and the epidermis (bracket)was neverobserved on the uninjected side and failed to stain with neural-, muscle-, neural crest-, or epidermal-specific antibodies (data not shown). Bar, 75 pm (applies to [A] through ID]). (E and F) Embryos were injected with RNA as described above and treated with HUA at the beginning of gastrulation. In(E), asection from an injected embryo stained with the neuralspecific antibody is shown, while (F) shows an example of a section from an injected embryo stained with a muscle-specific antibody. The large diameter of the cells is evidence that cell division has been effectively blocked by the HUA treatment. Bar, 100 urn (applies to both El and W). Hoechst staining of nuclei showed that the density of cells was approximately the same on the control and injected sides (data not shown), indicating that the extra tissue represents additional cells and not an increase in cell diameter or extracellular matrix.|
|Figure 4. Effects of XofchdE on Tissue Differentiation in Xenopus Embryos (A-C) Embryos were injected once with XotchdE and B-gal RNA at the 4- to 6-cell stage, fixed at tadpole stages, stained with X-gal, sectioned in paraplast, and counterstained with hematoxylin and eosin. (A) An example in which RNA injection resulted in expression of XotchdE as marked by X-gal staining in the dorsal midline. Injected midljne cells have failed to differentiate normally (arrow), while cells from the uninjected side of the embryo have formed a notochord (N). Bar, 100 urn. (6) An example in which RNA injection resulted in XotchdEexpression in dorsal ectodermal derivatives. Expression results in neural tube hypertrophy (open arrow), a larger, disorganized otic vesicle (Ot), and larger numbers of interstial cells (closed arrow). (C)An example in which RNA injection resulted in expression in ventral ectodermal derivatives. Note the thickened lateral epidermal layer consisting Of XofchdE-expressing cells (arrowhead). The thickened layer, although somewhat disorganized, appears to be fully differentiated in that these cells labeled with the EpA epidermal marker in other experiments (Jones and Woodland, 1986). Bar, 170 urn (applies to both [B] and [Cl). (D and E) XotchdE or the control T6 RNA was injected into all four animal blastomeres of 4-ceil-stage embryos. At the neural plate stage, embryos were processed for whole-mount in situ hybridization using a probe to Xenopus twist (Hopwood et al., 1969). An example of twist staining in an embryo injected with T6 RNA is shown in (D), while (E) shows an example of twist staining in an embryo injected with XotchdE RNA. Note that twist staining of neural crest is absent in the XotchdE-injected embryo in (E), while the midline staining of the notochord and mesoderm is still present. The blastopore (BP), which denotes the posterior end of the embryo, is marked by the curved arrow. Straight arrows indicate the approximate olane of section for the samples shown in (F) and (G). iF and G) Tissue sections of embryos injected with RNA and stained for twist as in panels (D) and (E). The distinct twist-positive neural crest cells (NC) seen in the T6-injected animal in (F) are conspicuously absent from the neural plate (NP) in the XotchdE animal in (G). twist labeling remains in the notochord (N) and lateral mesoderm (M). Bar, 100 pm (applies to both [fl and [G]). (H and I) Embryos injected once with XotchdE and p-gal RNA were fixed at the neural plate stage and stained in whole mount with X-gal and an epidermal keratin in situ probe (see Experimental Procedures). The uninjected side of an embryo is shown in (H); the injected side as marked by X-gal staining is shown in (I). Note the sharp boundary of expression of the epidermal gene on the uninjected side in (H), while in regions of X-gal staining, expression of the epidermal gene is lost in (I), The blastopore (BP) is indicated with the curved arrow. The normal epidermal-neural plate (E/N) border is traced in with dotted lines.|
|Figure 5. XotchdE Increases the Response of Ectoderm to Induction In Vitro (A) Animal cap tissue from blastula embryos injected with XotchdE RNA or with T6 RNA as a control was induced to form neural tissue using Hensen’s node from a stage 4 chick embryo (Kintner and Dodd, 1991). The amount of neural tissue induced was measured using an RNAase protection assay for two neural transcripts, neurofilament 3 and N-CAM (Dixon and Kintner, 1969). Ectoderm from XotchdEinjected embryos forms more neural tissue in response to Hensen’s node (lane 2) than ectoderm from TG-injected embryos (lane 3). Note that the XotchdE animal caps cultured alone do show a very weak N-CAM signal (lane l), unlike what is observed in T6-injected ectoderm cultured alone (lane 4). However, the amount of neural tissue represented by this weak N-CAM signal does not account for the extra neural tissue generated in ectoderm injected with XotchdE and then treated with Hensen’s node. In this experiment, the increase in neural tissue between lanes 2 and 3 is 1.5fold when normalized to the elongation factor 1 a loading control. In other experiments, a 3-fold difference was seen. (B) Ectoderm was isolated at different stages from embryos injected with XotchdE or T6 RNA, treated with activin A, and then assayed after 24 hr in culture for the formation of muscle tissue by measuring the amounts of a musclespecific actin using an RNAase protection assay (Kintner and Dodd, 1991; see Experimental Procedures). The top set of lanes shows the results when ectoderm from TG-injected animal caps at different stages is treated with increasing doses of activin. Note that the ability to respond to activin induction is reduced at stage 11 and is almost completely lost by stage 11 S. The bottom set of lanes shows the results when ectoderm from stage-matched XOWUIEinjected embryos is treated with increasing doses of activin A. Note that ectoderm has a much greater response to activin at all doses when injected with XotchdE(P.S- and Sfold for the 50 and 500 pM activin treatments, respectively, in a stage 10 animal cap culture). In addition, note that ectoderm from XotchdE-injected embryos still responds to activin at stages when ectoderm from control embryos does not. For example, in ectoderm from stage 11.5 embryos, controls form very little mesoderm at 500 pM activin, while XotchdE caps still respond to 50 pM and show a robust induction at 500 pM (over lo-fold more than controls).|
|Figure 6. Comparison of Xotch, twist, and Epidermal Keratin Staining in Wild-Type Neural Plate Stage Embryos (A) Shown is a dorsal view of a stage 15 embryo processed for whole-mount in situ hybridization using a probe for Xotch. Anterior is to the left. Note that staining is restricted to twin dark stripes on the dorsal side, corresponding to the neural plate and the underlying somitic mesoderm. Staining is absent at the dorsal midline (notochord). Note also that the lighter staining on the lateral side of each dark stripe occurs within the ectoderm that gives rise to the tissues most affected by XotchdE RNA injection. The bracket marks the approximate region of the embryo from which the sections were taken for the comparisons shown in (6) through (D). (B-D) Sections through stage 14 embryos stained in whole mount for Xotch (B), twist(C), and epidermal keratin (D) RNA. Xotch staining is found in the neural plate (NP) but also extends laterally (limits are shown by arrows in [B]) into the cells that will make up neural crest as marked by twist expression in (C) (arrows) and the lateral epidermis as marked by epidermal keratin expression in(D). Bar, 125 pm (applies to [B] through [D]).|