QUICK SEARCH:   [advanced] Author: Keyword(s):
Home     Help     Feedback     Subscriptions     Archive     Search     Table of Contents

doi: 10.1242/10.1242/dev.00212

This Article Summary Full Text Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via HighWire Citing Articles via Google Scholar Google Scholar Articles by Woda, J. M. Articles by Artinger, K. B. Search for Related Content PubMed PubMed Citation Articles by Woda, J. M. Articles by Artinger, K. B.

Dlx proteins position the neural plate border and determine adjacent cell fates

Juliana M. Woda¹, Julie Pastagia^1,2, Mark Mercola^1,*, and Kristin Bruk Artinger^1,

¹ Department of Cell Biology, Harvard Medical School, Boston, MA 02115, USA
² Department of Oral Biology, Harvard School of Dental Medicine, Boston, MA 02115, USA
^* Present address: The Burnham Institute, Stem Cell and Regeneration Program, 10901 N. Torrey Pines Road, La Jolla, CA 92037, USA
Present address: Department of Craniofacial Biology, University of Colorado Health Sciences Center, Denver, CO, USA

View larger version (49K): [in a new window]   Fig. 1. Dlx gene expression becomes restricted to the ventral ectoderm. (A) At blastula stage (stage 9), Xdlx3 is expressed broadly throughout the ectoderm. Animal pole is oriented up. (B) By early gastrula stage (stage 10), Xdlx3 expression is restricted to the more ventral ectoderm (black arrowhead). Lateral view with dorsal oriented to the left. Dorsal lip is on the left (green arrowhead). (C) By the beginning of neurulation (stage 13), Xdlx3 expression is completely absent from the neural plate and is expressed throughout the non-neural ectoderm (dorsal view with anterior to the top). (D) Schematic of Dlx homeodomain constructs. The activating Dlx construct was made by ligating regions encoding the Dlx3 homeodomain (blue) to the VP16 activation domain (yellow). A conditional version was generated by fusion to the ligand-binding domain of the human glucocorticoid receptor (GR; green). Inhibitory constructs were made similarly using the Engrailed repressor domain (EnR; red). Identical constructs were made with the Dlx5 homeodomain. The homeodomains are highly conserved among Dlx family members; thus the fusion proteins are envisaged to regulate target genes of all family members comparably. See Materials and Methods for details of construct preparation.

View larger version (95K): [in a new window]   Fig. 2. Dlx activity restricts neural plate expansion but does not induce epidermal differentiation. ß-galactosidase and either EnR-Dlx3hd, EnR-Dlx5hd or VP16-Dlx3hd mRNAs were injected into one dorsal animal blastomere of 4-cell stage embryos. The embryos were then stained for the ß-galactosidase (as a lineage label; magenta stain) and assayed for Xsox2 (stage 13), NCAM (stage 17-18) or keratin (stage 17-18) expression by whole-mount in situ hybridization (blue stain), or by EpA immunostaining (stage 17-18) to reveal neural plate or epidermal differentiation. All views are dorsal with anterior to the top except M, which is lateral with anterior to the right and H and I, which are transverse histological sections through the neural tube. Dashed lines indicate the dorsal midline. (A) Xsox2 expression is the same on the injected and uninjected sides of control embryos injected with ß-galactosidase mRNA alone. (B) Injection of VP16-Dlx3hd mRNA reduces the Xsox2 domain on the injected side. (C,D) In contrast, embryos injected with EnR-Dlx3hd (C) or EnR-Dlx5hd (D) mRNA shows expanded Xsox2 expression on the injected side. (E-G) The NCAM domain was similarly reduced by or expanded by VP16-Dlx3hd and EnR-Dlx3hd, respectively. (H,I) Transverse sections through stage 25 embryos (H) the control embryo has a symmetrical, closed neural tube. (I) An embryo expressing EnR-Dlx3hd illustrates that the neural tube closed properly but was expanded on the injected side where ß-gal-positive cells populate the neural tube. Dorsal is to the top. (J-M) keratin expression; a marker expressed throughout the non-neural ectoderm. (J) Normal expression of epidermal keratin in a control embryo injected with ß-galactosidase mRNA alone. (K) Injection of VP16-Dlx3hd mRNA did not expand epidermal keratin expression. (L) Injection of EnR-Dlx3hd mRNA inhibited epidermal keratin expression. White arrows mark the loss of keratin expression in the injected region. (M) A lateral view illustrates the loss of keratin (blue) in the injected region (magenta ß-galactosidase stain). (N-P) VP16-Dlx3hd did not affect the epidermal epitope EpA (O), whereas EnR-Dlx3hd (P) prevented normal expression (N).

View larger version (57K): [in a new window]   Fig. 3. Rescue of EnR-Dlx3hd activity by full-length Dlx3. A constant amount ß-galactosidase and 50 pg of EnR-Dlx3hd mRNAs were injected unilaterally as in Fig. 2 along with increasing doses of mRNA encoding full-length Dlx3 (Dlx3 FL). (A-C) Embryos stained for Xsox2 show the typical expansion of the neural plate obtained with EnR-Dlx3hd (A), the rescue achieved in combination with Dlx3 FL (B), and the overexpression phenotype typical of Dlx3 FL-injected embryos (C). (D) Incidence of each of the phenotypes in A-C as a function of the molar ratio of mRNAs injected. Note that the incidence of each phenotype depends on the relative dose of Dlx3 FL mRNA injected.

View larger version (80K): [in a new window]   Fig. 4. Alterations in cell lineages that border the neural plate in embryos with localized expression of VP16-Dlx3hd or EnR-Dlx3hd. Embryos were injected unilaterally with EnR-Dlx3hd or VP16-Dlx3hd at the 8-16 cell stage, cultured until the appropriate developmental stage and then examined by in situ hybridization (blue stain). ß-galactosidase stain (magenta) indicates progeny of injected blastomeres. All embryos are shown as dorsal views with anterior to the top except for J and K which are lateral views, anterior to the right, and V-Y, anterior views, dorsal to the top. (A-F) Stage 13 embryos showing that markers of cells at the border of the neural plate (Xhairy2A and Xmsx-1) are ablated by VP16-Dlx3hd and shifted laterally by EnR-Dlx3hd. Arrows mark displacement caused by EnR-Dlx3hd relative to uninjected side of the embryo. (G-P) Stage 13 embryos showing identical effects on markers of neural crest precursors (Xsnail and Xslug). Lateral views of Xsnail expression illustrate the extent of the shift (J,K) seen in dorsal view (I, arrows). Transverse sections (O,P) showing Xslug expression (blue) and ß-galactosidase (magenta) illustrate that the size of the Xslug expression domain is unaltered but occurs at the lateral margin of the cells expressing the injected mRNA. (Q-U) Primary neurons (marked by N-tubulin) are also ablated by VP16-Dlx3hd and displaced laterally by EnR-Dlx3hd in stage 14 Xenopus (Q-S) and 2-somite stage zebrafish (T,U). (V-Y) Stage 18 embryos showing cranial placode precursors (marked by Xsix1) shifted medially or laterally by localized expression of VP16-Dlx3hd and EnR-Dlx3hd, respectively (W,X). Note that the anterior domain of Xsix1 is unaffected, even where widespread expression of EnR-Dlx3hd ablates Xsix1 more laterally (Y).

View larger version (57K): [in a new window]   Fig. 5. Dlx activity is required before the end of gastrulation to position the neural plate border and specify adjacent cell fates. Embryos were injected unilaterally with either VP16-Dlx3hd-GR or EnR-Dlx3hd-GR mRNAs as above and dexamethasone was then added at various times to examine the temporal requirements for Dlx activity. Embryos were assayed for Xsox2 (stage 15) and Xslug (stage 17) by in situ hybridization (blue stain). ß-galactosidase staining (magenta stain) marks the progeny of the injected blastomeres. In all cases, control embryos injected with the fusion protein constructs but cultured without dexamethasone remained unaffected (not shown). (A-D) Addition of dexamethasone at stage 6 expanded Xsox2 (A) or outwardly shifted Xslug (C) in embryos expressing EnR-Dlx3dh and ablated both markers in embryos expressing VP16-Dlx3hd (arrows in B and D). (E,F) In contrast, addition of dexamethasone at stage 11.5-12 to either EnR-Dlx3hd- or VP16-Dlx3hd-injected embryos did not affect Xsox2. (G) Similarly, stage 11.5-12 dexamethasone addition to EnR-Dlx3hd-injected embryos caused no change in Xslug. (H) However, stage 11.5-12 dexamethasone addition to VP16-Dlx3hd-injected embryos ablated Xslug, indicating that the neural crest remains sensitive to Dlx activity after gastrulation. (I) The time course of the dexamethasone effect on EnR-Dlx3hd-expressing embryos suggests that endogenous Dlx activity affects neural crest patterning before the end of gastrulation.

View larger version (76K): [in a new window]   Fig. 6. A positive role for Dlx function in non-neural ectoderm. (A-D) Embryos were injected unilaterally with EnR-Dlx3hd and ß-galactosidase (magenta stain). Broad domains of EnR-Dlx3hd expression dorsoventrally show maximal expansion of the neural plate is limited to approximately twice its normal width, as visualized by Xsox2 expression (A,B blue stain). Arrows in B show staining for ß-gal indicative of injected mRNA outside of the Xsox2 domain. Note that in these cases of broad EnR-Dlx3hd expression, Xslug expression was not reactivated at the border of the expanded neural plate (C,D), similar to ablation of Xsix1 (Fig. 4Y) and other markers (not shown). (E) Schematic showing isochronic stage 12 transplant of fluorescent dextran (pseudo colored green stain) injected donor neural plate tissue (N) to ventral ectoderm of host embryos. Host embryos were injected ventrally to express ß-galactosidase either alone or with EnR-Dlx3hd. Control grafts showed Xslug induction at stage 25 (photo; ventral view, anterior to top). (F-I) Controls show that donor neural plate explants cultured alone express no or only minimal levels of Xslug (F) or Xsix1 (G), unless taken from a more lateral region that included prospective epidermis (H, Xslug; I, Xsix1). Therefore, marker expression is indicative of neural crest (Xslug) or placode (Xsix1) induction. (J-N) High magnification views of grafts into control hosts that had been injected with ß-gal mRNA alone. The three panels shown are light field, fluorescent and merged images. Induction (blue stain) of Xslug (J-L) and Xsix1 (M,N) was observed in areas adjacent to and overlapping both donor (green) and host tissue. In particular, note expression (arrows in J,K,M, arrowheads in L,N) within and adjacent to host tissue expressing the injected ß-galactosidase (magenta stain). Sections (L) show Xslug expression (blue) adjacent to donor tissue (green). (O-S) High magnification views of grafts into experimental hosts that had been injected with EnR-Dlx3hd and ß-gal mRNAs. Unlike controls (J-N), Xslug and Xsix1 are not induced adjacent to host tissue that had been injected with EnR-Dlx3hd and ß-galactosidase mRNAs (magenta). P shows a graft inserted adjacent to both injected and uninjected host tissue. Note that induction was prevented near cells that expressed EnR-Dlx3hd (red arrowhead) but occurred where host tissue lacked injected mRNAs (black arrowhead).

View larger version (27K): [in a new window]   Fig. 7. Model for role of Dlx in positioning the neural plate border and patterning adjacent cell fates. (A) We postulate that a reciprocal (inhibitory) interaction between Dlx and neural plate factors refines an initial neural plate: non-neural ectoderm bias along the mediolateral axis. The initial bias is established earlier, possibly by BMP, Wnt and/or FGF signaling. This reciprocal interaction leads to a sharpening of the border between the neural plate and non-neural ectoderm and specifies the precise position along the mediolateral axis. (B-F) Schematics of ectoderm along the mediolateral axis illustrate the function of Dlx factors under normal and manipulated conditions. (B) Under normal conditions, Dlx activity is required in non-neural ectoderm (yellow) for short-range communication (arrows) that leads to induction of lateral primary neuron, neural crest and cranial placode precursors. (C) Local inhibition of Dlx activity within the region expressing injected EnR-Dlx3hd (red bars) causes the neural plate to expand. Border region cell lineages are induced laterally where the expanded neural plate contacts Dlx-positive non-neural ectoderm. (D) When Dlx3 activity is inhibited beyond the intrinsic limit of neural plate expansion, the neural plate expands maximally but is separated from Dlx-positive epidermis. Short-range communication is attenuated or not initiated, resulting in absence of neural crest, cranial placodes and lateral primary neurons. (E) Overexpression of Dlx activity (with VP16-Dlx3hd or full-length Dlx3) inhibits neural plate induction but cannot initiate epidermal differentiation (Fig. 2); thus, the neural plate narrows but short-range signaling and normal cell fate specification does not occur in the border region. (F) Cranial placodes can be shifted medially in cases where a small region of ectopic Dlx activity separates competent neural plate and non-neural ectoderm, suggesting that placode-inducing signals can traverse a short region of interposing tissue.