RIPPLY3 is a retinoic acid-inducible repressor required for setting the borders of the pre-placodal ectoderm

Abstract

Retinoic acid signaling is a major component of the neural posteriorizing process in vertebrate development. Here, we identify a new role for the retinoic acid receptor (RAR) in the anterior of the embryo, where RAR regulates Fgf8 expression and formation of the pre-placodal ectoderm (PPE). RARα2 signaling induces key pre-placodal genes and establishes the posterolateral borders of the PPE. RAR signaling upregulates two important genes, Tbx1 and Ripply3, during early PPE development. In the absence of RIPPLY3, TBX1 is required for the expression of Fgf8 and hence, PPE formation. In the presence of RIPPLY3, TBX1 acts as a transcriptional repressor, and functions to restrict the positional expression of Fgf8, a key regulator of PPE gene expression. These results establish a novel role for RAR as a regulator of spatial patterning of the PPE through Tbx1 and RIPPLY3. Moreover, we demonstrate that Ripply3, acting downstream of RAR signaling, is a key player in establishing boundaries in the PPE.

Fig. 1.

Expression of Tbx1 and Ripply3 across developmental time. (A,B) Whole-mount in situ hybridization of Tbx1 (A) and Ripply3 (B) gene expression at Nieuwkoop and Faber developmental stages 13, 16, 18, 20, 22 and 26 (B). (C) QPCR showing Tbx1 gene expression over developmental time. (D) QPCR showing Ripply3 gene expression over developmental time. In C and D, the y-axis represents 2^–ΔCt values normalized to the housekeeping gene histone H4.

Fig. 2.

Ripply phylogenetic tree. Phylogenetic tree of Ripply sequences (aligned with MAFFT) generated using the PROml maximum likelihood-based method with global rearrangements. The scale bar represents the divergence distance of 0.3 amino acid substitutions per site of the Ripply sequence.

Fig. 3.

Tbx1 and Ripply3 expression are modulated by TTNPB and AGN193109 treatments. (A,B) QPCR of Tbx1 (A) and Ripply3 (B) expression from whole embryos treated with TTNPB (RAR-specific agonist), AGN193109 (RAR-specific antagonist) or control vehicle (0.1% ethanol). The y-axis represents fold induction relative to solvent control value at each stage. 2^–ΔΔCt values were normalized to histone H4. Asterisks represent differences between control and treated embryos that were statistically significant (P<0.05). (A) Tbx1 is induced by TTNPB early, and repressed by TTNPB in later stages. (B) Ripply3 is induced by TTNPB and repressed by AGN193109 across all developmental stages.

Fig. 4.

RARα2 is required for expression of PPE genes. Embryos were injected unilaterally at the two- or four-cell stage. The injected side is on the right, as indicated by the magenta β-galactosidase lineage tracer staining. (A,D,G,J) Control expression of Tbx1, Ripply3, Six1 and Eya1. (B,E,H,K) 10 ng RARα2 MO reduced expression of Tbx1 (14/15 embryos), Ripply3 (9/10), Six1 (6/11) and Eya1 (8/9). (C,F,I,L) 10 ng RARα2 MO + 0.5 ng RARα2 mRNA rescued expression of Tbx1 (5/11 embryos), Ripply3 (8/11), Six1 (11/14) and Eya1 (13/17).

Fig. 5.

Double whole-mount in situ hybridization reveals spatial relationship of PPE genes at stage 18. (A-C) Ripply3 (stained with BM Purple) is lateral to Tbx1, Six1 and Eya1 (stained with BCIP/Tetrazolium Blue), overlapping with Tbx1 and the posterior placodal (pp) area of Eya1. (D) Ripply3 (stained with Fast Red) is lateral to but does not overlap with Fgf8 (stained with BM-Purple). (E-G) Tbx1 (stained with BM Purple) is lateral to Fgf8, Six1 and Eya1 (all stained with Fast Red) in stage 18 embryos, overlapping significantly with Fgf8, and in the posterior placodal (pp) area of Eya1.

Fig. 6.

Lateral view of double wish staining and schematics. (A-G) Lateral views of double whole-mount in situ hybridization from Fig. 5, including schematics of the overlapping or non-overlapping PPE genes. pp, the posterior placodal area; pPrV, profundal placodal area. (H) Relationship of PPE genes at stage 18.

Fig. 7.

FGF8 establishes the PPE, but TBX1 maintains Fgf8 expression. All embryos were injected unilaterally at the two- or four-cell stage. The injected side is on the right. (A,C) Control expression of Tbx1 and Ripply3. (B,D) Embryos injected with 30 ng Fgf8 splice MO; Tbx1 (8/16 embryos) and Ripply3 (19/23) expression were knocked down or knocked out. (E,G,I,K) Embryos injected with 10 ng Tbx1 MO showed knockdown of Fgf8 (12/16 embryos), Ripply3 (17/21), Eya1 (21/26) and Six1 (25/36) expression. (F,H,J,L) 10 ng Tbx1 MO + 0.5 ng Tbx1 mRNA rescued Fgf8 (5/8), Ripply3 (6/8), Eya1 (6/8) and Six1 (8/11) expression.

Fig. 8.

RIPPLY3 inhibits the ability of TBX1 to activate T-Box response elements. (A) Diagram of T-Box response element TK-luciferase reporter (TBRE-TK-Luc). The Brachyury palindromic element is present in two copies, preceding the TK minimal promoter and luciferase. The essential GTG core is indicated in red. Inset: diagram of RIPPLY3, TBX1 and GROUCHO interactions, including the WRPW and FPVQ domains. (B) Whole-embryo luciferase assay reflecting Tbx1 transcriptional activity in the presence or absence of wild-type or mutant Ripply3. Relative light units were normalized to total protein (Milnes et al., 2008). Error bars represent biological replicates (multiple pools of five embryos derived from the same female frog). Statistics are relative to reporter alone (^*P<0.05). Tbx1 induces activity about threefold, whereas Ripply3 represses activity to basal levels when co-injected with Tbx1. Microinjection of the mutant Ripply3^{WRPQ→AAAA, FPVQ→AAAA} mRNA does not repress Tbx1 induction of reporter activity.

Fig. 9.

Knockdown of RIPPLY3 causes posterolateral shift of PPE gene expression. Embryos were injected at the two- or four-cell stage and the injected side is on the right. (A-H) Anterior (A,C,E,G) and dorsal (B,D,F,H) views of control expression of Fgf8, Tbx1, Eya1 and Six1. (I-P) Anterior (I,K,M,O) and dorsal (J,L,N,P) views of embryos injected with 50 ng Ripply3 MO 1+2 showing a posterolateral shift of Fgf8 (21/29 embryos), Tbx1 (14/19), Eya1 (12/18) and Six1 (12/20).

Fig. 10.

Ripply3 gain of function causes loss of Tbx1 expression. Embryos were injected at the two- or four-cell stage and the injected side is on the right. (A,B) Anterior (A) and dorsal (B) views of embryos injected with 0.25 ng Ripply3 mRNA showing a knockdown of Tbx1 (11/11). (C,D) Anterior (C) and dorsal (D) views of embryos injected with 0.25 ng Ripply3^WRPW→AAAA mRNA showing no change in Tbx1 (15/16) expression. (E,F) Anterior (E) and dorsal (F) views of embryos injected with 0.25 ng Ripply3^FPVQ→AAAA mRNA showing no change in Tbx1 expression (10/11).

Fig. 11.

Wiring diagram of RAR signaling and PPE genes. All interactions found in this paper, and specific to the PPE region, are diagrammed. RAR signaling induces Tbx1 and Ripply3 (blue lines). Ripply3 (red lines) regulates the PPE boundary of Tbx1, Fgf8, Six1 and Eya1 expression. Tbx1 (purple lines) and Fgf8 (green lines) engage in a positive-feedback loop, and both are required for Ripply3, Six1 and Eya1 expression. Previously published PPE interactions are also noted with circled numbers and the pathways indicated by broken lines (Ahrens and Schlosser, 2005; Brugmann et al., 2004; Schlosser et al., 2008).