Lamar E and Kintner C (2005), The Notch targets Esr1 and Esr10 are differenti...

XB-ART-1554

Development 2005 Aug 01;13216:3619-30. doi: 10.1242/dev.01937.

Show Gene links Show Anatomy links

The Notch targets Esr1 and Esr10 are differentially regulated in Xenopus neural precursors.

Lamar E , Kintner C .

???displayArticle.abstract???
The HES family of bHLH repressors plays a key role in regulating the differentiation of neural precursors in the vertebrate embryo. Members of the HES gene family are expressed in neural precursors as targets of the Notch signaling pathway, but how this occurs in the context of neurogenesis is not known. Here, we address this issue by identifying enhancers driving Notch-dependent gene expression of two Hes5-like genes expressed in Xenopus called Esr1 and Esr10. Using frog transgenesis, we identify enhancer elements driving expression of Esr1 and Esr10 in neural precursors or in response to ectopic expression of the proneural protein, Xngnr1. Using deletion and mutation analysis, we define motifs required for enhancer activity of both genes, namely Notch-responsive elements and, in the case of Esr10, E-box motifs. We find that Esr1 and Esr10 are differentially regulated both in terms of Notch input and its interaction with heterologous factors. These studies reveal inputs required for proneural expression of genes encoding bHLH repressors in the developing vertebrate nervous system.

???displayArticle.pubmedLink??? 16077089
???displayArticle.link??? Development

Species referenced: Xenopus laevis
Genes referenced: dlc dll1 eef1a2 esr1 hes1 hes4 hes5 hes5.2 hes5.5 neurog2 notch1 sms sult1a1 tbx2

???attribute.lit??? ???displayArticles.show???

	Fig. 2. Esr1 and Esr10 are expressed in proneural domains. Esr1, Esr10 and Delta are expressed in the neural tube (A,D,G, white arrowhead), cranial ganglia (A,D,G, arrow) and anterior neural tube (B,E, arrowhead). Esr10 is also expressed in the tailbud (D, black asterisk) and somitomeres (D, white asterisk). Delta1 is also seen in somitomeres (G, white asterisk), although the Notch ligand predominantly expressed in the tailbud is Delta2 (Jen et al., 1997). At neurula stages, Hairy2 is barely detectable in the neural tube (J) but expressed in presumptive neural crest (K, arrowhead). Hairy2 expression in the eye (K, arrow) precedes that of Esr1 and Esr10. Misexpression of mRNA encoding Xngnr1 induces Esr1 (C), Esr10 (F) and Delta1 (I); C′, F′ and I′ show uninjected sides. Hairy2 is not upregulated by misexpressed Xngnr1 (L; L′, uninjected side). Turquoise stain in C,F,I,L reflects activity of the lacZ tracer gene. (M, left) RNase protection assay showing that expression of Xngnr1 in neuralized animal caps analyzed at stage 12 induces Esr1 and Esr10 that can be inhibited by expression of SuH, and in the case of Esr1, further increased by co-injection of ICD. EF1a expression serves as a loading control; `Embryos' indicates staged-matched controls. Quantification (right) shows fold increases in Esr1 and Esr10 relative to their respective `noggin only' control, which is set arbitrarily to 1.
	Fig. 3. Compact elements drive neural Esr gene expression.(A) Promoters of Esr1, Esr7, Esr10 and Hairy2 (Davis et al., 2001) show high and moderate homology of S1 and S2, respectively, in the SPS (green). All exhibit a conserved CCAAT motif (blue). GFP expressed by deletion mutants of Esr1 (B,C-E) and Esr10 (B,F-H) in transgenic frogs, followed by whole-mount in situ hybridization, indicates that short elements drive Esr gene expression in the neural tube (D,G, arrowheads). Esr10/Dra also drives somitomeric (G, arrow) and tailbud (G, asterisk) GFP expression. Deletion to a Hin3 (E) site attenuates Esr1 GFP, although expression remains restricted to neural tissue (E, arrowhead). Deletion to a Pst site (H) abrogates Esr10 neural expression, although diffuse somitomeric expression (H, arrow) remains. Activities using the neural tube (NT) as a reference are summarized in (B, right; see Table 1 for details). Sections through the neural tube of stage 20 Esr1/RV transgenic embryos (J) show GFP-positive cells in the ventricular zone in a pattern similar to the endogenous gene (I). Also summarized (B, right) are data reported in Figs 4, 6 and 8 and Table 2 that are relevant to responses to ectopic Xngnr1 (NA; not assayed).
	Fig. 4. Proximal elements constitute Esr1 and Esr10 proneural enhancers. Sequences driving neural GFP expression of Esr1 and Esr10 constructs are shown schematically in F. (A-E) Xngnr1 mRNA (ngn) with a lacZ tracer mRNA was injected into embryos made transgenic with sequences flanking Esr1, Esr10 and Hairy2. Embryos were stained for GFP by in situ hybridization. GFP expression driven by Esr1/RV (A) and Esr10/Dra (C) is induced by Xngnr1. Esr1/Hin3 (B), Esr10/Pst (D) and the 500 bp H2 flanking sequence (E) are not, indicating that they lack elements responsive to proneural input.
	Fig. 5. An intact SPS is required for Esr1 and Esr10 expression. S1 (I, left) is highly conserved in Esr1, Esr10 and homologous genes, and matches the optimal RTGRGAR consensus determined by Tun et al. (Tun et al., 1994). S2 of Esr1, Esr10 and several E(spl) homologs is less conserved (mismatches in red). S2 is reported as the bottom strand. Su(H) sites within the SPS of Esr1 (B,C) and Esr10 (F,G) were mutated individually (mS1 or mS2) by changing G5 to a C, and GFP expression in transgenics was monitored by in situ hybridization and compared with wild-type controls (A,E). Neural and somitomeric Esr10 expression required two intact Su(H) sites (F,G), while neural Esr1 expression required only S1 (B,C). Injection of Xngnr1 (ngn; injected side down) mRNA could not rescue GFP expression in embryos carrying S1 mutations of Esr1 (Esr1/RvmS1) (D) or Esr10 (Esr10/DramS1) (H). (J) Luciferase activity of HeLa cells transfected with Esr1/RV SPS mutants showed that whereas mS1 abrogated transcription, mS2 had no effect.
	Fig. 6. Esr10 proneural enhancer activity requires intact E-boxes. HeLa cells were transfected with expression vectors for ICD, Xngnr1 and E47 (N/E), or both, and with luciferase vectors driven by wild-type (Esr10/Dra) or mutated (Esr10/DramE1E2) elements, shown schematically above (A). In the case of Esr10/Dra, ICD and N/E synergistically activate transcription approximately three times more over ICD alone (A, left). Synergy was lost when E-boxes were mutant (A, right). E-box motifs were also required for GFP expression driven by Esr10/Dra in transgenic frogs (compare C with B). (D,E) Injection of Xngnr1 mRNA with a lacZ tracer into Esr10/DramE1E2 transgenic embryos (E) could not activate enhancer activity as was seen with controls.(F) EMSA showing that Xngnr1 (N) and E47 (E) proteins shift an E2 oligo; shifts were competed by 10× and 100× cold competitor (WT) but not by similar increases mutant E2 oligos (Mut) or oligos corresponding to a binding site of a heterologous activator (Vax) (Mui et al., 2005). O, oligo; R, reticulocyte lysate; N/E, Xngnr1 plus E47. Complexes formed by E47 homodimers (Ex2) are of higher mobility than those formed by Xngnr1/E47 (N/E) heterodimers. ns, nonspecific complexes attributable to reticulocyte proteins.
	Fig. 7. The Esr1 enhancer does not require E-boxes and responds to Notch through two loci. (A) Luciferase activity of Esr1/RV and Hin3 fragments co-transfected with activated Notch (ICD) plus or minus Xngnr1 (ngn). (B) Luciferase activity of Esr1/RV and Esr1/Hin3 vectors co-transfected with ICD compared with proximal elements from mouse Hes1 (Jarriault et al., 1995) and Xenopus Hairy2 (Davis et al., 2001), Esr10/Dra and a vector containing eight multimerized Su(H) sites (Ling et al., 1994). Cells were transfected simultaneously with equal levels of ICD (100 ng/well) relative to the reporter (100 ng/well). (C) Luciferase activity of Esr1/RV co-transfected with increasing (25 ng/well and 100 ng/well) levels of ICD compared with a construct in which all upstream Su(H) sites are mutant (Esr1/RVmS3-5) or the Esr1/Hin3 deletion mutant. Unlike the wild-type reporter, luciferase activity of the Su(H) and Hin3 mutant constructs saturates at low (25 ng) ICD levels. (D) An S4 mutation results in loss of transcription similar to Esr1/RVmS-35.
	Fig. 8. Esr1 enhancer activity requires upstream Su(H) sites in vivo. (A,B) GFP expression in frogs transgenic with enhancer elements containing mutant E-boxes (Esr1/RVmE1E2E3) versus wild-type controls. Wild-type (C) and E1E2 mutant (D) embryos were injected with mRNA encoding Xngnr1 (ngn) and stained for GFP. GFP expression in frogs with mutant enhancers is unchanged relative to controls. (E,F) Transgenic frogs bearing Esr1 enhancer elements mutant in upstream Su(H) sites (Esr1/RVmS3-5) show greatly attenuated GFP activity (E) relative to controls (A), and activity is not inducible following Xngnr1 injection (F). (G) Within the S3-5 cluster, mutations within S4 (H), which is conserved in sequence and position in numerous Esr1 homologs, greatly attenuate enhancer activity.
	hes5.1 (hes family bHLH transcription factor 5) gene expression in Xenopus laevis embryo, assayed via in situ hybridization, NF stage 19-21, lateral view, anterior left, dorsal up (A) and anterior view (B).
	esr10 (enhancer of split related 10) gene expression in Xenopus laevis embryo, assayed via in situ hybridization, NF stage 19-21, lateral view, anterior left, dorsal up (D) and anterior view (E).
	dll1 (delta-like 1) gene expression in Xenopus laevis embryo, assayed via in situ hybridization, NF stage 19-21, lateral view, anterior left, dorsal up (G) and anterior view (H).