Identification of target genes for the Xenopus Hes-related protein XHR1, a prepattern factor specifying the midbrain–hindbrain boundary

Developmental Biology 283 (2005) 253 – 267 www.elsevier.com/locate/ydbio

Identification of target genes for the Xenopus Hes-related protein XHR1, a prepattern factor specifying the midbrain–hindbrain boundary Hitomi Takadaa,c,1, Daisuke Hattoria,c,1,2, Atsushi Kitayamab, Naoto Uenob, Masanori Tairaa,c,* a Department of Biological Sciences, Graduate School of Science, University of Tokyo, Japan Department of Developmental Biology, National Institute for Basic Biology, 38 Nishigonaka Myodaijicho, Okazaki 444-8585, Japan c Core Research for Evolutional Science and Technology (CREST), Japan Science and Technology Agency, 7-3-1 Hongo, Bunkyo-ku, Tokyo 113-8654, Japan

b

Received for publication 16 July 2004, revised 2 April 2005, accepted 16 April 2005

Abstract The midbrain – hindbrain boundary (MHB) acts as a local organizer in the development of the CNS in vertebrates. Previously, we identified an MHB-specific bHLH-WRPW transcriptional repressor gene, Xenopus Hes-related 1 (XHR1), which is initially expressed in the presumptive MHB (pre-MHB) region at the early gastrula stage. To better understand the gene cascades involved in MHB formation, we investigated the genes downstream from XHR1 by differential screening using a Xenopus cDNA macroarray and a dexamethasone (DEX)inducible, dominant-negative transcriptional activator construct of XHR1 (XHR1-VP16-GR). Among the newly identified candidate target genes of XHR1 were Enhancer of split-related genes (ESR1, ESR3/7, and ESR9) and Xenopus laevis cleavage 2 (XLCL2). XHR1-VP16-GR induced the expression of the ESR genes and XLCL2 as well as Xdelta1, Xngnr1, and XHR1 itself in the presence of DEX even after pretreatment with the protein synthesis inhibitor, cycloheximide. This suggests that these genes are direct targets of XHR1. XHR1knockdown experiments with antisense morpholino oligos and ectopic expression of wild-type XHR1 revealed that XHR1 is necessary and sufficient to repress ESR genes in the pre-MHB region. Misexpression of the ESR genes in the pre-MHB region repressed the MHB marker gene, Pax2, suggesting that the repression of the ESR genes by XHR1 is at least partly required for the early development of the pre-MHB. Our data also show that XHR1 is not activated by Notch signaling, differing from ESR genes. Taken together, we propose a model in which XHR1 defines the pre-MHB region as a prepattern gene by repressing those possible direct target genes. D 2005 Elsevier Inc. All rights reserved. Keywords: Xenopus; Midbrain – hindbrain boundary; Prepattern gene; bHLH-WRPW; XHR1; ESR1; ESR3/7; ESR9; XLCL2; Cl2; Lbh; Xdelta1; Xngnr1; Notch

Introduction The development of multicellular organisms requires many steps of spatially and temporally coordinated induction, patterning, regionalization, and cell fate decisions. In the development of the vertebrate nervous system, the Spemann –Mangold organizer in amphibians or its equiv* Corresponding author. Department of Biological Sciences, Graduate School of Science, University of Tokyo, Japan. Fax: +81 3 5841 4434. E-mail address: [email protected] (M. Taira). 1 These authors equally contributed to this work. 2 Present address: Department of Biological Chemistry, University of California at Los Angeles, Los Angeles, CA 90095, USA. 0012-1606/$ - see front matter D 2005 Elsevier Inc. All rights reserved. doi:10.1016/j.ydbio.2005.04.020

alent tissue in other vertebrates is responsible for the neural induction and initial patterning of the resulting neuroectoderm (De Robertis et al., 2000; Harland and Gerhart, 1997). At later stages, the neural plate or neural tube undergoes further patterning processes, which lead to the formation of morphologically and functionally distinct regions of the central nervous system (CNS) (Bally-Cuif and Hammerschmidt, 2003; Lumsden and Krumlauf, 1996). In the course of CNS development, various sets of neuronal and glial cells arise from undifferentiated progenitor cells at defined times and in defined regions, and eventually create the intricate neuronal network of this nervous system (Qian et al., 2000).

254

H. Takada et al. / Developmental Biology 283 (2005) 253 – 267

The patterning and regionalization process of the vertebrate CNS is also attributed to the activity of local organizers that are situated in the neural plate or neural tube, and induce subsets of brain structures. The midbrain – hindbrain boundary (MHB) is one of the best studied local organizers, inducing the tectum from the midbrain and the cerebellum from the anterior part of the hindbrain, as has been shown in the chick and mouse. The molecules responsible for the induction properties of the MHB have been identified as Fgf8 and Wnt1, and their expression depends on the homeobox genes, Pax2/5/ 8 and En1/2. These secreted molecules and transcription factors are also necessary for the formation and maintenance of the MHB region, as has been shown by mutant analyses in mice and zebrafish (Martinez, 2001; Rhinn and Brand, 2001; Wurst and Bally-Cuif, 2001). Other MHB genes that are thought to be necessary for the establishment of the MHB are the Xenopus gene, XHR1, encoding the proline-basic helix – loop – helix (bHLH) transcriptional repressor (Shinga et al., 2001), and the zebrafish genes including the POU domain transcription factor gene Spg/POU2, the Xenopus ortholog of which is Oct25 (Belting et al., 2001), the zinc-finger protein gene Bts1 (Tallafuss et al., 2001), and the homeobox genes Iro1 and Iro7 (Glavic et al., 2002; Itoh et al., 2002). Thus, multiple genes are involved in specifying the MHB as a local organizer for neural patterning. Although the mechanisms underlying the induction properties and establishment of the MHB have been extensively studied, the initial step in MHB formation is not yet thoroughly understood. The MHB has been thought to be defined by two mutually repressing homeodomain transcription factors, Otx2 and Gbx2, which are expressed in the region anterior and posterior to the presumptive MHB (pre-MHB), respectively, and which control the expression of MHB genes such as Pax2, En2, and Fgf8 (Joyner et al., 2000; Millet et al., 1999; Simeone, 2000). In our previous expression pattern screening for early neural genes, we identified a novel MHB gene, XHR1, in Xenopus. We found that XHR1 expression begins in the pre-MHB region at the early gastrula stage before the expression of the previously identified MHB marker genes, Pax2, En2, and Fgf8. XHR1 expression only partially overlaps that of Otx2 and Gbx2 at this stage, implying that interactions between Otx2 and Gbx2 are not necessary for the initiation of XHR1 expression (Shinga et al., 2001). This observation is supported by a genetic study with mice in which the Otx2 and Gbx2 genes were both knocked out, demonstrating that these two genes are not required for the initiation of MHB gene expression, but instead are involved in refining the positioning of the pre-MHB region (Li and Joyner, 2001). Thus, the initial step in pre-MHB formation and the gene cascade from the pre-MHB to the MHB have not yet been clarified. It has been proposed that the pre-MHB region carries unique characteristics, in which neurogenesis is actively

repressed. The zebrafish MHB gene her5, which is a homolog of XHR1, is required to repress neurogenin 1 and the Cdk inhibitor p27 xic1 in order to maintain the preMHB neuron-free (Geling et al., 2003). In the mouse, when the hairy/Enhancer of split [E(spl)] genes, HES1 and HES3, were knocked out, premature neuronal differentiation was observed in the MHB region. This in turn led to the premature termination in this region of the expression of Wnt1, Fgf8, Pax2/5/8, and En1/2, which subsequently caused the loss of the midbrain and anterior hindbrain structures (Hirata et al., 2001). These data suggest that the temporal repression of neuronal differentiation in the MHB region is critical for the inductive activity of MHB. It is therefore important to investigate the mechanism of MHB formation as one of the earliest events in neural patterning, and also to examine the relationship between cell fate decisions and patterning in the CNS. To address these subjects, we investigated the target genes of the earliest MHB-specific transcriptional repressor, XHR1, during the specification of the pre-MHB region. We have previously shown that a dominant-negative activator form of XHR1, XHR1-VP16, disrupts the morphological features of the MHB and reduces the expression of the MHB marker genes Pax2 and En2. Conversely, misexpression of XHR1 expands En2 expression, suggesting that XHR1 plays an important role in the specification and positioning of the MHB. In this study, we undertook a large-scale screening for XHR1 target genes, using a Xenopus cDNA macroarray with hormone-inducible XHR1-VP16, designated XHR1-VP16-GR. Our functional analyses using mRNA and antisense-morpholino oligo injections suggest that XHR1 acts as a prepattern factor in pre-MHB development by repressing various genes which include E(spl)-related genes induced by Notch signaling (Davis and Turner, 2001), a Notch ligand, Xdelta1, and a possible coactivator gene, XLCL2/Lbh (Briegel and Joyner, 2001; Paris and Philippe, 1990).

Materials and methods Manipulations of Xenopus embryos Artificial fertilization and rearing of pigmented and albino embryos were performed as described previously (Shinga et al., 2001). Embryos were staged according to Nieuwkoop and Faber (Nieuwkoop and Faber, 1967). Embryos injected with mRNA for enhanced green fluorescent protein (eGFP) as a tracer (see below) were dissected at the late gastrula to early neurula stage with fine forceps under a fluorescent dissection microscope (Leica) to isolate the dorsoanterior tissue that contains the pre-MHB region (see Fig. 1C). The dissected pieces were treated with collagenase (0.75 mg/mL) in 1  MBS (modified Barth’s solution) for 3 min at room temperature to separate the

H. Takada et al. / Developmental Biology 283 (2005) 253 – 267

255

Fig. 1. Preparations for macroarray screening. (A) Pax2 is downregulated by XHR1-VP16-GR only in the presence of DEX (indicated by arrow). mRNAs for XHR1-VP16-GR (12.5 pg) and nh-gal were coinjected into the dorsoanimal region of the left blastomere at the four-cell stage. Half the injected embryos were treated with DEX at the early gastrula stage (stage 10.5 – 11) (right) and the other half were untreated (left). Following nh-gal staining (red), WISH was performed with a Pax2 probe at the early neurula stage (stage 13 – 14; purple). The images are dorsal views with the anterior to the top (the orientation of the following figures is the same as shown here unless otherwise stated). (B) Pax2 expression levels and gastrulation-arrest rates upon DEX treatment at different stages. Embryos were coinjected as described in panel (A) with mRNAs for XHR1-VP16-GR or h-globin (100 pg) and nh-gal, treated with DEX from the indicated stages to stage 13 – 14, and subjected to WISH for Pax2 expression. Colored bars indicate Pax2 expression levels on the injected side. Percentages of embryos with gastrulation arrest are shown on the right side of the graph. N, number of embryos. Each datum is from two or more individual experiments. (C) A schematic representation of the dissection procedure used to isolate the anterior neuroectoderm containing the pre-MHB region. Embryos were injected with mRNA at the four-cell stage, cut into halves in the middle of the body length at the late gastrula to early neurula stage (stage 12.5 – 13). The anterior neuroectoderm (ANE) together with the underlying anterior endomesoderm (AEM) was dissected and separated by collagenase digestion. (D, E) Dissected fragments containing the pre-MHB region as ascertained by XHR1 expression visualized by WISH (D), or by RT-PCR of extracted total RNA for XHR1 expression (E). Histone H4 was used as the loading control.

neuroectoderm containing the pre-MHB region from the underlying endomesoderm. Plasmid construction and in vitro mRNA synthesis The human glucocorticoid hormone receptor ligandbinding domain (GR) (nucleotides 1666– 2463; accession no. NM000176) and the VP16 activation domain (nucleotides 418 –651; accession no. U89963) were PCR-amplified with Pfu DNA polymerase and the following sets of primers carrying the indicated restriction enzyme sites (underlined): 5V-GCTCTAGAACGCGTTCTGAAAATCCTGGTAA3V (GR-F1, XbaI); 5V-GGACTAGTGCTAGCTCACT-

TTTGATGAAACAGAA-3V (GR-R1, SpeI – NheI); 5VCCGCTCGAGGCCCCCCCGACCGATGT-3V (VP16-F2, XhoI); and 5V-GCTCTAGACCCACCGTACTCGTCAA-3V (VP16-R2, XbaI). PCR products were cleaved with the appropriate restriction enzymes and cloned sequentially into the XbaI and XhoI sites of the pCS2+ vector to construct pCS2 + VP16-GR. A BamHI – SphI fragment of pCS2 + XHR1-VP16 was cloned into the pCS2 + VP16-GR vector cleaved with the same set of restriction enzymes to construct pCS2 + XHR1-VP16-GR (note that SphI is an internal site of VP16). The coding sequences of ESR1 and XLCL2 cDNA were Pfu-PCR-amplified from the National Institute for Basic Biology (NIBB) expressed sequence tag (EST)

256

H. Takada et al. / Developmental Biology 283 (2005) 253 – 267

clones, and cloned into the NcoI and EcoRI sites of the pCS2+AdN vector (Mochizuki et al., 2000), and EcoRI and XhoI sites of the pCS2+ vector to construct pCS2+ESR1 and pCS2+XLCL2, respectively. Template preparation and in vitro synthesis and quantification of mRNAs were performed as described previously (Hiratani et al., 2001). Other plasmids used for mRNA synthesis were pSP64-Xhm (h-globin) (Krieg and Melton, 1984), pCS2+eGFP, pCS2+nh-gal (nuclear h-galactosidase), pCS2+XnotchICD (Chitnis et al., 1995), and pCS2+Su(H)1DBM (Wettstein et al., 1997).

0.1% SDS at 55-C for 5 and 55 min, and with 0.2  SSC, 0.1% SDS at 65-C for 70 and 80 min. The membranes were exposed to an imaging plate which was analyzed with BAS 5000 (Fuji). The relative intensity of each clone was quantified and normalized against the global background signal (ArrayGauge). Whole-mount in situ hybridization

Embryos were injected with mRNAs (3 –10 nl/blastomere) and cultured until the embryos reached the appropriate stage. nh-gal mRNA (30 pg/blastomere) was coinjected as a lineage tracer and the enzymatic activity of nh-gal was visualized with Red-Gal (Research Organics) as substrate. Dexamethasone (DEX; 20 mM in ethanol) and cycloheximide (CHX; 10 mg/ mL in water) were added to the incubation medium to final concentrations of 10 AM and 10 Ag/mL, respectively. Antisense morpholino oligos which were designed to complement 25-nucleotide sequences upstream from the start codon of XHR1 pseudoalleles, XHR1A/B and A’ (XHR1-MO2, 5VGGTAAGATGATGATGATGATGAAGA-3V; and XHR1MO3, 5V-GGTAAGATGATGATGACGATGATGA-3V, respectively), a five-base-mismatched control oligo for both XHR1-MO2 and XHR1-MO3 (XHR1-5mmMO, 5VGGTAACATAATGATGAAGAAGACGA-3V; mismatched nucleotides are underlined), and a nonspecific control oligo were purchased from Gene Tools. XHR1-MO2 and XHR1MO3 were used as a mixture of equal amounts, hereafter called XHR1-MOs.

Whole-mount in situ hybridization (WISH) was performed with pigmented or albino embryos according to the method of Harland (Harland, 1991), using an automated AIH101 or AIH201 system (Aloka). Digoxigenin (DIG)- or fluorescein isothiocyanate (FITC)-labeled antisense RNA probes were synthesized from either linearized plasmid DNAs or PCR products, using T7 RNA polymerase. Pigmented embryos were bleached with 10% H2O2 in MeOH after a chromogenic reaction with BM Purple (Roche) and extensive washing with maleate buffer. Stained embryos were cleared with benzyl benzoate/benzyl alcohol (BB:BA = 2:1) as indicated. For double in situ hybridization, fixed albino embryos were hybridized with a mixture of DIG- and FITC-labeled probes. For the first round of staining, hybridized embryos were incubated with anti-FITC antibodies conjugated with alkaline phosphatase for staining with 5-bromo-4-chloro-3-indolyl phosphate (BCIP, Roche; turquoise color). Before the second round of staining, alkaline phosphatase was inactivated by sequential treatments of BCIP-stained embryos with 25%, 50%, 75%, 100%, 75%, 50%, and 25% methanol in maleate buffer for 5 min each, except for the 100% treatment which was for 1 h. The embryos were then incubated with antiDIG antibodies conjugated with alkaline phosphatase and stained with BM Purple (purple color).

cDNA macroarray analysis

RT-PCR

Five cDNA macroarray membranes were prepared with a total of 48,000 clones from Xenopus laevis normalized cDNA libraries constructed from stage 10.5 (5760 clones), stage 15 (19,200 clones), and stage 25 (23,040 clones) embryos. ESTs of these clones were deposited in DDBJ/ GenBank. 32P-Labeled cDNA probes were prepared as follows. Total RNA was extracted from dissected pre-MHB regions using a guanidinium isothiocyanate-acid phenol method (Chomczynski and Sacchi, 1987) with some modifications. Total RNA (5.6 – 15.7 Ag) was reversetranscribed with oligo dT primer and four dNTPs containing [32P]dCTP for 30 min at 37-C in a 29-Al reaction mixture, then incubated for another 2 h after the addition of cold 0.5 mM dCTP. cDNA probes were purified with Probe Quant G50 (Pharmacia), denatured, and hybridized to cDNA macroarray membranes in hybridization solution [5  saline sodium citrate (SSC), 5  Denhardt’s, 0.1% sodium dodecyl sulfate (SDS), 0.1 mg/mL salmon sperm DNA] at 65-C overnight. Filters were washed four times: with 2  SSC,

Reverse transcription-PCR was performed as described previously (Osada et al., 2003). The primers and PCR cycles used were as follows: histone H4, forward, 5V-CGGGATAACATTCAGGGTATCACT-3V and reverse, 5V-ATCCATGGCGGTAACTGTCTTCCT-3V, 22 cycles; and XHR1 (3V untranslated region [UTR]), forward 5VCTTTGAAGGGTGCAAGA-3V and reverse 5V-GAGGGACATGGAAGTCT-3V, 30 cycles. The XHR1-3VUTR primer set was used to detect the mRNA of endogenous XHR1 but not that of injected XHR1 constructs.

Injection of mRNA and antisense morpholino oligos

Results Identification of candidate target genes of XHR1 using a differential cDNA macroarray To identify downstream target genes of the transcriptional repressor protein XHR1 by differential cDNA

H. Takada et al. / Developmental Biology 283 (2005) 253 – 267

macroarray screening, a dominant-negative transcriptional activator of XHR1 was used to upregulate XHR1 target genes that are otherwise repressed by endogenous XHR1. This was required to avoid the effects of XHR1-VP16 on the mesoderm, because we have observed in previous experiments that XHR1-VP16 can inhibit gastrulation when ectopically expressed in the mesoderm. For this purpose, and in order to minimize experimental variation, we generated a hormone-inducible XHR1-VP16, designated XHR1-VP16-GR, in which XHR1-VP16 is fused to GR (Kolm and Sive, 1995). We first tested whether XHR1-VP16-GR acts as a dominant-negative construct in the presence of DEX, as has been shown for XHR1-VP16 (Shinga et al., 2001). XHR1-VP16-GR mRNA, together with nh-gal mRNA as a tracer, was injected into the dorsal left blastomere of four-cell-stage Xenopus embryos. Early neurula embryos with nh-gal staining in the pre-MHB region were selected for the analysis of the expression levels of the MHB genes Pax2 and En2 by WISH, comparing the injected side (left) and the non-injected side (right). As shown in Fig. 1A, there was no detectable difference in the expression levels of Pax2 (Fig. 1A, left) or En2 (not shown) between the left and right sides without DEX. When injected embryos were treated with DEX at the gastrula stage (stages 10.5 –11), a remarkable reduction in Pax2 (Fig. 1A, right) and En2 (data not shown) expression was observed on the injected side. These data suggest that XHR1-VP16-GR functions as a dominant-negative XHR1 to reduce MHB marker gene expression in a DEX-dependent manner, consistent with our previous observation that XHR1-VP16 abolishes MHB marker expression. To minimize the undesired effects of activated XHR1-VP16-GR on earlier development, such as gastrulation arrest, which was observed with high frequency when DEX was added at the 32-cell stage (stage 6), and to ensure that the effect of XHR1 on the reduction of Pax2 expression was sufficient in DEX-treated embryos, we added DEX at the early- to mid-gastrula stages (stage 10.5– 11) (Fig. 1B). We assumed that some XHR1 target genes might be expressed in a relatively small area around the pre-MHB region, or that some other target genes might also be expressed abundantly in another region separate from the pre-MHB region. To maximize the difference in mRNA levels of target genes in embryos expressing XHR1-VP16GR with DEX (referred to as DEX+) compared with DEXuntreated (referred to as DEX ) or uninjected control embryos, we dissected out a small piece of anterior neuroectodermal tissue containing the pre-MHB region (Fig. 1C). To select embryos that expressed XHR1-VP16GR in the pre-MHB region, and also allow us to dissect out only the region that expressed XHR1-VP16-GR, GFP mRNA was coinjected as a tracer. As shown in Figs. 1D and 1E, WISH and semi-quantitative RT-PCR verified the dissection of the pre-MHB region: the pre-MHB pieces predominantly contained XHR1-expressing regions (Fig.

257

1D) and had more concentrated endogenous XHR1 mRNA than the carcass (Fig. 1E). To identify genes that were enriched in the DEX+ samples compared with the DEX and uninjected samples, we screened Xenopus cDNA macroarray membranes containing 48,000 clones by hybridization with labeled cDNA probes prepared from total RNA extracted from the dissected DEX+, DEX , or uninjected samples. The intensity of each clone with ‘‘DEX+’’, ‘‘DEX ’’, or ‘‘uninjected’’ probe was measured, and plotted as DEX+ versus DEX for a single set of experiments, and as DEX+ versus uninjected for two individual sets of experiments (Fig. 2). These data were processed separately, and those clones that displayed the same trends in the DEX+/DEX data sets and DEX+/uninjected data sets were selected for further analysis (Table 1). In screening for upregulated genes, clones were selected that showed > 20% upregulation in DEX+ versus DEX data and > 40% upregulation in DEX+ versus uninjected data, and had an intensity of more than 3.0 arbitrary units in DEX+. Because most clones in our macroarray have ‘‘sibling’’ clones that belong to the same contig based on their 5Vand 3V-EST sequences, we also examined the reproducibility of the results among those sibling clones (Table 1). In screening for downregulated genes, clones with a DEX+/DEX ratio lower than 1, a DEX+/uninjected ratio lower than 0.6, and an intensity higher than 7.0 arbitrary units in uninjected samples were selected, and the reproducibility of the results for clones in each contig was examined (Table 2). Based on our criteria, we chose the following 14 clones as upregulated genes to be analyzed further (described in terms of homologous genes): (1) human elongation of very long fatty acids-like 1 (hElovl1); (2) E(spl)-related (ESR) 7 (ESR7, the same as ESR3); (3) inhibitor of differentiation 2 (Id2); (4) ESR1; (5) ESR2; (6) midkine; (7) X. laevis cleavage 2 (XLCL2); (8) ESR9; (9) no homolog; (10) ESR1; (11) no homolog; (12) Xhox7.1/Msx1; (13) X. laevis

Fig. 2. Results of macroarray. Representative data for 9600 clones from a stage 15 library are shown. Intensity of each clone was plotted for DEX+ versus uninjected (A), and DEX+ versus DEX (B). Most dots aligned well along or in parallel with the y = x line, indicating that large numbers of clones exhibit no differential expression.

ESR1

XLCL2 ESR9b

ESR1b

Homologous gene names are cited from the EST data base (NIBB). Note that most of the genes on the arrayed membranes had 2 or more spots, forming contigs with each other. In the column of Results of WISH, + indicates that upregulation was assessed in the DEX-treated embryos and indicates that no change in their expression was observed. *This contig contains a gene which is highly related to ESR3/7, and referred to as ESR3/7b.

+ + + + + + + + + +

7 4* 3 2 3 13 2 2 4 5 3 8 6 13 4 2 3 2 2 8 2 2 3 3 3 6 2 2 6 4 3 2 3 6 1 2 2 5 3 3 2 7 909 1648 – 1678 – – 309 1364 1290 1676 – – – – 56% 99% – 86% – – 100% 64% No hit 99% – – – – 76% 100% 100% 83% 99% 89% 100% 97% No hit 100% No hit 87% 99% No hit Elovl1 ESR3/7

1 2 3 4 5 6 7 8 9 10 11 12 13 14

Elovl1 (human) ESR3/7 Id2 ESR1 ESR2 midkine XLCL2 ESR9 Unknown ESR1 Unknown Xhox7.1/Msx1 Xseb-4 Unknown

NM_022821 AF146088 AB019520 AF383157 AF383158 U06048 Z14122 AJ009282 – AF383157 – X58773 AF2223427 –

e 119 2e 75 e 71 3e 62 2e 28 5e 66 e 57 e 152 N/A 2e 85 N/A e 136 0.0 N/A

DEX Uninj. vs. DEX+

Identity based on ESTs Accession no. Homologous gene Identified gene Clone no.

Table 1 Summary of the selected clones upregulated by XHR1-VP16-GR

E-value

Identity based on the entire sequence

The length of clone (bp)

No. of upregulated clones

vs. DEX+

+ +

Results of WISH

H. Takada et al. / Developmental Biology 283 (2005) 253 – 267

Total no. of sibling clones

258

RNA recognition motif-containing protein seb-4 (Xseb-4); and (14) no homolog (Table 1). Notably, five of the 14 clones are members of the ESR/Hes family to which XHR1 belongs. We also chose four clones to represent the downregulated genes (described in terms of homologous genes): (15) no homolog; (16) Zic3; (17) repulsive guidance molecule A (RGM); and (18) Zic1. Expression analysis of candidate XHR1 target genes in whole embryos To assess whether the selected genes are actually upregulated in the region where XHR1-VP16-GR is expressed in a DEX-dependent manner, and to examine the expression patterns of these genes, we performed WISH analysis with embryos that had been injected with XHR1-VP16-GR and nh-gal mRNAs on the left side, then treated with DEX (hereafter referred to as XHR1-VP16-GR/+DEX). Twelve of fourteen ‘‘upregulated’’ clones showed enhanced expression in nh-galpositive regions in early neurula embryos, including the pre-MHB region (Table 1). All of the four ‘‘downregulated’’ clones showed a reduction in their expression in nh-gal-positive regions induced by XHR1-VP16-GR/ +DEX (Table 2). These data demonstrate the successful detection of mRNA levels in injected embryos using our macroarray procedures with dissected pre-MHB tissue fragments. We also examined the normal expression patterns of those clones by observing the right sides of embryos, where mRNA had not been injected. Because XHR1 is thought to act as a transcriptional repressor, the expression of its target genes should be repressed in the MHB region where XHR1 is expressed, and normal gene expression levels in the neural plate should be reasonably high. Therefore, we omitted clones 1, 3, 6, 12, and 13 based on their expression patterns, and clone 5 (ESR2) because of its very low level of expression (data not shown). For further analysis, we chose clones 2, 4, 7, 8, and 10, and determined their entire cDNA sequences to identify the following genes: clone 2, the same as ESR3/ 7 and a possible pseudoallele of ESR3/7, referred to ESR3/7b (accession no. AB211545; see Table 1); clone 4, a possible pseudoallele of ESR1, referred to as ESR1b (accession no. AB211546); clone 7, the same as XLCL2; clone 8, a possible pseudoallele of ESR9, referred to as ESR9b (accession no. AB211547); and clone 10, the same as ESR1. As reported previously, ESR1 is expressed in an arc shape in the anterior neural plate, and also in three longitudinal stripes on each side of the posterior neural plate, with a gap between the anterior and posterior regions of expression (Schneider et al., 2001). ESR1b, ESR3/7b, and ESR9b showed almost the same expression pattern as ESR1 (Deblandre et al., 1999) (see Figs. 3 and 4; ESR3/ 7b, data not shown). We assume that ESR1b, ESR3/7b,

H. Takada et al. / Developmental Biology 283 (2005) 253 – 267

259

Table 2 Summary of the selected clones downregulated by XHR1-VP16-GR Clone no.

Homologous gene

Accession no.

Identity based on ESTs

E-value

No. of downregulated clones Uninj. vs. DEX+

DEX

15 16 17

None Zic3 Repulsive guidance molecule A Zic1

– AB005292 BC045008

No hit 91% 81%

N/A 0.0 3e 82

9 3 3

AF022927

81%

0.0

4

18

Total no. of sibling clones

Results of WISH

8 2 2

11 3 3

+ + +

4

6

+

vs. DEX+

In the column of Results of WISH, + indicates that downregulation was assessed in the DEX-treated embryos.

and ESR9b are basically the same as ESR1, ESR3/7, and ESR9, respectively, and, therefore, we used the ESR1b, ESR3/7b, and ESR9b clones in the following studies, represented as ESR1, ESR3, and ESR9 hereafter just for simplicity. XLCL2 was first identified as a gene that is expressed in the cleavage stages (Paris and Philippe, 1990), but its developmental expression and function have not yet been reported in Xenopus. Interestingly, we found that XLCL2 is specifically expressed in the anterior neural plate (Fig. 3). To assess whether these candidate target genes are expressed in the region in which XHR1 is expressed, we used twocolor WISH with either the ESR1 or XLCL2 probe together with the XHR1 probe. As shown in Fig. 3, XHR1 expression (turquoise) just fits into the gap between the arc-shaped and striped regions of ESR1 expression (purple), and appears to border the posterior edge of the region of XLCL2 expression (purple) at the early neurula stage. These complementary expression patterns supports the idea that these ESR genes and XLCL2 are target genes repressed by XHR1, and are therefore not expressed in the region where XHR1 is expressed. Direct activation by XHR1-VP16-GR We next tested whether ESR1, ESR3, ESR9, and XLCL2 are direct target genes of XHR1. If a gene is directly activated by XHR1-VP16-GR, this activation does not require protein synthesis. This indicates that the basic

Fig. 3. The results of two-color WISH for ESR1 (left column, purple) or XLCL2 (right column, purple) and XHR1 (turquoise). Arrows indicate XHR1 expression in the pre-MHB region.

DNA-binding domain of XHR1 can bind to the enhancer of the gene, regulating its expression. In this experiment, we also tested whether or not Xngnr1, Xdelta1, and XHR1 are directly activated by XHR1-VP16-GR, because the preMHB region is known to be a neuron-free zone (Geling et al., 2003; Hirata et al., 2001), and because XHR1 might be autoregulated. We treated injected embryos at the gastrula stage (stage 10.5 –11) with a protein synthesis inhibitor, cycloheximide (CHX), 30 min before DEX treatment (referred to as XHR1-VP16-GR/+CHX/+DEX). This CHX treatment condition has been shown to inhibit protein synthesis-dependent indirect induction by a hormoneinducible transcription factor or activin treatment (Yamamoto et al., 2003). Curiously, the treatment of embryos with CHX by itself completely abolished ESR1 expression, whereas it rather upregulated Xngnr1 in the interstripe regions and XHR1 (Figs. 4Aa,e,f, third panels from the left), and distorted the ESR9, XLCL2, and Xdelta1 expression pattern (Figs. 4Ab,c,d, third panels). We speculate that some positive or negative regulators may turn over rapidly and therefore disappear quickly in the presence of CHX. Although CHX treatment itself disturbs expression patterns of these genes, DEX treatment in the presence of CHX resulted in an extensive increase in the expression of ESR1, ESR9, XLCL2, Xdelta1, Xngnr1, and XHR1 (Fig. 4A, fourth panels) as well as ESR3 (data not shown) in the nh-galpositive area, indicating that XHR1-VP16-GR directly activated these genes. Consistent with this observation, in the pre-MHB region of wild-type embryos, Xdelta1 is expressed at lower levels and Xngnr1 expression is not detected (Chitnis et al., 1995; Ma et al., 1996) (indicated by arrows in Figs. 4Ad,e, first panel). These data suggest that ESR1, ESR3, ESR9, XLCL2, Xngnr1, and Xdelta1 are direct target genes of XHR1, and that XHR1 is autoregulated. Activation of Xngnr1 by XHR1-VP16-GR/+DEX raised the question of whether or not repression of MHB genes is mediated by premature neurogenesis in the MHB region, as has been suggested in mice and zebrafish (Geling et al., 2003; Hirata et al., 2001). We tested this possibility using the proneural gene Xash3 (Zimmerman et al., 1993), the motor neuron marker Xlim3 (Taira et al., 1993), and the neuronal differentiation marker b2-tubulin (Good et al., 1989). As shown in Fig. 4B, none of them were strongly upregulated by XHR1-VP16-GR/+DEX, though Xlim3 was slightly activated (Fig. 4Ba) similar to Xngnr1 (Fig. 4Ae),

260

H. Takada et al. / Developmental Biology 283 (2005) 253 – 267

H. Takada et al. / Developmental Biology 283 (2005) 253 – 267

and Xash3 was rather downregulated (Fig. 4Bb), suggesting that reduction of MHB gene expression by XHR1-VP16GR/+DEX is not only due to premature neurogenesis in the pre-MHB region. XHR1 is necessary and sufficient for the repression of ESR genes in the MHB region To assess whether the wild-type XHR1 is sufficient for the repression of the ESR genes, we performed misexpression experiments in regions where the ESR genes are normally expressed. We injected wild-type XHR1 mRNA together with nh-gal mRNA into the dorsal left blastomere at the four-cell stage and examined the expression of ESR1, ESR3, and ESR9 by WISH. The endogenous expression of ESR1 (Fig. 5Aa), ESR3, or ESR9 (data not shown) was abolished in the nh-gal-positive region, indicating that XHR1 can act as a repressor of these ESR genes. We also observed that the misexpression of XHR1 greatly reduced the expression of XLCL2 (Fig. 5Ab) as well as that of Xngnr1 and Xdelta1 (data not shown). To determine whether XHR1 is required for the repression of the candidate target genes in the pre-MHB, we performed loss-of-function analysis with antisense morpholino oligos directed against XHR1 mRNA (XHR1MOs). Embryos were injected with XHR1-MOs together with nh-gal mRNA in the dorsal left blastomere and were subjected to WISH to examine expression patterns. As shown in Figs. 5Ba,aV, injection of XHR1-MOs upregulated ESR1 expression in the MHB region at the neurula stage, notably filling the gap between the anterior arc-shaped region of expression and the medial stripe of posterior expression, and extending the intermediate stripe anteriorly. We next tested the specificity of XHR1-MOs. As shown in Fig. 5Bb, the injection of XHR1-5mmMO did not lead to the upregulation of ESR1 in the pre-MHB region, suggesting that XHR1-MOs act specifically when the sequence of the MO matches the target sequence perfectly. To perform rescue experiments, embryos were injected with XHR1MOs into the dorsal left blastomere at the four-cell stage, incubated until the eight-cell stage, then injected with different amounts of XHR1 mRNA that did not contain the target sequences of XHR1-MOs together with nh-gal mRNA into the dorsal left blastomere at almost the same point corresponding to the site of injection for MOs. Injection of XHR1 mRNA, but not of globin mRNA, reversed the effects XHR1-MOs on ESR1 expression (Figs. 5Bc,d). The percentage of embryos showing ectopic ESR1

261

expression in the pre-MHB region was 36% (8/22; i.e., 1 minus 14/22) for XHR1-MOs + XHR1 mRNA, which is lower than the 81% (30/37) for XHR1-MOs and the 88% (14/16) for XHR1-MOs + globin mRNA, indicating that XHR1-MOs elicited the upregulation of ESR1 in the preMHB by specifically inhibiting the endogenous XHR1 function. Similar to ESR1, ESR9 was upregulated in the preMHB region by XHR1-MOs (Figs. 5Be,eV), while ESR3 was weakly upregulated because its normal expression level is very low (data not shown). These data suggest that XHR1 is necessary and sufficient to repress the expression of the ESR genes in the pre-MHB region. XHR1-MOs affect the expression of Xdelta1 and Pax2 XHR1-MOs distorted the Xdelta1 expression to narrow the gap at the MHB (Fig. 5Bg), but failed to upregulate the expression of XLCL2 or Xngnr1 in the pre-MHB region in contrast to the ESR genes (see Discussion). We further examined the effects of XHR1-MOs on the expression of the MHB gene Pax2, which is thought to be a positive regulator of pre-MHB development and an indirect target of XHR1 (Shinga et al., 2001). The injection of XHR1-MOs resulted in the downregulation of Pax2 expression in the early neurula, suggesting that XHR1 is required for the proper early establishment of the pre-MHB. The reduction in Pax2 expression, however, was not complete in the majority of the injected embryos (Fig. 5Bi). This partial reduction in Pax2 expression by XHR1-MOs showed tendency to be rescued by injection of XHR1 mRNA but not by injection of globin mRNA (Figs. 5Bk,l). The relatively weak effect of loss-of-function of XHR1 induced by XHR1-MOs on Pax2 expression is consistent with the previous observation that injection of the derepression-type dominant-negative constructs of XHR1, mb-XHR1 (in which the basic DNAbinding region was point-mutated) and XHR1-DWRPW (in which the C-terminal WRPW repression domain was deleted) downregulated Pax2, but not as strongly as the activator-type dominant-negative XHR1, XHR1-VP16 (Shinga et al., 2001). As in the case of mb-XHR1 and XHR1-DWRPW, the downregulation of Pax2 and ESR1 by XHR1-MOs was no longer detected at the late neurula or tailbud stage (data not shown; see Discussion). Misexpression of ESR1 disrupts MHB formation The results described above suggest that XHR1 directly represses ESR1, at least in the medial and intermediate parts

Fig. 4. The effects of XHR1-VP16-GR on the expression of possible target genes and neuronal markers. (A) ESR1, ESR9, XLCL2, Xdelta1, Xngnr1, and XHR1 are activated by XHR1-VP16-GR in the presence of CHX. Embryos were injected into the dorsal left blastomere at the four-cell stage with XHR1-VP16-GR mRNA (100 pg), and treated with or without CHX for 30 min before the addition (or not) of DEX at the gastrula stage (stage 11), as indicated. The expression of ESR1 (a), ESR9 (b), XLCL2 (c), Xdelta1 (d), Xngnr1 (e), and XHR1 (f) were analyzed at stage13 by WISH. (B) The effects of XHR1-VP16-GR on the expression of neuronal markers. The expression of Xlim3 (a), Xash3 (b), and b2-tubulin (c) was analyzed at stage 13 by WISH. Fractions indicate the proportion of the presented phenotype per total number: numbers in black, normal expression; numbers in red, upregulation; numbers in green, disrupted expression; numbers in blue, downregulation. Black arrows in Ad,e, pre-MHB region; red arrow in Ba, ectopic expression; black arrow in Bb, reduced expression.

262

H. Takada et al. / Developmental Biology 283 (2005) 253 – 267

Fig. 5. The effects of gain-of-function or loss-of-function of XHR1 on expression gene. (A) Misexpression of XHR1 represses ESR1 and XLCL2 expression. XHR1 mRNA (12.5 – 25 pg) was injected into the dorsal left blastomere at the four-cell stage. ESR1 (a) and XLCL2 (b) expression was examined by WISH at stage 13 – 14. Each stripe in which primary neurons arise is numbered in the expression of ESR1. 1, Medial stripe; 2, intermediate stripe; 3, lateral stripe. Arrows, reduced expression. (B) Effects of XHR1-MOs on the expression of ESR1 (a – d), ESR9 (e, f), Xdelta1 (g, h), and Pax2 (i – l). XHR1-MOs (a, e, g, i) or XHR1-5mmMO (b, f, h, j) was injected at the four-cell stage, and globin (c, k; negative control) or XHR1 (d, l) mRNA was injected with nh-gal mRNA at the eight-cell stage. (aV, eV) Magnification of the same embryo as shown in panels a and e. Black arrows, reduced expression; red arrows, ectopic expression. Fractions in black, normal expression; in blue, downregulation; in red, upregulation. Injected doses: MOs, 6 or 3 (i, j) ng; globin or XHR1 mRNA, 3 – 10 (d), or 0.3 – 1 (l) pg.

of the pre-MHB region. To examine whether the repression of ESR1 by XHR1 is required for the proper development of the MHB, we misexpressed ESR1 in the pre-MHB region and examined its effects on MHB gene expression. As shown in Fig. 6, ESR1 as well as ESR3 and ESR9 (data not shown) strongly reduced Pax2 expression, suggesting that the repression of ESR1 by XHR1 is necessary for the proper expression of MHB genes. Furthermore, misexpression of ESR1 also downregulated XHR1 expression. Because XLCL2/Lbh has been suggested to function as a coactivator (Briegel and Joyner, 2001) and its expression is restricted to the anterior neural plate, this protein might have some roles in neural patterning, and perhaps disturbs MHB

formation. Therefore, we misexpressed XLCL2 in the preMHB region and assayed for MHB gene expression. However, neither XHR1 nor Pax2 expression was affected by XLCL2 (data not shown), implying that a transcription factor(s) functioning with XLCL2, if any, might be absent in the pre-MHB region. Notch signaling does not activate XHR1 expression Many bHLH-WRPW transcriptional repressors, including ESR1, ESR3, and perhaps ESR9 are activated by Notch signaling and play a role in the maintenance of neural stem cell and progenitor cell fates (Davis and Turner, 2001).

H. Takada et al. / Developmental Biology 283 (2005) 253 – 267

263

Target genes of XHR1 in the pre-MHB region

Fig. 6. ESR1 downregulates Pax2 and XHR1 expression. ESR1 mRNA was injected at 100 – 400 pg per blastomere. Pax2 and XHR1 expression was analyzed at stage 13. Arrows, reduced expression. Fractions in black, normal expression; in blue, downregulation.

Therefore, it is of interest to examine whether the induction of XHR1 is dependent on the Notch signaling pathway. To address this question, we analyzed the expression of XHR1 under conditions in which Notch signaling is activated by the overexpression of a constitutively active Notch, Notch ICD, or is repressed by a dominant-negative mutant of Su(H)1, designated Su(H)1DBM, which does not bind DNA but still associates with Notch ICD (Wettstein et al., 1997). As shown in Fig. 7, the overexpression of ICD did not lead to the upregulation of XHR1, but rather to a dramatic reduction in XHR1 expression without affecting the expression of a pan-neural gene, SOX2, whereas it upregulated ESR1 expression as reported previously (Wettstein et al., 1997). When Notch signaling was inhibited by the overexpression of Su(H)1DBM, ESR1 expression was downregulated as expected (Wettstein et al., 1997), whereas XHR1 expression was unaffected (data not shown). These data suggest that XHR1 is not a direct target of Notch-ICD-Su(H) signaling.

The following lines of evidence suggest that ESR1, -3, and -9 are direct targets of XHR1: (1) XHR1-VP16-GR upregulated the ESR genes in the presence of DEX even after treatment with CHX (designated XHR1-VP16-GR/ +CHX/+DEX) (Fig. 4A); (2) overexpression of wild-type XHR1 led to the downregulation of the ESR genes (Fig. 5A); (3) knockdown experiments using XHR1-MOs resulted in the ectopic expression of the ESR genes only in the pre-MHB region, which occurs in the gap between the medial stripe and arc-shaped regions of ESR expression (Fig. 5B); and (4) the effect of XHR1-MOs seemed to be specific because coinjection of XHR1 mRNA rescued, though partially, the XHR1-MOs phenotype (Fig. 5B; only assayed for ESR1). Therefore, of the 14 genes that we initially selected as possible target genes for XHR1, these ESR genes are most likely to be direct targets of XHR1. Although there is no clear evidence that XHR1 is required for the repression of XLCL2, Xngnr1, and Xdelta1 in the pre-MHB region, as assayed by MO injection, we still think that those genes are also good candidates for direct targets of XHR1 based on their responsiveness to XHR1VP16-GR as XLCL2, Xngnr1, and Xdelta1 were strongly upregulated by XHR1-VP16-GR/+CHX/+DEX, similar to ESR1 (Fig. 4A), and on their expression patterns complementary to that of XHR1 (Fig. 3; see Figs. 4Ad,e). Given that XLCL2, Xngnr1, and Xdelta1 are direct targets of XHR1, the question remains: why are XLCL2, Xngnr1, and Xdelta1 not or less upregulated by microinjection with

Discussion In this study, we undertook large-scale screening for genes regulated by XHR1 in order to decipher the molecular mechanisms of early neural patterning and pre-MHB specification. Our data strongly suggest that bHLH-WRPW transcriptional repressor genes, ESR1, ESR9, and perhaps ESR3 are direct target genes of XHR1, and that XLCL2, Xdelta1, Xngnr1, and XHR1 can also be direct targets for XHR1. As discussed below, we propose that XHR1 demarcates the pre-MHB region from the surrounding neuroectoderm as a prepattern gene, which was originally defined in Drosophila notal development as a gene determining two-dimensional pattern formation in the epithelium (Calleja et al., 2002).

Fig. 7. Notch signaling does not activate XHR1 expression. Notch-ICD mRNA was injected at 250 – 1000 pg per blastomere. The expression of ESR1, XHR1, and Sox2 was analyzed at stage 13. Arrow, reduced expression. Fractions in white, normal expression; in blue, downregulation; in red, upregulation.

264

H. Takada et al. / Developmental Biology 283 (2005) 253 – 267

XHR1-MOs, in contrast to ESR1? We conjecture that there might be other repressors that are redundant for XHR1 function in repressing XLCL2, Xngnr1, and Xdelta1 in the pre-MHB region, or that there might be differences in the sensitivity of ESR1 and those three genes to the level of XHR1 protein. For example, XLCL2, Xngnr1, and Xdelta1 may be repressed by lesser amounts of XHR1 than is ESR1 and ESR9, because the translation of XHR1 mRNA is probably not completely repressed by MOs. In the CHX experiments (Fig. 4A), we have noticed that Xngnr1 and Xdelta1 are not strongly upregulated by XHR1-VP16-GR when treated with DEX alone (XHR1VP16-GR/+DEX). Because the E(spl) family members downstream from Notch signaling are thought to repress neuronal differentiation (Cau et al., 2000; Schneider et al., 2001; Takke et al., 1999), one possible explanation for this result is that XHR1-VP16-GR/+DEX strongly upregulates E(spl)-related genes, the products of which in turn inhibit the upregulation of Xngnr1 and Xdelta1 by XHR1VP16-GR/+DEX. Those repressors are likely to be ESR genes which are activated by XHR1-VP16-GR/+DEX (Table 1).

In contrast to ESR1, ectopic expression of XLCL2 (the Xenopus ortholog of mouse Lbh; Briegel and Joyner, 2001) could not repress MHB gene expression; however, we found that anterior neural marker Otx2 was ectopically expressed in XLCL2 injected area (data not shown), indicating that XLCL2 play some roles in AP patterning in neural plate. It may be possible that XHR1 defines the posterior edge of the midbrain region by repressing XLCL2 expression. Interactions between XHR1 and Notch In the neural plate of the Xenopus embryo, Notch is expressed in a wide area (Chitnis et al., 1995), including the pre-MHB region (Shinga et al., unpublished). One of the Notch ligand genes, Xdelta1, is expressed in a more discrete pattern, with higher levels of expression along the midline and in two other regions on the lateral surfaces, and lower expression in the pre-MHB region (Figs. 4Ad and 5Bh). These observations indicate that the pre-MHB is not completely devoid of Notch and Xdelta1 expression; in other words, the pre-MHB could be competent for ESR1 activation (Fig. 8Ba). This point of view explains the

Fig. 8. Models of the pre-MHB formation by XHR1. (A) Possible gene interactions. XHR1 demarcates the pre-MHB region by repressing ESR1, ESR3, ESR9, Xngnr1, Xdelta1, XLCL2, and yet identified genes (?). XHR1 is likely to repress own transcription. (B) Prepatterns in the neural plate. Combination of prepatterns by Notch and Xdelta1 (a; expression patterns are simplified) and XHR1 (b) specifies expression patterns of ESR genes (c) and Pax2 (d). See Discussion for more detail.

H. Takada et al. / Developmental Biology 283 (2005) 253 – 267

ectopic expression of ESR1 in the pre-MHB region as stripes when XHR1 is knocked down. Our data have suggested possible regulatory interactions which are summarized in Fig. 8A. In this model, XHR1 downregulates ESR1, Xngnr1, and Xdelta1 in the pre-MHB region to inhibit both upstream and downstream from Notch signaling. In addition, our data have suggested that XHR1 is not activated by Notch signaling (Fig. 7), and that ESR1 as well as XHR1 itself can inhibit XHR1 expression (Fig. 6; data not shown). This mutual inhibition possibly creates the boundary between the XHR1-expressing pre-MHB region and the adjacent proneural region in which Delta-Notch signaling is operated. Negative autoregulation of XHR1 may be to maintain a proper level of XHR1 expression as has been suggested in a model system (Becskei and Serrano, 2000). Mechanisms of pre-MHB specification and regionalization involving XHR1 function In the chick, mouse, and zebrafish, a repressive interaction between Otx2 and Gbx2 demarcates the MHB region, and is required for the maintenance but not for the induction of MHB gene expression (Broccoli et al., 1999; Joyner et al., 2000; Katahira et al., 2000; Li and Joyner, 2001; Millet et al., 1999; Simeone, 2000). We have previously shown that XHR1 is expressed in the pre-MHB region before the Otx2 – Gbx2 border is created. Moreover, the boundary between Otx2 and Gbx2 is first detected on the lateral side of the neural plate at the late gastrula stage (stage 12) (Glavic et al., 2002), where Pax2 expression is first detected, and then converges to the midline. In contrast, the expression of XHR1 starts in the center of the dorsal neuroectoderm at the early gastrula stage and its expression in the midline gradually vanishes (Shinga et al., 2001). These observations suggest the existence of at least two separate pathways for the development of the pre-MHB. This is consistent with the existence of multiple genetic cascades for MHB formation, as has been suggested for zebrafish (Tallafuss et al., 2001). We have shown that Pax2 is downregulated by XHR1VP16-GR/+DEX (Fig. 1A), but this is probably not merely due to premature neuronal differentiation as assayed by Xngnr1, Xlim3, Xash3, and b2-tubulin expression (Figs. 4Ae, 4B). Consistent with the results of XHR1-VP16-GR experiments, loss-of-function by XHR1-MOs does not upregulate Xngnr1 and b2-tubulin, but downregulates Pax2 in the pre-MHB region (Fig. 5). Because loss-offunction of mouse Hes1 and Hes3 or zebrafish her5 disrupts MHB formation through premature neurogenesis (Geling et al., 2003; Hirata et al., 2001), it is also possible that bHLHWRPW repressors other than XHR1 inhibit neurogenesis in the pre-MHB region. The reduction of Pax2 expression by XHR1-MOs is thought to be mediated, at least in part, by the ectopic expression of ESR genes, because ESR genes misexpression

265

inhibits Pax2 expression (Fig. 6). However, the reduction in Pax2 expression in the more lateral regions of the pre-MHB, where the ESR genes are not ectopically expressed, could be mediated by other factors yet to be identified. Therefore, XHR1 could act as an upstream component of Pax2 expression (Fig. 8A). Pax2 is thought to act as a positive regulator that maintains the expression of its own gene, as well as that of other MHB genes (Picker et al., 2002; Wurst and Bally-Cuif, 2001), which may include XHR1. In zebrafish, loss-of-function of a homolog of XHR1, her5, by antisense MOs upregulates neurogenin 1 and the Cdk inhibitor p27 xic1 in the pre-MHB region, but does not affect the expression of the MHB gene pax2.1 (Geling et al., 2003). It has also been shown that Her5 directly binds to CATGTG (Class B site; Fisher and Caudy, 1998) in a regulatory element of the neurogenin 1 gene to repress its expression (Geling et al., 2004). To investigate whether her5 is a functional equivalent of XHR1 in the MHB formation, it will be needed to examine whether XHR1 directly inhibits p27 Xic1 and whether her5 directly inhibits the orthologs of ESR1, ESR3, ESR9, and Xdelta1. Function of XHR1 as a prepattern gene The neural plate becomes regionalized along the anteroposterior and mediolateral axes during development. This two-dimensional patterning provides us a model system to study prepattern genes, as has been shown in Drosophila notal development, in which many transcription factors, such as hairy, iroquois, and pannier, are involved. These genes define not only the region where neurons arise but also notal morphology such as muscle attachment sites and pigmentation patterning (Calleja et al., 2002). In vertebrates, it has been shown that Xrx1, Zic2, Xiro3 (in Xenopus), her5, her3 (in zebrafish), and perhaps Hes3 (in mouse) function as prepattern genes for neurogenesis in the neural plate to maintain undifferentiated states by inhibiting the proneural gene neurogenin, and the Notch ligand gene delta, and to promoter cell proliferation by inhibiting the CDK inhibitor p27 Xic1 (Bally-Cuif and Hammerschmidt, 2003; Brewster et al., 1998; Hans et al., 2004; Hirata et al., 2001). Other than neurogenesis, Tcf4 and Irx2 have been shown to function as prepattern genes for muscle patterning in the limb bud and cerebellum development in the neural plate, respectively (Kardon et al., 2003; Matsumoto et al., 2004). These prepattern genes start to be expressed before the morphogenesis starts, and combinations of gene expression define subregions called prepatterns. According to our data, we propose that XHR1 functions as a prepattern gene for the pre-MHB region as follows. First, XHR1 starts to be expressed before the expression of ESR1, ESR9, and MHB genes start (Shinga et al., 2001). Second, our data with XHR1-MOs revealed that the preMHB region has a potential for striped expression of ESR1 and ESR9 (Figs. 5Ba,e). This potentiality is probably

266

H. Takada et al. / Developmental Biology 283 (2005) 253 – 267

conferred by the expression of Notch and Xdelta1 as continuous stripes as shown in Fig. 8Ba. This prepattern overlaps with that of XHR1 expression as shown in Fig. 8Bb to bring about the expression pattern of ESR1 and ESR9 (Fig. 8Bc). Concomitantly, the MHB gene Pax2 is expressed in the pre-MHB region where XHR1 defines by inhibiting ESR genes and so on (Fig. 8Bd). Thus, XHR1 appears to act as a prepattern gene to execute the initial step of pre-MHB formation as one of the early prepatterns in the vertebrate neuroectoderm, because XHR1 expression starts at the beginning of neural induction and patterning at the early gastrula stage.

Acknowledgments We thank J. Gurdon for pRN3-eGFP; D. Turner, R. Rupp, and J. Lee for pCS2+ and pCS2+nh-gal; R. Harland for Xenopus En2; Y. Sasai for Xenopus Sox2; A. Brandli for Xenopus Pax2; K. Zimmerman for Xash3; P. Good for h2tubulin; and C. Kintner for Xdelta1, Notch-ICD and Su(H)1DBM plasmids. We thank J. Morokuma for helping with the procedures for macroarray screening; M. Itoh, I. Hiratani, and S. Ohmori for their generous technical assistance; M. Inamori and A. Takahashi for the use of the automated WISH system; M. Itoh for the double WISH technique; T. Nohno and Y. Sasai for the double WISH protocols; and Y. Bessho and R. Kageyama for valuable discussions. We also thank the members of Ueno’s laboratory for their kind accommodation of a student (D. H.) and their useful discussions. This work was supported in part by a Grant-in-Aid for Scientific Research from the Ministry of Education, Science, Sports, and Culture of Japan; and by the Toray Science Foundation, Japan.

References Bally-Cuif, L., Hammerschmidt, M., 2003. Induction and patterning of neuronal development, and its connection to cell cycle control. Curr. Opin. Neurobiol. 13, 16 – 25. Becskei, A., Serrano, L., 2000. Engineering stability in gene networks by autoregulation. Nature 405, 590 – 593. Belting, H.G., Hauptmann, G., Meyer, D., Abdelilah-Seyfried, S., Chitnis, A., Eschbach, C., Soll, I., Thisse, C., Thisse, B., Artinger, K.B., Lunde, K., Driever, W., 2001. Spiel ohne grenzen/pou2 is required during establishment of the zebrafish midbrain – hindbrain boundary organizer. Development 128, 4165 – 4176. Brewster, R., Lee, J., Ruiz i Altaba, A., 1998. Gli/Zic factors pattern the neural plate by defining domains of cell differentiation. Nature 393, 579 – 583. Briegel, K.J., Joyner, A.L., 2001. Identification and characterization of Lbh, a novel conserved nuclear protein expressed during early limb and heart development. Dev. Biol. 233, 291 – 304. Broccoli, V., Boncinelli, E., Wurst, W., 1999. The caudal limit of Otx2 expression positions the isthmic organizer. Nature 401, 164 – 168. Calleja, M., Renaud, O., Usui, K., Pistillo, D., Morata, G., Simpson, P., 2002. How to pattern an epithelium: lessons from achaete-scute regulation on the notum of Drosophila. Gene 292, 1 – 12.

Cau, E., Gradwohl, G., Casarosa, S., Kageyama, R., Guillemot, F., 2000. Hes genes regulate sequential stages of neurogenesis in the olfactory epithelium. Development 127, 2323 – 2332. Chitnis, A., Henrique, D., Lewis, J., Ish-Horowicz, D., Kintner, C., 1995. Primary neurogenesis in Xenopus embryos regulated by a homologue of the Drosophila neurogenic gene Delta. Nature 375, 761 – 766. Chomczynski, P., Sacchi, N., 1987. Single-step method of RNA isolation by acid guanidinium thiocyanate – phenol – chloroform extraction. Anal. Biochem. 162, 156 – 159. Davis, R.L., Turner, D.L., 2001. Vertebrate hairy and Enhancer of split related proteins: transcriptional repressors regulating cellular differentiation and embryonic patterning. Oncogene 20, 8342 – 8357. Deblandre, G.A., Wettstein, D.A., Koyano-Nakagawa, N., Kintner, C., 1999. A two-step mechanism generates the spacing pattern of the ciliated cells in the skin of Xenopus embryos. Development 126, 4715 – 4728. De Robertis, E.M., Larrain, J., Oelgeschlager, M., Wessely, O., 2000. The establishment of Spemann’s organizer and patterning of the vertebrate embryo. Nat. Rev., Genet. 1, 171 – 181. Fisher, A., Caudy, M., 1998. The function of hairy-related bHLH repressor proteins in cell fate decisions. BioEssays 20, 298 – 306. Geling, A., Itoh, M., Tallafuss, A., Chapouton, P., Tannhauser, B., Kuwada, J.Y., Chitnis, A.B., Bally-Cuif, L., 2003. bHLH transcription factor Her5 links patterning to regional inhibition of neurogenesis at the midbrain – hindbrain boundary. Development 130, 1591 – 1604. Geling, A., Plessy, C., Rastegar, S., Strahle, U., Bally-Cuif, L., 2004. Her5 acts as a prepattern factor that blocks neurogenin1 and coe2 expression upstream of Notch to inhibit neurogenesis at the midbrain – hindbrain boundary. Development 131, 1993 – 2006. Glavic, A., Gomez-Skarmeta, J.L., Mayor, R., 2002. The homeoprotein Xiro1 is required for midbrain – hindbrain boundary formation. Development 129, 1609 – 1621. Good, P.J., Richter, K., Dawid, I.B., 1989. The sequence of a nervous system-specific, class II beta-tubulin gene from Xenopus laevis. Nucleic Acids Res. 17, 8000. Hans, S., Scheer, N., Riedl, I., Weizsacker Ev, E., Blader, P., CamposOrtega, J.A., 2004. her3, a zebrafish member of the hairy-E(spl) family, is repressed by Notch signalling. Development 131, 2957 – 2969. Harland, R.M., 1991. In situ hybridization: an improved whole-mount method for Xenopus embryos. In: Kay, B.K., Peng, H.B. (Eds.), Methods in Cell Biology, vol. 36. Academic Press, San Diego, CA, pp. 685 – 695. Harland, R., Gerhart, J., 1997. Formation and function of Spemann’s organizer. Annu. Rev. Cell Dev. Biol. 13, 611 – 667. Hirata, H., Tomita, K., Bessho, Y., Kageyama, R., 2001. Hes1 and Hes3 regulate maintenance of the isthmic organizer and development of the mid/hindbrain. EMBO J. 20, 4454 – 4466. Hiratani, I., Mochizuki, T., Tochimoto, N., Taira, M., 2001. Functional domains of the LIM homeodomain protein Xlim-1 involved in negative regulation, transactivation, and axis formation in Xenopus embryos. Dev. Biol. 229, 456 – 467. Itoh, M., Kudoh, T., Dedekian, M., Kim, C.H., Chitnis, A.B., 2002. A role for iro1 and iro7 in the establishment of an anteroposterior compartment of the ectoderm adjacent to the midbrain – hindbrain boundary. Development 129, 2317 – 2327. Joyner, A.L., Liu, A., Millet, S., 2000. Otx2, Gbx2 and Fgf8 interact to position and maintain a mid – hindbrain organizer. Curr. Opin. Cell Biol. 12, 736 – 741. Kardon, G., Harfe, B.D., Tabin, C.J., 2003. A Tcf4-positive mesodermal population provides a prepattern for vertebrate limb muscle patterning. Dev. Cell 5, 937 – 944. Katahira, T., Sato, T., Sugiyama, S., Okafuji, T., Araki, I., Funahashi, J., Nakamura, H., 2000. Interaction between Otx2 and Gbx2 defines the organizing center for the optic tectum. Mech. Dev. 91, 43 – 52. Kolm, P.J., Sive, H.L., 1995. Efficient hormone-inducible protein function in Xenopus laevis. Dev. Biol. 171, 267 – 272. Krieg, P.A., Melton, D.A., 1984. Functional messenger RNAs are produced by SP6 in vitro transcription of cloned cDNAs. Nucleic Acids Res. 12, 7057 – 7070.

H. Takada et al. / Developmental Biology 283 (2005) 253 – 267 Li, J.Y., Joyner, A.L., 2001. Otx2 and Gbx2 are required for refinement and not induction of mid – hindbrain gene expression. Development 128, 4979 – 4991. Lumsden, A., Krumlauf, R., 1996. Patterning the vertebrate neuraxis. Science 274, 1109 – 1115. Ma, Q., Kintner, C., Anderson, D.J., 1996. Identification of neurogenin, a vertebrate neuronal determination gene. Cell 87, 43 – 52. Martinez, S., 2001. The isthmic organizer and brain regionalization. Int. J. Dev. Biol. 45, 367 – 371. Matsumoto, K., Nishihara, S., Kamimura, M., Shiraishi, T., Otoguro, T., Uehara, M., Maeda, Y., Ogura, K., Lumsden, A., Ogura, T., 2004. The prepattern transcription factor Irx2, a target of the FGF8/MAP kinase cascade, is involved in cerebellum formation. Nat. Neurosci. 7, 605 – 612. Millet, S., Campbell, K., Epstein, D.J., Losos, K., Harris, E., Joyner, A.L., 1999. A role for Gbx2 in repression of Otx2 and positioning the mid/hindbrain organizer. Nature 401, 161 – 164. Mochizuki, T., Karavanov, A.A., Curtiss, P.E., Ault, K.T., Sugimoto, N., Watabe, T., Shiokawa, K., Jamrich, M., Cho, K.W., Dawid, I.B., Taira, M., 2000. Xlim-1 and LIM domain binding protein 1 cooperate with various transcription factors in the regulation of the goosecoid promoter. Dev. Biol. 224, 470 – 485. Nieuwkoop, P.D., Faber, J., 1967. Normal table of Xenopus laevis (Daudin). North Holland, Amsterdam. Osada, S., Ohmori, S.Y., Taira, M., 2003. XMAN1, an inner nuclear membrane protein, antagonizes BMP signaling by interacting with Smad1 in Xenopus embryos. Development 130, 1783 – 1794. Paris, J., Philippe, M., 1990. Poly(A) metabolism and polysomal recruitment of maternal mRNAs during early Xenopus development. Dev. Biol. 140, 221 – 224. Picker, A., Scholpp, S., Bohli, H., Takeda, H., Brand, M., 2002. A novel positive transcriptional feedback loop in midbrain – hindbrain boundary development is revealed through analysis of the zebrafish pax2.1 promoter in transgenic lines. Development 129, 3227 – 3239. Qian, X., Shen, Q., Goderie, S.K., He, W., Capela, A., Davis, A.A., Temple, S., 2000. Timing of CNS cell generation: a programmed sequence of

267

neuron and glial cell production from isolated murine cortical stem cells. Neuron 28, 69 – 80. Rhinn, M., Brand, M., 2001. The midbrain – hindbrain boundary organizer. Curr. Opin. Neurobiol. 11, 34 – 42. Schneider, M.L., Turner, D.L., Vetter, M.L., 2001. Notch signaling can inhibit Xath5 function in the neural plate and developing retina. Mol. Cell. Neurosci. 18, 458 – 472. Shinga, J., Itoh, M., Shiokawa, K., Taira, S., Taira, M., 2001. Early patterning of the prospective midbrain – hindbrain boundary by the HES-related gene XHR1 in Xenopus embryos. Mech. Dev. 109, 225 – 239. Simeone, A., 2000. Positioning the isthmic organizer where Otx2 and Gbx2 meet. Trends Genet. 16, 237 – 240. Taira, M., Hayes, W.P., Otani, H., Dawid, I.B., 1993. Expression of LIM class homeobox gene Xlim-3 in Xenopus development is limited to neural and neuroendocrine tissues. Dev. Biol. 159, 245 – 256. Takke, C., Dornseifer, P., v Weizsacker, E., Campos-Ortega, J.A., 1999. her4, a zebrafish homologue of the Drosophila neurogenic gene E(spl), is a target of NOTCH signalling. Development 126, 1811 – 1821. Tallafuss, A., Wilm, T.P., Crozatier, M., Pfeffer, P., Wassef, M., Bally-Cuif, L., 2001. The zebrafish buttonhead-like factor Bts1 is an early regulator of pax2.1 expression during mid – hindbrain development. Development 128, 4021 – 4034. Wettstein, D.A., Turner, D.L., Kintner, C., 1997. The Xenopus homolog of Drosophila Suppressor of Hairless mediates Notch signaling during primary neurogenesis. Development 124, 693 – 702. Wurst, W., Bally-Cuif, L., 2001. Neural plate patterning: upstream and downstream of the isthmic organizer. Nat. Rev., Neurosci. 2, 99 – 108. Yamamoto, S., Hikasa, H., Ono, H., Taira, M., 2003. Molecular link in the sequential induction of the Spemann organizer: direct activation of the cerberus gene by Xlim-1, Xotx2, Mix.1, and Siamois, immediately downstream from Nodal and Wnt signaling. Dev. Biol. 257, 190 – 204. Zimmerman, K., Shih, J., Bars, J., Collazo, A., Anderson, D.J., 1993. XASH-3, a novel Xenopus achaete – scute homolog, provides an early marker of planar neural induction and position along the mediolateral axis of the neural plate. Development 119, 221 – 232.