Previous studies have identified a minimal 434 bp enhancer from the promoter region of the Ciona Brachyury gene (Ci-Bra), which is sufficient to direct a notochord-specific pattern of gene expression. Here we present evidence that a Ciona homolog of snail (Ci-sna) encodes a repressor of the Ci-Bra enhancer in the tail muscles. DNA-binding assays identified four Ci-Sna-binding sites in the Ci-Bra enhancer, and mutations in these sites cause otherwise normal Ci-Bra/lacZ transgenes to be misexpressed in ectopic tissues, particularly the tail muscles. Selective misexpression of Ci-sna using a heterologous promoter results in the repression of Ci-Bra/lacZ transgenes in the notochord. Moreover, the conversion of the Ci-Sna repressor into an activator results in the ectopic induction of Ci-Bra/lacZ transgenes in the muscles, and also causes an intermixing of notochord and muscle cells during tail morphogenesis. These results suggest that Ci-Sna functions as a boundary repressor, which subdivides the mesoderm into separate notochord and tail muscle lineages.

Brachyury encodes a sequence-specific transcriptional activator that is essential for notochord differentiation in a variety of chordates, including ascidians, zebrafish, Xenopus, chickens and mice (e.g. Hermann and Kispert, 1994; Kispert et al., 1995). In an effort to determine the mechanisms underlying the specification of the notochord, we have conducted a detailed analysis of Brachyury gene regulation in the ascidian, Ciona intestinalis.

Ciona offers a number of advantages for this type of analysis. First, it is possible to obtain efficient expression of transgenic DNA in electroporated embryos (Corbo et al., 1997a). Second, the Ciona Brachyury gene (Ci-Bra) is expressed exclusively in prospective notochord cells, beginning with their clonal restriction at the 64-cell stage of embryogenesis (Yasuo and Satoh, 1994; Corbo et al., 1997a). In contrast, vertebrate Brachyury genes exhibit two distinct expression patterns, including ubiquitous expression throughout the presumptive mesoderm followed by a more restricted pattern in the notochord (e.g. Schulte-Merker and Smith, 1995). Although there has been considerable progress in identifying the cis regulatory elements responsible for the early induction of Xbra2 expression (Latinkic et al., 1997), comparatively little is known about the late, notochord-restricted pattern (e.g. Clements et al., 1996).

The only notochord-specific cis regulatory element known for any chordate Brachyury gene was identified in Ciona (Corbo et al., 1997a). The minimal Ci-Bra enhancer is 434 bp in length and appears to be activated by a combination of Suppressor of Hairless {Su(H)} and basic helix-loop-helix (bHLH) proteins (Corbo, 1997). These factors can activate the Ci-Bra enhancer in all of the mesoderm lineages, including the notochord, tail muscles and trunk mesenchyme. However, one or more repressors appear to restrict the Ci-Bra expression pattern to the notochord. Here, we present evidence that the Ciona snail gene (Ci-sna) encodes a transcriptional repressor, which subdivides the embryonic mesoderm into separate notochord and muscle lineages.

Previous studies have shown that Snail functions as a repressor in the early Drosophila embryo to subdivide the presumptive mesoderm and neurogenic ectoderm (Kosman et al., 1991; Leptin, 1991; Ip et al., 1992). snail is initially expressed in the prospective mesoderm of precellular embryos (Kosman et al., 1991; Leptin, 1991; Alberga et al., 1991). The Snail protein excludes the expression of neuroectodermal specification genes, such as rhomboid (Ip et al., 1992) and single minded (Kasai et al., 1992), in the mesoderm and restricts their expression to the presumptive neurogenic ectoderm. Thus, in Drosophila, mesoderm specification is achieved through a combination of transcriptional activators and repressors, which promote mesoderm differentiation and exclude alternative cell fates, respectively.

snail homologs have been described in a variety of vertebrates, including zebrafish, Xenopus, chickens and mice (e.g. Hammerschmidt and Nüsslein-Volhard, 1993; Ros et al., 1997; Jiang et al., 1997). In zebrafish embryos, sna-1 expression is gradually excluded from the axial mesoderm and restricted to paraxial lineages, which give rise to the somites (Hammerschmidt and Nüsslein-Volhard, 1993). This expression pattern is consistent with the possibility that sna-1 helps establish a boundary between the axial and paraxial mesoderm.

We recently reported the isolation and characterization of a Ciona snail gene, Ci-sna (Corbo et al., 1997b). The gene is expressed in the tail muscles of the tadpole, as well as the trunk mesenchyme, the progenitors of the adult body wall muscles, the cerebral vesicle and the lateral ependymal cells of the spinal cord. This complex Ci-sna expression pattern is complementary to the Ci-Bra pattern. Moreover, truncated Ci-Bra/lacZ transgenes exhibit ectopic expression in most or all of these Ci-sna lineages (Corbo et al., 1997a), thereby raising the possibility that Ci-Sna functions as a repressor of the Ci-Bra expression pattern.

Here we present evidence that Ci-Sna directly represses Ci-Bra expression. Gel shift assays identified four different Ci-Sna protein-binding sites within the 434 bp notochord-specific Ci-Bra enhancer. Mutations in either one of two of these sites, sna1 or sna2, cause an otherwise normal Ci-Bra/lacZ transgene to be derepressed in the tail muscles. Repression is restored in these ectopic lineages upon insertion of a synthetic Snail protein-binding site, s2, from Drosophila. Additional evidence for Ci-Sna/Ci-Bra regulatory interactions was obtained by misexpressing Ci-sna in the notochord. This was achieved by attaching the Ci-sna-coding region to Ci-Bra regulatory sequences. Ectopic expression of Ci-sna results in the down-regulation of Ci-Bra/lacZ transgenes in the notochord. Finally, a Ci-SnaVP16 fusion protein induces ectopic expression of Ci-Bra/lacZ transgenes in the tail muscles and spinal cord. The fusion protein generates a mutant phenotype, whereby notochord cells and tail muscles fail to separate during the intercalary movements that occur during tail morphogenesis. Taken together, these results suggest that Ci-Sna directly represses Ci-Bra expression to subdivide the embryonic mesoderm into dorsal (axial) and lateral (paraxial) lineages.

Animal husbandry

Adult Ciona intestinalis were collected at Half Moon Bay, the Oakland Marina and Bodega Bay in California. Animals were also purchased from the Marine Biological Laboratory collecting facility at Woods Hole, Massachusetts. Some of the animals used in the most recent experiments were kindly provided by Dr Nori Satoh in Japan. Mature eggs and sperm were collected from gravid adults and insemination was done in filtered sea water. Embryos were cultured at 15°C.

Site-directed mutagenesis

XhoI-PstI genomic DNA fragments that contain either the full-length 434 bp Ci-Bra enhancer, or the truncated 251 bp enhancer were used as templates for mutagenesis (Corbo et al., 1997a). Mutagenized forms of the Ci-sna-coding sequence were prepared with a NotI-NotI DNA fragment containing a full-length Ci-sna cDNA (Corbo et al., 1997b). Both DNAs were cloned into the pBluescript II SK+ vector (Stratagene, La Jolla, CA). Site-directed mutagenesis was done essentially as described by Kunkel (1985). Uracil-containing single-stranded DNA was prepared using the CJ236 strain of E. coli and the VCSM13 helper phage (also purchased from Stratagene). The following mutagenic oligonucleotides were used to delete the sna1 and sna2 sites, respectively:

AATAGCGACAAACCTACGTCACAATACAAA (sna1)

TAACACGTCACAATAGTGACGTCATATC (sna2)

The following nucleotides were deleted between the underlined TA in the sna1 region and the underlined AG in the sna2 region:

TATCTGGTGTT (sna1) and CACTTG (sna2)

The following mutagenic oligonucleotides were used to insert the Drosophila s2 Snail-binding site in the truncated 251 bp Ci-Bra enhancer and the mutagenized enhancer lacking the sna2 site, respectively:

TTTTGACATGTCAATCACCTTGCTGGGCAAAATCGGAAAC CA (251 bp–s2)

CACGTCACAATACACCTTGGTGACGTCATAT (sna2–s2)

The following mutagenic oligonucleotides were used to disrupt the central three zinc fingers (ZF) in the C-terminal region of the Ci-sna-coding sequence (Corbo et al., 1997b). The two Cys residues within each ZF were converted into Phe residues (indicated by the underlined residues):

GTAGCAAGTTCAATTTCAAAATATTCAAGAAAGAGTGCAG (ZF 2)

CGCTGCCATGCGAATTCCATATTTTTGGTAAAGCGTTCTC (ZF 3)

AGAAACCATACCAATTCACCGTGTTCTCCAGAGCTTTCGC (ZF 4)

Spacer sequences were inserted between sna1 and the Su(H) sites, and between sna2 and Su(H). This was accomplished by creating synthetic polylinker sequences using the following mutagenic oligonucleotides:

TTGACATGTCAATCAGTCGACGCATGCCCGCGGACTAGTA AATCGGAAACCAAG (sna1)

ATGGGAAAGTAACACGTCGACGCATGCCCGCGGACTAGT GTCACAATACACTTG (sna2)

A 233 bp ClaI-NcoI cDNA fragment from the Drosophila knirps-coding region (Nauber et al., 1988) was cloned in the pGEM-T-Easy vector (Promega, Madison, WI) and a 309 bp SphI-SpeI DNA fragment containing both knirps and vector sequences was used as the final spacer.

Gel shift assays

Binding assays were done with synthetic DNA fragments that contain the sna2-, sna3- or sna4-binding sites from the Ci-Bra enhancer. These fragments were prepared by annealing the following complementary oligonucleotides:

sna2: ACGTCACAATACACTTGGTGACGT and TATGACGT-CACCAAGTGTATTGTG

sna3: AAAACAAACACAAGGTGTTCGAT and CTGGATCG-AACACCTTGTGTTTG

sna4: CTGGATGCCACCACCTACGGCGC and AAGTGCGC-CGTAGGTGGTGGCAT

The resulting double-stranded DNA fragments were labeled with [γ-32P] ATP using T4 polynucleotide kinase (Sambrook et al., 1989). Various cold competitor DNAs were used in the binding assays. These included the wild-type sna2, sna3 and sna4 sequences, a mutant form of sna2 (sna2MUT), sna 1 and sna1MUT, and the Drosophila s2 sequence. The following oligonucleotides were used to prepare these DNA fragments:

sna2MUT: AACACGTCACAATAGTGACGTCAT + TGATATG-ACGTCACTATTGTGACG

sna1: CAAACCTTATCTGGTGTTACGTCA + ATTGTGACGT-AACACCAGATAAGG

sna1MUT: AATAGCGACAAACCTACGTCACAA + TGTATT-GTGACGTAGGTTTGTCGC

s2: GATCTGCGGCACCTTGCTGGGCAG + GATCCTGCCC-AGCAAGGTGCCGCA

Binding assays were done with a glutathione-S-transferase-Ci-Sna (GST-Ci-Sna) fusion protein that contains the C-terminal 159 amino acids, which includes all five zinc fingers (Corbo et al., 1997b). This fusion protein was prepared with the pGEX-KG expression vector using the HB101 strain of E. coli. The GST-Ci-Sna fusion protein was purified from crude bacterial extracts using glutathione-agarose beads (Smith and Johnson, 1988).

The fusion protein was preincubated with 10 pmol/μl of cold competitor DNAs in a solution containing 10 mM Hepes (pH 7.9), 50 mM KCl, 0.1 mM EDTA, 0.1 mM EGTA, 0.1% NP-40, 0.5 mM DTT, 0.5 mM PMSF, 1 mM benzamidine, 2 μg/ml leupeptin, 2.5 μg/ml pepstatin A, 0.5 μg/ml BSA, 0.1 μg/ml poly[dI-dC] at room temperature for 5 minutes. Aliquots containing 0.2 pmol/μl of the labeled sna2 DNA were added and incubated at room temperature for 15 minutes. Afterwards, the reaction mixtures were fractionated on a 5% polyacrylamide gel and subsequently prepared for autoradiography (Ip et al., 1991).

Construction of transgenic DNAs

The 3.5 kb, −434 bp, and −251 bpCi-Bra/lacZ transgenes are described in Corbo et al. (1997a). Mutagenized Ci-Bra enhancer sequences (see above) were excised from the pBluescript vector using XhoI and PstI. These DNA fragments were replaced with the corresponding region of the wild-type −251 bp Ci-Bra/lacZ transgene.

Ci-sna expression plasmids were prepared as follows. Wild-type or mutagenized Ci-sna cDNAs were excised from the pBluescript vector using KpnI and SacI. The resulting fragment was used to replace the lacZ reporter gene present in a truncated −251 bp Ci-Bra/lacZ transgene (see Corbo et al., 1997a). The truncated Ci-Bra enhancer directs Ci-sna expression in both the notochord and tail muscles.

The preparation of the Ci-snaVP16 recombinant-coding sequence involved the use of the VP16 activation domain (see Rusch and Levine, 1997). A synthetic BamHI site was introduced in the C terminus of the Ci-sna-coding region via site-directed mutagenesis. The VP16 activation domain was ligated to the C-terminal Ci-sna-coding region contained within an EcoRV-BamHI fragment. This fragment contains the five zinc fingers, but lacks the putative repression domain located in more N-terminal regions of the protein. The resulting Ci-snaVP16 recombinant was then placed downstream of a ClaI site, approximately 1 kb downstream of the translation start site of the lacZ open reading frame. The lacZ-Ci-snaVP16 fusion gene contains a nuclear localization sequence (NLS) present in the modified lacZ reporter gene.

Electroporation and in situ hybridization assays

In situ hybridization assays were done with digoxigenin-labeled antisense RNA probes exactly as described by Corbo et al. (1997a).

Transgenic DNAs were electroporated into Ciona embryos as described previously (Corbo et al., 1997a). Briefly, aliquots containing ∼50 μg of each lacZ reporter DNA was dissolved in 500 μl of 0.77 M mannitol. Co-electroporations involved the use of 50 μg of either the Ci-Bra or Ci-sna expression vectors. The DNAs were electroporated into embryos ∼25-35 minutes after fertilization and the embryos were allowed to develop until tailbud stages. The embryos were fixed in formaldehyde and stained for β-galactosidase activity, as described by Corbo et al. (1997a). A typical experiment involves the electroporation of at least 100-200 stained embryos and each of the results presented in this study was obtained in at least three separate experiments.

In situ hybridization assays reveal Ci-sna expression in the progenitors of the primary and secondary tail muscles by the onset of gastrulation (Corbo et al., 1997b). It is also expressed in the primordia of the cerebral vesicle and lateral ependymal cells of the spinal cord. Ci-sna expression is lost in the central nervous system (CNS) during neurulation, but staining persists in the primary and secondary tail muscles during intercalation of the notochord and extension of the tail (arrowheads, Fig. 1A,B). A transverse, dorsal view of stained embryos clearly reveals the subdivision of the mesoderm into dorsal and lateral lineages. The dorsal-most, or axial, mesoderm gives rise to the notochord (Fig. 1B, arrow), while the lateral mesoderm differentiates into definitive tail muscles (Fig. 1B, arrowheads). Subsequent experiments were conducted to determine whether Ci-sna helps specify this boundary between the notochord and tail muscles.

Fig. 1.

Localization of Ci-sna RNAs. Fertilized eggs were dechorionated and allowed to develop to tailbud stages prior to fixation. The embryos were hybridized with a digoxigenin-labeled Ci-sna antisense RNA probe and stained with alkaline phosphatase. (A) Lateral view of a stained embryo showing Ci-sna expression in tail muscles (arrowhead). (B) Dorsal view of a stained embryo. Strong Ci-sna expression is observed in the lateral tail muscles (arrowhead), but there is no detectable staining in the notochord (arrow).

Fig. 1.

Localization of Ci-sna RNAs. Fertilized eggs were dechorionated and allowed to develop to tailbud stages prior to fixation. The embryos were hybridized with a digoxigenin-labeled Ci-sna antisense RNA probe and stained with alkaline phosphatase. (A) Lateral view of a stained embryo showing Ci-sna expression in tail muscles (arrowhead). (B) Dorsal view of a stained embryo. Strong Ci-sna expression is observed in the lateral tail muscles (arrowhead), but there is no detectable staining in the notochord (arrow).

Ci-Sna binds to specific sites within the minimal Ci-Bra enhancer

Previous studies identified a minimal 434 bp enhancer from the Ci-Bra promoter region that mediates notochord-restricted expression of green fluorescent protein and lacZ transgenes in electroporated embryos (Corbo et al., 1997a). The enhancer appears to be activated by a combination of bHLH proteins and a Ciona homolog of the Drosophila Suppressor of Hairless {Su(H)} protein (Corbo, 1997; Corbo et al., unpublished data). A truncated Ci-Bra enhancer mediates ectopic expression in the trunk mesenchyme and tail muscles. These results suggest that bHLH-Su(H) synergy permits expression of the Ci-Bra enhancer in most mesodermal lineages, including the tail muscles, notochord and trunk mesenchyme, while one or more spatially localized repressors restricts Ci-Bra expression to the notochord.

As a first step towards assessing the role of Ci-Sna in Ci-Bra regulation, gel shift assays were conducted using a GST-Ci-Sna fusion protein that contains the carboxy-terminal 159 amino acid residues of Ci-Sna, and includes all five zinc fingers present in the full-length protein (Fig. 2). These experiments identified four different Ci-Sna-binding sites in the Ci-Bra enhancer, sna1, sna2, sna3 and sna4. The distal-most site, sna1, is located in the interval between −434 bp and −251 bp, which was previously shown to be important for restricted expression in the notochord (Corbo et al., 1997a).

Fig. 2.

Gel shift assays. (A) A truncated GST-Ci-Sna fusion protein was incubated with a 32P-labeled sna2 synthetic DNA fragment. Protein-DNA complexes were fractionated on a polyacrylamide gel and visualized by autoradiography. Lane 1. No cold competitor DNA was added. A strong protein-DNA complex is observed (arrowhead). Lane 2. Same as lane 1 except that the GST-Ci-Sna fusion protein was preincubated with a 50 molar excess of unlabeled sna2 competitor DNA. There is a substantial reduction in the levels of specific protein-DNA complexes. Lane 3. Same as lane 2 except that the competitor DNA corresponds to a mutant form of sna2 (sna2MUT). The mutant DNA causes little or no inhibition of sna2/Ci-Sna complexes. Lane 4. The GST-Ci-Sna fusion protein was preincubated with a 50 molar excess of a DNA fragment containing the sna1 sequence. There is a substantial inhibition in the formation of sna2/Ci-Sna complexes, although the sna1 sequence might have a slightly lower binding affinity than sna2 (compare with lane 2). Lane 5. Same as lane 4 except that a mutant sna1 DNA was used as the competitor. There is no discernible block in the formation of sna2/Ci-Sna complexes. Lane 6. Same as lanes 4 and 5 except that the competitor DNA corresponds to the Drosophila Snail s2-binding site. The s2 sequence leads to a complete inhibition in the formation of sna2/Ci-Sna complexes, suggesting that it has a higher affinity for the GST-Ci-Sna fusion protein than either sna1 or sna2. (B) Gel shift assays with labeled sna3 and sna4 probes. The GST-Ci-Sna fusion protein forms a shifted complex (arrowhead) with these sequences in the absence of competitor DNA (lanes 1 and 2). Increasing amounts of unlabeled sna3 and sna4 competitor DNA reduces GST-Ci-Sna/sna3 complexes. (C) Sequences of the DNA fragments used for the binding assays. The boxed regions in the sna1 and sna2 fragments correspond to the sequences that were deleted in sna1MUT and sna2MUT, respectively. These sequences were also deleted in Ci-Bra/lacZ transgenes used for the experiments shown in Fig. 3B,C. The boxed region of the s2 DNA fragment corresponds to the sequence that was used in Fig. 3F. (D) DNA sequence of the minimal 434 bp Ci-Bra enhancer (see Corbo et al., 1997a). The enhancer appears to be activated by a combination of three E box elements, E1, E2 and E3 (E1 overlaps with the sna2 site), and two Su(H)-binding sites {Su(H)1 and Su(H)2}. Point mutations in these putative activator sites reduce or eliminate staining mediated by Ci-Bra/lacZ transgenes (Corbo, 1997; Corbo et al., unpublished data). Gel shift assays identified four Ci-Sna-binding sites, sna1, sna2, sna3 and sna4. As seen for the Drosophila Snail protein, the Ci-Sna-binding sites are related to E box sequences. (E) Summary of factor-binding sites. The rectangles correspond to putative activator sites, while the circles represent potential repressor elements (Ci-Sna sites). As mentioned above, the sna2 site overlaps the E1 E box sequence. The organization of the Ci-Bra enhancer is very similar to that observed for the rho neuroectoderm enhancer (NEE; Ip et al., 1992).

Fig. 2.

Gel shift assays. (A) A truncated GST-Ci-Sna fusion protein was incubated with a 32P-labeled sna2 synthetic DNA fragment. Protein-DNA complexes were fractionated on a polyacrylamide gel and visualized by autoradiography. Lane 1. No cold competitor DNA was added. A strong protein-DNA complex is observed (arrowhead). Lane 2. Same as lane 1 except that the GST-Ci-Sna fusion protein was preincubated with a 50 molar excess of unlabeled sna2 competitor DNA. There is a substantial reduction in the levels of specific protein-DNA complexes. Lane 3. Same as lane 2 except that the competitor DNA corresponds to a mutant form of sna2 (sna2MUT). The mutant DNA causes little or no inhibition of sna2/Ci-Sna complexes. Lane 4. The GST-Ci-Sna fusion protein was preincubated with a 50 molar excess of a DNA fragment containing the sna1 sequence. There is a substantial inhibition in the formation of sna2/Ci-Sna complexes, although the sna1 sequence might have a slightly lower binding affinity than sna2 (compare with lane 2). Lane 5. Same as lane 4 except that a mutant sna1 DNA was used as the competitor. There is no discernible block in the formation of sna2/Ci-Sna complexes. Lane 6. Same as lanes 4 and 5 except that the competitor DNA corresponds to the Drosophila Snail s2-binding site. The s2 sequence leads to a complete inhibition in the formation of sna2/Ci-Sna complexes, suggesting that it has a higher affinity for the GST-Ci-Sna fusion protein than either sna1 or sna2. (B) Gel shift assays with labeled sna3 and sna4 probes. The GST-Ci-Sna fusion protein forms a shifted complex (arrowhead) with these sequences in the absence of competitor DNA (lanes 1 and 2). Increasing amounts of unlabeled sna3 and sna4 competitor DNA reduces GST-Ci-Sna/sna3 complexes. (C) Sequences of the DNA fragments used for the binding assays. The boxed regions in the sna1 and sna2 fragments correspond to the sequences that were deleted in sna1MUT and sna2MUT, respectively. These sequences were also deleted in Ci-Bra/lacZ transgenes used for the experiments shown in Fig. 3B,C. The boxed region of the s2 DNA fragment corresponds to the sequence that was used in Fig. 3F. (D) DNA sequence of the minimal 434 bp Ci-Bra enhancer (see Corbo et al., 1997a). The enhancer appears to be activated by a combination of three E box elements, E1, E2 and E3 (E1 overlaps with the sna2 site), and two Su(H)-binding sites {Su(H)1 and Su(H)2}. Point mutations in these putative activator sites reduce or eliminate staining mediated by Ci-Bra/lacZ transgenes (Corbo, 1997; Corbo et al., unpublished data). Gel shift assays identified four Ci-Sna-binding sites, sna1, sna2, sna3 and sna4. As seen for the Drosophila Snail protein, the Ci-Sna-binding sites are related to E box sequences. (E) Summary of factor-binding sites. The rectangles correspond to putative activator sites, while the circles represent potential repressor elements (Ci-Sna sites). As mentioned above, the sna2 site overlaps the E1 E box sequence. The organization of the Ci-Bra enhancer is very similar to that observed for the rho neuroectoderm enhancer (NEE; Ip et al., 1992).

DNA-protein complexes obtained with sna2 and the GST-Ci-Sna fusion protein are specifically inhibited with a 50-fold molar excess of unlabeled sna1 or sna2 DNA (Fig. 2A). However, mutant forms of sna1 (sna1MUT) or sna2 (sna2MUT) do not significantly block the shifted complexes (Fig. 2A). A high-affinity binding site for the Drosophila Snail protein, s2 (Ip et al., 1992), also mediates efficient competition in these assays (Fig. 2A).

Similar assays were used to identify two additional Ci-Sna-binding sites in the 434 bp Ci-Bra enhancer, sna3 and sna4 (Fig. 2B). Ci-Sna/sna3 protein-DNA complexes are blocked by the addition of increasing amounts of unlabeled sna3 and sna4 DNAs (Fig. 2B). The sequences of the wild-type and mutant Ci-Sna-binding sites used in these assays are shown in Fig. 2C.

The Ci-Bra enhancer sequence (Fig. 2D) contains two Ci-Su(H)-binding sites and three putative bHLH-binding sites (E boxes). The E1 E box overlaps with the sna2-binding site, which is similar to the situation observed for the Drosophila Snail protein whereby half the binding sites also include an E box consensus sequence (CANNTG). The sna1 and sna2 sites map quite close to Ci-Su(H) activator sites (Fig. 2E), which are essential for the expression of Ci-Bra transgenes in the notochord. A number of experiments were conducted to determine whether the sna1 and sna2 sites function as repressor elements that restrict the Ci-Bra pattern to the notochord.

Expression of mutagenized Ci-Bra transgenes in electroporated embryos

As shown previously (Corbo et al., 1997a), the 434 bp Ci-Bra enhancer directs robust expression of a lacZ transgene in the notochords of electroporated embryos (Fig. 3A, arrow). Staining is detected in both the primary notochord cells that arise from the A4.1 blastomere, as well as the secondary, B4.1-derived notochord cells located in the posterior regions of the tail. The wild-type enhancer also exhibits variable expression in the trunk mesenchyme (Fig. 3A, unfilled arrow).

Fig. 3.

Ci-Sna-binding sites are essential for the normal Ci-Bra expression pattern. Embryos were grown to mid-tailbud stages after electroporation with various Ci-Bra/lacZ transgenes (see diagrams below each panel) and the expression patterns were visualized by enzymatic staining. Black arrows indicate staining in the notochord; black arrowheads, staining in the tail muscles. Embryos are oriented with anterior to the left. (A) Expression pattern obtained with the wild-type 434 bp Ci-Bra/lacZ transgene. Staining is detected in the notochord and trunk mesenchyme (white arrow). (B) Expression pattern obtained with a mutant form of the 434 bp Ci-Bra enhancer, which contains a deletion within the sna1-binding site (note ‘X’ in the sna 1 site in the diagram). Staining is no longer restricted to the notochord, but is also detected in the tail muscles. (C) Expression pattern obtained with a mutant 434 bp Ci-Bra enhancer that contains a deletion within the sna2 site. Staining is detected in both the notochord and tail muscles. (D) Same as C except that the mutagenized Ci-Bra enhancer was modified to include a synthetic Drosophila s2 site in place of the native sna2 sequence. A normal lacZ staining pattern is observed. (E) Expression pattern obtained with a truncated −251 bp Ci-Bra/lacZ transgene. Staining is observed in both the notochord and tail muscles (see Corbo et al., 1997a). (F) Same as E except that the synthetic Drosophila s2 sequence was inserted 5′ of the Su(H)1 site. This restores a normal staining pattern. (G) Staining pattern obtained with a modified 434 bp Ci-Bra/lacZ transgene containing a 309 bp spacer sequence between the sna1 and Su(H)1 sites. Staining is detected in all mesodermal lineages, including the notochord and tail muscles. Staining is also detected in the cerebral vesicle of the CNS (open arrowhead). (H) Same as (G) except that the spacer was inserted between the Su(H)2 and sna2 sites. The reporter gene is expressed in all mesodermal lineages, as well as some of the ependymal cells of the spinal cord (open arrowhead).

Fig. 3.

Ci-Sna-binding sites are essential for the normal Ci-Bra expression pattern. Embryos were grown to mid-tailbud stages after electroporation with various Ci-Bra/lacZ transgenes (see diagrams below each panel) and the expression patterns were visualized by enzymatic staining. Black arrows indicate staining in the notochord; black arrowheads, staining in the tail muscles. Embryos are oriented with anterior to the left. (A) Expression pattern obtained with the wild-type 434 bp Ci-Bra/lacZ transgene. Staining is detected in the notochord and trunk mesenchyme (white arrow). (B) Expression pattern obtained with a mutant form of the 434 bp Ci-Bra enhancer, which contains a deletion within the sna1-binding site (note ‘X’ in the sna 1 site in the diagram). Staining is no longer restricted to the notochord, but is also detected in the tail muscles. (C) Expression pattern obtained with a mutant 434 bp Ci-Bra enhancer that contains a deletion within the sna2 site. Staining is detected in both the notochord and tail muscles. (D) Same as C except that the mutagenized Ci-Bra enhancer was modified to include a synthetic Drosophila s2 site in place of the native sna2 sequence. A normal lacZ staining pattern is observed. (E) Expression pattern obtained with a truncated −251 bp Ci-Bra/lacZ transgene. Staining is observed in both the notochord and tail muscles (see Corbo et al., 1997a). (F) Same as E except that the synthetic Drosophila s2 sequence was inserted 5′ of the Su(H)1 site. This restores a normal staining pattern. (G) Staining pattern obtained with a modified 434 bp Ci-Bra/lacZ transgene containing a 309 bp spacer sequence between the sna1 and Su(H)1 sites. Staining is detected in all mesodermal lineages, including the notochord and tail muscles. Staining is also detected in the cerebral vesicle of the CNS (open arrowhead). (H) Same as (G) except that the spacer was inserted between the Su(H)2 and sna2 sites. The reporter gene is expressed in all mesodermal lineages, as well as some of the ependymal cells of the spinal cord (open arrowhead).

Mutations in the sna1 site cause an otherwise normal Ci-Bra/lacZ transgene to be misexpressed in the tail muscles (Fig. 3B, arrowheads). This disruption in sna1 does not affect staining in the notochord (Fig. 3B, arrow). A similar expression pattern is obtained with a truncated Ci-Bra enhancer that lacks the sna1 site (Fig. 3E; see Corbo et al., 1997a). These results are consistent with the notion that sna1 corresponds to the negative response element located in the distal region of the Ci-Bra enhancer.

Additional mutagenesis experiments suggest that the sna2 site is also important for the normal, notochord-restricted staining pattern. Mutations in sna2 cause a derepression of the Ci-Bra/lacZ staining pattern that is virtually indistinguishable from that obtained with mutations in the sna1 site (Fig. 3C; compare with 3B). Again, staining is no longer restricted to the notochord, but is also detected in the tail muscles. These results suggest that both sna1 and sna2 are required for efficient repression. In contrast, a single Snail site appears to be sufficient to repress the rhomboid lateral stripe enhancer in Drosophila (Gray et al., 1994; see Discussion).

Mutations in the sna1 and sna2 sites result in the derepression of Ci-Bra/lacZ transgenes in the tail muscles, and possibly trunk mesenchyme. However, as noted above, the normal Ci-Bra enhancer exhibits variable expression in the mesenchyme (e.g. Fig. 3A) and, consequently, we will focus our attention on the repression of Ci-Bra in the tail muscles. Moreover, mutations in the sna3 and sna4 sites do not cause an obvious derepression of the Ci-Bra/lacZ staining pattern in tail muscles (data not shown), so subsequent experiments were restricted to the manipulation of the critical sna1 and sna2 sites.

Further evidence that Ci-Sna directly represses the Ci-Bra enhancer stems from cis complementation experiments, whereby a synthetic Snail-binding site was inserted in defective Ci-Bra enhancers lacking either the native sna1 or sna2 sites. The insertion of the Drosophila s2 sequence (Ip et al., 1992) in a defective Ci-Bra enhancer lacking the sna2 site restores an essentially normal staining pattern, whereby the modified Ci-Bra/lacZ transgene exhibits restricted expression in the notochord (Fig. 3D; compare with 3C). Similar results were obtained with a truncated Ci-Bra enhancer lacking the sna1 site (Fig. 3F). A synthetic s2 site at the 5′ end of the truncated enhancer restores the normal, notochord-restricted staining pattern (Fig. 3F, arrow; compare with 3E).

Spacer sequences were inserted between sna1 and Su(H)1, and Su(H)2 and sna2 (Fig. 3G, H) to determine whether Ci-Sna functions as a short-range repressor, as seen in Drosophila (Gray et al., 1994; Gray and Levine, 1996). Increasing the distance between sna1 and Su(H)1 results in a severe derepression of the Ci-sna/lacZ staining pattern (Fig. 3G). A similar derepression is observed when the same spacer sequence was placed between the Su(H)2 and sna2 sites (Fig. 3H). Staining is also detected in the cerebral vesicle (unfilled arrowheads, Fig. 3G, H), which is a normal site of Ci-sna expression (Corbo et al., 1997b; Erives et al., 1998). These results suggest that the critical sna1 and sna2 sites must map close to the Su(H) activator sites in order for Ci-Sna to repress Ci-Bra expression in the tail muscles.

Trans-complementation assays

Ci-Sna was misexpressed in the notochord by attaching the Ci-sna-coding sequence to a truncated 251 bp Ci-Bra enhancer (Ci-Bra*) that mediates expression in all mesodermal lineages, including the notochord, tail muscles and trunk mesenchyme (see Fig. 3E). Fertilized eggs were co-electroporated with a mixture of transgenic DNAs. One of the transgenes corresponds to the truncated Ci-Bra enhancer driving the Ci-sna gene (Ci-Bra*/Ci-sna). The other is a reporter gene that contains the 434 bp Ci-Bra enhancer attached to the lacZ-coding sequence (Ci-Bra/lacZ). Co-electroporation of these transgenes results in the repression of the reporter gene and the loss of lacZ staining (Fig. 4B). In contrast, there is no repression of the Ci-Bra/lacZ reporter gene when it is co-electroporated with an expression vector containing the truncated Ci-Bra* enhancer driving a mutagenized Ci-sna-coding sequence (sna* * *) (Fig. 4A). The mutagenized Ci-Sna protein contains point mutations in three of the five C-terminal zinc fingers.

Fig. 4.

Misexpression of Ci-Sna in the notochord. Tailbud stage embryos were co-electroporated with Ci-Bra/lacZ reporter genes and Ci-Bra*/Ci-sna expression vectors that misexpress Ci-Sna in the notochord. Embryos were stained to reveal β-galactosidase expression and are oriented with anterior to the left. (A) Tailbud stage embryo that was co-electroporated with the 434 bp Ci-Bra/lacZ reporter gene and a mutagenized expression vector (Ci-Bra*/Ci-sna* * *) containing the truncated 251 bp Ci-Bra enhancer attached to a mutant Ci-sna-coding region lacking three of the C-terminal zinc fingers. The lacZ reporter gene exhibits normal expression in the notochord. (B) Same as A except that the reporter gene was co-electroporated with the normal Ci-Bra*/Ci-sna expression vector, which encodes a wild-type Ci-Sna protein. There is a complete loss of staining in the notochord. (C) Same as B except that the Ci-Bra*/Ci-sna expression vector was co-electroporated with a mutagenized Ci-Bra/lacZ reporter gene, which contains a deletion in the critical sna2 site. lacZ staining is observed in the notochord and tail muscles, suggesting that the loss of the sna2 site renders the Ci-Bra enhancer refractory to repression by Ci-Sna.

Fig. 4.

Misexpression of Ci-Sna in the notochord. Tailbud stage embryos were co-electroporated with Ci-Bra/lacZ reporter genes and Ci-Bra*/Ci-sna expression vectors that misexpress Ci-Sna in the notochord. Embryos were stained to reveal β-galactosidase expression and are oriented with anterior to the left. (A) Tailbud stage embryo that was co-electroporated with the 434 bp Ci-Bra/lacZ reporter gene and a mutagenized expression vector (Ci-Bra*/Ci-sna* * *) containing the truncated 251 bp Ci-Bra enhancer attached to a mutant Ci-sna-coding region lacking three of the C-terminal zinc fingers. The lacZ reporter gene exhibits normal expression in the notochord. (B) Same as A except that the reporter gene was co-electroporated with the normal Ci-Bra*/Ci-sna expression vector, which encodes a wild-type Ci-Sna protein. There is a complete loss of staining in the notochord. (C) Same as B except that the Ci-Bra*/Ci-sna expression vector was co-electroporated with a mutagenized Ci-Bra/lacZ reporter gene, which contains a deletion in the critical sna2 site. lacZ staining is observed in the notochord and tail muscles, suggesting that the loss of the sna2 site renders the Ci-Bra enhancer refractory to repression by Ci-Sna.

These results suggest that ectopic expression of Ci-Sna in the notochord (and perhaps mesenchyme) leads to the repression of the Ci-Bra/lacZ transgene (see below). An additional control experiment was done to strengthen this conclusion. Embryos were co-electroporated with the Ci-Bra*/Ci-sna expression vector (which encodes the normal Ci-Sna protein) and a mutagenized Ci-Bra/lacZ reporter gene lacking the sna2 site. Intense staining is observed in the notochord, tail muscles and trunk mesenchyme (Fig. 4C), similar to the pattern obtained with the same reporter gene in the absence of the Ci-Bra*/Ci-sna expression vector (e.g. Fig. 3C). Thus, loss of the sna2 site renders an otherwise normal Ci-Bra/lacZ transgene immune to repression by ectopic Ci-Sna products in the notochord.

The misexpression of Ci-Sna does not cause an obvious mutant phenotype (e.g. Fig. 4B), suggesting that the endogenous Ci-Bra gene is not repressed. It is conceivable that the minimal 434 bp Ci-Bra/lacZ reporter gene, which was used in the preceding experiments (Fig. 4A,B), is sensitized for repression by Ci-Sna. Perhaps repression of the ‘full-length’ Ci-Bra promoter and endogenous gene depends on higher threshold levels of the Ci-Sna protein than those produced by the truncated 251 Ci-Bra*/Ci-sna expression vector. To test this possibility, co-electroporation assays were done with a Ci-Bra/lacZ reporter gene that contains 3.5 kb (rather than 434 bp) of the Ci-Bra 5′ regulatory region (Fig. 5C,D). Co-electroporation of this reporter gene with the wild-type Ci-Bra*/Ci-sna expression vector results in only a slight reduction of lacZ staining in the notochord (Fig. 5D). Parallel experiments with the same Ci-Sna expression vector and the minimal 434 bp Ci-Bra/lacZ reporter gene results in the loss of lacZ staining (Fig. 5B), as seen previously (Fig. 4B). The mutagenized Ci-Bra*/Ci-sna* * * expression vector fails to repress both the 434 bp and 3.5 kb Ci-Bra/lacZ reporter genes (Fig. 5A,C, respectively).

Fig. 5.

Ectopic Ci-Sna products do not repress the full-length Ci-Bra promoter region. Embryos were co-electroporated with Ci-Bra/Ci-sna expression vectors and Ci-Bra/lacZ reporter genes. Embryos were stained and oriented as in Figs 3, 4. (A) The minimal 434 bp Ci-Bra/lacZ reporter gene was co-electroporated with a mutagenized Ci-Bra*/Ci-sna* * * expression vector lacking three of the Ci-Sna zinc fingers. A normal Ci-Bra staining pattern is observed in the notochord. (B) Same as A except that the 434 bp Ci-Bra/lacZ reporter gene was co-electroporated with the normal Ci-Bra*/Ci-sna expression vector, which encodes a wild-type Ci-Sna protein. There is complete repression of the reporter gene in the notochord. (C) A full-length 3.5 kb Ci-Bra/lacZ reporter gene was co-electroporated with the mutagenized Ci-Bra*/Ci-sna* * * expression vector. Intense staining is detected in the notochord. (D) Same as C except that the normal Ci-Bra*/Ci-sna expression vector was used. The wild-type Ci-Sna protein results in only a slight reduction in the staining directed by the 3.5 kb Ci-Bra/lacZ reporter gene (compare with the 434 bp reporter gene in B).

Fig. 5.

Ectopic Ci-Sna products do not repress the full-length Ci-Bra promoter region. Embryos were co-electroporated with Ci-Bra/Ci-sna expression vectors and Ci-Bra/lacZ reporter genes. Embryos were stained and oriented as in Figs 3, 4. (A) The minimal 434 bp Ci-Bra/lacZ reporter gene was co-electroporated with a mutagenized Ci-Bra*/Ci-sna* * * expression vector lacking three of the Ci-Sna zinc fingers. A normal Ci-Bra staining pattern is observed in the notochord. (B) Same as A except that the 434 bp Ci-Bra/lacZ reporter gene was co-electroporated with the normal Ci-Bra*/Ci-sna expression vector, which encodes a wild-type Ci-Sna protein. There is complete repression of the reporter gene in the notochord. (C) A full-length 3.5 kb Ci-Bra/lacZ reporter gene was co-electroporated with the mutagenized Ci-Bra*/Ci-sna* * * expression vector. Intense staining is detected in the notochord. (D) Same as C except that the normal Ci-Bra*/Ci-sna expression vector was used. The wild-type Ci-Sna protein results in only a slight reduction in the staining directed by the 3.5 kb Ci-Bra/lacZ reporter gene (compare with the 434 bp reporter gene in B).

These results indicate that the Ci-Bra*/Ci-sna expression vector directs sufficient levels of the Ci-Sna protein to repress the 434 bp Ci-Bra/lacZ reporter gene, but not the 3.5 kb Ci-Bra/lacZ reporter. Perhaps the full-length Ci-Bra 5′ regulatory region contains multiple enhancers that mediate expression in the notochord, so that higher levels of Ci-Sna are required for repression. In situ hybridization assays suggest that the endogenous Ci-Bra gene is not repressed by the Ci-Bra*/Ci-sna expression vector in electroporated embryos (data not shown).

A Ci-SnaVP16 fusion protein disrupts the notochord/muscle boundary

Electroporation assays were conducted with a chimeric Ci-Sna protein that contains the VP16 activation domain in place of the Ci-Sna repression domain (see diagrams below the photomicrographs in Fig. 6). The Ci-SnaVP16-coding sequence was placed under the control of the ‘full-length’ Ci-sna promoter region, which includes the first 2.2 kb of 5′ flanking sequence and extends 3′ of the transcription start-site through the first intron of the transcribed region (Erives et al., 1998). This Ci-snaVP16 fusion gene was co-electroporated with the 434 bp Ci-Bra/lacZ reporter gene (Fig. 6). Normally, the reporter gene is expressed only in the notochord and sporadically in the trunk mesenchyme (e.g. Figs 3A, 4A, 5A). However, the Ci-snaVP16 transgene mediates ectopic activation of the Ci-Bra reporter in putative tail muscles (arrowheads, Fig. 6B,D). This ectopic expression is somewhat variable in intensity, but is nonetheless seen in many of the normal tail muscles. Moreover, the electroporated embryos exhibit a consistent mutant phenotype, whereby the tail is reduced in length and the tail muscles appear to be disorganized. These results suggest that the Ci-snaVP16 transgene alters the expression of endogenous genes and causes a concomitant disruption in the separation of the notochord-muscle boundary.

Fig. 6.

A Ci-SnaVP16 fusion protein disrupts tail morphogenesis. Embryos were co-electroporated with the 434 bp Ci-Bra/lacZ reporter gene, and an expression vector that contains the Ci-SnaVP16 recombinant-coding region under the control of the normal Ci-sna promoter region. The fusion protein contains the Ci-Sna DNA-binding domain, and the VP16 activation domain in place of the normal repression domain. (A) This co-electroporated embryo exhibits a relatively mild phenotype, perhaps due to the incorporation of only small amounts of the Ci-snaVP16 transgene. (B) High magnification view of the boxed region in A. The Ci-Bra/lacZ reporter gene exhibits ectopic expression in the tail muscles (arrowheads), in addition to normal expression in the notochord (arrows). (C) An embryo that exhibits a somewhat more severe mutant phenotype. There is a substantial failure of tail elongation. (D) High magnification view of the boxed region in C. The Ci-Bra/lacZ reporter gene is expressed in the notochord (arrows) and is also misexpressed in the tail muscles (arrowheads).

Fig. 6.

A Ci-SnaVP16 fusion protein disrupts tail morphogenesis. Embryos were co-electroporated with the 434 bp Ci-Bra/lacZ reporter gene, and an expression vector that contains the Ci-SnaVP16 recombinant-coding region under the control of the normal Ci-sna promoter region. The fusion protein contains the Ci-Sna DNA-binding domain, and the VP16 activation domain in place of the normal repression domain. (A) This co-electroporated embryo exhibits a relatively mild phenotype, perhaps due to the incorporation of only small amounts of the Ci-snaVP16 transgene. (B) High magnification view of the boxed region in A. The Ci-Bra/lacZ reporter gene exhibits ectopic expression in the tail muscles (arrowheads), in addition to normal expression in the notochord (arrows). (C) An embryo that exhibits a somewhat more severe mutant phenotype. There is a substantial failure of tail elongation. (D) High magnification view of the boxed region in C. The Ci-Bra/lacZ reporter gene is expressed in the notochord (arrows) and is also misexpressed in the tail muscles (arrowheads).

We have presented evidence that Ci-Sna restricts the expression of a minimal 434 bp Ci-Bra enhancer to the notochord of Ciona embryos. Repression in tail muscles depends on paired Ci-Sna-binding sites located on either side of Su(H) activator sites. We propose that Ci-Sna functions as a boundary repressor that subdivides the mesoderm into dorsal and lateral lineages. This boundary appears to be disrupted in electroporated embryos that express a Ci-SnaVP16 fusion protein.

Mechanisms of repression

Systematic mutations in each of the four Ci-Sna-binding sites within the 434 bp Ci-Bra enhancer suggest that the sna1 and sna2 sites are critical for restricted expression in the notochord, while mutations in sna3 and sna4 result in only very weak derepression in the tail muscles (see Fig. 3 and data not shown). Repression appears to depend on tight linkage between the sna1 and sna2 sites and Su(H) activator sites. The insertion of spacer sequences between sna1 and Su(H)1 or between Su(H)2 and sna2 results in a severe derepression of Ci-Bra/lacZ transgenes in the tail muscles (see Fig. 3). However, it is conceivable that this derepression results from alterations in the overall organization of the Ci-Bra enhancer. The insertion of the same spacer sequences in different regions of the Ci-Bra enhancer also result in ectopic expression (S. Fujiwara, unpublished results). Perhaps the precise organization of the cis regulatory elements within the Ci-Bra enhancer is essential for normal expression, similar to the situation seen for the TCRα enhancer in mammalian T cells (e.g. Bruhn et al., 1997).

Previous studies on the rhomboid (rho) neuroectoderm enhancer (NEE) suggest that a single Snail repressor site is sufficient to block Dorsal-bHLH synergy and inhibit expression in the ventral mesoderm (Gray et al., 1994). The minimal 300 bp NEE contains four high-affinity Snail-binding sites and removal of all but one of the sites, s2, does not substantially alter the normal staining pattern in transgenic embryos (Ip et al., 1992; Gray et al., 1994). The s2 site is located near the middle of the enhancer and is closely linked to two of the four Dorsal activator sites. However, s2 continues to mediate repression even when positioned upstream of the distal-most Dorsal site. The only obvious constraint on s2 function is that it must map within ∼50 bp of a Dorsal activator site (Gray et al., 1994).

There may be differences in the way that the Drosophila Snail and Ci-Sna proteins mediate repression. Intact sna1 and sna2 sites appear to be required for the repression of Ci-Bra in the tail muscles. Neither site is sufficient for repression, despite the tight linkage of each site with Su(H) activator sites. There are several potential explanations for these apparent differences in the Drosophila and Ciona Snail repressors. First, it is possible that the rho NEE contains cryptic low affinity sites (Ip et al., 1992). These sites may act in concert with s2 to surround Dorsal activator sites, similar to the situation observed for the Ci-Bra enhancer. Second, the Drosophila Snail repression domain contains two potential binding motifs for the putative corepressor protein, CtBP (Boyd et al., 1993; Schaeper et al., 1995; Sollerbrant et al., 1996; Nibu et al., 1998). In contrast, Ci-Sna contains just a single copy of this motif (P-DLS-K). Perhaps efficient repression by Snail depends on the recruitment of at least two CtBP corepressors to the template DNA. Finally, the Ciona Su(H) activator may be more robust than the Drosophila Dorsal activator and require multiple repressors for inhibition.

Mutant phenotypes

A Ci-SnaVP16 fusion protein appears to interfere with muscle differentiation when expressed under the control of the normal Ci-sna promoter region, which is activated prior to the onset of Ci-Bra expression. The VP16 moiety presumably converts the Ci-Sna repressor into an activator without altering the target specificity of the normal protein (e.g. Sze et al., 1997; Rusch and Levine, 1997). By analogy to the Drosophila Snail repressor, it is conceivable that Ci-Sna normally promotes muscle differentiation by blocking the expression of inappropriate tissue determinants in muscle lineages. Consistent with this view is the finding that the Ci-sna/Ci-snaVP16 expression vector results in the ectopic induction of Ci-Bra/lacZ transgenes in the tail muscles, where they are normally never expressed. The simplest interpretation of the mutant phenotype is that there is a breakdown in the notochord/muscle boundary, and an intermixing of notochord and muscle cells during tail morphogenesis.

This mutant phenotype represents the first use of electroporation and gene fusion methods to assess gene function in an ascidian, although it has been reported that antisense oligonucleotides can produce specific mutant phenotypes in a distantly related ascidian, Molgula (Swalla and Jeffery, 1996). In addition, mutant phenotypes have been obtained in Halocynthia by RNA injection (e.g. Wada et al., 1997). In principle, the targeted expression of modified transgenes offers several advantages over these other methods. Most notably, RNA injection usually results in the ectopic expression of a given gene in a broad assortment of tissues, thereby obscuring issues of cell autonomy. As we characterize more tissue-specific enhancers in Ciona, it should be possible to misexpress regulatory genes in precise patterns.

Snail as a boundary repressor

Ectopic expression of Ci-Sna in the notochord does not produce an obvious mutant phenotype (Fig. 5) and in situ hybridization assays suggest that the endogenous Ci-Bra gene is not repressed (S. Fujiwara, unpublished results). Moreover, Ci-Sna fails to repress a lacZ transgene containing the full-length, 3.5 kb Ci-Bra promoter region. These observations raise the possibility that Ci-Sna might not be sufficient to establish or maintain the muscle-notochord boundary. Alternatively, it is conceivable that the expression vector used to misregulate Ci-sna in the notochord (the truncated 251 bp Ci-Bra enhancer) results in only low levels of Ci-Sna, which are sufficient to repress the minimal 434 bp Ci-Bra enhancer but are insufficient to repress the full-length gene. Evidence for this latter possibility stems from the observation that the misexpression of Ci-Sna using the full-length, 3.5 kb Ci-Bra promoter region causes a disruption in notochord intercalation (S. Fujiwara, unpublished results).

Despite these considerations, the preponderance of the evidence presented in this study suggest that Ci-Sna functions as a boundary repressor, which helps subdivide the embryonic mesoderm into distinct muscle and notochord lineages (Fig. 7B). Previous studies have shown that the Drosophila Snail repressor helps establish a boundary between the embryonic mesoderm and neurogenic ectoderm (Fig. 7A; see Leptin, 1991; Kosman et al., 1991). We propose that one or more members of a vertebrate snail-related gene family may also act as boundary repressors to help delineate the axial and paraxial mesoderm.

Fig. 7.

Ci-Snail is a boundary repressor. We propose that Snail proteins function as evolutionarily conserved repressors to establish boundaries between adjoining tissues. (A) In Drosophila, snail is expressed in ventral regions of early embryos where it establishes the limits of the presumptive mesoderm. The Snail repressor excludes the expression of neuroectodermal specification genes, such as rho and single minded, in the ventral mesoderm, and restricts their expression to lateral regions that form the neurogenic ectoderm. (B) This study provides evidence that Ci-Sna functions as a repressor, which creates a boundary between the notochord and tail muscles. The conversion of Ci-Sna into an activator (Ci-SnaVP16) appears to cause a breakdown in this boundary and results in severe disruptions in tail morphogenesis. It is conceivable that vertebrate snail genes, such as the sna-1 gene in zebrafish, also functions as boundary repressors to subdivide the embryonic mesoderm into axial (notochord) and paraxial (somites) lineages.

Fig. 7.

Ci-Snail is a boundary repressor. We propose that Snail proteins function as evolutionarily conserved repressors to establish boundaries between adjoining tissues. (A) In Drosophila, snail is expressed in ventral regions of early embryos where it establishes the limits of the presumptive mesoderm. The Snail repressor excludes the expression of neuroectodermal specification genes, such as rho and single minded, in the ventral mesoderm, and restricts their expression to lateral regions that form the neurogenic ectoderm. (B) This study provides evidence that Ci-Sna functions as a repressor, which creates a boundary between the notochord and tail muscles. The conversion of Ci-Sna into an activator (Ci-SnaVP16) appears to cause a breakdown in this boundary and results in severe disruptions in tail morphogenesis. It is conceivable that vertebrate snail genes, such as the sna-1 gene in zebrafish, also functions as boundary repressors to subdivide the embryonic mesoderm into axial (notochord) and paraxial (somites) lineages.

We thank N. Satoh and D. Martasian for providing animals, and A. Di Gregorio for critically reading the manuscript. This work was funded by a grant from the NSF (IBN-9514138).

Alberga
,
A.
,
Boulay
,
J. L.
,
Kempe
,
E.
,
Dennefeld
,
C.
and
Haenlin
,
M.
(
1991
).
The snail gene required for mesoderm formation in Drosophila is expressed dynamically in derivatives of all three germ layers
.
Development
111
,
983
992
.
Boyd
,
J. M.
,
Subramanian
,
T.
,
Schaeper
,
U.
,
La Regina
,
M.
,
Bayley
,
S.
and
Chinnadurai
,
G.
(
1993
).
A region in the C-terminus of adenovirus 2/5 E1a protein is required for association with a cellular phosphoprotein and important for the negative modulation of T24-ras mediated transformation, tumorigenesis and metastasis
.
EMBO J
.
12
,
469
478
Bruhn
,
L.
,
Munnerlyn
,
A.
and
Grosschedl
,
R.
(
1997
).
ALY, a context-dependent coactivator of LEF-1 and AML-1, is required for TCRα enhancer function
.
Genes Dev
.
11
,
640
653
.
Clements
,
D.
,
Taylor
,
H. C.
,
Herrmann
,
B. G.
and
Stott
,
D.
(
1996
).
Distinct regulatory control of the Brachyury gene in axial and non-axial mesoderm suggests separation of mesoderm lineages early in mouse gastrulation
.
Mech. Dev
.
56
,
139
149
.
Corbo
,
J. C.
(
1997
).
Molecular aspects of notochord and nervuous system development in the ascidian, Ciona intestinalis. Ph.D. Dissertation, UCSD
.
Corbo
,
J. C.
,
Levine
,
M.
and
Zeller
,
R. W.
(
1997a
).
Characterization of a notochord-specific enhancer from the Brachyury promoter region of the ascidian, Ciona intestinalis
.
Development
124
,
589
602
.
Corbo
,
J. C.
,
Erives
,
A.
,
Di Gregorio
,
A.
,
Chang
,
A.
and
Levine
,
M.
(
1997b
)
Dorsoventral patterning of the vertebrate neural tube is conserved in a protochordate
.
Development
124
,
2335
2344
.
Erives
,
A.
,
Corbo
,
J. C.
and
Levine
,
M.
(
1998
).
Lineage-specific regulation of the Ciona snail gene in the embryonic mesoderm and neuroectoderm
.
Dev. Biol
.
194
,
213
225
.
Gray
,
S.
,
Szymanski
,
P.
and
Levine
,
M.
(
1994
).
Short range repressors permit enhancers to function autonomously within complex promoters
.
Genes Dev
.
8
,
1829
1838
.
Gray
,
S.
and
Levine
,
M.
(
1996
).
Short-range repressors mediate both quenching and silencing in complex loci in the Drosophila embryo
.
Genes Dev
.
10
,
700
710
.
Hammerschmidt
,
M.
and
Nüsslein-Volhard
,
C.
(
1993
).
The expression of a zebrafish gene homologous to Drosophila snail suggests a conserved function in invertebrate and vertebrate gastrulation
.
Development
119
,
1107
1118
.
Herrmann
,
B. G.
and
Kispert
,
A.
(
1994
).
The T genes in embryogenesis
.
Trends Genet
.
10
,
280
286
.
Ip
,
Y. T.
,
Kraut
,
R.
,
Small
,
S.
,
Levine
,
M.
and
Rushlow
,
C.
(
1991
).
The dorsal morphogen is a novel DNA binding protein that interacts with a long-range repression element in Drosophila
.
Cell
64
,
439
446
.
Ip
,
Y. T.
,
Park
,
R.
,
Kosman
,
D.
and
Levine
,
M.
(
1992
).
The dorsal gradient morphogen regulates stripes of rhomboid expression in the presumptive neuroectoderm of the Drosophila embryo
.
Genes Dev
.
6
,
1728
1739
.
Jiang
,
R.
,
Copeland
,
N. G.
,
Gilbert
,
D. J.
,
Jenkins
,
N. A.
and
Gridley
,
T.
(
1997
).
Genomic organization and chromosomal localization of the mouse snail (Sna) gene
.
Mammalian Genome
8
,
686
688
.
Kasai
,
Y.
,
Nambu
,
J. R.
,
Lieberman
,
P. M.
and
Crews
,
S. T.
(
1992
).
Dorsal-ventral patterning in Drosophila: DNA binding of snail protein to the single-minded gene
.
Proc. Natl. Acad. Sci. USA
89
,
3414
3418
.
Kispert
,
A.
,
Koschorz
,
B.
and
Herrmann
,
B. G.
(
1995
).
The T protein encoded by Brachyury is a tissue-specific transcription factor
.
EMBO J
.
14
,
4763
4772
.
Kosman
,
D.
,
Ip
,
Y. T.
,
Levine
,
M.
and
Arora
,
K.
(
1991
).
The establishment of the mesoderm-neuroectoderm boundary in the Drosophila embryo
.
Science
254
,
118
122
.
Kunkel
,
T. A.
(
1985
).
Rapid and efficient site-specific mutagenesis without phenotypic selection
.
Proc. Natl. Acad. Sci. USA
82
,
488
492
.
Latinkic
,
B. V.
,
Umbhauer
,
M.
,
Neal
,
K. A.
,
Lerchner
,
W.
,
Smith
,
J. C.
and
Cunliffe
,
V.
(
1997
).
The Xenopus Brachyury promoter is activated by FGF and low concentrations of activin and suppressed by high concentrations of activin and by paired-type homeodomain proteins
.
Genes Dev
.
11
,
3265
3276
.
Leptin
,
M.
(
1991
).
twist and snail as positive and negative regulators during Drosophila mesoderm development
.
Genes Dev
.
5
,
1568
1576
.
Nauber
,
U.
,
Pankratz
,
M. J.
,
Kienlin
,
A.
,
Seifert
,
E.
,
Klemm
,
U.
and
Jäckle
,
H.
(
1988
).
Abdominal segmentation of the Drosophila embryo requires a hormone receptor-like protein encoded by the gap gene knirps
.
Nature
336
,
489
492
.
Nibu
,
Y.
,
Zhang
,
H.
and
Levine
,
M.
(
1998
).
Unrelated short-range repressors interact with CtBP in the Drosophila embryo. Science
(in press).
Ros
,
M. A.
,
Sefton
,
M.
and
Nieto
,
M. A.
(
1997
).
Slug, a zinc finger gene previously implicated in the early patterning of the mesoderm and the neural crest, is also involved in chick limb development
.
Development
124
,
1821
1829
.
Rusch
,
J.
and
Levine
,
M.
(
1997
).
Regulation of a dpp target gene in the Drosophila embryo
.
Development
124
,
303
311
.
Sambrook
,
J.
,
Fritsch
,
E. F.
and
Maniatis
,
T.
(
1989
).
Molecular Cloning, A Laboratory Manual
.
Cold Spring Harbor Laboratory Press, 2nd Edition
.
Schaeper
,
U.
,
Boyd
,
J. M.
,
Verma
,
S.
,
Uhlmann
,
E.
,
Subramanian
,
T.
and
Chinnadurai
,
G.
(
1995
).
Molecular cloning and characterization of a cellular phosphoprotein that interacts with a conserved C-terminal domain of adenovirus E1A involved in negative modulation of oncogenic transformation
.
Proc. Natl. Acad. Sci. USA
92
,
10467
10471
.
Schulte-Merker
,
S.
and
Smith
,
J. C.
(
1995
).
Mesoderm formation in response to Brachyury requires FGF signalling
.
Curr. Biol
.
5
,
62
67
.
Smith
,
D. B.
and
Johnson
,
K. S.
(
1988
).
Single-step purification of polypeptides expressed in Escherichia coli as fusions with glutathione S-transferase
.
Gene
67
,
31
40
.
Sollerbrant
,
K.
,
Chinnadurai
,
G.
and
Svensson
,
C.
(
1996
).
The CtBP binding domain in the adenovirus E1A protein controls CR1-dependent transactivation
.
Nucleic Acids Research
24
,
2578
2584
.
Swalla
,
B. J.
and
Jeffery
,
W. R.
(
1996
).
Requirement of the Manx gene for expression of chordate features in a tailless ascidian larva
.
Science
274
,
1205
1208
.
Sze
,
J. Y.
,
Liu
,
Y.
and
Ruvkun
,
G.
(
1997
).
VP16-activation of the C. elegans neural specification transcription factor UNC-86 suppresses mutations in downstream genes and causes defects in neural migration and axon outgrowth
.
Development
124
,
1159
1168
.
Wada
,
H.
,
Holland
,
P. W.
,
Sato
,
S.
,
Yamamoto
,
H.
and
Satoh
,
N.
(
1997
).
Neural tube is partially dorsalized by overexpression of HrPax-37: the ascidian homologue of Pax-3 and Pax-7
.
Dev. Biol
.
187
,
240
252
.
Yasuo
,
H.
and
Satoh
,
N.
(
1994
).
An ascidian homolog of the mouse Brachyury (T) gene is expressed exclusively in notochord cells at the fate restricted stage
.
Dev. Growth Differ
.
36
,
9
18
.