A cDNA clone for a sea cucumber T-box gene was isolated and characterized. Based on molecular phylogenetic analysis it is concluded that the putative gene Hl-Tbr, encoded by the cDNA clone, is a T-box gene of the T-brain subfamily and hence a homolog of the mouse T-brain-1 (Tbr-1) as well as of the Xenopus Eomesodermin. In situ hybridization analysis of whole mount specimens showed that Hl-Tbr was expressed in the invaginated cells at the early gastrula stage and the expression of the gene was scarcely detectable by the end of the late gastrula stage.
INTRODUCTION
T-box genes encode a family of transcription factors that are characterized by a putative DNA-binding domain (Herrmann and Kispert, 1994; Bollag et al., 1994; Papaioannou and Silver, 1998), known as the T-domain. Studies have shown that they are conserved in a wide variety of animals and appear to play a crucial role in development (Herrmann and Kispert, 1994; Bollag et al., 1994; Papaioannou and Silver, 1998; Smith, 1999). The present study first reports the isolation and characterization of a sea cucumber T-box gene, and its expression during gastrulation.
MATERIALS AND METHODS
The sea cucumber Holothuria leucospilota was used. Fertilization was done as noted previously (Maruyama, 1980), and embryos were reared at 27–28°C. Primers for PCR amplification were 5′-TA(C/T)AT(C/T)CA(C/T)CC(C/T)CC(A/C/G/T)GA(C/T)TC(A/C/G/T)CC-3′ as the sense-strand oligonucleotide and 5′-(A/G)AA(A/C/G/T)GC(C/T)TT(A/C/G/T)GC(A/G)AA(A/C/G/T)GG(A/G)TT-3′ as the antisense oligonucleotide (Tagawa et al., 1998). Target fragments were amplified from a Holothuria leucospilota 25-hr gastrula cDNA library constructed in λ ZAPII (Stratagene). PCR products were sequenced with an ABI PRISM dye primer cycle sequencing kit (Perkin Elmer). Using PCR-derived clones, the cDNA library was screened under high stringency conditions. A cDNA clone was obtained and used for sequencing. Both strands for the cDNA clone were sequenced. In situ hybridization of whole-mount specimens proceeded with DIG-labeled sense (control) and antisense RNA probes, synthesized in vitro from a full-length cDNA clone.
Molecular phylogenetic relationships among T-box gene products were inferred by the neighbor-joining method (Saitou and Nei, 1987), analyzed using the software package PHYLIP (Felsenstein, 1985). Alignments of multiple protein sequences were made using CLUSTAL W (Thompson et al., 1994). Regions of gaps and questionable homology were excluded from the analyses. Accession numbers in the DDBJ, EMBL or GenBank nucleotide sequence databases for the sequences were: ascidian As-T2, D83265; mouse m-Tbx6, U57331; Xenopus X-vegT, U59483; mouse m-Tbx5, U57330; mouse m-Tbx4, U57329; Drosophila Dm-omb, M81796; mouse m-Tbx2, U15566; Eomesodermin (Xenopus Eomesodermin), U75996; human hu-Eomesodermin, AJ010280; mouse m-Eomesodermin, AF013281; human hu-Tbr-1, U49250; mouse m-Tbr-1, U49251; Drosophila DmTrg, S74163; ascidian As-T, D16441; starfish ApBra, AB018527; sea urchin HpTa, D50332; Xenopus Xbra, M77243; amphioxus AmBra-1, X91903; acorn worm PfBra, AB004912. Accession numbers in SWISS-PROT database were: mouse m-Tbx1, P70323; mouse m-Tbx3, P70324; zebrafish Zf-T, Q07998; amphioxus AmBra-2, P80492; mouse m-T, P20293.
RESULTS AND DISCUSSION
Using a PCR-based method (see Materials and Methods), I isolated a cDNA clone for a sea cucumber T-box gene (Hl-Tbr) from the Holothuria leucospilota gastrula cDNA library. The nucleotide and deduced amino acid sequences of the cDNA clone for Hl-Tbr are shown in Fig. 1. The cDNA was 3363 base pairs long and had an open reading frame that predicted a polypeptide of 679 amino acids, in which nucleotide A is present at position −3 as in most translation initiation sites in echinoderms (Mankad et al., 1998). The T-domain was evident in the middle portion of the predicted protein (Fig. 1), and was highly conserved when compared with T-domains of other T-box gene products (data not shown).
On the basis of the molecular phylogenetic analysis, the T-box gene family is grouped into at least five subfamilies, Brachyury (T), Tbx1, Tbx6, Tbx2 and Tbr1 (Papaioannou and Silver, 1998): the Tbr1 subfamily will be called the T-brain subfamily in this study (see below). For example, the Brachyury subfamily contains two T-box genes isolated from sea urchins (HpTa; Harada et al., 1995) and starfish (ApBra; Shoguchi et al., 1999) of echinoderms (see Fig. 2).
To determine the subfamily to which the sea cucumber homolog (Hl-Tbr) belongs, I aligned 117 amino acid sites of the T-domains based upon maximum similarity, by which molecular phylogenetic analysis was performed by the neighbor-joining method (Saitou and Nei, 1987). As seen in Fig. 2, Hl-Tbr is grouped as a T-brain subfamily gene. This subfamily is comprised of mouse and human T-brain-1 or Tbr-1 (Bulfone et al., 1995), Eomesodermin (Xenopus Eomesodermin) (Ryan et al., 1996), mouse Eomesodermin (Wattler et al., 1998; Ciruna and Rossant, 1999; Hancock et al., 1999), human Eomes (Yi et al., 1999), and Hl-Tbr. This clade was supported by the highest bootstrap value (100%; Felsenstein, 1985). In addition, it appears that the Tbr1 group or the T-brain sub-family except for the sea cucumber gene is comprised of two subgroups, Tbr-1 and Eomesodermin (see Fig, 2). From these results, it is concluded that the sea cucumber gene, Hl-Tbr, is a homolog of Tbr-1 genes (mouse Tbr-1 and human Tbr-1) as well as of Eomesodermin genes (Xenopus Eomesodermin, mouse Eomesodermin and human Eomes).
It is well-known that these vertebrate T-brain genes are expressed in a region of the developing forebrain; in mouse Tbr-1 (Bulfone et al., 1995), in Xenopus Eomesodermin (Ryan et al., 1998) and in mouse Eomesodermin (Ciruna and Rossant, 1999; Hancock et al., 1999). In addition, a transient expression mainly during gastrulation is also reported in the Eomesodermin subgroup (Ryan et al., 1996; Ryan et al., 1998; Ciruna and Rossant, 1999; Hancock et al., 1999), while no such expression is reported in the Tbr-1 subgroup (mouse Tbr-1).
Expression of the sea cucumber T-brain gene, Hl-Tbr, in embryos under gastrulation was examined by whole-mount in situ hybridization (Fig. 3). No signals above background levels were detectable when control (sense) probes were used. Transcripts of Hl-Tbr were detected in the invaginated cells of early gastrulae (Fig. 3A, B). It appears that the territory of the gene expression may demarcate that of the initial invagination (cf. Fig. 3A, B). In advanced gastrulae (Fig. 3C), the signal is seen only around the middle-to-upper portion of the archenteron. I note that some archenteric cells persisted in exhibiting more or less intense signals (see Fig. 3C). Then, by the end of the late gastrula, signals were barely detectable in the embryos (Fig. 3D). In addition to the unambiguous expression in the initially invaginated cells of the archenteron, a very weak staining, sometimes, that of background level, was noticed in animal hemispheres, especially on one side of the lateral gastrula-wall (data not shown); this pattern of staining disappeared in advanced gastrulae, too. An animal polar region in the gastrulae did not exhibit such signals which may suggest a specialized region (for example, the apical plate region as is well-known in sea urchin embryos).
The expression of the sea cucumber T-brain subfamily T-box gene during gastrulation was described in this study. The predominant expression in the invaginated cells at the initial step of archenteron formation may suggest a role for the gene in this morphogenetic process as well as fate decision processes of the cells, giving rise to larval mesoderm and probably some portion of the larval endoderm. In addition, I note that the expression pattern of the starfish T-brain gene, Ap-Tbr, during gastrulation has been revealed (Shoguchi et al., 2000), and that the predominant expression pattern of Ap-Tbr at the early gastrula stage quite resembles that of Hl-Tbr.
Acknowledgments
The author is indebted to Prof. N. Satoh for his guidance and assistance during the present study. Molecular cloning experiments were carried out at Department of Zoology, Kyoto University and the author is grateful to the staff of the department. The author also thanks Dr. H. Takahashi for valuable suggestions on molecular techniques, Dr. K. Tagawa for providing the degenerate primers, and Mr. E. Shoguchi for help in preparing the molecular phylogenetic tree.