Results and discussion

Large-scale EST analyses of Ciona intestinalis

For the last several years, we have conducted large-scale EST analyses of mRNAs expressed in Ciona intestinalis embryos and adults (Satou et al. 2002a, 2002b). The first series of the analyses included the five developmental stages of fertilized eggs (Nishikata et al. 2001), cleaving embryos (16–110 cell stages) (Fujiwara et al. 2002), tailbud embryos (Satou et al. 2001), swimming larvae (Kusakabe et al. 2002), and young adults about 2 weeks after metamorphosis (Ogasawara et al. 2002), and testis (Inaba et al. 2002). The second series expanded the first one by adding more data on the previously examined developmental stages, and by adding new data on mRNAs expressed in gastrulae/neurulae, and in the gonad (ovary), endostyle, neural complex, heart, and blood (coelomic) cells of the adult. The number of ESTs so far obtained is summarized in Table 1. A total of 480,753 ESTs, comprising 241,519 ESTs of the 3′ end and 239,234 ESTs of the 5′ end, have been obtained. The 3′ ESTs were used to examine the overlap of the clones in order to categorize the cDNAs into independent (non-duplicated) clusters (groups of clones). Altogether, after examining the overlap of all developmental stages and adult organs, the 241,519 3′ ESTs fell into 17,834 clusters. The reading of the Ciona intestinalis draft genome has predicted 15,832 protein-coding genes in the genome (Dehal et al. 2002), suggesting that 85% or more transcripts of this ascidian have been identified by the EST analyses.

Table 1. ESTs of transcripts expressed in Ciona intestinalis
Full size table

EST counts and temporal expression pattern of Ciona developmental genes during embryogenesis

Because the libraries were not amplified or normalized, the appearance of cDNA clones (or EST counts) occurs in proportion to their abundance at the corresponding developmental stage. Therefore, if the EST counts of a certain gene are determined at the six different developmental stages, they may represent the temporal expression pattern of the gene. Table 2 summarizes the temporal expression patterns of 14 genes so far reported in Ciona embryos as well as five developmentally regulated genes, and their EST counts at the six developmental stages. For example, Ci-pem is a Ciona intestinalis homolog of Cs-pem (Ciona savignyi-posterior end mark) (Yoshida et al. 1996). pem is so named because the maternal transcripts of the gene are localized at the posterior end of fertilized eggs and early embryos. The transcript is segregated into the posterior-most blastomeres at early gastrulae, and eventually inherited by a few endodermal strand cells, which are likely to be the presumptive germ cells (Fujimura and Takamura 2000). Therefore, it is expected that a large amount of pem maternal transcripts would be present in eggs, and that this amount would decrease as development proceeds, but that some transcripts would still remain at the larval stage. As shown in Table 2, the EST count of Ci-pem was maximum at the egg stage and decreased gradually during embryogenesis, but a few counts were still evident at the larval stage, and the count finally disappeared in the adult. Thus, the EST count of Ci-pem corresponded fairly well with the temporal expression pattern of the gene.

Table 2. EST counts of Ciona genes in cDNA libraries derived from whole animals/embryos at six different developmental stages
Full size table

Ci-macho1 is a Ciona intestinalis homolog (Satou et al. 2002c) of a muscle-determinant gene (macho-1) encoding a Zic-related zinc finger transcription factor of another ascidian, Halocynthia roretzi (Nishida and Sawada 2001). macho-1 is expressed only maternally, whereas Ci-macho1 is expressed maternally and zygotically, and the zygotic expression begins in cells of the CNS at the early tailbud stage (Satou et al. 2002c). As is evident in Table 2, transcripts (as indicated by EST counts) of Ci-macho1 were detected at the egg stage, not detected at the cleavage and gastrula stages, and then detected again at the tailbud and larval stages. Another example is the Ci-sna or snail gene of Ciona intestinalis (Fujiwara et al. 1998). This gene has been reported to be expressed zygotically and transiently at the cleavage and early gastrula stages in blastomeres giving rise to larval muscle and mesenchyme cells. The EST counts of Ci-sna were detectable in the cleaving embryos and gastrulae/neurulae, indicating a good correlation between Northern blot and EST counts.

Ci-Bra is a Brachyury gene, which is essential for the notochord differentiation of the ascidian embryo (Corbo et al. 1997). The gene begins to be expressed at the 64-cell stage and its expression is downregulated by the late tailbud stage. Table 2 indicates that the EST of Ci-Bra was obtained at the cleavage, gastrulae/neurulae and tailbud stages only. This is accordance with the known temporal expression pattern of Ci-Bra. In addition, the Ci-fibrinogen-like gene is one of the Ci-Bra targeted genes (Hotta et al. 2000) that are expressed in the differentiating notochord cells in the tailbud embryo (Imai et al. 2002a). It had a detectable EST count only at the tailbud stage (Table 2). This is another example of a good correlation between the temporal expression pattern of the gene and the EST count. Here we describe in detail 6 of the 14 developmentally regulated genes listed in Table 2. For the remaining eight genes, there was also a good correlation between the temporal expression profile determined by Northern blotting or in situ hybridization and the EST counts.

Table 2 also includes five examples of EST counts of developmentally regulated genes whose function is still unknown. For example, the ID 13304 gene encodes a protein with no sequence similarity to known proteins. Its EST count was detectable only in eggs, suggesting that the gene is expressed only maternally. The ID 11523 gene also encodes a protein with no significant similarity with known proteins. Because its EST count was detectable at the cleavage stage only, it is highly likely that this gene is expressed zygotically and transiently in cleaving embryos. Another example is the ID 02714 gene, whose EST count was detectable only in young adults, suggesting that this gene is not expressed during embryogenesis.

EST counts and spatial expression of Ciona developmental genes in organs and/or tissues of the adult

In addition to analyzing transcripts expressed at the six developmental stages, we carried out large-scale EST analyses of transcripts expressed in organs and/or tissues of Ciona adults (Table 1). Analysis of the EST counts of each gene suggests its spatial expression pattern. For example, Etani and Nishikata (2002) reported that Ci-Nut is expressed specifically in the larval neural tube. As shown in Table 3, the EST count of Ci-Nut was detectable in the adult gonad, indicating that the gene is also expressed in the gonad and thus may play role(s) in gametogenesis. Ci-CA8 encodes a type of cytoplasmic actin. The fact that its EST count was only detectable in the testis suggests that it is specifically expressed there, and thus that it may function in sperm formation. Both Ci-MHC (myosin heavy chain) 2 and Ci-TPM (tropomyosin) 3 showed high EST count scores in the heart library. At present we have not examined ESTs derived from the body-wall-muscle, and therefore the possibility remains that Ci-MHC2 and Ci-TPM3 are also expressed there. However, it is likely that the two genes are strongly expressed in the heart. The EST count of a gene encoding an orphan bHLH protein was detectable in the neural complex, while that of Ci-FoxI was detectable in the blood cells (Table 3). Thus, Ci-orphan bHLH may be specific to the neural complex, and Ci-FoxI specific to blood cells.

Table 3. EST counts in cDNA libraries derived from the Ciona adult organ or tissue
Full size table

On the other hand, the EST counts of β-catenin, Ci-HIF, orphan Fox, and Ci-HMG1/2 were detectable in various libraries, although the testis library showed no EST counts for Ci-HIF or orphan Fox, and the endostyle library lacked detectable EST counts for Ci-HIF and Ci-HMG1/2. These genes are expressed more ubiquitously in order to play roles in various tissues and organs of Ciona adults.

It is sometimes difficult to prepare samples of certain isolated tissues or organs without contamination by other tissues. Therefore, the possibility remains that a given library (and therefore the EST counts) contains some transcripts originating from unrelated tissues. Nevertheless, on the whole, the present large-scale EST analysis of mRNAs expressed in various organs and tissues of the Ciona adult strongly suggests that the EST counts reflect the spatial expression pattern of a given gene. Further EST analyses of mRNAs expressed in the epidermis, body-wall muscle, pharyngeal basket, and digestive system are now underway to clarify the expression patterns of Ciona intestinalis genes.

Conclusion

Altogether, very large-scale EST analyses of Ciona intestinalis provide information about the profiles of gene expression. In many cases, these analyses can be used in place of Northern blot analyses. Even if it does not strictly coincide with the result of Northern blot analyses, the profile of the EST counts clearly suggests the temporal and spatial expression patterns of a given gene. In addition, the information about EST counts may be used in various ways. For example, if a gene shows detectable EST counts only at the tailbud stage, it may be sufficient to examine its spatial expression pattern only at this stage by in situ hybridization, without performing in situ hybridization in embryos at other developmental stages.

The draft genome of Ciona intestinalis has been sequenced (Dehal et al. 2002). During a series of studies to identify developmentally regulated genes in the Ciona intestinalis genome (e.g., Satou et al. 2003; Wada et al. 2003), we have concluded that the strategy of sequencing the genome is more precise than Southern blot analyses for determining the number of copies of a given gene in the Ciona genome. For example, Southern blot analysis of Ci-Tbx6 yielded only one major band, suggesting that Ci-Tbx6 is a single-copy gene (Mitani et al. unpublished). In contrast, the Ciona intestinalis genome survey demonstrated the presence of four copies of the Ci-Tbx6 gene (Dehal et al. 2002). This means that Southern blot analysis is not required any longer in this ascidian. In addition, demonstrated by the present study, Northern blot analysis is not generally required any longer to analyze gene expression patterns in Ciona intestinalis.

The final goal of our research on Ciona intestinalis is first to obtain a complete set of information on the temporal and spatial expression of every developmentally regulated gene and then to discover the molecular pathways that govern the complex network of the expression of these genes. Ultimately, when large-scale in situ hybridization can be performed to cover nearly all the Ciona intestinalis genes, Ciona may provide an experimental system for exploring the genetic network that governs the formation of the basal chordate body plan.