Enhancer detection in the ascidian Ciona intestinalis with transposase-expressing lines of Minos



Germline transgenesis with a Tc1/mariner superfamily Minos transposon was achieved in the ascidian Ciona intestinalis. Transgenic lines that express transposases in germ cells are very useful for remobilizing transposon copies. In the present study, we created transposase-expressing lines of Minos in Ciona. A Ciona gene encoding protamine (Ci-prm) is expressed in the testes and sperm. Transgenic lines expressing Minos transposase in the testes and sperm were created with a cis-element of Ci-prm, and used for enhancer detection. Double-transgenic animals between transposase lines and a transgenic line with an enhancer detection vector passed on several independent enhancer detection events to subsequent progeny. This technique allowed us to isolate transgenic lines that express GFP in restricted tissues. This system provides an easy and efficient method for large-scale enhancer detection in Ciona intestinalis. Developmental Dynamics 237:39–50, 2008. © 2007 Wiley-Liss, Inc.


Ciona intestinalis (hereafter referred to as Ciona) is a cosmopolitan species of ascidians. Ascidians are members of a subphylum group, Tunicata, which belongs to the phylum Chordata together with vertebrates and cephalochordates. This phylum shares characteristics such as a notochord and a dorsal hollowed neural tube (Satoh, 1994, 2003). Ascidians display the basic chordate body plan. Among ascidians, Ciona is a suitable organism upon which to conduct biological studies based on genetic approaches. The life cycle of Ciona is about 2 to 3 months. This relatively short life cycle enables culturing in laboratories; in fact, successful inland culturing systems of Ciona have been reported (Hendrickson et al., 2006; Joly et al., 2007). Furthermore, the draft genome of Ciona has been sequenced (Dehal et al., 2002). The Ciona genome is comparatively small and compact, approximately 160 Mbp per haploid, and contains 15,852 protein-coding genes. The genome size and gene number are comparable to that of Drosophila melanogaster. Such genomic compactness is a splendid advantage for researchers experimenting with genetic approaches.

Genetic approaches in Ciona have been developed very recently (Moody et al., 1999; Nakatani et al., 1999). A recent technical innovation in Ciona genetics was the achievement of germline transgenesis with a Tc1/mariner superfamily Minos transposon (Sasakura et al., 2003a, b). The Tc1/mariner superfamily is a famous transposon family because of their loose host specificity (Plasterk et al., 1999; Sasakura et al., 2007). In fact, several Tc1/mariner transposons have shown their activity in both protostomes and deuterostomes (Loukeris et al., 1995a, b; Klinakis et al., 2000a, b; Shimizu et al., 2000; Zhang et al., 2002; Drabek et al., 2003; Pavlopoulos and Averof, 2005; Metaxakis et al., 2005). Genetic techniques such as germline transgenesis, enhancer detection (also called “enhancer trapping”), and insertional mutagenesis have been established in Ciona with Minos (Sasakura et al., 2003b, 2005; Awazu et al., 2004).

Enhancer detection is a kind of “local effect,” in which reporter genes in the transposon are entrapped by endogenous enhancers near the transposon insertions, and the expression patterns of reporter genes are affected according to the enhancers (O'Kane, 1998). Enhancer detection with transposons is useful for detecting endogenous enhancers as well as for creating tissue- and organ-specific marker lines. For efficient enhancer detection, special transposon vectors with a minimal or basal promoter showing very weak expression are utilized. In Ciona, a minimal promoter from Ci-TPO, which encodes thyroid peroxidase (Ogasawara et al., 1999), has been used for enhancer detection, and various enhancer detection lines have been established (Awazu et al., 2007).

Transposon-based techniques have become more sophisticated by novel technical innovations. In Ciona, the first transposon-mediated transgenic lines and enhancer detection lines were created by co-microinjection or co-electroporation of transposon DNA and transposase mRNA (Sasakura et al., 2003b; Matsuoka et al., 2005). The technique has been refined by the achievement of remobilization of transposons within the Ciona genome (Awazu et al., 2007). Recently, our group has reported that remobilization of Minos occurs by microinjection of transposase mRNA into eggs of transgenic lines that harbor transposon insertions in their genomes. This method has been applied to efficient enhancer detection. A transgenic line that harbors a concatemer of Minos vectors for enhancer detection has been created, and the transposons were remobilized by microinjection of transposase mRNA. Approximately 79% of transposase-injected animals passed on novel enhancer detection insertions to subsequent generations. Despite its power, one technical disadvantage of this method is that transposase mRNA has to be introduced by microinjection, which is relatively difficult and time-consuming. Further refining of this technique is still required.

An ultimate way to cause remobilization of transposons is to utilize a transgenic line that expresses transposase in its germ cells (hereafter we call such lines “transposase lines”). In this system, remobilization of transposons can be achieved by creating double-transgenic animals between a transposase line and a transgenic line that harbors transposon insertions in its genome. Transposase should be expressed in the germ cells of double-transgenic animals, and this should cause the cutting-and-pasting of transposons to create new insertions. The technical advantage of this method is the non-requirement of laborious procedures such as microinjection and electroporation. In Ciona, a transgenic line expressing Minos transposase in its germ cells has not yet been created due to several technical limitations. One is that a cis-element responsible for expression in germ cells has not been isolated. Namely, a cis-element that promotes transposase expression in germ cells is required. The second reason is that a highly efficient and easy transformation technology has not been established for systems other than the Minos system. To create Minos transposase lines, utilization of Minos itself is not ideal, because Minos transposase would make Minos insertions in the genome unstable. Another transformation technique is necessary. One candidate would be I-SceI-mediated transgenesis as reported in C. savignyi (Deschet et al., 2003). This method is also applicable to C. intestinalis (Awazu et al., 2007; Sasakura, 2007). However, this method is not convenient and our trials so far have failed.

Recently, we reported that electroporation of circular plasmid DNA caused plasmid insertion in Ciona (Matsuoka et al., 2005). In this technique, DNA that has been electroporated into Ciona eggs is inserted in their genome without exogenous enzymes (i.e., transposases and restriction enzymes), and the insertions are passed on to subsequent generations. This technique, called “electroporation-mediated insertion,” is less efficient than transposon-mediated transformation (Sasakura, 2007). The advantage of this method is convenience, in that electroporation is the only procedure that can introduce DNA into hundreds of eggs within one hour (Corbo et al., 1997). This method can be utilized to create a transposase line of Minos.

In the present study, we isolated a cis-element that drives genes in sperm. With this cis-element, we have created transgenic lines that express Minos transposase in germ cells with electroporation-mediated insertion. These lines were used to create enhancer detection lines by crossing with a transgenic line that harbors a concatemer of the Minos enhancer detection vector in its genome (Awazu et al., 2007). With this method, efficient enhancer detection was achieved.


Ciona Protamine Gene Is Expressed in Testes and Sperm

The creation of Minos transposase lines requires isolation of a cis-element responsible for expression in germ cells. Among the many genes expressed in germ cells, we took notice of the gene that encodes protamine, a sperm-specific histone (reviewed in Eirin-Lopez et al., 2006), for several reasons. First, a transgenic line that expresses transposase in the sperm has an advantage over those expressing transposase in eggs. After creation of double-transgenic animals with a transgenic line that has transposon insertions, we have to isolate their germ cells with transposase expression to cross with wild-type counterparts. Although Ciona is hermaphroditic, maturation of sperm is much faster and much more frequent than that of eggs. This means that a line expressing transposase in the sperm can achieve more rapid screening of newly created insertions than those with transposase expression in eggs. Second, creation of a transposase line of another Tc1/mariner transposon, Sleeping Beauty (SB), with a protamine cis-element has been described in mice (Fischer et al., 2001). Successful creation of transposase lines with this cis-element suggested that this method would be suitable for our purposes in the present study. Sperm-specific expression of a protamine gene is conserved among organisms, and thus protamine of Ciona may also be expressed in sperm. An orthologous Ciona gene encoding protamine has been described (Ci-prm, gene model ci0100153586 in JGI Ciona genome database ver1.0; Lewis et al., 2004). EST analyses indicate that Ci-prm is strongly expressed in the testes (Ghost Database: http://hoya.zool.kyoto-u.ac.jp/cgi-bin/gbrowse/ci).

The expression of Ci-prm in sperm was confirmed by in situ hybridization. Strong signals were observed in the testes, which were hybridized with antisense Ci-prm probes (Fig. 1A). Testes hybridized with sense probes did not show such signals (Fig. 1B). A cross-section indicated that the middle region of the testes emitted the signal, whereas the peripheral regions did not show visible signals. Spermatids that form flagella are gathered densely in the middle region of the testes (Fig. 1C, Fl). Therefore, we concluded that Ci-prm is expressed in the testes and cells at the later stages of spermatogenesis. This gene can be utilized to create a transposase line.

Figure 1.

Expression of Ci-prm. A: A testis hybridized with antisense Ci-prm probe. Signals are shown in purple. Scale bar = 20 μm. B: A testis hybridized with sense Ci-prm probe. No signal is detected. C: Hematoxylin staining of a testis. Spermatids that are forming flagella (Fl) are gathered at the central region. Scale bar = 20 μm. D: GFP expression in testes of the Tg[MiCiTnIGCiprmG]2 transgenic line. Testes are outlined by white lines. GFP signals are observed at the central region of the testes. Scale bar = 100 μm. E,F: Sperm of Mi[CiTnIGCiprmG]2. A bright field image is shown in E and a fluorescent image in F. Note that all of the sperm show GFP signals in the head and tail. Scale bar = 10 μm.

A cis-Element of Ci-prm Is Responsible for Expression in the Male Germ Cells

To isolate the cis-element of Ci-prm, an approximately 2-kbp-long upstream segment of Ci-prm was amplified by PCR. To determine whether this element can drive gene expression in sperm or its primordial cells, this element was fused with the gfp reporter gene, and this cassette was subcloned into pMiCiTnIG to create pMiCiTnIGCiprmG (Fig. 2A). pMiCiTnIG contains the upstream sequence of the Ciona troponin I gene (Davidson and Levine, 2003), and this construct expresses GFP in both larval and adult muscle cells. Two stable transgenic lines of pMiCiTnIG did not express GFP in the testes (data not shown). Therefore, the CiTnIG cassette was used as a selection marker of transgenic animals before sperm maturation.

Figure 2.

The structure of two transposon constructs: (A) pMiCiTnIGCiprmG and (B) pSBFr3dTPORCiprmMiTP. Inverted terminal repeats (ITRs) of transposons are shown by black arrowheads. pBluescript backbones are not included in this diagram. NLS, nuclear localization signal; Ter, SV40 transcription termination sequences; MiTP, Minos transposase cDNA. An enhancer from an intron of Ci-musashi (Fr3 in Awazu et al., 2004) was included in this construct to enhance expression of DsRed from the Ci-TPO promoter.

We have created several transgenic lines having insertions of MiCiTnIGCiprmG in their genomes. All of these lines showed GFP expression in the testes (Fig. 1D). GFP signals were observed in the central region but not in the peripheral regions of the testes. This pattern is consistent with the pattern of in situ hybridization of Ci-prm, suggesting GFP expression in the primordial cells of sperm. To support this, we confirmed that fully matured sperm present in the sperm duct also showed GFP expression (Fig. 1E). All sperm of heterozygous Tg[MiCiTnIGCiprmG] transgenic lines expressed GFP. These sperm were used to fertilize wild-type eggs. Half of the progeny showed GFP expression in the muscle (data not shown), supporting the heterozygosity of MiCiTnIGCiprmG insertions in these lines.

Enhancer Detection With Transposase Lines

The results mentioned above showed that the 2-kbp-long upstream region of Ci-prm can drive the expression of downstream genes in sperm and its primordial cells. To create transposase lines of Minos, the cis-element of Ci-prm was fused with transposase cDNA, and this cassette was then subcloned into an SB vector, pSBFr3dTPOR-DestF, to create pSBFr3dTPORCiprmMiTP (Fig. 2B). The Fr3dTPOR cassette is a selection marker of the transposon vector that expresses RFP in the endostyle, peripharyngeal bands, retropharyngeal bands, and muscle after metamorphosis (Fig. 3A). The SB element (Ivics et al., 1996) was first chosen to increase the efficiency of the transformation of Ciona, but further experiments showed that SB is not sufficiently active in Ciona to cause germline transformation (Sasakura et al., 2007). Thus, transformation with this vector was carried out by electroporation-mediated plasmid insertion (Matsuoka et al., 2005), and some electroporated animals transmitted the transgene to the subsequent generation. The transformation frequency was 9.7% of the electroporated animals (n = 41). Among them, we used two transgenic lines, namely Ju[SBFr3dTPORCiprmMiTP]1 and 2. Sperm maturation of the Ju[SBFr3dTPORCiprmMiTP] lines seemed normal (Fig. 3C) and their sperm was fertile, suggesting that Minos transposase did not have a toxic effect on sperm formation or function. RT-PCR of these transgenic lines was performed to examine whether transposase mRNA is expressed in the testes, and strong PCR bands appeared in the testes and intestine tissues (Fig. 3D). Faint expression was also observed in the pharyngeal gill and endostyle (Fig. 3D). Taking the activity of the cis-element of Ci-prm into consideration, this PCR result suggests the expression of transposase in the testes.

Figure 3.

Transgenic lines utilized for enhancer detection screening. A: RFP expression of Ju[SBFr3dTPORCiprmMiTP]2. RFP is expressed in the endostyle, peripharyngeal band, retropharyngeal band and muscle cells. En, endostyle; Mu, muscle; Or, oral siphon; PB, peripharyngeal band; RB, retropharyngeal band. B: GFP expression of Mu[MiTSAdTPOG]1. GFP is expressed in the anterior and posterior termini of the endostyle (arrows) and in the retropharyngeal band (RB). C: Sperm of Ju[SBFr3dTPORCiprmMiTP]1. Their overall structure is normal. D: Transposase is expressed in tissues including the testes of transposase lines. MiTP, Minos transposase; DsRed, RFP. GAPDH is used as positive controls. “RT-“ is a negative control without reverse transcription. T, testes and intestine; E, endostyle; G, pharyngeal gill. Total RNA isolated from these tissues was subjected to RT-PCR. Because complete isolation of testes from intestine is difficult, RNA was isolated from their complex. Ju[SBFr3dTPORCiprmMiTP]1 and 2 are shown as Line1 and Line2, respectively. “Wild” indicates RNA isolated from wild-type animals. W, RNA isolated from whole body. T, RNA isolated from testes and intestine. For wild types, poly (A) selected RNA was used for RT-PCR. The lengths of PCR bands are given in the right-hand column.

Ju[SBFr3dTPORCiprmMiTP]1 and 2 were used for enhancer detection, and both lines showed similar results. The screening procedure is described in Figure 4. Mu[ISMiTSAdTPOG]1 is a transgenic line harboring a concatemer of the Minos enhancer detection vector, pMiTSAdTPOG, in its genome (Awazu et al., 2007). pMiTSAdTPOG contains a minimal promoter of Ci-TPO, and shows weak expression at the anterior and posterior termini of the endostyle and in the retropharyngeal band at the juvenile stage (Fig. 3B; Ogasawara et al., 1999).

Figure 4.

Schematic diagram of enhancer detection screening utilizing transposase lines. To obtain a double-transgenic F1 generation, a transposase line was crossed with a transgenic line that harbors an enhancer detection Minos vector. The sperm of double-transgenic animals was used to fertilize wild-type eggs to obtain the F2 generation. From the F2 generation, animals with altered GFP expression patterns were isolated and further cultured to establish enhancer-detected lines. Double-transgenic F2 animals can be used to isolate enhancer-detected F3 generations by crossing with wild types. The frequency of each genotype is not considered.

Ju[SBFr3dTPORCiprmMiTP] lines were crossed with Mu[ISMiTSAdTPOG]1 to obtain double-transgenic F1 animals. Double-transgenic animals were isolated at the juvenile stage by GFP/RFP expression, and cultured until sperm maturation. Their sperm was isolated and used to fertilize wild-type eggs to obtain the F2 generation. GFP expression in the F2 family was observed at the larval and juvenile stages. Five F2 families were subjected to enhancer detection screening. GFP expression patterns observed in regions other than the termini of the endostyle and the retropharyngeal band were classified as enhancer detection patterns. Progeny between Mu[ISMiTSAdTPOG]1 and wild types showed no enhancer detection patterns (data not shown). Therefore, if F2 families showed enhancer detection patterns, the enhancer detection events must have been caused by transposase.

We could not detect GFP expression at the larval stage in F2 families (data not shown); however, animals with altered GFP expression patterns were observed at the juvenile and adult stages. All of the screened F2 families contained juveniles with enhancer detection patterns. Interestingly, various expression patterns were observed in one family (Fig. 5), suggesting the occurrence of multiple enhancer detection events in the sperm of double-transgenic animals. In contrast to the multiplicity of enhancer detection events in a family, the total number of animals with enhancer detection patterns at the juvenile stage in each family was very low. A rough estimation of the frequency suggests that approximately 1.0–4.2% (n = 2 families) of juveniles showed enhancer detection patterns. Both RFP-positive and -negative F2 animals showed enhancer detection patterns. An RFP-negative F2 animal derived from a sperm that did not contain transposase cDNA. Therefore, such sperm may have gained transposase at a certain stage of spermatogenesis (see Discussion section).

Figure 5.

An example of juveniles with enhancer detection patterns derived from one double-transgenic animal. All of them show different expression patterns, indicating that they inherit different enhancer-detection events. Note that both RFP-positive and -negative animals inherit enhancer detection events. At, atrial siphon; Bl, blood cells; En, endostyle; Ep, epidermis; Ga, ganglion; Gi, gill; Int, intestine; Or, oral siphon; PB, peripharyngeal band; PE, peribranchial epithelium; St, stigmata; Sto, stomach. A: A juvenile showing GFP expression in the oral siphon, endostyle, and stigmata. Scale bar = 100 μm. B: A juvenile showing GFP expression in the oral siphon, peripharyngeal band, endostyle, middle part of pharyngeal gill, and stomach. C: A juvenile showing GFP expression in the oral siphon muscle, epidermis, peripharyngeal band, endostyle, and peribranchial epithelium. D: A juvenile showing GFP expression in the blood cells and intestine. E: A juvenile showing GFP expression in the endostyle, ganglion, intestine, peribranchial epithelium, and stigmata. F: A juvenile showing GFP expression in the epidermis. Epidermis at ampullae showed strong signal.

F2 families contained animals that harbored both transposase and transposon insertion (Fig. 4). These double-transgenic animals were crossed with wild types to obtain an F3 generation, from which those with altered GFP expression patterns were screened. As in the F2 screening, an F3 family contained animals with various enhancer detection patterns (Fig. 5), and only a very few animals had an enhancer detection pattern.

Animals with enhancer-detection events at the juvenile stage were cultured until sperm maturation, and GFP expression was examined in the adult tissues. We detected GFP expression in the neural complex, endostyle, pharyngeal gill, stomach, intestine, peribranchial epithelium, cardiac muscle, pericardium, oviduct, and oocytes (Table 1, Fig. 6). A considerable number of animals showed GFP expression in oocytes, suggesting the presence of highly frequent enhancers responsible for expression in oocytes in the Ciona genome. These animals were crossed with wild types to obtain the next generation. GFP expression patterns at the larval and juvenile stages were observed. We could not detect GFP expression at the larval stage. Fifteen families showed GFP expression after metamorphosis (Table 1, Fig. 7). GFP expression was observed in the epidermis, neural tissues, dorsal tubercle, endostyle, pharyngeal gill, peribranchial epithelium, intestine, esophagus, and muscle. Expression patterns at the juvenile stage correlated with those at the adult stage in 10 out of 15 lines. In four lines (E6, 9, 23, and 24), GFP expression observed at the juvenile stage was not detected at the adult stage (Table 1). The frequency of F3 animals with enhancer detection GFP expression was measured in five families. As shown in Table 2, the frequency was around 50%, and this ratio matches Mendel's law for a single heterozygous locus. Therefore, we concluded that most of the enhancer detection events were caused by a single locus of transposon insertion. Two F3 families (E10 and E41) had two independent enhancer detection events, suggesting that a single sperm from their F1 ancestor had two independent enhancer detection insertions.

Table 1. GFP Expression Patterns in Enhancer Detection Lines
IDGFP expression patterns
Adult stageLarval stageJuvenile stage
E6OocytesNo expressionEndostyle, peribranchial epithelium
E9Not expressedNo expressionGill
E10Somatic cells of ovary, ovidact, pericardium, ganglionNo expressionOral/atrial siphon muscle, ganglion, esophagus
E17EndostyleNo expressionEndostyle, dorsal tubercle
E21Oocytes, stigmataNo expressionStigmata
E23OocytesNo expressionEndostyle
E24OocytesNo expressionEndostyle
E27Oocytes, endostyleNo expressionEndostyle
E34Oocytes, endostyleNo expressionEndostyle
E35Oocytes, stomach, stigmataNo expressionStigmata
E38Stigmata, languet, endostyle, ganglion, intestine, oocytesNo expressionGill
E39Intestine, cardiac muscle, stigmata, peribranchial epithelium, ganglionNo expressionGill, endostyle
E40Peribranchial epithelium, tentacle, endostyle, languetNo expressionEndostyle, intestine
E41Ganglion, endostyle, intestineNo expressionEndostyle, stigmata
E42Ganglion, endostyle, intestineNo expressionEndostyle, peribranchial epithelium
Figure 6.

GFP expression patterns of enhancer-detection lines in adult tissues. A: Line E10 showing GFP expression in the pericardium. B: Line E10 showing GFP expression in the somatic cells of the ovary. C: Line E24 showing GFP expression in oocytes. Note that B and C show different expression patterns. D: Line E15 showing GFP expression in the pharyngeal gill. E: Line E21 showing GFP expression in zone D of the stigmata (Shimazaki et al., 2006). F: Line E35 showing GFP expression in the stigmata. G: Line E17 showing GFP expression in zone 1 (a bracket) of the endostyle. The RFP signal is derived from pSBFr3dTPORCiprmMiTP. H: Line E27 showing GFP expression in zone 7 (a bracket) of the endostyle. I: Line E41 showing GFP expression in zones 2–8 (brackets) of the endostyle. Scale bars = 100 μm for A,C–I, and 200 μm for B.

Figure 7.

GFP expression patterns of enhancer-detection lines at the juvenile stage. A: A juvenile of Line E6 expressing GFP in the endostyle and gill. Scale bar = 50 μm. B: A juvenile of Line E10 expressing GFP in the muscle of oral and atrial siphons. C: A juvenile of Line E17 expressing GFP in the endostyle, peripharyngeal band, and retropharyngeal band. D: A juvenile of Line E21 expressing GFP in the stigmata. Note that a part of the stigmata does not express GFP (arrows). The RFP signal is derived from pSBFr3dTPORCiprmMiTP. E: A juvenile of Line E27 expressing GFP in the endostyle. F: A juvenile of Line E39 expressing GFP in the peribranchial epithelium, stigmata, and endostyle. G: A juvenile of Line E40 expressing GFP in the peribranchial epithelium, endostyle, and intestine. H: A juvenile of Line E41 expressing GFP in the endostyle. I: A juvenile of Line E42 expressing GFP in the ganglion, peribranchial epithelium, stigmata, endostyle, peripharyngeal band, and esophagus. At, atrial siphon; En, endostyle; Ga, ganglion; Gi, gill; Int, intestine; Oe, esophagus; Or, oral siphon; Pe, peribranchial epithelium; PB, peripharyngeal band; RB, retropharyngeal band; and St, stigmata.

Table 2. Frequency of GFP-Positive Animals in the F3 Generation
Line IDGFP positive animals (%)n

The low frequency of juveniles showing enhancer detection patterns means that most of the progeny from a double-transgenic animal showed the same GFP expression pattern as that of Mu[ISMiTSAdTPOG]1 at the juvenile stage (Fig. 3B). Although the expression pattern was the same, some showed stronger GFP expression. This suggests that these animals possess remobilized transposon insertions. Some of these animals were cultured, and their GFP expression pattern was observed at the adult stage. Of these, 66.6% (n = 15) showed GFP expression, which was not observed in adults of the Mu[ISMiTSAdTPOG]1 line (Table 3). Therefore, in these animals, newly created transposon insertions entrapped enhancers responsible for expression at the adult stage. Eight of the 15 animals showed GFP expression in oocytes, supporting the presence of high-frequency enhancers responsible for expression in oocytes.

Table 3. GFP Expression Patterns in Animals Showing the Same Expression Pattern as Mu[ISMiTSAdTPOG] 1 at the Juvenile Stage*
IDGFP expression patterns
Adult stageLarval stageJuvenile stage
  • *

    The GFP expression pattern of Mu[ISMiTSAdTPOG] 1 at the juvenile stage is described as “TPO pattern.”

E1Neural complexNo expressionTPO pattern
E3Oocytes, endostyleNo expressionTPO pattern
E4OocytesNo expressionTPO pattern
E5Not expressedNo expressionTPO pattern
E7OocytesNo expressionTPO pattern
E11OocytesNo expressionTPO pattern
E13OocytesNo expressionTPO pattern
E14Not expressedNo expressionTPO pattern
E15OillNo expressionTPO pattern
E19Not expressedNo expressionTPO pattern
E20OocytesNo expressionTPO pattern
E22OocytesNo expressionTPO pattern
E25Oocytes, endostyleNo expressionTPO pattern
E28Not expressedNo expressionTPO pattern
E44Not expressedNo expressionTPO pattern


The present study reports transgenic lines expressing Minos transposase in Ciona intestinalis. The cis-element of the protamine gene of Ciona was used to express transposase in male germ cells. Enhancer detection was carried out by creating double-transgenic animals between transposase lines and a transgenic line that has an enhancer detection vector. Enhancer detection lines with novel GFP expression patterns were isolated from their progeny. By this achievement, an easy experimental system for enhancer detection screening has been established in Ciona, as an emerging experimental system for studies of developmental genetics.

Ciona Protamine Gene Is Expressed in Testes and Sperm

We have obtained several pieces of evidence to show that Ci-prm is expressed in sperm. The mRNA of Ci-prmwas detected at the central part of the testes, the position at which cells at the later stages of spermatogenesis such as spermatids gather. Ci-prm expression was undetectable in the peripheral regions of the testes. Similarly, a cis-element of Ci-prm drove GFP expression at the center of the testes and in fully matured sperm. These data suggest that Ci-prm, like the protamine genes of other animals, is expressed at the later stages of spermatogenesis. In mice, it has been reported that protamine shows strong expression after meiosis (reviewed in Tanaka and Baba, 2005). Our finding that Ci-prm was expressed at the region where spermatids gather (Fig. 1) strongly suggests that Ci-prm is expressed after meiosis, although future analyses are required to determine this issue.

GFP expression pattern in sperm of Tg[MiTnIGCiprmG] transgenic lines is suggestive to know the expression manner of Ci-prm in the testes. All sperm in these lines showed GFP expression, while half of the sperm had gfp cDNA in their genome because these transgenic lines are heterozygous with respect to the transposon insertions. This means that sperm that did not have the protamine-gfp cassette somehow obtained gfp mRNA or proteins prior to or after meiosis. One suggestion is that spermatids share cytoplasm via cytoplasmic bridges after meiosis, and gfp mRNA or protein from gfp cDNA-positive spermatids is able to move between all cells. In mammals, the presence of cytoplasmic bridges between spermatids is well known (Braun et al., 1989). Cytoplasmic bridges have also been observed in a tunicate, Dolioletta (Tunicata: Thaliacea) (Holland, 1989), and this phenomenon might be conserved in Ciona. Similarly, the presence of RFP-negative F2 transgenic animals with enhancer detection patterns suggests transposase expression in the same manner as GFP (Fig. 5).

Comparison of the Methods of Enhancer Detection in Ciona

Three methods for enhancer detection have been reported in Ciona intestinalis. One is microinjection of a transposon construct for enhancer detection and transposase mRNA (microinjection-mediated enhancer detection); another is remobilization of Minos for enhancer detection in the genome by microinjection of transposase mRNA (remobilization by microinjection); and the last is utilization of a transposase line as reported in the present study (we call it the “jump-starter” system, because transgenic lines expressing transposase are called “jump-starter” lines).

Microinjection-mediated enhancer detection is the simplest method (Sasakura et al., 2003; Awazu et al., 2004, 2007). Its technical advantage is that we have only to prepare a transposon vector for enhancer detection to perform this method. However, this method involves two technical disadvantages. One is that the frequency of enhancer detection is the lowest among the three methods. A previous study using pMiTSAdTPOG reported that approximately 5.4% of microinjected animals transmit enhancer detection insertions (Awazu et al., 2007). The other disadvantage is that microinjection of both DNA and RNA is required. DNA is toxic compared with RNA, and a considerable portion of injected embryos will die due to unhealthy development. This means that microinjection of both DNA and RNA into a large number of eggs is required to establish the requisite number of enhancer detection lines.

Remobilization by microinjection has been developed to overcome these disadvantages (Awazu et al., 2007). In this method, transposase mRNA is injected into unfertilized eggs, and these eggs are fertilized with sperm of the Mu[ISMiTSAdTPOG]1 line, which has multiple copies of a transposon vector for enhancer detection in its genome. These injected animals are crossed with wild types, and progeny with altered GFP expression patterns are screened. The advantage of this method is the high frequency of enhancer detection. Approximately 79% of the injected animals transmitted enhancer detection insertions. On the other hand, the two technical disadvantages are that a “mutator” line must be created prior to starting this method, and that microinjection is required.

Compared to the other two methods, utilization of a transposase-expressing line has several advantages. The best advantage is that dechorionation or microinjection, which often interferes with normal development, is not required. In this method, isolation and mixing of germ cells is the only step required to create founder animals. Another advantage is that several enhancer detection insertions are inherited from one founder. All of the double-transgenic animals transmitted several enhancer detection insertions to the subsequent generation. Therefore, screening of a few founders may yield plentiful enhancer detection lines. This sufficiently reduces the labor required for the creation and culturing of founders.

The technical limitation of the present study is that only a few percent of progeny from the double-transgenic animals inherit enhancer detection insertions. This indicates that there are few progeny for one enhancer detection event. Such a small population limits the following culturing to establish the enhancer detection line. In contrast, the other two methods provide many more progeny with an enhancer detection insertion sufficient for further culturing.

We have to estimate the timing when remobilization of transposons will overcome the technical disadvantage. Low frequency of inheritance of a single enhancer detection insertion indicates that only a few sperm carried it. This means that cells containing an enhancer detection insertion are rarely amplified through mitosis. We have shown that Ci-prm is expressed at later stages of spermatogenesis such as spermatids. Thus, transposase expression and transposition of transposons may start at roughly the same time. We observed some juveniles from the same family that showed the same GFP expression patterns (data not shown). The occurrence of independent enhancer detection events having identical GFP expression patterns in the same family is unlikely. Rather, it is likely that one enhancer detection insertion might have been amplified by cell division, such as meiosis. This suggests that a few transpositions might have occurred in the late spermatogonia and spermatocytes. As the spermatogenesis proceeds, transposition may cease because condensation of chromatin inhibits the access of transposase to the genome DNA. Therefore, transposition may occur during the late-spermatogonia to early-spermatid stages. To increase the number of animals inheriting a single enhancer-detection insertion, frequency of transposition before meiosis has to be increased. For this purpose, another cis-element that drives the expression of downstream genes in spermatogonia is necessary.

The “jump-starter” system is suitable for large-scale enhancer detection rather than isolation of a specific line with a purposeful expression pattern, because we can obtain several animals with independent enhancer detection events from a single double-transgenic animal. For example, we here assume an isolation of 16,000 independent enhancer detection lines, the same frequency as the presence of genes in the Ciona genome (Dehal et al., 2002). If we carry out microinjection-mediated methods, the numbers of eggs for which microinjection is required would range from 21,894 (remobilization by microinjection) to 296,000 (microinjection-mediated enhancer detection). Because injected embryos often die due to damage caused by microinjection, the actual number of eggs would have to be increased several-fold. Microinjecting, culturing, and screening such large numbers of potential founders is time-consuming. If we carry out this experiment by the “jump-starter” system, the number of founders can be reduced to 8,000 or less. These founders can be generated without any artificial treatment, guaranteeing healthy development of manipulated animals. Therefore, the “jump-starter” method is suitable for large-scale enhancer detection.

Tendencies of Enhancer Detection in Ciona

In this study, we noticed that enhancer detection with the Ci-TPO promoter has two tendencies. One is that many enhancer detection lines show GFP expression in oocytes (Table 1). This suggests that the Ciona genome contains high-frequency enhancers responsible for maternal expression. This is consistent with the fact that many genes are maternally expressed. For example, EST analyses suggest that 3′-EST of Ciona fertilized eggs consist of over 8,000 clusters, which correspond to half the number of protein-coding genes in the Ciona genome (Kawashima et al., 2005). Therefore, enhancers of maternally expressed genes may be easily detected by enhancer detection. In contrast to oocytes, an enhancer detection event for sperm has not been observed. It is estimated that a much smaller number of genes is expressed in sperm than in eggs (Hozumi et al., 2004; Inaba et al., 2007). This suggests that the Ciona genome contains low-frequency genes and enhancers responsible for expression in sperm, which are rarely entrapped by transposon insertions.

The second tendency is that very few enhancer detection lines showed GFP expression during early embryogenesis or at the larval stage (Awazu et al., 2007). One possibility is that enhancers for expression during embryogenesis and adult stages have different preferences for promoters, and the Ci-TPO promoter is less responsible for enhancers during early embryogenesis. This possibility has to be clarified by enhancer detection with a minimal promoter of a gene that shows expression during early embryogenesis. An alternative possibility is that enhancers responsible for early embryogenesis are located at a much lower frequency than those for the adult stage. Kawashima et al. (2005) has suggested that as embryonic and larval development proceeds, embryos and larvae express a smaller and smaller number of gene species. The ratio increases slightly during metamorphosis, and a novel and/or different set of genes is expressed to form the adult body. This implies that more genes are expressed after metamorphosis than at the embryonic/larval stages, supporting the second possibility. These two possibilities must be examined in future studies.

Introduction of a new method is necessary to improve germline transgenesis technologies with Minos in Ciona intestinalis. Among several methods, creation of transposase-expressing lines has been desired because of its convenience for creating new insertions. In the present study, we have successfully generated transposase lines. These lines were used for enhancer detection, and an efficient method for enhancer detection was developed. The transposase lines described here will be valuable tools for carrying out sophisticated genetic techniques like local hopping, gene trapping, and insertional mutagenesis as well as enhancer detection in Ciona intestinalis.


Biological Materials

Wild-type Ciona intestinalis was collected from or cultivated in Oura Bay (Shizuoka), Onagawa Bay (Miyagi), Otsuchi Bay (Iwate), Maiduru Bay (Kyoto), and Usa Bay (Kochi). Sperm and eggs were collected by cutting the sperm duct and egg duct, respectively. Transgenic animals were cultured by an inland system described elsewhere (Joly et al., 2007). Testes from wild-type animals were fixed with 4% formaldehyde in 0.5 M NaCl, 0.1 M MOPS (pH 7.5), dehydrated by successive treatment with ethanol and butanol, and embedded in paraffin. Sections that were 7-μm-thick were prepared from the specimens, and they were stained with Delafield's hematoxylin and eosin.


pSPCiprmG: An approximately 2-kb segment of the 5′ upstream region from Ci-prm was amplified by PCR with primers 5′-cgggatccgatatttctgtgaattcacttg-3′ and 5′-cgggatccatttcggctcaaaaacaatttc-3′. The PCR product was digested with BamHI and inserted into the BamHI site of pSP-eGFP to create pSPCiprmG.

pMiCiTnIG: An 889-bp segment of the 5′ upstream region from Ci-TnI was ligated into the pCES vector using Sph1 and Not1 sites. This was cut with SphI and BglII, blunted with T4 DNA polymerase and ligated into the blunted PstI site of pMiLRneo to create pMiCiTnIG.

pSPCiprmMiTP: The 5′ upstream region from Ci-prm was inserted into the BamHI site of pSP-MiTP to create pSPCiprmMiTP.

pMiCiTnIGCiprmG: A CiprmG cassette was amplified from pSPCiprmG with primers 5′-gaactcgagcagctgaagcttg-3′ and 5′-gcagatctgatggccgctttgac-3′, and was subcloned into the SmaI site of pMiCiTnIG to create pMiCiTnIGCiprmG.

pSPFr3dTPOR: The Fr3dTPO fragment of pSPFr3dTPOG (Matsuoka et al., 2005) was amplified with 5′-gaactcgagcagctgaagcttg-3′ and 5′-cattctaatttgtctcgttttattttgtg-3′ and subcloned into the blunted BamHI site of pSPnDsRed to create pSPFr3dTPOR.

pSBFr3dTPOR: The Fr3dTPOR cassette was amplified from pSPFr3dTPOR with primers 5′-gaactcgagcagctgaagcttg-3′ and 5′-gcagatctgatggccgctttgac-3′ and the PCR product was ligated into the EcoRV site of pT/HB to create pSBFr3dTPOR.

pSBFr3dTPORDestF: An RfC1 fragment of the Gateway vector conversion kit (Invitrogen) was inserted into the blunted BglII site to create pSBFr3dTPORDestF.

pSBFr3dTPORCiprmMiTP: The Ci-prm-MiTP fusion cassette of pSPCiprmMiTP was subcloned into pSBFr3dTPORDestF with the gateway system (Invitrogen) to create pSBFr3dTPORCiprmMiTP.

Transgenic Lines

The transgenic line Mu[ISMiTSAdTPOG]1 was previously described (JM[ISMiTSAdTPOG]1 in Awazu et al., 2007). Tg[MiCiTnIG]1 and Ju[SBFr3dTPORCiprmMiTP]1 and 2 were created by electroporation-mediated transgenesis of 60 μg of pMiCiTnIG and pSBFr3dTPORCiprmMiTP, respectively (Matsuoka et al., 2005; Sasakura, 2007). Tg[MiCiTnIG]2 and Tg[MiCiTnIGCiprmG] lines were created by co-electroporation of 80μg of in vitro synthesized transposase mRNA and 60 μg of pMiCiTnIG and pMiCiTnIGCiprmG, respectively (Matsuoka et al., 2005; Sasakura, 2007). Electroporation was done with GenePulser Xcell™ (Bio-Rad) at the parameters of 20 ms, 50 V.

Screening of Enhancer Detection Lines

Mu[ISMiTSAdTPOG]1 and Ju[SBFr3dTPORCiprmMiTP]1-2 were crossed to obtain the F1 generation. Among the F1 generation, GFP- and RFP-positive animals were screened at the juvenile stage and cultured further. Sometimes GFP signals were invisible at the juvenile stage because of strong RFP fluorescence. In this case, RFP-positive animals were selected for further culturing. When these animals had mature sperm, the sperm was isolated surgically to inseminate wild-type eggs to obtain F2 families. GFP/RFP signals of F2 families were observed at the larval, juvenile, and adult stages. GFP-positive animals were first selected with a fluorescent binocular stereo microscope, and their GFP expression pattern was observed and photographed with a fluorescent uplight microscope. F2 animals with distinct GFP expression patterns were isolated to establish enhancer detection lines.

In Situ Hybridization

Digoxigenin (DIG)-labeled sense and antisense probes were synthesized using the full-length protamine coding sequence (cDNA clone GC35p16) with a DIG RNA labeling mixture (Roche), and T3 RNA polymerases (Roche) or T7 RNA polymerases (Invitrogen), respectively. Testes were dissected from adult Ciona and fixed overnight with 4% formaldehyde in 0.5 M NaCl, 0.1 M MOPS (pH 7.5) at 4°C. Specimens were preserved in 80% ethanol at −30°C until use. After washing with PBST (PBS containing 0.1% Tween 20), specimens were partially digested with 10 μg/ml proteinase K (Merck) in PBST for 50 min at 37°C. Hybridization and washing of the excess probes were carried out basically according to Ogasawara et al. (2002). After visualization of hybridization signals with NBT-BCIP, specimens were dehydrated through successive treatment with ethanol and butanol, and were embedded in paraffin. Sections that were 7-μm-thick were prepared from the specimens.


Total RNA was isolated from the testes and intestine complex, the gills, and the endostyles of Ju[SBFr3dTPORCiprmMiTP] lines and wild-type animals by AGPC methods (Sambrook et al., 1989). Residual DNA was digested with DNaseI (Takara). Reverse transcription was performed with SuperscriptII reverse transcriptase (Invitrogen). PCR was performed with AccuTaq DNA polymerase (Sigma). PCR primers for GAPDH, Minos transposase, and DsRed were 5′-tcggaatcaacggtttcggacg-3′ and 5′-cgatgacacggttgctgtatcc-3′, 5′-tagattagaattgtgtaacgtc-3′ and 5′- ccggatccatggttcgtggtaaacctattt-3′, 5′-aaggtacccatggcctcctccgagaacgtc-3′ and 5′-ctacaggaacaggtggtggc-3′, respectively. PCR parameters were 30 s at 98°C, then 40 cycles of 30 s at 94°C, 20 s at 55°C, 2 m at 68°C, followed by 10 m at 68°C for final extension.


We thank Kazuko Hirayama, Yasuo Kasuga, Yasutaka Tsuchiya, Toshihiko Sato, Hideo Shinagawa, Yoshiko Harada, Norio Miyamoto, and members of the Shimoda Marine Research Center at the University of Tsukuba for their kind cooperation with our study. We thank all members of the Maizuru Fishery Research Station of Kyoto University, the International Coast Research Center of the Ocean Research Institute of the University of Tokyo, the Education and Research Center of Marine Bioresources at Tohoku University, and Prof. Shigeki Fujiwara for the collection of Ciona adults. Dr. Stephene Ekker and Dr. Charalambos Savakis are acknowledged for their kind provision of SB and Minos. Dr. Brad Davidson is acknowledged for generously providing the pMiCiTnIG vector and Tg[MiCiTnIG] transgenic lines. We thank Dr. Koichi Kawakami for meaningful discussions. This study was supported by Grants-in-Aid for Scientific Research from JSPS and MEXT to Y.S., N.S., and K.I. Y.S. was supported by NIG Cooperative Research Program (2006-A72 and 2007-B01).