Activity of the medaka translation elongation factor 1α-A promoter examined using the GFP gene as a reporter

Authors

  • Masato Kinoshita,

    1. Division of Applied Biosciences, Graduate School of Agriculture, Kyoto University, Sakyo, Kyoto 606-8502,
    Search for more papers by this author
    • *Author to whom all correspondence should be addressed.

  • Shuichi Kani,

    1. Division of Biological Science, Graduate School of Science, Nagoya University and
    Search for more papers by this author
  • Kenjiro Ozato,

    1. Division of Biological Science, Graduate School of Science, Nagoya University and
    2. Laboratory of Freshwater Fish Stocks, Bioscience Centre, Nagoya University, Nagoya 464-8601, Japan.
    Search for more papers by this author
  • Yuko Wakamatsu

    1. Division of Biological Science, Graduate School of Science, Nagoya University and
    2. Laboratory of Freshwater Fish Stocks, Bioscience Centre, Nagoya University, Nagoya 464-8601, Japan.
    Search for more papers by this author

Abstract

The translation elongation factor 1α (EF-1α) is known to have several isoforms, which are expressed in a tissue- and stage-specific manner. Two genes encoding EF-1α exist per haploid genome in the medaka. In the present study, the promoter activity of the 5′-flanking region of the medaka EF-1α-A gene, an isoform of EF-1α, was characterized using transgenic techniques. First, using CAT gene as a reporter, it was revealed that about 1.8 kbp 5′-flanking sequence from the transcription initiation site of EF-1α-A was sufficient for high-level promoter activity. Second, the green fluorescent protein (GFP) gene fused to this region was introduced into medaka eggs using the microinjection method. Three germline transgenic individuals (one male and two female) were mated with non-transgenic medaka to obtain F1 offspring. In the case of embryonic and adult F1 transgenic individuals, GFP fluorescence was observed in almost all the tissues examined (e.g. kidney, liver, heart, gill, ovary, and testis), except for the skeletal muscle. In the case of F2 transgenic embryos derived from F1 transgenic males and non-transgenic females, the fluorescence was observed from the early gastrula stage. On the other hand, in the case of F2 transgenic embryos derived from F1 transgenic females and non-transgenic males, the fluorescence was observed even at the 1-cell stage, suggesting that this region is transcriptionally active during oogenesis. The usefulness of the EF-1α-A promoter as a tool for introducing foreign proteins into oocytes is discussed.

Introduction

Translation elongation factor 1α (EF-1α) is a ubiquitous protein present in abundance in all types of cells ( Negrutskii & Elskaya 1998). EF-1α promotes guanosine triphosphate (GTP)-dependent binding of aminoacyl-transfer ribonucleic acid (tRNA) to ribosomes during peptide chain elongation as part of the translational elongation complex. Recent studies of EF-1α reveal that it has several isoforms that are expressed in a tissue- and stage-specific manner during development. For example, two EF-1α isoforms exist in humans, designated as EF-1α and EF-1α-2 ( Knudsen et al. 1993 ). The latter is expressed in the brain, heart and skeletal muscle and the former in other tissues. In Xenopus laevis, three active EF-1α genes (42Sp50, EF-1αO, and EF-1αS) have been characterized ( Dje et al. 1990 ; Abdallah et al. 1991 ). The expression of 42Sp50 is predominant in immature oocytes and then switches to that of EF 1αO and EF 1αS during early embryonic development. These features of EF-1α suggest that the promoter regions of EF-1α isoforms are responsible for promoting gene expression in a tissue- and stage-specific manner.

To date, the structure of the fish EF-1α gene has been studied in zebrafish and medaka. Gao et al. (1997) reported that the zebrafish has only a single gene encoding EF-1α per haploid genome. In contrast, it was reported that the medaka has two genes encoding EF-1α per haploid genome, designated as EF-1α-A and EF-1α-B ( Kinoshita et al. 1999 ). In the preliminary experiments to examine the promoter activity of these two isoforms, it was revealed that while the EF-1α-A promoter was active in 1-day-old embryos (4-somite stage), the EF-1α-B promoter was not ( Kinoshita et al. 1999 ). These findings suggest that the two isoforms of the medaka EF-1α gene are expressed in a tissue- and stage-specific manner as reported for the case of humans and X. laevis.

In the present study, our aim was to characterize the promoter activity of the 5′-flanking sequence of one isoform of the medaka EF-1α gene, EF-1α-A, using transgenic techniques. First, we determined the promoter region of the EF-1α-A gene using the chloramphenicol acetyltransferase (CAT) gene as the reporter. Second, the promoter activity of this region was investigated in transgenic progenies using the green fluorescent protein (GFP) gene as the reporter. In these transgenic progenies, we found that EF-1α-A promoter induced gene expression in embryos and in almost all the tissues examined, except for the skeletal muscle in adult individuals.

In addition, GFP fluorescence was first observed at the early gastrula stage in transgenic progenies derived from the crossing of transgenic males with non- transgenic females, indicating that the zygotic expression of EF-1α-A began before this stage. On the other hand, GFP fluorescence was observed even at the 1-cell stage in the transgenic progenies derived from the crossing of transgenic females with non-transgenic males, indicating that the transgene product was provided in the oocytes as a maternal factor. The tissue specificity of medaka EF-1α-A promoter activity, the time of commencement of zygotic expression of the transgene, and providing the transgene product in oocytes are discussed. The usefulness of the EF-1α-A promoter as a tool for introducing foreign proteins into oocytes as maternal factors is also discussed.

Materials and Methods

Plasmid construction

A deletion series of about 4 kbp of the 5′-flanking sequence of the transcription initiation site of EF-1α-A: Medaka EF-1α-A gene containing the 5′-flanking sequence was cloned into EMBL 3 (Stratagene, La Jolla, CA, USA) and sequenced ( Kinoshita et al. 1999 ). About 4 kbp of the 5′-flanking sequence from the transcription starting site was amplified by polymerase chain reaction (PCR). The primers used for the PCR were as follows. Forward primer: GCGTCGACGTAAAACGACGGCCAGT (a part of the M13–20 primer sequence of pBluescript KS II +). Reverse primer: GCGTCTAGAGGATCCTTTGTATGTTTCTGTTCCAA. SalI or XbaI and BamHI cleavage sites were designated for the cloning of the PCR product into a plasmid vector. The PCR amplification cycles consisted of 94°C for 90 s, 50°C for 120 s and 72°C for 150 s. The PCR product was inserted into the SalI and XbaI sites of pCAT-basic vector (Promega, Madison, WI, USA). The vector was digested with SphI and XbaI, and progressively deleted by exonuclease III digestion.

In the case of pEF-1α-A-GFP, because Δ1793 shown in Fig. 1 exhibited almost full promoter activity, the fragment from – 1936 to + 853 was amplified by PCR with KOD DNA polymerase (Toyobo, Osaka, Japan). The primers used for the PCR were as follows. Forward primer: TCCAAAGGACTACTTCCTGT. Reverse primer: GGTCTAGAGGTACCTTTGTATGTTTCTGTTCCAA. pCMX-SAH/Y145F ( Ogawa & Umezono 1998), which contained the modified GFP gene, was digested with NarI and blunted with KOD DNA polymerase before digestion with KpnI. The PCR product was then digested with KpnI and inserted into the blunted and KpnI site of pCMX-SAH/Y145F.

Figure 1.

Activity of various deletion mutants of medaka EF-1α-A promoters. Deleted regions of the promoter are depicted on the left side of the diagram. The data represent the average of two experiments. For each experiment, each of the constructs was injected into six to 11 fertilized eggs. The activities of the constructs were normalized to the amount of the CAT gene remaining in 3-day-old embryos. The CAT activities derived from the deletion constructs are expressed relative to the activity derived from the non-deletion construct. Details are described in Materials and Methods. Non-deletion construct, a construct containing about 4 kbp of the 5′-flanking sequence from the transcription initiation site, Δ1793, Δ1570, Δ1067, Δ777, Δ491, Δ211, and Δ + 458 constructs contain sequences from + 853 to – 1793 bp, – 1570 bp, – 1067 bp, – 777 bp, – 491 bp, – 211 bp, and + 458 bp, respectively. Arrowheads show the positions of the putative Sp1-binding element.

Microinjection

An inbred strain of the medaka with the wild-type body color (HNI-I) was maintained under controlled lighting conditions (14 h light and 10 h dark) at 26°C. Plasmids of the deletion series of 4 kbp of the 5′-flanking sequence of EF-1α-A, followed by CAT and pEF-1α-A-GFP in a circular form in phosphate-buffered saline were injected at a concentration of 25 ng/μL into the cytoplasm of fertilized eggs before the first cleavage, as described previously ( Kinoshita et al. 1996 ).

CAT assay

Each deletion plasmid was injected into six to 11 eggs for each experiment. The eggs were incubated at 26°C for 3 days. Then, the 3-day-old embryos were collected into a microcentrifuge tube and homogenized with 400 μL of 250 m M Tris-HCl (pH 8.0). Half of the homogenate was centrifuged and 50 μL of the resulting supernatant was used for the CAT assay as previously described ( Kinoshita et al. 1996 ). The other half of the homogenate was used for extraction of DNA as previously described ( Kinoshita et al. 1996 ). To evaluate the amount of CAT gene remaining in the embryos, dot-blot analysis was carried out with [32P]-labeled CAT gene and the intensity of the radioactivity calibrated with Fuji BAS 2000. CAT activity was normalized to the amount of CAT gene remaining in the embryos. The data represent the average of two independent experiments.

Production of transgenic F1 and F2 carrying EF-1α-A-GFP

pEF-1α-A-GFP was injected into fertilized medakafish eggs, which were then incubated at 26°C and raised to adulthood. The founder fish (F0) were mated with the wild-type strain and about 200 offspring obtained from each pair were examined for GFP fluorescence and for detection of the GFP gene by PCR analysis at the 3-day embryo stage (described in the following). The founder fish, whose offspring showed GFP fluorescence and carried the GFP gene, were determined as being germ-line-transformed individuals. In the F1 generation from each transformed F0 individual, an F1 embryo exhibiting GFP fluorescence and carrying the GFP gene was raised to adulthood and mated with a non-transgenic individual to produce the F2 transgenic generation. The onset of zygotic expression of the transgene was examined in F2 embryos obtained from the crossing of an F1 transgenic male with a non-transgenic female, providing the gene product (GFP) in the oocytes was examined in the F2 embryos obtained from the crossing of a non-transgenic male with an F1 transgenic female.

Preparation of frozen sections

Ovarian and testicular tissues from the adult fish were fixed in phosphate-buffered saline (PBS)-based 4% paraformaldehyde for 8 h at 4°C and infiltrated in PBS-based 20% sucrose for 3 h at 4°C. Then, the samples were covered with OTC compound (Miles, Elkhart, IN, USA) and quickly frozen in liquid nitrogen. Frozen samples were stored at – 80°C until sectioning ( Sasado et al. 1999 ). Sections (10–30 μm thick) were prepared on a cryostat and placed on poly L-lysine-coated slides. After rinsing with Tris-buffered saline (TBS; 10 m M Tris-HCl, pH 7.4, 150 m M NaCl), the sections were incubated in 0.5 μg/mL 4′,6-diamidino-2-phenylindole (DAPI) in TBST (TBS including 0.1% Tween-20) for 5 min to visualize the nuclei and rinsed. The sections were then mounted in glycerol for microscopy.

Observation of fluorescence

The green fluorescence of GFP in the embryos and adult tissues was observed under a fluorescence microscope with a fluorescein isothiocyanate (FITC) filter or a stereoscopic microscope with a GFP filter (long-pass type).

Histochemical observation of frozen sections was carried out under a fluorescence microscope. Green and blue fluorescence images were obtained separately using a Chroma filter set (83000; Chroma, Brattleboro, VT, USA).

Detection of GFP gene by PCR

Deoxyribonucleic acid isolation from 3-day-old embryos or hatchlings was carried out according to Amsterdam et al. (1995) . In the case of 3-day-old embryos, the chorions were physically destroyed with needles before proteinase K treatment. For the detection of the GFP gene, PCR was performed using a GFP-gene-specific primer pair (SAH-A and HY145F-B), which yielded a 294 bp fragment. The DNA sequences of the primers were as follows: SAH-A, CCCTGGCCCACACTAGTGACCACCTTCGCTTACGGC; and HY145F-B, CATGATGTATACGTTGTGGCTGTTGAAGTTGTA. In the case of detection of the EF-1α-A gene, an EF-1α-A-gene-specific primer pair (EF-810F and EF-1160R), which yielded a 519 bp fragment, was used. The DNA sequences of the primers were as follows: EF-810F, CAGGACGTCTACAAAATCGG; and EF-1160R, AGCTCGTTGAACTTGCAGGCG.

Results

Detection of the promoter region of the medaka EF-1α-A

We had already cloned about 4 kbp of the 5′-flanking sequence of the medaka EF-1α-A and found that this region had a strong promoter activity when it was examined in 1-day-old embryos ( Kinoshita et al. 1999 ). In the present study, the length of the 5′-flanking sequence, which was sufficient to induce promoter activity, was screened using the CAT gene as the reporter ligated to various lengths of fragments deleted from the 5′-flanking sequence. The fusion constructs were injected into fertilized eggs and the CAT activities were assayed in 3-day-old embryos ( Fig. 1). When the promoter activity of the non-deletion construct containing about 4 kbp of the 5′-flanking sequence of EF-1α-A was defined as 100%, Δ1793 showed a similar level of activity, 93%. Subsequent deletions to Δ777 gradually decreased the promoter activity to 25%. Additional deletions to Δ211 again increased the activity to 46%, suggesting the existence of a suppressing element(s) between – 1067 and – 211. Δ + 458 exhibited no promoter activity. These data indicate that the DNA sequence 1793 bp upstream from the transcription initiation site was sufficient for the high-level expression of the promoter activity of EF-1α-A.

Production of transgenic medaka

To investigate the characteristics of the EF-1α-A promoter, pEF-1α-A-GFP that contained the fragment from – 1936 to + 853 and the GFP gene was injected into 150 fertilized eggs. Ninety-five of them hatched and 48 of the hatchlings grew to maturity, which were then mated with the non-transgenic individuals to obtain F1 embryos ( Table 1). Three of these 48 individuals (one male and two females) produced F1 offspring that showed GFP fluorescence. They were tentatively named F0-TG-male, F0-TG-female-1 and F0-TG-female-2. Thirty-seven percent (106 of the 290 individuals tested) of the F1 embryos derived from the crossing of the F0-TG-male with a non-transgenic female showed GFP fluorescence from the early gastrula stage, all of which inherited the transgene as revealed by PCR analysis (data not shown). Eighty-three (85 of the 102 individuals tested) and 61% (83 of the 136 individuals tested) of F1 embryos, derived from the crossing of the F0-TG-female-1 and F0-TG-female-2 with non-transgenic males, showed fluorescence even from the 1-cell stage, but some of these embryos lacked the transgene (data not shown). Thus, the expression pattern of GFP during the early developmental stage (from the 1-cell to early gastrula stage) and the existence of the transgene depended on the sex of the transgenic parent crossed with a non-transgenic individual.

Table 1.  Production of germ cell transmitter (F0)
 Eggs
injected

Hatch
Sexually
matured
Germ cell
transmitter
No.15095483

Promoter activity in F1 individuals

Figures 2 and 3 show the expression pattern observed in F1 individuals derived from crossing an F0 transgenic male (F0-TG-male) with a non-transgenic female. Actually, after the early gastrula stage, no difference in the expression pattern was observed between F1 embryos derived from the transgenic male and non-transgenic female pair and those derived from the transgenic female and non-transgenic male pairs. Moreover, in general, no difference in the expression pattern was observed among embryos and adults in each F1 transgenic line and among three different F1 transgenic lines. However, the intensity of green fluorescence at the same stage of embryos or the same tissue of adults varied slightly among the three lines (data not shown).

Figure 2.

Expression of GFP gene driven by the EF-1α-A promoter in a transgenic fish during embryonic development. F1 individuals, derived from crossing of an F0 transgenic male (F0-TG-male) with a non-transgenic female were examined by fluorescence microscopy. The sample was heterozygous for the GFP gene. Fluorescence of green fluorescent protein (GFP) was observed ubiquitously at the late gastrula stage (A). Intense fluorescence was observed in the embryonic body at the 9-somite stage (B-1), in which organs such as the forebrain, midbrain, hindbrain, optic vesicle (B-2), notochord, and somites (B-3) could be identified in the fluorescence images. A hatched embryo showed intense GFP fluorescence in the lens, head, and yolksac (C). (D) External appearance of a juvenile individual (3 weeks after hatching). FB, forebrain; HB, hindbrain; MB, midbrain; Nt, notochord; OV, optic vesicle; Sm, somite. Bars, 0.5 mm (A,B-1,C), 0.05 mm (B-2,B-3), 5 mm (D).

Figure 3.

Expression of the GFP gene driven by the EF-1α-A promoter in various organs of adult F1 transgenic fish. An adult fish derived from crossing of an F0 transgenic male (F0-TG-male) with a non-transgenic female was examined for green fluorescent protein (GFP) fluorescence. (A) Gill, (B) kidney, (C) heart, (D) spleen, (E) liver, (F) gut. (G) The brain showed weaker fluorescence than other organs, thus longer exposure was needed to obtain a fluorescence image with a 3 charged coupled device camera when compared with that for other organs. (H) External appearance of the tail part of a juvenile fish. Because the skeletal muscle did not exhibit any fluorescence, the vertebral column fluorescence could be observed. Testis (I-1) showed intense fluorescence. (I-2) Frozen section of a testis. Ovary (J-1) also showed intense fluorescence. (J-2) Frozen section of an oocyte. Po, post-vitellogenesis stage; Pr, previtellogenesis stage; V, vitellogenesis stage. Bars, 1 mm; except I-2, J-2, 0.2 mm.

High-level expression of GFP was observed ubiquitously in the embryonic body from the late gastrula ( Fig. 2A) to the 9-somite stage ( Fig. 2B-1). At the 9-somite stage, some organs, such as the optic vesicle, forebrain, midbrain, hindbrain ( Fig. 2B-2), notochord and somite ( Fig. 2B-3), could be identified on the fluorescence images. As the somites developed, the fluorescence intensity began to decrease in the somites, finally disappearing from the skeletal muscle before hatching ( Fig. 2C). In hatchlings, intense fluorescence was observed in the yolk sac, lens and gill, but only weak fluorescence in the brain ( Fig. 2C). In adult individuals, the fluorescence was apparent externally on the skin and, in particular, the lens showed such intense fluorescence that it was a useful external marker to identify transgenic individuals ( Fig. 2D).

On anatomical study of the adult fish, GFP was observed in most of the organs examined, except for the skeletal muscle. In the gill lamella, the fluorescence was so intense that it could be observed through the branchial mantle ( Figs 2D,3A). Intense fluorescence was also observed in the kidney ( Fig. 3B). In the heart ( Fig. 3C), weak fluorescence was observed in the entire organ. In these two organs, no fluorescence was observed in the blood vessels. In the spleen, a mosaic pattern of fluorescence was observed. Namely, the fluorescence was observed in the white pulp, but not in the red pulp ( Fig. 3D). The fluorescence was also observed ubiquitously in the liver ( Fig. 3E), gut ( Fig. 3F), brain ( Fig. 3G), and skin (data not shown), although it was relatively weak. Fluorescence was observed in the soft rays of the anal fin and the vertebral column through the skeletal muscle, indicating that the skeletal muscle did not show any fluorescence ( Fig. 3H). The absence of fluorescence in the skeletal muscle was also confirmed by examination of the excised tissues (data not shown).

Intense fluorescence was observed in the genital organs ( Fig. 3I,J). In the testis, fluorescence was observed on the rugged surface ( Fig. 3I-1), which corresponded to patchy areas of fluorescence at the outer-most part of the frozen section ( Fig. 3I-2). These areas are known to contain spermatogonia in the medaka testis ( Shibata & Hamaguchi 1988). In the present study, however, accurate identification of cell types in these areas with GFP fluorescence was difficult because of the thickness of the frozen sections. In the ovary, intense fluorescence was observed in globular structures on the surface ( Fig. 3J-1), which were determined to be oocytes in the frozen sections ( Fig. 3J-2). The fluorescence was observed in the entire oocyte in the previtellogenesis stage, in which the fluorescence was more intense in the germinal vesicle than the cytoplasm. In the post-vitellogenesis stage, the fluorescence was pushed away by yolk granules to the part of the cell beneath the cortex. No fluorescence was observed in the oil droplets and yolk granules in the oocytes.

Intense fluorescence in 1-cell stage embryos

Polymerase chain reaction analysis in the F2 generation showed that transgenic and non-transgenic individuals segregated at a 1 : 1 ratio in each transgenic line (data not shown). Thus, it was confirmed that F1 individuals were heterozygous for the transgene and that the transgene was transmitted according to Mendelian law.

In F2 embryos derived from crossing F1 transgenic females with non-transgenic males, all individuals showed intense fluorescence even at the 1-cell stage ( Fig. 4A), although half of these F2 embryos lacked the GFP gene because of heterozygosity of F1 transgenic females for the GFP gene, as mentioned earlier. At one-third epiboly, fluorescence was observed in all embryos ( Fig. 4B) and 1 day after fertilization (6-somite stage, Fig. 4C).

Figure 4.

Detection of green fluorescent protein (GFP) fluorescence in F2 embryos. The left column (TG female) indicates images of F2 embryos derived from crossing F1 transgenic females with non-transgenic males. The right column (TG male) indicates images of F2 embryos derived from crossing F1 transgenic males with non-transgenic females. The inset in (A) is a higher magnification view without excitation. In (D,F,H), the embryo on the left inherited the GFP gene, but the one on the right did not. The arrowhead indicates the embryonic shield that showed rather intense fluorescence. Bar, 1 mm.

However, the fluorescence became progressively fainter in half of the embryos during successive developmental stages and disappeared from the tissues except in the yolk sac in 7-day-old embryos (right figure of Fig. 4D). In the remaining embryos, the fluorescence remained in the lens and head except in the skeletal muscle (left figure of Fig. 4D). The number of embryos that showed fluorescence in the lens among 7-day-old embryos (precardial cavity formation stage) is shown in Table 2. Green fluorescent protein fluorescence was observed in 131 embryos out of the 261 embryos tested (right column of Table 2). To investigate the cause of the disappearance of fluorescence, each of the 7-day-old embryos was subjected to PCR analysis to determine the existence of the GFP gene. As shown in Fig. 5, the GFP gene was not detected in embryos in which the fluorescence disappeared, but it was in embryos in which the fluorescence persisted. Thus, it is likely that the product (GFP) of the GFP gene driven by the EF-1α-A promoter presented in the eggs as a maternal factor and persisted in the embryos at least until 7 days after fertilization. This possibility was confirmed by examining the expression of the GFP gene in F2 embryos derived from crossing F1 transgenic males with non-transgenic females in the following experiment.

Table 2.  Mendelian segregation of green fluorescent protein (GFP) expression in the F2 generation
 F2 from TG male*F2 from TG female
Stage/Expression++
  1. The transgenic F1 was mated with the wild-type strain to obtain the F2 generation.

  2. *F2 individuals derived from crossing transgenic F1 males with non-transgenic females.

  3. †F2 individuals derived from crossing transgenic F2 females with non-transgenic males.

  4. ‡The expression of GFP in the lens was observed by fluorescence microscopy.

Number of138129131130
7-day-old embryos(53%)(47%)(50%)(50%)
Figure 5.

Detection of inheritance of the green fluorescent protein (GFP ) gene in the F2 generation. Deoxyribonucleic acid was extracted from each 7-day-old embryo of the F2 generation. For the detection of the GFP gene, a GFP-gene-specific primer pair that yielded a 294 bp fragment was used. A medaka-EF-1α-A-gene-specific primer pair that yielded a 519 bp fragment was also used to confirm successful DNA extraction. M, 100 bp ladder. F2 from a transgenic (TG) male, 7-day-old embryos of F2 generation derived from crossing of the F1 transgenic male and non-transgenic female. F2 from a TG female, 7-day-old embryos of F2 generation derived from crossing an F1 transgenic female with a non-transgenic male. Lanes 1–5 and 11–15, embryos that lacked GFP fluorescence in the lens. Lanes 6–10 and 16–20, embryos that showed GFP fluorescence in the lens.

Zygotic expression of the GFP gene

Green fluorescent protein fluorescence in F2 embryos, derived from crossing F1 transgenic males with non-transgenic females, was observed during embryonic development ( Fig. 4). No embryo showed any fluorescence until the late blastula stage ( Fig. 4E). At the early gastrula stage (one-third epiboly), about half the embryos began to show a weak fluorescence in the entire blastoderm and a rather intense fluorescence in the shield ( Fig. 4F, left part), but the other half did not ( Fig. 4F, right part). This fluorescence became more intense in the embryonic body in 1-day-old embryos ( Fig. 4G). In 7-day-old embryos (precardial cavity formation stage), the fluorescence was observed in the head, particularly in the lens ( Fig. 4H, left part) in 53% of the embryos, but not in 47% of the embryos ( Fig. 4H, right part; Table 2). It was also confirmed by PCR analysis that the fluorescence- positive embryos inherited the GFP gene, while the fluorescence-negative ones did not ( Fig. 5). These observations suggest that the zygotic expression of the GFP gene began at the stage immediately before the early gastrula stage, supporting previous speculation that GFP is already accumulated in 1-cell stage embryos produced using a transgenic female as a parent.

Discussion

Promoter region essential for gene expression

In the present study, we demonstrated that about 1.8 kbp in the 5′-flanking sequence from the transcription initiation site of the medaka EF-1α-A gene were required to promote high-level expression. In contrast to this, only 277 bp flanking the first exon of EF-1α are required for strong promoter activity in the zebrafish ( Gao et al. 1997 ). Thus, the much longer segment in the 5′-flanking sequence was required for effective promoter activity in medaka. This discordance in the length of the essential 5′-flanking sequence between medaka and zebrafish may in part be due to the experimental method used. Gao et al. (1997) used cultured cell lines to evaluate the promoter activity, whereas we used embryos. In transgenic zebrafish, Higashijima et al. (1997) reported that they used fragments of the α-actin or β-actin gene of zebrafish, 3.9 or 17 kbp in the upstream region, respectively, to express the GFP gene in a manner identical to endogenous expression. While the reason is not yet clear, a much longer 5′-flanking sequence may be required to induce foreign gene expression in transgenic individuals than in cultured cell lines.

EF-1α-A promoter was active in early embryos

The observation that GFP expression began at the early gastrula stage raises the problem of determining the starting stage of zygotic expression of transgenes in fish. Many studies addressed this problem using transgenic medaka ( Chong & Vielkind 1989; Tsai et al. 1995 ; Hamada et al. 1998 ) and zebrafish ( Stuart et al. 1988 ; Amsterdam et al. 1995 ) and examining the transient expression of foreign genes driven by various ubiquitous promoters. All these studies showed that foreign gene expression started at the mid-blastula stage. Recently, Linney et al. 1999 also reported that the expression of a transgene became noticeable 5–7 h (gastrula stage) after fertilization of embryos derived from a transgenic male in zebrafish.

Taking into consideration that there is a time lag between the start of transcription and the appearance of detectable GFP fluorescence, the present results are consistent with those obtained using medaka and zebrafish in which the expression of exogenous genes began at approximately the midblastula stage. This finding indicates that the promoter of medaka EF-1α-A becomes active at the time of onset of zygotic expression. Thus, one characteristic of the medaka EF-1α-A promoter is to induce gene expression in early embryos from the mid-blastula stage, which continues in the later stages, except in the skeletal muscle.

EF-1α-A promoter is active in a wide range of adult organs

In the present study, it was shown that GFP expression in adult individuals was strong in the gill, lens, kidney, ovary and testis, and weak in the heart, spleen, liver, gut, brain, and skin. No expression was observed in the skeletal muscle. In mammals, it is well known that there are several isoforms of EF-1α, which are expressed in a tissue-specific manner in adults. For example, an isoform of EF-1α, designated as S1 in rats and mice ( Lee et al. 1993 ), or EF-1α-2 in humans ( Knudsen et al. 1993 ) and rabbits ( Kahns et al. 1998 ), is mainly expressed in the skeletal muscle, heart, and brain. On the other hand, the isoform designated as EF-1α in humans is weakly expressed in the heart and skeletal muscle. An isoform designated EF-1α-1 in rabbits is expressed in all tissues except in the skeletal muscle. In the case of medaka, we have found that there are two EF-1α genes per haploid genome, designated as EF-1α-A and EF-1α-B ( Kinoshita et al. 1999 ). At present, it is not clear whether any correlation exists between the mammalian and fish isoforms of EF-1α. However, it is likely that the expression pattern of the medaka EF-1α-A is similar to that of human EF-1α or rabbit EF-1α-1, because the medaka EF-1α-A was also not expressed in the skeletal muscle and was only weakly expressed in the brain and heart. Thus, EF-1α-A may be a medaka homolog of human EF-1α and rabbit EF-1α-1. It would be of interest to determine whether the medaka EF-1α-B shows an expression pattern similar to that of EF-1α-2 in rats, mice, and humans.

Medaka EF-1α-A promoter is transcriptionally active during oogenesis

The ovary exhibited strong GFP expression in transgenic females. The frozen sections revealed that GFP was observed in pre- and post-vitellogenesis stage oocytes ( Fig. 3J). In the F1 and F2 generations produced by crossing of transgenic females with non-transgenic males, GFP was observed in all of the 1-cell stage embryos, regardless of whether they carried the GFP gene or not ( Figs 4,5). These findings indicate that the medaka EF-1α-A promoter is active during oogenesis and the transgene product is provided into oocytes.

Recently, Linney et al. (1999) produced transgenic zebrafish lines carrying the GFP gene driven by the regulatory sequences of the X. laevis EF-1α gene, and reported that GFP fluorescence was observed in oocytes and embryos before the onset of zygotic expression in progenies from female transgenics.

Taking into consideration the activity of the EF-1α-A promoter during oogenesis, it is possible to use this promoter as a tool for studies of the functions of maternal factors during early developmental stages. Before zygotic gene expression starts, the early development of animals is known to depend on maternal factors produced by maternal genes during oogenesis ( Davidson 1986; Olsen et al. 1997 ; Yoon et al. 1997 ; Bashirullah et al. 1998 ; Ikenishi 1998; Wolpert et al. 1998 ). A conventional method to investigate the functions of these maternal factors is to modify the phenotypes after direct injection of the molecules of interest into the eggs. The transgenic method may be applied as another tool to modify both the quality and quantity of maternal factors that accumulate as transgene products in oocytes.

Usefulness of GFP-tagged medaka lines

Animal strains with unique genetic markers that can be detected easily are convenient to use as donor strains for transplantation experiments. Some natural genetic markers, such as body colors and allozymes, have been used in cellular transplantation experiments to generate chimeras ( Lin et al. 1992 ; Wakamatsu et al. 1993 ) and also in nuclear transplantation experiments ( Niwa et al. 1999 ). However, these natural markers cannot be detected before the stages at which the embryos express these genetic markers as phenotypes. Transgenic medaka were generated in the current study using GFP as an exogenous genetic marker, which made it easy to identify donor cells or the activity of donor nuclei in living individuals from the early embryonic stages. Currently, medaka transgenic lines homozygous for pEF-1α-A-GFP are being produced by mating heterozygous individuals generated in this study, the progenies of which are marked with intense fluorescence from the 1-cell stage.

Acknowledgements

We thank Dr Umezono and Dr Ogawa for providing pCMX-SAH/Y145F. This work was partially supported by grants from the Bio-Design Project by the Ministry of Agriculture, Forestry and Fisheries, Japan (BDP-00-IV-1-8) and from the Research for the Future Program of the Japan Society for the Promotion of Science.

Ancillary