Index References

DNA-based transposable elements, also called terminal-inverted-repeat elements, transpose directly from DNA to DNA, unlike retrotransposons, LINEs (long interspersed nuclear elements) or SINEs (short interspersed nuclear elements) that move via RNA intermediates. Enzymes that catalyze the transposition reaction of DNA-based elements, usually encoded by genes in the elements, are called transposases. Because the elements, substrates for transposases, are present as parts of the chromosomes, transposase proteins need to be transported into the nucleus after they are synthesized on the ribosomes in the cytoplasm. Many proteins that act in the nucleus are known to contain NLSs (Nair et al., 2003), and this is also the case for some highly active DNA-based elements such as Activator of maize (Boehm et al., 1995), Mu of maize (Ono et al., 2002), Tag1 of Arabidopsis (Liu and Crawford, 1998), mariner of Drosophila (Lohe et al., 1997) and BmTc1 of the silkworm (Mikitani et al., 2000).

The Tol2 element of the medaka fish Oryzias latipes (Koga et al., 1996; Koga and Hori, 2001) is a member of the hAT family that includes Activator of maize, hobo of Drosophila and Tam3 of snapdragon (Calvi et al., 1991; Atkinson et al., 1993). Twenty to 30 copies are present in the medaka fish genome and virtually all the copies are autonomous or potentially autonomous, carrying complete transposase genes (Koga and Hori, 2001). Another feature of this element is that chromosomal locations dif-fer from fish to fish even in a single local population (Koga and Hori, 1999), indicating that the elements are continuously moving under natural circumstances. However, its transposition frequency is not as high as that for Activator, transposition being rarely observed over one generation (Koga and Hori, 1999). This fact suggests that Tol2 has already acquired, in its evolution, a mechanism to repress transposition. One possible factor might be regulation of the localization of the transposase, because it might be assumed to have a significant role in determining the transposition frequency.

We started the present study for the purpose of determining whether the Tol2 transposase possesses a functional NLS. For this purpose, we prepared a plasmid carrying, as shown in Fig. 1, the SV40 promoter, the bacterial lacZ gene and a poly(A) addition signal, in this order. A cloning site was placed just after the transcription start codon (ATG) of the lacZ gene. This plasmid is a cassette into which a DNA fragment to be examined can be inserted.

View Details

Fig. 1. Cassette plasmid and cDNA portions tested. Note that the components are not strictly to scale. The top line illustrates the structure of the Tol2 element (DDBJ/EMBL/GenBank accession number D84375; Koga et al. 1996), which contains the terminal inverted repeats (white small triangles), internal inverted repeats (white laterally long triangles) and four exons (black arrows). Positions of the first and last nucleotides of the exons on the D84375 sequence are shown with numerals. The second line shows the mRNA for the Tol2 transposase (DDBJ AB031079; Koga et al., 1999). The positions of the first methionin codon and the stop codon are indicated. The next two lines are portions inserted into the cassette plasmid shown underneath. Fragments A and B represent indicated nucleotides of the AF031079 sequence. These portions were amplified by PCR and an ApaI recognition site and an XhoI site were then added at the 5' and 3' ends, respectively, by involving sequences for these sites in the PCR primers. The PCR products were once cloned into pHSG399 (Takeshita et al., 1987) and, after verification of the entire sequences, transferred to the cassette plasmid by digestion with ApaI or XhoI and subsequent ligation. The cassette plasmid was constructed by combining fragments from pSI (Promega corp.; DDBJ U47121) and pSV-β-Galactosidase (Promega corp.; DDBJ X65335). Further sequence alteration was also made using synthesized oligonucleotides. pSI provided the plasmid backbone, the SV40 promoter, and a poly(A) addition signal. The lacZ gene was taken from pSV-β-Galactosidase (abbreviated by pGa). The ranges used as components are shown with arrows and nucleotide positions on the sequences of the original plasmids. The two underlined hexanucleotide regions are cloning sites into which DNA fragments to be examined were inserted. The thick underlined trinucleotide regions are the translation start codon (ATG) and the stop codon (TAA).

For our experiments, it was a prerequisite that the cassette plasmid does not have any signal concerning nuclear transport or export, or that signals, if present, are appropriately balancing. We therefore introduced the intact cassette plasmid into the human HeLa cells and the mouse NIH/3T3 cells and confirmed them to be evenly stained (Fig. 2). This result indicates that our system meets the above requirement. Another necessary condition was that the system is capable of detecting an NLS. For this examination, we prepared a plasmid that carries a sequence for the NLS of the SV40 large-T antigen (amino acid sequence PKKKRKV; Kalderon et al., 1984; Lanford and Butel, 1984). We introduced the plasmid into cells. In accordance with the expectation, the blue color due to β-galactosidase was observed in the nucleus but not in the cytoplasm (Fig. 2).

View Details

Fig. 2. Detection of the fusion proteins. Cells were cultured in Dulbecco’s modified Eagle medium (DMEM) with 10% fetal bovine serum (FBS) and antibiotics, at 37°C in a humidified atmosphere of 5.0% CO2. DNA introduction was conducted using PolyFect Transfection Reagent (QIAGEN GmbH), with 2 × 105 cells per dish seeded in 60 mm dishes 24 h prior to the introduction. The total amount of DNA for each dish was 3.0 μg and 2.5 μg for HeLa and NIH/3T3 cells, respectively. After 24 h of incubation to allow DNA intake and gene expression, cells were washed twice with phosphate-buffered saline (PBS), and fixed with formamide. Staining of cells by supplying substrates for β-galactosidase was conducted using the β-Galactosidase Enzyme Assay System (Promega corp.). Three independent assays were conducted for each plasmid. In all the assays, 10 to 30% of cells survived the DNA introduction, and 5 to 20% of surviving cells were stained, exhibiting blue color in the nucleus and/or the cytoplasm. The staining pattern was sufficiently uniform in that more than 90% of the stained cells in every dish exhibited the same pattern. After the β-galactosidase staining, the cells were stained with DAPI, which binds specifically to DNA and therefore indicates the location of the nucleus when irradiated with fluorescent light. Cells exhibiting typical β-galactosidase staining patterns are shown here. A negative control experiment for the lacZ staining was also conducted using plasmid DNA not carrying the lacZ gene, resulting in no stained cells (data not shown).

To examine the subcellular distribution of the entire Tol2 transposase, we made a plasmid containing the entire sequence for the transposase in the cassette plasmid (Fig. 1, fragment A). With this construct the cytoplasm was stained far more densely than the nucleus (Fig. 2), suggesting that the cDNA contains a signal preventing transport of the product protein into the nucleus, or causing export to the cytoplasm. Because these two could not be distinguished in our present system, we termed the putative signal an extranuclear localization signal (ELS). A signal having the latter function is called a nuclear export signal (NES) (Wen et al., 1995; Bogerd et al., 1998; Fukuda et al., 1996). At present, however, the broader-sense term ELS seems to be more appropriate than NES.

To determine the location of the ELS in the amino acid sequence of the transposase protein, we partitioned the mRNA sequence into overlapping fragments and tested their ELS activity. By repeating this process in three rounds, we found that the region for amino acids 302–334 (Fig. 1, Fragment B) exhibits the ELS function (Fig. 2). Another less effective ELS was suggested to exist in the transposase, from the results with other fragments (data not shown), for which we have not conducted further mapping.

We used human and mouse cells in the present study. This is because the infrastructure for this kind of experiment, including cell lines, culture conditions, promoters and transformation systems, is well established for mammals. Actually, we have also tried the same experiments using available medaka fish cells. However, because the survival rate with DNA introduction was extremely low, we could not obtain data. Several conditions need to be improved in order to confirm the results with fish cells, and we are now targeting such improvements. There is information to support the view that the present phenomena might be observed also in fish. First, an intact Tol2 copy can produce a transposase in human, mouse and chicken cells, and the transposase catalyzes transposition there. Second, the pattern of excision footprints in these mammalian and avian cells was found to be similar to that observed in fish (Koga et al., 2003 and unpublished results).

It is a widely accepted idea that natural selection acts against high transposition activity of transposable elements, at least in the short time scale (Lohe et al., 1995; Hartl et al., 1997). Such natural selection would lead to a decrease in number or frequency of autonomous copies, favoring mutations that inactivate transposase genes. In accordance, accumulation of internally deleted, nonauton-omous copies is a phenomenon commonly observed with DNA-based transposable elements such as Activator of maize (Fedoroff et al., 1983), hobo of Drosophila (Streck et al., 1986) and P of Drosophila (O'Hare and Rubin, 1983; Nitasaka and Yamazaki, 1994). Another, though not mutually exclusive, response of transposable elements to natural selection would be to acquire a mechanism for repression of their own transposition frequency. An example can been seen with P of Drosophila in which transposition in somatic cells is repressed by alternative splicing of mRNA for its transposase (Laski et al., 1986). Inhibition of transposition by overproduction of transposase, as observed with mariner of Drosophila (Lohe and Hartl, 1996), may be another example. The present study suggested that the transposase of Tol2 contains an ELS. This could be an important factor, acquired by natural selection, to control the transposition frequency. A possible role would be to reduce the number of transposase molecules that catalyze the transposition reaction.

We are grateful to M. Kinoshita (Kyoto University), S. Sugiyama (Nagoya University), D. L. Hartl (Harvard University), H. Mitani (University of Tokyo) and M. Moore (InterMal) for helpful discussion. Human HeLa cells and mouse NIH/3T3 cells were obtained from the HSRRB of the Japan Health Sciences Foundation (registry no. JCRB9004 and JCRB0615, respectively). This work was supported by grant no. 10216205 to A. K. and no. 10216206 to A. T. from the MEXT of Japan, and Basic Science Research Grant from the Sumitomo Foundation to A. K.