Precise marker excision system using an animal-derived piggyBac transposon in plants.

Accurate and effective positive marker excision is indispensable for the introduction of desired mutations into the plant genome via gene targeting (GT) using a positive/negative counter selection system. In mammals, the moth-derived piggyBac transposon system has been exploited successfully to eliminate a selectable marker from a GT locus without leaving a footprint. Here, we present evidence that the piggyBac transposon also functions in plant cells. To demonstrate the use of the piggyBac transposon for effective marker excision in plants, we designed a transposition assay system that allows the piggyBac transposition to be visualized as emerald luciferase (Eluc) luminescence in rice cells. The Eluc signal derived from piggyBac excision was observed in hyperactive piggyBac transposase-expressing rice calli. Polymerase chain reaction, Southern blot analyses and sequencing revealed the efficient and precise transposition of piggyBac in these calli. Furthermore, we have demonstrated the excision of a selection marker from a reporter locus in T0 plants without concomitant re-integration of the transposon and at a high frequency (44.0% of excision events), even in the absence of negative selection.

Introduction

Gene targeting (GT) is a relatively new tool that is available for plant genome engineering (Paszkowski ; Lee ; Puchta, 2002; Terada ; Endo ; Yamauchi ; Saika ; Moritoh ; Ono ). We have succeeded in the introduction of targeted point mutations into the rice genome via GT (Endo ; Saika ); herbicide-tolerant and high-tryptophan rice plants were generated by changing the amino acids in acetolactate synthase (ALS) and the α-subunit of anthranilate synthetase (OASA2), respectively, via GT (Endo ; Saika ). GT rice plants that harbor these mutations acquired herbicide BS resistance and tolerance to the tryptophan analog 5MT, and allowed easy selection of GT cells on medium that contained BS or 5MT, respectively. Conversely, positive/negative selection has been used in most cases to isolate rare clones that have undergone targeted gene replacement by GT (Yamauchi ; Moritoh ; Ono ), as this system allows the modification of any gene of interest. However, the positive selectable marker gene should be excluded completely from the target locus to produce mutants in which only the desired mutations are introduced into the target gene.

To date, several methods, such as co-transformation, transposition, and site-specific recombination, have been employed successfully to remove selectable marker genes from transgenic plants (reviewed in Hohn ; Darbani ; Woo ). Among several transposable elements, the maize Ac/Ds transposon system is used widely; this system is reported to be capable of producing maker-free transgenic rice, tomato, tobacco and aspen plants (Goldsbrough ; Ebinuma ; Cotsaftis ). However, a major problem when using a transposable element system for the elimination of marker genes is the residual footprint at the excised site and the low transposition activity in heterologous hosts. Furthermore, the efficiency of marker-free transgenic plant generation using a transposable element system has been demonstrated to be low, because of the tendency for the transposon to reinsert elsewhere in the genome (Goldsbrough ; Ebinuma ; Cotsaftis ).

Marker excision systems using site-specific recombination that include the Cre/loxP system from bacteriophage P1 and the FLP/FRT system from Saccharomyces cerevisiae, have been well studied in plants (reviewed by Wang ; Woo ). A large number of studies has demonstrated that site-specific recombination methods allow the generation of marker-free plants at a high frequency in model plants and crop species (Woo ). Furthermore, Terada succeeded in Cre/loxP-mediated marker elimination from a GT locus. However, marker excision using site-specific recombination leaves dispensable sequences such as the recognition sequences of recombinase at the excised site. In mammalian cells, it has been reported that the recognition sequences of recombinase have the potential to affect adjacent gene expression (Meier ). Therefore, the development of an efficient and accurate marker excision system is an indispensable part of the production of desired point mutation plants via GT.

The piggyBac transposon derived from the lepidopteran cabbage looper moth Trichoplusia ni (Cary ) integrates into the host genome at TTAA elements and has been used for transgenesis and insertional mutagenesis in a variety of organisms that include Drosophila melanogaster (Horn ), chicken (Lu ), mouse (Ding ), and human (Wilson ). Additionally, the piggyBac transposon excises without leaving a footprint at the excised site TTAA elements (Cary ). Recently, Yusa have shown that the piggyBac transposon system enables excision of a selectable marker from a GT locus in the host genome without residual ectopic sequences in mammalian cells.

We have developed a stable and efficient Agrobacterium-mediated transformation system for rice (Toki ), and have recently established an efficient sequential monitoring system for stable transformation by visualization of cells using a non-destructive and highly sensitive visible marker [click beetle luciferase (Eluc)] in rice (Saika ). Using a stable and efficient transformation system and an effective sequential monitoring system for stable transformation of rice, we present evidence that the animal-derived piggyBac transposon is capable of accurate and effective transposase-mediated transposition in plant cells, and suggest that a high-frequency marker excision system for plant genomes could be established using the piggyBac transposon system.

Results and discussion

High-frequency transposase-mediated transposition of the animal-derived piggyBac transposon in plant cells

To test the use of the piggyBac transposon for effective marker excision in plants, we designed two assay systems in rice cells (Figure1). These systems allow the transposition of piggyBac transposon to be visualized as luminescence derived from reconstituted luciferase expression cassettes. The reporter constructs pBL1 and pBL2 (Figure1) carry a rice elongation factor-1α promoter::emerald luciferase (Eluc) cassette that contains the piggyBac transposon that harbors a rice actin terminator and a mutant cytosine deaminase (codA) gene [D314A, mutcodA, (Mahan )] expression cassette as a negative selection marker in the Eluc gene or in the intron in the 5′ untranslated region (5′ UTR), respectively. A codA gene converts the non-toxic 5-fluorocytosine (5-FC) into the toxic 5-fluorouracil (5-FU) (Perera ). Transgenic rice calli with pBL1 or pBL2 are Eluc negative and 5-FC sensitive. Using pBL1, if the piggyBac transposon is excised precisely by the activation of transposase, the Eluc gene is restored and the calli become Eluc positive and 5-FC resistant. However, if the piggyBac excision leaves a footprint, the Eluc gene is inactivated and the calli remain Eluc negative (Figure1a). Using pBL2, the piggyBac-excised calli become Eluc positive and 5-FC resistant regardless of whether the piggyBac excision leaves a footprint or not (Figure1b).

Figure 1

Schematic representation of excision assay to detect transposition of piggyBac transposon as luciferase luminescence in rice calli.(a) The pBL1 reporter constructs carry a Eluc expression cassette that contain a rice actin terminator and a mutcodA expression cassette as a negative selection marker in the Eluc gene. Transgenic rice calli with pBL1 are Eluc negative (−) and 5-FC sensitive (S). Upon precise transposon excision from the reporter locus, transgenic calli become Eluc positive (+) and 5-FC resistant (R). However, if the piggyBac excision leaves a footprint, the Eluc gene is inactivated and the calli remain Eluc negative.(b) The pBL2 reporter constructs carry a Eluc expression cassette that contain a rice actin terminator and a mutcodA expression cassette as a negative selection marker in the 5′ UTR intron. Transgenic rice calli with pBL2 are Eluc negative (−) and 5-FC sensitive (S). The piggyBac-excised calli become Eluc positive and 5-FC resistant, regardless of whether the piggyBac excision leaves a footprint or not. LB, left border; RB, right border.

Seven-day-old rice calli were infected with Agrobacterium that harbored pBL1 or pBL2 and selected against hygromycin sulfate for 4 weeks (Figure2). To evaluate the frequency of piggyBac transposition and the effect of 5-FC negative selection, transgenic lines with a single copy of the reporter constructs (pBL1-5, pBL1-10, pBL2-3 or pBL2-9) were selected by Southern blot analysis using a mutcodA probe (Figures3a,b, S1(A,B), S2(A,B) and S3(A,B)), and were transferred to N6D medium without the antibiotic meropenem, which kills Agrobacterium. If the transgenic lines carried the multi copy of reporter constructs, the transposon excision from one locus, but not other locus, did not confer 5-FC-resistance on the transgenic calli or plants. After 4 weeks of cultivation, transgenic lines pBL1-5, pBL1-10, pBL2-3 and pBL2-9 calli were transformed with Agrobacterium to introduce the control vectors (pBL1-5_c, pBL1-10_c, pBL2-3_c or pBL2-9_c) or an expression vector that encoded either insect piggyBac transposase [ePBase, (Tamura )] (pBL1-5_e, pBL1-10_e, pBL2-3_e and pBL2-9_e) or hyperactive piggyBac transposase, which carries seven amino acid substitutions (I30V, S103P, G165S, M282V, S509G, N570S, and N538K) [hyPBase, (Yusa )] (pBL1-5_hy, pBL1-10_hy, pBL2-3_hy and pBL2-9_hy) under the control of the maize polyubiquitin 1 promoter, and selected with or without 5-FC (Figure2). After 4 weeks of selection, Eluc luminescence was detected on hyPBase and ePBase transgenic rice calli but not on control calli. In addition, the Eluc signal observed on hyPBase rice calli was significantly higher than that on ePBase-expressing rice calli, which indicated that the frequency of hyPBase-inducible transposition was higher than that of ePBase-inducible transposition (Figures3c, S1(C), S2(C) and S3(C)). However, no significant differences in Eluc luminescence were seen between transgenic rice calli grown on N6D with and without 5-FC (Figures3c, S1(C), S2(C) and S3(C)). Furthermore, the Eluc luminescence detected in pBL1-5 and pBL1-10 transgenic rice calli was comparable with that of pBL2-3 and pBL2-9 transgenic rice calli (Figures3c, S1(C), S2(C) and S3(C)), which suggested that the piggyBac transposon was excised precisely from the Eluc expression cassette via the activation of PBase. Thus, pBL1 transgenic rice calli were used for further analysis.

Figure 2

Experimental scheme for piggyBac excision in rice.Transgenic lines with a single copy of reporter constructs were selected by Southern blot analysis, and were transformed with Agrobacterium to introduce the control vector or expression vector of ePBase or hyPBase under the control of the maize polyubiquitin 1 (Ubi-1) promoter. After a 4-week selection period, Eluc luminescence was detected on hyPBase- and ePBase-expressing rice calli. Eluc-positive calli derived from hyPBase-expressing calli were transferred to shoot regeneration medium with or without 5-FC. These plants were subjected to marker excision analysis.

Figure 3

Analysis of hyPBase- or ePBase-induced piggyBac excision from pBL1-5 locus in rice calli.(a) Structure of the pBL1 reporter construct. The emerald luciferase (Eluc) gene was disrupted by the piggyBac transposon (solid triangles) that contained a rice actin terminator and a cytosine deaminase (codA) expression cassette as a negative selection marker. LB, left border; RB, right border. A bar represents the DNA probe for Southern blot analysis.(b) Southern blot analysis using the codA probe shown in (a) and 2 μg of genomic DNA extracted from wild-type (Wt) and the excision assay reporter transgenic rice (pBL1) calli and digested with EcoRI.(c) Bright-field (BF) and LUC luminescence (LUC) images of control (pBL1-5_c, left panels), hyPBase-expressing (pBL1-5_hy, middle panels) and ePBase-expressing (pBL1-5_e, right panels) rice calli. LUC luminescence derived from an Eluc gene restored by piggyBac transposition are detected after a 4-week selection with (+) or without (–) 5-FC. The sensitivity of detection was enhanced for LUC luminescence in ePBase transgenic rice calli (lower panels).

Efficient and precise excision of piggyBac from the reporter locus

To confirm the Eluc expression results, polymerase chain reaction (PCR) and Southern blot analysis were performed with genomic DNA extracted from three independent chimeric calli with Eluc-negative and Eluc-positive pBL1-5_e and pBL1-5_hy, with or without 5-FC treatment. Using PCR analysis, a 5.9-kb band that was specific to the original pBL1 vector was detected by the primers illustrated in Figure4(a) in all transgenic lines (Figure4b). A 1.65-kb band from the piggyBac-excised fragment was detected in pBL1-5_hy transgenic calli regardless of 5-FC treatment, but was observed only in pBL1-5_e transgenic calli that had been treated with 5-FC (Figure4b). Southern blot analysis with a specific mutcodA probe revealed 8.2-kb fragments derived from the full-length reporter construct in all transgenic lines (Figure4c,d). In addition to 8.2-kb fragments, 3.9-kb fragments expected from the piggyBac-excised reporter cassette were detected using the 3′-flanking region of the Eluc gene probes in pBL1-5_hy transgenic calli treated with or without 5-FC (Figure4c,e). Only faint 3.9-kb fragments were observed in pBL1-5_e transgenic calli even in the presence of 5-FC selection (Figure4e), while no 3.9-kb fragments were detected from pBL1-5_c (Figure4e). Similarly, the full-length reporter-specific fragments were recognized in pBL1-10 all transgenic lines, while, the piggyBac-excised reporter-specific fragments were detected in pBL1-10_hy transgenic calli but not pBL1-10_c and pBL1-10_e transgenic calli (Figure S4).

Figure 4

Molecular analysis of hyPBase- or ePBase-induced piggyBac excision in rice calli.(a, c) Structure of the original pBL1 vector and piggyBac-free reporter locus. Arrowheads represent PCR primers. Bars represent DNA probe fragments used for Southern blot analysis.(b) PCR analysis of genomic DNA isolated from wild-type (Wt), control (pBL1-5_c), hyPBase (pBL1-5_hy) and ePBase (pBL1-5_e) transgenic calli treated with (+) or without (−) 5-FC. Bands of 1.65 kb (open arrowhead) that resulted from piggyBac excision are detected in addition to the original 5.9 kb pBL1 bands (filled arrowhead). Primer pairs are as shown in (a).(d, e) Southern blot analysis using a specific codA (D) or 3′-flanking region of Eluc gene probe (e) in wild-type (Wt), control (pBL1-5_c), hyPBase (pBL1-5_hy) and ePBase (pBL1-5_e) transgenic calli treated with (+) or without (−) 5-FC. When digested with AscI/PacI, bands of 3.9 kb (open arrowhead) derived from the Eluc gene restored by piggyBac transposition were detected in addition to the original 8.2 kb pBL1 bands (filled arrowhead).(f) Expression of hyPBase (left) and ePBase (right) in wild-type (Wt), control (C), hyPBase (pBL1-5_hy) and ePBase (pBL1-5_e) transgenic calli treated with (+) or without (−) 5-FC. Relative transcript levels were normalized to OsAct1 mRNA. Error bars represent ± standard deviation (SD) of three individual experiments.(g) Nucleotide sequences of the original pBL1 (top) and piggyBac excision site in the Eluc reporter cassette (bottom) in hyPBase and ePBase transgenic rice calli. Open boxes highlight the duplicated TTAA sequence that flank the piggyBac transposon.

The hyperactive mutant of PBase that contained seven amino acid substitutions has been reported to show a 17-fold increase in excision activity, and increased protein expression compared with wild-type PBase in mouse embryonic stem (ES) cells (Yusa ). Therefore, our results also might be explained by high levels of hyPBase protein expression under the control of the maize polyubiquitin 1 promoter, although quantitative RT-PCR analysis revealed that transcript levels of the hyPBase and ePBase genes in pBL1-5_hy and pBL1-5_e transgenic rice calli were comparable (Figure4f).

In mouse ES and in induced pluripotent stem (iPS) cells, piggyBac-excision-derived footprints were detected in 5% (Wang ), 9% (Woltjen ) or 0.8% (Yusa ) of PBase-mediated excision events. No difference in footprint frequency between hyPBase- and insect PBase-induced piggyBac transposition was seen in mouse ES cells (Yusa ). To determine if the piggyBac transposon leaves a footprint at the excised site in the rice genome, we analyzed the sequence of the 1.65-kb PCR products; 20 clones from each transgenic line of pBL1-5_e, pBL1-5_hy, pBL1-10_e and pBL1-10_hy were sequenced; a TTAA element was restored after transposon removal in all cases (Figure4g). These results indicated that rice cells are just as conducive as animal cells to piggyBac transposition. In mammalian cells, the piggyBac transposon system allows efficient excision of a selectable marker from a GT locus in the host genome without changing any nucleotide sequence (Yusa ). Thus, the development of improved marker excision strategies using piggyBac to eliminate selectable marker genes from GT loci seems a promising strategy.

High-frequency marker excision via piggyBac transposition in T0 plants

Eluc-positive calli derived from pBL1-5_hy and pBL1-10_hy transgenic calli were transferred to shoot regeneration medium with or without 5-FC. Ten plants from pBL1-5_hy-3 line and 20–30 plants from pBL1-10_hy-2, -3 and -4 were regenerated under 5-FC selection, while 20–40 plants from pBL1-5_hy-1–7 and pBL1-10_hy-1–4 were regenerated without 5-FC selection, respectively. PCR analysis using genomic DNA extracted from leaves of regenerated plants showed that a 1.65-kb band from the piggyBac-excised fragment, but not the original pBL1 vector-specific 5.9-kb fragment, was detected in 100 and 75.0% of pBL1-5_hy and pBL1-10_hy 5-FC-resistant regenerated plants, respectively (Tables1 and SI, and Figure S5). A higher proportion of T0 plantlets (on average, 71.7 and 91.0%; range, 44.4–90.0 and 80.0–96.0% of pBL1-5_hy and pBL1-10_hy, respectively, Tables2 and SII) contained the 1.65-kb band and lacked the 5.9-kb fragment without 5-FC selection at the time of regeneration, which indicated efficient marker excision in hyPBase transgenic plants at the callus stage. However, 40.0 and 20.6% of pBL1-5_hy and pBL1-10_hy 5-FC-resistant regenerated plants and 41.0 and 9.0% of pBL1-5_hy and pBL1-10_hy regenerated plants treated without 5-FC, respectively, were shown to harbor a re-integrated piggyBac transposon, as a mutcodA-specific fragment was detected by PCR analysis (Figure S5). Namely, the re-integration frequency of excised piggyBac was 40.0 and 25.4% in pBL1-5_hy and pBL1-10_hy regenerated plants treated with 5-FC and 56.0 and 9.9% in pBL1-5_hy and pBL1-10_hy regenerated plants without 5-FC, respectively. These results suggested that 5-FC negative selection neither enriched piggyBac-excised transgenic rice cells nor suppressed re-integration of the piggyBac. The negative selection by codA might be not stringent or not sufficient for facilitation of the growth of piggyBac-excised rice cells.

Table 1

PCR analysis of piggyBac excision events with 5-FC treatment in pBL1-5_hy-expressing T0 plants

Line no.	No. of T0 plants analyzed	piggyBac excision from reporter locus			Frequency of piggyBac excision
		Without CodA	With CodA	Total	Without re-integration	With re-integration	Total
3	10	6	4	10	60.0	40.0	100

Table 2

PCR analysis of piggyBac excision events without 5-FC treatment in pBL1-5_hy-expressing T0 plants

Line no.	No. of T0 plants analyzed	piggyBac excision from reporter locus			Frequency of piggyBac excision
		Without CodA	With CodA	Total	Without re-integration	With re-integration	Total
1	36	4	12	16	11.1	33.3	44.4
2	40	12	22	34	30.0	55.0	85.0
3	20	3	14	17	15.0	70.0	85.0
4	30	24	3	27	80.0	10.0	90.0
5	21	4	10	14	19.0	47.6	66.7
6	20	5	6	11	25.0	30.0	55.0
7	29	10	12	22	34.5	41.4	75.9
Average					30.7	41.0	71.7

Furthermore, we performed Southern blot analysis with genomic DNA extracted from each three T0 plants of three independent lines with pBL1-5_hy (line nos. 2, 4 and 5), which contained the piggyBac-excised fragment and lacked the original pBL1 vector-specific fragment and the mutcodA-specific fragment as far as analyzed by PCR. In all analyzed T0 plants, a fragment derived from the piggyBac-excised reporter cassette was detected using the 3′-flanking region of the Eluc gene probes (Figure5a,c), whilst the full-length reporter construct and the re-integrated piggyBac transposon were absent using the 3′-flanking region of the Eluc gene probes and a specific mutcodA probe, respectively (Figure5a,b). However, the fragment derived from full-length reporter construct was only detected in the pBL1-5_c T0 plant. Self-pollinating T1 progeny of these pBL1-5_hy T0 plants were obtained for further analyses.

Figure 5

Southern blot analysis of the reporter locus in pBL1-5_hy T0 plants.Southern blot analysis with genomic DNA extracted from pBL1-5 and each three T0 plants of three independent lines with pBL1-5_hy (line nos. 2, 4 and 5) was performed using a specific codA (a) or 3′-flanking region of the Eluc gene probe (b) shown in Figure4(c). Filled and open arrowheads represent the original 5.9 kb pBL1 bands and 1.65 kb bands derived from the piggyBac-excised reporter cassette, respectively.

It has been reported that 60% of piggyBac excision events are not accompanied by re-integration of the transposon in mouse ES cells without negative selection (Wang ). In addition, previous studies have demonstrated that piggyBac preferentially re-integrates close to the excision site (Wang ; Li ), a finding that suggested that it might be difficult to segregate re-integrated piggyBac transposons from the donor site in T1 progeny. Recently, Li have exploited an excision-competent/integration-defective PBase. Thus, negative selection may not be absolutely required for the generation of marker-free rice plants using the piggyBac transposon system. Accordingly, our results suggested that the piggyBac transposon system could be used widely to eliminate the selectable marker in several plant species, regardless of whether the negative selection via 5-FC was effective or not.

Transposons have become valuable tools for genetic manipulation in many organisms, such as insects (Horn , 2003), vertebrates (Ding ; Wilson ; Yusa ; Hackett ), and plants (Koncz ; Sugimoto ; An ). To date, transposable elements have been thought not to be able to transpose effectively in heterologous species. In addition, many transposons, including Drosophila P elements, are non-functional in heterologous species, a finding that suggests that host factors are needed for transposon activity (Handler ). However, there have been multiple reports of successful transposition of the moth-derived piggyBac transposon in heterologous animal species (Horn ; Ding ; Wilson ; Lu ). Sarkar reported the presence of many piggyBac-like sequences in the genomes of a phylogenetically diverse range of organisms that included fungi, plants, insects, crustaceans, urochordates, amphibians, fishes and mammals. In this report, we have shown that piggyBac transposons can also transpose effectively and accurately in rice cells, a finding that suggested that host factors or additional regulatory factors might be widely conserved in many organisms or, alternatively, are not required for piggyBac transposition in rice cells.

The production of marker-free transgenic plants in T1 progeny

PCR analysis was performed with genomic DNA extracted from 34 T1 progeny from pBL1-5_hy-4 to investigate further whether marker-free plants were obtained in T1 progeny. Twenty-one T1 progeny were each found to contain a 1.65-kb band from the piggyBac-excised fragment, but not from the original pBL1 vector-specific 5.9-kb fragment and a mutcodA-specific fragment. In addition, a hyPBase-specific fragment was absent from 11 T1 progeny, a finding that indicated that eight out of the 34 T1 progeny were marker-free transgenic plants that contained piggyBac-excised reporter cassettes and lacked hyPBase expression cassettes.

Furthermore, two (line nos. 1 and 2) and six plants (line nos. 3–8) of pBL1-5_hy-4 T1 progeny with or without the hyPBase expression construct, respectively, were selected and were subjected to Southern blot analysis. Southern blot analysis with the 3′-flanking region of the Eluc gene probes exhibited the presence of 3.9-kb piggyBac-excised reporter fragments in the pBL1-5_hy-4 T0 plant and in all pBL1-5_hy-4 T1 progeny, but not pBL1-5_c T0 plants, and showed the full-length reporter construct only in pBL1-5_c T0 plants (Figure6a–c). Furthermore, no additional fragment derived from a re-integrated piggyBac transposon was observed in all transgenic plants (Figure6b). In agreement with the results of the PCR, the lack of hyPBase expression constructs was seen in pBL1-5_hy-4 T1 progeny (line nos. 3–8, Figure6e). These results suggested that marker-free transgenic progeny that lack the PBase expression cassette can be obtained via segregation by crossing.

Figure 6

Analysis of piggyBac excision and re-integration in T1 progeny.(a) Structure of the original pBL1 vector and piggyBac-free reporter locus. Bars represent DNA probe fragments used for Southern blot analysis.(b, c) Southern blot analysis using a specific codA (b) or 3′-flanking region of Eluc gene probe (c) in wild-type (Wt), T0 regenerated plants that carry pBL1 (pBL1-5) or hyPBase-expressing vector (pBL1-5_hy-4), T1 progeny of pBL1-5_hy-4 (pBL1-5_hy-4 T1).(d) Structure of hyPBase-expressing vector. Bars represent DNA probe fragments used for Southern blot analysis.(e) Southern blot analysis using a specific hyPBase gene probe shown in (d) and 2 μg of genomic DNA extracted from wild-type (Wt), T0 regenerated plants that carry pBL1 (pBL1-5) or hyPBase-expressing vector (pBL1-5_hy-4), T1 progeny of pBL1-5_hy-4 (pBL1-5_hy-4 T1) and digested with SpeI.

Ebinuma reported the production of marker-free transgenic tobacco and aspen plants using the multi-autotransformation (MAT) vector system, in which the selectable marker is composed of a chimeric isopentenyl transferase (ipt) gene inserted into the maize transposable element Ac, although the frequency of marker-free transgenic tobacco plants was about 100-fold lower compared with that of hyPBase-mediated marker excision in this study. About 10% of Ac elements excised from the transgenic tomato genome without re-integration, or re-integrated into a sister chromatid that was subsequently lost by somatic segregation (Belzile ), whilst 60 or 44% of piggyBac were shown to be lost from the genome after excision in mouse ES cells and rice cells (this study), respectively, even in the absence of negative selection (Wang ).

Emelyanov reported the first use of Ac/Ds transposable elements from plant hosts in transgenesis in animals by using zebrafish and human cell lines. However, until now, experimental evidence that transposons from animal hosts can indeed be utilized in plant species has been lacking. Here, we present evidence using a highly sensitive visible marker that animal-derived piggyBac transposons can transpose effectively and accurately in plant cells. Taken together, our data provide convincing evidence that the piggyBac transposon is functional and can be used as a marker excision system from GT loci in rice plants (Figure S6). Our results suggest that the piggyBac transposon system has several advantages over other marker excision systems currently utilized in plants and that it could be applied universally not only to model plants but also to important crops.

Experimental procedures

Vector construction

The pBL1 reporter vector (Figure1a) was constructed as follows: (i) The Eluc cDNA was amplified by PCR from pELuc-test (TOYOBO) using primer sets Eluc-F NcoI/Eluc-R ScaI and cloned into the NcoI/blunt ended site between the rice elongation factor-1α promoter (Pef) + first intron + Ω and the transcription terminator of the rice heat shock protein 16.9b gene (Thsp16.9b) in pENTR L1/L2 (Life Technologies), yielding pE(L1-L2)Pef:ELUC:Thsp16.9b. (ii) A 680-bp artificially synthesized fragment that contained the piggyBac inverted-repeat transposable element (IVR) and multi-cloning site (HpaI–IVR–AvrII–AatII–IVR–HpaI) was cloned into the HpaI site (GTGAAC to GTTAAC at +399 bp, silent mutation) introduced by site-directed mutagenesis with the Quick change mutagenesis kit (Agilent Technologies) in the coding region of Eluc gene, yielding pE(L1-L2)Pef:ELpbUC:Thsp16.9b. (iii) The mutant codA (D314A) gene was obtained with the Quick change mutagenesis kit (Agilent Technologies) using the primers 5′-gtctgctttggtcacgatgccgtcttcgatccgtggtatc-3′ and 5′-gataccacggatcgaagacggcatcgtgaccaaagcagac-3′. (iv) A 3.7 kb fragment that contained the rice actin terminator (Tact), cauliflower mosaic virus 35S (CaMV35S) promoter (P35S), a mutant codA gene and the Agrobacterium nopaline synthase (NOS) terminator (Tnos) was digested with AvrII/AatII and integrated into pE(L1-L2)Pef:ELpbUC:Thsp16.9b, yielding pE(L1-L2)Pef:ELpCodAbUC:Thsp16.9b. (v) The 8.4-kb fragment that contained Pef:ELpCodAbUC:Thsp16.9b was re-cloned into the pZD202Hyg vector [a derivative of pZD202 (Kwon ), with a hygromycin resistance (HPT gene) cassette (2X CaMV35S promoter:HPT:T35S/Tnos) to SacII/HindIII site of the multi-cloning site of pZD202] using a Gateway LR clonase reaction (Life Technologies).

The pBL2 reporter constructs (Figure1b) were constructed as follows: (i) The piggyBac IVR fragment was cloned into a HpaI site (GTAC to GTTAAC) introduced by site-directed mutagenesis with the Quick change mutagenesis kit (Agilent Technologies) in the 5′-UTR intron of the rice elongation factor-1α promoter, yielding pE(L1-L2)Pefpb:ELUC:Thsp16.9b. (ii) A 3.7-kb fragment that contained Tact:P35S:mutcodA:Tnos was digested with AvrII/AatII and integrated into pE(L1-L2)Pefpb:ELUC:Thsp16.9b, yielding pE(L1-L2)PefpCodAb:ELUC:Thsp16.9b. (iii) The 8.4-kb fragment that contained PefpCodAb:ELUC:Thsp16.9b was re-cloned into the pZD202Hyg vector using a Gateway LR clonase reaction (Life Technologies).

Expression vectors for hyPBase and ePBase were constructed as follows: (i) A 4.7-kb fragment that contained the maize polyubiquitin1 promoter + hyPBase or ePBase + the Arabidopsis ribulose-bisphosphate carboxylase small subunit gene terminator (TrbcS) was cloned into the HindIII/PacI site in pPZP200MCS [a derivative of pPZP200 (Hajdukiewicz ), with the addition of restriction sites (PacI–AscI–SnaBI) to the multi-cloning site of pPZP200], yielding pPZP/Pubi:hyPBase:TrbcS or pPZP/Pubi:ePBase:TrbcS. (ii) A kanamycin resistance (nptII gene) cassette that harbored the open reading frame (ORF) under the control of the rice actin promoter (Pact) and the Arabidopsis polyA-binding protein terminator, was digested with AscI/SnaBI and cloned into pPZP/Pubi:hyPBase:TrbcS or pPZP/Pubi:ePBase:TrbcS, yielding pPN/hyPBase and pPN/ePBase, respectively. (iii) For construction of the control vector, a cassette of P35S:nptII:Tnos in pENTR L1-L4 (Life Technologies), Pact:GFP:Tact in pENTR R4-R3 (Life Technologies), P2x35S:HPT:Tnos in pENTR L3-L2 (Life Technologies) was combined and cloned into the pZD202 vector (Kwon ) using Multisite-gateway Pro (Life Technologies) in accordance with the manufacturer's protocol (Life Technologies).

Agrobacterium-mediated transformation

The binary vectors described above were transferred into Agrobacterium tumefaciens strain EHA105 (Hood ) by electroporation. Agrobacterium-mediated transformation of rice (O. sativa L. cv. Nipponbare) was performed as described previously (Toki, 1997; Toki ).

Genomic DNA extraction and Southern blot analysis

Genomic DNA was extracted from rice calli using Nucleon PhytoPure (GE Healthcare) in accordance with the manufacturer's protocol. Firstly, 2 μg of genomic DNA was digested with EcoRI or AscI/PacI or SpeI and fractionated in a 1.0% agarose gel. Southern blot analysis was performed in accordance with a standard protocol. Specific DNA probes for the mutcodA gene and 3′-flanking region of the Eluc gene were synthesized with a PCR digoxigenin (DIG) probe synthesis kit (Roche Diagnostics) in accordance with the manufacturer's protocol, using the primers mutcodA (5′-ggcccatggtgtcgaataacgctttacaaac-3′ and 5′-cccgagctctcaacgtttgtaatcgatggct-3′); Eluc (5′-ccaggtactccgacaacaacc-3′ and 5′-ggtgttcctgtcgatgatcttg-3′); hyPBase (5′-atcagatgcctgaggatggac-3′ and 5′-acctgctgttcttcaggacctc-3′).

PCR amplification

Genomic DNA was extracted from small pieces of rice calli using Agencourt chloropure (BECKMAN COULTER) in accordance with the manufacturer's protocol. PCR amplifications were performed with PrimeSTAR GXL DNA polymerase (TaKaRa) using the primer sets as follows: (i) for analysis of piggyBac excision, Pef-F 417 (5′-acaactatatagaccggtgcaaagtg-3′) and Eluc-R 589 (5′-ggttgttgtcggagtacctgg-3′); and (ii) for analysis of piggyBac re-integration, codA D314A-F XbaI (5′-tctagaatggtgtcgaataacgc-3′) and codA D314A-R BamHI (5′-ggatcctcaacgtttgtaatcga-3′).

Observation of LUC luminescence

Rice calli were treated with 0.2 mm Beetle luciferin potassium salt (Promega). LUC luminescence images were taken using a high resolution photon counting camera (C2400-700 VIM camera, Hamamatsu Photonics, Japan) with a 10-min exposure time, and processed using Aquacosmos (Hamamatsu Photonics).

RNA extraction and quantitative RT-PCR analysis

Total RNA was extracted from calli and seedlings of rice calli using an RNeasy Plant Mini Kit (QIAGEN). Quantitative reverse transcription (qRT)-PCR was performed with a Power SYBR Green PCR Master Mix (Life Technologies) and an ABI7300 (Life Technologies) in accordance with the manufacturers' protocols. Primer pairs for qRT-PCR were designed using primer3plus software (http://www.bioinformatics.nl/cgi-bin/primer3plus/primer3plus.cgi) and are as follows: OsAct1 (5′-cattgctgacaggatgagcaa-3′ and 5′-gggcgaccaccttgatctt-3′); hyPBase (5′-gtgatgacctgcagcagaaa-3′ and 5′-gctcacgttgtggctgtaga-3′); and ePBase (5′-tgttttgacggaccccttac-3′ and 5′-tgcggtttaccggtactttc-3′).