Excision of DNA fragments with the piggyBac system in Chrysanthemum morifolium

Chrysanthemum morifolium is one of the most popular ornamental plants in the world. However, as C. morifolium is a segmental hexaploid, self-incompatible, and has a sizable heterologous genome, it is difficult to modify its trait systematically. Genome editing technology is one of the attractive methods for modifying traits systematically. For the commercial use of genetically modified C. morifolium, rigorous stabilization of its quality is essential. This trait stability can be achieved by avoiding further genome modification after suitable trait modification by genome editing. Since C. morifolium is a vegetatively propagated plant, an approach for removing genome editing tools is required. In this study, we attempted to use the piggyBac transposon system to remove specific DNA sequences from the C. morifolium genome. Using the luminescence as a visible marker, we demonstrated that inoculation of Agrobacterium harboring hyperactive piggyBac transposase removes inserted 2.6 kb DNA, which harbors piggyBac recognition sequences, from the modified Eluc sequence.


Introduction
Chrysanthemum morifolium Ramat. is the most economically valuable flower worldwide.It is mainly a hexaploid (2n= 6×= 54) with the loss or gain of several chromosomes (Shibata and Kawata 1986).Genome size of C. morifolium would be fluctuated between cultivars and generally considered to have a large genome (12.3-24.7 Gbp, https://www.asteraceaegenomesize.com/#(Accessed Oct 17, 2021)).The recent study using flow cytometry showed that one of the C. morifolium cultivar, 'Sei-Marin, ' have 7.93 Gb genome (Nakano et al. 2021).The origin of C. morifolium is also complicated, and more than six wild Chrysanthemum species are suggested to be involved in the origin of C. morifolium (Ma et al. 2020).Additionally, most C. morifolium cultivars are selfincompatible (Wang et al. 2014), sexually reproduced for breeding, and propagated asexually for commercial production.Therefore, using genetic markers for breeding a broad range of traits is quite challenging, and the production of cultivars is expensive.Transgenic technology is attractive for adding novel traits to C. morifolium.Several interesting transgenic C. morifolium have been developed, with traits such as insect resistance (Shinoyama et al. 2002) and naturally impossible violet and blue flower colors (Noda et al. 2013(Noda et al. , 2017)).For the commercial use of transgenic C. morifolium, it is legally necessary to confer sterility to these plants to avoid gene transfer in the area where it can pollinate wild species (Aida et al. 2018(Aida et al. , 2020)).Genome editing is one of the potential methods to systematically modify traits, such as sterility.We have previously demonstrated the application of the clustered regularly interspaced short palindromic repeats (CRISPR)-associated protein 9 Abbreviations: Cas9, CRISPR-associated protein 9; CRISPR, clustered regularly interspaced short palindromic repeats; Eluc, emerald luciferase; hyPBase, hyperactive piggyBac transposase; IVR, inverted-repeat transposable element; MS, Murashige and Skoog; NLS, nuclear localization signal; nos, nopaline synthase; npt II, neomycin phosphotransferase II; P35S, cauliflower mosaic virus 35S promoter; PcUbi, Petroselinum crispum ubiquitin4-2; RS, recognition site; SpCas9, Cas9 from Streptococcus pyogenes; Thsp, terminator of Arabidopsis heat shock protein 18.2; Tpab5, terminator of Arabidopsis polyadenylate-binding protein 5. a Present address: Institute of Crop Science, National Agriculture and Food Research Organization, 2-1-2 Kannondai, Tsukuba, Ibaraki 305-8518, Japan Published online June 13, 2023 (Cas9) system in C. morifolium (Kishi-Kaboshi et al. 2017).Recently, Shinoyama et al. (2020) successfully knocked out DMC1 by genome editing and conferred male and female sterility in C. morifolium, which will be essential for cultivating transgenic C. morifolium in the open field under guidelines and directives of the Cartagena Protocol on Biosafety.The next step will be to create commercially available genetically modified C. morifolium that is incapable of exogeneous gene transfer.After the intended traits have been developed, it is necessary to restrict the nuclease activity used for genome editing, such as Cas9 or TALEN, as residual enzyme activity has a risk of unintended genome editing on off-target sites during a long vegetative propagation period.The removal of these exogeneous genes for genome editing enzymes from genome-edited C. morifolium is one of a possible method to avoid the risk of unintended genome editing, including off-target mutation.Therefore, we used the piggyBac transposon system to excise unnecessary DNA sequences from the C. morifolium genome.To the best of our knowledge, this study is the first to undertake such a task.
DNA transposons are genetic elements that can move from one location to another in the host genome.The piggyBac transposon was isolated from the cabbage looper moth Trichoplusia ni (Cary et al. 1989).It is unique because it rarely leaves a trace of transposon insertion on the host genome (Fraser et al. 1996).Hyperactive piggyBac transposase (hyPBase) has a 17fold excision rate from that of the original piggyBac transposase, and its frequency of footprint occurrence is low (approximately 1%) (Yusa et al. 2011).In plant, the piggyBac system with hyPBase was first used to rice (Nishizawa-Yokoi et al. 2014).We used emerald luciferase (Eluc) luminescence to monitor DNA excision from transgene-containing transposon recognition sites to evaluate whether the piggyBac transposon system works in C. morifolium.

Vector construction
The pBI121 vector was used as the backbone in this report.The pBI121 vector was digested with HindIII and EcoRI, and assembled with PCR products using Gibson Assembly Master Mix (NEB, https://www.nebj.jp/).
The artificially synthesized 346 bp DNA fragment of the Arabidopsis alcohol dehydrogenase 5′UTR+Castor bean catalase intron+Simian virus 40 NLS+FLAG tag and the hyPBase fragment amplified by PCR were introduced into the XbaI/SacI site of pE(L1-L2) PcUbi:Tathsp, yielding pE(L1-L2) PcUbi_ADHintNLSflaghyPBase_Tathsp.The pBIN_int_ hyPBase contains an intron sequence from catalase in Ricinus communis before NLS, FLAG tag, and a hyPBase sequence, which is codon-optimized for eudicots.The hyPBase nucleotide sequences in pBIN_hyPBase and pBIN_int_hyPBase are almost identical except for one nucleotide between the FLAG tag and hyPBase sequences.There was no amino acid difference.The inserted sequences in these vectors were confirmed by the Sanger sequencing method.All oligonucleotide sequences used in this work are indicated in Table 1.

Luciferin treatment and detection of luminescence
Whole leaf or transgenic calli were sprayed with D-luciferin solution (300 mg l −1 D-luciferin, 20 mM sodium phosphate, pH 7.0).The luminescence was visualized with an imaging system at Chemi-Hi Resolution mode and an exposure time of 120 min (Chemidoc MP and Image Lab 4.0 software, Bio-Rad, CA, USA).

DNA analysis
Leaves from transgenic plants or calli from Agrobacteriuminoculated samples were collected and stored at −30°C.
Genomic DNA was extracted using a Plant DNA Isolation Reagent (Takara clontech).We used PrimeStarGXL DNA polymerase (Takara clontech) to amplify the DNA.Short Eluc DNA fragments from Ag-hyPBase-inoculated E-RS-luc plants were excised from the agarose gel, purified with Nucleospin Gel and PCR Clean-up kit (Macherey-Nagel), and sequenced using a capillary DNA analyzer.All oligonucleotide sequences are indicated in Table 1.

Results and discussion
In this study, we constructed the system which visualize the DNA excision event by luminescence.First, we generated the pBIN_PcUbi_Eluc_Thsp vector containing the Emerald luciferase (Eluc) gene between the parsley ubiquitin4-2 (PcUbi) promoter and Arabidopsis heat shock protein 18.2 terminator (Thsp) as a positive control (Figure 1A).Transgenic C. morifolium plants (Eluc plant) was generated using the pBIN_PcUbi_Eluc_Thsp vector.
We obtained 20 Eluc plant lines, which were confirmed by PCR to possess the Eluc sequence.We evaluated luciferase activity in leaves from 14 transgenic Eluc plant lines.Five plant lines (#52, #72, #74, #78, and #79) showed clear luminescence, especially two of them (#74 and #79) showed strong luminescence after D-luciferin treatment (Figure 1B).This indicated that the PcUbi-Eluc-Thsp construct and D-luciferin treatment could be used to monitor Eluc activity in C. morifolium.
Next, we generated a pBIN_PcUbi_E-RS-luc_Thsp vector, which contained the E-RS-luc construct between the PcUbi promoter and the Thsp.The E-RS-luc construct have the Eluc sequence with the 2.6 kbp insert DNA, pTpab5nptIIb sequence (Figure 1C).The Eluc sequence in the EL-pTpab5nptIIb-UC sequence could not be translated as luciferase by insertion.When the piggyBac transposase excises DNA sandwiched between two piggyBac IVRs from the genome, the Eluc sequence was recovered and could be translated as luciferase.Transgenic plants of C. morifolium (E-RS-luc plants) were generated using the pBIN_PcUbi_E-RS-luc_Thsp vector.We obtained eight Eluc plant lines (#1, #4, #5, #8, #9, #10, #13, and #14), which were confirmed by PCR to possess 5.5 kbp entire PcUbi_E-RS-luc_Thsp sequence (Supplementary Figure S1).
We incorporated a hyPBase expression cassette on binary vectors to apply piggyBac transposase in C. morifolium.It is known that the transgene expression level in rice callus is increased when the first intron of the castor bean catalase gene was inserted in the transgene (Tanaka et al. 1990).We prepared two binary vectors, pBIN_int_hyPBase and pBIN_hyPBase (NLS, nuclear localization signal; Figure 2A, B), with or without intron sequence from the castor bean catalase gene just after the start codon of the NLS-FLAG-hyPBase open reading frame, with the expectancy of enhancement of transgene expression.
Leaf disks from the Eluc #74 and E-RS-luc #10 plants were inoculated with A. tumefaciens (Ag-hyPBase and Ag-int-hyPBase).Then, it was transformed with either the pBIN_hyPBase or the pBIN_int_hyPBase vector.The Ag-hyPBase-inoculated E-RS-luc #10 plants showed punctate luminescence sites in leaf disks.In contrast, mock-inoculated E-RS-luc plants did not show luminescence at two weeks post-inoculation (Figure 3A).The Ag-hyPBase-inoculated E-RS-luc #1 and #10 plants began to show a punctate luminescence site in the leaf disk one week after inoculation (Figure 3B).The E-RSluc #12 plant was confirmed by PCR to contain truncated E-RS-luc sequence but not to have intact PcUbi_E-RS-luc_Thsp construct (Supplementary Figure S1) and used as a negative control for luminescence.The luminescence sites were more clearly visible in Ag-hyPBase-inoculated E-RS-luc #1 and #10 plants four weeks after inoculation.We counted the number of leaves with luminescence after Agrobacterium inoculation (Figure 3C).The E-RSluc plants #1 and #10 consistently showed luminescence after inoculation with Ag-hyPBase.On the other hand, luminescence was not detected in the Ag-int-hyPBaseinoculated leaves of E-RS-luc plants.
Next, we performed PCR analysis to amplify the Eluc or E-RS-luc sequences for the DNA extracted from leaf disks inoculated with Ag-hyPBase.A small fragment with electromobility similar to the Eluc plant was detected in Ag-hyPBase-inoculated E-RS-luc #1, #10, and #12 plants.In contrast, no such fragment was detected in the E-RS-luc plants inoculated with Ag-int-hyPBase (Figure 4A).This result was consistent with the luminescence of E-RS-luc leaf disks inoculated with Ag-hyPBase or Ag-int-hyPBase (Figure 3B).The DNA band of the PCR-amplified E-RS-luc sequence was faint in E-RS-luc #12 plant.A smaller DNA band than E-RS-luc PCR fragments were shown in Ag-hyPBase or Ag-int-hyPBase-inoculated samples (Figure 4A).This suggested that the E-RS-luc #12 plants had truncated T-DNA, which had the Eluc sequence but would lack the PcUbi promoter sequence, in its genome.After the Ag-hyPBase inoculation, the RS sequence was excised from the E-RSluc #12 plant, but the recovered Eluc sequence might not be functional.
We recovered and sequenced the small DNA fragments of E-RS-luc plants inoculated with Ag-hyPBase from the agarose gel.The conjugated Eluc sequence after the excision of sequences sandwiched between RS sequences from E-RS-luc sequence is expected to differ from that in the Eluc plant (Figure 4B).Sequences from all E-RS-luc plants inoculated with Ag-hyPBase showed a TTA A site (Figure 4C) that did not exist in the Eluc plant.These results indicated that the DNA fragment was derived from E-RS-luc constructs and that the DNA fragment between two piggyBac RS sequences was excised from the E-RS-luc sequence.Despite our expectation that the castor bean catalase first intron would enhance transgene expression, inoculation of Ag-int-hyPBase did not result in excising DNA from the E-RS-luc plant.The intron-mediated enhancement was used to boost transgene expression in various plants, but the enhancement mechanism in monocot and dicot plants would be different (Laxa 2017).For C. morifolium leaf segments, the castor bean catalase first intron did not enhance the expression of hyPBase.It is also possible that the splicing of the inserted intron led irregular transcription of hyPBase gene.It might be possible that other intron sequences and/or other intron position enhance transgene expression in C. morifolium.
We tried to gain regenerated shoots from the E-RS-luc leaf segments, which have an RS-excised-Eluc sequence.We used 89 leaf segments and obtained one shoot from the E-RS-luc #10 plant inoculated with Ag-hyPBase in one experiment.We performed two independent experiments, but the regeneration rates were very low in both experiments.In addition, overgrowth of Agrobacterium often killed the regenerated shoot.Finally, we obtained one shoot from the E-RS-luc #1 plant, two from the E-RS-luc #10 plant inoculated with Ag-hyPBase, and four from the E-RS-luc #1 plant inoculated with Ag-int-hyPBase.However, none had a short DNA fragment by PCR analysis of the E-RS-luc sequence.It would be required to use more leaf segments to obtain the regenerated shoots with the RS-excised-Eluc sequence from the E-RS-luc leaf segments and to restrict growth of Agrobacterium in tissue culture.
In this study, we performed excision of DNA sequences from the C. morifolium genome using the piggyBac system.The occupancy of the luminescence site in inoculated leaf disks suggests that the frequency of DNA excision is not very high as expected.However, we can detect excision events on our usual transformation scale using 100-400 leaf disks.The leaf segments with luminescent spots were observed in more than 50% of the Ag-hyPBase-inoculated E-RS-luc leaves of plants #1 and #10.The regeneration efficiency from the E-RSluc transgenic plant inoculated with Ag-hyPBase was not as good as the usual transformation process using non-transgenic C. morifolium (data not shown).A larger-scale experiment will yield DNA-excised plants using our E-RS-luc transgenic plant.We maintained transgenic C. morifolium plants vegetatively in a plant box and used them for analysis.Continuous cultivation under aseptic conditions and consecutive transformation processes might badly influence regeneration.We believe it is possible to generate a more efficient system to obtain DNA-excised shoots with the improvement of the Agrobacterium inoculation and regeneration conditions and the development of a more efficient piggyBac expression system with an intron enhancer in C. morifolium tissue culture.Otherwise, it would be possible to perform genome editing first and then remove the genome editing tool using piggyBac with an inducible expression system in one transformation event to avoid low regeneration efficiency.
We used approximately 1.8-kbp DNA fragments to evaluate hyPBase activity in C. morifolium.The length of Cas9 from Streptococcus pyogenes (SpCas9) is 4.1 kbp.The lengths of commonly used plant promoters and terminators are as follows: 0.3 kbp for cauliflower mosaic virus 35S promoter, 0.9 kbp for PcUbi promoter, 0.3 kbp for nos promoter, 0.3 kbp for A. thaliana hsp terminator, and 0.3 kbp for nos terminator.Therefore, a minimal length of approximately 5 kbp is required to express SpCas9.The original piggyBac element is 2.5 kbp (Cary et al. 1989).HyPBase has been demonstrated to excise 4.3-kbp DNA sequences in rice (Nishizawa-Yokoi et al. 2014).The piggyBac can transfer 14.3 kbp and 100 kbp DNA sequences in mouse cells (Ding et al. 2005;Li et al. 2011).Therefore, it will be possible to excise the Cas9 expression cassette used for genome editing in C. morifolium.We believe that the excision of a genome editing tool from genome-edited C. morifolium will be one of the solutions applied to generate mutant plants in C. morifolium by genome editing for commercial use.

Figure 1 .
Figure 1.The generation of Eluc and E_RS_luc transgenic C. morifolium plants.A: Structures of pBIN_PcUbi_Eluc_Thsp to generate transgenic Eluc plants.B: The location of each transgenic leaf, with its line numbers, is shown in the left upper panel (a).A photograph of the leaves is shown in the right upper panel (b).Luminescence in leaves of transgenic Eluc plant ((c) left bottom, 20 min exposure; (d) right bottom, 120 min exposure).Leaves from independent transgenic Eluc plants were treated with D-luciferin.Bars indicate 1 cm length.C: Structures of pBIN_PcUbi_E-RS-luc_Thsp to generate transgenic E_ RS_luc plants.Eluc, emerald luciferase; IVR, inverted repeat of piggyBac transposon; LB, left border; nos pro, nos promoter, npt II; neomycin phosphotransferase for kanamycin resistance, PcUbi, ubiquitin4-2 promoter from Petroselinum crispum; RB, right border; Thsp, terminator of Arabidopsis heat shock protein 18.2; Tnos, nos terminator; TPab5, terminator of Arabidopsis polyadenylate-binding protein 5.

Figure 2 .
Figure 2. Structures of hyPBase expressing construction and scheme of the excision of a DNA sequence.A: Structures of the binary vectors to make transgenic A. tumefaciens.LB, left border; RB, right border; PcUbi, ubiquitin4-2 promoter from Petroselinum crispum; nos pro, nos promoter, npt II; neomycin phosphotransferase for kanamycin resistance, Tnos, nos terminator; hyPBase, hyperactive piggyBac transposase; int, intron of caster bean catalase; Thsp, terminator of Arabidopsis heat shock protein 18.2.B: Scheme of the excision of a DNA sequence from the plant by hyPBase.In the E-RSluc transgenic plant, the Eluc gene construct was divided by the RS region and inactive.When hyPBase excises the RS region, the Eluc gene construct becomes intact and active.The position of the primers used to amplify the E-RS-luc or Eluc DNA fragment is indicated by arrow heads.

Figure 3 .
Figure 3. Luminescence in E-RS-luc plants inoculated with Ag-hyPBase.A: Luminescence of leaf segments of Eluc #74 and E-RS-luc #10 plants inoculated with mock, Ag-hyPBase, or Ag-int-hyPBase at two weeks post-inoculation.The left panel (a) indicates the position of each plant and the treatment.Luminescence (middle panel, b) and visible (right panel, c) images are shown.The position of the luminescent site in the E-RS-luc plant is indicated by red arrowheads.Bars indicate 1 cm length.B: Luminescence of Eluc #74 and E-RS-luc #1, #10, and #12 plants inoculated with Ag-hyPBase, or Ag-int-hyPBase at 1 (b and c) and 4 (d and e) weeks post-inoculation (wpi).The positions of each plant and the treatment are indicated in the left panel (a).Luminescence (middle panel, b and d) and visible (right panel, c and e) images are shown.The position of the luminescent site in the E-RS-luc plant is indicated by red arrowheads.The position of the accumulated luminescent outflow from Eluc plants is indicated by black arrowheads.Bars indicate 1 cm length.C: Number of leaf segments with luminescence spots per used leaf segment.PC, positive control; NC, negative control.

Figure 4 .
Figure 4. Eluc DNA fragments were excised from the E-RS-luc plant inoculated with Ag-hyPBase.A: Eluc and E-RS-luc size DNA fragments were detected in E-RS-luc plants inoculated with Ag-hyPBase at two weeks post-inoculation.The position of the E-RS-luc size DNA fragments is indicated by the black arrow.The position of the Eluc size DNA fragments is indicated by the red arrow.B: Alignment of DNA fragments surrounding RS insertion sites of Eluc and E-RS-luc fragments.The red color and under bar indicate the position of the TTAA site recognized by piggyBac transposase.C: Sanger sequencing of the PCR-amplified Eluc fragment from E-RS-luc plants inoculated with Ag-hyPBase.DNA bands were excised from the agarose gel and analyzed.The position of the TTAA site is indicated by the red line.

Table 1 .
Oligonucleotide sequences used in this study.
Copyright © 2023 Japanese Society for Plant Biotechnology