2025 年 94 巻 4 号 p. 453-463
We developed a “synchronized germination system” for large-scale transformation of a model tomato cultivar, ‘Micro-Tom’, and used it to generate large quantities of uniform explants for transformation. Following initial germination inhibition in a medium with high concentrations of sugar and other compounds, over 80% of the selected seeds germinated synchronously after being transferred to a germination-induction medium (GI medium). In addition, we used the “synchronized germination system” to investigate the appropriate cotyledon age and selection agent for efficient transformation. To study the effect of cotyledon developmental stages on transformation, ‘Micro-Tom’ seeds were treated with the “synchronized germination system”. Cotyledons cultured for 1, 2, 4, and 6 days in GI medium were transformed using Agrobacterium-pCAMBIA1305.1 for β-glucuronidase (GUS) reporter expression. The results showed a significantly higher GUS-positive shoot formation rate for cotyledons cultured for 2 days compared to those cultured for 4–6 days. Furthermore, using GUS and the anthocyanin regulatory gene VlmybA1-2 as reporters, the effects of kanamycin, hygromycin, and G-418 on shoot selection were compared using 2-day cultured cotyledons. Selection with hygromycin and G-418 proved more efficient than kanamycin, with significantly lower percentages of shoots lacking GUS or VlmybA1-2 traits under G-418 and hygromycin selection. The “synchronized germination system” developed in this study has been applied to large-scale transformation using activation tagging vectors, and a large number of transformants was successfully obtained.
Tomato (Solanum lycopersicum L.) is one of the most important vegetable crops and one of the major products of the food industry worldwide. The high variability of tomato fruits, ranging from the cherry type to the large round or elongated berry type, can be used in fresh markets and processed products such as pastes, juices, sauces, and powders. Tomato fruits are particularly rich in nutritional compounds, such as lycopene, β-carotene, vitamin C, flavonoids, and hydroxycinnamic acid derivatives, which have various health benefits (Agarwal and Rao, 2000; Rao and Agarwal, 2000; Riso et al., 2006; Sies et al., 1992).
Because of their economic importance, relatively small genome size, and ease of transformation, tomatoes have been widely used in plant genetics, physiology, and pathology studies (Dan et al., 2006). However, when ordinary tomato cultivars are used for these studies, the large size of the plants, which are approximately 1 m in height, and the relatively long life cycle of 90–110 days from seed germination to fruit ripening make them unsuitable for laboratory cultivation (Saito et al., 2011).
‘Micro-Tom’ is a miniature dwarf determinate tomato cultivar originally bred for home gardening purposes (Scott and Harbaugh, 1989). It differs from standard tomato cultivars primarily in two recessive genes that confer a dwarf phenotype (Meissner et al., 1997). ‘Micro-Tom’ has some features of model plants such as Arabidopsis thaliana such as small size, allowing up to 1,357 plants·m−2 to be cultivated, as well as a short life cycle of 70–90 days from germination to fruit ripening (Meissner et al., 1997). In functional genomics, various techniques including Fox hunting (Ichikawa et al., 2006), activation tagging (Weigel et al., 2000), and CRISPR-based mutagenesis (He et al., 2024), are utilized. It is important to develop a high-throughput transformation system to apply these techniques to tomato research. Meissner et al. (1997) developed functional genomics techniques for tomatoes, such as gene tagging and promoter trapping, and achieved large-scale Agrobacterium tumefaciens-mediated ‘Micro-Tom’ transformation. Further, Mathews et al. (2003) reported the generation of a large number of ‘Micro-Tom’ transformants via activation tagging. However, the resulting transformants have not yet been deposited with organizations that distribute genetic resources, so they are not easily accessible. The National Bio-Resource Project (NBRP) in Japan provides large-scale mutant ‘Micro-Tom’ bioresources, distributes activation tag lines, and makes various omics databases available (Saito et al., 2011). However, the number of activation tag lines available to date is insufficient for saturation mutagenesis.
Several groups have reported improving the efficiency of Agrobacterium-mediated genetic transformation in ‘Micro-Tom’ using cotyledons as explants for transformation (Chetty et al., 2013; Dan et al., 2006; Nonaka et al., 2019; Park et al., 2003; Pino et al., 2010; Sun et al., 2006). These studies examined the effects of Agrobacterium strains, Agrobacterium concentrations, plant culture conditions, and binary vector type on transformation efficiency. In these studies, the cotyledons of 7–12 day-old seedlings were used for transformation. However, most of these studies defined cotyledon age as the number of days from the start of seed culture. Therefore, it remains unclear at which developmental stage the cotyledons are usable. Furthermore, these studies mainly used kanamycin to select transformed cells; they did not compare kanamycin to other selection agents such as G-418 or hygromycin.
In this study, a synchronized germination system for ‘Micro-Tom’ was developed and used to study the effects of the cotyledon stage on Agrobacterium-mediated genetic transformation. This study also found that hygromycin and G-418 were more effective than kanamycin for selecting transformed shoots. Moreover, the synchronized germination system used in this study can be used for large-scale transformation of functional genomics, such as activation tagging.
Seeds of ‘Micro-Tom’ (TOMJP00001) were provided by the University of Tsukuba Gene Research Center through the NBRP of the Japan Agency for Medical Research and Development (AMED). They were sown in pots and grown at 25 ± 3°C under fluorescent light (16 h of 27 μmol·m−2·s−1 PPFD/8 h of 0 μmol·m−2·s−1 PPFD) for seed production. The harvested seeds were soaked in 0.1 M HCl for 15 min, rinsed with tap water, dried overnight, and stored at 4°C in sealed containers with silica gel.
Germination inhibition and inductionTomato seeds were soaked in 70% ethanol for 2 min and sterilized in 200 mL of 1% sodium hypochlorite solution with 50 μL polyoxyethylene sorbitan monolaurate for 10 min. After sterilization, the seeds were washed five times with sterile water and sown in the medium. Media were dispensed in 40 mL portions into 200 mL culture bottles, and 20–30 seeds were sown in each culture bottle. First, 1/2 MS medium (containing half the concentration of MS inorganic salts, MS vitamins, 30 g·L−1 sucrose, and 8 g·L−1 agar) and germination induction medium (GI medium) (containing only water solidified with 8 g·L−1 agar) were used for the germination of ‘Micro-Tom’ seeds. They were cultured at 25°C under fluorescent light (16 h of 27 μmol·m−2·s−1 PPFD/8 h of 0 μmol·m−2·s−1 PPFD). Next, various concentrations of NaCl, sucrose, and mannitol were added to the GI medium for germination inhibition tests. Seeds were incubated at 25°C in the dark, and 20–30 seeds were sown for each treatment and experiment. Our preliminary results showed that seeds that did not germinate by day 5 of incubation rarely germinated after day 6. Thus, in the germination inhibition test, germination was observed on day 5 and each experiment was repeated 3–4 times. After germination inhibition, the seeds were washed with sterile water and transferred to the GI medium at 25°C under fluorescent light (16 h of 27 μmol·m−2·s−1 PPFD/8 h of 0 μmol·m−2·s−1 PPFD). Twenty-four hours after transferring the seeds to the GI medium, seed germination was observed. Seed germination was defined as the emergence of a radicle protruding from the seed coat. Statistical analysis of germination inhibition tests and differences in germination rates between the different treatments were realized by performing the Shirley-Williams nonparametric test.
Vector construction and Agrobacterium strainsAgrobacterium strains GV3101pMP90RK (Koncz and Schell, 1986), AGL0 (Lazo et al., 1991), and AGLS were used for transformation. AGLS was generated by introducing the helper plasmid pSuperAgro2 (Inplanta Innovations Inc., Kanagawa, Japan) into the AGL0 strain. This plasmid carries ACC deaminase and GABA transaminase genes, which were reported to be effective for the Agrobacterium-mediated transformation of plant cells (Ezura et al., 2008; Nonaka et al., 2008a, b, 2019). The combinations of Agrobacterium strains and binary vectors used in this study, as well as the selection agents and vector applications, are shown in Table 1, and an overview of the genes in their T-DNA region is shown in Figure 1.
Combinations of Agrobacterium strains, binary vectors, and selection agents, and purposes of the respective vectors used.
Schematic representation of the T-DNA region of the vectors used in the transformation of ‘Micro-Tom’. Black arrows indicate the position of the primers used for PCR. The number under each black arrow represents the length (bp) of each PCR product. LB: left T-DNA border, 35T: CaMV polyA signal, HPT: hygromycin B phosphotransferase gene for hygromycin selection, 2×35S: CaMV35S promoter with the dimer of the enhancer region, 35S: CaMV35S promoter, GUSplus: improved version of the GUS reporter gene that has better heat stability than the conventional GUS gene (Chen et al., 2010), NosT: nopaline synthase gene terminator of Agrobacterium, RB: right T-DNA border, NPTII: neomycin phosphotransferase type II gene for kanamycin/G-418 selection, NosP: nopaline synthase gene promoter of Agrobacterium, 4×35S: CaMV35S promoter with the tetramer of enhancer region, TE: translational enhancer of A. thaliana AGP21 gene 5′-untranslated region (Matsui et al., 2012), Vlmyb: anthocyanin regulator VlmybA1-2 from Vitis labruscana (Koshita et al., 2008), HT: A. thaliana heat shock protein 18.2 gene terminator (Matsui et al., 2014), MasP: Agrobacterium mannopine synthase gene promoter, Ori: replication origin of Escherichia coli, AmpR: aminoglycoside phosphotransferase gene for ampicillin/carbenicillin selection for bacteria, ipT: Agrobacterium T-cyt gene terminator.
pCAMBIA vectors were provided by CAMBIA (Black Mountain, ACT, Australia), except for pCAMBIA2305.1. To construct pCAMBIA2305.1, a HindIII/BstEII fragment containing the CaMV35S promoter and the β-glucuronidase (GUS) gene from pCAMBIA1305.1 (GenBank Accession No. AF234297; Fig. 1) were ligated into the HindIII/BstEII-digested pCAMBIA2301 backbone. To construct pKAN43Ag8-Vlmyb, the VlmybA1-2 gene (AB073011) was amplified from pBI121-VImybA1-2 (Koshita et al., 2008) using the primers 5′-TCAAAGGAGAAAAGATGGAGAGCTTAGGAGTTAG-3′ and 5′-CTTCATATTGGCGCGCCTCAGATCAAGTGATTTACTTGTG-3′ and assembled into an NsiI/AscI-digested pKAN43Ag8-GFP backbone using an In-Fusion HD Cloning Kit (Takara Bio Inc., Otsu, Japan). pKAN43Ag8-GFP and pKANAC3K are unpublished plasmids, but have been deposited with Addgene (#178182 and #221545, respectively). The plasmid pSKI074 (Weigel et al., 2000) was provided by the Salk Institute for Biological Studies and is available from Addgene (plasmid #11571). The sequences of pKANACH (Shimizu and Hashimoto, 2013) were deposited in GenBank (Accession No. AB777653).
Agrobacterium-mediated transformationAgrobacterium was grown on solid LB medium with the selective agents in Table 1 for 2–3 days, suspended in a medium containing MS inorganic salts, B5 vitamins, and 30 g·L−1 sucrose (pH 5.8), and adjusted to an OD600 value of approximately 0.2. The transformation of ‘Micro-Tom’ and its culture media was based on the methods of Chetty et al. (2013), Dan et al. (2006), Park et al. (2003), and Sun et al. (2006), with some modifications. Further, the compositions of plant growth regulators in the media were based on Dan et al. (2006). After 5 days of culture in a medium containing 120 g·L−1 sucrose, the seeds were transferred to the GI medium. Cotyledons at 1, 2, 4, and 6 days after culture in the GI medium were then cut transversely (Fig. S1A and B) on a 7-cm diameter filter paper soaked with 3.5 mL of an Agrobacterium suspension. Cotyledon cutting was performed using a Feaser stainless steel surgical blade No. 14 (FEATHER Safety Razor Co., Ltd., Osaka, Japan) and KFI K-34 Tweezers (KOWA forceps industry Co., Ltd., Nagoya, Japan). The filter paper with explants was placed on a dry sterile filter paper to remove the excess Agrobacterium suspension and then placed onto a co-culture medium [MS medium inorganic salts, B5 vitamins, 2 mg·L−1 zeatin, 0.1 mg·L−1 indole-3-acetic acid (IAA), 200 μM acetosyringone, and 8 g·L−1 agar (pH 5.8)] dispensed into a sterile 9-cm Petri dish.
After three days of co-culture, the infected explants on filter paper were transferred to a selection medium [MS inorganic salts, B5 vitamins, 2 mg·L−1 zeatin, 0.1 mg·L−1 IAA, antibiotics as shown in Table 1, one tablet·L−1 Augmentin 250RS (GlaxoSmithKline K.K., Tokyo, Japan), with 30 g·L−1 sucrose (pH 5.8)] and incubation for 3–4 weeks. The selection medium was solidified with 8 g·L−1 agar or 3.2 g·L−1 gellan gum. Shoots formed on this medium were transferred to a shoot elongation medium [MS inorganic salts, B5 vitamins, 0.5 mg·L−1 zeatin, antibiotics (Table 1), one tablet·L−1 Augmentin 250RS, 30 g·L−1 sucrose, and 8 g·L−1 agar (pH 5.8)] and incubated for 1–2 months. Once the shoots elongated to around 2 cm, they were transferred to a rooting medium [MS inorganic salts, B5 vitamins, 1 mg·L−1 indole-3-butyric acid, 50 mg·L−1 kanamycin (or 10 mg·L−1 hygromycin, in the case of pCAMBIA1305.1 and pKANACH), one tablet·L−1 Augmentin 250RS, 30 g·L−1 sucrose, 8 g·L−1 agar (pH 5.8)] and incubated for 2–4 weeks before acclimatization. Acclimatized plants were grown under the same conditions as described above. Transgene presence in these plants was confirmed by kanamycin resistance assays, as described by Pino et al. (2010) with some modifications, involving spraying with 400 mg·L−1 kanamycin 2–4 times during the vigorous growth phase, except in plants transformed using hygromycin.
To evaluate the effect of cotyledon age and selection agent (kanamycin vs. hygromycin) on transformation, 37 explants were infected for each treatment in each experiment, and the number of shoot-forming and GUS-positive shoot-forming explants were counted. The experiments were repeated 3–6 times and the collected data was analyzed by performing Tukey-Kramer’s multiple range test or Welch’s t-test. To evaluate the effect of kanamycin vs. G-418 on transformation, 66–106 explants were infected for each treatment in each experiment, and the number of shoot-forming and anthocyanin-accumulating shoot-forming explants were counted. Each experiment was repeated three times and the data collected was analyzed by Welch’s t-test.
GUS activity assayFor GUS staining, 2–3 mm leaf pieces from regenerated shoots transformed with pCAMBIA1305.1 and pCAMBIA2305.1 were used. The samples were incubated for 24 h at 37°C in microtubes with a 100 μL staining solution containing 50 mM sodium phosphate buffer (pH 7.0), 1 mM 5-bromo-4-chloro-3-indolyl-β-D-glucuronide (X-Gluc), 20% (v/v) methanol, 0.3% (v/v) Triton X-100, 0.1 mM potassium ferrocyanide, and 0.1 mM potassium ferricyanide (Kosugi et al., 1990; Nishihara et al., 1993). After incubation, the tissues were washed, and endogenous pigments were removed by repeated immersion in 70–99% ethanol.
Transgene detectionDNA was extracted according to the methods described by Moriya (https://www.an.shimadzu.co.jp/sites/an.shimadzu.co.jp/files/pim/pim_document_file/an_jp/applications/application_note/18495/p0007.pdf; June 29, 2025) and Takatori et al. (2022), with some modifications. Tomato leaves (approximately 5 × 5 mm) were homogenized in 50 μl of CTAB buffer [2% (w/v) hexadecyltrimethylammonium bromide, 2 M NaCl, 50 mM EDTA, 20 mM Tris-HCl, 0.5% (v/v) 2-mercaptoethanol] and incubated at 94°C for 10 min. Subsequently, homogenized leaves were centrifuged at 14,000 × g for 1 min. The resulting supernatant was then diluted 1:20 with 1/10 TE buffer and subjected to PCR. Each 10 μL reaction mixture consisted of 1× Quick Taq® HS DyeMix (TOYOBO Co., Ltd., Osaka, Japan), 0.2 μM of each primer, and 1 μL of the diluted crude DNA extract. PCR amplification was done in a thermal cycler with the following conditions: DNA polymerase activation for 2 min at 95°C, and 40 cycles of 30 s at 94°C, 30 s at 60°C, and 30 s at 68°C. Primers 5′-GAACTGGTGAACGACGGACT-3′ and 5′-TGCCAAAGCCCTTGAAGTAG-3′, which were designed from the pCAMBIA1305.1 sequence, were used to confirm the presence of the GUS gene. Similarly, the primers 5′-TTCGCAAGACCCTTCCTCTA-3′ and 5′-GCAAGGCTTTGGAGAACTTG-3′, which were designed from the pKAN43Ag8-Vlmyb sequence, were used to detect the VlmybA1-2 gene. To detect the T-DNA of pKANAC3K and pKANACH, the primers 5′-AGTCGTGTCTTACCGGGTTG-3′ and 5′-CCTGACGAGCATCACAAAAATCG-3′ were used. These primers were designed based on the common sequence of the T-DNA region of pKANAC3K and pKANACH (Fig. 1), which represent the replication origin sequence of Escherichia coli. The PCR products were analyzed using agarose gel electrophoresis.
Anthocyanin detectionLeaves of transgenic plants expressing pKAN43Ag8-Vlmyb were sampled and stored at −80°C until use. For anthocyanidin analysis, 100 mg frozen samples were ground into a fine powder and then extracted with 1,000 μl of 1% HCl-MeOH. The soluble fraction was then obtained via centrifugation at 10,000 rpm for 10 min and hydrolysed with a 2-fold volume of 2 N HCl at 100°C for 120 min to obtain anthocyanin aglycones, which were then subjected to HPLC analysis as described by Hashimoto et al. (2002), Uddin et al. (2001, 2002), and Shimizu et al. (2011) and Takatori et al. (2015).
The samples were analyzed using an HPLC system (JASCO Corp., Tokyo, Japan) with a TSKgel ODS-80Ts QA Column (4.6 mm i.d. × 150 mm; Tosoh Corp., Tokyo, Japan), PU-2089 gradient pump, MX-2082-32 mixer, AS-2051 autosampler, MD-2010 diode array HPLC detector, and a CO-2060 column oven. The mobile phase consisted of solvent A (1.5% phosphoric acid) and solvent B (1.5% phosphoric acid, 5% tetrahydrofuran, 20% formic acid, and 25% acetonitrile). Further, the composition of the mobile phase was varied from 70% solvent A and 30% solvent B to 30% solvent A and 70% solvent B over a period of 35 min using computer software. Detection at 525 nm was performed in a 10-μL injection volume with the flow rate set at 0.8 mL·min−1. The anthocyanin content in the transformant leaves was estimated by comparing the total area under the peaks of the samples at 525 nm with the anthocyanidin standards of delphinidin, malvidin, and petunidin (Cayman Chem. Co., Ann Arbor, MI, USA).
In this study, sterilized seeds were first cultured on 1/2 MS medium and incubated at 25°C under fluorescent light. Seeds grown in this medium showed non-uniform germination initiation (Fig. S2A, with dispersed germination initiated on days 3–5). It was hypothesized that the inorganic salts and sucrose in the MS medium inhibited germination. Sterile seeds were then cultured on a GI medium containing 8 g ·L−1 agar without salts and sugars. As a result, the seeds germinated earlier than on 1/2 MS medium, as shown in Figure S2B; however, germination initiation was dispersed on days 2–5. Since the residual sodium hypochlorite from seed sterilization may have inhibited germination, sterilized seeds were soaked overnight in sterile water before sowing on the GI medium; however, germination initiation was still dispersed on days 2–5 (Fig. S2C).
Next, we conducted experiments based on the hypothesis that seed germination could be synchronized by first inhibiting germination in a medium with a high salt or sugar concentration, followed by germination induction via transfer to the GI medium. In the NaCl germination inhibition assay, 17.2 seeds per 20 seeds (88.8%) germinated on GI medium containing 0.75 g·L−1 NaCl in an average of four replicates (Fig. 2A). However, the germination rate was significantly reduced on 1.5–3.0 g·L−1 NaCl medium compared to 0.75 g·L−1 NaCl, and was completely inhibited at 6–24 g·L−1 NaCl. When the seeds cultured in 6–24 g·L−1 NaCl were transferred to the GI medium, a 16.3–18.5 per 20 seeds (81.3–92.5%) germination rate was observed (Fig. 2B). In the GI medium supplemented with 15 g·L−1 sucrose, a 17.0 per 20 seeds (85.0%) germination rate was observed (Fig. 2C). The germination rate on 30 g·L−1 sucrose was significantly reduced compared to that in 15 g·L−1 sucrose; germination was 0–0.3 per 20 seeds at sucrose concentrations of 60–240 g·L−1. When seeds were cultured in 60–240 g·L−1 sucrose and then transferred to the GI medium, a 16.0–18.0 per 20 seeds (80.0–90.0%) germination rate was observed (Fig. 2D). In a 15 g·L−1 mannitol medium, 18.5 per 20 seeds (92.5%) germinated, but germination was reduced in a 30 g·L−1 mannitol medium, with a germination rate of 6 per 20 seeds (30.0%); germination was almost inhibited (0–1 per 20 seeds) at concentrations of 60–240 g·L−1 (Fig. 2E). The seeds grown in 60 g·L−1 and 120 g·L−1 mannitol resulted in 17.5–18.3 per 20 seeds (87.5–92.4%) germination, respectively, after transfer to GI medium; however, the seeds cultured at 240 g·L−1 mannitol did not germinate on the GI medium (Fig. 2F).
Germination inhibition and synchronized germination of ‘Micro-Tom’. (A) Inhibition of germination using NaCl. (B) Induction of germination in the GI medium after NaCl (6–24 g·L−1) treatment. (C) Inhibition of germination using sucrose. (D) Induction of germination in the GI medium after sucrose (60–240 g·L−1) treatment. (E) Inhibition of germination using mannitol. (F) Induction of germination in the GI medium after mannitol (60–240 g·L−1) treatment. The inhibition of germination was determined 5 days after treatment, while germination in the GI medium was determined one day after culturing and 20 seeds were sown in each treatment per experiment. The values provided are the mean ± SE (n = 4). The values with asterisks indicate significant decreases relative to the values obtained for 0.75 g·L−1 NaCl in A, 15 g·L−1 sucrose in C, and 15 g·L−1 mannitol in F (Shirley-Williams nonparametric test at the one-tailed significance level of 5%).
The effect of the duration of germination inhibition on synchronized germination was studied by incubating the seeds in a 120 g·L−1 sucrose medium for 1–7 days and then transferring them to the GI medium. One day after transfer, germination rates varied from 8.7 to 24.0 seeds per 30 seeds (28.9–80.0%) after 1–2 days of germination inhibition (Fig. 3). These germination rates were significantly lower than those observed for seeds incubated for 7 days. This suggests that at least three days of incubation on 120 g·L−1 sucrose medium was required for synchronized germination. Figure S2D shows the germination-induced seeds after incubation with 120 g·L−1 sucrose, exhibiting uniform germination. On the second day of germination induction, the hypocotyl elongated and the seed coat became transparent, allowing the inner cotyledons to appear (Figs. S1A and S2D).
Effect of the germination inhibition duration using 120 g·L−1 sucrose on the induction of seed germination in the GI medium. Seeds were incubated in 120 g·L−1 sucrose for 1–7 days and transferred to the GI medium. Germination was measured one day after culturing in the GI medium and 30 seeds were sown in each treatment per experiment. The values provided represent the mean ± SE (n = 3). Further, values with asterisks indicate significantly decreased germination relative to the 7-day value (Shirley-Williams nonparametric test at the one-tailed significance level of 5%).
After culturing the seeds on a 120 g·L−1 sucrose medium, they were transferred to the GI medium and cultured for 1, 2, 4, and 6 days. Their cotyledons were then infected with Agrobacterium to determine the effect of each cotyledon developmental stage on transformation efficiency. The Agrobacterium strain AGL0-pCAMBIA1305.1 was used with a selection medium containing 25 mg·L−1 hygromycin and 8 g·L−1 agar. The average number of explants that formed shoots per 37 infected explants was 17.6 (47.6%) for day 1 cotyledons, 23.7 (64.0%) for day 2 cotyledons, 8.7 (23.4%) for day 4 cotyledons, and 5.7 (15.3%) for day 6 cotyledons (Fig. 4A). The shoot formation rate was significantly higher for day 1 and day 2 cotyledons than for day 4 and day 6 cotyledons. The number of explants that formed GUS-positive shoots was also significantly higher for day 1 cotyledons than for day 6 cotyledons and was significantly higher on day 2 than on day 4 and day 6 cotyledons (Fig. 4B). DNA was then extracted from GUS-positive shoots for transgene confirmation via PCR; all shoots tested showed GUS gene amplification (Fig. S3A). GUS staining images of the shoots transformed with pCAMBIA1305.1 are shown in Figure S4C.
Effect of cotyledon age on shoot regeneration and GUS-positive shoot production in Agrobacterium infected ‘Micro-Tom’ cultured in a selection medium with 25 mg·L−1 hygromycin. (A) Percentage of explants that formed shoots per 37 infected explants. (B) Percentage of explants that formed GUS-positive shoots per 37 infected explants. The values represent mean ± SE (n = 3–5). Different letters indicate significantly different values based on Tukey–Kramer’s multiple range test at P < 0.05.
These results indicate that cotyledons at earlier germination stages have a higher GUS-positive shoot formation efficiency than cotyledons at more advanced stages. There was no significant difference in the transformation efficiency between the cotyledons on days 1 and 2. However, because the cotyledons on day 1 were still undeveloped, white, stiff, and not easy to excise from the seed, cotyledons on day 2 were used in subsequent experiments.
Evaluation of GUS gene selection efficacies of kanamycin and hygromycinThe effects of the antibiotics used for transformed shoot selection were investigated based on the formation of GUS-positive shoots. In this experiment, day 2 cotyledons were used, and pCAMBIA1305.1 and pCAMBIA2305.1 were used as binary vectors for hygromycin and kanamycin selection, respectively. The selection medium, supplemented with 25 mg·L−1 hygromycin or 100 mg·L−1 kanamycin, was solidified with 8 g·L−1 agar. This experiment was conducted simultaneously with the experiment presented in Figure 4. Specifically, 80.2% and 63.0% of the explants regenerated shoots with kanamycin and hygromycin selection, respectively, with no significant differences between them (Fig. 5A). GUS-positive shoot formation rates were 30.6% and 55.9% for kanamycin and hygromycin selection, respectively, again showing no significant difference (Fig. 5B). However, the percentage of explants with GUS-negative shoots per explant with shoots was 70.7% for kanamycin selection and 18.3% for hygromycin selection, which was significantly lower than that for kanamycin selection (Fig. 5C).
Effect of selection agents on the transformation of ‘Micro-Tom’. (A) Percentage of infected explants that formed shoots upon selection using hygromycin or kanamycin. (B) Percentage of infected explants that formed GUS-positive shoots upon selection using hygromycin or kanamycin. (C) Percentage of shoot-forming explants that formed GUS-negative shoots upon selection using hygromycin or kanamycin. (D) Percentage of infected explants that formed shoots upon selection using G-418 or kanamycin. (E) Percentage of infected explants that formed purple shoots upon selection using G-418 or kanamycin. (F) Percentage of shoot-forming explants that formed only non-purple shoots upon selection using G-418 or kanamycin. The values provided represent the mean ± SE (n = 3), and significant differences were determined by performing Welch’s t-test (**P < 0.01; *P < 0.05).
In this study, AGLS-pKAN43Ag8-Vlmyb (Table 1; Fig. 1) was used for VlmybA1-2 overexpression in ‘Micro-Tom’, and the effect of 100 mg·L−1 kanamycin and 30 mg·L−1 G-418 on shoot formation was evaluated from the rate of anthocyanin-accumulated shoot formation in the selection medium containing 3.2 g·L−1 gellan gum.
Within 1–2 weeks after transformation, purple calluses appeared on the surface of the explants (Fig. S4A), and within 4–5 weeks, shoots with purple leaves or stems were observed among the regenerated shoots (Fig. S4B). Although a small amount of anthocyanins can accumulate in wild-type ‘Micro-Tom’ leaves (Mathews et al., 2003), purple pigmentation throughout the leaf and stem is a specific feature of regenerated shoots following AGLS-pKAN43Ag8-Vlmyb infection. When leaf extracts from purple shoots were examined using HPLC, accumulation of delphinidin, malvidin, and petunidin was detected, whereas no anthocyanidins were detected in wild-type leaves (Fig. S5). These purple shoots exhibited traits similar to transgenic ‘Micro-Tom’ overexpressing the tomato anthocyanin regulator ANT1 (Mathews et al., 2003). DNA was extracted from purple shoots for transgene confirmation via PCR, and all purple shoots tested showed amplification of the VlmybA1-2 gene (Fig. S3B). Thus, the purple shoots are thought to accumulate anthocyanins because of the expression of the VlmybA-1 gene.
In kanamycin selection, green shoots predominated, and some purple shoots were observed among the green shoots (Fig. S4D). In contrast, vigorous purple shoot regeneration was observed more frequently in the G-418 selection (Fig. S4E). In the G-418 selection, shoots were obtained from 75.3% of the explants, whereas the shoot formation rate was 92.5% in the kanamycin selection; however, there was no significant difference between them (Fig. 5D). The percentage of explants with purple shoots was 68.3% in the G-418 selection and 34.2% in the kanamycin selection; the percentage in the G-418 selection was significantly higher than for the kanamycin selection (Fig. 5E). The percentage of explants that formed non-purple shoots among the explants that formed shoots was 9.3% for the G-418 selection and 61.7% for the kanamycin selection; that for the G-418 selection was significantly lower (Fig. 5F). When G-418-selected shoots were transplanted to the rooting medium, all the shoots successfully rooted, and 64 of them were acclimatized (Fig. S4F and G). However, the shoots that showed high anthocyanin expression levels showed poorer leaf and stem development than the wild type shoots.
Functional genomics studiesSeveral experiments were conducted using the activation tagging vector, pKANACH (Shimizu and Hashimoto, 2013). As a result, a total of 926 hygromycin-resistant plants were acclimatized from 2,626 explants (Table 2). Thereafter, 12 acclimatized plants were selected for transgene confirmation via PCR, and of these 12 acclimatized plants, 11 showed transgene amplification (Fig. S3D). For transformations using the activation tagging vectors pSKI074 and pKANAC3K, the acclimatized plants were sprayed with 400 mg·L−1 kanamycin to remove any false-positive (escape) plants. Three days after the application of kanamycin, the leaves of escaped plants (or plants with low transgene expression) turned yellow; the yellow areas eventually turned white (Fig. S6A). Upon transformation with pSKI074, 15.8–31.7% of the infected explants produced acclimatized plants in four independent experiments (Table 2). The percentage of explants that produced kanamycin spray-resistant shoots ranged from 10.1 to 23.4% (Table 2). When pKANAC3K was used for transformation, transgenic plants were selected using G-418. Thereafter, a total of 5,558 cotyledon explants were infected several times, and finally 3,916 (70.5%/explants) independent acclimatized transformants were obtained (Table 2). When sprayed with kanamycin, 3,713 (66.8%) kanamycin-resistant plantlets were obtained (Table 2). Twelve regenerated plants, upon infection with AGL0-pKANAC3K transformants, were selected for transgene confirmation via PCR. All kanamycin-spray-resistant plants showed transgene amplification (Fig. S3C), whereas no amplification was observed in kanamycin-spray-sensitive plants (data not shown).
Production of transgenic ‘Micro-Tom’ using activation tagging vectors.
Some acclimatized plants showed abnormal morphology, with thickened leaves, a dark leaf color, and reduced fertility (Fig. S6B). It has been reported that polyploid plants were regenerated at a relatively high frequency following ‘Micro-Tom’ transformation (Nonaka et al., 2019). The abnormal individuals observed in this study were consistent with the polyploid tomato morphology reported by Hirai and Ezura (2012). When polyploid-like plants were removed from the kanamycin-spray-resistant pKANAC3K transformants via visual observation, the formation rate of transformants with diploid-like traits was 44.3% (representing 2,464 independent plants) (Table 2).
Synchronized seed germination was induced on a GI medium after incubation with 6–24 g·L−1 NaCl, 60–240 g·L−1 sucrose, and 60–120 g·L−1 mannitol. In this study, seeds were harvested under the same growing conditions for the germination inhibition and induction experiments. Synchronization of germination by NaCl or mannitol sometimes failed to inhibit or induce germination, depending on the cultivation conditions of the seed parents. In contrast, the synchronization of germination by 120 g·L−1 sucrose was consistently stable. In addition, sucrose is the most commonly used sugar in tissue culture and is thought to have few deleterious effects (Yaseen et al., 2013). Therefore, this study used germination synchronization with 120 g·L−1 sucrose for transformation.
Generally, during germination, seeds convert polysaccharides to monosaccharides to increase intracellular osmotic pressure, absorb water from the surroundings, and swell radicle cells, leading to germination (Mitsuhashi et al., 2004). The high osmotic pressure conditions in this study may have inhibited the water uptake of the seeds and inhibited germination. After germination inhibition, radicles emerged from the seeds within 8–12 hours of culture in the GI medium, suggesting that the preparatory processes for germination, such as the synthesis of gibberellins and conversion of polysaccharides to hexose in the seeds, are in progress during inhibition.
Several studies on ‘Micro-Tom’ transformation used 7–12-day-old seedlings (Chetty et al., 2013; Dan et al., 2006; Meissner et al., 1997; Nonaka et al., 2019; Park et al., 2003; Sun et al., 2006). Sun et al. (2006) achieved a high transformation efficiency using a ‘Micro-Tom’ cotyledon with fully expanded cotyledons and first leaves. Davis et al. (1991) reported that in tomato transformation, mature leaves had a higher transformation efficiency than immature leaves. Rai et al. (2012) examined the effect of cotyledon age on transformation using cotyledons from 3–9 days after germination in the tomato cultivar ‘H-86’. They found that cotyledons at 6 days after germination had the highest transformation efficiency and that their efficiency decreased if the cotyledons were older or younger than 6 days. It has been proposed that the synthesis of vir inducers involved in Agrobacterium infection of plant cells may be higher in mature explants than in immature explants (Rai et al., 2012). Davis et al. (1991) also reported a high rate of transformation and reduced necrosis when mature tomato leaves were used.
In tissue culture, younger explants tend to be less differentiated and more metabolically active, which generally leads to higher regenerative capacity and transformation efficiency (Mazumdar et al., 2010). Pino et al. (2010) reported that 8-day-old cotyledons had a higher transformation efficiency than 10- and 12-day-old cotyledons in ‘Micro-Tom’ transformation. In our study, cotyledons 2 days after germination induction showed higher transformation efficiencies than cotyledons 4–6 days after germination induction. However, it is possible that cotyledons at more advanced developmental stages could have been transformed more efficiently than younger cotyledons, given that our methods differed from those used in previous studies.
Nonaka et al. (2019) and Chetty et al. (2013) investigated the effect of different Agrobacterium strains on the transformation efficiency of ‘Micro-Tom’. Their results showed that the shoot formation rate during the selection stage was closely correlated with the final transformation efficiency. This study compared the effects of cotyledon developmental stages on transformation during the shoot formation stage, but it did not assess the efficiency of producing acclimatized transformants. However, shoots derived from cotyledons at each developmental stage underwent rooting and subsequent acclimatization under identical conditions. Thus, the final transformation efficiency likely reflected the rate of GUS-positive shoot formation observed at the shoot formation stage.
VlmybA1-2 is an anthocyanin regulator gene from grapevines that causes anthocyanin accumulation in plant cells when overexpressed. Its expression can be visually monitored without any special equipment by using fluorescent proteins or substrates such as X-Gluc, and this has been reported to be useful as a visual marker in kiwifruit and tobacco transformation (Koshita et al., 2008). Purple calluses expressing VlmybA1-2 have been obtained from tomatoes, but plant regeneration from these calluses has not been reported (Koshita et al., 2008). In this study, anthocyanin-accumulating tomato plants were successfully obtained through VlmybA1-2 gene expression, demonstrating the utility of this gene as a marker for evaluating transformed shoot production.
When the VlmybA1-transgenic shoots were transferred to a rooting medium, normal rooting was observed (Fig. S4F). However, among the positive transformants, growth inhibition was observed in plants that accumulated high levels of anthocyanins. Cerqueira et al. (2022) reported that tomatoes overexpressing ANT1 showed thinner purple leaves, a lower seed germination rate, suppressed side branching, increased chlorophyll concentration, and lower photosynthesis rates than the wild-type. A similar phenomenon may occur in ‘Micro-Tom’, which overexpresses the VlmybA1-2 gene, thereby reducing plant vigor.
Kanamycin has mainly been used for the selection of transformed cells in the transformation of ‘Micro-Tom’ cotyledons (Chetty et al., 2013; Dan et al., 2006; Nonaka et al., 2019; Park et al., 2003; Pino et al., 2010; Qiu et al., 2007; Sun et al., 2006). In this study, we found that selection was also possible using hygromycin and G-418, in addition to kanamycin (Fig. 5). It was also suggested that the use of hygromycin or G-418 may reduce escape or the production of shoots with low transgene expression in the selection medium (Fig. 5C and F). Kanamycin selection was reported to be less efficient than hygromycin or G-418 in obtaining transgenic Cucumis sativus (Tabei et al., 1994), Liriodendron species (Li et al., 2022), and Vanda species (Gnasekaran et al., 2014). In contrast, Dan et al. (2006) compared selection efficiency with kanamycin and glyphosate for the transformation of ‘Micro-Tom’ cotyledons, reporting no significant difference in the efficiency of producing transformed plantlets or the escape rate. Khuong et al. (2013) also successfully transformed ‘Micro-Tom’ leaves and regenerated transformants using Basta selection.
In this study, the effects of selection agents on transformation efficiency were evaluated at the shoot formation stage. However, the effects of these different selection agents on the production of acclimatized transformants were not investigated. Nonetheless, as shown in Table 2, experiments using activation tagging vectors resulted in the successful generation of a large number of transformants with kanamycin, G-418, or hygromycin selection. Therefore, ‘Micro-Tom’ can produce transformants with a wide range of selective agents. This may be useful for co-transformation with vectors carrying different resistance genes. In experiments where large numbers of transformants are produced, the number of transformants obtained per explant is important; however, the time required to obtain transformants is also important. Selection with hygromycin and G-418, as shown in this study, could reduce the number of false-positive shoots and contribute to reducing the transformation workload. It is difficult to make a simple comparison between our findings and those of previously reported studies because of differences in Agrobacterium strains, binary vectors, selection methods, and transformation efficiency calculation methods. Table S1 summarizes the differences in the transformation methods and efficiency between our study and previous reports on ‘Micro-Tom’ transformation.
In conclusion, this study successfully developed a synchronized germination system for the tomato model cultivar, ‘Micro-Tom’. We cultured seeds in a medium containing germination inhibitors such as sucrose to suppress germination and subsequently transferred the seeds to a medium free from these inhibitors. The cotyledons derived from this system proved to be suitable for large-scale transformations, including activation tagging.
We wish to thank the National Bio-Resource Project for providing ‘Micro-Tom’ seeds. We are also grateful to Dr. Shozo Kobayashi and Dr. Hiroshi Yakushiji (National Agricultural Research Organization) for providing the VlmybA1-2 gene and Dr. Detlef Weigel (Salk Institute for Biological Studies) for providing the pSKI074 plasmid.
