1.1. Background

The global population reached 7.7 billion in 2019 ^1,2. This rapid population growth is believed to contribute not only to environmental destruction and desertification but also to critical issues that threaten humanity’s survival, including energy depletion and food shortages. With respect to food, it has been argued that, to ensure human well-being, food production must increase at a rate that exceeds population growth ³. Amid the rising food demand accompanying population growth, the outlook for a food supply sufficient to meet future population increases remains uncertain because of various factors, including crop failures caused by climate change and increased demand for new crops such as biofuel feedstocks. In response to these projections, efforts are underway to develop crop varieties with higher yields or improved adaptability to harsh conditions through breeding. Additionally, sensors are being incorporated into crops to obtain hormonal information for optimizing growth control and evaluating responses to stress and environmental conditions ⁴.

Efficient molecular delivery into plant cells is a key technology supporting both plant breeding and plant sensing applications.

1.2. Methods for Introducing Substances

Methods for introducing substances such as genes or sensors into crops (plants) include the Agrobacterium-mediated method, which uses Agrobacterium, a pathogen that causes diseases in dicotyledonous plants ^5,6; the electroporation method, which applies a high voltage to cells to create reparable micro-pores through which molecules are introduced ⁷; and the particle gun method ^8,9, in which target molecules are adsorbed onto submicron-sized metal particles (e.g., gold or tungsten) and propelled at high speed using high-pressure helium or explosives to perforate cell tissues and introduce the molecules into cells.

The Agrobacterium-mediated method enables efficient and stable genetic transformation in plants susceptible to Agrobacterium infection ¹⁰. However, it presents challenges in plants resistant to infection, where transformation efficiency is lower, and substances other than genes are difficult to introduce. Electroporation creates pores in target cells to facilitate molecular delivery, enabling the introduction of substances other than genes. However, in plant cells, this method requires the preparation of protoplasts by removing the cell wall. Regenerating whole plants from protoplasts is impossible or inefficient ^11,12. Furthermore, optimal conditions for cell wall removal and processing time must be determined for each cell type, making the procedure cumbersome. The particle gun method has the advantage of being applicable to a wide range of plant species, but it also has the drawback of causing substantial damage to the target tissue, resulting in low transformation efficiency ¹³. Therefore, understanding the respective contributions of electrical and mechanical stimuli is important for developing an effective transformation technology for plant cells and tissues.

In this study, we propose a novel method for forming perforations in plant cell walls and membranes to facilitate molecular delivery. The method combines mechanical stimulation imparted by the jet generated during the collapse of electrically induced bubbles with electrical stimulation from the voltage applied for bubble generation. First, the influence of device structure on the magnitude of the mechanical and electrical stimuli was investigated. Next, genes were introduced into plant cells and tissues, and differences in delivery characteristics associated with variations in device structure were evaluated.

3.1. Experimental Setup

The experimental setup is shown in Fig. 2. In this system, the voltage and current between the bubble injector and the counter electrode were measured using an oscilloscope (DS5614A, Iwatsu Electric Corporation, Japan). The behavior of the generated bubbles was observed using a high-speed camera (HPV-X2, Shimadzu Corporation, Japan). Arabidopsis and BY-2 cells were used as the targets for genetic transformation. A pulse power supply (PLT1500, BEX Corporation, Japan) was employed as the power source, forming an electrical circuit in which the bubble injector, counter electrode, and a 1 k\(\Omega\) resistor (HS75 1K J, Arcol, UK) were connected in series. The bubble injector was connected as the anode, while the counter electrode served as the cathode. The counter electrode was positioned 10–20 mm from the tip of the bubble injector and immersed approximately 10 mm into the solution. Because the counter electrode was located sufficiently far from the injector tip, the electric field distribution was primarily determined by the geometry in the vicinity of the injector tip ¹⁴. Therefore, the counter electrode could be approximated as an electrode located at an effectively infinite distance. The voltage waveform between the bubble injector and the counter electrode was measured using a high-voltage probe. The current waveform was measured using a current probe installed in the electrical circuit. Both voltage and current signals were recorded using an oscilloscope. The microscopy observation conditions were as follows. Fluorescence observations were performed using a stereomicroscope (SMZ800N, Nikon Corporation, Japan) equipped with a fluorescence adapter (OLFAS, BEX Corporation, Japan) and a Royal Blue excitation filter. A long-pass filter was used for fluorescence imaging. Band-pass filter observations were performed using an inverted microscope (ECLIPSE Ti series, Nikon Corporation, Japan). For green fluorescent protein (GFP) fluorescence observation, the inverted microscope was equipped with a GFP filter set. The optical characteristics of this filter set consisted of an excitation wavelength of 495 nm and an emission wavelength of 505 nm, corresponding to the absorption and emission peaks of GFP, respectively. All fluorescence images were captured under identical imaging conditions.

3.2. Genetic Transformation Experiments in Arabidopsis

The preparation procedure for the solid medium used for Arabidopsis culture is described as follows. First, 5 g of MES (MES, FUJIFILM Corporation, Japan) was dissolved in 80 mL of ultrapure water (hereinafter referred to as the MES solution). Next, 5.85 g of KOH was dissolved in 95 mL of ultrapure water. This solution was added dropwise to the MES solution to adjust the pH to 5.7, yielding the MES-KOH solution. In this study, 2 L of culture medium was prepared. One package of Murashige and Skoog Medium Mixed Salts (for 1 L, FUJIFILM Corporation, Japan), 10 mL of MES-KOH solution, 200 mg of myo-inositol, 20 g of sucrose, and 30 g of agar powder were added. The total volume was adjusted to 2 L with ultrapure water. The resulting mixture was sterilized by autoclaving at 120°C for 20 min. After sterilization, 2 mL of the medium was dispensed into each well of a 24-well plate (Multiple Well Plate, 24-well, flat bottom, Corning) and allowed to cool and solidify at room temperature. Arabidopsis seeds were surface-sterilized before sowing. The seeds were placed in a tube, and 1 mL of 70% ethanol was added. The suspension was mixed by pipetting and allowed to stand for 15 min. The ethanol was then removed, and 1 mL of 90% ethanol was added. After mixing by pipetting, the suspension was allowed to stand for 10 min. The ethanol was then removed, and one seed was sown onto the surface of the medium in each well. After seeding, the plates were sealed, wrapped with aluminum foil to protect them from light, and stored at 5°C for at least 3 days for cold treatment. Following the cold treatment, the plates were transferred to an incubator (LH-60PFD12-DT, Nippon Medical & Chemical Instruments Corporation, Japan) and cultured at 22°C under a 14 h light period for 2 weeks (Fig. 2(a)).

The transformation solution was prepared as follows. For simplicity, 2 L of solution was prepared in this study. One package of Murashige and Skoog Medium Mixed Salts was added, together with KCl and mannitol to achieve final concentrations of 0.2 mol/L and 1 mol/L, respectively. The total volume was adjusted to 2 L with ultrapure water. A plasmid solution (pCR4_p35S::\(\Omega\)::GFP, pCR4_p35S::\(\Omega\)::GFP-NLS, BEX Corporation, Japan) was added to the wells containing cultured Arabidopsis at a final concentration of 100 μg/mL. One Arabidopsis plant cultured in a single well was defined as one sample. Using the bubble injector, bubbles were generated twice, and a jet stream was applied once to each of the two true leaves to achieve genetic transformation. After treatment, the samples were observed using a fluorescence microscope equipped with a long-pass filter. Subsequently, the presence or absence of gene expression was confirmed by reverse transcription PCR (RT-PCR), as described later. Transformation efficiency was calculated as the percentage of wells exhibiting gene expression among the total number of Arabidopsis plants (4 wells) subjected to transformation in each experiment. The experiment was repeated three times, and the average value was used as the transformation efficiency.

3.3. RT-PCR Method

First, total RNA was isolated from Arabidopsis. The leaves were homogenized using a BioMasher II (Nippi, Japan). The Arabidopsis leaves were placed into the BioMasher II tube, and 10 μL of RAI solution from the RNA isolation kit (NucleoSpin RNA Plant, Macherey-Nagel, Germany) was added. The pestle was then rotated manually to crush the leaf tissue. Subsequently, RNA was isolated from the resulting leaf lysate using the RNA isolation kit according to the manufacturer’s instructions. The cDNA Synthesis Kit (PrimeScript II 1st Strand cDNA Synthesis Kit, Takara Bio, Japan) was used to convert the isolated RNA into cDNA (reverse transcription reaction). For cDNA library construction, oligo(dT) primers were used to prime cDNA synthesis. PCR amplification of the first-strand cDNA was performed using oligo(dT) primers with a TaKaRa PCR Thermal Cycler Dice Standard TP650. A 10 μL reaction was performed in a 0.2 mL PCR tube by combining 1 μL of mRNA (100–250 ng), 1 μL of 10 mM each dNTP, 1 μL of 50 μM oligo(dT), and RNase free water to make up the final volume. The tube was incubated at 65°C for 5 min. It was then removed and immediately chilled on ice. To the mixture, 4 μL of 5\(\times\) PrimeScript™ II Buffer, 0.5 μL of RNase Inhibitor (40 U/μl), and 1 μl of PrimeScript™ II RTase (200 U/μl) were added. The reaction was performed at 42°C for 60 min, followed by enzyme inactivation by incubation at 95°C for 5 min and subsequent cooling on ice.

PCR amplification of GFP and Actin was performed using gene-specific primers and cDNA as the template with a thermocycler (TP650). The reaction mixture consisted of 5 μL of KOD One PCR Master Mix (TOYOBO, Japan), 0.5 μL of a gene-specific primer mix (BEX Co., Ltd.), 1 μL of template cDNA, and distilled water to make up a final volume of 10 μL. PCR was programmed for 30 cycles consisting of denaturation at 98°C for 10 s, annealing at 60°C for 5 s, and elongation at 68°C for 10 s, followed by cooling on ice. After the reaction, one-tenth of the assay mixture was subjected to agarose gel electrophoresis (Mupid®-2plus, ADVANCE Co., Japan) using 1\(\times\) TAE buffer following the usual procedure. The size and amounts of the PCR products were estimated by analyzing 1 μL samples using 1.5% agarose gel (wt/vol) electrophoresis followed by GelRed staining (Biotium, Inc., U.S.A.).

3.4. Genetic Transformation Experiments in BY-2 Cells

For simplicity, 100 mL of solid medium was prepared in this study. First, two bags of Murashige and Skoog Medium Mixed Salts were dissolved in ultrapure water at a rate of two bags per 1 L, and 50 mL of the resulting solution was aliquoted.

Next, 3 g of sucrose, 400 mg of myo-inositol, and 4 mg of thiamine hydrochloride were dissolved in ultrapure water to prepare a total volume of 10 mL, and 250 μL of this solution was added. Subsequently, 250 μL of an 80 mg/mL KH\(_2\)PO\(_4\) solution and 45.2 μL of a 2 mmol/L 2,4-dichlorophenoxyacetic acid (2,4-D) solution were added. The 2,4-D solution was prepared by dissolving the compound in a small amount of ethanol, followed by dilution to the desired volume with ultrapure water. Next, 0.7 g of agar powder was added, and the total volume was adjusted to 100 mL with ultrapure water. The pH was then adjusted to 5.7 using KOH, and the mixture was sterilized by autoclaving at 120°C for 20 min. After sterilization, the solution was dispensed into Petri dishes and allowed to cool at room temperature until solidified.

After solidification, approximately one teaspoon of BY-2 cells was inoculated onto the surface of the medium and cultured at 25°C in an incubator (Personal Cool Incubator, Nippon Medical Instrument Corporation, Japan). When liquid medium was used, it was prepared using the same composition but without agar. Approximately 4 teaspoons of callus from the cultured BY-2 cells were collected and suspended in 1 mL of liquid medium. The suspension was then centrifuged at \(500 \times g\) for 5 min using a centrifuge (Tabletop Refrigerated Centrifuge Model 2000, Kubota Corporation, Japan).

After centrifugation, the supernatant was carefully removed without aspirating the cells. Subsequently, 240 μL of transformation solution was added to adjust the cell suspension volume to approximately 500 μL. Then, 5 μL of plasmid solution was added to achieve a final concentration of 100 μg/mL. After thorough mixing by pipetting, the suspension was aliquoted into four PCR tubes at 30 μL per tube. In this study, each PCR tube was treated as a single sample. Transformation was subsequently performed by introducing bubbles into each suspension 10 times using a bubble injector. After treatment, the cells were cultured under static conditions for one day and then observed using a fluorescence microscope equipped with a bandpass filter and a long-pass filter. Transformation was defined as “present” if one or more fluorescently expressing cells were observed within a single sample. Transformation efficiency was calculated as the percentage of samples exhibiting fluorescent expression among the four samples subjected to transformation in each experiment. This experiment was repeated three times, and the average value was used as the transformation efficiency.

3.5. Power Supply Condition

The power supply conditions used to generate a single bubble are shown in Fig. 3. The applied voltage, pulse width, and pulse interval were set to 700 V, 10 μs, and 2 μs, respectively. A single bubble was generated by applying electrical pulses to the bubble injector 25 times under these conditions.

4.1. Bubble Growth Behavior

In this study, two types of bubble injectors were used. The first consisted of a glass tube with an outer diameter of 1010 μm and an inner diameter of 500 μm, into which a tungsten wire with a diameter of 400 μm was inserted, with the tip positions aligned (hereafter referred to as the electrical injector). The second consisted of a glass tube with an outer diameter of 300 μm and an inner diameter of 110 μm, into which a tungsten wire with a diameter of 100 μm was inserted, with the tip positions aligned (hereafter referred to as the mechanical injector). The rationale for referring to these as the electrical injector and the mechanical injector is described in Section 4.2.

Figure 4 shows the bubble injector tip and the bubble growth process observed using a high-speed camera for the electrical injector, whereas Fig. 5 shows the corresponding results for the mechanical injector. The figures also display the bubble growth curve over time together with the time-dependent changes in voltage and current between the bubble injector and the counter electrode. As shown in Figs. 4 and 5, the electrical injector generated bubbles with a smaller maximum volume than the mechanical injector, whereas the time required for bubble generation was longer. These results indicate that, in the electrical injector, the active tungsten wire electrode remained in contact with the solution for a longer period, delaying bubble-induced current interruption (Fig. 4(b), 0.15 ms; Fig. 5(b), 0.01 ms). Consequently, a state conducive to current flow was maintained. This result indicates that the electrical injector consumed a larger amount of electrical energy during bubble generation. Furthermore, the larger bubble observed with the mechanical injector may be attributed to the earlier onset of electrical discharge. Previous studies have reported that electrical discharge occurs when the tip of the tungsten wire becomes completely covered by a bubble ^18,19. Because the thinner tungsten wire used in the mechanical injector was covered by the bubble more rapidly than that used in the electrical injector, electrical discharge was considered to occur earlier. As electrical discharge supplied additional energy for bubble expansion, the earlier onset of discharge may have contributed to the larger maximum bubble volume observed with the mechanical injector.

4.2. Quantification of Electrical and Mechanical Stimuli

As described in Section 4.1, the terms “electrical injector” and “mechanical injector” are used throughout this study for convenience. The electrical injector denotes the injector exhibiting relatively higher current magnitude and electrical energy, whereas the mechanical injector denotes the injector exhibiting relatively higher kinetic energy. These designations are intended solely to distinguish the dominant characteristics of the injectors and do not imply a complete separation of electrical and mechanical effects.

To quantify the stimulus levels delivered by the electrical injector and mechanical injector, the electrical stimulus was characterized in terms of electrical energy, while the mechanical stimulus was characterized in terms of kinetic energy.

Electrical energy \(E_{e}\) was calculated using Eq. (1):

On the other hand, the kinetic energy \(E_{k}\) was calculated using Eq. (2):

The jet mass, m, was calculated from the solution density and the maximum bubble volume. It was assumed that, upon bubble collapse, a volume of solution equal to the maximum bubble volume is drawn into the tip of injector and forms the jet. Accordingly, the jet mass \(m\) is expressed by Eq. (3):

Substituting Eq. (3) into Eq. (2) yields Eq. (4).

From Eq. (1), the electrical energy \(E_{e}\) supplied during bubble generation was calculated to be 0.024 J for the electrical injector and 0.019 J for the mechanical injector. In contrast, the kinetic energy \(E_{k}\) imparted to the target by the bubble injector was calculated using Eq. (4). Although the jet velocities of the electrical and mechanical injectors were nearly identical (5\(\pm\)0.2 m/s and 5\(\pm\)0.3 m/s, respectively), the kinetic energies differed because of the difference in the maximum bubble volume \(V_{b}\). As a result, the kinetic energy was calculated to be 6.25 μJ for the electrical injector and 22.5 μJ for the mechanical injector. Therefore, the electrical injector delivered a greater amount of electrical energy to the system.

4.3. Results of Genetic Transformation Experiments in Arabidopsis

The results of the genetic transformation experiments in Arabidopsis are shown in Fig. 6. Figs. 6(a) and (b) show fluorescence images of Arabidopsis obtained after the experiments. In Fig. 6(b), green fluorescence, presumed to indicate GFP expression, was observed within the region outlined by the black line. However, similar green fluorescence was occasionally observed in the control samples. This fluorescence is likely attributable to the autofluorescence of plant tissues. Therefore, gene expression analysis was performed following RNA isolation and cDNA synthesis, and the results are shown in Fig. 6(c). Fig. 6(c) shows the amplification band obtained using gene-specific primers. Lanes 1, 3, 5, and 7 correspond to amplification with Actin primers, whereas lanes 2, 4, 6, and 8 correspond to amplification with GFP primers. The Actin primers produced a 657-bp amplicon and were used as a positive control for RNA isolation and RT-PCR. Lanes 1, 2, 5, and 6 represent PCR reactions performed using RNA extracted from the plant samples as the template. In contrast, lanes 3, 4, 7, and 8 represent PCR reactions performed using cDNA synthesized by reverse transcription of RNA extracted from the plant samples as the template. Lanes 1–4 correspond to the control samples, which were not subjected to genetic transformation, whereas lanes 5–8 correspond to the transformed samples. These results suggest that GFP was expressed in the transformed cells. Furthermore, as shown in Fig. 6(c), samples in which the 720-bp band was detected in lane 8 were defined as having undergone “successful transformation of GFP.” The transformation efficiencies calculated based on this criterion are shown in Fig. 6(d). Comparison of the electrical and mechanical injectors revealed that the electrical injector achieved a higher gene transformation efficiency in Arabidopsis.

4.4. Results of Genetic Transformation Experiments in BY-2 Cells

The results of the genetic transformation experiments in BY-2 cells are shown in Fig. 7. Fig. 7(a) shows fluorescence microscopy images acquired using a long-pass filter. Green fluorescence was observed within the cells in the region indicated by the red frame. Higher-magnification observations of this region obtained using a band-pass filter and a 20\(\times\) objective lens are shown in Figs. 7(b) and (c). As shown in Fig. 7(c), BY-2 cells exhibiting bright fluorescence throughout the cell were observed. Subsequently, genetic transformation experiments were performed on BY-2 cells using both the electrical injector and the mechanical injector. The transformation efficiencies obtained with the two injectors are compared in Fig. 7(d). As shown in Fig. 7(d), the electrical injector achieved a higher transformation efficiency in BY-2 cells than the mechanical injector.