A. chinensis ‘Qihong’ (QH) and A. valvata ‘Duie’ (DE) cultured in MS medium (Murashige and Skoog basal medium containing 30 g L^-1 sucrose and 8 g L^-1 agar, adjusted to pH 5.8 before autoclaving.) were used as experimental materials. The plantlets were maintained in a growth chamber under controlled conditions: temperature 24-26 °C, photoperiod 16-h light/8-h dark, and light intensity of approximately 2500 lx. Subculturing was performed every four weeks to maintain vigorous growth and ensure material consistency. Healthy, uniformly developed shoots were selected as explant sources for subsequent transformation experiments. Both materials mentioned above originated from the Kiwifruit Germplasm Repository of the College of Horticulture Science, Zhejiang A&F University, Lin’an District, Hangzhou City, Zhejiang Province.

Two Agrobacterium strains were used in this study: A. tumefaciens GV3101 and A. rhizogenes K599. Both strains were obtained from Weidi Biotechnology Co., Ltd. (Shanghai, China). Each strain was stored as 50% glycerol stocks at -80 °C and revived on selective solid media prior to use. GV3101 was cultured on yeast extract peptone (YEP) medium, while K599 was cultured on tryptone yeast medium. Antibiotics were added to maintain plasmid selection: kanamycin (50 mg L^-1) and rifampicin (50 mg L^-1) for GV3101, and kanamycin (50 mg L^-1) and streptomycin (50 mg L^-1) for K599. Single colonies were inoculated into liquid medium of the same composition and grown at 28 °C with shaking at 200 rpm for 12 h before being used for transformation assays.

The binary vector 1380-GFP was used for all transformation experiments. This plasmid carries the GFP (green fluorescent protein) encoding gene under the control of a constitutive promoter and a kanamycin resistance gene for selection of transgenic tissues. The plasmid was introduced into A. tumefaciens GV3101 and A. rhizogenes K599 by the freeze-thaw method and verified by colony PCR prior to use.

Antibiotics used in this study included kanamycin sulfate, streptomycin sulfate, and rifampicin, all purchased from Sangon Biotech Co., Ltd. (Shanghai, China). In plant tissue culture, kanamycin was used as a selective agent at concentrations of 100 mg L^-1 for shoot induction medium and 50–100 mg L^-1 for callus selection, while cefotaxime (cef, 100–400 mg L^-1) was applied to eliminate residual Agrobacterium following co-cultivation.

All chemical reagents used in this study were of analytical grade and purchased from Sangon Biotech Co., Ltd. (Shanghai, China). Key reagents included MgCl₂ (CAS: 7786-30-3), NaCl (CAS: 7647-14-5), NaOH (CAS: 1310-73-2), CaCl₂ (CAS: 10043-52-4), DMSO (CAS: 67-68-5), citric acid (CAS: 77-92-9), betaine (CAS: 107-43-7), acetosyringone (AS, CAS: 2478-38-8), CTAB (CAS: 57-09-0), MES (CAS: 4432-31-9), 6-benzylaminopurine (6-BA, CAS: 1214-39-7), naphthaleneacetic acid (NAA, CAS: 86-87-3), indole-3-butyric acid (IBA, CAS: 133-32-4), indole-3-acetic acid (IAA, CAS: 87-51-4), and trans-zeatin (TZ, CAS: 1637-39-4).

The following media were prepared and used during bacterial culture and plant transformation: TY Solid Medium: 3 g/L Yeast extract, 5 g/L Tryptone, 10 mM CaCl₂, 15 g/L Agar. TY Liquid Medium: 3 g/L Yeast extract, 5 g/L Tryptone, 10 mM CaCl₂. YEP Solid Medium: 10 g/L Yeast extract, 10 g/L Tryptone, 5 g/L NaCl, 15 g/L Agar. YEP Liquid Medium: 10 g/L Yeast extract, 10 g/L Tryptone, 5 g/L NaCl. MS Basal Medium: 4.43 g/L MS salts, 30 g/L Sucrose, 8 g/L Agar, pH 5.8. Shoot Induction Medium: MS Basal Medium supplemented with 100 mg/L cef, 100 mg/L kanamycin, 3 mg/L TZ, and 0.5 mg/L IAA, pH 5.8. Co-cultivation Medium: MS Basal Medium supplemented with 100 μM AS, pH 5.8. Root Induction Medium: MS Basal Medium supplemented with 400 mg/L cef, 0.6 mg/L IBA, pH 5.8. MES Infiltration Buffer: 10 mM MgCl₂, 10 mM MES, 100 μM AS, pH 5.8. MS Infiltration Buffer: 4.43 g/L MS salts, 20 g/L Sucrose, 100 μM AS, 5.88 g/L Citric Acid, 0.5 mg/L Betaine, pH 5.2. Citric acid was included to stabilize pH during infiltration, while betaine served as an osmoticum to enhance bacterial viability and T-DNA delivery.

For nucleic acid extraction, the CTAB buffer consisted of 4 g CTAB, 16.364 g NaCl, 20 mL 1 M Tris–HCl (pH 8.0), and 8 mL 0.5 M EDTA, with the final volume adjusted to 200 mL using distilled water.

Stock solutions of plant growth regulators (6-BA, NAA, IBA, IAA, and TZ) were prepared at 2 mg mL^-1 by dissolving in a small amount of 1 M NaOH and diluting with sterile water. Acetosyringone stock (50 mg mL^-1) was prepared in DMSO. All stock solutions were sterilized by filtration (0.22 μm) and stored at -20 °C until use.

The GV3101, harboring the binary vector 1380-GFP, was employed to establish transformation systems for both A. valvata and A. chinensis tissue-cultured plantlets. The transformation procedure included bacterial activation, preparation of infiltration suspension, explant infiltration, co-cultivation, and selective regeneration.

Bacterial activation and preparation of infiltration suspension: GV3101 stored at -80 °C was streaked on YEP solid medium and incubated at 28 °C for 48 h. A single colony was inoculated into 800 μL YEP liquid medium containing 50 mg/L kanamycin and 50 mg/L rifampicin, followed by shaking incubation at 28 °C and 200 rpm for 12 h. The bacterial culture was expanded to 100 mL under the same conditions until the optical density reached OD₆₀₀ = 0.8-1.2. Cells were collected by centrifugation (5000 × g, 10 min) and resuspended in MS infiltration buffer to a final OD₆₀₀ = 0.8 for subsequent infection.

Explant preparation and agro-infiltration: Healthy, fully expanded leaves from in vitro A. chinensis plantlets were excised into 1 × 1 cm segments and used as explants. The explants were immersed in the prepared bacterial suspension for 25 min with gentle agitation to ensure even contact and bacterial infiltration. After infiltration, the explants were blotted on sterile filter paper to remove excess bacterial solution and reduce contamination during co-cultivation. Co-cultivation: The infiltrated explants were transferred onto co-cultivation medium and incubated in darkness at 26 °C for 2 days to facilitate T-DNA transfer. After co-cultivation, explants were rinsed several times with sterile water containing 200 mg/L cefotaxime to remove residual Agrobacterium, blotted dry, and transferred to the selection medium.

Selection and regeneration: Explants were cultured on shoot induction medium containing the appropriate antibiotics and plant growth regulators, under controlled environmental conditions of 24-26 °C, 16-h light/8-h dark photoperiod, and subcultured every 4 weeks. Developing callus tissues from wound sites were maintained on fresh selection medium for continuous proliferation and subsequent shoot induction. Emerging shoots were excised and transferred onto root induction medium to promote root formation under the same environmental conditions.

The K599, harboring the binary vector 1380-GFP, was used for the genetic transformation of A. valvata tissue-cultured plantlets. The transformation process included bacterial activation, preparation of infiltration suspension, explant infiltration, co-cultivation, and selective regeneration culture.

Bacterial activation and preparation of infiltration suspension: K599 was retrieved from -80 °C glycerol stock and streaked onto TY solid medium, followed by incubation in darkness at 28 °C for 48 h. A single colony was inoculated into 800 μL TY liquid medium containing 50 mg/L kanamycin and 50 mg/L streptomycin, and cultured at 28 °C and 200 rpm for 12 h. The bacterial culture was expanded to 100 mL under the same conditions until OD₆₀₀ reached 0.8-1.2. Cells were collected by centrifugation (5000 × g, 10 min) and resuspended in MES infiltration buffer to an OD₆₀₀ = 0.8 for infiltration. The bacterial suspension was then incubated in the dark at room temperature for 2–3 h to induce virulence gene expression prior to plant tissue infiltration.

Explant preparation and infiltration: Healthy, uniformly growing A. valvata tissue-cultured plantlets were selected as explant sources. Fully expanded young leaves were excised into 1 × 1 cm segments and immersed in the prepared bacterial suspension for 25 min with gentle shaking to allow bacterial adhesion and T-DNA transfer. After infiltration, the explants were blotted on sterile filter paper to remove excess bacterial liquid and prevent overgrowth during co-cultivation.

Co-cultivation: The infiltrated explants were transferred onto co-cultivation medium and incubated in darkness at 26 °C for 2 days. Following co-cultivation, explants were rinsed several times with sterile water containing 200 mg/L cefotaxime, blotted dry on sterile filter paper, and transferred to selective medium for further culture.

Regeneration: Explants were cultured on MS medium containing kanamycin, cefotaxime, and plant growth regulators (0.6 mg/L IBA) for 14 weeks, after which they were transferred to hormone-free MS medium supplemented only with kanamycin and cefotaxime. Cultures were maintained at 24-26 °C under a 16-h light/8-h dark photoperiod and subcultured every 4 weeks. Under K599-mediated transformation, the rol gene cluster in the Ri plasmid endogenously induces root primordia formation without the need for exogenous hormones. Consequently, the induction of transgenic hairy roots occurred directly on the selection medium, where callus proliferation and root differentiation were allowed to proceed simultaneously until distinct hairy roots developed from the wound sites.

A high-pressure propagation-mediated, tissue culture-free transformation system was established for A. valvata using the K599 carrying the 1380-GFP vector. In this method, infiltration and induction were performed directly on the branches of intact plants under greenhouse conditions, enabling in planta transformation without the need for sterile tissue culture.

Preparation of bacterial suspension: The preparation of the K599 bacterial suspension was conducted as described above in Section 2.6, using MES infiltration buffer and adjusting the bacterial concentration to OD₆₀₀ = 0.8 prior to infiltration. Agro-infiltration of intact plants: Healthy A. valvata plants with a stem diameter of 5–9 mm was selected as recipients. On each target branch, longitudinal wounds (approximately 1.0-1.5 cm in length, penetrating to the cambium without damaging the pith) were made using a sterile scalpel. Sterile filter paper strips (1.5 × 3 cm) were soaked in the bacterial suspension and tightly wrapped around the wounded area for 20 min in the dark to facilitate bacterial infiltration and T-DNA delivery. After treatment, the filter paper was removed, and excess bacterial solution on the wound surface was carefully blotted with sterile absorbent paper.

Immediately after infiltration, the treated wound region was enclosed with a black rooting box (polyethylene, approximately 5 × 5 × 8 cm, with side ventilation slits for gas exchange) filled with sterile substrate (vermiculite:matrix = 1:3, particle size 2–4 mm, autoclaved at 121 °C for 30 min). The rooting box was securely fixed to the stem to maintain contact between the infected site and the substrate. The entire plant was cultivated under greenhouse conditions at 24-26 °C, 70-80% relative humidity, and a 16-h light/8-h dark photoperiod. The substrate was kept moist but not waterlogged, and the system was monitored regularly to maintain stable humidity and temperature.

Induction of transgenic tissues: The rol genes on the Ri plasmid of K599 induced cell dedifferentiation and root primordia formation at the infiltrated wound sites without the addition of exogenous hormones. Under greenhouse propagation conditions, transgenic callus and hairy roots gradually developed from the cambial region at the infected sites. Samples of these tissues were collected for molecular and fluorescence analysis once sufficient growth was observed.

Transgenic shoots (experimental group) and wild-type shoots (control group) were thoroughly washed with sterile water. Leaf samples were transversely and longitudinally sectioned using a blade. The sections were observed and imaged using a confocal laser scanning microscope (OLYMPUS FV3000). The excitation wavelength was set at 488 nm, and emission was collected between 505–550 nm. Images were captured using a 10× objective lens. Calculate the genetic transformation efficiency of both kiwifruit species using the following formula: Genetic transformation efficiency = (Number of GFP-positive explants/Total number of explants infected with Agrobacterium) × 100%, in order to quantitatively evaluate the effectiveness of the transformation system. Transformation efficiency was calculated from three independent biological replicates, each consisting of 30–40 explants. Data are presented as mean ± standard deviation (SD).

Genomic DNA was extracted from putative transgenic and wild-type tissues of A. chinensis and A. valvata using a modified CTAB method. Sample preparation and lysis: Fresh leaves or transgenic tissues (approximately 100 mg) were placed in 2 mL centrifuge tubes containing one sterile steel bead. Samples were flash-frozen in liquid nitrogen and ground to a fine powder using a tissue homogenizer (30 Hz, 60 s). Subsequently, 1.0 mL of pre-warmed (65 °C) CTAB extraction buffer (adjusted to pH 8.0 before use) was added, and samples were mixed thoroughly and incubated in a 65 °C water bath for 2 h to ensure complete cell lysis and nucleic acid release. Phase separation and DNA precipitation: After incubation, tubes were centrifuged at 12, 000 rpm for 10 min. Approximately 700 μL of the supernatant was transferred to a new tube, mixed with an equal volume of chloroform:isoamyl alcohol (24:1, v/v), and gently inverted several times. The mixture was centrifuged again at 12, 000 rpm for 10 min, and 600 μL of the upper aqueous phase was transferred to a new tube. An equal volume of isopropanol was added to precipitate DNA, and samples were incubated on ice (or at -20 °C) for 5–10 min. DNA was collected by centrifugation at 12, 000 rpm for 10 min, washed with 1 mL of 70% ethanol, and centrifuged again for 5 min. The supernatant was discarded, and the DNA pellet was air-dried until translucent. Finally, the DNA was dissolved in 30 μL of nuclease-free water (ddH₂O) and stored at -20 °C until use. PCR amplification and verification: The extracted DNA was used as a template for PCR analysis to verify the presence of the GFP transgene. PCR reactions were performed using 2× Rapid Taq Master Mix (Vazyme, China) and the following primer pair specific to the GFP sequence: Forward primer (1380-GFP-F): 5′-TAGTTCATCCATGCCATGTGTA-3′; Reverse primer (1380-GFP-R): 5′-CTTCAAGAGCGCCATGCCTGA-3′. The PCR amplification was carried out in a total volume of 25 μL, including 12.5 μL of 2× Master Mix, 1 μL of each primer (10 μM), 2 μL of genomic DNA template, and 8.5 μL of ddH₂O. The cycling conditions were as follows: initial denaturation at 95 °C for 3 min, followed by 35 cycles of denaturation at 95 °C for 15 s, annealing at 52 °C for 15 s, and extension at 72 °C for 15 s, with a final extension at 72 °C for 5 min. Amplification products were analyzed by 1% agarose gel electrophoresis (150 V, 150 mA, 25 min) and visualized under UV illumination. PCR-positive bands corresponding to the expected fragment size were further purified and sequenced to confirm the presence of the GFP transgene. The expected band size of PCR products was purified and sequenced. The obtained sequences were aligned with the reference GFP coding region using SnapGene software to verify sequence existence.

The expression of the GFP reporter gene in putative transgenic tissues of A. chinensis and A. valvata was examined using both handheld fluorescence illumination and confocal laser scanning microscopy. For preliminary screening, regenerated shoots, callus, and hairy roots were observed under a handheld blue-light excitation source equipped with a green emission filter in a dark environment. The appearance of bright green fluorescence at wound sites or throughout the tissues was recorded as a preliminary indicator of successful transformation, while wild-type tissues were used as negative controls. For detailed cellular-level analysis, fluorescence-positive samples were further examined using a confocal laser scanning microscope (OLYMPUS FV3000, Japan). Fresh leaf or root sections were washed with sterile water and mounted on glass slides. Imaging was performed at an excitation wavelength of 488 nm and an emission range of 505–550 nm, using identical settings for both transgenic and control samples. The presence of uniform green fluorescence within the cytoplasm of transgenic cells was used to confirm the functional expression of the GFP transgene.

To establish a stable and reproducible A. tumefaciens-mediated genetic transformation system for kiwifruit, A. chinensis and A. valvata were selected as representative genotypes. These two species differ significantly in regenerative ability and physiological properties, providing a suitable contrast for system development. Leaf discs from in vitro-grown shoots were infiltrated with GV3101 harboring the 1380-GFP binary vector. The entire process included bacterial activation, infiltration, co-cultivation, callus induction, and regeneration culture (Figure 1A). After infiltration and co-cultivation for 2 days, the explants were transferred to selection medium containing kanamycin and cefotaxime. Callus formation was observed at wound sites of both genotypes approximately four weeks after infiltration (Figure 1B). GFP fluorescence appeared in portions of the induced callus, indicating successful T-DNA delivery. In A. chinensis, the GFP-positive callus gradually developed into green, compact tissues, from which shoot primordia emerged after approximately 12 weeks of culture, followed by complete shoot regeneration at around 16 weeks and subsequent rooting within 20 weeks in selection medium (MS medium containing kanamycin, cef, 3mg L^-1 TZ and 0.5 mg L^-1 IAA) (Figure 1B). In contrast, A. valvata formed proliferative GFP-positive callus with strong growth potential but did not produce adventitious shoots during the same period (Figure 1B). These results demonstrate that the established GV3101-mediated system can achieve stable gene introduction and expression in both A. chinensis and A. valvata, though the morphogenic response differs between genotypes.

A dual verification system combining fluorescence-specific detection and molecular biological validation was employed to confirm the successful integration and expression of the GFP reporter gene in transformed kiwifruit tissues (Figure 2). This two-tier system effectively eliminated false positives caused by residual Agrobacterium cells or non-specific amplification, thereby ensuring the reliability of transformation results. For fluorescence-specific detection, sections of transgenic tissues from A. chinensis and A. valvata were examined under a confocal laser scanning microscope using wild-type tissues as negative controls. Imaging parameters were set precisely to distinguish GFP fluorescence from chlorophyll autofluorescence, excitation at 488 nm, emission at 505–550 nm for GFP, and a chlorophyll autofluorescence peak at approximately 685 nm. The results revealed that all cells of transgenic tissues exhibited bright and uniform green fluorescence, with no evidence of chimerism or mosaic expression (Figure 2A). In contrast, no GFP signal was detectable in wild-type tissues, confirming that the observed green fluorescence originated exclusively from functional GFP protein expression rather than background autofluorescence. For molecular biological validation, genomic DNA was extracted from both transgenic and wild-type leaves. PCR amplification with GFP-specific primers (Forward primer (1380-GFP-F): 5′-TAGTTCATCCATGCCATGTGTA-3′; Reverse primer (1380-GFP-R): 5′-CTTCAAGAGCGCCATGCCTGA-3′) produced a clear, single band at the expected size (460 bp) in transgenic samples, whereas no amplification was detected in wild-type controls (Figure 2B). The PCR-positive fragments were purified and subjected to Sanger sequencing, and alignment of the obtained sequences using SnapGene software showed 100% identity with the reference GFP coding sequence (Figure 2C). These results confirmed accurate and stable integration of the target gene into the kiwifruit genome and its functional expression at the cellular level. Quantitative evaluation based on GFP and molecular verification indicated a transformation efficiency of approximately 24% in A. chinensis and 4% in A. valvata (Figure 2D).

To expand the available transformation systems for kiwifruit and assess the compatibility of A. valvata with A. rhizogenes, a tissue culture-based genetic transformation system mediated by strain K599 carrying the 1380-GFP vector was developed. This system complements the A. tumefaciens-mediated approach and provides an efficient route for studying root-specific gene function in A. valvata. In this system, A. rhizogenes K599 exhibited strong infectivity toward A. valvata explants. Following infiltration and co-cultivation, wound sites on A. valvata leaf explants developed visible callus after approximately 4 weeks, with a callus induction rate exceeding 80%. The callus enlarged progressively, displaying a friable texture and light-yellow color typical of active cell proliferation (Figure 3A). Comparative hormone regulation experiments demonstrated a clear influence of exogenous hormones on callus developmental fate. When explants were cultured on selection medium supplemented with the same cytokinin–auxin combinations used in the A. tumefaciens-mediated system, the transgenic callus continued to proliferate without organ differentiation, forming compact tissue masses (Figure 3B, 4–14 weeks). In contrast, when transferred to hormone-free MS medium, the same transgenic callus began to undergo spontaneous organogenic transition, producing numerous hairy roots from the transgenic callus within 4 weeks and by the eighteenth week, transgenic hairy roots developed from the surface of callus, with an induction rate of 68.7% (Figure 3B, 18 weeks). Fluorescence microscopy analysis revealed strong and uniformly distributed GFP signals in both proliferating callus and induced hairy roots, confirming stable expression of the transgene (Figure 3C). The fluorescent tissues showed no signs of chimerism, and GFP expression was particularly evident in root elongation zones, corresponding to regions of active cell division.

A reproducible tissue culture-free transformation system was successfully established for A. valvata using A. rhizogenes K599 under high-pressure propagation conditions. The system exhibited stable performance across independent trials, achieving both high transformation efficiency and reliable gene integration in woody tissues (Figure 4A). After 21 days of cultivation, apparent callus formation was observed at the wound sites of inoculated stem (Figure 4B). Fluorescence detection using confocal laser scanning microscopy revealed that approximately 50% of the induced callus and hairy roots exhibited bright green fluorescence (Figure 4C). The GFP signals were evenly distributed throughout the cytoplasm, confirming stable expression of the reporter gene and the absence of chimerism. The transformation efficiency and GFP verification results demonstrate that this high-pressure propagation-based, tissue culture-free system enables rapid and stable gene introduction into woody stems. Compared with traditional aseptic transformation methods, it markedly shortens the experimental cycle, maintains tissue vitality, and simplifies operation without compromising transformation stability.