A simple and efficient gene functional analysis method for studying the growth and development of peach seedlings

Abstract

Stable genetic transformation of peach [Prunus persica (L.) Batsch] still faces many technical challenges, and existing transient expression methods are limited by tissue type or developmental stage, making it difficult to conduct functional analysis of genes regulating shoot growth. To overcome this dilemma, we developed a three-step method for efficient analysis of gene functions during peach seedling growth and development. This method resulted in transformation frequencies ranging from 48 to 87%, depending on the gene. From transformation of germinating seeds to phenotyping of young saplings took just 1.5 months and can be carried out any time of year. To test the applicability of this method, the function of three tree architecture-related genes, namely PpPDS, PpMAX4, and PpWEEP, and two lateral root-related genes, PpIAA14–1 and −2, were confirmed. Since functional redundancy can challenge gene functional analyses, tests were undertaken with the growth-repressor DELLA, which has three homologous genes, PpDGYLA (DG), PpDELLA1 (D1), and −2 (D2), in peach that are functionally redundant. Silencing using a triple-target vector (TRV2-DG-D1-D2) resulted in transgenic plants taller than those carrying just TRV2-DG or TRV2. Simultaneously silencing the three DELLA genes also attenuated the stature of two dwarf genotypes, ‘FHSXT’ and ‘HSX’, which normally accumulate DELLA proteins. Our study provides a method for the functional analysis of genes in peach and can be used for the study of root, stem, and leaf development. We believe this method can be replicated in other woody plants.

Introduction

Peach is an economically important species in the Rosaceae family. Peach could be used as a model fruit tree in the Rosaceae family due to its relatively short juvenile period, a small genome size (approximately 230 Mb/haploid) and the availability of a high-quality genome sequence. However, functional gene analysis in peach lags behind other species of its family because there are no in vitro regeneration and transformation methods for generating transgenic peach plants. To the best of our knowledge, the generation of transgenic peach has been reported in four published articles, but most of these methods have proven to be impossible to repeat [1, 2]. Several methods of transient expression have been developed to analyse the functions of certain classes of genes, such as those related to volatile compound production, ABA biosynthesis, or anthocyanin production [3–6]. However, these methods have a narrow application scope. For example, the transience of the overexpression precludes observation of phenotypes that regulate organ growth and development. Recently, composite plants of peach were generated by infecting seedlings with Agrobacterium rhizogenes to induce transgenic adventitious hairy roots [7].

Virus-induced gene silencing (VIGS) is a powerful tool for functional gene analysis. VIGS induces RNA-mediated posttranscriptional gene silencing [8] and can be used for reverse genetics studies much like RNA interference (RNAi) and T-DNA insertional mutagenesis. As a tool, VIGS has many advantages over conventional techniques, such as ease of manipulation, high effectiveness and independence from stable transformation. VIGS technology has been successfully applied in many crops, including cotton [9], litchi [10], and pear [11].

The extensive application of VIGS may be attributed to its two outstanding merits: the systemic spread of gene silencing and long-term gene silencing. There are two ways that viruses or virus-induced siRNA spread throughout infected plants: cell-to-cell and/or long-distance through the phloem [12]. These mechanisms also function in VIGS experiments, such that photo-bleaching of young leaves can be seen far away from the infection sites when silencing phytoene desaturase (PDS) gene in tomato [13]. This systemic spread of gene silencing supports the VIGS-induced long-term silencing. Senthil-Kumar and Mysore [14] reported that VIGS was maintained for more than two years and could be transmitted to progeny seedlings in Nicotiana benthamiana and tomato. Bennypaul et al. [15] also reported that mosaic virus (BSMV)-based vector systems result in VIGS inheritance in wheat.

Peach tree architecture greatly impacts the productivity of orchards and has recently been attracting more attention as a breeding target [16]. Many traits related to peach tree architecture, such as branch angle, plant height and pendulous branch growth, were studied, and some key genes for these traits were identified [17–19] However, functional analysis of these key genes was conducted using heterologous expression. For instance, the functions of the WEEP gene, responsible for pendulous branch growth, and of the GID1c gene, responsible for plant height, were confirmed by silencing their homologous genes in plum [18, 19]. The function of PpeTAC1, which regulates branch angle, was analysed using the attac1 mutant of Arabidopsis [17]. Up to now, there are no methods to support functional gene analysis in peach tree architecture.

The tobacco rattle virus (TRV)-derived vector system has been applied in a wide range of species [20]. In this study, we employed a TRV-based vector to infect newly germinated peach seeds and successfully demonstrated the functions of genes related to tree architecture. Using this method, we efficiently confirmed the function of the DELLA, DGYLA, MAX4, and WEEP genes involved in plant height, branch outgrowth, and pendulous branch growth. In addition, PpIAA14, involved in lateral root formation, was cloned into pSAK277 and overexpressed by transformation of newly germinated peach seeds. We think that our reported method will be a powerful tool for analysing gene functions related to the growth and development of peach seedlings.

Results

Silencing of the peach PDS gene using TRV vector

The PDS gene encodes a phytoene desaturase enzyme and is commonly used to evaluate the silencing efficiency of endogenous gene. PDS silencing results in photo-bleaching of leaves by inhibiting carotenoid biosynthesis. To ensure that the TRV vector system could trigger virus-induced gene silencing in peach, the homologous gene of PDS (PpPDS) was identified from the peach genome and targeted for silencing. Three days post-infiltration (DPI), the remaining cotyledon of each peach seed infected with Agrobacterium carrying the TRV-PpPDS exhibited a photo-bleached phenotype (Fig. S1, see online supplementary material). By 15 DPI, 48.33% of TRV-PpPDS infected seedlings showed leaf photobleaching (Fig. 1A, Table 1). The leaves exhibiting mild and severe bleaching were collected for RT-qPCR analysis. The results showed that the completely white leaves had the lowest PpPDS transcript levels, while leaves exhibiting mild bleaching had a higher transcript level than the white leaves but a lower transcript level than control (Fig. 1B). These results demonstrated that the TRV vector system works in peach seedlings.

Figure 1.

Silencing of PpPDS resulted in photo-bleached leaves in peach seedlings. A Seedlings infected with TRV2 (left) and TRV2-PpPDS (middle and right). The photos were taken 15 days post-infiltration. B Expression level of PpPDS in leaves exhibiting mild and severe bleaching. The values represent the average of three biological replicates The error bars indicate means ± standard deviation (SD) (*P < 0.05; ***P < 0.001).

Table 1.

Frequency of seedlings with expected phenotype

Gene	No. of infected seeds	No. of plants with expected phenotype	Frequency / %
PpPDS	36/36/45	15/18/24	48.33 ± 5.69
PpMAX4	36/28/28	30/24/26	87.33 ± 5.13
PpWEEP	45/35/40	24/19/24	55.80 ± 3.60
PpDGYLA	85/20/20	44/10/9	49.00 ± 3.61
PpDGYLA-DELLA1-DELLA2	75/20/20	44/10/10	53.00 ± 5.20
PpIAA14–1	30/30/40	20/19/25	64.17 ± 2.47
PpIAA14–2	35/34/38	24/18/22	60.00 ± 8.19

Frequency (%) = (Number of plants showing expected phenotype/Total number of infected plants)^*100%. Three biological replicates were conducted.

Silencing of PpMAX4 affects branch outgrowth and root system of peach seedlings

Strigolactones play a negative role in branch outgrowth. Downregulation of MAX4, a key gene in strigolactone biosynthesis, increases the number of lateral branches and adventitious roots in poplar [21]. In this study, only one homologous gene of MAX4 (PpMAX4) was identified in the peach genome, with a similar gene structure compared to AtMAX4 in terms of exons and introns distribution (Fig. S2, see online supplementary material). By 14 DPI, the seedlings infected with TRV2 had no lateral branches, while nine TRV2-PpMAX4 seedlings showed growth of almost all lateral buds (Fig. S3, see online supplementary material). The average number of lateral branches per node were calculated 25 DPI with TRV2 or TRV2-PpMAX4. Peach seedlings infected with TRV2-PpMAX4 had an average of 0.8 lateral branches per node, while the uninfected and TRV2 plants had an average of 0.05 lateral branches per node (Fig. 2A and B). The frequency of PpMAX4 gene silencing reached 87.33% (Table 1). Shoot tips were collected 10, 15, 20, and 25 DPI for RT-qPCR analysis. Compared with the controls, the transcript levels of PpMAX4 were significantly downregulated in peach seedlings infected with TRV2-PpMAX4 at all days analysed (Fig. 2C). The PpMAX4 silencing was stable for 25 days as the seedling grew. In addition, PpMAX4-silenced plants (10 days old) had more root fresh weight compared to the TRV2 control (Fig. 2D and E). These results demonstrated that silencing of PpMAX4 increases the lateral number and root fresh weight of peach seedlings.

Figure 2.

Silencing of PpMAX4 increased the number of lateral branches and lateral roots. A Lateral branching in wild-type seedlings (left) and seedlings infected with TRV2 (middle) and TRV2-PpMAX4 (right). The photos were taken 28 days post-infiltration (DPI). B Analyse the number of lateral branches at each node of the peach seedling 28 DPI (by comparing the branch number to the internode number). C Expression levels of PpMAX4 in shoot tips collected 10, 15, 20, and 25 DPI. The values represent the average of three biological replicates (^****P < 0.0001). D Lateral roots of seedlings 10 DPI. E Fresh roots weight of seedlings 10 DPI. B and E Error bars indicate means ± standard deviation (SD) from three biological replicates. ^**** indicates significant differences at P < 0.0001 between TRV2 and TRV2-PpMAX4.

Silencing of PpWEEP results in a pendulous growth of peach seedlings

A pendulous or weeping architecture of peach branches is caused by a loss-of-function mutation in PpWEEP [19], which seems to alter normal gravitropic responses. In this study, silencing of the PpWEEP gene resulted in a pendulous growth phenotype in peach seedlings grown in the dark, but not under light (Fig. 3A). The dark-dependent phenotype may occur because phototropism of shoot tips is strong enough to mitigate the loss of PpWEEP function in the light. The angle of the main stem to the vertical was analysed in seedlings grown in the dark. Peach seedlings infected with TRV2-PpWEEP had an average angle of 24°, while the uninfected and TRV2 plants had an average angle of 3° (Fig. 3B). The frequency of PpWEEP gene silencing was 55.8% (Table 1). At 10, 15, 20, and 25 DPI, shoot tips were collected for RT-qPCR analysis. Compared with the WT and TRV2 controls, PpWEEP transcript levels were significantly downregulated in peach seedlings infected with TRV2-PpWEEP (Fig. 3C). Silencing also confirmed that PpWEEP plays a key role in upright tree architecture, while its loss results in pendulous branch growth in peach.

Figure 3.

Silencing of PpWEEP resulted in a phenotype of pendulous branch growth in peach seedlings. A Upright and pendulous growth of dark-grown seedlings. Uninfected wild-type seeds and seeds inoculated with Agrobacterium harboring the empty TRV2 vector and TRV2-PpWEEP were cultured in pots under dark conditions. The photos were taken at 15DPI. B Measure of the angle of the main stem to the vertical. Error bars indicate means ± standard deviation (SD) from three biological replicates. ^**** indicate significant differences at P < 0.0001 between TRV2 and TRV2-PpWEEP.C Expression level of PpWEEP in shoot tips collected at 10, 15, 20, and 25 days post infiltration. The values represent the average of three biological replicates (^*P < 0.05; ^***P < 0.0001).

Silencing of PpDGYLA, PpDELLA1, and −2 increases the height of peach seedlings

DELLA proteins are well-known repressors of GA signaling. Our previous work suggested that peach has three homologues of DELLA, namely PpDGYLA (DG), PpDELLA1 (D1) and −2 (D 2), which play negative roles in internode elongation of peach [22, 23]. In view of the functional redundancy among the three peach DELLA genes, three fragments from the coding sequences of DG, D1, and D2 were amplified and then simultaneously inserted into a TRV2 vector (Fig. S4, see online supplementary material). Peach seedlings are different in genetic material and show to some extent variation in plant height, which raise the bar to analyse the change of plant height induced by TRV2-DG or TRV2-DG-D1-D2. In this study, there was no significant discrepancy in plant height among these seedlings grown from seeds infected with TRV2, TRV2-DG, and TRV2-DG-D1-D2. To reduce the natural variation in plant height, we designed two experiments: peach seedlings were cultured in the dark or were grown in the light and treated with paclobutrazol (PBZ, a GA biosynthesis inhibitor). Both of these treatments efficiently reduced the variation of plant height in wild-type peach seedlings (Fig. S5, see online supplementary material). The plant height of dark-grown seedlings from seeds inoculated with Agrobacterium carrying the silencing vectors were analysed 10 DPI. In the dark, the TRV2-DG-D1-D2 plants were taller than the TRV2-DG, and the TRV2-DG plants were taller than uninoculated control (Fig. 4A and D). RT-qPCR showed that the three peach DELLA genes were downregulated in the TRV2-DG-D1-D2 plants, while only PpDGYLA was downregulated in the TRV2-DG plants (Fig. 4C). Similar results were observed when peach seedlings were treated with 300 mg/L PBZ (Fig. 4B, E, and F).

Figure 4.

Silencing of DELLA genes increased plant height. Seedlings were cultured in the dark (A) or treated with PBZ (B). The photos were taken 10 DPI (A) and 15DPI (B). D and E Analysis of plant height of dark-grown (D) and PBZ-treated (E) seedlings. C and F Expression levels of PpDG, PpD1, and PpD2 in shoot tips collected 5, 15 days (C, dark-grown), and 25 days (F, PBZ-treated) post-infiltration (^*P < 0.05; ^****P < 0.0001). The values represent the average of three biological replicates. D and E Error bars indicate means ± standard deviation (SD) from three biological replicates. ^*** and ^**** indicate significant difference at P < 0.001 and P < 0.0001 between TRV2 and TRV2-PpDGYLA or TRV2-DG-D1-D2.

Silencing of the DELLA genes disrupts genetically dwarfed peach seedlings

Our previous work demonstrated that a mutation in the gibberellic acid receptor PpGID1c causes a GA-insensitive dwarf phenotype in peach through its increased accumulating of DELLA protein [23]. To confirm this conclusion, peach pits collected from two dwarf genotypes, ‘HongShouXing (HSX)’ and ‘FengHuaShouXingTao (FHSXT)’ were used for VIGS assay. Plant height was analysed 15 DPI. In the two dwarfed backgrounds, TRV2-DG-D1-D2 plants were taller than the TRV2-DG plants, and the TRV2-DG plants were taller than the genotype controls. The change was most drastic in ‘FHSXT’, with the plant height of TRV2-DG-D1-D2 plants extremely higher than the TRV2-DG and TRV2 control plants (Fig. 5A and B). RT-qPCR showed that the three peach DELLA genes were downregulated in dwarf peach seedlings infected with TRV2-DG-D1-D2, while only PpDGYLA was downregulated in the TRV2-DG plants (Fig. 5C and D). Together, these results demonstrated that the TRV vector system could silence DELLA genes, resulting in increased plant height of peach seedlings.

Figure 5.

The dwarf phenotypes of ‘FHSXT’ and ‘HSX’ were attenuated by silencing of peach DELLA genes. A The plant height of seedlings. Photos were taken 15 DPI. B Analysis of plant height of the ‘FHSXT’ and ‘HSX’ dwarf genotypes. Error bars indicate means ± standard deviation (SD) from three biological replicates. ^*** and ^**** indicate significant differences at P < 0.001 and P < 0.0001 between TRV2 and TRV2-PpDGYLA or TRV2-DG-D1-D2. Expression levels of PpDGYLA, PpDELLA1, and PpDELLA2 in shoot tips from ‘HSX’ (C) and ‘FHSXT’ (D) collected 25 DPI. The values represent the average of three biological replicates (^**P < 0.01; ^***P < 0.001; ^****P < 0.0001).

Overexpression of PpIAA14–1 and PpIAA14–2 mutants affected lateral root development in peach

Previous work demonstrated that a gain-of-function mutation in the IAA14 gene stops lateral root formation in Arabidopsis [24]. In this study, two homologous genes, PpIAA14–1 and PpIAA14–2, were identified from the peach genome. According to Fukaki et al. [24], a single nucleotide substitution was introduced into PpIAA14–1 and PpIAA14–2, which converted the proline at position 102 (PpIAA14–1) or 108 (PpIAA14–2) into a serine (Fig. S6, see online supplementary material). To boost the frequency of transient expression, removal of one cotyledon was tested using the common Agrobacterium strain GV3101. Our result showed that removal of one cotyledon significantly improved transformation efficiency (Fig. S7, see online supplementary material). Then, 35S-PpIAA14–1^P102S-GFP and 35S-PpIAA14–2^P108S-GFP were delivered into the germinated seeds. At four DPI, the lateral root phenotypes were analysed. Compared to the GFP control, the number of lateral roots significantly decreased in peach seedlings expressing PpIAA14–1^P102S-GFP or PpIAA14–2^P108S-GFP (Fig. 6A). Among these 200 seedlings infected with PpIAA14–1^P102S-GFP and PpIAA14–2^P108S-GFP, there are 160 seedlings having a lateral root number ranging from 0 to 10. One hundred seedlings were infected by Agrobacterium tumefaciens containing 35S-GFP and 88 seedlings had 21–56 lateral roots (Fig. 6B). Analysis of the transcript levels of PpIAA14–1^P102S-GFP and PpIAA14–2^P108S-GFP using RT-qPCR showed that GFP was overexpressed in roots expressing PpIAA14–1^P102S-GFP and PpIAA14–2^P108S-GFP, but the transcript levels of PpIAA14–1 or PpIAA14–2 were downregulated (Fig. 6C). These results demonstrated that overexpression of PpIAA14–1^P102S or PpIAA14–2^P108S could significantly reduce lateral root formation in peach seedlings.

Figure 6.

Overexpressing PpIAA14–1^P102S and PpIAA14–2^P108S inhibited the number of lateral roots in peach. A Whole seedlings removed from soil 14 days post-infection. B Analysis of lateral root numbers at 14 DPI. In these seedlings, the minimum number of lateral roots was 0, while the maximum was 56. The frequency for each range indicated by color are shown. C Transcript levels of IAA14–1 and IAA14–2 were analysed in the main root 10 days post-infection. The values represent the average of three biological replicates. (^****P < 0.0001) (EV represent pSAK277-GFP).

Discussion

Peach is one economically important species that has long proven recalcitrant to in vitro regeneration and genetic transformation [2, 7]. Existing transient expression methods have a narrow range of application. Therefore, the development of a method to support functional gene analysis in peach is of paramount importance. The TRV vector system is widely used to silence endogenous genes in many plant species. The TRV vector system has been used to analyse genes regulating fruit firmness in peach through transient fruit transformation [25]. In this study, our results showed that six genes (PpPDS, PpMAX4, PpDGYLA, PpDELLA1, −2, and PpWEEP) were efficiently silenced for a relatively long time (25 days in the TRV2-PpMAX4 assay) in peach seedlings, with effects in multiple vegetative tissues as the seedlings developed. Compared to existing transient expression methods [3–6], long-term gene silencing could be achieved with this method, which is essential for analysing genes involved in shoot development. Xu et al. [7] and Wu et al. [26] reported a method by which transgenic adventitious hairy roots could be obtained. Compared with the method we developed, this needs a procedure of tissue culture and restricted to the application on root. Together, our study provides a simple and effective tool for functional gene analysis in peach.

Gene silencing efficiency is a crucial indicator that determines whether the TRV vector system is suitable for its use in peach. In this study, we analysed genes regulating tree architecture-related traits (plant height, branch number, and lateral root number). These traits generally belong to quantitative character. The expected change induced by silencing PpMAX4, PpWEEP, PpDGYLA, PpDELLA1, and PpDELLA2 is not a clear qualitative change. To enhance the reliability of the results, several ways were used to reduce variation of the observed traits. For peach seedlings within 30 d age, almost no branches outgrow. Therefore, lateral branches in PpMAX4 silencing assay were calculated 25 DPI. To analyse the gene function of PpDGYLA, PpDELLA1, and −2, we reduced the range of plant height variation by using the seedlings cultured in dark conditions or treated with PBZ. In addition, to overcome the challenge posed by functional redundancy, the TRV2-DG-D1-D2 vector was constructed and the three DELLA genes were silenced simultaneously. This result is similar to the use of the TRV vector system to efficiently knockdown 12 members of the highly homologous E2 gene family [27]. By these methods, assays obtained a relatively high gene silencing frequency in peach, ranging from 48.33% to 87.33%.

TRV-induced gene silencing is strongly affected by many factors [28]. Our results showed that genetic background is another factor affecting gene silencing frequency of the TRV vector system. We collected peach pits from two dwarf peach genotypes ‘HSX’ and ‘FHSXT’. In the TRV-DG and TRV2-DG-D1-D2 assay, we found that the frequency of gene silencing was significantly higher in ‘FHSXT’ than in ‘HSX’. Peach is both self-pollinated and cross-pollinated. This means that seeds from the same plant may carry genetic differences, leading to greater variation in growth and development of seedlings, which can threaten functional gene analysis efforts. However, we also believe that using peach seedlings increases the likelihood of identifying genes that play key roles in the development of specific traits. The influence of a single gene on plant phenotype is complex and changes in different genetic backgrounds. The phenotypes of a mutant allele are not a simple result of a single gene mutation, but are instead due to interactions of the allele with other genes and the environment [29]. Therefore, a gene with a powerful function in most genetic backgrounds is most likely to play a key role in a specific trait.

In conclusion, we provide a simple pipeline schematic for our method (Fig. 7). In Step 1, part of the seed coat and one cotyledon are removed from an imbibed seed. In Step 2, the germinated seeds with one cotyledon are inoculated with Agrobacterium using vacuum infiltration. In Step 3, the infected seedlings were transferred into soil for phenotyping. This transformation method had three distinct advantages, especially for TRV-induced gene silencing in peach seedlings. Firstly, the method is highly efficient. Our results showed a greater than 48.33% gene silencing frequency that could be accomplished within 1.5 months and carried out at any time of year. Secondly, gene functions related to growth and development of the shoot, root, and leaf can be analysed using this method. Thirdly, this method can likely be applied to other woody plants, such as plum and almond. However, we do wish to pursue greater transformation frequency and to further study the duration, stability, and heritability of the gene silencing achieved with this method.

Figure 7.

A simple procedure for transformation of peach seedlings.

Materials and methods

Plant material and growth conditions

Peach pits were collected from the wild peach Maotao and two dwarf genotypes, ‘HSX’ and ‘FHSXT’, which are maintained at the Fruit Tree Germplasm Repository of Henan Agricultural University (Henan Province, China). ‘HSX’ and ‘FHSXT’ are two dwarfing varieties in peach. Their dwarfing traits are caused by the accumulation of DELLA proteins. These peach pits were stored at 4°C for more than 60 days before seeds were taken from the pit and immersed in distilled water for 1 day. A portion of the seed coat near the radicle was removed, and then the seeds were placed on wet filter paper for 7 days at 24 ± 1°C in the dark. During germination, the radicle would elongate, and germinated seeds with a radicle of 1 ± 0.2 cm in length were selected for transformation.

Vector construction

The tobacco rattle virus (TRV)-derived vectors pTRV1 and pTRV2 used in gene silencing assays were described in [30]. Portions of the coding sequences of PpPDS (1–167 bp), PpMAX4 (1236–1500 bp), PpWEEP (185–387 bp), and PpDGYLA (150–300 bp) were amplified and inserted into the vector pTRV2. Three fragments cloned from PpDGYLA (150–300 bp), PpDELLA1 (150–300 bp), and PpDELLA2 (150–300 bp) were amplified and connected into a fragment, named DG-D1-D2, which was then inserted into the vector pTRV2. The primers used to amplify the fragments are listed in Table S1 (see online supplementary material) along with each Gene ID.

The binary vector pSAK277, containing a 35S-GFP cassette, was used for gene overexpression assays. The coding sequences of PpIAA14–1 and PpIAA14–2 were cloned. According to a previous study [24], overlapping PCR was used to convert the proline at amino acid (aa) position 102 in PpIAA14–1 and at position 108 in PpIAA14–2 to serine. The primers used in the overlapping PCR assay are listed in Table S1 (see online supplementary material). After mutation, the pSAK277-PpIAA14–1^P102S-GFP and pSAK277-PpIAA14–2^P108S-GFP vectors were constructed. cDNA from the peach cultivar ‘QiuMiHong’ was used as the template for the PCRs for subsequent cloning.

Vacuum infiltration

The prepared vectors were delivered into A. tumefaciens GV3101, which then were cultured at 28°C on LB plates containing 50 mg/L kanamycin or spectinomycin. Positive clones were selected and cultured in 400 μL LB liquid medium (containing 50 mg/L kanamycin or spectinomycin) overnight. Then, 100 μL of the A. tumefaciens cultures were added to 100 mL LB medium (containing 50 mg/L kanamycin or spectinomycin) and incubated at 28°C with shaking at 180 rpm for 12 h.

The A. tumefaciens culture was centrifuged at 5000 rpm to collect agrobacterium cells, which were then suspended in MES suspension buffer (100 μM acetosyringone, 10 mM MgCl₂, 10 mM MES; pH 5.6). The concentration was adjusted to OD_600nm = 0.7–0.9. Then, the A. tumefaciens suspense was incubated at room temperature for 3 hours.

After germination on filter paper for 7 days, seeds with a radicle of 1 ± 0.2 cm in length were selected and then the rest of the seed coat and one cotyledon were removed from the germinated seeds. Seeds with greenish plantule were then immersed in the prepared A. tumefaciens suspension and placed under vacuum (−0.8 Mpa) for 20 min. After vacuum infiltration, seeds were rinsed in sterile water, cultured on filter paper for two more additional days, then transferred into soil. In TRV2-PpPDS, TRV2-PpMAX4, and pSAK277-PpIAA14–1^P102S/PpIAA14–2^P108S assays, seedlings were cultured at room temperature and normal light conditions. In TRV2-PpWEEP, seedlings were cultured at room temperature and dark conditions. In TRV2-PpDG and TRV2-DG-D1-D2 assay, seedlings were cultured at light or dark conditions.

RT-qPCR analysis

In TRV2-PpMAX4, TRV2-PpWEEP, TRV2-PpDG, and TRV2-DG-D1-D2 assay, shoot tips were collected to analyse transcript level. In TRV2-PpPDS assay, to confirm the relationship between phenotype and gene expression level, whole leaves exhibiting mild and severe bleaching were collected, respectively. In pSAK277-PpIAA14–1^P102S/PpIAA14–2^P108S assay, root tip was collected to analyse transcript level. Samples of each technical replicates were prepared using more than three plants. Three biological replicates were conducted in each assay of RT-qPCR analysis. RNA was prepared and first-strand cDNA synthesis was carried out using the Total RNA Rapid Extraction Kit (Huayueyang Biotechnology, Beijing, China) and HiScript III RT SuperMix for RT-qPCR (+gDNA wiper) (Vazyme, Nanjing, China). A 20-μL reaction for quantitative RT-qPCR contained 1 × ROX reference dye, 10 μL of 2 × SYBR premix Ex Taq II (TaKaRa), 0.4 μM of each primer, and 100 ng of template cDNA. The peach EF2 gene (Prupe.4G138900) was used as the reference for normalization [31]. The primers used for RT-qPCR are listed in Table S1 (see online supplementary material).

Scoring the phenotypes of TRV-induced gene silencing and PpIAA14 overexpression

The plants following TRV-induced gene silencing were scored according to wild-type and expected phenotypes. The frequency of gene silencing was determined according to Senthil-Kumar and Mysore [14]. The frequency of gene silencing was calculated by comparing the number of plants with an expected phenotype to the total number of plants infected with TRV constructs. The phenotype and frequency of PpPDS gene silencing were determined according to Senthil-Kumar and Mysore [14]. The expected phenotypes of PpMAX4, PpWEEP, and PpDELLA were different, as described below. For PpMAX4, the frequency of gene silencing was calculated 25 days after infiltration. The expected phenotype of PpMAX4-silencing was a seedling with a greater number of lateral branches than the average branch number of WT and TRV2 (0.05 branches/a node). For PpWEEP, the frequency of gene silencing was calculated 15 days after infiltration. The expected phenotype of PpWEEP-silencing was seedlings with an angle larger than the average angle of WT and TRV2 (3°). For PpDELLA, the frequency of gene silencing was calculated 15 days after infiltration. The expected phenotype of silencing the three PpDELLA genes was seedlings with a plant height higher than the average plant height of WT and TRV2 (seedling grown in the dark, 5.7 cm; in the light, 2.1 cm; and seedlings of dwarf peach genotype, 2.4 cm).

The expected phenotype of the plants inoculated with pSAK277-PpIAA14–1^P102S-GFP and pSAK277-PpIAA14–2^P108S-GFP was seedlings with a lower average number of lateral roots compared with WT and GFP control seedlings (32.75 and 31.92).

Statistical analysis

The Statistical Product and Service Solutions (SPSS) software (IBM Co., Armonk, NY, USA) was used to conduct the statistical analyses. All experimental data were tested by Student’s t-test. Prism 8.0.1 software (GraphPad, San Diego, CA, USA) was used to generate figures.

Author contributions

L.Z., Y.S., and X.H. performed the experimental work and generated the data; J.F. and J.C. conceived and designed the experiments; L.G., X.Z., and B.T. analysed the data; W.W., B.T., Z.L., and X.Y. prepared materials; H.Z., X.W., and X.L. provided resources. J.C. and J.F. wrote the paper. All authors read and approved the final manuscript.

Data availability statement

All relevant data are included in the article and its supporting materials.

Conflict of interest

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

Supplementary informations

Supplementary data is available at Horticulture Research online.