Beginning in September 2019, samples of the fruit, roots, leaves, flowers, and stems of a 6-year-old “Chun Xue” peach tree were obtained from the Horticulture Experimental Station of Shandong Agricultural University. Tissue samples were stored at −80°C before RNA extraction.

The transgenic tomato variety was “Micro Tom” purchased from Nanjing Fengtai Horticultural Seed (Nanjing, China) Co. Tomato seeds were sterilized, soaked in warm water, germinated on Murashige and Skoog (MS) medium for 3 days, and then used for Agrobacterium infection. The seeds used in the germination rate experiment were treated in the same way. Annual hickory peach seedlings were provided by the National Apple Engineering Center Laboratory. A 3:1:1 mixture of grass charcoal, vermiculite, and perlite was used as a substrate for each tomato material and annual peach seedling planted in a 12-cm-wide square plastic culture bowl. Tomato and annual peach seedlings were cultivated in a growth chamber in which the temperature, relative humidity, photoperiod, and photosynthetic photon flux density were 26°C/20°C, 65%, 14 h/10 h (day/night), 300 μmol m^–2 s^–1, respectively. The leaves of the annual peach seedlings were sampled in separate periods after spraying with 100 μM ABA, at 4°C, root soaking with 10% PEG, and watering with 150 mM NaCl. Thirty milliliters of 150 mM NaCl solution was applied to the transgenic tomato substrate every afternoon and the wild type (WT) was treated with water for 12 days. The leaves of transgenic tomatoes and the WT were then taken for physiological data analysis.

RNA from the samples was extracted using the RNA Pure Plant Kit (DNase I) (CW0559) and first strand cDNA was obtained using the SuperRT cDNA Synthesis Kit (CW0741) for qRT-PCR. An Ultra SYBR Mixture Kit (CW0957) was used for qRT-PCR analysis, and the primers are listed in Supplementary Table 1. PpActin (Prupe.6G163400) is an internal reference gene of peach (Zhao et al., 2020), and SlAction (Solyc11g005330) is the internal reference gene of tomato (Hu et al., 2021). These kits were purchased from Kangwei Century Biotechnology (Jiangsu) Co., and primer synthesis was performed by Biotech Bioengineering (Shanghai) Co.

Primers (5′- ATGGCATTGTTTTCTATGGCCA-3′; 5′- CACAGATTCAGTTTCATACAGCATCG-3′) were designed by searching the NCBI database using the cDNA sequence Prupe.4G030400 (PpTPS1). The amplification reaction was performed using the cDNA of “Chun Xue” peach as the template, and the amplified products were sequenced and compared with the NCBI database. PpTPS1 was cloned into the plant expression vector PBI121-GFP using SmaI single enzyme digestion (Supplementary Figure 1). The recombinant plasmid was transformed into GV3101 Agrobacterium using the freeze-thaw method, and the transformed Agrobacterium was then used to infect the “Micro Tom” tomato explants (Jian et al., 2019). After induction of germination, rooting, and identification, overexpressed transgenic tomatoes were obtained.

Tomato leaves were taken after treatment to measure the relative water content, relative electrical measurements, and MDA contents, and the analysis was performed as described previously by Roger (2001). The proline, soluble protein, and soluble sugar contents were measured in the treated tomato leaves, according to Zhuang et al. (2019).

NCBI was used to design specific primers for PpABI5 (5′-GAGAGTGGTGAATGGGGGTG-3′; 5′-ACCTTTTCCACCGGTCCATC-3′), ligate the amplified PpABI5 fragment with the TRV2 plasmid using restriction enzymes EcoRI and BamHI, and then transfer it into the Agrobacterium GV3101. The PpABI5-TRV2 and TRV2 Agrobacterium liquid were mixed with the Agrobacterium liquid of TRV1 in a 1:1 ratio, and they were injected into the green ripe peach fruit, followed by vacuum treatment for 10 min and dark treatment for 1 day. The sample was then left under normal lighting condition for 3 days. The pulp was used as a sample for qRT-PCR analysis (Zhang et al., 2015).

The peach yeast two-hybrid library was established. Then PpTPS1-pGBKT7 vector was constructed and transferred it into Y2H and cultured on -Trp, -Trp/X-α-gal medium at 30°C to verify the autonomous activation activity. Afterward, the library was initially screened for interacting genes, and the screened PpABI5 (Prupe.7G112200.1) was inserted into the pGADT7 vector (Zhang et al., 2021). The Matchmaker Yeast Two-Hybrid System User Manual was followed to perform mutual verification again.

PpABI5 was cloned into a pGEX-6P-1 vector containing a GST tag, and PpTPS1 was cloned into a pET32a vector containing a His-tag. The vectors were transformed into BL21 DE3 Escherichia coli, and 1 mM isopropyl β-D-thiogalactoside (IPTG) was used to induce the expression of the fusion protein in the bacteria on a shaker at 37°C (Zhao et al., 2020). After the induced PpTPS1-His fusion protein and PpABI5-GST protein were purified by dialysis, the protein was mixed according to the combination of PpTPS-His/GST and PpTPS-His/PpABI5-GST, purified by a GST purification column, and then subjected to Western blot analysis. The GST/His-tagged Protein Purification Kit (P2262/P2229), His/GST Tag Mouse Monoclonal Antibody (AF5060/AF5063), and goat anti-mouse IgG (A0216) used in the experiment were purchased from Shanghai Biyuntian Biotechnology Co.

The PpABI5-pSPYNE and PpTPS1-pSPYCE recombinant plasmids were constructed and transformed into Agrobacterium tumefaciens GV3101. The two bacterial solutions were mixed at a 1:1 ratio to transfect onion epidermal cells for 30 min, and PpABI5-pSPYNE + pSPYCE and PpTPS1-pSPYCE + pSPYNE, 1:1 mixed bacterial liquid, were used as the two groups of controls (Chen et al., 2018). After 48 h of culture in MS medium, the cells were observed with a laser-scanning confocal microscope (LSM880).

PpTPS1^100–400 (primers: 5′-AAAGGCAGAGCCATGCTTAG-3′; 5′-TAAAGAGACCCTCGCGGGCT-3′) was amplified using 5′−100 to −400 bp genomic DNA of the 2-kb upstream promoter of PpTPS1 as a template. FastDigest BstBI was used to linearize the plasmid, transform it into yeast one-hybrid (Y1H) Gold, and detect the lowest concentration of AbA that inhibits the growth of bait yeast strains on Ura medium at 30°C. Next, the lowest concentration of AbA was used for the Y1H screening library (Wang et al., 2020). The screening results were sequenced and compared with the sequences in NCBI.

PpTCP1, PpTCP13, and PpTCP15 were inserted into the pGreenII0029 62-SK vector, and the promoter fragment PpTPS1^100–400 was inserted into the pGreenII0800-Luc vector (Walter et al., 2004). The vectors were then transformed into A. tumefaciens strain GV3101 pSoup. A mixture of four Agrobacterium strains was injected on the back of tobacco leaves: PpTCP1/PpTCP13/PpTCP15-62SK + PpTPS1-Luc, PpTCP1/PpTCP13/PpTCP15-62SK + Luc, and 62SK + PpTPS1-Luc. After 2–3 days, 1 mM D-luciferin and sodium salt were applied to the backs of the tobacco leaves, and an in vivo imaging system was used to detect fluorescence.

Biological triplicates were performed for each sample, and the data are expressed as the mean ± standard error (SE). Two-way ANOVA was performed using SPSS for Windows 24 when applicable. The qRT-PCR data were analyzed by Duncan’s test. A significance level of P < 0.05 was used.

The leaves, flowers, stems, roots, fruit, and seeds of 6-year-old “Chun Xue” peach trees were used as samples to perform qRT-PCR analysis to study the expression patterns of PpTPS1 in different tissues of peaches. The results showed PpTPS1 had the highest expression in fruit, followed by leaves, and the lowest expression was observed in flowers, indicating that PpTPS1 may be widely involved in a variety of physiological activities in plants (Figure 1A).

Most TPSs genes contain abundant promoter cis-elements related to abiotic stress, but no further research has been conducted. To initially explore whether PpTPS1 is induced by abiotic stress, we analyzed its expression pattern under exogenous ABA, low-temperature, salt, and drought treatments. The results showed that the target gene was significantly induced under the treatment of 100 μM ABA, 4°C, 10% PEG, and 150 mM NaCl. Interestingly, the expression pattern of PpTPS1 under NaCl and PEG treatment was similar to that under ABA treatment, and its expression level increased rapidly at first, reached a maximum at 3 h and then decreased (Figure 1B). This result indicated that PpTPS1 could play a role in most abiotic stresses.

The amplification products of “Chun Xue” were sequenced and compared with the NCBI, and the sequences were consistent with PpTPS1 (Genbank ID: OL855977). The results of RT-PCR (Supplementary Figure 2) and qRT-PCR analyses, showed that PpTPS1 was successfully overexpressed in tomatoes, and three overexpressing tomato lines were obtained (Figure 2A). We treated transgenic and WT “Micro TOM” tomatoes that were grown for 5 weeks and grew consistently with 150 mmol⋅L^–1 NaCl solution for 12 days, and the control group was treated with water. As shown in Figure 2B, the leaves of WT tomatoes treated with NaCl showed curling, yellowing, and necrosis, while the transgenic tomatoes grew better, and the higher the expression of PpTPS1 was, the better the plants grew; the control group exhibited no significant difference. We measured the relative water content of transgenic and WT tomatoes after salt treatment, and the results demonstrated that the relative water content of transgenic tomatoes was significantly higher than that of WT tomatoes. The relative electrical and malondialdehyde (MDA) content measurements showed that WT tomatoes suffered significantly more damage than transgenic tomatoes, and the trend was consistent with the expression level (Figure 2C). This result suggests that salt treatment may cause a relatively large disruption of the cellular structure of WT tomatoes compared with transgenic tomatoes, disrupting the ionic balance, but the exact mechanism needs further experimental verification.

Content of osmotic adjustment substances can be used as an evaluation index to measure salt tolerance. Numerous studies have shown that plants can increase their resistance to abiotic stress by increasing the content of osmotic adjustment substances. To preliminarily explore the mechanism of PpTPS1 transgenic tomatoes in response to salt stress, we measured the content of its osmotic adjustment substances. The results showed that the proline, soluble protein, and soluble sugar contents of transgenic tomatoes after NaCl treatment were significantly higher than those of the WT tomatoes, especially OE4 (Figure 2D). Therefore, we speculate that the increased content of osmotic regulators may be one reason transgenic tomatoes have better salt resistance.

To explore whether PpTPS1 responds to ABA, we analyzed the seed germination rate of PpTPS1-overexpressing tomatoes under ABA treatment. The results showed that there was no significant difference between the seed germination rate of WT and transgenic tomatoes on a normal MS medium. On 1 μM ABA + MS medium, the seed germination rate of transgenic tomatoes was significantly lower than that of the WT (Figure 2E), and it was opposite to the expression trend. This result shows that overexpression of PpTPS1 can increase the sensitivity of plants to ABA.

After testing for autonomous activation activity, we found PpTPS1 had no autonomous activation activity (Figure 3A). We used PpTPS1-BD as the “bait” to screen its interacting proteins in the library. From the preliminary screening results (Supplementary Table 2), we found that PpABI5 may interact with PpTPS1, and previous studies of ABI5 have shown that it is involved in salt stress. We cloned the full-length PpABI5, constructed the PpABI5-AD vector, and cotransformed Y2H competent yeast with PpTPS1-BD. The results showed that PpTPS1 can interact with PpABI5 (Figure 3B). To further verify the interaction between PpTPS1 and PpABI5, we conducted in vitro protein interaction verification (pull-down) and bimolecular fluorescence complementation (BiFC) analysis (Figures 3C,D), and the results supported the above conclusions.

To verify whether PpABI5 responded to salt stress, we analyzed the expression pattern of PpABI5 after exogenous ABA and NaCl treatment. The results showed that PpABI5 can be induced by salt stress, which is consistent with our hypothesis and the results of previous studies (Figure 3E).

Because of the interaction between PpTPS1 and PpABI5, we were curious about the expression of SlABI5 in PpTPS1-overexpressing tomatoes. The qRT-PCR results showed that the expression of SlABI5 in transgenic tomatoes was significantly lower than that in WT tomatoes (Figure 4A). Then we constructed the PpABI5-TRV2 vector and successfully silenced PpABI5 in peach fruit (Figure 4B). The qRT-PCR results showed that the expression of PpTPS1 in peach fruit was significantly increased after PpABI5 was silenced compared with the control (Figure 4C). This result indicated that there is an antagonistic relationship between PpTPS1 and PpABI5.

To explore the regulatory relationship of PpTPS1 upstream genes, we analyzed the 2-kb promoter sequence upstream of ATG, the translation start site of the PpTPS1 gene, and found that there were multiple TCP-binding motifs, G (T/C) GGNCCCAC motifs, in the 118–378 bp fragment of the promoter (Cubas, 2002; Kosugi and Ohashi, 2002; Li et al., 2005). Previous studies revealed that signal transmission between TCP transcription factors and ABA played an important role in salt stress (Tatematsu et al., 2008). The experiments in the present study showed that the use of 200 ng mL^–1 AbA can inhibit the growth of the PpTPS1^100–400-pAbAi “bait” yeast strain (Figure 5A). A Y1H screen library was performed using PpTPS1^100–400-pAbAi as bait, and the initial screening results contained three TCP family genes: PpTCP1 (Prupe.3G240200.1), PpTCP13 (Prupe.3G252600.1), and PpTCP15 (Prupe.5G191200.1) (Supplementary Table 3). Further verification revealed that the PpTPS1^100–400-pAbAi “bait” yeast transferred into PpTCP1-AD, PpTCP13-AD, and PpTCP15-AD plasmids grew normally on SD/-Ura-Leu/300 ng⋅mL^–1 AbA, indicating that PpTCP1, PpTCP13, and PpTCP15 could bind to the PpTPS1 promoter (Figure 5B).

To explore how PpTCP1, PpTCP13, and PpTC15 regulate the expression of PpTPS1, we performed a dual-luciferase assay. The results showed that the fluorescence intensity of the regions where PpTPS1^100–400-Luc coexpressed with PpTCP1-62SK, PpTCP13-62SK, and PpTCP15-62SK increased significantly (Figure 5C), indicating that PpTCP1, PpTCP13, and PpTCP15 could promote the expression of PpTPS1.

We analyzed the 1.5 kb promoter region upstream of PpTCP1, PpTCP13, and PpTCP15, and the results showed that these three genes all contained the cis-acting element ABRE, which is involved in the ABA response (Table 1). We then performed qRT-PCR analysis on peach seedlings treated with ABA. As shown in Figure 6, ABA significantly induced the expression of PpTCP1, PpTCP13, and PpTCP15, but the expression patterns of the three were different. We speculated that these genes could play different roles in the process of ABA signaling.