The seed potatoes were purchased from Gansu Ailan Potato Seed Industry Co. Ltd. The potato plantlets “Solanum tuberosum L. cv. Atlantic” and the tobacco plant (N. benthamiana L.) were provided by the Molecular Biology Laboratory of College of Life Science and Technology in Gansu Agricultural University, where the experiment was carried out from April to October 2019.

The potato plantlets were propagated by subculturing using single-node cuttings on Murashige and Skoog (MS) basal medium containing 3% sucrose and 0.45% agar and grown in an illuminating incubator providing a light: dark regimen of 16: 8 h and a light intensity of 20000lx at 25 ± 2°C. Micro-tubers were screened and multiplicated on MS media containing 8% sucrose and 0.45% agar under dark conditions at 25 ± 2°C. Tobacco plants were cultured in an environmentally controlled growth chamber with a 16 h light/8 h dark cycle at 25°C. The relative humidity was maintained at 60–70% and was used for subcellular location analysis.

The potato tubers used for wound healing were washed and stored at 5°C for further analysis. The tubers of uniform size and without injury were wounded and healed after BTH treatment according to the method described by Jiang et al. (2019). Healing tissues samples (2 mm depth) were collected from the wounded surface after healing for 0, 1, 3, 5, 7, and 14 days. All the samples were frozen in liquid nitrogen and stored at −80°C for subsequent experiments.

The distribution of cellular Ca²⁺ in healing tissues was based on the method described by De Freitas et al. (2012) with some modification. The tissue blocks of 1 mm³ cut from the healed region were incubated in 2.5% glutaraldehyde and sucked to vacuum. Then, the tissues were rinsed using 0.1 M sodium cacodylate trihydrate buffer containing 2% potassium antimonite five times, each time for 4 h at 4°C, the tissues were post-fixed in 1% osmic acid for 2 h, and washed 5 min again by sodium cacodylate trihydrate buffer. Then, the tissues were dehydrated in ethanol with various concentration gradients and embedded in epoxy resin. 1–2 μm sections were prepared and dyed with uranium acetate and lead citrate. For observation of Ca²⁺, the transmission electron microscope (TEM) (Leica SP8, Germany) was used.

The presence of Ca²⁺ in the cytosolic was determined via staining with Fluo-3-acetoxymethyl ester (Fluo-3-AM), according to the protocol described by Markulin et al. (2019). The sections of healing tissues (0.3–0.5 mm) at 4 and 72 h were incubated in 10 μM Fluo 3-AM for 24 h at 4°C, washed twice with phosphate-buffered saline and examined with a fluorescence microscope (BX61 LSM 800, Olympus, Japan) using excitation filter at 488 nm and emission filter at 515–565 nm. Fluorescent pictures of cytosolic Ca²⁺ were obtained under 10× magnification.

Total RNA was isolated from the transgenic tubers using a simple Total RNA Kit (Cat. No. DP419, TIANGEN208 Biotech, China). The RNA integrity was determined using 1% agarose gel, the concentration and purity were established at an absorbance of 260 nm and a 260/280 ratio, respectively. First-strand cDNA synthesis was reverse transcribed using the TIAN script RT Kit211 (Cat. No. KR116, TIANGEN Biotech, China) according to the manufacturer’s instructions.

The obtained cDNA was used in an expression assay of StCDPKs by real-time quantitative PCR (qRT-PCR) on the Light Cycler 96 SW 1.1 instrument. The cDNA concentration of the transcript was measured and diluted to 100 ng/μL as a template for qRT-PCR. The qRT-PCR reaction consisted of 1 μL cDNA template (ca.0.1 μg cDNA), 10 μL 2× Super Real PreMix Plus (with SYBR Green), 0.4 μL 50× ROX Reference Dye, 0.6 μL primers, and 7.4 μL RNase-Free ddH₂O. The elongation factor 1-alpha 1 [ef1a, (NM_001273486.1)] was used as an internal control gene. qRT-PCR was performed with the following conditions: 94°C for 900 s, with 1 cycle; 95°C for 30 s, 55°C for 20 s with 40 cycles, and finally an extension step for 30 s at 72°C. The relative expressional levels of each gene were calculated using the 2^–ΔΔC(t) method compared to that of 0 h (Livak and Schmittgen, 2001). Primer sequences used for RT-qPCR are shown in Table 1.

The full-length cDNA sequence of StCDPK14 was obtained from the National Center for Biotechnology Information¹ with StCDPK14 as a query (XM_006342017.2). The number of EF-hand Ca²⁺ binding structures was predicted using the simple modular architecture research tool (SMART) program.² The prediction of myristoylation and palmitoylation sites was performed by myristoylator³ and CSS-Palm3.0,⁴ respectively. The conserved domain analysis was performed using SMART (see text footnote 2). Prediction of interacting proteins with StCDPK14 was constructed by the search tool for the retrieval of interacting genes/proteins (STRING) software.⁵

The coding sequences of StCDPK14 gene without a stop codon were amplified by PCR and subcloned into the pEGFP vector, in frame with the GFP sequence, resulting in StCDPK14-GFP vectors under the control of the CaMV 35S promoter. The primers used are listed in Table 2. The GFP fusion construct was mixed with the membrane and nucleus marker and co-transformed into N. benthamiana leaves by Agrobacterium tumefaciens infiltration. The leaf discs near the injection site were cut 48 h after infiltration and the lower epidermis was selected to observe signals of GFP. Fluorescence signals were visualized at 488 nm and detected under a confocal laser scanning microscope (Leica SP8, Germany).

The StCDPK14 coding sequence was amplified using the primers listed in Table 2, the product was then cloned into the pHellsgate8 vector using gateway cloning technology and resulted in an interference-expression (pHellsgate8-StCDPK14) construct that was transformed into A. tumefaciens LBA4404 according to the method described by Zhang et al. (2018). The potato tubers obtained from sub-culture were removed as buds and cut into slices of 1–2 mm, and were infected by Agrobacterium containing pHellsgate8-StCDPK14 and the empty pHellsgate8 plasmid. The infected slices were placed on MS solid media at 28°C in the dark for 48 h and after that, they were transferred into differentiation media for culture in a light chamber (16 h light/8 h dark with a light intensity of 20,000lx) at 25°C. When the new buds were generated from the center callus of the potato slice, they were transferred into rooting MS medium supplemented with 75 mg/L kanamycin and 200 mg/L carbenicillin for screening kanamycin-resistant transformed plants. After 1–2 months, the regenerated plantlets were acclimatized and grown in flasks under the condition of a photoperiod of 16/8 h light/dark at 25°C.

The genomic DNA of the transgenic plants was isolated using a plant genomic DNA isolation kit (Cat. No. DP305, TIANGEN Biotech, China). The kanamycin-resistant potato plants were screened using the neomycin phosphate (NTP II) gene with a pair of primers to detect positive transformations of the StCDPK14 transgenic lines. The DNA from wild-type potato plants was used as a negative control and the pHellsgate8-StCDPK14 as the experimental set. A PCR was performed as described in the above section. The positive and rooting plants were chosen for further culture and transgenic tubers were obtained after approximately 3 months of growth. In this experiment, the transgenic tubers from three interference-expression lines were mixed and used for transcript level of StCDPK14 and StRbohs and ROS content. The transgenic tubers from the line of StCDPK14-D were only used for the observation of suberin deposition.

The measurement of O₂^– and H₂O₂ content in healing tissues was performed using the commercial kits (Suzhou Comin Biotechnology Co. Ltd.) according to manufacturer’s instruction. For O₂^–, 0.1 g healed tissue was homogenized in extracted solution, centrifuged at 12,000 rpm for 20 min, and then the supernatants were mixed with four kinds of solution. After centrifugation of the mixture at 8000 rpm, the supernatants were prepared for measurement. For H₂O₂, 0.1 g healed tissue was homogenized in 1 mL acetone and centrifuged at 8000 × g at 4°C for 10 min. The supernatants were removed and added into a reaction solution, incubated for 5 min at room temperature and used for determination. The absorbance of the reaction to determine O₂^– and H₂O₂ content was measured at 415 and 530 nm, respectively. The O₂^– and H₂O₂ content were calculated and expressed as μmol⋅g^–1 FW and nmol⋅g^–1 FW, respectively.

The suberin deposition in transgenic tuber wound-healing tissue was microscopically detected by staining with toluidine blue and neutral red according to the method of Jiang et al. (2019). Six tubers of transgenic and wild-type control were used to observe the suberin deposition using a microscope (BX53, Olympus, Japan).

The yeast-two-hybrid analysis was conducted according to the Matchmaker Gold Yeast-Two-Hybrid System User Manual. Full length StCDPK14 was inserted into the pGBKT7 (GAL4 DNA-binding domain cloning vector) bait plasmid (pGBKT7-CDPK14), and the coding region of StRbohB was cloned into the vector of pGADT7 (GAL4 activation domain cloning vector) (pGADT7-RbohB). Both plasmids were then co-transformed into the yeast strain Y2HGold. Primers used in this assay are listed in Table 2. Mediums lacking Leu-Trp and Leu-Trp-His were used for selecting positive interactions.

A BiFC assay was conducted as described by Zhou et al. (2018). The coding region of StCDPK14 was cloned into the pSAT1-nVenus-N (PE3308) vector, resulting in nVenus-StCDPK14, and the coding sequence of StRbohB was cloned into pSAT1-cCFP-N (PE3449), resulting in StRbohB-cCFP. Primers used are listed in Table 2. Transient expression of protoplasts was detected via the polyethylene glycol-mediated transformation method. Confocal laser scanning microscope (Olympus FV 1000, Japan) was used to visualize fluorescence.

In addition, the coding region of StCDPK14 was cloned into the pEarleyGate201 vector, the EF-hand motifs of StRbohB were cloned into the pEarleyGate202 vector, resulting in pEarleyGate201-CDPK14-YN and pEarleyGate202-RBOHB-YC. The plasmid constructs were expressed in N. benthamiana leaves by Agrobacterium infiltration. The fluorescence was then visualized by a confocal laser scanning microscope (LeciaSP8, Germany). Primers used are listed in Table 2.

All the above experiments above were performed in triplicate. Data are expressed as the means (±) of three biological replicates in each treatment. Statistical significance was examined using the least significant difference (LSD) when P < 0.05 with statistical product and service solutions (SPSS) 21.0 software. All the charts were drawn using OriginPro 8.5.

The TEM observation showed that Ca^{2 +} precipitate particles in control and BTH-treated healing tissues distributed in large quantity in cytoplasm, and occasionally in cellular Ca^{2 +} sink, such as mitochondria, endoplasmic reticulum, vacuoles, and plasmids. Additionally, large amounts of Ca²⁺ distribution were also found in the nucleus (Figure 1). However, it is not sure whether the cellular Ca²⁺ concentration was elevated by BTH treatment. Therefore, the further observation of cellular Ca²⁺ concentration by Fluo-3-AM was performed.

To evaluate the effect of BTH treatment on Ca²⁺ concentration at 4 and 72 h of wound healing of potato tubers, the fluorescence intensity that stained with Fluo-3-AM was further visualized on the wounded tubers. The results showed that the fluorescent granules in BTH-applied and control tubers were clearly observed, whereas the non-stained controls did not show fluorescence. After 4 or 72 h of wound healing, the fluorescence of cytosolic Ca²⁺ levels in BTH treatment was increased compared to that in the control, indicating that BTH markedly increased Ca²⁺ levels in healing tissue of potato tuber (Figure 2).

A total of 23 CDPK genes in tubers treated with BTH were isolated and their expression profiles in healing tissues were assessed by qRT-PCR (Figure 3). During the early stage of healing (0–1 day), StCDPK1/3/4/5/6/10/12/15/23 were BTH-inducible, whereas the others were not affected by the elicitor. The expression of StCDPK4 and StCDPK10 in BTH-treated tissues both showed a peak on the first day of healing. During the middle and late stages of healing (3–14 days), StCDPK1/8/9/10/14/15/18/19 were upregulated in BTH-treated tissues. Moreover, the StCDPK2/5/6/7/21 were only increased by BTH during the late stage of healing (5–14 days), and they were observed to be gradually increased except for StCDPK6. StCDPK12 was not observed to be BTH-inducible during the late stage of healing.

Among these genes, StCDPK14 was induced significantly by BTH in comparison with other members, which was 3.1-fold, 8.3-fold, 6.4-fold, 10-fold, and 20-fold of the control after 1, 3, 5, 7, and 14 days of healing, respectively. Additionally, according to the RNA sequencing analysis of healing tissues treated with BTH, StCDPK14 was similarly induced (Supplementary Table 1), where the fragments per kilobase of exon model per million reads mapped value (FPKM) was also found to be upregulated the most, with an increase of 4.1-fold in comparison with the control. These data suggested that StCDPK14 might play a critical role during wound healing induced by BTH. Therefore, the StCDPK14 was selected for further experiments to reveal the molecular mechanism.

The information acquisition related to proteins based on bioinformatics analysis could provide an important foundation for further functional dissection of potato CDPKs. Comparison of the sequence of the StCDPK14 protein with those from other species including Arabidopsis thaliana, Solanum lycopersicum, and Oryza sativa in the GenBank and Phytozome databases indicated that it shared a significant similarity with AtCDPK29 (90%), SlCDPK29 (89%), and OsCDPK19 (88%), respectively (Figure 4A). They shared the conserved variable domain at the N-terminal, Ser/Thr kinase domain, junction domain, and four EF-hands motif domains, suggesting that StCDPK14 was equipped with the complete domain structure of a kinase protein. Moreover, StCDPK14 was predicted to interact with StRbohA, StRbohB, and StRbohC (Figure 4B), therefore, it was tempting to speculate that an interaction between them might occur.

The different and specific subcellular locations of CDPKs may provide the potential for isoform-specific differences in mediating diverse cellular functions. To detect the subcellular localization of StCDPK14, the fusion protein of StCDPK14-eGFP was created and transformed into N. benthamiana leaves via the A. tumefaciens mediated method. Confocal micrographs displayed that the StCDPK14-eGFP fusion protein was targeted to the membrane and nucleus, and the GFP was ubiquitously expressed throughout the cell of N. benthamiana plants, suggesting that the StCDPK14 protein was membrane- and nucleus-associated (Figure 4C).

We successfully generated transgenic plant and tubers as shown in Figure 5. The amplification of expected 600 bp DNA fragment using NTP II gene specific primers appeared in four lines (StCDPK14-A, B, D, N), but not in wild-type line. Further confirmation of StCDPK14 expression in the successfully interference plants (StCDPK14-B, D, N) indicated that the transcript level of StCDPK14-D was noticeably inhibited compared to the other two lines.

To determine whether a decrease in StCDPK14 expression occurred throughout the healing stage in transgenic potato tubers, the expression of wild-type and interference-expression tubers was compared. StCDPK14 displayed a similar expression tendency in wild-type and interference-expression tubers; the interference-expression tubers had a lower transcript level during the first 24 h and the late stage of healing (Figure 6). Obviously, the expression of StCDPK14 peaked at 8 h of healing, which was 56.7% lower than that of the wild type. However, another peak of StCDPK14 expression level was observed at 14 days of healing and was 50.6% lower in interference-expression tubers than the wild-type tubers. These data indicated that the StCDPK14 in interference-expression potato tubers was suppressed.

To illustrate whether the interference-expression of StCDPK14 impacted the Rbohs genes, the expression levels of StRbohs (A–H) in transgenic tubers were also examined (Figure 7A). The interference-expression of StCDPK14 resulted in a marked decrease in the transcript levels of StRbohs during wound healing, including the early 24 h of healing. The expression of StRbohA-H in the wild type reached maximum levels ranging from 0.5 to 5.6 during wound healing, whereas the expression in the StCDPK14 transgenic tubers was lower than that. Interestingly, the inhibition effect on StRbohB is the most obvious throughout the whole period of healing, especially within the first 24 h of healing, which was remarkably inhibited by 13.1, 9.7, and 7.4-fold at 4, 8, and 12 h of healing under the interruption of StCDPK14. However, the expression of StRbohA/C/D/E/H (except for StRbohB) was not significantly inhibited in the later stage of healing. This result indicated that the interference-expression of StCDPK14 affected the StRbohs expression in the early healing stage of potato tubers and the effect on StRbohB was the most significant.

In the StCDPK14 transgenic line, a gradually reduced O₂^– content along with wound healing was observed, and the control showed a gradually increased tendency, instead (Figure 7B). The maximum difference in O₂^– content between interference-expression and wild-type tubers was displayed at 21 days of healing, which was 81.2% lower than the wild type. However, the H₂O₂ content in the two groups peaked at 4 h and 3 days of healing. After StCDPK14 was interrupted, H₂O₂ levels showed a notable decrease compared to that of the wild type. A significant decrement of 52.8 and 35% lower than the wild type was revealed at 4 h and 3 days of healing, respectively. These results indicated that the interference-expression of StCDPK14 suppressed O₂^– and H₂O₂ production during healing in tubers.

To evaluate the suberin deposition on the wounded surface of tubers, observation of tuber sections was performed by fluorescent microscopy. The results revealed that the interruption of StCDPK14 had a distinct effect on suberin deposition (Figure 8). The captured fluorescent signal meant a deposition of suberin in the wild-type control and interference-expressed tubers during healing. Obviously, the suberin deposition in the StCDPK14-interference tubers was less than that of the wild-type tubers at each time point of tuber healing, and the maximum difference was observed at 14 days. Thus, the interference of StCDPK14 caused a reduction of suberin deposition in wounded tubers.

The protein interaction prediction showed that StCDPK14 could interact with StRbohA, StRbohB, or StRbohC, namely, these proteins might be the substrates of StCDPK14 (Figure 4B). Our previous transcriptomic analysis indicated StCDPK14 and StRbohB were induced the most after BTH treatment (Supplementary Table 1). And above results showed that the transcript level of StRbohB was also dramatically reduced the most when StCDPK14 was interrupted. Therefore, a yeast-two-hybrid screen between StCDPK14 and StRbohB (PGSC0003DMG400024754) was performed to identify the interaction. The full length of StCDPK14 was fused to the GAL4 DNA binding domain of the bait vector to create the construct. For the verification of the interaction with StRbohB, the coding regions of each protein were introduced into the GAL4 activation domain of the prey vector. After the co-transformation into the Y2HGold yeast strain, the protein-protein interaction between them was reconstructed. The yeast-two-hybrid result showed that the fusion protein of StCDPK14 with StRbohB was expressed in medium lacking Leu-Trp-His and blue colonies are observed on medium with addition of X-a-gal (Figure 9A), suggesting that StCDPK14 interacted with StRbohB.

Moreover, a BiFC assay was selected to further verify the interaction of StCDPK14 with StRbohB. StCDPK14 specifically interacted with StRbohB and localized to the plasma membrane (Figure 9B). Rboh was reported to be phosphorylated at N-terminal extension with EF-hand motifs by CDPK. To further confirm the interaction between StCDPK14 and EF-hand motifs of StRbohB, another BiFC assay was performed to identify the interaction. As expected, StCDPK14 specifically interacted with the EF-hand motifs of StRbohB (Figure 9C), which indicated that the potential interaction sites exist at N-terminal of StRbohB.