Transcription factor SlWRKY50 enhances cold tolerance in tomato by activating the jasmonic acid signaling

Abstract

Tomato (Solanum lycopersicum) is a cold-sensitive crop but frequently experiences low-temperature stimuli. However, tomato responses to cold stress are still poorly understood. Our previous studies have shown that using wild tomato (Solanum habrochaites) as rootstock can significantly enhance the cold resistance of grafted seedlings, in which a high concentration of jasmonic acids (JAs) in scions exerts an important role, but the mechanism of JA accumulation remains unclear. Herein, we discovered that tomato SlWRKY50, a Group II WRKY transcription factor that is cold inducible, responds to cold stimuli and plays a key role in JA biosynthesis. SlWRKY50 directly bound to the promoter of tomato allene oxide synthase gene (SlAOS), and overexpressing SlWRKY50 improved tomato chilling resistance, which led to higher levels of F_v/F_m, antioxidative enzymes, SlAOS expression, and JA accumulation. SlWRKY50-silenced plants, however, exhibited an opposite trend. Moreover, diethyldithiocarbamate acid (a JA biosynthesis inhibitor) foliar treatment drastically reduced the cold tolerance of SlWRKY50-overexpression plants to wild-type levels. Importantly, SlMYC2, the key regulator of the JA signaling pathway, can control SlWRKY50 expression. Overall, our research indicates that SlWRKY50 promotes cold tolerance by controlling JA biosynthesis and that JA signaling mediates SlWRKY50 expression via transcriptional activation by SlMYC2. Thus, this contributes to the genetic knowledge necessary for developing cold-resistant tomato varieties.

Introduction

Low temperature or cold stress is a major abiotic stress hampering global vegetable productivity and inflicting major economic losses on growers (Chinnusamy et al. 2007). Tomato (Solanum lycopersicum), being a vegetable crop driven from the tropics and subtropics, is highly vulnerable to low temperatures (Liu et al. 2012). However, prior research has demonstrated that the cold tolerance of domesticated tomatoes can be significantly enhanced by using rootstocks derived from the wild tomato species Solanum habrochaites, which is native to high-altitude regions and in which jasmonic acid (JA) plays an essential role (Ntatsi et al. 2017; Wang et al. 2023). JA, as a class of lipid-derived phytohormones, affects many fundamental functions of plants and plays versatile roles in stress responses including those to low temperature (Kazan and Manners 2008; Sharma and Laxmi 2015). Exogenous JA can increase cold tolerance in tomato, trifoliate orange (Poncirus trifoliata), and loquat (Eriobotrya japonica) fruit by increasing polyamine, glycine betaine (GB), and the activity of the antioxidant enzyme (Cao et al. 2009; Ding et al. 2021; Ming et al. 2021). The inducer of C-repeat binding factor (CBF) expression–CBF (ICE–CBF) pathway is cold responsive, and JAZMONATE ZIM-DOMAIN 1/4 (JAZ1/4) in Arabidopsis (Arabidopsis thaliana) and JAZ-B-box (BBX37) module in apple (Malus domestica) are suggested to interact with ICE1 and decrease or induce the transcriptional function of ICE1, respectively (Hu et al. 2013; An et al. 2021). Therefore, identifying factors tailoring JA biosynthesis and signal transduction will provide impetus to the development of cold-resilient tomato cultivars.

An increasing number of studies have revealed a fundamental JA biosynthesis and transduction module. α-Linolenic acid (α-LeA) serves as the initial substrate. The plastid membrane releases it through the action of phospholipase A1 (PLA1) before it is converted to 12-oxo-phytodienoic acid (OPDA) via the continuous action of 13-lipoxygenase (LOX), allene oxide synthase (AOS), and allene oxide cyclase (AOC). Then, OPDA is reduced and undergoes 3 cycles of β-oxidation to become JA (Wasternack and Song 2017). The isoleucine (Ile)-conjugated JA (JA-Ile) molecule is the most biologically active, fine-tuning JA signaling and perception (Fonseca et al. 2009). When stimuli occur in plants, higher JA-Ile promotes the F-box protein CORONATINE INSENSITIVE1 (COI1) that forms a functional E3 ubiquitin ligase (SCF^COI1)-mediated degradation of JAZ repressors and finally causes the release of transcription factors (TFs) MYC2 and activates downstream JA-responsive genes. Conversely, steady-state JAZ proteins suppress MYC2 from performing its transcriptional function (Chini et al. 2007; Kazan and Manners 2008; Eremina et al. 2016). MYC2 is classified as a TF with a basic helix–loop–helix (bHLH) structure that specifically recognizes G-box and G-box variants present in the promoter regions of its target genes. By this mechanism, MYC2 regulates various aspects of the JA pathway (Chini et al. 2009; Kazan 2015). In banana (Musa acuminata) fruit, MaMYC2 participated in MeJA-induced cold tolerance by coordinating with MaICE1 (Zhao et al. 2013). Ming et al. (2021) found a regulatory mechanism consisting of PtrMYC2-PtrBADH-l (betaine aldehyde dehydrogenase-like), which regulates GB biosynthesis, increasing trifoliate orange insensitivity to cold stress. Moreover, directly or indirectly, AtMYC2 exerts a positive or inhibitory function in JA signaling by regulating TF-encoding genes, like ERFs, WRKYs, and MYBs (Dombrecht et al. 2007). Among them, a large number of TFs play a versatile function in controlling the JA signaling pathway by controlling the expression of biosynthesis and signal transmission genes. NaWRKY3 and NaWRKY6 can promote the transcription of JA biosynthesis–related genes (LOX, AOS, AOC, and OPR) in Nicotiana benthamiana after herbivorous attack, thus promoting the accumulation of JAs (Skibbe et al. 2008). Tomato ethylene response factors 15 and 16 (SlERF15/16) promote JA biosynthesis by activating the LOXD and AOC under herbivore infestation (Hu et al. 2021). NAC (NAM, ATAF1/2, and CUC2) TF VaNAC17 in grape (Vitis amurensis) can modulate endogenous JA accumulation in drought stress (Su et al. 2020). Moreover, in rubber tree (Hevea brasiliensis), the ICE-like TF HbICE2 participated in JA-regulated cold resistance (Chen et al. 2019). However, TFs involving JA biosynthesis in tomato cold-response regulation remain largely unknown.

Transcriptional regulation is the prominent possession by which all organisms control their gene transcription in response to internal and external stimuli. TFs with their target genes act in transcriptional regulatory networks by binding to sequence-specific DNA and inducing or inhibiting gene expression (Wani et al. 2021). To date, many TFs have been revealed in cold stress response, like CBFs, WRKYs, NACs, MYBs, and bHLHs (Zhu 2016; Kidokoro et al. 2022; Zheng et al. 2023). The WRKY family is one of the largest plant-specific TF families, and the name originated from the most prominent amino acid sequence feature in the WRKY domain (Bakshi and Oelmüller 2014). The WRKY domain is a region of approximately 60 amino acids that are highly conserved across different species. It contains a conserved heptapeptide sequence, WRKYGQK, at its N-terminus, followed by a zinc-finger motif at its C-terminus (Eulgem and Somssich 2007). WRKY TFs are generally divided into 3 groups: Group I contains 2 WRKY domains, and Group II contains only 1 WRKY domain. Both groups have a C2H2 zinc-finger motif, but Group III contains 1 WRKY domain and the zinc-finger motif is C2-HC. WRKY preferentially binds to the conserved cognate binding site C/TTGACT/C, called W-box, in the promoter of target genes (Ciolkowski et al. 2008). WRKY has been demonstrated to be involved in growth and development processes, like seed germination, flowering, and leaf senescence (Jiang et al. 2014; Ma et al. 2020; Niu et al. 2020). Additionally, research implies that WRKY TFs are responsible for coping with external stimulus, including cold stress (Jiang and Yu 2009). For example, in bermudagrass (Cynodon dactylon), CdWRKY2 in Group II targets the promoter of CdCBF1 and promotes cold resistance (Huang et al. 2022). In rice (Oryza sativa) plants, overexpression (OE) of OsWRKY71 and OsWRKY76 from Group II contributes to the adaptation of cold stress (Yokotani et al. 2013; Kim et al. 2016). Similarly, transgenic grapevine (V. amurensis) callus overexpressing VaWRKY12 and VaWRKY33 from Group II shows improved cold tolerance (Sun et al. 2019; Zhang et al. 2019). Consistently, Niu et al. (2012) suggested that in transgenic Arabidopsis, OE of TaWRKY19 from Group I also displays a cold-insensitive phenotype. Contrastingly, Zou et al. (2010) revealed that AtWRKY34 in Group I adversely interferes with the cold tolerance ability of mature pollen in Arabidopsis by suppressing the CBF signal cascade. In tomato, 81 WRKY genes were characterized by genome-wide computational analysis, of which at least 27 genes were differentially expressed in response to low-temperature stimulation (Wang et al. 2021). So far, the function of a few WRKY TFs in cold response of tomato has been studied. Tomato WRKY33 from Group I directly targets genes encoding kinases, TFs, and molecular chaperones, thus inducing cold resistance (Guo et al. 2022). Chen et al. (2018) found that plants with transiently silenced Group II SlWRKY50 showed sensitivity to cold stress. Nevertheless, our understanding of the precise mechanism of SlWRKY50 in regulating tomato cold resistance and downstream gene expression remains incomplete.

Here, we deciphered the molecular mechanism of SlWRKY50 in the cold resistance of tomato. The JA-inducible SlWRKY50 can directly interact with the promoter of SlAOS, leading to the activation of SlAOS expression and subsequent accumulation of JAs, ultimately enhancing the antioxidant activity of tomato plants at low temperatures. It is intriguing to note that JA signaling also regulates the transcription of SlWRKY50 through SlMYC2, a key regulator in JA signal transmission. Collectively, our work provides a mechanism by which a positive feedback loop mediated by SlWRKY50 and JA signals positively regulates the cold stress responses of tomato.

Results

Induced expression difference of SlAOS in cold-sensitive and cold-tolerant tomato genotype

JA is a key hormone in the fine-tuning of responses in tomato plants to low-temperature stimuli. Previously, we reported cold induction in the transcription levels of JA biosynthesis genes in cold-sensitive and cold-tolerant grafted tomato seedlings (Wang et al. 2023). Tomato SlAOS (Solyc04g079730), which encodes AOS, a key rate-limiting JA biosynthesis enzyme, was observed with obvious expression changes under cold stress, whose expression rose 12-fold in the cold-tolerant genotype compared with only 2.7-fold in the sensitive genotype after 3 h of 4 °C treatment. The expression analysis of SlAOS in 4 tissues of tomato showed that it was mainly expressed in the root, followed by the leaf, stem, and stem apex, and its expression in the cold-tolerant genotype was generally higher than that in the cold-sensitive genotype in the root and stem (Fig. 1A).

We conducted a more comprehensive expression analysis of SlAOS in cold-tolerant and cold-sensitive seedling leaves using reverse transcription quantitative PCR (RT-qPCR), taking samples at different times. In both genotypes, SlAOS expression increased steadily during the first 9 h and then displayed a dramatic induction between 9 and 12 h. This increase continued to the final time point tested (24 h). At all the time points, the transcription level of SlAOS in the cold-tolerant seedlings was notably stronger than that in the cold-sensitive genotype (Fig. 1B). Based on this preliminary expression analysis, we speculate that SlAOS could be a potential player in regulating tomato response to cold stress.

SlWRKY50 directly binds to the promoter of SlAOS to activate its expression

To characterize the regulatory TFs involved in the upregulation of SlAOS at low temperatures, promoter cis-element prediction, weighted gene coexpression network analysis (WGCNA), and gene relative expression analysis under cold treatment were performed. Promoter sequence in the region −2.0 kb from the transcription start site of SlAOS was used. We identified the WRKY-binding W-box (T/CTGACC/T) element or its core motif TGAC in the promoter region of SlAOS using the online perdition tool PlantCARE (Fig. 2A). Then, by analyzing our previous RNA-sequencing data, we obtained a gene dendrogram based on clustering the dissimilarity using consensus topological overlap. Each colored block represents a color-coded module that groups together highly connected genes. Our target gene SlAOS was found in the pink module (Fig. 2B). Additionally, a network map of SlAOS and its associated WRKY TFs was built (Fig. 2C). Among all the WRKYs, SlWRKY50 was highly coexpressed with SlAOS (weight = 0.2796), which was previously reported for inducing cold tolerance in tomato (Chen et al. 2018). Expression analysis showed that SlWRKY50 was roused at low temperature and that the transcription level in the cold-tolerant seedlings was significantly stronger than that in the sensitive seedlings (Fig. 2D). Therefore, we selected SlWRKY50 to further confirm its interactive activity with SlAOS and characterize its function under cold stress. Additionally, to confirm the subcellular localization of SlWRKY50, we expressed a yellow fluorescent protein (YFP)-fused version of SlWRKY50 transiently in N. benthamiana leaves. Results showed that the YFP signal of the control vector (35S: YFP) was evenly filled throughout the cell, whereas the YFP signal for SlWRKY50-YFP was specifically localized in the nucleus and colocalized with the nuclear marker (VirD2NLS: mCherry) (Fig. 2E).

Next, SlWRKY50 was tested to investigate whether it induces the transcription of SlAOS to regulate JA biosynthesis. The yeast 1-hybrid (Y1H) assay was utilized to examine the role of SlWRKY50 in gene activation. The assay revealed that yeast cells transformed with the SlWRKY50 prey and baits constructed using Pro_AOS⁻²⁰⁰⁰ and Pro_AOS⁻³⁵¹ grew well on the selective medium, but fewer yeast cells were observed on the plates transformed with empty vector (Fig. 3A). To further verify the role of SlWRKY50 in gene activation in vivo, we generated firefly luciferase (LUC) reporter constructs driven by SlAOS promoter and an effector using SlWRKY50. Transient expression assays demonstrated that the coexpression of the effector with the reporter led to greater activation of LUC expression than the control (Fig. 3B) and to a raised LUC/Renilla (REN) ratio in N. benthamiana leaves (Fig. 3C). This was further supported by GUS staining analysis of transiently transformed N. benthamiana, wherein the SlAOS promoter-driven GUS reporter gene was activated by SlWRKY50 effector (Supplemental Fig. S1A). To verify whether SlWRKY50 can directly bind the W-box elements in the SlAOS promoter, an electrophoretic mobility shift assay (EMSA) was conducted in vitro. The probe was designed using the W-box inside the Pro_AOS⁻³⁵¹ promoter. Incubation of the GST-SlWRKY50 protein with the labeled probe produced a change in mobility, whereas the addition of unlabeled competitor DNA resulted in a dosage-dependent reduction in electrophoretic mobility shift (Fig. 3D). SlWRKY50 protein did not form a protein–DNA complex when the mutated probe was added, confirming that SlWRKY50 specifically binds directly to the W-box element within the Pro_AOS⁻³⁵¹ region. Our data indicated that SlWRKY50 works upstream of the SlAOS and stimulates its expression in the process.

Tomato SlWRKY50 is a positive factor in cold tolerance

To clarify the function of SlWRKY50 in the cold resistance of tomato seedlings, we selected SlWRKY50-silenced plants (pTRV-SlWRKY50) with only 30% to 40% SlWRKY50 transcripts of the empty vector plants for the assessment of cold tolerance (Supplemental Fig. S2A). Compared with the pTRV plants, pTRV-SlWRKY50 displayed serious wilting and chilling-induced injury after 48-h cold treatment (Fig. 4A), consistent with the results of the physiological indicators. The maximum quantum yield of photosystem II (F_v/F_m) serves as a representative indicator of PSII performance. Before cold treatment, F_v/F_m of pTRV-SlWRKY50 plants was not significantly different from control plants. However, after cold stress treatment, compared with pTRV, the F_v/F_m value decreased markedly in the pTRV-SlWRKY50 (Fig. 4B). In contrast, the relative electrolyte leakage (REL), an important parameter to measure the cell membrane stability, was greater in the pTRV-SlWRKY50 plants than that in the control plants (Fig. 4C). Similar results were obtained for malondialdehyde (MDA) and H₂O₂ content (Fig. 4, D and E). Conversely, compared with pTRV plants, the activities of antioxidant enzymes superoxide dismutase (SOD), peroxidase (POD), and catalase (CAT) of pTRV-SlWRKY50 growing at 4 °C were significantly inhibited (Fig. 4, F to H).

We also generated SlWRKY50 OE lines and obtained 2 lines with significant upregulation of SlWRKY50 (OE#1 and OE#2) (Fig. 5, A and B). However, no offspring were obtained for the OE#2 line, likely due to the extremely high expression of the transgene (Supplemental Fig. S2B). Therefore, we used the T₁ generation of OE#1 and T₀ of OE#2 for further analysis and compared the physiological responses between the wild type (WT) (LA4024) and OE lines after 48-h exposure to cold treatment. As seen in Fig. 5C, WT plants showed severe wilting and leaf curling under cold stress, whereas SlWRKY50-OE plants were relatively less affected. This was further supported by F_v/F_m, REL, MDA, and H₂O₂ measurements, indicating that SlWRKY50-OE plants had a stronger tolerance to cold stress than WT plants (Fig. 5, D to G). For example, there was no difference in F_v/F_m under normal conditions between WT and SlWRKY50-OE plants; however, under cold stress, F_v/F_m values dropped to 0.29 for WT and 0.43 and 0.49 for SlWRKY50-OE (Fig. 5D). Additionally, the activities of antioxidant enzymes SOD, POD, and CAT in SlWRKY50-OE seedlings were significantly stronger than those of WT plants with cold treatment (Fig. 5, H to J). Notably, we selected 2 cold-response genes for validation: CBF1 (Solyc03g124110) and WRKY33 (Solyc09g014990), and the transcription of CBF1 was activated in SlWRKY50-OE plants under cold stress (Fig. 5K), but the activation of WRKY33 was not obvious (Fig. 5L). These results indicate that SlWRKY50 positively regulates the cold tolerance of tomato.

SlWRKY50 modulates cold-induced JA accumulation by regulating SlAOS expression

To further understand whether SlWRKY50 positively regulates cold stress via JA accumulation, we applied diethyldithiocarbamate acid (DIECA), a JA biosynthesis inhibitor, to SlWRKY50-OE plants, whereas ddH₂O was sprayed over other OE#1 plants as a control. As shown in Fig. 6, SlWRKY50-OE plants showed enhanced cold tolerance, but the application of DIECA largely diminished its cold tolerance. The phenotype of the SlWRKY50-OE plants sprayed with DIECA was similar to that of WT after 48 h of cold treatment (Fig. 6A), as reflected in F_v/F_m, REL, MDA, and H₂O₂ content and antioxidant enzyme activity (Fig. 6, B to H).

We then quantitatively measured the transcription level of SlAOS in pTRV-SlWRKY50, SlWRKY50-OE, and WT plants. At low temperature, the SlAOS expression level decreased in SlWRKY50-silenced plants compared with that of pTRV plants (Fig. 7A). SlAOS expression in SlWRKY50-OE plants was higher than in WT at normal and low temperatures, with the difference being more significant at low temperature (Fig. 7B). Moreover, when exposed to cold treatment, JA accumulation was significantly lower in pTRV-SlWRKY50 plants than in pTRV plants (Fig. 7C), and JA accumulation in SlWRKY50-OE plants was significantly higher than in WT. However, the difference was less after DIECA application (Fig. 7D). Our results further support the notion that JA is required for plant survival under cold stress. The concert of SlWRKY50-JA laid the foundation for producing cold-tolerant tomato lines.

SlMYC2 orchestrates JA signaling to regulate SlWRKY50 expression

Through analysis of publicly accessible ChIP-Seq data (PRJCA000395, National Genomics Data Center) (Du et al. 2017), we found that SlMYC2, a crucial TF of the JA signaling pathway, may target SlWRKY50. To test whether SlMYC2 can orchestrate JA signaling to control the transcription of SlWRKY50 that occurs in response to cold treatment, we first quantitatively measured the expression of SlWRKY50 in JA synthesis mutant (spr8), JA overaccumulation lines (35S: PS). Under normal growth conditions, spr8 plants showed a slightly lower level of SlWRKY50, whereas the 35S: PS plants showed higher levels of SlWRKY50 transcript than their genetic background cultivar Castlemart (CM). Cold treatment induced the mRNA abundance of SlWRKY50, but this induction was more obvious in the 35S: PS lines (Fig. 8A). Then, the Y1H assay indicated that SlMYC2 could interact with the SlWRKY50 promoter (Fig. 8B). Moreover, in the dual-LUC assay, coexpression of Pro_WRKY50: LUC and 35S: MYC2 showed stronger luminescence signal than the control (Fig. 8, C and D). Together with GUS staining results (Supplemental Fig. S1B), this confirmed that SlMYC2 acts as a transcriptional activator of SlWRKY50 in vivo. Next, we conducted an EMSA, and the results showed that SlMYC2 could interact with the G-box in the SlWRKY50 promoters, and the binding ability decreased with the increment of competitor levels in a dosage-dependent manner (Fig. 8E). We then compared the temporal expression patterns of SlWRKY50 and SlMYC2 in response to cold treatment. RT-qPCR assays revealed that the expression of SlMYC2 induced by cold treatment was largely abolished, and SlWRKY50 was largely suppressed in the jai1 mutant, which harbors a null mutation of the JA-Ile receptor in tomato (Li et al. 2004), indicating that the cold-induced expression of SlWRKY50 depends on JA signaling. Notably, although the expression of SlMYC2 and SlWRKY50 in WT plants was induced by cold, their induction timing was different. The expression of SlWRKY50 peaked at 6 h, whereas the expression of SlMYC2 peaked at 9 h after cold (Fig. 8F). These results suggest that SlWRKY50 activates the JA signaling pathway, and then SlMYC2, in turn, induces SlWRKY50 transcription.

Discussion

Tomato plants are susceptible to low temperature, but some wild species are relatively cold tolerant. For the genetically engineered improvement of this vegetable crop, it is crucial to comprehend the underlying mechanism of cold tolerance in tomatoes. Our previous research has suggested that JA, a stress phytohormone, plays a critical role in tomato cold tolerance (Wang et al. 2023). However, studies on the upstream regulators of the JA-signaling pathway are scarce. In this study, we demonstrated that SlAOS expression and late JA biosynthesis in tomato are transcriptionally regulated by SlWRKY50 under cold stress, and SlWRKY50 expression is further promoted by SlMYC2, a key TF in the JA signaling pathway, which provided a molecular link between cold stress and JA accumulation.

SlWRKY50 contributes to tomato cold tolerance by targeting SlAOS and JA accumulation

Transcriptional regulation is an important process by which all organisms regulate gene expression in response to changing internal and external conditions. The WRKY TF family is one of the most important TF families in sessile plant perception and adaptation to many adverse conditions, including harsh weather, salinity, and drought (Ding et al. 2015; Jha et al. 2019; Lv et al. 2020; Huang et al. 2022; Alabd et al. 2023). Until recently, only a few plant WRKYs have been reported to be key components in the regulation of cold response, like VaWRKY12 and VaWRKY33 in grapevine (Sun et al. 2019; Zhang et al. 2019), CdWRKY2 in bermudagrass (Huang et al. 2022), OsWRKY71 and OsWRKY76 in rice (Yokotani et al. 2013; Kim et al. 2016), and TaWRKY19 in Arabidopsis (Niu et al. 2012). Guo et al. (2022) found that ShWRKY33 possessing self-transcription function confers tomato cold tolerance, which exerts a critical W-box in its promoter. Herein, according to our transcriptome and RT-qPCR evaluations, we observed that under cold stress, the transcription level of SlWRKY50 in the cold-tolerant tomato genotype was markedly stronger than that in the cold-sensitive tomato (Fig. 2, C and D). Chen et al. (2018) found that the proline content and SOD activity decreased while MDA content increased in the SlWRKY50-silenced tomato plants. But in our study, in addition to silencing SlWRKY50, we also generated OE plants to further confirm the positive function of SlWRKY50 in tomato cold response. Based on our results, OE of SlWRKY50 enhanced cold tolerance whereas silencing of SlWRKY50 reduced cold tolerance in tomato, which can be assessed based on phenotype, physiological parameters including F_v/F_m, REL, MDA, H₂O₂ content, and antioxidant enzyme activity, and the expression of CBF1(Figs. 4 and 5). However, consistent with the results of Guo et al. (2022), in which a variation was detected in the critical W-box in the promoter of SlWRKY33, resulting in the loss of self-activation of this gene, this can also explain why the expression of SlWRKY33 did not significantly change in SlWRKY50 transgenic tomato plants (Fig. 5L). It is undeniable that WRKYs are not only involved in cold stress but also participate in many growth and metabolic processes. AtWRKY71 plays a prominent function in hastening the flowering process by directly activating FT and LFY (Yu et al. 2016). AtWRKY2 and AtWRKY34 regulate pollen development, germination, and tube growth (Lei et al. 2017). In this study, when the expression of SlWRKY50 was high, the growth of flower buds was suspended (Supplemental Fig. S2B), indicating that SlWRKY50 inhibits the reproductive growth of plants, likely due to the effect of JA accumulation; however, Zhai et al. (2015) only reported that JA could delay flowering but not influence differentiation. The mechanism underlying this phenotype is intriguing and warrants further investigation.

Plants have developed a sophisticated system to maintain homeostasis and generate protective compounds, like hormones, enzymes, and osmoprotectants, to mitigate the passive effects of cold stress (Li et al. 2021; Ming et al. 2021; Huang et al. 2022; Liu et al. 2023). Previous findings confirmed that JA is essential for enhancing the cold tolerance of tomato (Wang et al. 2023). JA is produced from α-LeA in the chloroplast membrane, catalyzed by LOXs, AOS, AOC, and OPR, and it is critical for balancing plant growth and stress response (Cao et al. 2009; An et al. 2021; Ding et al. 2021; Ming et al. 2021). The biosynthesis of JA is controlled at both transcriptional and posttranscriptional processes (Wasternack and Hause 2013). Many regulators have been identified in the regulation of JA biosynthesis, like the positive regulators NaWRKY3 and NaWRKY6 (Skibbe et al. 2008), SlERF15 and SlERF16 (Hu et al. 2021), VaNAC17 (Su et al. 2020), HbICE2 (Chen et al. 2019) and OsEIL1 (Ma et al. 2019), and the negative regulators GhWRKY70 (Xiong et al. 2019).

AOS is the enzyme that catalyzes the first specific step in the conversion of α-LeA to JAs in plants (Itoh et al. 2002). We found that cold-tolerant tomato seedlings accumulated a higher content of JA/JA-Ile in the leaves than cold-sensitive ones, which was attributed to the higher expression of SlAOS (Fig. 1B). However, the upstream regulators of AOS remain to be characterized in either model or nonmodel plants when exposed to cold stress. WRKY activates downstream genes by targeting the W-box element in the promoter (Ciolkowski et al. 2008). Herein, we found that SlWRKY50 positively regulates JA biosynthesis by targeting SlAOS. EMSA and Y1H assays demonstrated that SlWRKY50 binds to the SlAOS promoter, and based on LUC and GUS assay results, SlWRKY50 can activate its expression (Figs. 2 and 3). Moreover, our study showed that OE of SlWRKY50 significantly promotes SlAOS expression, resulting in increased JA accumulation and reduced damage during cold stress (Figs. 5 and 7). GhWRKY70 was suggested to passively regulate tolerance to soilborne fungus Verticillium dahliae by reducing the expression of JA-associated genes in cotton (Gossypium hirsutum) (Xiong et al. 2019). Our findings indicate that SlWRKY50 is capable of directly activating the JA signaling pathway in response to cold stress, thus shedding light on the interplay between TFs and plant hormones in regulatory networks.

JA signaling participates in SlWRKY50-mediated cold stress response in tomato

Experimental evidence shows that when exposed to multiple abiotic stresses (including cold and drought), the exogenous application of plant hormones alters the transcription levels of WRKY genes (Yan et al. 2014). For example, the synergistic interaction between abscisic acid (ABA)-inducible OsWRKY51 and OsWRKY71 genes suppresses gibberellic acid signaling in rice seed aleurone cells (Hwang et al. 2016). Similarly, AtWRKYs 18, 40, and 60 are involved in signaling pathways that are regulated by plant hormones salicylic acid, JA, and ABA (Chen et al. 2010). Surprisingly, according to the ChIP-Seq data provided by Du et al. (2017), SlMYC2 is expected to bind the SlWRKY50 promoter, which contains several G-box elements. MYC2 is a crucial regulator of JA signaling, known to form transcriptional modules with other TFs. In tomato, MYC2 was found to target a variety of TFs, indicating its high hierarchical level of function (Du et al. 2017). In Arabidopsis, MYC2 can directly regulate both JA biosynthesis and catabolism genes while also regulating octadecanoid-responsive AP2/ERF-domain TF 47 (ORA47), which in turn directly targets JA biosynthesis genes (Chen et al. 2016). Additionally, MYC2 is involved in the activation of several JA-dependent physiological processes in Arabidopsis (Dombrecht et al. 2007; Kazan and Manners 2008; Du et al. 2017). We provide evidence that a sharp increase in SlWRKY50 expression was due to transcriptional activation by MYC2. First, using tomato materials with different JA concentrations (LOX deleted mutant [spr8], overaccumulation JA [35S: PS], and their WT [CM]), we further suggested that plants with a stronger accumulation of JA biosynthesis had more SlWRKY50 expression (Fig. 8A). Second, Y1H, EMSA, and LUC transient expression experiments were used to confirm the association between SlMYC2 and the SlWRKY50 promoter (Fig. 8, B to E). Additionally, under low-temperature treatment, the transcript of SlWRKY50 reached the peak before MYC2, whereas in jai1 mutant, the induction of SlWRKY50 in response to cold treatment was not significant (Fig. 8F). Our previous research has shown that JA is rapidly induced in the early stages of cold stress (Wang et al. 2023), although the specific causes of early JA accumulation are yet to be determined. These findings suggest that the early accumulation of JA under cold stress can activate the expression of SlWRKY50. This triggers the transcription of SlAOS, resulting in the subsequent accumulation of JA during later stages and effectively regulating the expression of SlWRKY50 through signal transduction. Notably, a recent discovery suggested an autoregulatory negative feedback loop between MYC2 and MYC2-targeted bHLH, which negatively regulates MYC2-targeted genes and JA signaling (Liu et al. 2019). However, our findings indicated the presence of a self-regulatory positive feedback loop between JA and JA-targeted WRKY50, which positively regulates JA signaling transduction under cold stress. These findings shed light on the dynamic changes occurring in JA biosynthesis and on how plants maintain a balance between defense and growth.

Conclusion

In conclusion, our study has elucidated the role of SlWRKY50 as a TF in enhancing the cold tolerance of tomato plants. It was found that an early accumulation of JA triggers the expression of WRKY50 (although the underlying causes are to be determined). SlWRKY50 can activate SlAOS and promote subsequent JA accumulation, which further reinforces the expression of SlWRKY50 through the MYC2-WRKY50 pathway, establishing a self-amplifying feedback loop. The accumulated JA could activate antioxidant enzyme genes, leading to enhanced ROS scavenging and ultimately improving cold tolerance (Fig. 9).

Materials and methods

Plant materials

Tomato (S. lycopersicum) “LA4024” was used as the WT tomato in this study, and seeds were planted in sowing trays. The seedlings were raised in controlled growth chambers. After the seedlings grew 2 leaves, the soil was gently washed away from the roots and supplied with a hydroponic solution of Hoagland's solution at half strength, which was replaced every 5 d as described in our previous study (Wang et al. 2023).

To create the SlWRKY50-OE lines in the LA4024 background, the 1,080-bp coding sequence (CDS) without termination codon was PCR amplified from tomato cDNA using the OE-WRKY50-F and OE-WRKY50-R primers (Supplemental Table S1). The PCR product was digested with XbaI and SacI and inserted downstream of the d35S promoter in the binary plasmid vector pBSE403DsR, which carries a red fluorescent protein (RFP) tag for the identification of transgenic plants. The expression vectors were transformed into Agrobacterium tumefaciens strain GV3101 as described by Ming et al. (2021), and 2 independent homozygous OE lines were utilized for the experiment.

To create virus-induced gene silencing (VIGS) tomato plants, the cDNA fragment of SlWRKY50 was amplified via PCR using the gene-specific primers (VIGs-WRKY50-F; VIGs-WRKY50-R) listed in Supplemental Table S1. The amplified SlWRKY50 fragment was digested with SmaI and inserted into the corresponding sites of the pTRV2 vector. An empty pTRV2 vector was used as the control. The constructs were separately transformed into A. tumefaciens strain GV3101, and tomato plants were transformed according to Ming et al. (2021). After 21 d, fully grown leaves were collected and checked with RT-qPCR for the selection of silenced plants.

RNA extraction and RT-qPCR analysis

For RNA extraction, TRIzol reagent (Invitrogen, Carlsbad, USA) was utilized, and 500 mg of leaf samples was ground for this purpose. HiScript III RT SuperMix Kit (R323-01, Vazyme, Nanjing, China) was used for cDNA synthesis, and RT-qPCR was conducted on a QuantStudio 7 Flex system (Applied Biosystems, USA) with tomato actin as a reference and 4 biological replicates according to the methodology specified by Long et al. (2022). Supplemental Table S2 provides primers for RT-qPCR.

Subcellular localization analysis

The full-length CDS of SlWRKY50 without termination codon was amplified with the primer pair YFP-WRKY50-F and YFP-WRKY50-R (Supplemental Table S1) and was ligated into p101YFP vector. A. tumefaciens strain GV3101 was used for the transformation of N. benthamiana together with the plasmid expressing a nuclear marker, VirD2NLS: mCherry (Meng et al. 2022), and the subcellular localization of SlWRKY50 was observed under a confocal laser scanning microscope (TCS SP8; Leica Wetzlar, Germany) according to an established protocol (Ming et al. 2021). The excitation wavelength of the fluorescent signal is 514 (YFP) or 587 nm (mCherry), and the emission wavelength is 520 to 540 nm (YFP) or 610 to 630 nm (mCherry), respectively.

Preparation of protoplasts

The injected N. benthamiana leaves were cut into 0.5- to 1.0-cm-wide strips along the vein with a sharp blade and placed in the protoplast isolation solution (each 20-ml protoplast isolation solution contained 300-mg cellulose R10, 60-mg Macerozyme, 200-ul 2 M KCl, 800-ul 0.5 M MES [pH 5.7], 1.092-g mannitol, 200-ul 1 M CaCl₂, and 200-mg BSA). The strips were shaken in a 40-rpm shaker at room temperature for approximately 5 h. The enzymolysis products were filtered with a 300-mesh sieve and washed 2 to 3 times with W5 solution (each 500-ml W5 solution contained 3.375-g NaCl, 5.2-g CaCl₂, 0.938-ml 2 M KCl, and 1.5-ml 0.5 M MES [pH 5.7]); the protoplast can be obtained.

Y1H assay

Full-length or partial promoter fragments of SlAOS or SlWRKY50 were ligated into the pHis2 vector to construct baits. The plasmids were transformed into the Y187 yeast strain, and the inhibitory concentration of 3-amino-1,2,4-triazole (3-AT) was screened by spreading gradient concentration SD/-Leu/Trp plates to avoid self-activation. Then, the full length of SlWRKY50 or SlMYC2 CDS was inserted into the vector pGADT7 as prey. All the recombinant vectors were transformed into yeast strain Y187. Yeast cells were grown for 3 d at 28 °C on SD/-Trp/-Leu/-His medium and were added with or without 3-AT to avoid self-activation. The experimental procedure was performed as previously reported (Lv et al. 2020).

Dual-LUC assays and GUS histochemical staining

Full-length promoter fragments of SlAOS or SlWRKY50 were ligated into the pGreenII0800-LUC and pBI121 vector as reporter construct, whereas the full length of SlWRKY50 or SlMYC2 CDS was inserted into pGreenII 62-SK to generate an effector. The A. tumefaciens method was used to infiltrate N. benthamiana leaves using needleless syringes as previously reported by An et al. (2021). LUC and REN LUC activities were measured by using the Dual-Luciferase Reporter Assay System (Promega, Madison, WI, USA) on a microplate reader (Infinite 200 Pro; Tecan) as described by the manufacturer. GUS histochemical staining was carried out according to the method described by Lv et al. (2020). Then, chlorophyll was removed using 75% (v/v) ethanol, and the leaves were photographed.

EMSA

The full-length CDSs of SlWRKY50 and SlMYC2 were ligated into the pET-28a vector and digested with BamHI and SacI. The recombinant vector was expressed in Escherichia coli strain DE3 to induce WRKY50-GST and MYC2-GST fusion protein, respectively. Two single-strand DNA fragments (Supplemental Table S1) containing a cis-acting element that binds to the corresponding TF were synthesized based on the promoter fragment and labeled by Sangon Biotech (Shanghai) Co., Ltd. whereas unlabeled fragments were used as competitors. A binding assay was performed, and protein–DNA complexes were separated on 6% native-polyacrylamide gel and electroblotted onto a Hybond nylon membrane (Biosharp, Hefei, China), as reported by Liu et al. (2023).

Physiological measurements

F_v/F_m was measured by imaging PAM (IMAG-MAX/L, Germany), and REL was calculated as the ratio of EC1/EC2. MDA content, H₂O₂, and CAT were measured by using detection kits from the Nanjing Jiancheng Institute of Biological Engineering, China. The physiological parameter was measured by following protocols provided in our previous study (Wang et al. 2023).

Quantification of JA contents

JA was measured with detection kits (ml036359 for JA; Mlbio, Shanghai, China) by the ELISA method. Briefly, 100 mg of tissue was homogenized with 1 ml of ice-cooled extraction buffer (100 mM potassium phosphate buffer containing 1% PVP [w/v], pH 7.8), and readings were recorded at 450 nm (OD₄₅₀) on the microplate reader (Tecan), according to Ming et al. (2021).

Statistical analysis

Plots were generated using Rstudio 4.03, and data were analyzed using Duncan's Multiple Range Test, Student's t-test, or ANOVA. At least 3 biological replicates were used for the calculation of Sd, and figures were processed with Adobe Photoshop 6.0 software (Adobe Systems, San Jose, CA, USA) for making final figures.

Accession numbers

Sequence data from this article can be found in the reference tomato genome (https://solgenomics.net, SL2.5), and the accession numbers are listed in Supplemental Table S2.

Acknowledgments

We thank Prof. Chuanyou Li (Institute of Genetics and Developmental Biology, Chinese Academy of Sciences, Beijing, China) for providing the spr8, jai1 mutant, 35S: PS lines, and their background parent used in this study. We thank PD. Hamza Sohail for critically reading our manuscript.

Author contributions

L.W., Z.B., and B.O. planned and designed the research; L.W., G.C., X.S., and H.C. performed the experiments; L.W. and G.L. analyzed the data; and L.W., Z.B., and B.O. wrote the manuscript.

Supplemental data

The following materials are available in the online version of this article.

Supplemental Table S1. PCR primer sequences used for vector construction.

Supplemental Table S2. List of primer sequences used for RT-qPCR analysis.

Supplemental Table S3. List of primer sequences used for EMSA analysis.

Supplemental Figure S1. SlWRKY50 transcriptionally activates SlAOS, and SlMYC2 transcriptionally activates SlWRKY50.

Supplemental Figure S2. Information about transgenic plants of SlWRKY50-OE and SlWRKY50-silenced efficiency of pTRV-SlWRKY50 plants.

Supplemental Data Set 1. WGCNA in pink module.

Funding

This research was supported by the National Key R&D Program of China (2019YFD1001900 and 2022YFE0100900), the Natural Science Foundation of Hubei Province of China (2019CFA017), the Key Research and Development Project of Hubei Province (2021BBA239), and the Fundamental Research Funds for the Central Universities (2662023YLPY008).

Data availability

All the data generated in this study are included in this publishing article and its Supplemental data.

References

Author notes