CALMODULIN1 and WRKY53 Function in Plant Defense by Negatively Regulating the Jasmonic Acid Biosynthesis Pathway in Arabidopsis

Jasmonic acid (JA) is an important hormone that functions in plant defense. cam1 and wrky53 mutants were more resistant to Spodoptera littoralis than in the wild-type (WT) Arabidopsis group. In addition, JA concentration in cam1 and wrky53 mutants was higher compared with the WT group. To explore how these two proteins affect the resistance of Arabidopsis plants, we used a yeast two-hybrid assay, firefly luciferase complementation imaging assay and in vitro pull-down assay confirming that calmodulin 1 (CAM1) interacted with WRKY53. However, these two proteins separate when calcium concentration increases in Arabidopsis leaf cells. Then, electrophoretic mobility shift assay and luciferase activation assay were used to verify that WRKY53 could bind to lipoxygenases 3 (LOX3) and lipoxygenases 4 (LOX4) gene promoters and negatively regulate gene expression. This study reveals that CAM1 and WRKY53 negatively regulate plant resistance to herbivory by regulating the JA biosynthesis pathway via the dissociation of CAM1-WRKY53, then the released WRKY53 binds to the LOXs promoters to negatively regulate LOXs gene expression. This study reveals WRKY53′s mechanism in insect resistance, a new light on the function of WRKY53.


Introduction
Plant resistance to insect feeding can be improved through the following aspects: perception of herbivorous insects; excitation of early signal; hormone-mediated signal transduction; metabolism (mainly the synthesis of anti-insect substances); and phenotypic changes [1].
Arabidopsis shows an increase in cytosolic calcium concentration following Spodoptera littoralis herbivory in leaves [2]. However, plant signaling pathways are complex and different members in the same family may cause opposite effects. For example, CML37 and CML42 are two Ca 2+ ion response factors that regulate JA biosynthesis to change the tolerance to insects. CML42 is a positive regulator, while CML37 is a negative regulator [3,4]. Calmodulins (CAMs) also act as important calcium sensors and play important roles in plant stress, such as AtCaM3 which regulates heat shock, AtCaM4 participates in salt stress and negatively affects freezing tolerance [5][6][7][8][9]. However, it is not clear whether CAMs are involved in JA-regulated Arabidopsis plant defense against Spodoptera littoralis.
Many WRKYs participate in the JA biosynthesis pathway. NaWRKY3 and NaWRKY6 improve JA and JA-Ile levels by increasing the expression of JA biosynthesis genes, including LOX, AOS, AOC, and OPR [14]. HvLOX1 in barley (Hordeum vulgare) contains W-boxes in its promoter and appears to interact with WRKY transcription factors [15]. In Arabidopsis, the binding of JAV1 and WRKY51 to the AOS promoter depends on the presence of a W-box [16]. The Arabidopsis genome encodes 74 WRKYs, which are divided into three groups based on the number of WRKY domains and the pattern of the zinc-finger motif [17,18]. WRKY53 participates in many signaling pathways in Arabidopsis, including leaf senescence, drought stress, abiotic stress and biotic stress signaling [19][20][21][22][23]. However, whether WRKY53 promotes the JA biosynthesis pathway remains unexplored.
Thus, in the current study, we demonstrated how CAM1 and WRKY53 negatively regulate Arabidopsis resistance according to a yeast two-hybrid assay, firefly luciferase complementation imaging assay, in vitro pull-down assay, electrophoretic mobility shift assay and luciferase activity assay. CAM1-WRKY53 complex breaks down due to high calcium concentration, then WRKY53, which detaches from the CAM1-WRKY53 complex, negatively regulates JA content by negatively regulating LOXs gene expression.

CAM1 and WRKY53 Maybe Negatively Regulate the JA Biosynthesis Process When Resistant to Spodoptera Littoralis
Because Spodoptera littoralis herbivory increased calcium concentrations in Arabidopsis mesophyll cells, we reasoned that CAMs might take part in Ca 2+ signal transduction. Many studies have shown that CAM1 functions in plant resistance. CAM1 interacts with and activates GAD1, leading to the accumulation of GABA, which is involved in many plant stress responses [24,25]. CAM1 is also involved in abscisic acid-mediated ROS accumulation, stomatal closure, and leaf senescence [26]. Furthermore, changes in CAM1 and CAM4 expression lead to changes in NO levels in Arabidopsis leaves [8]. By contrast, CAM2 is involved in pollen development [27], CAM3 mainly participates in the heat shock response in Arabidopsis and CAM7 promotes photomorphogenesis [6,28]. Therefore, we selected CAM1 for further analysis.
Since CAMs interact with group II WRKYs and group III WRKY53 [29,30] and WRKYs activate LOXs transcription, we measured five group II WRKY gene (WRKY11, WRKY15, WRKY17, WRKY21, WRKY39) expression, and WRKY53 gene expression in WT Arabidopsis and the cam1 mutant ( Figure S1). Only WRKY53 was significantly expressed in the cam1 mutant compared with that in WT Arabidopsis.
To confirm the effect of CAM1 and WRKY53 in Arabidopsis on the resistance of Spodoptera littoralis, we inoculated WT, cam1 and wrky53 Arabidopsis plants with Spodoptera littoralis larvae and the WT group was the control group. After 7 days, we measured the length and weight of Spodoptera littoralis larvae. Both growth indicators were lower in cam1 and wrky53 plants at different levels than in control plants (Figure 1a-c). Additionally, the cam1 and wrky53 mutant leaves were also eaten less than WT plants (Figure 1d). Because JA plays an important role in plant defense, we tested JA concentration in WT, cam1 and wrky53 plants. The results showed that JA concentration was significantly higher in cam1 plants and wrky53 plants than in WT plants. (Figure 1e). These results indicate that JA is important in plant defense against Spodoptera littoralis in Arabidopsis, CAM1 and WRKY53 maybe negatively regulate this JA biosynthesis process according to the changes in JA concentration. To further investigate this hypothesis, we conducted the following experiments.
according to the changes in JA concentration. To further investigate this hypothesis, we conducted the following experiments.

CAM1 Interacts with WRKY53
Because the growth of the insect in both mutants was reduced and the JA content was increased in cam1 and wrky53 plants, we reasoned that the roles of CAM1 and WRKY53 are closely related. WRKYs interact with various CAMs [29,30]. To determine whether CAM1 and WRKY53 interact, we carried out yeast two-hybrid (Y2H), firefly luciferase complementation imaging (LCI) and in vitro pull-down assays. Since WRKY53-BK showed self-activation in the Y2H assay, we used the N-terminus (residues 1-217) of WRKY53 as the bait protein and CAM1 as the prey protein. CAM1 and WRKY53 interacted in the Y2H system, as confirmed in three independent experiments (Figure 2a). To further confirm this interaction, we injected N. benthamiana leaves with different combinations of Agrobacterium cultures harboring CAM1-Cluc and WRKY53-Nluc in four separate quadrants of each leaf (Cluc/Nluc, CAM1-Cluc/Nluc, Cluc/WRKY53-Nluc and CAM1-Cluc/WRKY53-Nluc). Leaf quadrants harboring CAM1-Cluc/WRKY53-Nluc had stronger fluorescent signals than the other three quadrants (Figure 2c). CAM1 also interacted with WRKY53 in an in vitro pull-down assay. It showed that both CAM1-GST and GST combined with glutathione beads, once HIS fusion protein is bound to these proteins, it can be shown in the anti-HIS pulldown. Additionally, anti-HIS pulldown showed that WRKY53-HIS interacted with CAM1-GST in the pull-down assay, whereas WRKY53-HIS failed to interact with GST ( Figure 2b). These results indicate that CAM1 interacts with WRKY53.
For example, AtWRKY7, a typical group IId WRKY protein, contains a conserved CaMbinding domain (CaMBD) with the amino acid sequence 72-VAVNSFKKVISLLGRSR-88 [29]. WRKY53 is a WRKY group III WRKY protein. Compared with the amino acid sequence of WRKY7, WRKY53 also contains a conserved CaMBD with the amino acid sequence 57-VKQIVSSYERSLLLLNW-73: 57 V, 67 S, 68 L and 69 L are conserved ( Figure 2d). To determine whether these conserved amino acids play key roles in the interaction between CAM1 and WRKY53, we mutated these amino acids to R. Only the simultaneous mutation of all four amino acids significantly affected the binding of CAM1 and WRKY53 (Figure 2e), which suggested that 57 V, 67 S, 68 L and 69 L form a fixed structure that combines with CAM1.

WRKY53 Negatively Regulates LOX3 and LOX4 Expression
Analysis of hormonal data from the control groups indicated that the levels of JA were higher in cam1 and wrky53 than in WT plants (Figure 1e). This result indicates that CAM1 and WRKY53 negatively regulate the JA biosynthesis pathway. Arabidopsis contains four LOXs, which function in the first steps of the JA biosynthesis pathway. The promoters of LOX3 and LOX4 each contain W-boxes, a motif recognized by WRKY proteins. Therefore, to further clarify the role of WRKY53 in regulating the JA biosynthesis pathway, we performed EMSA to detect interactions between WRKY53 and the W-boxes in the LOX3 and LOX4 promoters and confirmed that WRKY53 bound to the W-boxes in the promoters of both genes ( Figure S2). Because the binding between WRKY53 and LOX3 was stronger than that between WRKY53 and LOX4, we focused on LOX3 in subsequent analysis; the results for LOX4 are shown in Figure S3.
In a luciferase activity assay, WRKY53 pGreenII 62-SK interacted with the LOX3 promoter pGreenII 0800-LUC and WRKY53 pGreenII 62-SK negatively regulated LOX3 pGreenII 0800-LUC expression. However, when CAM1 pGreenII 62-SK was added to the system, LOX3 pGreenII 0800-LUC expression returned to the level of the control group (Figure 3a). To further explore the roles of calcium and CAM1 in the binding of WRKY53 to the LOX3 promoter, we conducted two experiments. First, we performed an EMSA using the first W-box in the LOX3 promoter. When CAM1-MBP was added to the assay system, the interaction between WRKY53-MBP and the W-box decreased. However, when CAM1-MBP and calcium were added to the assay system simultaneously, the results were similar to those of samples containing only WRKY53-MBP and the W-box (Figure 3b). Second, we performed a pull-down assay, which showed that calcium affected the bonding strength of the CAM1-WRKY53 complex (Figure 3c). These results indicate that in healthy leaves (Ca 2+ in lower concentration), CAM1 is combined with WRKY53, whereas WRKY53 is rarely bound to the LOX3 promoter. When Spodoptera littoralis attacks Arabidopsis leaves, calcium levels increased and CAM1 and WRKY53 separate, allowing WRKY53 to interact with the LOX3 promoter. In addition, we demonstrated that WRKY53 negatively regulates LOX3 expression and the JA biosynthesis pathway (Figures 1e and 3a).

WRKY53 Negatively Regulates LOX3 and LOX4 Expression
Analysis of hormonal data from the control groups indicated that the levels of JA were higher in cam1 and wrky53 than in WT plants (Figure 1e). This result indicates that CAM1 and WRKY53 negatively regulate the JA biosynthesis pathway. Arabidopsis contains four LOXs, which function in the first steps of the JA biosynthesis pathway. The show that CAM1 interacts with WRKY53. In the Y2H assays, the N terminus of WRKY53 (residues 1-217) was used as the bait vector. (d) The amino acids outlined in red were conserved amino acids in WRKY53 compared to WRKY7. (e) WRKY53 is the unmutated protein; in WRKY53-1, 57 V was mutated to R; in WRKY53-2, 67 S was mutated to R; in WRKY53-3, 58 L was mutated to R; in WRKY53-4, 59 L was mutated to R; in WRKY53-5, 67-69 SLL was mutated to RRR; and in WRKY53-6, four amino acids were mutated to RRRR. shows that WRKY53 pGreenII 62-SK negatively affects LOX3 expression, and CAM1 pGreenII 62-SK decreased the negative regulation. Error bars denote ± SEM, columns labeled with different letters are significantly different at p < 0.05, Dunnett's C (variance not neat). (b,c) EMSA and pulldown assay showing the effects of calcium and CAM1 on the binding between WRKY53 and the first W-box in the LOX3 promoter. In (b), a hot probe refers to a biotin-labeled probe and a cold probe to an unlabeled probe (200-fold the concentration of the hot probe). Ca 2+ concentration is 10 −2 mM.

Discussion
In this study, we describe how CAM1 and WRKY53 affect the resistance of Arabidopsis plants by negatively regulating the JA biosynthesis pathway (Figure 4). (a) shows that WRKY53 pGreenII 62-SK negatively affects LOX3 expression, and CAM1 pGreenII 62-SK decreased the negative regulation. Error bars denote ± SEM, columns labeled with different letters are significantly different at p < 0.05, Dunnett's C (variance not neat). (b,c) EMSA and pulldown assay showing the effects of calcium and CAM1 on the binding between WRKY53 and the first W-box in the LOX3 promoter. In (b), a hot probe refers to a biotin-labeled probe and a cold probe to an unlabeled probe (200-fold the concentration of the hot probe). Ca 2+ concentration is 10 −2 mM.

Discussion
In this study, we describe how CAM1 and WRKY53 affect the resistance of Arabidopsis plants by negatively regulating the JA biosynthesis pathway (Figure 4).
We found that the cam1 and wrky53 mutants were more resistant to Spodoptera littoralis than the WT Arabidopsis plants. Due to the importance of JA in insect resistance [10], we detected the JA concentration in three plants' leaves. Compared with the WT group, the JA concentration in the cam1 mutant was increased, but the difference was not significant, and the JA concentration in the wrky53 mutant was significantly increased. Higher JA concentration enables plants to make faster defense responses [31]. Thus, we explored why JA concentration increased in these two mutants.
Notably, changes in intracellular Ca 2+ concentrations regulate the JA biosynthesis pathway. When plants are wounded by insects, CML37 and CML42 influence the JA biosynthesis pathway [3,4]. Based on the results in Figure 1, we speculate that CAM1 may also be involved in regulating the JA biosynthesis pathway. Additionally, we found that CAM1 was more closely related to WRKY53 by screening several WRKYs ( Figure S1). Because WRKYs can interact with CAMs [29,30], we reasoned that WRKY53 interacts with CAM1. Y2H, LCI, and in vitro pull-down assays confirmed this (Figure 2a-c). In addition, as WRKY53 is a group III WRKY, four conserved amino acids, 57 V, 67 S, 68 L, and 69 L, may form a fixed structure to combine with CAM1 (Figure 2d,e). Although previous studies have reported the interaction between CAMs and WRKYs, their participation in signaling pathways in plants has not been studied.  We found that the cam1 and wrky53 mutants were more resistant to Spodoptera littoralis than the WT Arabidopsis plants. Due to the importance of JA in insect resistance [10], we detected the JA concentration in three plants' leaves. Compared with the WT group, the JA concentration in the cam1 mutant was increased, but the difference was not significant, and the JA concentration in the wrky53 mutant was significantly increased. Higher JA concentration enables plants to make faster defense responses [31]. Thus, we explored why JA concentration increased in these two mutants.
Notably, changes in intracellular Ca 2+ concentrations regulate the JA biosynthesis pathway. When plants are wounded by insects, CML37 and CML42 influence the JA biosynthesis pathway [3,4]. Based on the results in Figure 1, we speculate that CAM1 may also be involved in regulating the JA biosynthesis pathway. Additionally, we found that CAM1 was more closely related to WRKY53 by screening several WRKYs ( Figure S1). Because WRKYs can interact with CAMs [29,30], we reasoned that WRKY53 interacts with CAM1. Y2H, LCI, and in vitro pull-down assays confirmed this (Figure 2a-c). In addition, as WRKY53 is a group III WRKY, four conserved amino acids, 57 V, 67 S, 68 L, and 69 L, may form a fixed structure to combine with CAM1 (Figure 2d,e). Although previous studies have reported the interaction between CAMs and WRKYs, their participation in signaling pathways in plants has not been studied.
Because WRKY53 can bind to the W-box, we reasoned that WRKY53, a WRKY family transcription factor, should bind to the W-boxes in the LOX3 and LOX4 promoters. Because WRKY53 can bind to the W-box, we reasoned that WRKY53, a WRKY family transcription factor, should bind to the W-boxes in the LOX3 and LOX4 promoters. Indeed, an EMSA performed to verify this showed that WRKY53 bound to the W-boxes in the LOX3 and LOX4 promoters ( Figure S2). Because of the upregulation of JA concentrations in cam1 and wrky53 compared to the WT (Figure 1e), we suggested that CAM1 and WRKY53 participated in JA biosynthesis because WRKY53 negatively regulated LOX3 and LOX4 expression. Additionally, a luciferase activity assay showed that WRKY53 is a negative regulator of LOX3 and LOX4. However, this negative regulatory effect was interrupted by the addition of CAM1 (Figures 3a and S3b). Therefore, we speculated that CAM1 and LOXs promotors could competitively combine with WRKY53. Referring to the results of Figure S2, the W-box1 of LOX3 promotor, which had the strongest binding ability with WRKY53, was used as the representative for the following study. We found that CAM1 and WRKY53 bound together when calcium was not added, while when calcium concentration increased, the protein interaction between CAM1 and WRKY53 was unbound, allowing WRKY53 to bind to the LOXs promoter and negatively regulating their expression (Figure 3b,c). When Arabidopsis was attacked by Spodoptera littoralis, calcium concentration increased in leaf cells [2]. Additionally, this was the reason why JA concentration was higher in the wrky53 mutant (Figure 1e). For the cam1 mutant, the results of larval length, weight gain, and JA concentration were between the data of WT and wrky53 mutant (Figure 1). We speculated that CAM1 was the upstream of WRKY53 and might be involved in a variety of signaling pathways.
Few studies have focused on the role of WRKY53 in plant defense against Spodoptera littoralis. Our findings thus enrich the understanding of the functions of WRKY53. However, CAM1 may take part in other pathways to affect Arabidopsis plant resistance to insects, and other calcium sensors may also interact with WRKY53 to participate in insect resistance, which requires further research.

Spodoptera Littoralis Egg Hatching and Inoculation
Spodoptera littoralis eggs were used as insect materials. Their hatched conditions are 28 ± 1 • C and 50 ± 10% relative humidity. After 1-2 days, each plant was inoculated with one larva with identical lengths and weights. Seven days later, the insect's lengths and weights were measured. Each group contained about 30 insects.

In Vitro Pull-Down Assay
The CDS of CAM1 was cloned into pGEX-4T-1 with BamHIand SalI and the CDS of WRKY53 was cloned into pET28A with NdeI and SalI. GST, CAM1-GST and WRKY53-HIS were transformed into E. coli Rosetta (DE3) cells for protein expression. Protein GST was used as the bait protein. The bindings of CAM1-GST with WRKY53-HIS were detected by immunoblot analysis using anti-GST and anti-HIS antibodies. The pulldown buffer was 1% NP40, 150 mM NaCl, 50 mM Tris-HCl, 1 mM EGTA, pH7.5.

Electrophoretic Mobility Shift Assay (EMSA)
The W-boxes of the LOX3 and LOX4 promoters were used to generate 3 -biotin-labeled probes. WRKY53-MBP and MBP (as a control) were used in EMSAs with a LightShift Chemiluminescent EMSA Kit (Thermo Fisher Scientific, Waltham, MA, USA) following the manufacturer's protocol. The probe sequences in the EMSA were shown in Table S1.

Measuring JA Levels in Leaves
JA levels were measured in Arabidopsis leaves in three experimental groups (WT group, cam1 group, wrky53 group). The internal standard was H 2 JA [32,33], and analysis was performed on an Agilent 1290 system (Agilent Technologies Co. Ltd., Palo Alto, CA, USA) with AB SCIEX-6500Qtrap (Anheuser-Busch, Saint Louis, MO, USA). Each group had three biological replicates and every replicate had more than 30 plants.

Statistical Analysis
Changes in JA concentration and increases in larval length and weight were examined by Dunnett's C (variance not neat) (p < 0.05).

Conclusions
In this study, we revealed the role of CAM1 and WRKY53 in Arabidopsis plant defense response. In healthy plants, CAM1 interacts with WRKY53 and WRKY53 does not bind to the LOXs promotors. However, in insect-wounded plants, WRKY53 is isolated from the CAM1-WRKY53 complex and WRKY53 negatively regulates LOXs gene expression. These results suggest that WRKY53 is a negative regulator of Arabidopsis plant defense against Spodoptera littoralis.
(Tsinghua-Peking Joint Center for Life Sciences and MOE Key Laboratory of Bioinformatics, School of Life Sciences, Tsinghua University) for providing the Cluc, Nluc, pGreenII 0800-LUC and pGreenII 62-SK plasmids; and Diqiu Yu (Xishuangbanna Tropical Botanical Garden, Chinese Academy of Sciences, China) for providing the wrky53 mutant. All authors have approved the manuscript for publication and no conflict of interest exist.

Conflicts of Interest:
The authors declare no conflict of interest.