Introduction

Sub-functionalization or neo-functionalization of genes occurs during evolution through duplication events in the genome. Gene duplications leading to genetic redundancy can allow the acquisition of new functions, interactions, and expression patterns to benefit plant adaptive responses to the changing environmental stimuli1,2,3. Examples of severe genetic redundancy can be observed in plant hormone signaling modules1,2. Genetic redundancy in plant hormone receptors makes it difficult to study and characterize their complex signaling networks3. A paradigmatic example can be found in the case of a lipid-derived plant hormone, (+)-7-iso-jasmonoyl-L-isoleucine (JA-Ile, 1, Fig. 1a), which is perceived by COI1-JAZ co-receptors consisted of F-box protein CORONATINE INSENSITIVE 1 (COI1) and JASMONATE-JIM DOMAIN (JAZ) repressor proteins. Subsequent degradation of JAZ through the 26S-proteasome induces the expression of jasmonate (JA)-responsive genes, including defenses against necrotrophic pathogens and herbivorous insects in Arabidopsis thaliana and several crop plants4,5,6. Concurrently, the COI1-JAZ-mediated JA signaling regulates growth inhibition, fertility, and secondary metabolite production. The molecular basis of the multiple effects of JA-Ile partially lies in the severe genetic redundancy of the 13 JAZ gene family members encoded in the Arabidopsis genome3,7,8.

Fig. 1: Chemical library of coronatine stereoisomers.
figure 1

a Chemical structures of JA-Ile (1), coronatine (2), coronafacic acid (3), and coronamic acid (4). b Chemical structure of stereochemical isomers of 2 (2a-2o). c Schematic representation of selective binding of 2-stereoisomer (2e) with COI1-JAZ co-receptors. COR (2) can interact with most of the COI1-JAZ pairs, whereas the stereoisomer of COR (2e) has a selective affinity with some COI1-JAZ pairs.

Differences in spatiotemporal expression of JAZ genes during growth and development lead to redundant and complementary JA signaling in Arabidopsis. Genetic redundancy seriously challenges canonical genetic approaches to functionally dissect JAZ gene function as the effect of a single JAZ knock-out in Arabidopsis can be easily complemented by other members of the JAZ gene family3. Tremendous efforts have been made to dissect the redundant function of JAZ genes in jasmonate signaling. Genetic studies using Arabidopsis mutant lines lacking multiple JAZ genes (jaz1/3/4/9/10 quintuple mutant as jazQ, jaz1/2/3/4/5/6/7/9/10/13 decuple mutant as jazD, jaz1/2/3/4/5/6/7/8/9/10/13 as jazU, and recently reported jaz1/2/3/4/5/6/7/9/10/11/12 undecuple mutant) are pioneering works to understand the function of JAZ family genes9,10,11,12. Recently, genetic studies using the model liverwort Marchantia polymorpha L have attracted attention as an alternative approach to analyze the jasmonate signaling because no genetic redundancy is observed in the key JA signaling pathway components, COI1, JAZ, and MYC13,14,15,16,17. In addition, chemical genetics is a powerful and promising alternative. Chemical genetics approaches have been successful in analyzing genetically redundant signaling in other hormonal pathways, such as those of abscisic acid and strigolactones, in which specific hormonal agonists activated distinctive receptor subtypes triggering specific responses18,19,20,21,22,23,24. Transient inhibition/degradation of a single JAZ protein in a fully developed plant by using chemicals may therefore disclose the specific function of a particular JAZ gene. Thus, the chemical genetic approach can complement the genetic approach25,26.

Here we report that the analysis of the chemical library (Fig. 1b) harboring all the stable stereochemical isomers of JA-Ile mimic, coronatine (COR, 2, Fig. 1a), provided the agonists of selective affinity among the COI1-JAZ co-receptor pairs in two plant species, A. thaliana and Solanum lycopersicum. Significantly, the stereoisomer 2e (Fig. 1b) caused protein degradation of the JAZ9 repressor to activate JAZ9-dependent jasmonate signaling in A. thaliana. The use of 2e enabled the analysis of a unique gene expression pattern governed by a single JAZ repressor, JAZ9, among the genetically redundant JAZ family genes.

Results

Chemical synthesis of all possible stereochemical isomers of coronatine

The use of a focused chemical library of 2-stereoisomers (Fig. 1b) can be a promising approach to look for novel, specific ligands of COI1-JAZ co-receptor pairs. Stereochemical isomers of 2 have the same size exclusion volume as naturally occurring 2 and slightly different 3D orientation of functional groups in a molecule, thus they can be accommodated in the ligand-binding pocket of COI1 with modulation of interaction network to each co-receptor pair leading to selectivity among JAZs (Fig. 1c).

The chemical library of all the stable 2-stereoisomers was constructed by coupling stereoisomers of coronafacic acid (CFA; 3) and coronamic acid (CMA; 4) (Fig. 1a). Enough amount of the stable stereoisomers of 3 was achieved by gram-scale racemic CFA synthesis and subsequent optical resolution, epimerization, and separation of diastereomers (Supplementary Fig. 1a)27,28,29. The optical purities of four CFAs (3-3c) were determined by chiral HPLC analyses30. Four CFAs (3-3c) were coupled with four CMAs (4-4c) which were synthesized by the reported method31, and subsequent deprotection provided all the stable 16 stereochemical isomers of coronatine (2 and 2a-2o in Fig. 1a, b, and Supplementary Figs. 1b and 2).

Stereochemical isomers demonstrated unique selectivity for COI1-JAZ co-receptor pairs of two plant species, Arabidopsis thaliana and Solanum lycopersicum

We examined the affinities of 16 stereochemical isomers of 2 with COI1-JAZ co-receptors of two plant species, Arabidopsis thaliana (encoding 13 JAZ proteins), as a representative model dicot, and Solanum lycopersicum (encoding 13 JAZ proteins), as a model plant belonging to the Solanaceae family (Supplementary Fig. 3). Chemical screening using in vitro fluorescence anisotropy (FA) assay was performed for all of the functional combinations of COI1-JAZ co-receptor pairs8,32.

For Arabidopsis COI1-JAZ pairs, in vitro FA assay was carried out for all of the functional combinations of AtCOI1-AtJAZ1/2/3/4/5/6/9/10/11/12 (Supplementary Fig. 3)33. AtCOI1-AtJAZ7/8/13 pairs were excluded in all of the following assays because AtJAZ7/8/13 lack the short conserved canonical degron sequence and thus cannot form the AtCOI1-1/2-AtJAZ complex34,35. Instead of full-length AtJAZ proteins, we used the fluorescein-labeled AtJAZ degron short peptides (Fl-AtJAZPs), including the ligand-interacting domain of full-length AtJAZ proteins33,36. FA assay provided the binding affinity (Kd) of each stereochemical isomer of 2 against AtCOI1-AtJAZ pairs (Fig. 2a, and Supplementary Figs. 4 and 5). The results of the FA assay were summarized as an affinity’heatmap’ profile of 16 stereochemical isomers (Fig. 2a). Based on their affinity, stereochemical isomers were categorized into four groups; no affinity (2a, 2b, 2g, 2j, 2k, 2m, 2n, and 2o), non-selective affinity (2, 2d, 2h, 2i, 2l), moderately selective affinity (2c), and highly selective affinity (2e, 2f) to some of the AtCOI1-AtJAZ combinations (Fig. 2a and Supplementary Fig. 5). In vitro pull-down assays using Fl-AtJAZP and GST-AtCOI1 protein confirmed the highly selective affinity of 2e and 2f (Supplementary Fig. 6). Surprisingly, 2e has a selective affinity to the AtCOI1-AtJAZ9 pair in FA and pull-down assays. Docking and subsequent in silico molecular dynamics (MD) simulations demonstrated that 2e, as well as 2, can be accommodated in the binding pocket of COI1-InsP8-JAZ9 complex (Supplementary Fig. 7). In this MD simulation, we used inositol octakisphosphate (InsP8) instead of inositol pentakisphosphate (InsP5) because previous reports showed that InsP8 also formed stable complex with COI1-2-JAZ instead of phosphates (PO4)37,38. In planta AtJAZ9-selectivity of 2e was tested by using a GUS reporter assay using A. thaliana reporter lines, 35S: JAZ1-GUS, 35S: JAZ9-GUS, 35S: JAZ10-GUS, and 35S: JAZ12-GUS (Fig. 2b and Supplementary Fig. 8). All JAZ-GUS fusion proteins were degraded in 2-treated seedlings. In contrast, 2e caused protein degradation in the AtJAZ9-GUS reporter line, whereas no degradation was observed in the AtJAZ1-GUS and AtJAZ10-GUS reporter lines. In addition, 2e-treatment caused degradation of JAZ9-GFP protein in 35S: JAZ9-GFP, but not of JAZ1-GFP protein in 35S: JAZ1-GFP (Supplementary Fig. 9). These results confirmed that 2e-treatment caused the selective ‘protein knock-out (degradation)’ of JAZ9 protein in A. thaliana.

Fig. 2: Affinity heatmaps of coronatine stereoisomers on COI1-JAZ co-receptors of A. thaliana and S. lycopersicum.
figure 2

a Heatmap analyses of Kd value of all stereoisomers for functional AtCOI1-AtJAZ pairs. White means “+not determined” due to a low increase of FA (Supplementary Fig. 4). The color bar represents the Kd value of each stereoisomer. Kd value of each stereoisomer which increased FA value at 1 µM was calculated by dose-response. b Evaluation of GUS activity in the roots of 4-day-old 35S: JAZ1-GUS, 35S-JAZ9-GUS, and 35S: JAZ10-GUS seedlings. Images of each seedling stained with 5-bromo-4-chloro-3-indolyl glucuronide were pretreated for 2 h with or without ligand (1 μM). Similar results were obtained in three independent experiments. Scale bar, 1 mm. c Heatmap analyses of Kd values of all stereoisomers for SlCOI1-SlJAZ pairs. White means “not determined” due to a low increase of FA (Supplementary Fig. 10). The color bar represents the Kd value of each stereoisomer. Kd value of each stereoisomer which increased FA value at 1 µM was calculated by dose-response curves.

For SlCOI1-SlJAZ pairs of S. lycopersicum, in vitro FA assay was also carried out for SlCOI1-SlJAZ1-11/13 (Fig. 2c). Phylogenetic characterization and functional annotation of SlJAZs were reported previously (Supplementary Fig. 3)32. Based on this result, we reported the affinity between 1/2 and SlCOI1-SlJAZ pairs using GST-SlCOI1 and fluorescein-labeled SlJAZ degron short peptides (Fl-SlJAZPs)39. According to this previous report, SlJAZ12 was excluded in the FA assay because it does not have a conserved JAZ degron sequence32. The stereochemical library showed distinct selectivity against SlCOI1-SlJAZ pairs (Fig. 2c and Supplementary Figs. 10 and 11), as with the case of AtCOI1-AtJAZ pairs. Among them, remarkable selectivity was found for the stereochemical isomers 2e and 2f. 2e has selectivity against SlCOI1-SlJAZ5/6/8, and 2f has selectivity against SlCOI1-SlJAZ5/6/8/13. Unfortunately, we cannot examine the affinity of stereoisomers with the SlCOI1-SlJAZ7P pair in the FA assay because the severe non-specific binding was observed in the absence of ligand. Interestingly, isomers 2e and 2f also showed selectivity against AtCOI1-AtJAZ pairs (Fig. 2a).

Through these examinations, we found two stereochemical isomers, 2e and 2f, showing distinct selectivity for COI1-JAZ co-receptor pairs of two plant species. These results demonstrated that the stereochemical library of 2 is a promising chemical tool for discovering ligands with selective affinity for some subtypes of the genetically redundant COI-JAZ co-receptor family in higher plants.

A stereochemical isomer induced defenses without growth inhibition through selective activation of jasmonate signaling in A. thaliana

To examine the power of stereochemical isomers to dissect the jasmonate signaling pathway, we focused on 2e, with selective affinity to the AtCOI1-AtJAZ9 pair (Fig. 2a). We explored the biological effect of 2e and its mode of action on A. thaliana.

2e showed no growth inhibition on 4-week-old A. thaliana, whereas 2 caused severe growth inhibition (Fig. 3a, b, and Supplementary Fig. 12). In addition, no growth inhibition was observed on root by the addition of 2e (Supplementary Fig. 13). However, 2e, as well as 2, moderately activated resistance response against the infection of fungal pathogen Alternaria brassicicola compared to the mock-treatment (Fig. 3c, d, and Supplementary Fig. 14)36. This 2e-mediated defense trait was impaired in jaz9-1, in which other members of the JAZ family protein likely complemented the original function of JAZ9 protein in presence of non-selective ligand, and 2e cannot cause their degradation (Fig. 3e and Supplementary Fig. 15) This result was in clear contrast to 2, which caused a close defense trait in both Col-0 and jaz9-1 (Fig. 3c–e and Supplementary Fig. 15). These effects of 2e on A. thaliana strongly suggested that 2e can induce a resistance response by activating the AtJAZ9-mediated partial jasmonate signaling. Reverse transcription-qPCR (RT-qPCR) analyses demonstrated that 2e-treatment in 4-week-old A. thaliana (without pathogen infection) caused the upregulation of defense-related genes, such as ORA59, ERF1, PDF1.2, and HEL, without affecting the expression of JA marker genes, such as MYC2 and VSP1/2 (Fig. 3f). In the downstream of JA signaling, ORA59, ERF1, PDF1.2, and HEL belong to ERF/ORA branch, and MYC2 and VSP1/2 belong to the MYC2-branch (Supplementary Fig. 16). Thus, 2e selectively upregulated the ERF/ORA branch of JA-signaling without affecting MYC-branch. In addition, 2e-mediated expression of ORA59 and PDF1.2 was impaired in coi1-1, which confirmed the COI1-dependent mode-of-action of 2e (Supplementary Fig. 17).

Fig. 3: In planta analyses of stereoisomers in Arabidopsis thaliana.
figure 3

a, b Growth inhibition assay using ligand-treated 4-week-old A. thaliana. a Representative images of the 2- (50 µM) and 2e-treated (50 µM) plants. Similar results were obtained in three independent experiments (Supplementary Fig. 12). Scale bar: 1.0 cm. b Quantification of fresh weight of aerial part of the ligand-treated plants (n = 8). Significant differences were evaluated by one-way ANOVA/Tukey HSD post hoc test (p < 0.05). Values are mean ± SE. ce Fungal infection assay using c, d 4-week-old Col-0 and e 4-week-old jaz9-1. Plants were treated with mock solution, 2 or 2e (50 μM) and infected with A. brassicicola (n = 7). c Representative images of leaves infected with A. brassicicola are shown in (d) and (e). Quantification of fungal spores was undertaken 5 days after inoculation; the results are depicted using box plots; horizontal lines are medians, boxes show the interquartile range, and error bars show the entire data range. Significant differences were evaluated by one-way ANOVA/Tukey HSD post hoc test (p < 0.05). Experiments were repeated three times with similar results. Values are mean ± SE. f Gene expression analysis in 4-week-old A. thaliana with or without compound treatment (50 µM, n = 4) without pathogen inoculation. Each sample was flash-frozen after 8 h from definitive treatment. Similar results were obtained in three independent experiments. Significant differences were evaluated by one-way ANOVA/Tukey HSD post hoc test (p < 0.05). Values are mean ± SE.

Chemical and genetic AtJAZ9 knock-out showed distinct gene expression profiles

Previous reverse genetic studies demonstrated that jaz9-1 has no significant phenotype regarding root-growth inhibition and defense response10,11. In contrast, our 2e-induced protein knock-out of JAZ9 caused activation of the defense trait. To examine the distinct difference between chemically induced protein degradation or loss-of-function of JAZ9 by 2e and genetic knock-out in jaz9-1, RNA sequencing (RNAseq) analyses were carried out for a comprehensive comparison of gene expression profiles between 6-day-old seedlings of WT (Col-0) treated by 2/2e and jaz9 mutant (jaz9-1) treated by mock (Supplementary Data 1).

To analyze the entire gene expression profiles, we performed principal component analysis, Venn diagram analyses, and heatmap analyses of upregulated genes with >2.5-fold changes (Fig. 4a–c). In principal component analysis, no significant difference was observed between gene expression profiles of the mock-treated jaz9-1 and the mock-treated Col-0 (Fig. 4a), suggesting that in the jaz9-1 mutant, most functions of JAZ9 as a transcriptional repressor would be complemented by other partially redundant members of the JAZ family. In contrast, a clear difference was found between the gene expression profiles of 2e-treated Col-0 and mock-treated jaz9-1 (Fig. 4a). This result showed a distinct difference in gene expression profiles between genetic knock-out and chemically induced protein degradation (or loss-of-function).

Fig. 4: Gene expression analyses of 2- and 2e-treated Col-0 and jaz9-1 seedlings.
figure 4

a Principal component analysis of upregulated genes in each compound-treated 6-day-old Arabidopsis seedlings by RNAseq analyses (fold change >2.5, FDR < 0.1). Black, blue, red, and green dots showed three independent replicates of each treatment. b Venn diagram (FDR < 0.1) indicates the number of upregulated genes at least 2.5-fold change in response to different treatments. c Heatmap illustrating changes in gene expression of jasmonate response to different treatments (FDR < 0.1). Each GO ID is as follows: Regulation of JA mediated signaling pathway: 2000022, Response to JA: 0009753, Defense response: 0006952, Metabolic process: 0008152. d Venn diagram (FDR < 0.1) indicating the number of downregulated genes at least 2.5-fold change in response to different treatments and GO enrichment analysis. e Gene expression analyses of cell wall organization-related genes by RT-qPCR in Col-0 and coi1-1 seedlings (n = 4) with or without compound treatment (1 µM) for 8 h. Similar results were obtained in three independent experiments. Significant differences were evaluated by one-way ANOVA/Tukey HSD post hoc test (p < 0.05). Values are mean ± SE. f Venn diagram (FDR < 0.1) indicating the number of upregulated genes at least 2.5-fold change in response to different treatments in Col-0 and jaz9-1. g RT-qPCR analyses in Col-0 and jaz9-1 seedling (n = 4) with or without compound treatment (1 µM). Each relative expression fold was calculated using an average expression level in Mock- or 2e-treated Col-0 or jaz9-1 seedlings. Similar results were obtained in three independent experiments. Significant differences were evaluated by Student’s t-test (**p < 0.01, *p < 0.05; ns, not significant). Values are mean ± SE.

We further examined the unique gene expression caused by 2e. Detailed analyses of the gene expression profiles shown in the Venn diagram and heatmap also showed that 2e caused a part of 2-mediated gene expression: 2 upregulated expressions of 950 genes, of which 2e upregulated 103 genes (Fig. 4b). A significant difference was observed between the results of 2e-induced chemical degradation (or knock-out) and genetic knock-out of JAZ9: 91 of 103 gene upregulations caused by 2e were not observed in jaz9-1 (Fig. 4b and Supplementary Fig. 18 as a list of 91 genes). Heatmap analysis (Fig. 4c) showed that 2 upregulated major JA signaling-related genes, such as JAZ, JA-responsive genes, defense-related genes, and metabolism-related genes, whereas 2e primarily upregulated expressions of ethylene-mediated defense-related genes (see also Supplementary Figs. 19 and 20).

To our surprise, a remarkable difference in expression profiles was observed in downregulated genes: 2e caused downregulation of gene expression only for eight genes, whereas 2 caused downregulation for 614 genes (Fig. 4d and Supplementary Fig. 19). GO analysis of 2-repressed genes showed that GO terms related to cell wall organization/biogenesis were significantly repressed (Fig. 4d). RT-qPCR analyses showed that 2 repressed the gene expression of cell wall organization/biogenesis-related genes such as EXLA2, XTH19, and XTH23, which were previously reported as essential genes for plant root growth40,41, in a COI1-dependent manner, but 2e did not affect their expression (Fig. 4e). The reason why 2e did not cause growth inhibition could be explained by the lack of gene downregulation of these cell-wall-related genes.

A stereoisomer caused AtJAZ9-dependent gene expressions

The fact that mode of action of 2e is dependent mainly on JAZ9 was confirmed by a comparison of RNAseq data on 2e-treated Col-0 and 2e-treated jaz9-1 (Fig. 4f). In jaz9-1, 2e activated the expression of 100 genes, and 52 of 100 genes were upregulated in 2e-treated Col-0, of which three were upregulated in mock-treated jaz9-1 (Fig. 4f). Thus, expressions of 49 genes (Supplementary Fig. 20a as a list of 49 genes) appear to be induced independently of JAZ9 in the 2e-treated jaz9-1. The 49 genes include JA-marker genes such as JAZ1, MYC2, and AOS, and defense-related genes belonging to the ERF/ORA branch, such as ORA59, ERF1, PDF1.3, and PDF1.2 (Supplementary Fig. 21a). However, the RT-qPCR analysis showed that the expression levels of most of the 49 genes found in 2e-treated jaz9-1 were not statistically significant compared to those in 2e-treated Col-0, while the expression levels of ERF/ORA-genes which are upregulated by 2e in Col-0 were significantly suppressed in 2e-treated jaz9-1 (Fig. 4g and Supplementary Fig. 21b). These results showed that 2e upregulated the expression of defense-related genes in a mostly JAZ9-dependent manner. This result agrees with the result in Fig. 3e, in which 2e caused no remarkable resistance response against the pathogen in the jaz9-1 mutant (Fig. 3e and Supplementary Fig. 15). To test if 2e activity may also require JAZ10, as previously described for the JA-Ile selective agonist N-coronafacyl ent-coronamic acid O-phenyloxime (NOPh)36, the requirement for JAZ10 in 2e-induced responses was analyzed. 2e did not affect the growth and upregulated defense-related genes belonging to the ERF/ORA branch in the jaz10-1 mutant (Supplementary Fig. 22). Moreover, no growth inhibition and expressions of marker genes belonging to ERF/ORA branch were observed in 2e-treated jaz12-1 (Supplementary Fig. 23). These results rule out the role of JAZ10 in 2e-mediated regulation and they suggested that 2e upregulated the gene expressions mostly in a JAZ9-dependent manner.

We successfully developed the chemical tool 2e which mediates JAZ9-selective degradation in planta. This 2e-mediated activation leads to the expression of genes normally repressed by JAZ9 in A. thaliana. Examination of differential gene expression induced by JAZ9 protein degradation (or loss-of-function) revealed the unique function of JAZ9, which did not affect the downregulation of genes, including cell-wall-related ones which play important roles in the plant root growth, and caused upregulation of defense-related genes through the EIN3/EIL1-ERF/ORA59 pathway40,41,42,43. The results highlight the power of a stereochemical library of 2 to screen the JAZ-selective activator of jasmonate signaling in other plant species.

Discussion

The chemical genetic approach is a promising way to dissect the plant hormone signaling of high genetic redundancy25,26. Previous studies have reported success in abscisic acid and strigolactone using small molecules screened from chemical libraries to elucidate the function of receptor subtypes belonging to genetically redundant receptor families18,23,24. Chemical libraries composed of planar molecules have been successfully used to screen hit compounds in drug discovery and plant chemical biology26,44,45. Chini et al. recently discovered a general antagonist of AtCOI1-AtJAZ using a chemical library46. To the best of our knowledge, this was the only successful example of the chemical library screening of COI1-JAZ co-receptor agonist/antagonist. By contrast, we have screened a targeted library, based on stereoisomers of 2, for highly subtype-selective ligands of 10 AtCOI1-AtJAZ and 11 SlCOI1-SlJAZ co-receptor pairs. 2e/2f, in addition to high selectivity for AtJAZ9, also showed selective affinity with SlCOI1-SlJAZ5/6/8, which are the closest tomato homologs of AtJAZ9 belonging to Group V (Supplementary Fig. 3). The chemical biology community focused on the conventional chemical libraries being composed of planar molecules, whereas we envisioned a chemical library composed of the stereoisomers of three-dimensional structures. These two approaches are complementary, and we demonstrated that stereochemical libraries could provide hit compounds that cannot be obtained from a chemical library of planar molecules.

In the present study, we have successfully identified AtCOI1-AtJAZ9-selective agonist 2e from the stereochemical library of 2. Taking advantage of the JAZ9-selective mode-of-action of 2e, we successfully find that JAZ9 plays a unique function in jasmonate signaling: JAZ9 does not participate in the gene downregulation caused by 2. Among the genes specifically repressed by 2, but not by 2e, we focused on genes related to cell wall formation and biosynthesis, such as EXLA2, XTH19, and XTH23 (Fig. 4e). Previous reports suggested that these genes play important roles in root growth: overexpression lines of EXLA2 and double knock-out mutant of XTH19 and XTH23, which affect the root growth and lateral root formation40,41,47. Thus, the current results answered the question, why 2e did not affect the growth inhibition (Fig. 3a, b). In addition, using 2e, we demonstrated that pathogen-resistance through ERF/ORA-branch JA-signaling is controlled by JAZ9-meditated jasmonate signaling. These findings agree with our previous results, in which NOPh has a selective affinity with AtCOI1-AtJAZ9/10 pairs to enhance resistance response against pathogen A. brassicicola without causing growth inhibition36. NOPh selectively activates ERF/ORA-branch to unpair growth and defense through COI1/JAZ9-EIN3/EIL1-ERF/ORA59 signaling pathway42. In contrast to NOPh, 2e mostly caused the selective degradation of single JAZ repressor, thus enabling analysis of gene expression profiles regulated by a single JAZ. 2e is specific mainly for JAZ9, although RNAseq suggest that it might target other JAZ at least partially and weakly. However, since the 2e-mediated expression level of JAZ9-independent JA-responsive genes is low (Fig. 4g and Supplementary Fig. 21b), we concluded that 2e induces gene expression mainly in a JAZ9-selective manner. 2e effectively reduces the redundancy of the COI1-JAZ co-receptors and it led to the identification of specific genes most likely regulated by the JAZ9 repressor (Fig. 4b and Supplementary Fig. 18 as a list of genes).

Previous achievements demonstrated that the limitations of the genetic approach, such as redundancy or lethality of essential genes, could be overcome by chemical genetics48,49. We hypothesized that transient inhibition/degradation of a single JAZ protein in fully developed plants may disclose the function of a single JAZ gene within the redundant JAZ gene family. Our results demonstrated the uniqueness of “chemical mutants” in which gene redundancy does not seem to compensate the transiently degraded protein. We showed that this chemical genetic approach is complementary to genetic approaches25,26. The current result paves the way for the rational design of JAZ-subtype selective agonists in crops and edible plants, which will elucidate the role of each JAZ gene in the highly redundant JAZ family of these plants.

Methods

Chemical syntheses

Details of chemical syntheses are described in Supporting Information.

Fluorescence anisotropy measurement

Fluorescence anisotropy titration experiments were performed at 25 °C in 50 µL of Tris buffer (50 mM Tris-HCl, pH 7.8, 100 mM NaCl, 10% Glycerol, 0.1% Tween20, 20 mM 2-mercaptoethanol (2-ME) and 1 µM inositol-1,2,4,5,6-pentakisphosphate, IP5) containing 100 nM COI1 and 100 nM Fl-JAZPs using a 96-well clear-bottom plate unless otherwise noted. Each ligand was added to the solution and incubated at 25 °C until the r-value was not-fluctuated, and anisotropy intensities were measured (λexem = 485 nm/516 nm). r values were calculated using the following equation: r = (IVV – G × IVH)/(IVV + 2G × IVH), IVV and IVH are the fluorescence intensities observed through polarizers parallel and perpendicular to the polarization of the exciting light, respectively, and G is a correction factor to account for instrumental differences in detecting emitted compounds. Fluorescence anisotropy was recorded on an EnVision 2105 (PerkinElmer, USA).

Pull-down assay of AtCOI1-AtJAZ peptide

Purified GST-COI1 (5 nM) with ASK1 and each compound (5 µM) are dissolved in 350 µL of Tris buffer and added Fl-JAZPs (10 nM). After incubation on ice for 1 h, the mixture was combined with an anti-Fluorescein antibody (GENETEX, GTX26644, 0.2 µL) and incubated at 4 °C for 5 h with rotation. After incubation, the mixture was combined with Protein G beads (Bio-Rad, 1614023, 10 µL in 50% Tris buffer slurry) and incubated at 4 °C for 1 h with rotation. After incubation, the sample was washed five times with PBS-T. The washed beads were resuspended in 35 μL of SDS-PAGE loading buffer containing dithiothreitol (DTT, 100 mM). After heating for 10 min at 60 °C, the samples were subjected to SDS-PAGE and analyzed by western blotting. SDS-PAGE and western blotting were carried out using a Mini-Protean III electrophoresis apparatus (Bio-Rad, Hercules, CA). Chemiluminescence was observed on Amersham Imager 680 (Cytiva, USA).

Plant materials and growth conditions

Arabidopsis thaliana Col-0 is the genetic background of wild-type and mutant lines used in this study. For biological assay using seedlings, surface-sterilized seeds were grown under a 16-h-light (75 μmol m−1 s−1; cool-white fluorescent light)/8-h-dark cycle at 22 °C on a half-strength Murashige-Skoog (MS) plate (2% sucrose, 0.4% gellan gum) or liquid medium (0.5% sucrose) in a Biotron LPH-240SP growth chamber (Nippon Medical & Chemical Instruments Co., Ltd., Osaka, Japan) after vernalization in the dark at 4 °C for 2 days. For biological assay using adult plants, seeds were grown directly in soil or 1:1 soil/vermiculite under a 12-h-light (75 µmol m−1 s−1, cool-white fluorescent light)/12-h-dark cycle at 22 °C.

For the 35S-JAZ1-GFP line, Amplified 35S-JAZ3-GFP from pGWB5-JAZ25,50 was inserted into the SalI-XhoI site of pET-32a (Invitrogen). pET-32a-35S-JAZ1-GFP was created by an in-fusion reaction between amplified ORF of JAZ1 and pET-32a-35S-GFP amplified by inverse PCR of pET-32a-35S-JAZ3-GFP. The adapter (EcoRI-SalI-XhoI) was inserted into the HindIII-SacI site of pGWB5-JAZ2. And then, 35S-JAZ1-GFP in prepared pET-32a-35S-JAZ1-GFP was introduced into the SalI-XhoI site of the adapted pGWB5. For the 35S-JAZ9-GFP line, amplified JAZ9 was introduced into pET-32a-35S-GFP by the in-fusion reaction. Amplified 35S-JAZ9-GFP was introduced into the SalI-XhoI site of the adapted pGWB5.

Agrobacterium GV3101, containing these constructs, was used to transform Col-0 plants by floral dipping (Clough & Bent, 1998, conducted by INPLANTA INNOVATIONS Inc., Japan). Transgenic lines were checked for Hygromycin resistance, and lines carrying one insertion were driven to homozygous and used for further experiments.

The mutant lines have been previously described: coi1-1, jaz9-1 (SALK_004872), and jaz10-1 (SAIL_92_D08)51,52,53,54. The marker lines 35S: JAZ1-GUS, 35S: JAZ9-GUS, and 35S: JAZ10-GUS were previously described4,55.

GUS-reporter assay

Four-day-old seedlings on a half-strength MS plate were transferred to a half-strength MS liquid medium in the presence or absence of each compound. After treatment for 2 h, each seedling was fixed with 90% aqueous acetone for 20 min on ice. Each seedling immersed in GUS staining buffer (50 mM phosphate buffer; pH 7.0; 1 mM K4Fe(CN)6; 1 mM K3Fe(CN)6; 0.1% Triton X-100; 2 mM 5-bromo-4-chloro-3-indolyl β-D-glucuronic acid (X-Gluc, Biomedical Science, Japan)). After staining at 37 °C for two days, each seedling was pictured with an E-520 digital camera (Olympus Corp., Japan).

To quantify the GUS activity, 9–11 roots of ligands-treated seedlings were flash-frozen, and homogenized with extraction buffer (50 µL; 50 mM phosphate buffer, pH 7, 10 mM EDTA, 10 mM 2-mercaptoethanol, 0.1% sarcosyl (N-lauroylsarcosine sodium salt; >94%, Sigma-Aldrich) and 0.1% Triton X-100). After centrifugation (15,000 × g, 4 °C, 10 min), each supernatant was collected. 10 µL extracts were incubated with a 40 µL extraction buffer containing 1 mM 4-MUG (methylumbelliferyl-β-D-glucuronide hydrate; ≥98%, Sigma-Aldrich). 10 µL samples were collected at t = 0 and t = 1 h. Then the reaction was quenched with 90 µL of 0.2 M Na2CO3. Fluorescence (Ex./Em. = 365/460 nm) was detected using the spectrophotometer Infinite M200Pro (TECAN, Switzerland). Each fluorescence intensity was normalized by the amounts of proteins determined by the Bradford assay (595 nm).

Chemical treatment for biological assay using adult plants

For growth inhibition assay, 2 µL of an aqueous solution containing each compound were dropped on the rosette center of a leaf of 2-week-old plants. After incubation for 6 days, each compound was applied again similarly. Pictures of plants were taken after 6 days of definitive treatment. The fresh weight of the aerial part was measured, and then metabolite analyses were performed in the same way as seedlings.

For gene expression analysis, leaves of 3-week-old plants were treated with each compound. After incubation for 6 days, each compound was applied again similarly. After treatment for 8 h, each leaf was flash-frozen, and then total RNA extraction and RT-qPCR were performed in the same way as seedlings.

Root growth inhibition analysis using seedlings

Two-day-old seedlings on half-strength MS plate were transferred on half-strength MS plate in the presence or absence of each compound. After treatment for 4 days, each seedling was pictured with an E-520 digital camera (Olympus Corp., Japan), and the root length of these seedlings was measured using Image J 1.45S software (http://imagej.net/Welcome).

Fungal infection analyses

Seeds were grown directly in the soil. As previously described, three leaves (leaf numbers 4, 5, and 6) of 3-week-old plants were inoculated with A. brassicicola suspension of 5 × 106 spores/ml PDB (Difco)56. Compound was applied twice at 5 ho before pathogen inoculation and at the same time of pathogen inoculation. Pictures of infected plants were taken 5 days after inoculation. Quantification of spores was carried out in a hemocytometer under a light microscope (Leica). Five to eight inoculated leaves of at least six different plants were pooled for each biological sample, and four to six independent biological replicates were measured for each treatment. Results were analyzed by one-way ANOVA/Tukey HSD post hoc test (p < 0.05). The infection assay was repeated three times with similar results. Data are shown as mean ± SEM.

RNAseq analysis

Six-day-old seedlings (5–8 seedlings/sample) grown on half-strength MS liquid medium were treated with half-strength MS liquid medium in the presence or absence of each compound. After treatment for 8 h, each sample was flash-frozen, and then total RNA was extracted using RNeasy Mini Kit (Qiagen Co. Ltd., Germany). RNAseq libraries were prepared using the low-cost and easy RNA-SEQ (Lasy-Seq) method described previously57. Paired-end sequencing was performed using HiSeq X Ten (Illumina, San Diego, CA). RNA-seq analyses were performed with three biological replications, and the data was analyzed using DESeq2.

Low-quality reads and adapter were trimmed using trimmomatic version 0.39 (http://www.usadellab.org/cms/?page=trimmomatic) with ‘ILLUMINACLIP:TruSeq3-SE.fa:2:30:10 LEADING:3 TRAILING:3 SLIDINGWINDOW:4:15 MINLEN:36’ options. HISAT2 (http://daehwankimlab.github.io/hisat2/) version 2.2.1 was used to align the reads for reference genome (https://phytozome.jgi.doe.gov/pz/portal.html#!info?alias=Org_Athaliana) with ‘–max-intronlen 5000’ option. Aligned reads within gene models were counted using featureCounts version 2.0.1 (http://subread.sourceforge.net/) with ‘--fracOverlap 0.5 -O -t gene -g ID -s 1 --primary’ options. Differentially expressed genes were identified using R version 4.0.4 (https://www.r-project.org/) and DESeq2 version 1.30.1 (https://bioconductor.org/packages/release/bioc/html/DESeq2.html) package.

Heat map analyses were performed with an online tool heatmapper58. The normalized log2 values were used to compare the transcriptomic changes using MapMan 3.6.0RC1 (Supplementary Fig. 19)59. Gene ontology (GO) enrichment analyses were carried out using the TAIR website (https://www.arabidopsis.org/tools/go_term_enrichment.jsp).

RT-qPCR of Arabidopsis thaliana seedlings

six-day-old seedlings (5–8 seedlings/sample) grown on half-strength MS liquid medium were treated with half-strength MS liquid medium in the presence or absence of each compound. After treatment for 2 or 8 h, each sample was flash-frozen, and then total RNA was extracted using ISOGEN (NIPPON GENE, Japan). Then first-strand cDNA was gained with ReverTra Ace reverse transcriptase (Toyobo, Japan) with oligo-dT primers. A StepOnePlus Real-Time PCR System (Life Technologies, USA) was used for quantitative PCR (all primers sequences for RT-qPCR in Supplementary Data 2). Polyubiquitin 10 was used as a reference gene. To select homozygous coi1-1, grown 3-day-old heterozygous coi1-1 on a half-strength MS plate containing 10 µM JA was transferred to a half-strength MS liquid medium. After 3 days of culturing, each seedling was treated with each compound.

Statistics and reproducibility

All the experiments were performed three or more times with independent samples and reproduced with similar results. Quantitative values are expressed as the mean values ± S.D. or S.E. Student’s t-test was performed between two groups using Microsoft Office Excel. Analysis of variance (ANOVA) was performed among more than two groups using the CoStat (version 6.400) software.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.