Arabidopsis MYB24 Regulates Jasmonate-Mediated Stamen Development

The phytohormone jasmonates (JAs) regulate various defense responses and diverse developmental processes including stamen development and fertility. Previous studies showed that JA induces CORONATINE INSENSITIVE 1-mediated degradation of JA ZIM-domain (JAZ) proteins, and activates the MYB transcription factors (such as MYB21 and MYB24) to regulate stamen development. In this study, we further uncover the mechanism underlying how MYB24 interacts with JAZs to control JA-regulated stamen development. We show that N-terminus of MYB21/24 interacts with 10 out of 12 JAZ proteins while both N-terminus and C-terminus of MYB24 are involved in dimerization of MYB21 and MYB24. Interestingly, male sterility of the JA-deficient mutant opr3 can be rescued by suitable level of the MYB24 overexpression but not by excessive high level of MYB24. Surprisingly, overexpression of MYB24NT, but not MYB24CT, could cause male sterility. These results provide new insights on MYB factors in JA-regulated stamen development.

We previously showed that JAZ1/8/11 interact with MYB21/MYB24 in yeast and plants . In this study, we further showed that MYB21 and MYB24 interact with most JAZs via their N-terminal R2R3 domains, and both the N-terminus and C-terminus of MYB24 mediate the dimeric interactions of MYB21 and MYB24. Proper overexpression of MYB24 partially restores male fertility of opr3. Overexpression of N-terminus of MYB24, but not C-terminus, causes male sterility in wild-type. Furthermore, young flower buds from myb21 myb24 myb57 accumulate more jasmonic acid than that of wild-type.

Plant Materials and Growth Conditions
The Arabidopsis mutants opr3 (Stintzi and Browse, 2000) and myb21 myb24 myb57 (Cheng et al., 2009) were described previously. Arabidopsis thaliana seeds were disinfected, germinated on Murashige and Skoog (MS) medium, stored at 4 • C for 3 days, and transferred to a growth room for another 7 days before being transferred to soil under a 16 h (22-24 • C)/8 h (17-19 • C) light/dark photoperiod. Nicotiana benthamiana seeds were sown in soil and grown under a 16 h (26 • C)/8 h (22 • C) light/dark photoperiod.

Yeast Two-Hybrid Assay
The full lengths of coding regions of MYB21, MYB24, 12 JAZs, and truncated domains of MYB24 were individually fused with the activation domain (AD) in pB42AD, or the DNA binding domain (BD) in pLexA. The primer pairs used for vector construction are listed in Supplementary Table 1. Yeast transformation and protein-protein interaction assays were performed according to the Matchmaker LexA Two-Hybrid System (Clontech) as described previously  Firefly Luciferase (LUC) Complementation Imaging (LCI) Assay JAZ5, MYB21, MYB24, MYB24NT, and MYB24CT were individually inserted into pCAMBIA-nLUC or pCAMBIA-cLUC for fusion with the N-terminal half of LUC (nLUC) or C-terminal half of LUC (cLUC). The primers used to construct the LCI vector are listed in Supplementary Table 1. Agrobacterium tumefaciens cells containing the indicated plasmids were co-infiltrated into N. benthamiana leaves and LUC activity was detected as described previously .

Generation of Transgenic Plants
To obtain transgenic plants overexpressing MYB24, the coding sequence of MYB24 was amplified and cloned using XbaI and SacI into pCAMBIA1301 under the control of the CaMV 35S promoter. The construct was transformed into OPR3/opr3 heterozygous plants using the Agrobacterium-mediated floral dip method. The primers used for vector construction are presented in Supplementary Table 1. pCAMBIA-MYB24NT-nLUC and pCAMBIA-MYB24CT-nLUC were transformed into Arabidopsis Col-0 wild-type to generate MYB24NT-and MYB24CToverexpressing plants.

Quantitative Real-time PCR
Young flower buds of plants were collected for total RNA extraction and subsequent reverse transcription. Quantitative real-time PCR was performed with RealMasterMix (SYBR Green I; Takara; Bio Inc., Otsu, Japan) using the ABI 7500 Real-Time PCR System (Applied Biosystems, Foster City, CA, United States), and ACTIN8 was used as an internal control. The primers used are listed in Supplementary  Table 1.

Flower Phenotype Analysis and Pollen Germination Assay
For flower phenotype analysis, flowers of each genotype at stage 13 were photographed under a microscope. Pollen germination assay was conducted as described previously . Pollen grains were germinated on pollen germination medium [1 mM MgSO 4 , 5 mM CaCl 2 , 5 mM KCl, 10% (w/v) sucrose, 0.01% boric acid, and 1.5% agar at pH 7.5], incubated for 10 h at 22 • C in the dark, and observed under a microscope.

Jasmonic Acid Measurement
Five hundred milligrams of young flower buds before floral stage 13 from 5-week-old wild-type and the myb21 myb24 myb57 mutant was harvested, and jasmonic acid was extracted and quantified as described previously using a liquid chromatography-tandem mass spectrometry system (Cheng et al., 2009).

MYB21 and MYB24 Interact with Multiple
JAZs via Their N-terminus MYB21 and MYB24 were divided into MYB21NT and MYB24NT harboring the R2R3 DNA BD, and MYB21CT and MYB24CT containing the C-terminal NYW G / S M / V DD I / L W S / P transcriptional activation motif (Figures 1A,B). Binding domain-fused MYB21CT and MYB24CT exhibited autoactivation, while MYB21NT and MYB24NT did not. Binding domain-fused MYB21NT and MYB24NT were used to detect interactions with AD-fused JAZ proteins. The results ( Figure 1C) showed that MYB21NT and MYB24NT interact with JAZ1, JAZ2, JAZ3, JAZ4, JAZ5, JAZ6, JAZ8, JAZ10, JAZ11, and JAZ12, but not with JAZ7 or JAZ9, demonstrating that MYB21 and MYB24 interact with multiple JAZs through their N-terminus.
We next performed a firefly LCI assay to test the interactions of JAZ5 with MYB21/24 in plant. JAZ5 was fused with nLUC, while MYB21 and MYB24 were fused with cLUC. The results showed that co-infiltration of JAZ5-nLUC/cLUC-MYB21 or JAZ5-nLUC/cLUC-MYB24 in N. benthamiana leaves resulted in strong LUC signals while the negative control did not (Figure 1D), suggesting that MYB21 and MYB24 interact with JAZ5 in plant.

MYB24 Overexpression Restores Stamen Development in opr3
We next examined whether MYB24 overexpression could escape from inhibition by multiple JAZs to rescue stamen development and fertility in the JA-deficient mutant opr3. As shown in Figure 2A, opr3 exhibited unelongated filaments, indehiscent anthers, and inviable pollen grains at floral stage 13. MYB24 expression decreased in young opr3 flower buds ( Figure 2D).  MYB24 overexpression in opr3 (opr3 MYB24OE37, with fivefold to sixfold of wild-type level) restored filament elongation, anther dehiscence, and pollen viability (Figure 2A). Further, the opr3 MYB24OE37 plants were able to set seeds (Figures 2B,C). These results suggest that overexpression of MYB24 could partially restore stamen development and fertility in opr3.

Excess Expression of MYB24 Cannot Rescue the Fertility of the opr3 Mutant
Previous studies showed that strong MYB24 overexpression inhibits stamen development (Yang et al., 2007;Song et al., 2011). We therefore tested whether excess expression of MYB24 could restore fertility in opr3. As shown in Figure 3, transgenic plants that expressed excessive amounts of MYB24 (∼75-fold of the wild-type level) were male sterile. Further, transgenic opr3 plants that expressed excessive amounts of MYB24 (∼70-fold of the wild-type level) were still male sterile, suggesting that excess expression of MYB24 could not restore stamen development and fertility in opr3 plants.

The N-terminus and C-terminus of MYB24 Are Involved in Dimeric Interactions
We next investigated the dimeric interactions of MYB21 and MYB24 in detail. As shown in Figure 4A, Y2H analysis showed that BD-fused MYB24NT interacted strongly with AD-fused MYB24NT, and weakly with MYB24CT. Further, LCI assay exhibited that the co-expression of MYB24NT-nLUC/cLUC-MYB21, MYB24CT-nLUC/cLUC-MYB21, MYB24NT-nLUC/ cLUC-MYB24, and MYB24CT-nLUC/cLUC-MYB24 resulted in strong LUC activity, while the negative controls did not (Figures 4B,C). These results demonstrate that both N-terminus

MYB24NT Overexpression Causes Male Sterility
We next examined whether the overexpression of MYB24NT and MYB24CT could dominantly repress stamen development and male fertility. MYB24NT overexpression (∼20-40-fold of the wild-type level) inhibited stamen development, including filament elongation, anther dehiscence, and male fertility (Figures 5A,B and Supplementary Figure 1A), while all MYB24CT transgenic lines showed no obvious influence on male fertility ( Figure 5C and Supplementary Figure 1B). These data indicate that overexpression of N-terminus of MYB24 could dominantly repress stamen development and fertility.
MYB57 in turn regulate JA biosynthesis in young flower buds (before floral stage 13). The jasmonic acid content of young myb21 myb24 myb57 mutant flower buds was approximately twofold of the wild-type level (Figure 6), suggesting that MYB21, MYB24, and MYB57 negatively regulate JA biosynthesis as part of a negative feedback loop.

DISCUSSION
Jasmonate ZIM-domain proteins serve as repressors that target specific transcription factors and control their downstream pathways to modulate distinct JA responses for coordinated regulation of development, growth, and defense (Song et al., 2014b;Chini et al., 2016;Gimenez-Ibanez et al., 2017;Major et al., 2017). Our previous study showed that JAZ1, JAZ8, and JAZ11 interact with MYB21 and MYB24 . In this study, we further showed that MYB21 and MYB24 act through N-terminus to interact with 10 out of 12 JAZs (Figure 1), suggesting that most JAZs may act through interfering the DNA binding function of MYB21/24 to attenuate their function in regulating stamen development. Excess expression of MYB24 is unable to restore the stamen development of opr3, whereas suitable overexpression of MYB24 could recover stamen development and male fertility (Figures 2, 3). Exploring the downstream pathways of MYB24 would help to understand that the suitable MYB24 expression level is essential for proper stamen development. Both the N-terminal DNA BD and C-terminal transcriptional activation motif mediate dimerization of MYB21 and MYB24 (Figure 4). Determination of crystal structure of MYB21/24 will help to further elucidate the interaction. Interestingly, overexpression of N-terminus of MYB24, but not the C-terminus of MYB24, attenuates stamen development and male fertility (Figure 5). MYB24NT-OE2 with the highest expression level of MYB24NT confers the most severe male sterility (Figures 5A,B and Supplementary Figure 1A), suggesting that the expression level of MYB24NT is correlated with male sterility. It remains to study whether MYB24NT affects the dimeric interaction of MYBs to affect stamen development.
We also found that young flower buds of myb21 myb24 myb57 accumulated more jasmonic acid (Figure 6), suggesting that MYBs negatively regulate JA biosynthesis to attenuate JA-induced expression of MYBs and to elaborately regulate stamen development, and that the restored fertility in opr3 (Stintzi and Browse, 2000;Chehab et al., 2011) by suitable MYB24 overexpression (Figure 2) is not due to recovery of JA biosynthesis. It would be useful for understanding the MYB21/24 module in stamen development if the links between MYB21/24 and JA biosynthetic genes are elucidated.