The Mutation of Rice MEDIATOR25, OsMED25, Induces Rice Bacterial Blight Resistance through Altering Jasmonate- and Auxin-Signaling

Rice bacterial blight disease caused by Xanthomonas oryzae pv. oryzae (Xoo) is one of the most severe diseases of rice. However, the regulatory mechanisms of rice defense against Xoo remain poorly understood. The rice MEDIATOR25, OsMED25—a subunit of the mediator multiprotein complex that acts as a universal adaptor between transcription factors (TFs) and RNA polymerase II—plays an important role in jasmonic acid (JA)-mediated lateral root development in rice. In this study, we found that OsMED25 also plays an important role in JA- and auxin-mediated resistance responses against rice bacterial blight. The osmed25 loss-of-function mutant exhibited high resistance to Xoo. The expression of JA-responsive defense-related genes regulated by OsMYC2, which is a positive TF in JA signaling, was downregulated in osmed25 mutants. Conversely, expression of some OsMYC2-independent JA-responsive defense-related genes was upregulated in osmed25 mutants. Furthermore, OsMED25 interacted with some AUXIN RESPONSE FACTORS (OsARFs) that regulate auxin signaling, whereas the mutated osmed25 protein did not interact with the OsARFs. The expression of auxin-responsive genes was downregulated in osmed25 mutants, and auxin-induced susceptibility to Xoo was not observed in osmed25 mutants. These results indicate that OsMED25 plays an important role in the stable regulation of JA- and auxin-mediated signaling in rice defense response.


Introduction
Rice (Oryza sativa L.) is one of the most important crops worldwide and a model plant of monocotyledonous species with regard to molecular studies. Rice bacterial blight disease caused by the hemi-biotrophic pathogen Xanthomonas oryzae pv. oryzae (Xoo) is one of the most severe diseases of rice. Xoo is a vascular pathogen which remains within the xylem vessels throughout the course of the disease [1]. The two plant hormones, salicylic acid (SA) and jasmonic acid (JA), play important roles regarding resistance to biotrophic/hemibiotrophic and necrotrophic pathogens in plants, respectively, but were suggested to exert mutually antagonistic interactions in dicotyledonous plants [2]. However, recent studies revealed that JA also plays an important role in the defense response to Xoo in rice [3]. The JA-responsive valine-glutamine (VQ)-motif-containing protein, OsVQ13, positively affects defense signaling regulated by OsWRKY45, which is a central transcription factor (TF) in the SA-mediated defense response to Xoo in rice [4,5].
The accumulation of bioactive JA, JA-isoleucine, induces upregulation of OsWRKY45 and enhances resistance to Xoo in rice [6]. Additionally, many defense-related genes including OsWRKY45 and OsVQ13 are upregulated by both JA and the functional analogue of SA in rice, benzothiadiazole [5,7]. Furthermore, the rices ENHANCED DISEASE SUS-CEPTIBILITY1 (OsEDS1) and PHYTOALEXIN DEFICIENT4 (OsPAD4), which are rice homologues of Arabidopsis EDS1 and PAD4, both regulate SA-dependent defense response in Arabidopsis but play important roles in the JA-mediated defense response to Xoo in rice [8]. JA biosynthesis and expression of JA-responsive defense-related genes are suppressed in rice after inoculation of the virulent Xoo [9,10]. Treatment with cell wall-degrading enzymes extracted from Xoo activates JA signaling and induces Xoo resistance in rice [11]. JA induces the production of antibacterial compounds that suppress Xoo growth [12,13]. The monoterpene compound, γ-terpinene, damages the plasma membranes of Xoo cells [14], and geraniol suppresses expression of cell-division-related genes in Xoo [15]. Conversely, linalool exerts no antibacterial effects on Xoo; however, it acts as a signaling molecule to induce Xoo resistance in rice [16]. These results suggest that the defense mechanism via JA and SA signaling in rice is notably different from that of Arabidopsis, and that JA plays a crucial role in resistance of rice to Xoo.
These intricate JA signaling networks in rice are strictly regulated by TFs. The basic helix-loop-helix (bHLH)-type of TFs, OsMYC2, OsbHLH034, and RICE EARLY RESPON-SIVE TO JASMONATE1 (RERJ1) positively regulate JA-mediated defense response to Xoo in rice [17][18][19]. OsbHLH034 positively regulates biosynthesis of lignin, which is one of the most important components in the defense response to Xoo in rice [18,20,21]. RERJ1 interacts with OsMYC2 and positively regulates production of linalool by upregulating linalool synthase [19]. OsMYC2 is also activated under phosphate starvation conditions, resulting in the induction of resistance to Xoo in rice [22]. The OsMYC2 activity is suppressed by a complex that consists of rice JASMONATE ZIM (JAZ)-domain proteins (OsJAZs) and rice NOVEL INTERACTOR OF JAZ1 (OsNINJA1) [17,23]. The OsMYC2-interacting JAZ, OsJAZ8, acts as a repressor in the JA-induced defense response to Xoo in rice [24]. OsNINJA1-overexpressing rice plants are more susceptible to Xoo, due to the downregulation of OsMYC2-responsive defense-related genes [23]. The OsNINJA1-interacting protein OsSRO1a also acts as a mediator that suppresses OsMYC2-mediated defense response to Xoo in rice [25].
TFs, i.e., OsMYC2, do not directly interact with the general transcription machinery, including RNA polymerase II, during transcription. Rather, they bind to specific cisacting sequences present in the promoter regions of target genes through their DNAbinding domains and recruit the MEDIATOR multiprotein complex which binds TFs to the RNA polymerase II [26]. The MEDIATOR complex contains numerous subunits, and approximately 50 subunits have been found in the rice genomes [27]. The MEDIATOR25 (MED25) subunit was first isolated as a regulator involved in the flowering process, and it also plays important roles in other plant growth processes regulated by several plant hormones, including JA [26,28]. Recently, we reported that osmed25 loss-of-function rice mutants exhibited a JA-insensitive phenotype, and OsMED25 acts as a positive regulator in OsMYC2-mediated leaf senescence [29,30]. The OsMED25 directly interacts with OsMYC2 and positively regulates OsMYC2-responsive senescence-associated genes [30].
OsMED25 is involved in various JA signaling pathways; however, the involvement of OsMED25 in rice resistance to Xoo is unclear. In the present study, we investigated the role of OsMED25 in JA-mediated resistance to Xoo using an osmed25 mutant. We also investigated the involvement of auxin in OsMED25-mediated resistance of rice to Xoo.

Analysis of Xoo Resistance in osmed25 Mutants
Since we previously reported that OsMED25 is involved in root development and leaf senescence in rice through JA signaling [30], in the present study we investigated whether OsMED25 is involved in Xoo resistance. The osmed25 mutants were inoculated with virulent Xoo, and lengths of blight lesions were measured two weeks after the inoculation. Blight lesions in osmed25 mutants were significantly shorter than those in wild-type (WT) plants (Figure 1a  Data were analyzed using Student's t-test (n = 13 for WT; n = 8 for osmed25). The asterisk indicates a significant difference from WT plants at p < 0.05.

Interactions between OsMED25 and OsARFs
In Arabidopsis, AtMED25 regulates auxin signaling through its interaction with AUXIN RESPONSE FACTORs (AtARFs), which are key TFs for the regulation of auxin responses in Arabidopsis [26]. The auxin signaling was reported to negatively regulate Xoo resistance in rice [32]. Thus, we investigated whether OsMED25 interacted with OsARFs. The rice genome contains 25 OsARFs [33], but OsARF6, OsARF11, OsARF16, OsARF17, OsARF19, OsARF21, and OsARF25 were selected as those possibly interacting with OsMED25 because they are considered to be transcriptional activators of auxin signaling [33]. OsMYC2 was used as a positive control regarding interactions with OsMED25 [30]. As shown in Figure 3a, OsMED25 interacted with OsARF6, OsARF17, OsARF19, OsARF21, and OsARF25 in yeast cells. These OsMED25-interacting OsARFs in yeast cells were further confirmed to be interacting with OsMED25 in plant cells (Figure 3b). Conversely, the OsMED25-interacting OsARFs did not interact with mutated osmed25 protein in both yeast and plant cells (Figure 3a,b). Furthermore, no mKG fluorescence was observed in plant cells using mKGNor mKGC-empty vectors as negative controls (Supplementary Materials Figure S1).

Auxin-Mediated Suppression of Xoo Resistance in osmed25 Mutants
To determine whether OsMED25 was involved in auxin-mediated suppression of Xoo resistance, we performed a resistance test on osmed25 mutants after treatment with IAA. Following IAA treatment for 24 h, the plants were inoculated with virulent Xoo. Lengths of blight lesions were recorded at 16 days after inoculation. We found that the lengths of blight lesions in IAA-treated and -untreated osmed25 mutants were comparable ( Figure 5). However, the lengths of blight lesions among WT plants were significantly longer in IAA-treated than in IAA-untreated plants ( Figure 5).

Discussion
In our previous studies, we consistently observed that JA exerts beneficial effects regarding Xoo resistance in rice [5,7,[12][13][14][15][16][17][18][23][24][25]37]. In the present study, however, we observed that the JA-insensitive osmed25 mutants exhibited strong resistance to Xoo. We revealed that the expression of OsMYC2-mediated defense signaling tended to be downregulated due to the loss-of-function of OsMED25 while the expressions of some OsMYC2independent defense-related genes such as Bowman-Birk proteinase inhibitors and OsPrx26 were constitutively upregulated in osmed25 mutants. These results suggest that OsMED25 negatively regulates part of OsMYC2-independent defense signaling by forming complexes with uncharacterized TFs in rice. Identification of JA-mediated TFs regulated by OsMED25 is needed to comprehensively elucidate JA signaling in rice.
Transgenic rice plants overexpressing a Bowman-Birk proteinase inhibitor gene exhibited increased resistance to Xoo [38], suggesting that constitutive upregulation of some Bowman-Birk proteinase inhibitors in osmed25 mutants may result in increased resistance to Xoo. Rice class III peroxidases such as OsPrx38 and OsPrx114 play an important role in Xoo resistance by producing lignin, which is essential for the thickening of secondary cell walls [18,20]. OsPrx114 (PO-C1) is strongly induced by inoculation with avirulent Xoo and is secreted to the xylem lumen and walls from the xylem parenchyma cells in rice [20]. Xoo contacts living cells such as xylem parenchyma cells through the pit membranes separating the xylem lumen from the associated xylem parenchyma cells. Inoculation of an avirulent Xoo strain triggered the thickening of xylem secondary walls and reduced the pit diameter, thus resulting in the decreased Xoo access to the xylem parenchyma cells [20]. Lignin is an essential component of the plant response leading to the thickening of the xylem secondary walls. Therefore, lignin production by Prxs plays an important role in the rice defense response to Xoo [20]. This was supported by the increased lignin content in the osmed25 mutants, suggesting that OsPrx26 may be involved in the lignin biosynthesis of rice. Further studies such as analysis using transgenic rice plants overexpressing Bowman-Birk proteinase inhibitors or OsPrx26 are needed to clarify their respective effects in relation to Xoo resistance in rice.
In the present study, we also revealed that OsMED25 selectively interacts with Os-ARFs. In particular, OsMED25 interacted with OsARF6, OsARF17, OsARF19, OsARF21, and OsARF25, and further positively regulated the expression of auxin-responsive genes, suggesting that OsMED25 is involved in the regulation of auxin signaling by forming complexes with specific OsARFs. Furthermore, we revealed that loss-of-function of OsMED25 adversely affected shoot gravitropic response as shown in osmed25 mutants, suggesting the presence of OsMED25/OsARFs-dependent auxin responses in rice. OsMED25 contains a von Willebrand factor type A domain, a non-conserved middle domain, an activator-interacting domain (ACID), and a glutamine-rich domain [26]. The mutation site of the osmed25 protein is in the ACID, which is presumed to interact with TFs such as OsMYC2 [26,30]. In the present study, the ACID region was found to be important for the interaction between OsMED25 and OsARFs. All tested OsARFs contained a conserved N-terminal DNA-binding domain, a non-conserved middle region, and a C-terminal dimerization domain, regardless of the interaction with OsMED25, and belong to the same subgroup [33,39]. We did not find any characteristic region present in OsMED25-interacting OsARFs but did find a characteristic region present in OsMED25-noninteracting OsARFs. Three-dimensional analysis of OsARFs may provide further insight into the interactions between OsMED25 and OsARFs.
Auxins adversely affect Xoo resistance in rice [3]. Treatment of IAA, which is the main auxin of rice, promotes susceptibility to Xoo in rice [32]. In the present study, we found that the auxin-mediated suppression of Xoo resistance did not occur in osmed25 mutants as shown by the downregulated expression of auxin-responsive genes, indicating that osmed25 mutants exhibit an auxin-insensitive phenotype regarding auxin-mediated Xoo resistance. Auxin levels are reduced by overexpression of rice NONEXPRESSOR OF PATHOGENESIS-RELATED GENES1, OsNPR1, which is a positive regulator of Xoo resistance in rice [40]. Furthermore, transgenic rice plants overexpressing the cytochrome P450 gene OsCYP71Z2 also exhibit increased resistance to Xoo through suppression of auxin biosynthesis [41]. Xoo produces and secretes IAA and induces auxin signaling in rice to facilitate the infection process [42]. Taken together, these results suggest that insensitivity to auxin derived from Xoo and suppression of auxin signaling regulated by OsMED25, in addition to aberrant JA signaling, resulted in strong resistance to Xoo in osmed25 mutants.
OsMED25 also acts as a positive regulator of brassinosteroid signaling in rice, which plays a positive role in rice Xoo resistance [43,44]. In Arabidopsis, AtMED25 is also involved in the regulation of ethylene (ET) and abscisic acid (ABA) signaling [26]. In wheat, TaMED25 is involved in the regulation of ET signaling and adversely affects resistance to powdery mildew caused by the biotrophic fungus Blumeria graminis f. sp. tritici [45]. Both ET and ABA adversely affect Xoo resistance in rice [3]. Since OsMED25 is a multifunctional protein, it is involved in diverse hormone signaling pathways by forming complexes with unknown TFs and other mediator proteins in rice. Further experimental analyses are needed to clarify the regulation of hormone crosstalk through OsMED25 in rice.

Plant Materials, Auxin Treatment, and Bacterial Inoculation
The growth conditions of the rice plants (Oryza sativa L. 'Nipponbare') and Xoo (strain 7174) were the same as those previously described by Kashihara et al. [25]. The Xoo strain is virulent to Nipponbare. Fully opened fifth-leaf blades of the rice plants were inoculated using the clipping inoculation technique [46]. The lengths of the blight lesions were measured 14 or 16 days after inoculation. Treatment with 100 nM indole-3-acetic acid (IAA; Sigma-Aldrich, St. Louis, MO, USA) was performed following the method of JA treatment described by Yamada et al. [24]. To assess shoot gravitropic responses, 10-day-old seedlings were placed horizontally under dark conditions at 25 • C for 24 h.

Reverse Transcription-Quantitative PCR (RT-qPCR)
Total RNA was extracted from the fourth leaf blades using TRIzol reagent (Thermo Fisher Scientific, Witham, MA, USA). RT-qPCR was performed using TB Green Premix Ex Taq (Takara Bio, Kusatsu, Japan) in a Thermal Cycler Dice TP850 System (Takara Bio) according to the manufacturer's instructions. Four leaf blades were used per replicate. Analysis of the obtained data was performed following the method of Gomi et al. [47]. Sequences of gene-specific primers used for RT-qPCR are shown in Supplementary Materials Table S1.

Measurement of Lignin Content
The lignin content of fully opened fourth leaf blades from three-weeks-old rice plants was determined as described by Kashihara et al. [25]. Five leaf blades were used per replicate.

Bimolecular Fluorescence Complementation (BiFC) Assay
The Kusabira Green (mKG) system (MBL, Nagoya, Japan) for the BiFC assay was used [18]. To visualize the nuclei, rice histone H4 (Os10g39410) was fused with DsRed as previously described by Onohata and Gomi [18]. The constructed vectors were expressed in onion epidermal cells using a particle bombardment system (PDS-1000/He; BioRad, Hercules, CA, USA) following the method of Kim et al. [48]. We used a KEYENCE BIOREVO BZ-9000 microscope (KEYENCE, Osaka, Japan) to observe mKG and DsRed fluorescence. The conditions were the same as those previously described by Onohata and Gomi [18].

Supplementary Materials:
The following supporting information can be downloaded at https://www. mdpi.com/article/10.3390/plants11121601/s1; Table S1. Sequences of gene-specific primers used for RT-qPCR; Figure S1. BiFC experiments using empty vectors as negative controls in plant cells.