The Oc suspension cell line that was established from seedling roots of Oryza sativa L. indica type accession C5928 (Baba et al., 1986) was cultured in 50 mL of R2S liquid medium (Ohira et al., 1973) in a 120-rpm shaking incubator at 28°C, as previously reported (Cho et al., 2016; Yoon et al., 2021). After 2 weeks of culture, 20 mL of the upper liquid layer was decanted, and 10 mL of the remaining cell suspension was transferred to a fresh flask containing 40 mL of R2S culture medium.

The OsWRKY7 mutant (line number 5A-00022) was identified from T-DNA tagging lines generated in Oryza sativa var. japonica, cultivar ‘Dongjin’ (Jeon et al., 2000; Jeong et al., 2002; An et al., 2005a; An et al., 2005b; Jeong et al., 2006). Homozygous mutant plants from the T2 progeny were identified by PCR using genomic DNA isolated from the blade of the second leaf of each plant. The genotyping primers were GTGTCGGGTGCGTATTTAAAAAC (F), GGGAATTGGCGCTAAATCTGC (R), and atccagactgaatgcccacag (NGUS2) ( Supplementary Table 1 ). Null mutations of OsWRKY7 were obtained using the CRISPR/Cas9 method in ‘Donjin’ background. Homozygous mutants at T1 generation were detected through the subcloning and multiple sequencing of the mutated region using the genomic DNA extracted from the leaf blades of individual plants.

For overexpression of OsWRKY7, a full length of the gene was cloned by PCR using the specific primers ( Supplementary Table 1 ) with HindIII and SacI restriction enzyme sites. The gene was subcloned between the maize ubiquitin (ZmUbi) promoter and HA tag in the binary vector pGA3428 (Kim et al., 2009). The resulting vector was used to generate transgenic rice plants overexpressing OsWRKY7-HA. Produced plants were verified by western blot analysis using protein extracts from the leaf blades after 24 hours of M. oryzae infection. All plants were grown in the paddy field or in a controlled growth room under a long day conditions (i.e., 14 hours of light at 28°C/10 hours of darkness at 23°C, humidity of approximately 50%).

Oc cells were grown for 10 days in R2S solution complemented with 3% glucose as a carbon source. They were divided into two equal parts; one was added with sucrose and the other with glucose, obtaining a final sugar concentration of 3% in two replications. Samples were collected immediately before the transfer to new media (0 h) and 4 hours after sugar addition (4 h). Total RNA was extracted using the QIAGEN RNA purification kit (Hilden, Germany), and the purity and concentration of RNA were estimated using a NanoDrop 2000 spectrophotometer (Thermo scientific). RNA-seq analysis was performed based on protocols previously reported in Peng et al. (2021). The gene ontology (GO) terms of the differentially expressed genes were analyzed using the Rice Oligo Array Database (ricephylogenomics-khu.org). All the mapped genes were downloaded and filtered based on the following criteria: hyper p-value < 0.05 and query numbers > 10. Additional significance of the GO terms was denoted by a fold enrichment value > 1.5, which was obtained by dividing the query number by the query expectation value (Hong et al., 2020; Kim et al., 2021). The biotic stress category was obtained via the MapMan analysis toolkit 3.5.1 R2 (Thimm et al., 2004).

Total RNA was extracted from the leaves, roots, and Oc cells using RNAiso Plus (Takara, Japan). After combining 3 µg of RNA with 10 ng of oligo (dT), the first-strand cDNA was synthesized using 100 units of Moloney murine leukemia virus (M-MLV) reverse transcriptase (Promega, Madison, WI, USA), 100 units of RNasin (Promega), and 2.5 mM deoxyribonucleotide triphosphates. The gene expression levels were analyzed by quantitative RT-PCR using SYBR Green I, Prime Q-Matermix (2x) (GENET Bio, Daejeon, Korea) in a Rotor-Gene Q system (Qiagen, Hilden, Germany). Rice Actin1 (OsActin1) was used as an endogenous control to normalize the expression level of the respective genes. All experiments were repeated at least three times, and the relative transcript levels were calculated using the ΔΔCt method (Schmittgen and Livak, 2008; Yoon et al., 2022). The primers used in this study are listed in Supplementary Table 1 .

After selecting a potential target sequence with the CRISPR web tool, the RNA scaffold structure was analyzed using the RNAfold web server (univie.ac.at), and off-target sites were checked with the web tools (http://crispr.dbcls.jp; Naito et al., 2015 and CRISPOR (tefor.net); Concordet and Haeussler, 2018). A potential target sequence (5’-ATGTCATCGTACTTCTCCCA-3’) within the first exon of OsWRKY7 was cloned into the CRISPR vector pRGEB32 through the BsaI restriction site. The construct was transformed into the Agrobacterium tumefaciens strain LBA4404 using the freeze–thaw method (An, 1987). Transgenic rice plants were obtained via the Agrobacterium-mediated co-cultivation method (Lee et al., 1999).

The 3.081-kb promoter region (−3098 to −18 bp from the ATG start codon) of OsWRKY7 from cultivar ‘Nipponbare’ was cloned into the HindIII-digested pGA3452 vector (Kim et al., 2009) upstream of the luciferase (LUC) coding region using the 5X In-Fusion HD enzyme Premix (Takara, Japan), which generated the pOsWRKY7-LUC reporter vector. The 2.856-kb promoter portion (−2856 to −1 bp from the start ATG codon) of OsPR10a was inserted into pGA3452, which resulted in pOsPR10a-LUC. The ZmUbi-GUS (Cho et al., 2009) construct was used as the internal control. For effector constructs, the coding regions of OsWRKY7, OsWRKY28, and OsWRKY10 were amplified by PCR and were inserted between the ZmUbi promoter and 6X Myc of pGA3697 (Kim et al., 2009).

Protoplasts were prepared from Oc cells based on methods described in Cho et al. (2016) and He et al. (2016). In brief, Oc cells were incubated in an enzyme solution (2% cellulose RS, 1% macerozyme, 0.1% CaCl₂, 0.4 M mannitol, 0.1% MES, pH 5.7) for 4 hours at 28°C in the dark. After incubation, they were added with an equal volume of KMC solution (117 mM potassium chloride, 82 mM magnesium chloride, and 85 mM calcium chloride). The number of protoplasts was quantified with a hemocytometer to ensure that a cell density of 2 × 10⁶ cells mL⁻¹ was reached. The isolated protoplasts were transformed with 10 µg of each plasmid DNA using the PEG method (Yoo et al., 2007; Cho et al., 2022). ZmUbi : GUS construct (2 µg) was co-transfected as an internal control (Cho et al., 2009). The GUS activity was used to normalize the LUC activity.

The chromatin immunoprecipitation (ChIP) assay was performed using the protocols described in Haring et al. (2007) and Yoon et al. (2022) with some modifications. Firstly, the fresh leaves from the OsWRKY7-HA transgenic plants were collected and the proteins and chromatin were cross-linked using 3% formaldehyde. After the isolation of nuclei, DNA was broken into fragments of approximately 200–800 bp in length by sonication. Agarose beads of G and A proteins (Milipore; http://www.emdmilipore.com) were used to preclear the sheared chromatin for 1 hour at 4°C. Anti-HA monoclonal antibody (C29F4, https://www.cellsignal.com) was used for immunoprecipitation together with the precleared agarose beads for 4 hours at 4°C. The washing and reverse cross-linking steps described in Haring et al. (2007) were followed. The enriched chromatin was analyzed by qRT-PCR using the primers listed in Supplementary Table 1 . The obtained data were normalized using the fold enrichment method (Haring et al., 2007; Yoon et al., 2022).

The oligonucleotide (-808 to -831 from the start ATG of OsPR10a) was end labeled with biotin (Microgen, Korea). To produce MBP-tagged OsWRKY7 protein, the coding sequence of OsWRKY7 was amplified by PCR and cloned into the pMAL-c2X vector (Addgene plasmid # 75286; http://n2t.net/addgene:75286; RRID : Addgene_75286; walker et al., 2010) through the EcoR1 and Sal1 enzyme sites. The MBP-OsWRKY7 protein were extracted using amylose resin (E8021S, New England, Biolabs Inc.) The MBP-OsWRKY7 protein (1 µg) and biotin labeled ds-DNA fragment (750 fmol) were incubated in the reagent (KIT 0020148, Thermo scientific) containing 1× binding buffer, 10% glycerol, 5 mM MgCl₂, 0.5 µg poly (dI-dC), and 0.05% NP-40 in a final volume of 20 µL for 30 min at 25°C. After electrophoresis in 6% polyacrylamide gel, the DNA-protein complexes were transferred to membrane. The procedure for detection of immobilized nucleic acids was according to the protocol provided by the manufacture.

The Magnaporthe oryzae PO6-6 strain was cultured on an agar medium containing 8% V8 juice (Cambell’s Soup Company, USA) for 2 weeks under fluorescent light. The resulting spores were collected and diluted in distilled water to obtain the concentration of 5 × 10⁶ mL⁻¹ for spot inoculation. A solution containing 2 × 10⁵ spores per mL was sprayed onto the fully expanded healthy leaves of 6-week-old plants. Leaf samples were harvested and photographed to calculate the size of the lesions developed in the susceptible plants (Vo et al., 2019).

To identify genes that preferentially responded to exogenously provided sucrose, RNA-seq analysis was performed using Oc cell lines that had been maintained in suspension culture for a long period (Baba et al., 1986). Because the cells were grown in a culture medium containing sucrose, the carbon source was replaced with glucose to deplete sucrose when the cells were sub-cultured. Ten days after subculture, sucrose was added to a culture flask to obtain a final concentration of 3%. As a control, glucose was added to another culture flask to reach the same final concentration. Four hours after adding the sugars, cells were harvested and mRNAs were isolated to construct cDNA libraries. RNA-seq analyses of the libraries showed that the transcript levels of 1,336 genes were at least doubled by the sucrose addition during the 4-h culture period ( Supplementary Figure 1A ). Among these genes, 871 were also induced by glucose, leaving 465 that were sucrose preferential. The analyses also revealed that 1,148 genes were preferentially suppressed by sucrose ( Supplementary Figure 1B ). To validate the RNA-seq data, 12 upregulated and 12 downregulated genes were selected for qRT-PCR analysis, which confirmed that eight and 10 genes belonging to the two groups, respectively, were sucrose responsive ( Supplementary Table 2 ). GO analysis of the 465 genes preferentially induced by sucrose showed that the biological terms dominated by sucrose were protein amino acid phosphorylation, defense response, and transport ( Supplementary Figure 1C ).

Mapman analysis of all the sucrose up-regulated genes from the RNA-seq analysis showed that three groups of transcription factors were enriched in biotic stress ( Supplementary Figure 2 ). Because WRKY transcription factors are involved in defense response against pathogens, we selected all three WRKY genes (OsWRKY7, OsWRKY26, and OsWRKY108) that were sucrose inducible. qRT-PCR analysis showed that the transcript level of OsWRKY7 was increased by sucrose, reaching the maximum level at 2 hours after sucrose addition, and then rapidly declining to the basal level at 6 hours ( Figure 1A ). Glucose also induced the transcript level, but not significantly. Mannitol did not increase gene expression during the 8-h incubation time, which indicated that the expression of OsWRKY7 was preferentially induced by sucrose. The level of OsWRKY26 transcript started to increase at 2 hours after sucrose addition, reached the maximum at 4 hours, and then declined ( Figure 1B ). The transcript level was also increased by glucose addition. However, the gene was not induced at 2 hours after glucose addition, but its transcript was increased to approximately half of the sucrose-induced level at 4 hours and continuously increased at 6 and 8 h. The transcript level was not changed by mannitol. The above observations suggest that OsWRKY26 responded to both sucrose and glucose, but the gene was more rapidly induced by sucrose. Another WRKY gene, OsWRKY108, was also induced by sucrose. Compared to the other two genes discussed above, OsWRKY108 responded more rapidly to sucrose, reaching the maximum level at 0.5 hours after sucrose addition, and declining to the basal level at 2 hours ( Figure 1C ). This gene was weakly induced by glucose and did not respond to mannitol. The pathogen marker gene OsPR10a was also preferentially induced by sucrose, but not by glucose or mannitol ( Figure 1D ).

Because OsWRKY7 was specifically and transiently induced by sucrose, further tests were conducted to determine whether this gene was induced by other sugars. In addition to glucose, other monosaccharides such as galactose and fructose, as well as glucose and fructose added at the same time, did not induce gene expression above the control level during the 2-h incubation period ( Figure 2A ). Also, gene expression was not stimulated by maltose, a disaccharide that consists of two glucose molecules. The results suggest that OsWRKY7 is specifically induced by sucrose. Lowering sucrose concentration to 2%, 1%, and 0.5% did not influence the rate and level of OsWRKY7 induction ( Figure 2B ), however, further reductions to concentrations of 0.25% and 0.1% slightly reduced the sucrose inducibility.

To investigate whether gene expression was induced by sucrose in plants, 14-day-old rice seedlings grown hydroponically in Yoshida solution were left in the dark for 48 hours to reduce their endogenous sugar level. After this period, sucrose was added to the culture medium at the final concentration of 3%. Analysis of the OsWRKY7 transcript level in the roots showed that it increased significantly at 8 hours after sucrose addition ( Figure 2C ). Glucose also induced gene expression in the roots, but the induced levels were only slightly higher than those achieved by mannitol in the control.

The effect of mannoheptulose, which blocks hexose-mediated signaling (Chiou and Bush, 1998), was also tested. The results showed that the addition of 5 µM mannoheptulose to the sucrose solution did not interfere with the sucrose-induced expression of OsWRKY7 ( Figure 2D ). This observation suggests that hexoses produced from sucrose metabolism are probably not involved in the induction of the gene and that sucrose may function directly to promote OsWRKY7 expression.

To elucidate whether protein stability was involved in the sucrose response, MG132, which is an inhibitor of the 26S proteasomal protease (Genschik et al., 1998), was added to the sucrose solution. Analysis of the OsWRKY7 transcript level showed that the sucrose-induced expression of the gene was suppressed by 100 µM MG132 ( Figure 2E ). As a control, the sucrose-induced OsWRKY108 gene was also tested and it was shown that the expression of this gene was not affected by MG132 ( Supplementary Figure 3 ). This result suggests that a regulatory protein that represses OsWRKY7 expression is degraded by an E3 ligase during the sucrose treatment.

To examine whether the induction of OsWRKY7 by sucrose occurred at the transcriptional level, the 3.081-kb promoter sequence of OsWRKY7 was placed upstream of the LUC coding region and the construct was introduced into protoplasts prepared from Oc cells. As an internal control, the GUS coding region driven by the ZmUbi promoter was co-transfected. After dividing the treated cells into three portions, sucrose, glucose, and mannitol at the final concentration of 3% were added to each aliquot. After 15 hours of incubation, the samples were analyzed and it was shown that the LUC activity in the sucrose-added cells was considerably higher than that in the mannitol-added cells ( Figure 2F ). On the contrary, the induction of the reporter gene expression by glucose was only slightly above that achieved by mannitol. This experiment demonstrated that sucrose preferentially induced OsWRKY7 at the transcription initiation level.

Expression of OsWRKY7 was induced by M. oryzae 48 hours after infection ( Supplementary Figure 4A ). To study the functional roles of OsWRKY7, we identified T-DNA tagging line 5A-00022, in which T-DNA was inserted at 375 bp upstream of the start ATG codon of OsWRKY7 ( Supplementary Figure 4B ). Homozygous progeny of the oswrky7-1 mutant was used for characterization of the gene function. In the mutant line, the expression level of OsWRKY7 was decreased to approximately a half of the wild type (WT) level, which indicated that the inserted T-DNA reduced gene expression ( Supplementary Figure 4C ). When the plants were infected with M. oryzae, the lesions were longer and broader than those observed in the segregated WT plants ( Supplementary Figures 4D, E ), suggesting that OsWRKY7 is involved in the defense response against this pathogen.

To confirm this observation, knockout mutations were generated in the first exon of OsWRKY7 via the CRISPR/Cas9 system ( Figure 3A ). Among 28 independent transgenic plants, five were selected to determine the flanking regions of the target site. Two mutant lines were chosen: oswrky7-2 carrying a single-bp deletion and oswrky7-3 carrying a two-bp deletion ( Figure 3A , Supplementary Figures 5A, B ). The deletions generated frameshift mutations, causing early termination during translation of the OsWRKY7 transcript. The mutant plants grew normally without any significant phenotypic alteration in the plant height, grain numbers per panicle, panicle length, and 100-grain weight ( Figure 3B , Supplementary Figure 6 ). Interestingly, the OsWRKY7 transcript levels in the frameshift mutants were reduced compared to those in the WT plants, suggesting that the deletion mutations affected either the transcription rate or the stability of the transcript ( Figure 3C ). When the plants were inoculated with M. oryzae, the mutants displayed a phenotype that was more susceptible to the fungus compared with the WT controls ( Figures 3D, E ). These experiments confirmed that OsWRKY7 is involved in the defense response against M. oryzae.

To obtain transgenic rice plants that constitutively expressed OsWRKY7, the gene was placed after the ZmUbi promoter and the construct was introduced to rice embryonic calli. However, it was quite difficult to obtain transgenic plants, which suggested that the overexpression of OsWRKY7 was harmful to the plants. To overcome the difficulty, the HA tag was added at the end of the OsWRKY7 coding region, and the fusion molecule was placed after the ZmUbi promoter. In this way, several transgenic plants expressing OsWRKY7-HA were obtained, which indicated that attaching the tag reduced the harmful effect of the gene ( Figure 4A ). Plants #5 and #7, which expressed the fusion transcript and the HA-tagged protein at high levels, were selected ( Figure 4B and Supplementary Figure 7 ). After inoculating the mature leaves of the transgenic plants with M. oryzae, it was observed that the lesion lengths and widths at the infection sites were significantly smaller in the transgenic plants than in the WT control plants ( Figures 4C, D ). This observation confirmed that OsWRKY7 is a positive regulatory element in the defense response to M. oryzae.

The expression level of the genes that were induced in response to pathogen infection were then analyzed. The levels of the pathogen-responsive marker genes OsPR1a, OsPR1b, and OsPR10a, were significantly reduced in oswrky7 mutant plants ( Figures 5A-C ). On the contrary, expression levels of the pathogen responsive marker genes were markedly increased in the plants of expressing OsWRKY7-HA ( Figures 5D-F ). This result indicates that the markers function downstream of OsWRKY7. It was then tested whether SA or JA were involved in the induction of the marker genes by OsWRKY7. The mutations of this gene did not affect the expression of phenylalanine ammonia-lyase (OsPAL) or isochorismate synthase 1 (OsICS1), which are involved in SA biosynthesis ( Supplementary Figures 8A, B ). In addition, a key regulator of systemic acquired resistance, NPR1 homolog 1 (OsNH1), was also not affected in mutant ( Supplementary Figure 8C ). Likewise, the mutations did not influence the expression of genes encoding enzymes associated with JA biosynthesis, namely allene oxide synthase (OsAOS2), allene oxide cyclase 1 (OsAOC1), and oxophytodienoate reductase 3 (OsOPR3) ( Supplementary Figures 8D-F ). These results suggest that the signaling pathway of SA or JA did not mediate the induction of the pathogen-responsive marker genes by OsWRKY7.

To evaluate whether the OsWRKY7-induced PR genes were sucrose-inducible, hydroponically grown seedlings were treated with 3% sucrose at 14 days after germination (DAG). After 8 hours of incubation in the solution, the expression levels of the PR genes in the roots were measured. The assay showed that the transcript level of OsPR10a was considerably higher in the sucrose-treated seedlings than in the control seedlings treated with 3% mannitol ( Figure 6A ). Glucose treatment did not induce gene expression above the level observed in the control. This result is consistent with the data observed in Oc cells ( Figure 2D ). On the contrary, the expression of OsPR1a was induced by both sucrose and glucose ( Figure 6B ). The OsPR1b transcript levels in sugar solutions were similar to those observed in mannitol solution ( Figure 6C ). These experiments indicate that, among the three PR genes tested, only OsPR10a was preferentially induced by sucrose.

We tested whether OsPR10a expression in oswrky7 mutant was recovered by the sucrose treatment. The WT and oswrky7 plants that were grown hydroponically in Yoshida solution for 14 days were treated with mannitol or sucrose at the final concentration of 3%. After 9 hours treatment in the solutions, the root were sampled and the expression level of OsPR10a was measured. The experiment showed that sucrose significantly increased OsPR10a expression in the WT as observed in the previous experiments. Similarly, the transcript level of the gene in the mutant was also induced by sucrose, although the induced level was lower compared with WT ( Supplementary Figure 9 ). This observation suggests that there are other sucrose-inducible transcription factors that induce OsPR10a expression.

Because the expression of PR10a was significantly reduced in OsWRKY7 mutants, PR10a may be directly controlled by the transcription factor. To prove this hypothesis, a construct containing the 2.856-kb promoter region of OsPR10a was placed in front of the LUC gene. We also constructed another molecule in which the coding region of OsWRKY7 was placed under the ZmUbi promoter. Both molecules, pOsPR10a-LUC and pUbi-WRKY7, were co-transfected into the Oc cell protoplasts and incubated for 15 hours to express the introduced genes. The results showed that the LUC expression level in the protoplasts co-transferred with both molecules were significantly higher than that in the protoplasts transferred with pOsPR10a-LUC alone ( Figure 6D ). As controls, OsWRKY10 and OsWRKY28 were inserted downstream of the ZmUbi promoter. It has been previously reported that OsWRKY10 functions as a positive regulatory element for the expression of PR10a, whereas OsWRKY28 does not influence the PR gene (Ersong et al., 2021). In this study, when pUbi-WRKY28 was co-introduced with pOsPR10a-LUC, the luciferase activity was not increased above the background level that was observed from pOsPR10a-LUC alone ( Figure 6D ). On the contrary, co-transfection of pOsPR10a-LUC and pUbi-OsWRKY10 significantly induced the reporter gene expression ( Figure 6D ). The above observations showed that the transient assay resembled the previous results and suggested that OsWRKY7 directly promoted the expression of pOsPR10a-LUC.

To test whether the OsPR10a promoter region responded directly to sucrose, the pOsPR10a-LUC vector was introduced into the protoplasts and was incubated in the culture solution containing three different sugars. Although the LUC expression level in the protoplasts cultured in sucrose-containing medium was slightly higher than that in the mannitol control, a similar expression level was achieved for the protoplasts grown in the glucose-containing medium ( Figure 6E ). This suggests that sucrose did not induce the gene directly. It was also tested whether the increased expression of the OsPR10a promoter by the OsWRKY7 protein was sucrose dependent. To answer this question, both pOsPR10a-LUC and pUbi-WRKY7 were introduced into the protoplasts and the effects of sucrose were examined. The experiment showed that sucrose did not significantly increase the pOsPR10a promoter activity above the levels obtained from mannitol or glucose ( Figure 6E ). This suggests that the sucrose induction of PR10a was not due to the stability of the OsWRKY7 mRNA or protein during sucrose treatments.

To confirm that OsWRKY7 directly controlled the OsPR10a promoter, chromatin immunoprecipitation (ChIP) experiments were performed with transgenic plants expressing the OsWRKY7-HA protein. Primers were designed to target W-box elements, W-box like elements 1 (WLE1), and ASF1MOTIF, which all belong to the WRKY protein-binding TGAC core elements (Maleck et al., 2001; Hwang et al., 2008) ( Figure 7A ). In the ChIP in vivo assay, the P2 and P3 DNA fragments containing the WLE1 element were enriched ( Figure 7B ).

To further confirm the interaction between OsWRKY7 and the OsPR10a promoter, the oligonucleotide (-808 to - 831 from the start ATG) containing the WLE1 element within the P2 fragment were used for EMSA experiment using MBP tagged OsWRKY7 protein ( Figure 7C ). The experiment showed that OsWRKY7 protein bind to the sequence and the binding was reduced when the unlabeled probe was used as a competitor. Both ChIP and EMSA assays support the hypothesis that OsWRKY7 directly binds to the promoter region of OsPR10a.