Poaceae‐specific β‐1,3;1,4‐d‐glucans link jasmonate signalling to OsLecRK1‐mediated defence response during rice‐brown planthopper interactions

Summary Brown planthopper (BPH, Nilaparvata lugens), a highly destructive insect pest, poses a serious threat to rice (Oryza sativa) production worldwide. Jasmonates are key phytohormones that regulate plant defences against BPH; however, the molecular link between jasmonates and BPH responses in rice remains largely unknown. Here, we discovered a Poaceae‐specific metabolite, mixed‐linkage β‐1,3;1,4‐d‐glucan (MLG), which contributes to jasmonate‐mediated BPH resistance. MLG levels in rice significantly increased upon BPH attack. Overexpressing OsCslF6, which encodes a glucan synthase that catalyses MLG biosynthesis, significantly enhanced BPH resistance and cell wall thickness in vascular bundles, whereas knockout of OsCslF6 reduced BPH resistance and vascular wall thickness. OsMYC2, a master transcription factor of jasmonate signalling, directly controlled the upregulation of OsCslF6 in response to BPH feeding. The AT‐rich domain of the OsCslF6 promoter varies in rice varieties from different locations and natural variants in this domain were associated with BPH resistance. MLG‐derived oligosaccharides bound to the plasma membrane‐anchored LECTIN RECEPTOR KINASE1 OsLecRK1 and modulated its activity. Thus, our findings suggest that the OsMYC2‐OsCslF6 module regulates pest resistance by modulating MLG production to enhance vascular wall thickness and OsLecRK1‐mediated defence signalling during rice‐BPH interactions.

Glucosinolates (GSs) are a class of secondary metabolites containing nitrogen and sulfur that are characteristic of the Brassicaceae family (Schweizer et al., 2013;Wang et al., 2019). Increasing evidence suggests that upon herbivore attack, a JAZ-MYC module regulates the expression of key GS biosynthesis genes to control the accumulation of these toxic compounds (Schweizer et al., 2013;Wang et al., 2019). In Arabidopsis, the myc2/3/4 triple knockout mutant shows increased susceptibility to the feeding of generalist herbivores, suggesting that these MYC transcription factors redundantly modulate defences against insects in Brassicaceae plants (Fernandez-Calvo et al., 2011;Major et al., 2017). Biochemical analyses revealed that the MYC2/3/4 transcription factors regulate GS biosynthesis by directly associating with the promoters of several GS biosynthesis genes or interacting with MYB factors involved in regulating GS biosynthesis (Liao et al., 2020;Schweizer et al., 2013). However, how JAs help regulate defence responses against pests in the Poaceae family remains largely unknown.
Some BPH resistance genes, such as Bph6, Bph14, and Bph29, produce broad plant resistance to BPH by modulating phytohormone signal transduction pathways (Du et al., 2009;Guo et al., 2018;Wang et al., 2015). Indeed, phytohormones including JA, salicylic acid (SA), and brassinosteroids play important roles in regulating plant responses to BPH feeding (Pan et al., 2018;Xu et al., 2021). For example, the R2R3 MYB transcription factor OsMYB30 regulates the expression of the phenylalanine ammonia-lyase genes OsPAL6 and OsPAL8 to control BPH resistance and SA accumulation (He et al., 2020), providing a functional OsMYB30-OsPAL-SA module in rice BPH resistance.
Here, we report that mixed-linkage b-1,3;1,4-D-glucan (MLG) is a Poaceae-specific metabolite that acts downstream of JA signalling to regulate BPH resistance. Overproduction of MLG enhances BPH resistance and causes thickening of the vascular bundles. Moreover, MLG-derived oligosaccharides bind with high affinities to OsLecRK1 in vitro and modulate its activity in vivo, suggesting that MLGs may contribute to BPH resistance by enhancing vascular thickening and OsLecRK1-mediated defence signalling during the rice-BPH interaction.

Results
Mixed-linkage b-1,3;1,4-D-glucan is required for JAmediated BPH resistance To investigate the potential role of JAs in defence against BPH in rice, we measured endogenous JA levels in rice at various time points (0, 6, 12, and 24 h) during BPH infestation. As shown in Figure S1a, compared to the untreated controls (0 h), the levels of OPDA, JA, and JA-Ile in the leaf sheaths and stems of BPHinfested plants were elevated beginning at 6 h of BPH infestation, with a peak at 12 h after treatment. Reverse transcription quantitative PCR (RT-qPCR) analysis showed that the JA biosynthesis-related gene OsLOX2 and the JA signalling-related genes OsCOI1a and OsCOI1b were significantly upregulated at 6, 12, and 24 h after BPH application ( Figure S1b).
To further explore the roles of JAs in BPH resistance, we used CRISPR-Cas9 to target the third exon of OsLOX2, the fourth exon of OsCOI1a, and the third exon of OsCOI1b. We generated two independent oslox2 lines (designated oslox2-1 and oslox2-2) and two oscoi1a oscoi1b double mutants (designated oscoi1-1 and oscoi1-2) with knockout mutations of OsLOX2 or OsCOI1a and OsCOI1b, respectively ( Figure S2a,b). Compared to wild-type Nipponbare (NIP), oslox2-1 and oslox2-2 plants had significantly reduced levels of wounding-induced JA and JA-Ile ( Figure S2c). Exogenous application of MeJA at 5 and 10 lM significantly inhibited the root growth of NIP seedlings, whereas the oscoi1 mutants were less sensitive to MeJA-induced root inhibition than NIP ( Figure S2d,e).
We next examined the responses of NIP, oslox2 (oslox2-1 and oslox2-2), and oscoi1a/1b (oscoi1-1 and oscoi1-2) plants to BPH feeding. Four-week-old oslox2 and oscoi1 plants exhibited few morphological differences from NIP seedlings ( Figure S3a). However, compared to NIP plants, oslox2 and oscoi1 plants were hypersensitive to BPH infestation ( Figure S3a). The increased susceptibility of oslox2 and oscoi1 to BPH was further confirmed by measuring the survival rates of plants and the weight of honeydew excreted by BPH. Upon BPH feeding for 5 days, only 27% of oslox2-1 and oslox2-2 plants and 20% of oscoi1-1 and oscoi1-2 plants, survived compared to the 65% survival rate of NIP plants ( Figure S3b). Consistent with these results, the honeydew weights were significantly higher on oslox2 and oscoi1 plants than on NIP ( Figure S3b). When we sprayed NIP, oslox2, and oscoi1 plants with 100 lM MeJA prior to BPH infestation, this treatment significantly reduced the susceptibility of oslox2, but not oscoi1, to BPH feeding (Figure S3c,d).
To identify potential JA-regulated metabolites that change in abundance in response to BPH feeding, we performed liquid chromatography mass spectrometry (LC-MS)-based widely targeted and untargeted metabolomic profiling of NIP, oslox2, and oscoi1 plants that had undergone BPH treatment for 0 and 24 h. We used leaf sheath and stem samples because they are the major sites of BPH feeding. Our widely targeted metabolomics analysis identified 805 metabolites that matched known biochemicals (Dataset S1). Among these, the levels of 109 and 88 metabolites were significantly different in the NIP vs. oslox2-1 and NIP vs. oscoi1-1 groups, respectively ( Figure S4a and Dataset S1). The levels of 24 metabolites were significantly different in NIP upon BPH exposure compared to NIP without treatment ( Figure S4b; Dataset S1). In particular, the levels of four BPH-inducible flavonoids, including one flavonoid (tricin-4 0methylether-3 0 -O-glucoside), two flavonols (kaempferol-4 0 -Oglucoside and isorhamnetin-3,7-O-diglucoside), and one dihydroflavonol (dihydrokaempferol) were significantly lower in both oslox2 and oscoi1 compared to NIP ( Figure S4c-e and Dataset S1).
We then performed untargeted metabolome profiling to identify possible Poaceae-specific metabolites that might be affected during the BPH-rice interaction. We detected approximately 800 features in the leaf sheaths and stems of rice plants (Dataset S2), including 261 and 302 features with significantly different levels in the NIP vs. oslox2-1 and NIP vs. oscoi1-1 groups, respectively ( Figure 1a and Dataset S2). Upon BPH exposure, 46 features accumulated to significantly higher levels (compared with untreated plants) in NIP, but their levels were significantly lower in oslox2 and oscoi1 compared to NIP (Figure 1b Mixed-linkage b-1,3;1,4-D-glucans (MLGs) are Poaceae-specific b-D-glucans that are essential for several developmental and defence processes in rice Zhang et al., 2021). Therefore, we measured MLG contents in NIP, oslox2, and oscoi1 upon BPH feeding for 0 and 24 h, using endohydrolase-based high performance anion exchange chromatography (HPAEC) and quantitative assay kits. MLG levels increased in NIP after 24 h of BPH infestation (Figure 1e,f). Consistent with this finding, the BPH-responsive accumulation of MLG was significantly reduced in both oslox2 and oscoi1 compared to NIP (Figure 1e,f). To further validate the functional link between MLGs and the JA pathway, we performed a chemical complementation test of oslox2 and oscoi1 plants via exogenous application of 50 ppm MLGs. As expected, the BPH-sensitive phenotypes of the oslox2 and oscoi1 mutants were recovered by MLG treatment compared to the untreated controls ( Figure S3e,f).

Overproduction of MLGs enhances BPH resistance in rice
MLGs are primarily synthesized by members of the cellulose synthase-like CslF family (Burton et al., 2006;Lim et al., 2018). Genetic analyses suggested that the loss of OsCslF6 leads to significant decreases in MLG levels in rice cell walls (Vega-Sanchez et al., 2012). To identify the OsCslF genes involved in BPHinduced MLG accumulation, we examined the spatial and temporal expression patterns of various OsCslF gene family members. OsCslF2 and OsCslF6 were mainly expressed in leaf sheaths and stems, while OsCslF1, OsCslF3, OsCslF4, OsCslF7, OsCslF8, and OsCslF9 were abundantly expressed in roots ( Figure S5a). Upon BPH exposure, most OsCslF genes (except for OsCslF9) were upregulated at 6 h after BPH infestation, with Leaf sheaths and stems of 4-week-old NIP, oslox2-1, and oscoi1-1 plants at 0 (CK) and 1 days (BPH) after infestation with BPH were collected and subjected to untargeted metabolomic analysis, followed by analysis using XCMS Online software. The diagrams show the number of different features identified by one-way ANOVA (P < 0.01) in each group of NIP vs. oslox2-1 (261) or NIP vs. oscoi1-1 (302) (a) and features that accumulated to significantly higher (b) or lower (c) levels in NIP plants compared to oslox2-1 and oscoi1-1 plants upon BPH infestation (P < 0.01). (d) Untargeted metabolomic profiling showing potential metabolites that accumulated upon BPH infestation in 4-week-old NIP plants but whose levels were significantly lower in the oslox2-1 and oscoi1-1 mutants (P < 0.01 by one-way ANOVA). Standard-scores (Z-scores) were used to indicate the means of arbitrary peak abundance units (n = 4 biological replicates) of the corresponding samples. The first metabolite is b-D-glucan. (e) HPAEC analysis showing MLG levels in the leaf sheaths and stems of 4-week-old NIP, oslox2, and oscoi1 plants upon BPH infestation for 0 (CK) and 1 day (BPH). Three biological replicates were conducted with similar results, and representative data from one replicate are shown. (f) MLG levels in the leaf sheaths and stems of 4week-old NIP, oslox2, and oscoi1 plants upon BPH infestation for 0 (CK) and 1 day (BPH). Data are means AE SD (n = 3 biological replicates). Letters indicate significant differences between groups conducted by one-way ANOVA, P < 0.05. peak expression detected at 12 or 24 h after treatment ( Figure S5b). Among these genes, OsCslF6 transcript levels were significantly lower in the oslox2 and oscoi1 mutants compared to NIP ( Figure S5b). Moreover, at 6, 12 and 24 h after treatment with 100 lM MeJA, the transcript levels of OsCslF6 were induced, but at significantly lower levels in the oscoi1 mutants compared to wild-type NIP and oslox2 plants ( Figure S5c). These findings suggest that OsCslF6 plays a major role in MLG biosynthesis in the leaf sheaths and stems, which are the sites of BPH infestation in rice.
To verify the role of OsCslF6 in BPH resistance, we generated OsCslF6 knockout (f6Cas3 and f6Cas10) ( Figure S6a) and OsCslF6 overexpression (F6OE4 and F6OE8) transgenic lines ( Figure S6b). Four-week-old f6Cas3 and f6Cas10 plants showed dwarf phenotypes compared to wild-type plants, which is consistent with previous findings (Vega-Sanchez et al., 2012). F6OE4 and F6OE8 plants displayed stronger morphological changes than NIP at this stage ( Figure 2a). Insect feeding assays indicated that under growth room conditions, the f6Cas3 and f6Cas10 mutants were hypersensitive to 7 days of BPH infestation compared to wild-type NIP plants (Figure 2a), based on survival rates ( Figure 2b). By contrast, compared to NIP and f6Cas plants, the F6OE4 and F6OE8 lines were more resistant to BPH (Figure 2a,b). After 7 days of BPH feeding, only 40% of wild-type plants survived, whereas more than 75% of OsCslF6-overexpressing plants survived (Figure 2b). To determine the potential role of OsCslF6 in broad-spectrum resistance to planthoppers, we tested the phenotype of f6cas9 and F6OE plants with another phloemsucking herbivore, white-backed planthopper (WBPH, Sogatella furcifera Horvath). Compared with NIP, the f6cas9 mutants were more susceptible to WBPH, while the F6OE lines were more tolerant to WBPH feeding (Figure 2a,c).
To further confirm the role of OsCslF6 in JA-mediated defence against insects, we exposed 6-week-old field-grown NIP, f6Cas3, and F6OE8 plants to BPH. Consistent with the data from plants in the growth room ( Figure 2a,b), compared to wild-type plants, the f6Cas3 mutant was more sensitive and the F6OE8 plants were more resistant to BPH feeding in the field (Figure 2d), based on survival rates ( Figure 2e). Analysis using a quantitative assay kit revealed that MLG levels were much lower in the f6Cas3 and f6Cas10 mutants but significantly higher in the F6OE4 and F6OE8 lines compared to NIP ( Figure S6c). These findings suggest that OsCslF6 is important for resistance to BPH infestation.
To investigate the structural mechanism by which OsCslF6 affects insect resistance, we used transmission electron microscopy and confocal microscopy to compare the vascular bundle size and cell wall thickness in transverse sections of leaf sheaths from rice plants of different genotypes at the tillering stage. We observed that the wall thickness of sclerenchyma cells of f6Cas3 and F6OE8 plants did not significantly differ from that of NIP plants (Figure 2f,g). However, the vascular bundles of f6Cas3 were significantly smaller than those of wild-type plants, whereas the vascular bundles of F6OE8 plants were significantly larger (Figure 2h), as confirmed by measuring the areas of phloem and xylem ( Figure 2i). Transmission electron microscopy showed that the cell walls of sieve tubes and companion cells were significantly thinner in the f6Cas3 mutants, but significantly thicker in F6OE8 plants, compared to NIP (Figure 2h,i). These findings suggest that overexpressing OsCslF6 enhances cell wall thickness in vascular bundles, which contributes to improved resistance to insect herbivory.

OsMYC2 regulates OsCslF6 transcript levels and BPH resistance
AtMYC2 is a key transcription factor that is essential for JA signalling in Arabidopsis (Lorenzo et al., 2004). To better understand the molecular link between JAs and OsCslF6mediated BPH resistance in rice, we generated OsMYC2 knockdown (OsMYC2 RNAi-1 and OsMYC2 RNAi-2) and overexpression (OsMYC2 OE-1 and OsMYC2 OE-2) transgenic lines ( Figure S7a,b). We then measured OsCslF6 transcript levels in OsMYC2 RNAi and OsMYC2-OE plants by RT-qPCR. Before BPH exposure, OsCslF6 transcript levels were significantly higher in OsMYC2-OE seedlings than in wild-type NIP and after BPH exposure, OsCslF6 transcript levels were lower or higher in OsMYC2 RNAi and OsMYC2-OE seedlings than in NIP, respectively ( Figure S7c). Consistent with their OsCslF6 transcript levels, OsMYC2-OE and OsMYC2 RNAi seedlings had significantly higher or lower MLG contents than the wild type upon BPH treatment, respectively ( Figure S7d).
In Arabidopsis, AtMYC2 binds to G-box or G-box-like motifs in the promoters of JA-responsive genes to directly regulate their transcription (Dombrecht et al., 2007;Kazan and Manners, 2013). Bioinformatic analysis of the promoters of the OsCslF genes revealed two putative G-box elements in the promoter region of OsCslF6 ( Figure 3a) and one or two G-box or G-box like elements in the promoters of other OsCslF genes ( Figure S8a). In chromatin immunoprecipitation-quantitative PCR (ChIP-qPCR) assays, OsMYC2 bound to both G-box elements in the OsCslF6 promoter ( Figure 3b), but not the other OsCslF promoters ( Figure S8b). Electrophoretic mobility shift assays (EMSAs) confirmed that the G-boxes located at P1 and P2 of the OsCslF6 promoters are targets of OsMYC2 ( Figure 3c). The addition of unlabeled competitor probes markedly reduced the signal intensities of OsMYC2-P1/P2 complexes; this effect was diminished in reactions using mutant competitor probes ( Figure 3c). Finally, in dual-luciferase assays, co-expression with OsMYC2 significantly increased the activity of luciferase driven by the OsCslF6 promoter compared to the controls (Figure 3d and Figure S8c). These results suggest that OsMYC2 binds directly to the OsCslF6 promoter to control BPH-induced MLG accumulation.
To genetically link OsMYC2 to the cellular function of OsCslF6, we crossed OsMYC2 RNAi-1 with F6OE8 plants to generate OsMYC2-RNAi OsCslF6-OE double transgenic plants (designated R1OE8). As shown in Figure 3, overexpressing OsCslF6 rescued the increased BPH sensitivity of OsMYC2 RNAi-1 plants (Figure 3e, f), suggesting that OsMYC2 plays an indispensable role in regulating OsCslF6-mediated plant responses to BPH infestation. Consistent with their increased MLG contents, the two OsMYC2-OE lines exhibited enhanced tolerance to BPH feeding compared to the NIP control ( Figures S9a,b).

Natural variation in the OsCslF6 promoter affects MLGmediated BPH resistance
To investigate the possible contribution of natural variation in OsCslF6 to MLG-mediated BPH resistance in rice, we analysed the single-nucleotide polymorphisms (SNPs) in the promoter and coding regions of OsCslF6 from the 3K Rice Genomes Project . We did not detect nonsynonymous sequence variants in the OsCslF6 coding region; however, we identified 14 SNPs in its promoter (Figure 4a).
An analysis of regulatory elements suggested the potential importance of SNP8 (T/A alleles) and SNP9 (T/A alleles) at analyses showed that all six rice ARID genes, OsARID1 to OsARID6, were transcriptionally upregulated by BPH feeding ( Figure S10), supporting their potential involvement in regulating insect defence responses in plants.
We further analysed the distribution of the TT, AT, and AA alleles of SNP8 and SNP9 in the cultivated rice species from the 3K Rice Genomes Project. The AA allele was mainly found in indica . Letters indicate significant differences between groups conducted by one-way ANOVA, P < 0.05. rice, the TT allele was mainly found in japonica rice (Figure 4c), and the AT allele was present in both cultivars ( Figure 4c). Geographical localization analysis revealed that varieties with the AA allele were primarily distributed in tropical regions, whereas varieties with the TT allele were primarily distributed in temperate regions (Figure 4d).
Given that the AT-rich motif in the AA allele promoter is very close to the OsMYC2-binding G-box element (Figure 4b), we next explored the potential link between these two elements in regulating the transcription of OsCslF6. ChIP-qPCR assays revealed that compared to the G-box element alone, the presence of the AT-rich element strongly activated the association of OsMYC2 with the G-box element in the OsCslF6 promoter ( Figure 4e). Furthermore, luciferase activity assays showed that either the single mutation of SNP8 (TT to AT) or double mutation of SNP8/9 (TT to AA) significantly enhanced the OsMYC2-induced activation of OsCslF6 expression compared to the allele (Figure 4f). These results suggest that SNP8 and SNP9 in the OsCslF6 promoter likely affect its regulation.
To further test the effect of these SNPs on OsCslF6 function in BPH resistance, we randomly selected seven cultivars with the AA or TT allele from the 3K sequenced rice accessions for BPH feeding tests. Most accessions containing the AA allele showed more resistance after 13 days of BPH infestation than accessions containing the TT allele under natural growth conditions (Figure 5a), based on the survival rate of each accession (Figure 5b). Consistent with these BPH resistance phenotypes, most of the AA accessions also accumulated more MLGs than the TT accessions under normal conditions (Figure 5c). These findings demonstrate that SNP8 and SNP9 in the OsCslF6 promoter contribute to MLG-mediated BPH resistance in rice.

MLGs bind OsLecRK1 in vitro and stimulate its activity in vivo
MLGs are carbohydrate-based molecular patterns that can be perceived by plants to trigger pathogen-induced immune responses (Rebaque et al., 2021). Moreover, in response to Magnaporthe oryzae infection of rice, oligosaccharides released from MLGs act as DAMPs, which are perceived by OsCERK1 to activate immune signalling (Yang et al., 2021). We therefore hypothesized that MLG may also function as precursors of DAMPs that are recognized by cell surface OsLecRKs during rice-BPH interactions. To test this possibility, we first evaluated the association between MLG-derived oligosaccharides and OsLecRKs (e) ChIP-qPCR data showing the promoter fragments P1 (À415 to À215), P2 (À314 to À215), and P3 (À415 to À306) in (a; TT version) that were amplified from immunoprecipitated proteins pulled down by anti-HA antibodies. The OsACTIN1 promoter was used as a control. Data are means AE SD (n = 3 biological replicates). (f) Luciferase activity of rice protoplasts co-transfected with the effector plasmids (OsMYC2 and vector control) combined with reporters (control+OsCslF6; OsMYC2 + OsCslF6 TT; OsMYC2 + OsCslF6 AT; and OsMYC2 + OsCslF6 AA). The activity of protoplasts transfected with the empty effector construct was defined as 1. Data are means AE SD (n = 3 biological replicates). Letters indicate significant differences between groups conducted by one-way ANOVA, P < 0.05. neither ligand was bound by MBP-OsLecRK2-His recombinant protein (Figure 6b).
To confirm that MLGs participate in OsLecRK1-mediated defence signalling, we investigated the phosphorylation of the OsLecRK1 kinase in the presence or absence of MLG-based oligosaccharides in vitro. As shown in Figure 6c, compared to the mock control, BGTRIB and BGTETB stimulated phosphorylation of MBP-OsLecRK1-His recombinant protein, as detected by the anti-51-8 antibody. To examine whether MLGs activate OsLecRK1 in vivo, we treated protoplasts isolated from NIP with sterile water, 10 lM BGTRIB, and BGTETB, and harvested the protoplasts at 0, 0.5, 1, 2, and 3 h after treatment. Immunoblot analyses showed that the abundance of OsLecRK1-HA fusion protein increased upon BGTRIB and BGTETB exposure from 0.5 to 3 h and peaked at 0.5 or 1 h (Figure 6d), suggesting that MLGs rapidly activate the accumulation of OsLecRK1 in vivo.
To further examine the genetic relationship between OsCslF6 and OsLecRK1 during the rice-BPH interaction, we generated the F6OE4 oslecrk1/3 line by CRISPR-mediated knockout of OsLecRK1 and OsLecRK3 in the F6OE4 background. Phenotypic analyses showed that, in contrast to the improved growth of F6OE4 line, the 3-week-old seedlings of the F6OE4 oslecrk1/3 and oslecrk1/3 lines were smaller in size than the wild-type NIP plants under normal growth conditions (Figure 6e). Upon BPH feeding for 6 days, the oslecrk1/3 mutant was hypersensitive to BPH attack (Figure 6e), consisting with previous findings . In contrast, the F6OE4 oslecrk1/3 plants were also susceptible to BPH feeding, which abolished the enhanced tolerance seen in the F6OE4 line (Figure 6e). In addition, the enhanced resistance to WBPH of F6OE4 lines was abolished in F6OE4 oslecrk1/3, which was as susceptible to WBPH feeding as the oslecrk1/3 mutant (Figure 6e). These findings suggest that OsCslF6-mediated MLG biosynthesis acts upstream of OsLecRK1 in the rice response to  (13,16,18,26,32,35,and 57) and TT allele (60,65,75,92,103,111,and NIP) infested with BPH for 13 days. The survival rate was determined after 13 days of BPH treatment. Data are means AE SD (n = 3 biological replicates; 10 plants were used for each replicate). (c) MLG contents of rice subpopulations with the AA or TT allele under normal conditions. Data are means AE SD (n = 3 biological replicates). Letters indicate significant differences between groups conducted by one-way ANOVA, P < 0.05. pest infestation. Taken together, our results reveal that MLGderived oligosaccharides bind to and stimulate OsLecRK1 during rice-BPH interaction.

Discussion
The cell wall is primarily composed of polysaccharides including cellulose (b-1,4-D-glucose), callose (b-1,3-D-glucose), and MLGs. This structure serves as the first barrier defending plants against pest and pathogen invasion (Calderon-Cortes et al., 2012;Vorwerk et al., 2004). BPH is an herbivore that feeds on phloem sap in leaf sheaths through its stylet; therefore, the thickness and mechanical strength of cell walls are crucial for plant resistance to BPH (Guo et al., 2018). In response to planthoppers, callose synthase genes are upregulated in rice to enhance callose deposition in sieve tubes, preventing the insect from inserting its stylet and ingesting sap (Calderon-Cortes et al., 2012). If the insect manages to feed, the resulting cell wall fragments may act as DAMPs to stimulate cell surface-localized receptor proteins, thus initiating defence signalling (Erb and Reymond, 2019).
In this study, we discovered that the MLGs produced by OsCslF6 integrate both mechanisms by reinforcing cell wall thickness in vascular bundles and stimulating OsLecRK1-mediated defence signalling during the rice-BPH interaction. MLGs are a class of Poaceae-specific metabolites Ermawar et al., 2015), and it is conceivable that (like GSs in Arabidopsis) MLGs function downstream of JAs to help regulate defence responses against pests in plants of the Poaceae family, including rice.
Enhancing of physical barriers, including the sclerenchyma and vasculature, can prevent BPH stylets from reaching the phloem cells for feeding. In rice leaf sheaths, the sclerenchyma is located under the epidermis and above the vascular bundles. Bph3, Bph6, Bph9, Bph14, and Bph26, are strongly expressed in the vascular bundle cells (Du et al., 2009;Guo et al., 2018;Liu et al., 2015;Tamura et al., 2014;Zhao et al., 2016), although we do not know whether these genes regulate phloem strength in response to BPH feeding in rice. Bph30 is a rice BPH resistance gene encoding a novel protein with two leucine-rich domain that fortifies sclerenchyma cells to prevent BPH stylets from reaching the phloem (Shi et al., 2021), supporting the importance of sclerenchyma-mediated structural mechanisms in BPH resistance. Our results showed that OsCslF6-catalysed MLG deposition enhances BPH resistance by specifically modulating the cell wall thickness of vascular bundles, but not the sclerenchyma cells (Figure 2). Like Bph30, the OsCslF6-overexpressing lines are also resistant to WBPH (Figure 2a), indicating that fortifying the sclerenchyma or vascular bundle cells produces broad-spectrum pest resistance in rice.
Insect feeding or pathogen infection disrupts cell wall integrity; this mechanical damage is perceived by the plasma membranelocalized pattern-recognition receptors (PRRs; Reymond, 2021). In microbial infection, cell wall fragments are released into the extracellular space, where they act as DAMPs to bind to PRRs, eventually leading to immune responses and pathogen resistance (Gong et al., 2020;Gust et al., 2017;Reymond, 2021). OsLecRKs are cell surface PRRs conferring strong and broad-spectrum resistance to planthoppers in rice (Cheng et al., 2013a;Liu et al., 2015). Nevertheless, the biochemical mechanism by which OsLecRKs bind to cell wall components and contribute to DAMPmediated immunity responses is unknown.
Using MST and kinase activity assays, we established that the MLG-based oligosaccharides bind to recombinant OsLecRK1 protein with high affinities in vitro and activate its kinase activity in vivo (Figure 6a-d). Moreover, genetic analysis showed that deletion of OsLecRK1 and OsLecRK3 abolished the MLGmediated BPH and WBPH resistance phenotypes of the OsCslF6overexpressors (Figure 6e), suggesting that OsLecRK1 signalling is indeed downstream of MLG biosynthesis during the riceplanthopper interaction. Thus, our results imply that MLGderived oligosaccharides may act as DAMPs and are likely recognized by OsLecRK1, which induces downstream defence responses. However, it is still unclear whether the MLG-based oligosaccharides are cleaved by a specific rice endoglucanase, or an endoglucanase secreted in pest saliva. A recent study showed that during infection of rice, the fungal pathogen Magnaporthe oryzae secretes two endoglucanases, MoCel12A and MoCel12B, which release MLG-based oligosaccharides that are perceived by the OsCERK1 immune complex (Yang et al., 2021). To the best of our knowledge, no previous studies have identified salivary endob-1,3;1,4-glucanase in planthoppers, including BPH. Alternatively, MLG may function in rice-BPH interactions in an OsLecRK1independent manner; elucidating the signalling cascades downstream of OsLecRK1 will clarify whether this occurs. Thus, further investigation of the specific enzymes in distinct BPH subspecies and the functional kinase substrates of OsLecRK1 would increase our understanding of the molecular mechanisms underlying rice-BPH interactions.
Previous genome-wide association analyses identified 3502 SNPs and 59 loci associated with BPH resistance in rice , which may have important implications for the control of BPH. Here, we observed that SNP8 and SNP9 in the OsCslF6 promoter produced an AT-rich domain in the AA allele, but not in the AT or TT alleles (Figure 4a,b). This variation is likely associated with MLG deposition and BPH resistance in different cultivars selected from the 3K sequenced rice accessions ( Figure 5). It is worth noting that BPH cannot survive in temperate regions; the northern geographic limit of BPH overwintering is approximately 23-25°N (Jing et al., 2017;Xue et al., 2014;Zheng et al., 2021). Despite its capability for long-distance flight, the line between the overwintering vs. non-overwintering of BPH likely overlaps with the global distribution of the AA and TT alleles in the OsCslF6 promoter (Figure 4c,d), suggesting that SNP8/9 were positively selected to enhance MLG deposition and BPH resistance during the co-evolution of rice and BPH.
In multicellular eukaryotes, the AT-rich domain is a conserved DNA binding element targeted by ARID transcription factors and has a wide range of functions (Wilsker et al., 2005). In plants, the Arabidopsis AtARID1 and rice OsARID3 proteins play important roles in plant developmental processes, including sperm cell formation and shoot meristem development (Xu et al., 2015;Zheng et al., 2014). Given the importance of this AT-rich domain in modulating OsMYC2-activated OsCslF6 transcription (Figure 4f), we propose that other AT-rich domain-binding transcription factors, most likely OsARID proteins, are also involved in regulating MLG deposition in response to BPH attack. Consistent with this hypothesis, we showed that the transcripts of all six rice OsARID genes were significantly upregulated upon BPH feeding ( Figure S10), suggesting that the OsARID proteins may participate in BPH resistance, although their functional associations with OsCslF6 and MLG biosynthesis remain to be elucidated.
Finally, we propose a model for the role of OsCslF6-catalysed MLGs in regulating JA-mediated resistance in rice (Figure 7). According to our model, BPH feeding induces the accumulation of JAs, which are perceived by the co-receptors SCF OsCOI1 and OsJAZ. The ubiquitin ligase SCF OsCOI1 promotes OsJAZ ubiquitination and proteasomal degradation. The transcription factor OsMYC2, which is inhibited by OsJAZ proteins under normal growth conditions, activates the expression of OsCslF6 and the accumulation of MLGs to enhance cell wall thickness of vascular bundles, improving plant resistance to BPH (Figure 7). At the same time, the MLG-based oligosaccharides such as BGTRIA/B and BGTETB/C, which might be hydrolyzed by an unknown endoglucanase, are likely perceived by plasma membrane-localized OsLecRK1 (Figure 7) to activate in a key step in activating defence signalling during the rice-BPH interaction. MLG may also function in BPH defence responses in rice in an unknown OsLecRK1-independent manner. Moreover, natural variants of two alleles in the OsCslF6 promoter were correlated with MLGmediated BPH resistance. Based on biochemical analyses of these two SNPs, we propose that in response to BPH feeding, ARID  (Figure 7). Thus, our results suggest that CslF6 genes represent useful targets for molecular breeding to generate broad-spectrum pest-resistant cultivars of rice and other cereal crops.

Plant materials and growth conditions
The Oryza sativa cultivar Nipponbare (NIP) was used as the wild type and the background for genetic transformation in this study. Fourteen cultivated rice germplasm samples were used for natural variation analysis, including seven with TT alleles and seven with AA alleles (including NIP). The rice seeds were surface sterilized with 75% ethanol for 60 s, followed by 20% NaClO for 30 min. After eight rinses with sterile distilled water, the seeds were germinated on half-strength Murashige and Skoog (1/2 MS) medium at 28°C under a short-day photoperiod (10-h day/14-h night) with a light level of 1679 lux. Two-week-old seedlings were transferred to a greenhouse under a short-day photoperiod (10h day/14-h night) with a light level of 21 000 lux at 30°C.

Insect materials and BPH feeding assays
The BPH colony used in this study contained mixed biotypes of BPHs and was collected from rice fields in Guangdong. The insects were maintained in wooden cages (50 9 50 9 70 cm) with rice seedlings at 26-28°C under a 16-h light/8-h dark photoperiod. BPH feeding on rice was carried out as described previously (Guo et al., 2018). In brief, 4-week-old seedlings of different genotypes were infested with BPH. Plants were covered with a light-transmitting mesh, and 10 BPH third-instar nymphs per plant were released. BPH honeydew weights were measured as previously described (Chen et al., 2010). In brief, a Parafilm sachet was attached to the leaf sheath of each 4-week-old seedling, and a female BPH within 1 days after emergence was enclosed in the sachet. After 5 days, the insect was removed and the sachet was weighed.

Untargeted metabolomic profiling
For untargeted metabolomic profiling, metabolite extraction was performed as previously reported with minor modifications (Kang et al., 2019). Four biological replicates were collected for each genotype before and after BPH treatment. Freeze-dried samples were ground in liquid nitrogen, and each 100 mg sample of powdered plant tissue was extracted with 2 mL of precooled chloroform:methanol mixture (1:2) at 4°C overnight. After centrifugation, the supernatant was transferred to a new tube, and the extracts were dried under nitrogen gas. The dried samples were resuspended in 200 lL chloroform:methanol mixture (1:2) and filtered prior to LC-MS analysis.
LC-MS analysis was performed on an UPLC-ESI-Q-TOF system (UPLC, Shimadzu Nexera X2; MS, Sciex 5600 + Triple TOF). LC was Figure 7 Working model of the role of MLGs in regulating BPH resistance. BPH feeding induces the production of MLG-derived oligosaccharides such as BGTRIA/B and BGTETB/C, to activate plant defence responses by directly binding to the membrane-localized OsLecRK receptor kinases, or through an unknown OsLecRK-independent mechanism. Meanwhile, BPH triggers accumulation of jasmonates (JA-Ile), which are perceived by the co-receptors SCF OsCOI1a/1b (SCF COI1 ) and OsJAZ (JAZ). The ubiquitin ligase SCF COI1 promotes the ubiquitination of JAZ proteins and guides their proteasomal degradation. The transcription factor OsMYC2 (MYC2), which is inhibited by JAZ proteins under normal growth conditions, is released to activate the expression of OsCslF6 and the accumulation of MLGs. Natural variants in the AT-rich motif of the OsCslF6 promoter (SNP8/9) are positively associated with MLG deposition and BPH resistance. performed using a Dionex Acclaim C18 column (2.6 lm, 2.0 9 150 mm) with a mobile phase consisting of solvent A (ddH 2 O with 0.1% formic acid) and solvent B (acetonitrile with 0.1% formic acid). The gradient program was as follows: 0 min, 5% B; 0-5 min, 5% B-40% B; 5-20 min, 40% B-95% B; 20-22 min, 95% B; 22-25 min, 95% B-5% B; 25-30 min, 5% B. The flow rate was 0.35 mL/min; column temperature 40°C; injection volume 4 lL. The effluent was connected to the ESI-QTOF-MS system in TOF MS + IDA scan mode. The ESI source operation parameters were as follows: ion source, turbo spray; source temperature 550°C; ion spray voltage (IS) 5500 V (positive ion mode); ion source gas I (GSI), gas II (GSII), and curtain gas (CUR) were set to 55, 55, and 30 psi, respectively. The DP and CE values in TOF MS were set to 100 and 40 V, respectively, while the DP and CE values in IDA were set to 100 and 40 AE 15 V, respectively.
Data analysis was carried out using XCMS online software (Tautenhahn et al., 2012; https://xcmsonline.scripps.edu). Feature detection was performed using the centWave method. Retention time correction was performed using the obiwarp method. For alignments, Option was used with the following list parameters: retention time tolerance at 0.5 min, mass tolerance at 5 ppm. Statistical analysis was performed with a paired Welch t-test, and only features whose intensity increased or decreased >1.5-fold with P value < 0.01 were selected as significant features. Annotation and identification of features were based on m/z within 5 ppm tolerance for the XCMS-linked database METLIN search.

Quantitative analysis of MLGs
Endohydrolase-based measurement of MLG content was performed as previously described (Doblin et al., 2009). In brief, alcohol-insoluble residue (AIR) was prepared by grinding plant material in liquid N 2 with a mortar and pestle. Each 200-mg sample was combined with 10 mL of 50% ethanol and incubated in a boiling water bath for 5 min. Following centrifugation at 1800 g for 5 min, the supernatant was removed, and the residue was washed twice with 80% ethanol at room temperature, followed by two washes with 100% ethanol for 1 h. The ethanolsoluble fraction was removed, and the AIR was washed once with 100% ethanol and dried at À20°C under a vacuum for subsequent experiments. The AIR prepared during the early stage was resuspended in 5 mL NaPO 4 buffer containing 100 lL b-(1,3;1,4)-D-glucan endohydrolase (lichenase). The mixture was incubated for 15 min at 50°C with continuous mixing. The supernatant was collected (as the b-glucan hydrolase-released oligosaccharides) for HPAEC analysis and for quantitative analyses using a b-Glucan Assay kit (Megazyme, Dubliin, Ireland).

HPAEC analysis
The released oligosaccharides were separated by HPAEC on a CarboPacPA1 column (Dionex, Thermo Fisher Scientific, Waltham, MA, USA) equilibrated with 50 mM NaOAc in 0.2 M NaOH using a Dionex BioLC ICS 300 system (Dionex) equipped with a pulsed amperometric detector (PAD, Dionex, Thermo Fisher Scientific, Waltham, MA, USA). Oligosaccharides were eluted at 1 mL/min with a linear gradient of 50 mM NaOAc in 0.2 M NaOH to 350 mM NaOAc in 0.2 M NaOH over a 15 min period. Laminaribiose (Seikagaku, Tokyo, Japan) and cellodextrins (Sigma-Aldrich, Merck, Darmstadt, Germany) were run as standards.

ChIP and EMSA
Chromatin immunoprecipitation (ChIP) assays were performed as described (Yuan et al., 2017). Briefly, nuclear proteins were extracted from 2-week-old OsMYC2-OE transgenic rice leaf sheaths and stems. After coating with anti-HA (Abiocode, Agoura Hills, CA, USA; HY-K0201), the protein/DNA complexes were immunoprecipitated with Dynabeads Protein G (Invitrogen, Thermo Fisher Scientific, Waltham, MA, USA) for at least 4 h at 4°C. The precipitated DNA was purified using a DNA purification kit (Qiagen, Redwood City, CA, USA), and the enriched DNA fragments were subjected to qPCR using specific primers (Table S1). qRT-PCR was executed as follows: initial denaturation at 95°C for 5 min, followed by 40 cycles of PCR of denaturing at 95°C for 10 s and annealing at 60°C for 30 s. The OsACTIN1 promoter was used as a negative control.
For the electrophoretic mobility shift assay (EMSA), the amplified OsMYC2 CDS was fused in-frame with the His tags in vector pRSET-A. The recombinant His-OsMYC2 protein was expressed in Escherichia coli BL21 (DE3) cells and purified using Ni-NTA agarose (Invitrogen). All probes were labelled with biotin using a Biotin 3 0 End DNA Labeling Kit (Pierce, Thermo Fisher Scientific, Waltham, MA, USA). Unlabeled oligonucleotides were used as competitors in the binding assay. The oligonucleotide sequences of the probes are listed in Table S1. The EMSA reactions were performed using a LightShift Chemiluminescent EMSA Kit (Pierce) according to the manufacturer's instructions. An appropriate amount of His-MYC2 protein was incubated in binding buffer (50 ng/lL Poly(dIÁdC), 2.5% glycerol, 0.05% NP-40, 10 mM EDTA, and probes) in a total volume of 20 lL for 30 min at room temperature. After incubation, the binding reactions were loaded onto a 6% polyacrylamide gel and separated by polyacrylamide gel electrophoresis in 0.59 Tris-Borate-EDTA buffer at 4°C. The DNA-protein complex was transferred to a nylon membrane (Pierce). After cross-linking, biotin activity was detected according to the manufacturer's instructions of the EMSA kit.

MST assay
The MST assay was performed as described previously (Zhou et al., 2022). In brief, purified MBP-OsLecRK1-His and MBP-OsLecRK2-His proteins were labelled with reactive dyes using the Monolith RED-NHS (MO-L011). The concentration of labelled proteins was adjusted to 10 lM using the labeling buffer NHS buffer (pH 8.2) containing 130 mM NaHCO 3 and 50 mM NaCl. A serial dilution of various oligosaccharides (BGTRIA, BGTRIB, BGTETB, BGTETC, and CTE) ranging from 1.5 nM to 50 lM was prepared for mixing with the labelled proteins. Next, 90 lL of the prepared protein sample (concentration 10 lM) and 10 lL of the 300 lM dye solution were mixed by gently pipetting several times, then incubated for 30 min in the dark. The mixed samples were filled into standard treated capillaries and fluorescence scanning was performed on a Nano Temper Monolith NT.115 (50% MST power; 30% laser power), to determine binding affinities.

Phosphorylation assay
The in vitro phosphorylation assay was carried out as described by Zhou et al. (2022). MBP-OsLecRK1-His Recombinant proteins were incubated with the substrate oligosaccharides BGTRIB and BGTETB (10 lM) in reaction buffer (20 mM HEPES pH 7.4, 0.15 M NaCl, 10 mM MgCl 2 and 3 lL of 10 mM N 6 -substituted ATPcS) at 25°C for 15 min. Thereafter, the mixed solution was supplemented with 1.5 lL of 50 mM PNBM at 25°C for 1 h, followed by the addition of SDS loading buffer for stopping reaction. Finally, phosphorylated MBP-OsLecRK1-His protein was detected with a thiophosphate ester specific antibody 51-8 by immunoblot assay, whereas MBP-OsLecRK1-His was detected with anti-His antibody as a loading control.

Dual-luciferase reporter assay
To generate the effector and reporter constructs, the pUC119 and pGreenII-0800-LUC vectors were used. All primers used to generate constructs for the transient transactivation assays are listed in Table S1. The transient assays were performed in rice protoplasts via PEG-mediated transfection, and firefly LUC and Renilla LUC (REN) activities were analysed using the Dual-Luciferase reporter Assay System (Promega, Madison, Wisconsin, USA) and the SpectraMax i3x Multi-Mode Detection Platform (Molecular Devices, Shanghai, China) according to the manufacturer's manual. Relative LUC activity is expressed as the ratio of LUC to REN.

Statistical analysis
Statistical analyses throughout the study were carried out using GraphPad 8.3 software. The significance of differences among different groups was determined by one-way ANOVA with a Tukey's HSD test. The different letters above the figures indicate the statistical significance (P < 0.05). The numbers of samples are indicated in the figure legends.

Supporting information
Additional supporting information may be found online in the Supporting Information section at the end of the article.
Appendix S1 Additional description of methods. Dataset S1 Targeted metabolomic profiling data. Dataset S2 Un-targeted metabolomic profiling data. Figure S1 BPH feeding induces the accumulation of jasmonates. Figure S2 Generation and identification of oslox2 and oscoi1 mutants. Figure S3 Jasmonates contribute to resistance to BPH feeding in rice. Figure S4 Widely targeted metabolomic profiling showing the differential flavonoid contents in the JA-related mutants upon BPH feeding. Figure S5 Spatial and temporal expression patterns of members of the MLG synthetase OsCslF gene family. Figure S6 Generation and phenotypic analysis of OsCslF6 knockout and overexpression transgenic lines. Figure S7 OsMYC2 participates in the plant response to BPH infestation. Figure S8 OsMYC2 does not interact with the OsCslF3, OsCslF4, OsCslF7, OsCslF8, or OsCslF9 promoters in vivo. Figure S9 Overexpression of OsMYC2 enhances plant resistance to BPH infestation. Figure S10 Expression of the OsARID gene family upon BPH infestation. Figure S11 Generation of OsLecRK1 and OsLecRK3 double knockout lines.