Silencing a Simple Extracellular Leucine-Rich Repeat Gene OsI-BAK1 Enhances the Resistance of Rice to Brown Planthopper Nilaparvata lugens

Many plant proteins with extracellular leucine-rich repeat (eLRR) domains play an important role in plant immunity. However, the role of one class of eLRR plant proteins—the simple eLRR proteins—in plant defenses against herbivores remains largely unknown. Here, we found that a simple eLRR protein OsI-BAK1 in rice localizes to the plasma membrane. Its expression was induced by mechanical wounding, the infestation of gravid females of brown planthopper (BPH) Nilaparvata lugens or white-backed planthopper Sogatella furcifera and treatment with methyl jasmonate or abscisic acid. Silencing OsI-BAK1 (ir-ibak1) in rice enhanced the BPH-induced transcript levels of three defense-related WRKY genes (OsWRKY24, OsWRKY53 and OsWRKY70) but decreased the induced levels of ethylene. Bioassays revealed that the hatching rate was significantly lower in BPH eggs laid on ir-ibak1 plants than wild-type (WT) plants; moreover, gravid BPH females preferred to oviposit on WT plants over ir-ibak1 plants. The exogenous application of ethephon on ir-ibak1 plants eliminated the BPH oviposition preference between WT and ir-ibak1 plants but had no effect on the hatching rate of BPH eggs. These findings suggest that OsI-BAK1 acts as a negative modulator of defense responses in rice to BPH and that BPH might exploit this modulator for its own benefit.


Introduction
When attacked by herbivorous insects, plants recognize patterns associated with damage and with herbivores via pattern recognition receptors (PRRs) and respond by activating a series of early signaling events: increasing cytosolic Ca 2+ concentrations, triggering a burst of reactive oxygen species (ROS) and activating mitogen-activated protein kinase (MPK) cascades [1,2]. These early events trigger signaling pathways mediated by phytohormones consisting mainly of jasmonic acid (JA), ethylene (ET) and salicylic acid (SA), which reorganize the transcriptome, proteome and metabolome, respectively, of the plant and thereby enhance plants' resistance to herbivores [3][4][5][6].
Simple eLRR proteins have also been reported to regulate the defense responses of plants to pathogens. For example, some simple eLRR genes, such as SlLRP in Solanum lycopersicum [25], SbLRR and SbLRR2 in Sorghum bicolor [26,27] and NtLRP1 and NtLRR1 in Nicotiana tabacum [28,29], are responsive to pathogen infections. In pepper Capsicum annuum, a simple eLRR protein, CaLRR1, interacts with a hypersensitive-induced reaction (HIR) protein, CaHIR1, and two pathogenesis-related proteins, CaPR10 and CaPR4b, and then regulates the plant defense responses [30]. A simple eLRR protein in rice, OsLRR1, whose transcript levels are induced by pathogen infection, wounding and treatment with SA or JA, enters the endosomal pathway and interacts with plasma membrane-localized OsHIR1; the ectopic expression of OsLRR1 in Arabidopsis thaliana enhances the resistance of plants to Pseudomonas syringae pv. tomato DC3000 (Pst DC3000) by increasing the transcript levels of the defense genes related to the SA (AtPR1 and AtPR2)and JA (AtPDF1.2)mediated pathways [7,9]. Expressing OsLRR1 in Chinese cabbage Brassica rapa L. subsp. pekinensis enhances plants' resistance to bacterial soft rot by activating defense-related genes (PDF1, BrPGIP2, glucanase genes MERI5B and Meri-5 and hypersensitive cell death marker gene Athsr3) [7]. Moreover, OsLRR1 is required for XA21-mediated rice defense responses to Xoo [31]. However, whether and how simple eLRR proteins regulate plants' defenses to herbivores remains largely unknown.
Rice (Oryza sativa L.), one of the most important food crops in the world, suffers heavily from predation by many herbivorous insects. Brown planthopper (BPH) Nilaparvata lugens (Stål), a major hemipteran insect pest, damages rice mainly by feeding on phloem sap, laying eggs in plant tissues and transmitting viruses [32]. Previous studies have revealed that, when rice plants are infested by BPH, they respond by activating a complex signaling network, primarily MPK cascades and pathways mediated by JA, SA, ethylene and hydrogen peroxide (H 2 O 2 ) [33][34][35][36][37][38]; the defense responses, such as the accumulation of trypsin proteinase inhibitors (TrypPIs) and the release of herbivore-induced plant volatiles (HIPVs), may influence the performance of conspecific and non-conspecific herbivores directly, as well as indirectly, by regulating the role of the natural enemies of herbivores [22,[36][37][38][39][40][41][42]. Although some eLRR proteins play an important role in regulating rice plants' defense responses to pathogens and herbivores, as stated above, whether simple eLRR proteins are involved in the defense responses of rice to herbivores remains unclear.
In this study, we identified an herbivore-induced simple eLRR protein gene in rice, OsI-BAK1, which responds to infestation by gravid BPH females [41]. To investigate its role in BPH-induced defenses in rice, we obtained stably transformed rice plants rendered deficient in OsI-BAK1 by RNA interference (RNAi). Combining molecular biology, phytohormone analyses and bioassays, we demonstrate that OsI-BAK1 positively regulates the biosynthesis of BPH-induced ET and decreases the resistance of rice to BPH.

Isolation and Characterization of OsI-BAK1
The transcriptome data from rice plants infested by gravid females of BPH showed that the transcript level of a simple eLRR gene was significantly upregulated [41]. By cloning this gene, we found that its sequence was 100% identical to the previously reported OsI-BAK1 (brassinosteroid insensitive 1-associated kinase 1) [43]. The sequence analysis found that OsI-BAK1 includes an open reading frame of 657 bp; this frame, which encodes a protein of 219 amino acids with a predicted molecular weight of 23.64 kDa, contains a signal peptide and five LRR motifs (Figures 1a and S3). BLASTP searches in the NCBI showed that OsI-BAK1 shares a high similarity with simple eLRR proteins, such as ObLRR1like (96.92% identity) in Oryza brachyantha, BdLRR1 (94.76%) in Brachypodium distachyon, AtLRR2 (94.76%) in Aegilops tauschii subsp. Strangulate and TdLRR2-like-1 (87.61%) and TdLRR2-like-2 (87.61%) in Triticum dicoccoides (Figure 1b).
A subcellular localization assay using a fusion gene (OsI-BAK1 with green fluorescent protein) showed that the fluorescent signal emitted from the fusion protein was observed on the plasma membrane; this signal overlapped with the fluorescent signal emitted from the membrane-localized positive control protein AtPIP2A (Figure 1c). The results suggest that OsI-BAK1 was a plasma membrane-localized eLRR protein and may play a role in signal recognition and transduction.

Herbivore Infestation, Wounding and JA and Abscisic Acid Treatment Enhance Transcript Levels of OsI-BAK1
Compared with the corresponding controls, all of the measured treatments-BPH infestation, white-backed planthopper (WBPH) infestation, mechanical wounding and treatment with methyl jasmonate (MeJA) or abscisic acid (ABA)-enhanced the transcript level of OsI-BAK1 in plants ( Figure 2). Unlike BPH infestation and mechanical wounding, both of which induced the expression of OsI-BAK1 quickly starting 1.5 h after treatment (Figure 2a,b), the other three treatments induced the expression of OsI-BAK1 relatively later, starting 8 h after treatment (Figure 2c-e). Moreover, the effect of mechanical wounding on the expression of OsI-BAK1 lasted for less time, peaking at 12 h and returning to control levels at 48 h, compared to the effect of other treatments on the expression of OsI-BAK1 (in these, the expression of OsI-BAK1 peaked 48 h after treatment).

Silencing OsI-BAK1 in Rice
To investigate the function of OsI-BAK1 in herbivore-induced defense responses in rice, we obtained two homozygous OsI-BAK1-silenced lines (irIBAK1-1 and irIBAK1-2) with a single insertion ( Figure S4a, See Supplementary Materials). The transcriptional analysis showed that the transcript level of OsI-BAK1 in the two ir-ibak1 lines was only 6.9-8.5% of that in WT plants 0 and 8 h after BPH infestation (Figure 3a).
Based on the database from the Rice Genome Annotation Project (http://rice.plantbiology. msu.edu/index.shtml; accessed on 29 October 2019), five genes shared high similarities in nucleotide sequences with OsI-BAK1: LOC_Os01g59440 (92.31% identity), LOC_Os11g31540 (90.00%), LOC_Os01g17250 (89.51%), LOC_Os11g31530 (89.09%) and LOC_Os05g41230 (82.42%). Among these five genes, two (LOC_Os11g31530 and LOC_Os05g41230) showed such low transcript levels that their transcripts were not detected; the other three genes were not co-silenced by the RNAi construct in the transgenic plants ( Figure S4b-d). Therefore, the specificity of the RNAi sequence was high. Compared with WT plants, ir-ibak1 plants showed only a slight difference in growth phenotypes (Figure 3b   Mean transcript levels (+SE, n = 5) of OsI-BAK1 in rice leaf sheaths that were mechanically wounded (W, (a)), infested with gravid brown planthopper (BPH) females (b), gravid white-backed planthopper (WBPH) females (c) or treated with methyl jasmonate (MeJA) (d) or abscisic acid (ABA) (e). Noninfested, plants that were covered in an empty glass cylinder without BPH or WBPH; Con, control plants corresponding to mechanical wounding or MeJA or ABA treatment. Asterisks represent significant differences between the treatments and controls at each time point (* p < 0.05 and ** p < 0.01; Student's t-tests).

OsI-BAK1 Negatively Regulates the Transcript Level of Defense-Related WRKYs
MPK cascades and WRKY transcription factors (TFs) are key regulators in plant defense responses [44,45]. Thus, we detected the kinase activity of two MPKs, OsMPK3 and OsMPK6, and the transcript levels of OsMPK3 and OsMPK6 [46,47], as well as of three WRKY genes (OsWRKY24, OsWRKY53 and OsWRKY70), all of which have been reported to play an important role in the defense responses of plants to pathogens or herbivores [33,34,48,49] in ir-ibak1 and WT plants. The results showed that the transcript levels of OsMPK3 were higher in the ir-ibak1 lines than in WT plants at 1 h after BPH infestation (Figure 4a), whereas no difference was found in the transcript levels of OsMPK6 between WT and ir-ibak1 plants before or after BPH infestation ( Figure S5a). Although BPH infestation induced the kinase activation of OsMPK3, it did not induce the kinase activation of OsMPK6; no difference in the kinase activation of both OsMPK3 and OsMPK6 was observed between the WT and ir-ibak1 lines ( Figure S5b-d). Silencing OsI-BAK1 enhanced not only the BPH-induced transcript levels of OsWRKY24, OsWRKY53 and OsWRKY70 but, also, the constitutive transcript levels of OsWRKY53 (Figure 4b-d).

Silencing OsI-BAK1 Reduces the Accumulation of BPH-Induced Ethylene
The JA-, SA-, ET-, ABA-and H 2 O 2 -mediated signaling pathways regulate the resistance of rice to BPH [36,39,50]. Therefore, we investigated the changes in the concentrations of these six signal molecules in WT and ir-ibak1 plants before and after they were infested by BPH. A remarkable decrease in the ET levels in transgenic plants compared with WT plants was observed when plants were infested with gravid females of BPH for 24, 48 or 72 h (Figure 5a). Consistently, the transcript levels of two genes related to ET biosynthesis (OsACO1 and OsACS1) [51,52] and of one transcription factor, OsERF2, which might be related to the ET pathway [53], decreased in the ir-ibak1 lines compared with WT plants (Figure 5b-d). However, although BPH induced the accumulation of JA, JA-Ile, ABA and H 2 O 2 in plants, no difference in the levels of these signal molecules was observed between the WT and ir-ibak1 plants, except for the constitutive level of H 2 O 2 , which was 1.3-fold higher in irIBAK1-1 plants than in WT plants ( Figure S6a-d). Moreover, silencing OsI-BAK1 did not influence the SA accumulation in plants ( Figure S6e).

Silencing OsI-BAK1 Enhances the Resistance of Rice to BPH
The bioassays revealed that gravid BPH females preferred to lay eggs on WT plants over ir-ibak1 plants: the number of eggs laid on irIBAK1-1 and irIBAK1-2 plants was only 66.38% and 66.52% of the number laid on WT plants (Figure 6a,b). Moreover, the hatching rate of BPH eggs laid on ir-ibak1 plants was significantly lower than that of BPH eggs laid on WT plants (Figure 6c). However, silencing OsI-BAK1 did not influence the survival rate of BPH nymphs (Figure 6d).  To investigate whether the enhanced resistance of ir-ibak1 plants to BPH was due to their reduced emission of ET, we treated ir-ibak1 plants by spraying them with ethephon and carried out the bioassays as above. The data showed that the decrease in preference of BPH for ir-ibak1 plants was abolished when ir-ibak1 plants were individually sprayed thoroughly with 500-µM or 1-mM ethephon (Figure 7). However, spraying ethephon on ir-ibak1 plants with the same concentration did not affect the hatching rate of BPH eggs ( Figure S7).

Discussion
Simple eLRR proteins have been reported to play an important role in the defense responses of plants to pathogens by influencing the defense-related signaling pathways [9,30,54]. In this study, we found that a simple eLRR gene in rice, OsI-BAK1, is induced by BPH or WBPH infestation, mechanical wounding and treatment with MeJA or ABA. Silencing OsI-BAK1 increased the transcript levels of defense-related genes OsMPK3, OsWRKY24, OsWRKY53 and OsWRKY70 and decreased the BPH-induced ET levels, which subsequently enhanced the resistance of rice to BPH. These results demonstrate that OsI-BAK1 functions as an important negative regulator in the defense responses of rice to BPH.
Simple eLRR genes have been reported to exhibit specific expression patterns. For example, some simple eLRR genes, such as SlLRP in Solanum lycopersicum [25], SLRR and SbLRR2 in Sorghum bicolor [26,27] and OsLRR1 in rice [9], are reportedly induced by pathogen infection. Some simple eLRR genes are responsive to abiotic stresses: OsLRR1 and CaLRR1 are responsive to mechanical wounding [9,55], and NtLRR2 and CaLRR1 are responsive to salt treatment [29,55]. Moreover, some simple eLRR genes respond differently to phytohormone treatment. In pepper, for example, CaLRR1 is induced by ABA treatment but not by treatment with SA, MeJA or ET [55], whereas SbLRR2 is strongly induced by MeJA but not by SA or 1-aminocyclopropane-1-carboxylic acid (ACC) [27]. In addition, the transcript levels of several simple eLRR genes are reported to be tissue-specific. For example, NtLRR2 is mainly expressed in roots rather than in stems and leaves [29]. LRRop-1 from Arabidopsis is enriched in seeds and rosette leaves [56]. We observed that OsI-BAK1 was induced by mechanical wounding, infestation with BPH and WBPH and treatment with MeJA and ABA ( Figure 2). Together, these data suggest that each simple eLRR gene is expressed in particular tissues and responds differently to biotic and abiotic stresses, thereby playing different roles in plant growth, development and defenses against various stresses.
The subcellular localization of eLRR proteins has been researched extensively. Among these, LRR-RLKs and LRR-RLPs are plasma membrane-localized proteins [57,58], whereas PGIP-like proteins and leucine-rich repeat extensin-like proteins are cell wall-localized [59,60]. The subcellular localization of simple eLRR proteins is complicated. CaLRR1 in pepper, for instance, is localized to the cytoplasm and plasma membrane and can be secreted into the apoplastic space [30]. OsLRR1 is localized to the endosomal pathway [9]. These proteins may also be localized to the endoplasmic reticulum [28,56] or extracellular space [61]. Like the subcellular localization of NtLRR2 [29] and CaLRR1 [30], both of which play an important role in the resistance of plants to pathogens, OsI-BAK1 is localized to plasma membranes (Figure 1c). This result demonstrates that OsI-BAK1, in addition to its functions in grain filling, leaf development [43] and herbivore resistance reported in this study, may also play a role in plants' resistance to disease. This hypothesis needs to be confirmed in the future.
Both MPKs and WRKY transcription factors play a central role in herbivore-induced plant defense responses via modulating defensive signaling pathways, such as those mediated by JA, SA and ethylene [47,62]. Moreover, MPKs and WRKY transcription factors can regulate each other at the transcriptional and translational levels [34,48]. In this study, although silencing OsI-BAK1 increased the transcript level of OsMPK3, it did not influence the activity of OsMPK3 or the transcript level and activity of OsMPK6 (Figures 4a and S5). However, we did observe that silencing OsI-BAK1 upregulated the transcript levels of OsWRKY24, OsWRKY53 and OsWRKY70 (Figure 4b-d). Moreover, silencing OsI-BAK1 decreased the BPH-induced levels of ET (Figure 5a) but not JA, JA-Ile, ABA, SA and H 2 O 2 ( Figure S6). Given that both OsWRKY53 and OsWRKY70 regulate herbivore-induced levels of ET in rice [33,34,48], we think that the decrease in BPH-induced ET levels in ir-ibak1 plants is probably related to the change in transcript levels of OsWRKY53 and OsWRKY70. Why the levels of these WRKY transcripts and of ET in ir-ibak1 plants changed bears further study.
The ET-mediated signaling pathway plays an important role in the resistance of plants to herbivores [1,6]. In rice, it has been reported that the ET pathway negatively regulates the resistance of plants to BPH [36]: BPH preferred to feed and oviposit on the WT plants over plants with ET deficiency; moreover, in BPH, the levels of honeydew excretion were reduced, and in plants with ET deficiency, natural enemies became more attractive. Here, we observed a similar result: BPH preferred to oviposit on WT plants over ir-ibak1 plants. In addition, the exogenous application of ethephon on the ir-ibak1 lines abolished the difference in BPH oviposition preference (Figure 7), suggesting that the decreased oviposition preference of BPH for ir-ibak1 plants depends on the ET pathway (as we found previously) [36]. Interestingly, we did not find that the exogenous application of ethephon on ir-ibak1 plants restored the hatching rate of BPH eggs ( Figure S7), demonstrating that other signaling pathways or components regulated the survival of BPH eggs laid on ir-ibak1 plants. It will be interesting to identify these pathways or components in the future.
ET has been reported to negatively regulate root growth [63]. Moreover, in rice, the overexpression of OsERF2 reduces the root length [53]. The root length in ir-ibak1 plants was longer and the levels of ET and OsERF2 transcripts were lower than in WT plants. Thus, the longer root length phenotype exhibited in ir-ibak1 plants compared to WT plants may also be related to the impaired ET signaling pathway. Whether other components are involved in this process remains to be elucidated.
In summary, our results show that a plasma-membrane-localized simple eLRR protein, OsI-BAK1, functions as a negative regulator in the defense response of rice to BPH. BPH can benefit by exploiting this regulator, i.e., inducing the expression of OsI-BAK1, which then decreases the transcript levels of OsMPK3, OsWRKY24, OsWRKY53 and OsWRKY70 and increases the BPH-induced ET levels, changes that promote a susceptibility to BPH in the host plant. These findings demonstrate how BPH regulates host plant resistance, namely by influencing an early signaling factor.

Plant Growth and Insects
In this study, the rice (Oryza sativa L.) genotypes were Xiushui 11 (wild-type, WT) and transgenic lines with silenced OsI-BAK1 (ir-ibak1 lines) (see below). Pregerminated seeds of different lines were cultured in plastic culture dishes (diameter 90 mm, height 15 mm) in an incubator at 28 ± 2 • C with a 14-h light period. Ten-day-old seedlings were transferred to hydroponic boxes (20 L) containing a rice nutrient solution [64] and kept in a greenhouse (28 ± 2 • C, 14-h light, 55-65% relative humidity). Twenty to twentyfive days later, plants were transplanted into individual 300-mL hydroponic plastic pots containing a nutrient solution ( Figure S1). Plants were used for different experiments 4 d after transplantation. Colonies of BPH and white-backed planthopper (WBPH) Sogatella furcifera were originally collected from rice fields in Hangzhou, China and maintained on plants of TN1, a susceptible rice variety, for more than 10 years in a climate chamber (26 ± 2 • C, 14-h light, 65% relative humidity).
The homologs of OsI-BAK1 from other plant species were identified using BLASTP in the NCBI website (https://blast.ncbi.nlm.nih.gov/Blast.cgi; accessed on 10 September 2021). The full-length amino acid sequences of OsI-BAK1 were set as query sequences and blasted with the default parameters. The amino acid sequences were downloaded from the website and aligned by ClustalW in MEGA-X (pairwise alignment: gap opening penalty 10.00, gap extension penalty 0.10; multiple alignment: gap opening penalty 10.00, gap extension penalty 0.20; use negative matrix: off, delay divergent cutoff 30%) [65]. The alignment results were used to construct a neighbor-joining tree with default parameters similar to a previous study (scope: all selected taxa; statistical method: neighbor-joining; test of phylogeny: bootstrap method, 1000 times of replications; substitutions type: amino acid; model/method: Poisson model; rates among sites: uniform rates; pattern among lineages: same (homogeneous); gaps/missing date treatment: complete deletion; number of threads 3) [22].

Subcellular Localization Assay
For subcellular localization, the full length of OsI-BAK1 was inserted into pCAM-BIA1301 and fused with green fluorescent protein (GFP), yielding a transformation plasmid vector ( Figure S2a). Protoplast isolation and polyethylene glycol (PEG)-mediated transformation was performed following a method described previously [66]. Briefly, rice protoplasts were prepared from leaf sheaths of seedlings of rice variety Nipponbare at a concentration of 2 × 10 6 cells/mL. Ten micrograms of plasmid DNA (pCAMBIA1301-OsI-BAK1-GFP) was mixed with 100 µL of protoplasts and 110 µL of 40% PEG-calcium transfection solution (40% (w/v) PEG4000 in ddH 2 O containing 0.2 M of mannitol and 100 mM of CaCl 2 for 5 min. The reaction was stopped by adding 400 µL of W5 solution (0.1% glucose, 0.08% KCl, 0.9% NaCl, 1.84% CaCl 2 ·2H 2 O and 2-mM MES·KOH (pH 5.7)). The solution was centrifuged, the supernatant was removed and the remainder was incubated in the dark for 16-20 h at 26 • C. Transformed protoplasts were imaged under a confocal microscope (LSM 800, Zeiss, Oberkochen, Germany). The plasmid pBIN-AtPIP2A-mCherry [22] was used as a positive control for membrane localization. The chemical compounds mentioned above were all obtained from Sigma-Aldrich (Darmstadt, Germany).

RNA Extraction and Real Time-qPCR
In these experiments, five independent biological samples were used. Total RNA was extracted from 100 mg of rice leaf sheaths using the MiniBEST Plant RNA Extraction Kit (TaKaRa, Dalian, China). Five hundred nanograms of each total RNA sample was reversetranscribed by using PrimeScript™ RT Master Mix (TaKaRa, Dalian, China) according to the manufacturer's protocols. The real-time qPCR assay was performed on the CFX96 Real-Time System (Bio-Rad, Richmond, CA, USA), using TB Green ® Premix Ex Taq™ II or Premix Ex Taq™ (Probe qPCR) (TaKaRa, Dalian, China) following the manufacturer's protocols. A linear standard curve, threshold cycle number versus log 10 (designated transcript level), was built using a series of concentrations of a specific cDNA standard, and the relative expression levels of the transcripts of different genes were determined according to the standard curve. A rice actin gene, OsACTIN (TIGR ID: LOC_Os03g50885), was used as an internal standard to normalize the cDNA concentrations. All primers and probes of the target genes used for real-time qPCR are listed in Table S1.

Generation and Characterization of Transgenic Plants
To prepare the RNAi construct, a 520-bp fragment of OsI-BAK1 was cloned into the pCAMBIA1301-RNAi vector ( Figure S2b). This vector was inserted into rice variety Xiushui 11 using Agrobacterium tumefaciens-mediated transformation. Rice transformation, screening of the homozygous T 2 plants and identification of the number of insertions by Southern blot were performed following a method described previously [38]. The silencing efficacy of OsI-BAK1 in the ir-ibak1 lines was performed using real-time qPCR with primers irIBAK1-F (5 -GGCCTTCACCCTCTTCTACT-3 ) and irIBAK1-R (5 -GTCGTAGGTAGAGGCTCTGG-3 ). Two homozygous T 2 lines (irIBAK1-1 and irIBAK1-2) with a single insertion were selected and used in the following experiments.

Measurement of Plant Growth Parameters
Plant growth parameters of the 30-and 40-day-old WT plants and of the ir-ibak1 plants were measured. These plant growth parameters included the root length, shoot height, plant mass and chlorophyll content. The root length was measured from the shoot base to the longest root tip. The shoot height was measured from the shoot base to the longest leaf apex. The whole weight of the plant was regarded as the plant mass. For the chlorophyll content determination, the youngest three fully expanded leaves from each plant and three locations (at the tip, middle and base of the leaf) from each leaf were measured by a chlorophyll meter, SPAD-502 Plus (Tokyo, Japan). The experiments above were replicated 10 times.

BPH Bioassay
To investigate the effect of OsI-BAK1 on the hatching rate of the BPH eggs, the WT and ir-ibak1 plants were individually confined in glass cages, and 15 gravid BPH females were introduced into each cage ( Figure S1). Twelve hours later, the females were removed, and the number of newly hatched nymphs on each plant was counted every day until no newborn nymphs appeared for three consecutive days. The unhatched eggs in each plant were counted under a microscope to calculate the hatching rate. The experiments for each line were replicated 10 times.
To assess the influence of silencing OsI-BAK1 on the survival rate of BPH nymphs, the WT and ir-ibak1 plants were individually confined in glass cages, and 15 newly hatched BPH nymphs were released into each cage ( Figure S1). The surviving nymphs on each plant were recorded daily until all of them became adults. The experiments for each line were replicated 10 times.
To observe the impact of OsI-BAK1 on the BPH oviposition preference, pots with two plants-an irIBAK1-1 plant and a WT plant or an irIBAK1-2 plant and a WT plant-were individually confined in glass cages (see Figure S1 but with two plants). Fifteen gravid BPH females were released into each cage. Two days later, the insects were removed, and the number of eggs on each plant was counted under a microscope. The experiment was replicated 10 times.

Detection of MPK Activation
Plants of the WT and two ir-ibak1 lines were randomly assigned to BPH and control treatments. The outermost part of two leaf sheaths that were infested by gravid BPH females was harvested at 0, 1 and 3 h after BPH infestation. The experiments for each treatment at each time point were replicated five times. The sample containing five replications was mixed and ground together, and 100 mg of each sample was used to extract the total proteins and detect the MPK activation according to a method described previously [39]. The quantitative results were calculated after repeating this experiment for four times using ImageJ (https://imagej.nih.gov/ij/; accessed on 16 July 2021). The relative kinase activation of OsMPK3 and OsMPK6 was calculated by taking the signal intensity of the WT sample without BPH infestation (at 0 h) as 1 for normalization.

SA, JA, JA-Ile, ABA and H 2 O 2 Analysis
Plants of the WT and two transgenic lines were randomly assigned to BPH and control treatments. The outermost part of two leaf sheaths that were infested by gravid BPH females was harvested at 0, 8, 24 and 48 h after BPH infestation. The SA, JA, JA-Ile and ABA were extracted with ethyl acetate spiked with labeled internal standards ( 2 D6-JA, 2 D6-JA-Ile, 2 D4-SA and 2 D6-ABA) and analyzed using a HPLC/mass spectrometry/mass spectrometry system [22]. For the H 2 O 2 analysis, the outermost part of two leaf sheaths was collected at 0, 8, 12 and 24 h after BPH infestation. The H 2 O 2 concentration of each sample was measured using the Amplex Red Hydrogen Peroxide/Peroxidase Assay Kit (Invitrogen, Eugene, OR, USA) according to the instructions. The experiments for each treatment at each time point were replicated five times.

Ethylene Analysis
Plants of the WT and two ir-ibak1 lines, each with 15 gravid BPH females, were individually covered in sealed glass cylinders (diameter 4 cm, height 50 cm). ET production was measured at 24, 48 and 72 h after BPH infestation by gas chromatography according to the method [36]. The experiments for each line at each time point were replicated eight times.

Data Analysis
Two-treatment data were analyzed using Student's t-tests. Data from three or more treatments were compared using one-way ANOVA followed by Tukey's HSD post-hoc tests. All tests were carried out with SPSS software version 18 (IBM Corp., Armonk, NY, USA).