A tripartite rheostat controls self-regulated host plant resistance to insects – Nature

Plant materials and growth conditions

Nipponbare (designated as N14) is a model O. sativa japonica rice variety that contains the N14-susceptible allele of Bph14. N14 was used in early experiments to establish the principles of BISP–BPH14 interactions. The RI35 line (designated as Bph14) is a recombinant inbred line that contains the BPH resistance gene Bph14 (ref. ⁸). Minghui 63 (MH63) is the BPH-susceptible parent of RI35 (ref. ⁴⁴) and is a model variety for O. sativa indica rice breeding and genomics. MH63 and BPH14 were used for most studies to elucidate BISP–BPH14–NBR1 interactions owing to the agronomic importance of MH63. The following transgenic lines were developed in this study: N14–Bisp (pUBI::Bisp-Myc, N14 recipient), Bph14–Bisp (pUBI::Bisp–Myc, Bph14 recipient), Osrlck185 (OsRLCK185–Cas9, ZH11 recipient), Bph14–Osnbr1 (OsNBR1–Cas9, Bph14 recipient). The Osrlck185 mutants were generated in the ZH11 background⁴⁵; ZH11 is a readily transformable O. sativa japonica rice that BPH-susceptible genotype.

Insect materials and growth conditions

The BPH insects were maintained on susceptible cultivar Taichung Native 1 at 28 ± 2 °C with 60–80% relative humidity and a photoperiod of 16 h of light–8 h of dark. RNAi of BPH by dsRNA injection was developed as part of this study: GFP-RNAi (microinjected with dsGFP) and Bisp-RNAi (microinjected with dsBisp).

BPH bioassays on rice plants

To evaluate BPH resistance of transgenic and WT rice plants, sets of around 15 or 20 seeds from the same plant were sown in a plastic cup (10 cm in diameter, 20 cm in height). At the three-leaf stage, all plants were placed under a large gauze cover, and each seedling was infested with ten second- or third-instar BPH nymphs. When all of the susceptible plants had died (scored as 9), each seedling of the other cultivars or lines was given a score of 0, 1, 3, 5, 7, or 9 according to the degree of damage, as previously described^8,46,47, and the BPH resistance score was calculated. At least three replicates were used for each cultivar or line.

To measure the weight gain and honeydew excretion of BPHs, a newly emerged short-winged female adult and a Parafilm sachet were weighed using a 1/100,000 Shimadzu AUW120D electronic balance^11,12. The insect was placed in the pre-weighed sachet, which was then attached to the leaf sheath near the lower part of a rice plant. After 48 h, the insect and Parafilm sachet were weighed again. The difference between the first and second measurements of the weight of the insect was recorded as the BPH weight gain; the difference between the first and second measurements of the weight of the sachet was recorded as amount of honeydew excreted. At least ten replicates were used for each cultivar or line.

For the two-host choice tests, a transgenic plant and a WT plant were grown in a plastic cup as described elsewhere^10,11,12. At the four-leaf stage, the cup was covered with gauze, and 20 or 30 second- or third-instar nymphs were released into the cup (the exact number depended on the BPH resistance of the WT plant). The number of nymphs that settled on each plant was recorded at 6, 12, 24, 48, 72, 96, and 120 h after release (the moment of release was considered time 0). Ten cups were analysed for each cultivar or line examined.

To study BPH nymph survival on rice, second- or third-stage BPH nymphs were released (ten insects per plant), and the cups were covered with a light-transmitting mesh. The nymphs on each plant were counted 8 or 9 days after release. Survival rates were calculated as the number of surviving nymphs divided by the total number of nymphs released at the start of the experiment. Ten cups were analysed for each cultivar or line.

Plasmid construction

Routine molecular cloning techniques were used to prepare the constructs. The plasmids and primers used in this work are listed in Supplementary Table 3. All of the resulting recombinant vectors were sequenced.

To construct the plasmid for Y2H screening, the full-length Bph14 sequence was amplified from Bph14 leaf sheath cDNA and cloned into the Xma I site of the pGBKT7 vector, which produced the BPH14–BD vector.

To construct the plasmids for the Y2H assays, the coding sequences of Bisp (ORF sequence without its 1–25 amino acid residue signal peptide; amino acids 26–241), Bisp^26–124(amino acids 26–124) and Bisp^125–241 (amino acids 125–241) were amplified and cloned into the Xma I site of the pGADT7 vector to produce BISP–AD, BISP(26–124)–AD and BISP(125–241)–AD, respectively. The coding sequences of the CC, NB and LRR domains and full-length Bph14 were amplified from the BPH14-BD vector and cloned into the Xma I sites of the pGADT7 and pGBKT7 vector to produce CC–AD, NB–AD, LRR–AD, BPH14–AD, CC–BD, NB–BD, and LRR–BD, respectively. The coding sequence of N14 (Bph14 allele in the rice variety Nipponbare) was amplified from N14 leaf sheath cDNA and cloned into the Xma I site of the pGBKT7 vector to produce N14–BD.

Eight kinases that were differentially phosphorylated during BPH-infested and non-infested plants were chosen for this study. The coding sequences for the OsRLCK185, OsRLCK185^1–85 (amino acids 1–85), OsRLCK185^KD (kinase domain, amino acids 86–353), OsRLCK185^354–491 (amino acids 354–491), OsMAPKKKε, OsMKK6, OsMPK20, OsMPK16, OsCDPK14, OsCDPK20, and OsRLCK259 fragments were amplified from N14 leaf sheath cDNA and cloned into the pGBKT7 vector to produce OsRLCK185–BD, OsRLCK185(1–85)–BD, OsRLCK185^KD–BD, OsRLCK185(354–491)–BD, OsMAPKKKε–BD, OsMKK6–BD, OsMPK20–BD, OsMPK16–BD, OsCDPK14–BD, OsCDPK20–BD, and OsRLCK259–BD, respectively. The coding sequences of OsNBR1, OsATG8a, OsATG8b, OsATG8c, and OsATG8e were amplified from Bph14 leaf sheath cDNA and cloned into pGBKT7 and pGADT7 to produce OsNBR1–BD, OsATG8A–BD, OsATG8B–BD, OsATG8C–BD, and OsATG8E–BD, respectively.

Constructs used for rice protoplast transfection were generated using pCXUN-4×HA and pCXUN-4×Myc¹⁹. The coding sequences of the CC, NB and LRR domains, Bph14, N14, the N14 LRR domain (N14–LRR), Wrky72, Bisp (ORF sequence without its signal peptide), Bisp^26–124, Bisp^125–241, OsRLCK185, OsRLCK259, OsNBR1, OsNBR1 mutant N1, OsATG8a, OsATG8b, OsATG8c, and OsATG8e were amplified and cloned into the pCXUN-4×Myc vector to produce CC–MYC, NB–MYC, LRR–MYC, BPH14–MYC, N14–MYC, N14–LRR–MYC, WRKY72–MYC, BISP–MYC, BISP(26–124)–MYC, BISP(125–241)–MYC, OsRLCK185–MYC, OsRLCK259–MYC, N1–MYC, OsNBR1–MYC, MYC–OsATG8A, MYC–OsATG8B, MYC–OsATG8C, and MYC–OsATG8E, respectively. Meanwhile, the fragments for the GFP, CC, NB, LRR domains, Bph14, N14, the N14 LRR domain (N14–LRR), Bisp, Bisp^26–124, Bisp^125–241, OsRLCK185, and OsNBR1 were amplified and cloned into the pCXUN-4×HA vector to produce GFP–HA, CC–HA, NB–HA, LRR–HA, BPH14–HA, N14–HA, N14–LRR–HA, BISP–HA, BISP(26–124)–HA, BISP(125–241)–HA, OsRLCK185–HA, and OsNBR1–HA, respectively. Additionally, we generated a series of Osnbr1 fragments, including the isolated N1 (amino acids 1–755), N2 (amino acids 756–832), N3 (amino acids 1–782), N4 (amino acids 1–755 and 798–832), and mutant N5 (K13A), N6 (WL788/791AA) and N7 (K13A and WL788/791AA) fragments and cloned them into the pCXUN-4×HA vector to produce N1–HA, N2–HA, N3–HA, N4–HA, N5–HA, N6–HA, and N7–HA, respectively.

To analyse the subcellular localization of BISP, the coding sequence of Bisp (ORF sequence without its signal peptide) was cloned downstream of the maize (Zea mays) Ubiquitin 1 promoter, in-frame with GFP in the binary vector pCAMBIA1300, which produced the BISP–GFP construct.

For BiFC analysis, the coding sequences of Bph14, N14, Bisp (ORF sequence without its signal peptide), and Bisp^125–241 were cloned in-frame into the Xma I sites of the pUSYNE and pUSYCE vectors¹⁹, which produced the constructs BISP–YN, BISP(125–241)–YN, BPH14–YC, and N14–YC, respectively.

The Escherichia coli recombinant protein vectors used for the expression and purification of proteins in the phosphorylation activity assays, designated pET-MBP-His and pET-GST-His, were created by adding a C-terminal MBP or GST tag, respectively, to the pET-28a expression vector (EMD Biosciences, Novagen). The coding sequences of Bisp (ORF sequence without its signal peptide), Bisp^26–124 and OsRLCK185 were amplified and cloned into the pET-MBP-His and pET-GST-His vector, which produced OsRLCK185–MBP–His, BISP(26–124)–GST–His and BISP–GST–His constructs, respectively.

For the N. benthamiana leaf agroinfiltration experiments, the coding sequences of Bph14, N14, OsNBR1, and N1 (an OsNBR1 mutant) were cloned into the pEarleyGate 203 vector⁴⁸ to produce BPH14-203, N14-203, OsNBR1-203, and N1-203, respectively. The coding sequence of Bisp (ORF sequence without its signal peptide) was cloned into the pEarleyGate 201 and pEarleyGate 101 vector⁴⁸ to produce BISP-201 and BISP-101 (BISP–YFP), respectively.

To generate the CRISPR–Cas9 construct, the target sites for CRISPR–Cas9 were designed using the web-based tool CRISPR-P (v.2.0; http://crispr.hzau.edu.cn/CRISPR2/). The CRISPR–Cas9 binary construct was prepared as previously described⁴⁹, which produced the construct OsNBR1–Cas9.

Y2H screening

The Matchmaker Gold Yeast Two-Hybrid System (Clontech, 630489) was used to screen for BPH14-interacting proteins. Y2H Gold cells carrying BPH14–BD were mixed with 1 ml of a BPH salivary gland cDNA library and incubated overnight before plating on TDO (SD/-Leu-Trp-His)-selective medium. Candidate clones growing on TDO medium were confirmed following the manufacturer’s protocol (Clontech, 630489). The BPH cDNA library was constructed using the Make Your Own “Mate & Plate” Library System (Clontech, 630490) with a simple and highly efficient protocol. Total RNA was isolated from BPH salivary glands using TRIzol reagent (Takara, 9109), and cDNA was synthesized from the isolated mRNA using SMART cDNA synthesis technology (Clontech, 634926).

For the Y2H assay of protein interactions, the yeast strain AH109 was transformed with the indicated bait and prey plasmids according to the manufacturer’s instructions. Co-transformants were simultaneously plated on selection medium containing DDO (SD/-Leu-Trp), TDO (SD/-Leu-Trp-His) with the appropriate concentration of 3-amino-1,2,4-triazole (3-AT) or QDO (SD/-Leu-Trp-His-Ade) and incubated at 30 °C.

Cloning and sequencing of Bisp cDNA

The cDNA sequences of Bisp were obtained from BPH salivary gland transcriptomes⁵. To obtain the full-length counterparts of the truncated sequences in the transcriptomes, 5′- and 3′-RACE amplification was performed using a SMARTer RACE cDNA Amplification kit (Clontech, 634923) following the manufacturer’s instructions. Bisp-specific primers were obtained based on sequencing data. The RACE products were amplified, and the purified products were ligated into the pMD18-T vector (Takara, 6011) and sequenced.

Protein extraction, immunoblot analysis and in vivo co-IP assays

Proteins were extracted from the leaf sheaths of rice seedlings at the four-leaf stage. The leaf sheaths were wiped with a cotton ball soaked in alcohol to remove honeydew and ground in liquid nitrogen. Total protein was extracted from the leaf tissue in rice protein extraction buffer (50 mM Tris-HCl pH 7.5, 150 mM NaCl, 10% glycerol, 0.1% NP-40, plant protease inhibitor cocktail (Roche)). Equal amounts of total protein were analysed by SDS–PAGE and detected by immunoblotting using anti-actin (Abbkine, ABL1050; 1:3,000), anti-BISP (1:500), anti-OsWRKY72 (Beijing Protein Innovation, AbP80456-A-SE; 1:500) and anti-AtNBR1 (Agrisera, AS194281; 1:1,000) antibodies. For OsATG8 analysis, equal amounts of proteins were subjected to SDS–PAGE with 6 M urea⁵⁰ and detected by immunoblotting using anti-AtATG8A antibodies (Abcam, ab77003; 1:1,000). The anti-BISP and anti-NISP1 antibodies were prepared by expressing BISP and NISP1 (cloned into pET28a) in E. coli strain BL21 (DE3), respectively. The expressed recombinant proteins were collected and injected into two rabbits (DIA·AA Biotech). Ponceau S solution (Sigma-Aldrich, P7170) was used to stain the PVDF membrane and was used as a loading control. The intensities of protein signals were quantified using ImageJ.

Protein samples from rice protoplasts were prepared in rice protein extraction buffer (100 mM Tris-HCl pH 7.5, 1 mM EDTA, 5 mM MgCl₂, 0.5% (w/v) Triton X-100, with a plant protease inhibitor cocktail (Roche)). Total soluble proteins were extracted from rice protoplast samples, each in 100 µl of rice protoplast protein extraction buffer. Next, 10 µl of the extract was separated by SDS–PAGE and subjected to immunoblotting using anti-HA (MBL, M180-3; 1:1,000), anti-MYC (MBL, M192-3; 1:1,000) or anti-actin (Abbkine, ABL1050; 1:3,000) antibodies.

For the co-IP assays, rice protoplasts were incubated at 28 °C for 14–20 h after transfection. Total protein extracts were collected from the protoplasts in the rice protein extraction buffer as described above. The supernatants were incubated with 10 μl Protein G Agarose beads (Millipore, 16–266) and 1 μl anti-HA (MBL, M180-3) or 1 μl anti-MYC (MBL, M192-3) antibodies at 4 °C for 5 h, and detected using anti-HA or anti-HA mAb-HRP-DirecT (MBL, M180-7, clone: TANA2; 1:1,000), anti-actin and anti-MYC or anti-MYC mAb-HRP-DirecT (MBL, M192-7, clone: My3, 1:1,000) antibodies, respectively. The intensities of protein signals were quantified using ImageJ.

For the yeast protein immunoblots, total yeast protein was extracted as previously described⁵¹. In brief, single yeast colonies were resuspended in 4 ml YPDA liquid medium and grown to OD₆₀₀ = 0.6 at 30 °C. The pelleted yeast cells were resuspended in 1 ml distilled water, combined with 100 μl 0.2 M NaOH, incubated for 5 min at room temperature and pelleted and resuspended in 50 μl loading buffer. The proteins were boiled for 10 min at 95 °C and detected by immunoblotting with anti-HA or anti-MYC antibodies. Ponceau S solution was used to stain the PVDF membrane for loading control.

Rice protoplast isolation and transfection

Rice protoplasts were prepared from etiolated seedlings and transfected as previously described^11,19. About 3–10 μg of plasmid DNA was used to transfect around 2 × 10⁵ protoplasts using the PEG-mediated method. The fluorescent or epitope-tagged proteins were detected at 14–20 h after transfection.

Subcellular localization and BiFC analysis

For subcellular localization, the BISP–GFP plasmid and the nuclear marker bZIP63–RFP^11,19 were transformed into rice protoplasts. Following incubation at 28 °C for 14–20 h, the protoplasts were imaged using a confocal microscope (Leica, DMi8).

For BiFC analysis, protoplasts were transfected with the indicated expression vectors. Following incubation at 28 °C for 14–20 h, the protoplasts were imaged by confocal microscopy. BiFC fluorescent signals were quantified using previously described methods⁵². The intensity of YFP and CFP signals in individual protoplasts was determined using ImageJ software. The average YFP/CFP intensity ratio (per cent) of the whole cell was measured, and the means were calculated from five to ten independent cells. The fluorescent or epitope-tagged proteins were detected with anti-GFP (Roche, 11814460001; 1:1,000), anti-HA or anti-MYC antibodies.

RNA isolation, real-time and semi-quantitative RT–PCR analysis

Total RNA was isolated from rice leaf sheaths, BPHs or N. benthamiana leaves using TRIzol reagent (Takara, 9109). First-strand cDNA was synthesized using a PrimeScript RT Reagent kit with gDNA Eraser (Takara, RR047A) following the manufacturer’s instructions.

Real-time RT–qPCR was carried out using a CFX96 Real-Time system (Bio-Rad) with SYBR Green Supermix (Bio-Rad, 172–5274) following the manufacturer’s instructions. For RT–qPCR in rice, three housekeeping genes, Osubiquitin, Ostbp and Oshsp⁵³, were used as reference genes for calibration of real-time RT–PCR data. For RT–qPCR in BPH, NlRPS18, Nlubiquitin and NlACTB (which encodes β-actin) were used as reference genes. The sequences of the primers used are listed in Supplementary Table 3.

For the semi-quantitative RT–PCR assays, first-strand cDNA (5 μl) was diluted to 50 μl final volume with TE buffer, and 1 μl of the dilution was used as template for PCR amplification using gene-specific primers. The internal standard NbACTB was used. The number of cycles used for amplification with each primer pair was adjusted to ensure that the amplification was in the linear range. The PCR products were sequenced to ensure that they were derived from the targeted genes. The sequences of the primers used are listed in Supplementary Table 3.

Immunohistochemistry

For immunolocalization of BISP in rice leaf-sheath sections, N14 plants at the two-leaf stage were infested with ten fourth- or fifth-instar nymphs for 2 days or served as non-infested controls. The outer leaf sheaths were collected, fixed, dehydrated, and embedded in paraffin (Paraplast Plus, Sigma-Aldrich, P3683) as previously described^11,17. The sections were incubated in dimethylbenzene and an alcohol gradient to remove the paraffin and to gradually rehydrate. The sections were incubated with anti-BISP antibodies or pre-immune rabbit serum (prepared as described above, 1:100) at 4 °C in the dark overnight. The next day, the sections were washed 3 times (15 min each time) with PBST (PBS with 0.1% (v/v) Triton X-100), blocked with 5% bovine serum albumin in PBST at room temperature for 1 h, washed 3 times at 15-min intervals with PBST, and incubated with the secondary Cy3-conjugated goat anti-rabbit antibodies (Jackson ImmunoResearch, 111-165-003, 1:500) at 4 °C overnight. The sections were washed extensively with PBST at 15-min intervals and imaged under a confocal microscope (Leica, DMi8).

For immunolocalization in BPH salivary glands and guts, female or male insects were anaesthetized with carbon dioxide (CO₂). The salivary glands and guts were dissected in 0.65% NaCl, washed 3 times in PBST and fixed in 4% (v/v) formaldehyde in PBST at 4 °C overnight. The next day, the organs were extensively washed with PBST, incubated with anti-BISP antibodies, pre-immune serum or anti-NlSP1 antibodies⁵⁴ (1:100) at 4 °C in the dark overnight, and treated as described above. Nuclei were stained with 100 nM 4,6-diamidino-2-phenylindole (DAPI) at room temperature in the dark for 5 min. The salivary glands and guts were extensively washed with PBST and placed on a glass slide. Photographs were taken under a confocal microscope (Leica, DMi8). For the quantification of Cy3, the Cy3 and DAPI fluorescence signals in individual cells were quantified using ImageJ. The average Cy3/DAPI intensity ratio of each cell was measured, and mean ratios were calculated from five to ten independent cells.

RNAi and bioassay of BPHs after dsRNA injection

A 506-bp fragment of Bisp was amplified with the primers listed in Supplementary Table 3, including the T7 promoter sequence in the forward primer. After cloning the Bisp fragment into the pMD18-T vector (Takara, 6011), dsRNA was synthesized from PCR-generated DNA templates using a MEGAscript T7 Transcription kit (Ambion, AM1354). BPH nymphs in their third, fourth or fifth stage were injected with Bisp-RNAi or GFP-RNAi using a Nanoliter 2010 injector (World Precision Instruments) as previously described⁵. The efficiency of gene silencing was determined by RT–qPCR at 1–10 days after dsRNA treatment.

After the injected BPHs were placed and reared on rice plants for 12 h, the survival rate, weight gain and honeydew excretion of GFP-RNAi BPHs, Bisp-RNAi BPHs and non-injected control insects were measured as described above. Five replicates were set up for each treatment to measure BPH survival rates, and 20 replicates were performed for each treatment group to measure weight gain and honeydew excretion.

Plant transformation

To constitutively express Bisp in rice plants, the BISP–MYC plasmids were transformed into N14 or Bph14 plants through Agrobacterium-mediated transformation⁸. To knockout OsNBR1 by genome editing, the CRISPR–Cas9 plasmids OsNBR1–Cas9 were introduced into Agrobacterium tumefaciens strain EHA105 and transformed into Bph14 as described elsewhere⁸. Genomic DNA was extracted from these transformants, and PCR amplification was performed using primer pairs flanking the designed target site. The PCR products were sequenced to determine whether the gene editing was successful.

Quantification of free SA

Endogenous free SA levels were determined in rice leaf sheaths. Fifteen leaf sheaths from each plant or line were ground to a fine powder in liquid nitrogen as a replicate. Three biological replicates of each frozen sample were prepared. Each sample (100 mg) was added to 750 μl precooled extraction buffer (methanol:water:acetic acid, 80:19:1, v/v/v) supplemented with 3 µg naphthaleneacetic acid (Sigma-Aldrich, 35745) as an internal standard and shaken at 4 °C overnight. After centrifugation, the supernatant was filtered through a 0.22-μm nylon membrane and dried under a nitrogen flow at room temperature for at least 4 h. Finally, 200 μl methanol was added to dissolve the extracts. Endogenous free SA measurements were performed as described elsewhere⁵⁵.

Recombinant protein expression and purification in E.
coli and BISP antibody specificity analysis

The OsRLCK185–MBP–His, BISP–GST–His and BISP(26–124)–GST–His constructs were expressed in E. coli BL21 (DE3) cells. GST-tagged or MBP-tagged recombinant proteins were purified using Glutathione Sepharose beads (GE Healthcare, 17075601) and Amylose Resin (New England Biolabs, E8022S), respectively, according to the manufacturer’s instructions. Protein concentrations were determined by BCA protein assay (Beyotime, P0010S).

For the BISP antibody specificity analysis, the purified BISP–GST–His protein and total protein from whole insect and a twofold dilution series were separated on SDS–PAGE gels. BISP–GST–His was detected in immunoblots using anti-BISP and anti-His (GenScript, A00186; 1:2,000) antibodies, respectively. Ponceau S solution was used to stain the PVDF membrane and used as the total BPH protein loading control.

In vitro phosphorylation assays

To test the autophosphorylation of OsRLCK185, 1 µg purified recombinant OsRLCK185–MBP–His, BISP(26–124)–GST–His or BISP–GST–His was incubated in 25 µl kinase reaction buffer containing 50 mM HEPES (pH 7.5), 10 mM MgCl₂, 1 mM DTT, 1 mM ATP, and 2.5 µCi [γ-³²P]ATP at 30 °C for 10–30 min. The kinase reaction was subsequently analysed using a 10% SDS–PAGE gel for autoradiography. A similar gel was subjected to Coomassie Brilliant Blue staining as a control.

To test the phosphorylation of OsRLCK185 at the serine and threonine residues, phosphorylated OsRLCK185 was detected by immunoblotting with anti-phosphoserine/phosphothreonine antibodies (Millipore, 05-368, 1:1000) and anti-His antibody (GenScript, A00186).

Recombinant protein expression and purification in insect cells and BLI analyses

LRR, N14–LRR, BISP, BISP(26–124), and BISP(125–241) were cloned into pFastBac 1 with an N-terminal 6×His-SUMO tag and expressed in SF9 insect cells (Thermo Fisher Scientific, 11496015) at 27 °C. SF9 cells (500 ml) cultured in Sf-900 II SFM medium (Invitrogen, 10902088) were infected with 25 ml recombinant baculovirus and were cultured for 4 days at 27 °C. Cells were collected and re-suspended in PBS solution (2 mM KH₂PO₄, 8 mM Na₂HPO₄, 136 mM NaCl, and 2.6 mM KCl, pH 7.4). After sonication and centrifugation, recombinant LRR, N14–LRR, BISP, BISP(26–124), and BISP(125–241) proteins in the supernatant were purified with Ni-NTA beads (Novagen,70666-4) and then were dialysed in PBS solution.

The binding kinetics and affinities of LRR and N14–LRR with BISP, BISP(26–124) or BISP(125–241) was determined using an Octet RED96 system (FortéBio) as previously described^56,57. Purified LRR and N14–LRR were biotinylated with NHS-biotin (Sangon Biotech, C100212) according to the manufacturer’s instructions. All streptavidin-coated biosensors (FortéBio) were hydrated in PBS buffer (2 mM KH₂PO₄, 8 mM Na₂HPO₄, 136 mM NaCl, 2.6 mM KCl, 0.05% Tween 20, and 0.5% bovine serum albumin, pH 7.4) for 10 min. Biotinylated LRR and N14–LRR were diluted in PBS buffer to a final concentration of 20 μg ml⁻¹ and immobilized onto a streptavidin-coated biosensor. Biosensors with immobilized LRR and N14–LRR were diluted in binding buffer containing different concentrations of BISP, BISP(26–124) or BISP(125–241) for the binding kinetics analysis (association step). Subsequently, the streptavidin-coated biosensor was returned to PBS buffer for 180 s (dissociate step). PBS buffer was used in all BLI experiments to reduce nonspecific binding to the biosensors except for experiments in the presence of BISP, BISP(26–124) or BISP(125–241). The experiments included the following steps at 25 °C: (1) loading (60 s); (2) baseline (120 s); (3) immobilization of proteins onto sensors (60 s); (4) association (180 s); and (5) disassociation (180 s). The data were analysed, and sensor-grams were step-corrected and reference-corrected. Global fitting of the kinetic rates was performed, then the equilibrium dissociation constant (K_d), association (K_on) and dissociation (K_off) rate constants and their standard errors were analysed using FortéBio data analysis software (v.1.1.0.16, FortéBio). The assays were repeated two times for each affinity measurement.

For the competition assay, an Octet RED96 system (FortéBio) was used as previously described⁵⁸. Biotinylated LRR was diluted in PBS buffer to a final concentration of 20 μg ml⁻¹ and immobilized onto a streptavidin-coated biosensor. The association of BISP(26–124) or PBS was measured for 360 s at 25 °C. Next, the competitors (BISP, PBS or BISP(26–124)) were added to the BISP(26–124)-associated or PBS-associated samples. The experiments included the following steps at 25 °C: (1) loading (60 s); (2) baseline (120 s); (3) immobilization of protein onto sensors (60 s); (4) baseline (240 s); (5) association with BISP(26–124) or PBS (360 s); (6) disassociation (240 s); (7) association with the competitors BISP, PBS or BISP(26–124) (120 s); and (8) dissociation (180 s). The data were analysed using FortéBio data analysis software (v.1.1.0.16, FortéBio). The assays were repeated two times for each measurement.

MST assay

The MST assay was performed as previously described^59,60. The affinity of purified LRR or N14–LRR with BISP, BISP(26–124) or BISP(125–241) was measured using a Monolith NT.115 (Nanotemper Technologies). Proteins were fluorescently labelled using a Monolith Protein Labelling kit (RED-NHS 2nd Generation; Nanotemper Technologies, MO-L011) according to the manufacturer’s protocol, and the labelled protein used for each assay was about 20 nM. A solution of unlabelled BISP, BISP(26–124) or BISP(125–241) protein was diluted for an appropriate serial concentration gradient. The samples were loaded into capillaries (Nanotemper Technologies, MO-K022) after incubation at room temperature for 5 min. Measurements were performed in 20 μl of PBS buffer containing 0.05% Tween 20. The assays were repeated three times for each affinity measurement. Data analyses were performed using Nanotemper Analysis software provided by the manufacturer.

For the competition assays between LRR with BISP and BISP(26–124) by MST, 50 nM LRR proteins and 200 nM BISP(26–124) were combined in PBS buffer with 0.05% Tween 20 and incubated for 10 min to enable protein interactions. A series of different concentrations of BISP was then added to the LRR–BISP(26–124) mixture before measurement. The assays were repeated three times for each measurement. The dissociation constant (K_i) from the MST assays was calculated using the K_i Finder software from the Nanotemper website (https://nanotemper.my.site.com/explore/s/article/Can-competition-assays-be-performed-with-MST).

Inhibition of protein degradation in rice protoplasts

After PEG-mediated transformation, the rice protoplasts were placed in 1 ml W5 buffer containing 10 μM 3-MA (3-methyladenine, Selleck, S2767), 100 μM chloroquine (MCE, HY-17589A), 100 μM leupeptin (MCE, HY-18234A), or 20 μM E-64d (Aloxistatin, Sigma-Aldrich, E8640) to inhibit autophagy or 50 μM MG132 (Selleck, S2619) to inhibit the 26S proteasome. The treated protoplasts were incubated at 28 °C for 14–20 h before protein extraction. Anti-actin (Abbkine, ABL1050) was used as a reference to quantify total protein levels. Rice WRKY72 is degraded by the 26S proteasome pathway¹⁹ and was used as a positive control to ensure that the proteasome inhibitor MG132 was active. The rice HD1 protein is degraded in the dark by the autophagy pathway⁶¹, and OsNBR1 was used as a positive control to assure autophagy inhibitors were active. The intensity of protein signals in immunoblots was quantified by ImageJ.

Live-cell imaging of autophagic structures and virus-induced gene silencing in N.
benthamiana leaves

To observe autophagic structures in N. benthamiana leaves by confocal microscopy, A. tumefaciens strain GV3101 cells harbouring CFP–NbATG8F (designed as CFP–ATG8F)²⁷ and the constructed expression vectors, BISP-201, BPH14-203, N14-203, OsNBR1-203, and N1-203 were infiltrated into N. benthamiana leaves for 60 h. The samples were then infiltrated with or without the autophagy inhibitor ConA (first made in DMSO at 0.5 M, then 1:1,000 added) for 10 h before being examined under a confocal microscope (Leica, DMi8).

For virus-induced gene silencing (VIGS) assays, pTRV1 and pTRV2 or its derivatives were introduced into Agrobacterium strain GV3101. VIGS assays were performed as previously described⁶². In brief, the Agrobacterium cultures with pTRV1 and pTRV2 or its derivatives were mixed at a 1:1 ratio and were infiltrated into the leaves of six-leaf stage N. benthamiana plants. Silenced phenotypes appeared in the upper leaves about 10 days after infiltration. A. tumefaciens strain GV3101 cells harbouring CFP–NbATG8F and the constructed expression vectors BISP-101 (BISP–YFP) and BPH14-203 were infiltrated into the N. benthamiana upper leaves.

Transmission electron microscopy

For immunogold electron microscopy, N14 leaf sheaths infested with BPH for 24 h were cut into 0.1-mm pieces, fixed in precooled 3% paraformaldehyde with 0.1% glutaraldehyde (in 100 mM PBS, pH 7.4) and immediately placed under a vacuum until the samples were completely submerged in the liquid. The samples were transferred to fresh 0.1% glutaraldehyde and incubated at 4 °C overnight. The next day, the samples were washed 3 times with PBS (pH 7.4) for 30 min. The samples were cleared by successive washes with 30% and 50% ethanol for 60 min each at 4 °C, then 70%, 80%, 85%, 90%, and 95% ethanol for 30 min each at −20 °C, and finally 100% ethanol 3 times for 60 min each at −20 °C. After ethanol washes, the samples were embedded in Lowicry K4M resin stepwise, according to the manufacturer’s instructions. Ultrathin sections (70 nm) of the embedded tissues were cut with a diamond knife using an Ultracut E Ultramicrotome (Reichart-Jung) and collected on 3-mm copper (mesh) grids. Immunolabelling of leaf sections was done using standard procedures as previously described⁵⁹ with anti-BISP antibodies or pre-immune serum at 100 µg ml^–1 and anti-rabbit IgG gold-coupled secondary antibodies (Sigma-Aldrich, G7277- 4ML) at 1:50 dilution. The sections were then stained with 2% uranyl acetate (in 70% ethanol) for 15 min, washed in distilled water 3 times, stained with lead citrate for 12 min, washed in NaOH buffer, washed in distilled water 3 times, and examined under a JEM-1230 electron microscope.

For transmission electron microscopy observations, 2-week-old Bph14–Bisp transgenic lines, Bph14 and MH63 leaf sheaths were cut into 0.1-mm pieces, fixed in precooled 4% paraformaldehyde with 2% glutaraldehyde (in 100 mM PBS, pH 7.4) and immediately placed under a vacuum until the samples were completely submerged in the liquid. The samples were transferred to fresh 2% glutaraldehyde and incubated at 4 °C overnight. The next day, the samples were washed 5 times with PBS (pH 7.4) for 20 min and submerged in a 1% solution of osmium tetroxide at room temperature for 1–2 h until they turned completely black because of oxidation. The samples were washed 5 times with PBS at 20-min intervals. The samples were cleared by successive washes with 15%, 30%, 50%, and 70% ethanol for 30 min each, then 80%, 85%, 90%, and 95% ethanol for 20 min each, and finally 100% ethanol twice for 45 min each. After ethanol washing, the samples were embedded in Spurr resin stepwise as instructed by the manufacturer. Ultrathin sections (70 nm) of the embedded tissues were cut with a diamond knife using an Ultracut E Ultramicrotome (Reichart-Jung), collected on 3-mm copper (mesh) grids, stained with 2% uranyl acetate (in 70% ethanol) for 10 min, washed in distilled water 3 times, stained with lead citrate for 15 min, washed in NaOH buffer, and examined under a JEM-1230 electron microscope. To quantify autophagosomes, the double-membrane autophagosomes in every eight transmission electron microscope images were counted and served as one biological replicate; each genotype had five replicates.

BPH infestation

Seeds were sown in 10-cm diameter plastic cups and grown in a plant growth chamber (CONVIRON PGC2000) as described elsewhere¹¹. To analyse the relationship between BISP levels and autophagy, 4-leaf rice plants were infested with second- or third-instar BPH nymphs (10 per plant) at the selected time points (6, 12, 24, 48, and 72 h before the end of the experiment). Control groups were maintained in parallel but without BPH infestation. At the end of the treatments, leaf sheaths were wiped with cotton wool balls soaked in alcohol to remove honeydew, leaf sheaths were excised, frozen, and ground to a fine powder in liquid nitrogen, and proteins isolated. Immunoblotting was performed following the procedures described above in the section ‘Protein extraction, immunoblotting and in vivo co-IP assays’. All treatments were repeated three times.

To measure autophagy and BISP levels following BPH infestation, 4-leaf rice plants were infested with second- or third-instar BPH nymphs (10 per plant) for 24 h. The insects were then removed, and the plants were allowed to grow under standard conditions (described above). At the selected time points after BPH removal, honeydew was removed from leaf sheaths, leaves were excised, frozen, and ground to a fine powder in liquid nitrogen, and proteins isolated. Immunoblotting was performed following the procedures described above. All treatments were repeated three times.

For the autophagy inhibitor ConA treatment, the BPH-infested and non-infested rice plants were treated with or without ConA (10 μM) for 12 h before samples were collected. Leaf sheath total protein extraction and immunoblotting were performed according to the procedures described above. All treatments were repeated three times.

Agronomic performance test

To investigate the agronomic performance of Bph14 and Bph14–Bisp plants, plants were grown in a field in Wuhan, China, under routine management. At harvest, the middle plants in the central row of each plot were assessed for the following nine agronomic traits: plant height, sword leaf length, sword leaf width, number of panicles per plant, number of filled grains per panicle, number of filled grains per plant, percentage of filled grain, 1,000-grain weight, and grain yield per plant⁶³.

Statistical analysis

No statistical methods were used to predetermine the sample sizes. The experiments were not randomized, and investigators were not blinded to allocation during the experiments and outcome assessment. Quantification analyses of protein abundance were conducted using ImageJ software. All values are presented as the mean ± s.d. or mean ± s.e.m. as indicated. Data points are plotted on the graphs, and the number of samples is indicated in the corresponding figure legends. In all cases, sampling for replicates and between time points was performed by collecting material from new plants. All statistical analyses were performed by one-way analysis of variance with MS Excel and GraphPad Prism (v.8.0.1, GraphPad Software). Detailed information about statistical analysis values for all experiments is provided in Supplementary Table 4.

Reporting summary

Further information on research design is available in the Nature Portfolio Reporting Summary linked to this article.