Pre-germinated Xiushui 11 rice seeds were germinated in plastic bottles in a greenhouse at 28 ± 2°C under a 14 h light/10 h dark photoperiod. Each 7-day-old seedling was then transferred to a 750 mL plastic pot containing potting medium. After 25 days, the rice plants were ready for use in the subsequent experiments. Nicotiana benthamiana was kept in a climate chamber at 23 ± 1°C under a 16 h light/8 h dark photoperiod. After 4–5 weeks (five-leaf stage), the plants were used in the subsequent Agrobacterium tumefaciens-mediated transient transformation experiments. The original SBPH colonies were obtained from rice fields in Nanjing, China and maintained on rice seedlings in a climate chamber at 25 ± 1°C under a 14 h light/10 h dark photoperiod. To minimize interference from egg deposition, fourth-instar nymphs were injected and metamorphosed into fifth-instar nymphs 2 days later, and the latter were used for rice defense response analyses. Injected third-instar nymphs were used for the survival analysis. Newly emerged brachypterous female adults within 24 h were used for honeydew excretion measurement and electrical penetration graph (EPG) analysis, owing to the highest feeding capacity at this stage.

Total RNA was extracted from whole bodies of the first- to fifth-instar nymphs and newly emerged brachypterous male and female adults. It was also extracted from salivary glands, guts, ovaries, fat bodies, and integuments dissected from brachypterous female adults. For each biological replicate, 50 nymphs, 20 adults, and various tissues from 200 brachypterous female adults were collected. Each experiment was repeated in triplicate. Total RNA was isolated with the SV Total RNA Isolation System (Promega, Madison, WI, United States) according to the manufacturer’s instructions. After DNase treatment and quantification, 1 μg total RNA sample was reverse-transcribed with a PrimeScript RT-PCR Kit (TaKaRa Bio Inc., Dalian, China). The quantitative polymerase chain reaction (qPCR) reactions were performed with a TB Green™ Premix Ex Taq™ Kit (TaKaRa Bio Inc.) according to the manufacturer’s protocol and run in a LightCycler^® 480 II Real Time System (Roche Diagnostics, Basel, Switzerland). Elongation factor 2α (ef2) served as the internal normalization controls. The qPCR primers are listed in Supplementary Table 1. The gene expression levels were calculated by normalizing the target gene mRNA level to ef2 mRNA abundance via the 2^–ΔCt method (Chyla et al., 2019; Tian et al., 2021).

LsCaM and NlCaM (GenBank no. RZF46934.1 and AXY40108.1) were obtained by reverse transcription polymerase chain reaction (RT-PCR) using total RNA from SBPH and BPH. They were cloned into the pMD19-T vector (TaKaRa Bio Inc., Kusatsu, Japan), and sequenced. An In-Fusion HD Cloning Kit (TaKaRa Bio Inc.) was used to insert LsCaM into the pBINPLUS-green fluorescent protein (GFP) vector (subcellular localization and expression) via a single BamHI digestion site. LsPDI1 was cloned via homologous recombination into the pBINPLUS-mCherry vector (expression) via a single BamHI digestion site using the aforementioned kit. Primers are listed in Supplementary Table 1.

Primers containing the T7 promoter (Supplementary Table 1) were designed to clone a specific fragment (∼300–500 bp) in each target gene. The double stranded RNA (dsRNA) was synthesized and microinjected as previously described (Liu et al., 2010; Ji et al., 2017). To determine RNAi efficiency, target gene expression was measured at 2, 4, and 6 days after dsRNA injection. qPCR was repeated three times with six replicates per time, there were 20 insects per replicate.

Planthopper feeding behavior was recorded on a direct-current EPG system (Wageningen Agricultural University). The method used was previously described (Cao et al., 2013; Ji et al., 2017). The feeding behavior of newly emerged adult females 2 days after dsRNA injection was monitored for 6 h. There were 15 replicates per treatment. The recorded signals were analyzed with PROBE software (Wageningen Agricultural University). The durations of each sequential waveform event were measured for each insect. The average waveform duration per insect per waveform was calculated for each treatment.

Honeydew excretion reflects feeding activity in sap-sucking insects. To assess the effect of CaM knockdown on planthopper honeydew excretion, one rice stem was covered with an inverted transparent plastic cup placed on a filter paper. Three days after injection, a single newly emerged brachypterous female adult was transferred to the cup and left to feed for 2 days. The filter papers were then soaked in 0.1% (w/v) ninhydrin in acetone and oven-dried at 65°C for 30 min. The honeydew stains appeared as violet or purple spots because of their amino acid content (Lei et al., 2014; Ji et al., 2021). The areas of the ninhydrin-positive deposits were measured with ImageJ v. 1.8.0 (National Institutes of Health, Bethesda, MD, United States). There were 15 replicates per treatment.

The injected third-instar nymphs were left to recover on rice seedlings for 1 day. Healthy nymphs were randomly selected for the subsequent survival bioassay on rice plants or artificial diet. One rice stem per pot was confined within a glass cylinder (diameter, 2 cm; height, 8 cm) containing 20 third-instar nymphs. In the artificial diet experiment, 20 third-instar nymphs were introduced into individual feeding chambers (diameter, 2 cm; long, 9 cm) as described previously (Ji et al., 2017). The surviving nymphs were counted daily. The survival rate of each treatment and the corrected survival rates of nymphs with injected dsLsCaM, using nymphs with injected dsGFP as controls, on rice or artificial diet were calculated. The experiment was repeated five times.

The number of eggs laid per female adult was determined using a previously reported method (Ji et al., 2017). A single newly emerged dsRNA-treated female adult and two newly emerged untreated male adults were released into the same glass cylinder confining a single rice stem. The insects were allowed to feed for 8 days. The number of eggs laid by a female adult on each dissecting rice stem was counted under a microscope (Olympus SZ51, Olympus Corp., Tokyo, Japan). There were 15 replicates per treatment.

Signal peptides and transmembrane and EF-hand domains were predicted with SignalP v. 5.0,¹ TMHMM v. 2.0,² and PROSITE,³ respectively.

The complete open reading frame (CORF) of LsCaM was cloned into a pET-28a vector (Novagen, Madison, WI, United States). The LsCaM:pET-28a construct was transformed into Escherichia coli BL21 (DE3). The recombinant protein products were purified with Ni-NTA columns (Qiagen, Venlo, Netherlands) according to the manufacturer’s instructions. They were then concentrated with a YM-3 Microcon centrifugal filter device (EMD Millipore, Billerica, MA, United States).

The Ca²⁺-binding property of the recombinant CaM was determined by gel mobility shift assay as previously described (Anisuzzaman et al., 2010; Hattori et al., 2012; Ye et al., 2017; Tian et al., 2021). Briefly, 0.75 μg CaM was mixed with equal volumes of 0.5 mM CaCl₂, 0.15 mM CaCl₂, 0.05 mM CaCl₂, 0.015 mM CaCl₂, or 0.5 mM ethylenediaminetetraacetic acid (EDTA). Each mixture was incubated at 25°C for 30 min, combined with Laemmli sample buffer, and subjected to SDS-PAGE with no heating.

Proteins were extracted from the salivary glands of 100 fifth-instar SBPH or BPH nymphs, and from the leaf sheaths of rice steams confined within ventilated glass cylinders containing 200 fifth-instar nymphs that had been removed after 2 days. Leaf sheaths from rice stems in empty glass cylinders served as negative controls. The outer three leaf sheaths of each rice stem were harvested and pulverized in liquid nitrogen. Then 2 mL NP40 buffer (Beyotime Institute of Biotechnology, Shanghai, China) was added and the suspension was vortexed at 4°C for 20 min. Each sample was centrifuged at 15,200 × g and 4°C for 5 min. Each supernatant was collected and concentrated to 200 μL with a YM-3 Microcon centrifugal filter device (EMD Millipore, Billerica, MA, United States). The samples were subjected to SDS-PAGE on a 12% (w/v) gradient gel (Bio-Rad Laboratories, Hercules, CA, United States) and transferred onto a polyvinylidene difluoride (PVDF) membrane. Western blotting was performed using primary antibody (1:2,000 dilution) and visualized with the VersaDoc imaging system (Bio-Rad Laboratories) as previously described (Ji et al., 2017; Tian et al., 2021).

Empty pBINPLUS-GFP or pBINPLUS-mCherry vector or their recombinant plasmids harboring CaM or LsPDI1 were transformed by electroporation into A. tumefaciens GV3101. The A. tumefaciens was re-suspended in an infiltration buffer (10 mM MgCl₂, 500 mM MES, and 100 mM acetosyringone). At a final OD₆₀₀ = 0.4, the A. tumefaciens was infiltrated into N. benthamiana leaves with a needleless syringe as previously described (Dong et al., 2020; Ji et al., 2021). The CaM-GFP was expressed for 24 h, then LsPDI1-mCherry was infiltrated into the same region. GFP (488 nm) and mCherry (561 nm) fluorescence signals were observed under a Zeiss LSM750 confocal laser-scanning microscope (CLSM; Carl Zeiss AG, Oberkochen, Germany). At 24 h post-LsPDI1 infiltration, the infiltrated N. benthamiana leaves were collected, and vacuum-infiltrated with 1 mg mL^–1 3,3′-diaminobenzidine (DAB, pH 7.2), and incubated overnight in the dark, and decolorized with trichloroacetaldehyde. DAB staining was performed on 20 leaves.

Rice protoplasts were isolated and transfected as previously described (Zhang et al., 2011; Ji et al., 2021). Briefly, 10-day-old rice seedlings were cut into 0.5 mm segments, transferred to 0.6 M mannitol, and incubated in the dark for 30 min. The plasmolyzed tissues were transferred to a mixture comprising 1.5% (w/v) Cellulase R-10 (Yakult Honsha, Tokyo, Japan), 0.75% (w/v) Macerozyme R-10 (Yakult Honsha), 0.5 M mannitol, 10 mM MES (pH 5.7), 0.1% (v/v) bovine serum albumin (BSA), 10 mM CaCl₂, and 5 mM β-mercaptoethanol for cell wall degradation. Polyethylene glycol-mediated transfection was used to transform the recombinant vector with 2 × 10⁶ rice protoplasts. They were incubated in the dark for 16 h, and their green fluorescence was examined under a Zeiss LSM750 CLSM (Carl Zeiss AG).

Infiltrated tobacco leaf disks and planthopper-infested rice leaf areas were stained with aniline blue to visualize callose deposition as previously described (Bing et al., 2012; Fu et al., 2021). Briefly, treated samples were subjected to 0.05% (w/v) aniline blue staining for 2 h followed by 2.5 g/mL trichloroacetic acid decolorization. The treated samples were observed and photographed under a Zeiss LSM750 CLSM (Carl Zeiss AG) fitted with a UV filter. Fluorescence was quantified with ImageJ v. 1.8.0 (NIH). Each treatment was scored for 20 randomly selected microscopic fields. Each experiment was repeated six times.

Samples were prepared for H₂O₂ analysis as previously described (Lou and Baldwin, 2006; Ji et al., 2017, 2021; Ye et al., 2017). Each single rice stem was confined in a glass cylinder containing 30 fifth-instar nymphs that had been injected with the double stranded RNA of LsCAM (dsLsCAM) or GFP (dsGFP) 2 days earlier. The outer two leaf sheaths of each stem were harvested at various time points after the treatment. Infiltrated N. benthamiana leaf disks and rice leaf sheath samples were collected for H₂O₂ measurement. H₂O₂ concentration was determined with an Amplex-Red Hydrogen Peroxide/Peroxidase Assay Kit (Invitrogen, Carlsbad, CA, United States) (Lou and Baldwin, 2006). Each experiment was repeated six times.

Differences in insect bioassay, genes expression, and H₂O₂ content among treatments were analyzed by one-way ANOVA followed by Duncan’s multiple range tests for multiple treatment comparisons or by Student’s t-tests for pairwise treatment comparisons. Data analyses were performed using Statistica v. 6 (SAS Institute Inc., Cary, NC, United States).

The conserved salivary proteins CaM, enolase, stubble-2, placental protein 11 (PP11), α-N-acetylgalactosaminidase (NAGA), carboxylesterase, regucalcin, and trypsin-26 were selected for RNAi application in SBPH nymphs. Phenotypic variation was observed at 24 h intervals. Morphological defect and lethality were virtually undetectable in the dsGFP-injected SBPH nymph control throughout the test period. The suppression of LsCaM or Lsenolase transcription was lethal to SBPH (Figure 1). Compared with the dsGFP-injected SBPH nymphs, the dsLsenolase-injected and dsLsCaM-injected nymphs presented with significantly lower survival rates 3 and 4 days after dsRNA injection, respectively. At 7 days post-injection, their survival rates were 22 and 62%, respectively (Figure 1A). By contrast, suppression of the other six salivary protein-encoding genes did not cause any abnormalities in the SBPH nymphs (Figures 1A,B).

The qPCR analysis confirmed that the expression levels of all eight genes were markedly lower in the RNAi-treated SBPH nymphs than in the dsGFP-injected nymphs (Figures 1C,D and Supplementary Figure 1). The Lsenolase and LsCaM transcript levels were ∼70% lower at 2–6 days post-injection (Figures 1C,D).

To investigate whether Lsenolase and LsCaM are effectors, we used the EPG technique to study the effects of knocking down LsCaM and Lsenolase on SBPH feeding. EPG profiles the feeding behavior of piercing-sucking insects (Seo et al., 2009; Ji et al., 2017) and distinguishes the non-penetration (NP), pathway (penetration initiation, salivation, stylet movement, and extracellular activity near the phloem; PP), the phloem (N4), and the xylem (N5) phases. LsCaM silencing significantly prolonged NP and PP but significantly shortened N4 (Figure 2A). Silencing LsCaM also significantly reduced honeydew excretion (Figure 2B). Hence, food intake had decreased. By contrast, silencing Lsenolase affected neither the feeding behavior nor the amount of honeydew excreted (Figures 2A,B). Taken together, these results suggest that LsCaM is required for SBPH feeding, whereas enolase is not. The number of eggs laid by the LsCaM-silenced female was 82% lower than those laid by the dsGFP-injected female (Figure 2C).

A BLAST analysis revealed that the CORF of SBPH LsCaM had 96% identity with the CORF of the previous reported BPH NlCaM (Wang et al., 2018; Supplementary Figure 2A). Nevertheless, they had identical protein sequences. Thus, we used dsRNA synthesized from the LsCaM CORF sequence to knock down BPH NlCaM. Injection with dsLsCaM downregulated BPH NlCaM by 66–79% at 2–6 days post-injection (Figure 2D). Therefore, high RNAi efficiency of BPH NlCaM was achieved through dsLsCaM injection.

Consistent with the above finding on SBPH, relative to the dsGFP control, dsLsCaM injection into BPH significantly extended NP and PP but shortened N4 (Figure 2E). Silencing NlCaM (dsLsCaM) reduced by 35% the amount of honeydew excreted by BPH (Figure 2F). Moreover, the fecundity of NlCaM-silenced BPH female adult (dsLsCaM) had declined by 73% (Figure 2G). Thus, NlCaM is also implicated in BPH feeding, fecundity, and survival.

To test whether CaM influenced the SBPH and BPH survival on rice and whether this influence was related to the role of CaM in the planthopper-rice interaction, we compared the survival rate of nymphs feeding on different food matrices. CaM silencing generally reduced the survival rate of SBPH and BPH (Figure 3), and the effect was most pronounced on rice plants: 3 days after dsRNA injection, the survival rate of SBPH and BPH nymphs on rice dropped significantly, and it was 24% (SBPH) and 32% (BPH) at 7 days (Figures 3A,B). In contrast, the survival rate of nymphs with knocked down CaM was higher in insects raised on artificial diet than in those raised on rice; moreover, the survival rate of nymphs with silenced CaM raised on artificial diet was reduced slightly compared with that of control nymphs 5 days after the start of the experiment (Figures 3C,D). Compared with that of nymphs with silenced CaM raised on artificial diet, the corrected survival rate of nymphs with silenced CaM raised on rice was significantly lower 4–7 days post-injection (Figures 3E,F). These results demonstrate that CaM contributes to the survival of SBPH and BPH nymphs raised on rice.

Spatial expression analysis revealed that BPH and SBPH CaM were expressed in all tissues examined and highly expressed in salivary glands (Figures 4A,B). Temporal expression pattern revealed that BPH and SBPH CaM were expressed at all developmental stages (Figures 4C,D).

Calmodulin protein comprises 149 amino acids, and its molecular mass is 16.81 kDa. Previous studies showed that four and five unique peptides were detected in the CaM proteins derived from BPH and SBPH watery saliva proteomes, respectively (Huang et al., 2018; Fu et al., 2021; Supplementary Figure 2B). We hypothesized that BPH and SBPH secrete CaM as they feed on rice. We performed a western blot analysis to test this hypothesis. The antibody revealed ∼16 kDa bands in plants grazed by BPH or SBPH nymphs (Figures 4E,F; lane 3). These bands were absent in uninfested control plants (Figures 4E,F; lane 2). The same bands were also detected in BPH and SBPH salivary gland extracts (Figures 4E,F; lane 1). Hence, CaM is transferred from BPH and SBPH salivary glands to the host plant during feeding.

The CaM protein had no extracellular signal peptides or transmembrane domains but contained four EF-hand Ca²⁺-binding domains (Supplementary Figure 2B). We conducted an electrophoretic mobility shift assay to determine the Ca²⁺-binding capacity of the CaM protein. Recombinant CaM was produced in an E. coli expression system, mixed with various CaCl₂ concentrations, and subjected to sodium dodecyl sulfate–polyacrylamide gel electrophoresis (SDS-PAGE). CaM mobility was lower in the presence of 0.015–0.5 mM CaCl₂ than it was in the presence of 0.5 mM EDTA. Moreover, high CaCl₂ concentrations impeded CaM migration (Figure 4G). Similar electrophoretic mobility shift patterns were reported for other Ca²⁺-binding proteins (Hattori et al., 2012; Ye et al., 2017; Tian et al., 2021), but not for the BSA control protein. When CaM-GFP fusion protein was transiently expressed in rice protoplasts or N. benthamiana leaves, GFP fluorescence was detected in the cytoplasms and nuclei (Figure 4H and Supplementary Figure 3).

Small brown planthopper salivary LsPDI1 is an elicitor that induces Ca²⁺-dependent cell death, reactive oxygen species burst, and callose deposition in plants (Fu et al., 2021). Transient CaM-GFP and LsPDI1-mCherry expression was performed in N. benthamiana leaves to determine whether CaM influences plant immune responses. CaM-GFP expression significantly suppressed LsPDI1-induced cell death (Figure 5A), H₂O₂ accumulation (Figures 5B,C), and callose deposition (Figures 5D,E), whereas only GFP expression control did not (Figure 5). Taken together, these results imply that CaM may suppress plant immune responses by binding free Ca²⁺ in the plant cell cytoplasms and nuclei.

We also measured callose deposition and H₂O₂ accumulation in rice plants infested with dsGFP- or dsLsCaM-injected BPH or SBPH nymphs. Callose deposition and H₂O₂ concentration were significantly higher in the dsLsCaM-BPH-infested rice plants than they were in the dsGFP-BPH-infested plants (Figures 6A,B,D). The same also was true for SBPH (Figures 6A,C,E). Thus, secreted CaM by BPH and SBPH may suppress insect-induced callose deposition and H₂O₂ accumulation.