Introduction

The activation of innate immunity by membrane-localized receptors is a highly conserved mechanism in eukaryotes [1]. In sessile plants, the first line of defense against potential pathogenic microbes or insects relies on cell surface pattern-recognition receptors (PRRs) [2]. The plant genome encodes hundreds of PRRs, including receptor-like kinases (RLKs) and receptor-like proteins (RLPs) [2]. RLKs consist of different types of N-terminal ectodomains, a single transmembrane domain, and a cytoplasmic protein kinase domain, while RLPs have a similar structure but lack the kinase domain [3, 4]. As RLPs lack intracellular signaling domains, they are usually associated with SOBIR1 or SOBIR1-like adaptors to form bimolecular receptor kinases [4]. Despite the large number of PRRs encoded by plant genomes, only a few of them have been well characterized in terms of their function in recognizing microbial-associated molecular patterns (MAMPs), damage-associated molecular patterns (DAMPs), and herbivore-associated molecular patterns (HAMPs) [3, 5]. For instance, FLS2 is known to recognize bacterial flagellin [6], EFR can recognize bacterial elongation factor Tu [7], and EIX2 is responsible for recognizing the fungal ethylene-inducing xylanase [8].

Piercing-sucking insects, such as whiteflies, aphids, and planthoppers, damage plants by feeding or transmitting viruses. They inject a mixture of saliva into plant tissues, which aids their performance by binding to the leaked calcium, degrading extracellular DNA, and regulating plant hormonal signals [9–12]. In response, plants have evolved the ability to detect herbivores by sensing HAMPs present in insect saliva [5, 13]. This triggers pattern-triggered immunity (PTI), including mitogen-activated protein kinase (MAPK) cascades, reactive oxygen species (ROS), and hormone signaling [14–16]. Although a few salivary proteins have been reported to activate plant PTI, the specific PRRs that recognize these proteins and trigger plant immunity are still not well understood [17–19]. One exception is the inceptin receptor (INR), which is a legume-specific LRR-RLP recognizing the Vu-In in insect oral secretions [5]. Moreover, plant PRRs have been widely reported to be targeted by pathogen effectors [20–22]. It remains unknown how insects cope with the immune recognition system.

The whitefly Bemisia tabaci (Hemiptera: Aleyrodidae) and brown planthopper Nilaparvata lugens (Hemiptera: Delphacidae) are notorious pests globally. During the feeding process, both insects secrete abundant species-specific salivary proteins into host plants [23, 24]. The plants have been reported to initiate defense responses when they recognize insect feeding [17, 25]. However, the specific PRRs responsible for this process are unknown. Both insects can successfully ingest phloem saps with limited plant defenses. It remains unclear how insects attenuate plant defenses. Given that many defense mechanisms are conserved across plants [1, 26], it is interesting to investigate whether insect evolves a similar strategy to aid feeding in convergent evolution. In this study, we characterize the function of RLPs in plant immunity responsive to piercing-sucking insects, and demonstrate that B. tabaci and N. lugens independently evolve salivary proteins to regulate the posttranslational modification of RLPs.

Results

BtRDP is most abundantly expressed in salivary glands and secreted into host plants

According to the transcriptomic results, BtRDP (B. tabaci RLP-degrading protein; GenBank accession: MN738093) is the most abundantly expressed gene in salivary glands [24]. It is a secretory protein containing a signal peptide at the N-terminal, and totally 6 unique peptides from BtRDP are detected in B. tabaci watery saliva (Fig. S1). As there is no gene homolog identified in the NCBI nr database, BtRDP is subjected to BLAST analysis against the transcriptomic or genomic data of 30 insect species. The results show that RDP is distributed in all the analyzed Aleyrodidae species, while no RDP homolog is detected in other species (Table S1). Based on phylogenetic analysis, RDPs from five B. tabaci cryptic species are clustered in the same clade, showing a close relationship to RDP from Aleurocanthus spiniferus (Fig. S2).

BtRDP can be secreted into host plants, with two bands between 25-35kDa being detected in Nicotiana tabacum plants infested by B. tabaci, but not in the non-infested plants (Fig. 1a). The smaller high intensity band is the same size as that of BtRDP in salivary glands, and always presents in the B. tabaci-infested plants. For the larger lower intensity band, it presents in infested samples with lower frequency. It is possible that the larger band might be the BtRDP that undergoes modification after being secreted into plants, similar with another salivary effector previously reported [27]. However, we cannot exclude the possibility that this larger band is the non-specific N. tabacum protein induced by B. tabaci feeding. Western-blotting, qPCR and transcriptomic analysis demonstrate that BtRDP is nearly exclusively expressed in the salivary glands (Fig. 1b, c; Fig. S3a), which is significantly different from the actin or ribosomal 18S rRNA (Fig. S4a). In addition, BtRDP is mainly expressed in the adult stages, but lowly expressed in the egg, nymph, and pseudopupa stages (Fig. 1d; Fig. S3b), similar to another salivary protein BtFTSP [28] but different from actin or ribosomal 18S rRNA (Fig. S4b). Immunohistochemical (IHC) staining shows that BtRDP is specifically distributed in principle salivary gland, but not in the accessory salivary gland or other tissues (Fig. 1e). A few salivary proteins have been reported to attach to salivary sheath after secretion [29]. We also stain the salivary sheath secreted by B. tabaci using anti-BtRDP serum. The results show that BtRDP signal is mainly detected in the protuberant structure of sheath (Fig. S5), indicating that BtRDP attaches to salivary sheath after secretion.

Fig. 1.

BtRDP is important for B. tabaci feeding on tobacco plants

RNA interference is firstly employed to investigate the potential role of BtRDP in whitefly feeding, and dsBtRDP treatment efficiently and specifically suppresses the target gene at transcriptional and protein level (Fig. 2a; Fig. S6). We do not observe a significant difference in salivary sheath morphology between dsBtRDP- and dsGFP-treated B. tabaci (Fig. S7a). However, the salivary sheath length of dsBtRDP-treated B. tabaci is significantly shorter than that of the dsGFP-treated control (Fig. S7b). Moreover, dsBtRDP treatment exerts no influence on the whitefly survivorship (Fig. S7c). For reproduction analysis, the dsBtRDP-treated whiteflies oviposit less eggs than the control (Fig. 2b). The whitefly feeding behavior, which can be classified into nonpenetration (np), pathway duration (C), phloem salivation (E1), and phloem ingestion (E2) [30] (Fig. S8), is recorded using the electrical penetration graph (EPG) technique. Compared with dsGFP treatment, dsBtRDP-treated B. tabaci exhibits a dramatic decrease in phloem ingestion (Fig. 2c). Moreover, dsBtRDP-treated whitefly exhibits a significantly increased pathway duration phase and a slightly increased nonpenetration phase (Fig. 2c).

Fig. 2.

Salivary proteins are reported to function in intercellular space and/or extracellular space after secreted into host plants [29, 31]. To investigate the role of BtRDP in different subcellular location of host plants, we construct two transgenic N. tabacum overexpressing BtRDP, one with a complete coding region (oeBtRDP) and the other without signal peptide (oeBtRDP^-sp). The signal peptide of BtRDP may enable protein secretion into the extracellular space. The flag tag is fused to the C-terminal ends of recombinant proteins, respectively. The empty vector (EV) transgenic plants, which exert no significant influence on insect performance compared with the WT (Fig. S9), is used as a negative control. Fecundity assay demonstrates that both oeBtRDP and oeBtRDP^-sp transgenic plants are beneficial to whiteflies, with more eggs oviposited in these two overexpressing plants compared with the EV controls (Fig. 2d-i; Fig. S10). Two-choice experiments exert similar results, with more insects being settled and oviposited in oeBtRDP or oeBtRDP^-sp transgenic plants than the EV controls (Fig. 2d-i; Fig. S10). These results indicate that BtRDP is beneficial to whiteflies, regardless of its subcellular location in host plants.

Transcriptomic sequencing is performed on EV, oeBtRDP#1, and oeBtRDP^-sp#1 transgenic plants. PCA analysis demonstrates that the biological replicates of each transgenic plant can be well clustered (Fig. S11a). Totally 1253 and 1227 differentially expressed genes (DEGs) are identified when comparing oeBtRDP#1 and oeBtRDP^-sp#1 with EV plants, respectively (Table S2; Table S3). Enrichment analysis of DEGs in these two comparison groups exerts similar results, with pathways associated with MAPK signaling, environmental adaptation, signal transduction, and plant-pathogen interaction significantly enriched (Fig. S10b, c). There are 464 DEGs overlap in these two comparison groups (Fig. S11d). Interestingly, the majority of these DEGs exhibit similar expression pattern in oeBtRDP#1 and oeBtRDP^-sp#1 when compared with EV plants (Fig. S11e), suggesting that BtRDP affect some plant pathways regardless of its subcellular location. There are 1084 DEGs identified in comparison of oeBtRDP#1 and oeBtRDP^-sp#1 plants (Table S4). Enrichment analysis show that the DEGs associated with metabolism, diterpenoid biosynthesis, and plant-pathogen interaction are significantly enriched (Fig. S12), suggesting that subcellular location of BtRDP also has non-negligible influence on host plants.

BtRDP interacts with NtRLP4/NtSORBIR1 complex

To investigate the underlying mechanism of BtRDP in improving whitefly feeding, yeast two-hybrid (Y2H) screening against a N. benthamiana cDNA library is conducted using BtRDP^-sp as a bait. Totally seven proteins are found to potentially interact with BtRDP^-sp (Table S5), including an NbRLP4 (accession Niben261Chr07g1310001.1 in Niben261 genome). NbRLP4 contains a predicted N-terminal signal peptide, a malectin-like domain, a LRR domain, followed by a transmembrane domain. NbRLP4, as well as its closest homology in N. tabacum (NtRLP4; GenBank accession: XM_016601109), does not contain the kinase domain (Fig. 3a).

Fig. 3.

The relative expression of NtRLP4 in response to B. tabaci infestation is investigated. qRT-RCR results demonstrate that the transcript level of NtRLP4 is not significantly influenced by B. tabaci (Fig. S13). However, a decrease in protein level of NtRLP4 is detected by western-blotting assay (Fig. 3b). Point-to-point Y2H assays reveal that NtRLP4 without signal peptides and transmembrane domains interacts with BtRDP^-sp (Fig. 3c). To investigate the specificity of NtRLP4-BtRDP interaction, another RLP protein NtCf9 showing the highest sequence similarity to tomato resistance protein Cf9 [32], as well as three randomly selected salivary proteins from B. tabaci, are used as controls. The results show that BtRDP^-sp cannot interact with truncated NtCf9 without signal peptides and transmembrane domains (Fig. 3c). Also, NtRLP4_(23-541) cannot interact with the other three salivary proteins without signal peptides (Fig. S14a, b). In addition, it is the LRR domain, but not the malectin-like domain or the short intercellular region, that is required for the NtRLP4-BtRDP interaction (Fig. 3c). To investigate the shortest LRR sequence responsible for BtRDP interaction, LRR domain is divided into three segments. However, none of these segments interact with BtRDP^{- sp} (Fig. S14a). Co-immunoprecipitation (Co-IP) assays further confirm the specific interaction between BtRDP and NtRLP4 (Fig. 3d). It is the C-terminal myc tag-fused NtRLP4 (NtRLP4-myc), but not the NtCf9 (NtCf9-myc), that can be immunoprecipitated by C-terminal flag tag fused BtRDP (BtRDP-flag). In bimolecular fluorescence complementation (BiFC) assay, YFP fluorescence is observed when co-expressing the N-terminal nYFP tag fused NtRLP4 and N-terminal cYFP tag fused BtRDP^-sp (Fig. S14c). Furthermore, oeBtRDP transgenic plants that overexpressing BtRDP-flag are incubated with anti-flag beads. The results show that endogenous NtRLP4 can be immunoprecipitated, which is significantly different from the control (Fig. S14d). These results suggest that BtRDP interacts with NtRLP4.

The absence of kinase motifs in the C-terminal region of NtRLP4 suggests that it must function in cooperation with other proteins to initiate the signaling. The receptor-like kinase SOBIR1, which contained a kinase domain (Fig. S15a), is reported to be required for the function of several RLPs involved in the innate immunity [4]. Through Co-IP assays, our study demonstrates that NtRLP4 is capable of interacting with NtSOBIR1 (Fig. S15b).

NtRLP4 and NtSOBIR1 are resistant to B. tabaci infestation

The transgenic N. tabacum overexpressing NtRLP4 (oeRLP#1 and oeRLP#2, without any tags) is constructed (Fig. 3e). Compared with the EV control, both oeRLP lines are less attractive to B. tabaci, with less settling and ovipositing insects (Fig. 3f; Fig. S16a). Also, we observe the significantly decreased fecundity of B. tabaci in oeRLP plants (Fig. 3g; Fig. S16b). Similar results are observed after transient overexpression of NtRLP4-GFP in N. tabacum (Fig. S16c, d), indicating that NtRLP4 is resistant to B. tabaci infestation.

Thereafter, the EV and oeRLP#1 plants are subjected to transcriptomic sequencing (Fig. 17a). Totally 729 differentially expressed genes (DEGs) are identified, including 423 up-regulated and 306 down-regulated ones (Table S6). Enrichment analysis demonstrate that the majority of up-regulated DEGs are associated with plant-pathogen interaction, environmental adaptation, MAPK signaling pathway, and signal transduction (Fig. 17b). By contrast, the glutathione metabolism, carbohydrate metabolism, and amino acid metabolism pathways are significantly enriched in down-regulated DEGs (Fig. 17c). Noteworthily, there are many DEGs annotated as RLK/RLP or WRKY transcription factor, with the majority of them being significantly upregulated (Fig. S17d, e). These results suggest an increased defense in oeRLP plants. The crosstalk between salicylic acid (SA) and jasmonic acid (JA) pathways is critical for plant defenses [12]. The increased defense of oeRLP transgenic plant is partially caused by the altered hormonal signaling. The JA-associated genes FAD7 is significantly induced, while the SA-associated genes PAL and NPR1 are repressed in oeRLP plants (Fig. 3h). Transient overexpression of NtRLP4-GFP in N. tabacum plants via Agrobacterium infiltration exert similar results (Fig. S16e). Furthermore, NtRLP4-GFP significantly enhances the H₂O₂ accumulation (Fig. S16f).

We also investigate the function of NtSOBIR1 by transient overexpression and hairpin RNAi (Fig. S18). Overexpression of NtSOBIR1-GFP induces the cell death phenotype five days post agro-injection (Fig. S18b). Insects prefer to settle and oviposit on GFP-expressed N. tabacum than on NtSOBIR1-GFP-expressed control, even no obvious death phenotype is observed two days post agro-injection (Fig. S18c, d). For NtSOBIR1 silencing, no cell death phenotype is observed, while the NtSOBIR1-silenced leaves exhibit reduced growth (Fig. S18f). Compared with EV-silenced control, NtSOBIR1-silenced N. tabacum is slightly attractive to B. tabaci (Fig. S18g, h). Collectively, these results demonstrate that NtSOBIR1 is resistant to B. tabaci infestation.

BtRDP attenuates plant defenses by promoting the degradation of host RLP4

To investigate the interplay between NtRLP4 and BtRDP, we construct transgenic N. tabacum silencing NtRLP4 (RNAi-RLP#1 and RNAi-RLP#2) (Fig. S19a). Compared with EV plants, B. tabaci show a tendency to settle on NtRLP4-silenced plants and lay more eggs (Fig. S19b, c). Thereafter, dsGFP- and dsBtRDP-treated B. tabaci are reared on EV, RNAi-RLP#1, and RNAi-RLP#2, respectively. The dsBtRDP-treated B. tabaci produce less offspring than the control (decreased by 28.3%; P=0.0007) on EV plants (Fig. 4a), consistent with the results on WT plants (Fig. 2b). On RNAi-RLP#1 plant, the impair insect reproduction is still observed after dsBtRDP treatment, but the extent is slightly reduced (decreased by 23.7%; P=0.002) (Fig. 4a). On RNAi-RLP#2 plant, there is no statistical significance between dsGFP and dsBtRDP treatments (Fig. 4a). These results suggest that RNAi-RLP plants are beneficial for insects and can partially, but not completely, rescue the impaired feeding performance of dsBtRDP-treated insects.

Fig. 4.

To investigate the influence of BtRDP on NtRLP4, the GFP-tagged NtRLP4 (NtRLP4-GFP) is transiently co-expressed with mCherry-tagged BtRDP (BtRDP-mCherry), red fluorescent protein (RFP)-mCherry, and BtFTSP1-mCherry via agroinfiltration, respectively. All tags are fused to the C-terminus. BtRDP-mCherry does not influence the location patterns of NtRLP4-GFP (Fig. S20). However, a significant reduction in the fluorescent signal of NtRLP4-GFP is detected when co-expressed with BtRDP-mCherry, but not the controls (Fig. S20). The co-expression of BtRDP-mCherry/NtRLP4-GFP significantly reduces H₂O₂ accumulation (Fig. 4b) and suppresses the expression of the JA-associated genes FAD7 and PDF.12, while inducing the expression of the SA-associated genes PAL and NPR1 (Fig. 4c). Insect bioassays demonstrate that transient overexpression of BtRDP-mCherry rescues the negative effects on whitefly performance caused by NtRLP4-GFP overexpression (Fig. 4d, e). In host choice experiments, B. tabaci display a preference for settling on N. tabacum leaves co-expressing BtRDP-mCherry/NtRLP4-GFP over those co-expressing RFP-mCherry/NtRLP4-GFP (Fig. 4d). We observe almost no preference in B. tabaci feeding on N. tabacum plants co-expressing BtRDP-mCherry/NtRLP4-GFP and plants expressing GFP alone (Fig. S21).

The attenuated NtRLP4-GFP fluorescent signal (Fig. S20), along with the observed decrease in the protein level of NtRLP4 during B. tabaci infestation (Fig. 3b), leads us to speculate on whether BtRDP influences the accumulation of NtRLP4. To this end, the BtRDP^-sp-his and GFP-his proteins are prokaryotic expressed in Escherichia coli. Different concentrations of purified BtRDP^-sp-his and GFP-his proteins are infiltrated into N. tabacum plants overexpressing NtRLP4-myc. Interestingly, the protein level of NtRLP4-myc shows a slight decrease with increasing amounts of BtRDP^-sp-his, while no such effect is observed with GFP-his (Fig. S22). Moreover, the BtRDP^-sp-his protein does not have significant effect on the accumulation of another RLP NtCf9-like (Fig. S22).

Then, we detect the NtRLP4 level in oeBtRDP and oeBtRDP^-sp plants. Compared with the EV control, a reduced NtRLP4 abundance is observed in both oeBtRDP#1 and oeBtRDP#2 transgenic lines, but not the oeBtRDP^-sp plants (Fig. 4f), suggesting the importance of BtRDP secretion in extracellular space. BtRDP attaches to salivary sheath that localizes in plant apoplast (Fig. S5). The presence of signal peptide promotes the secretion of BtRDP into extracellular space. Therefore, we use the complete coding region of BtRDP in subsequent experiments. The NtRLP4-myc and BtRDP-flag are transiently co-expressed in N. benthamiana plants through Agrobacterium infiltration. The results show that BtRDP-flag significantly decreasing the protein level of NtRLP4-myc, but not the NtCf9-myc (Fig. 4g). qRT-PCR analyses demonstrate that the transcript level of NtRLP4 remains unchanged (Fig. S23). These findings indicate that the reduction in protein level of NtRLP4 is not caused by transcriptional changes but rather to degradation.

The ubiquitin system and autophagy are two major pathways regulating protein degradation [33]. At first, we employ MG132, a 26S proteasome inhibitor known to inhibit the degradation of ubiquitin-conjugated proteins [34]. The results show that MG132 efficiently prevents the degradation of NtRLP4-myc by BtRDP-flag (Fig. 4h). Then, we employ autophagy inhibitor 3-Methyladenine (3-MA) and bafilomycin A1 (BAF) that can block autophagy in plants [35]. However, the decreased protein level of NtRLP4-myc is still observed in the presence of 3-MA or BAF (Fig. S24). Additionally, NtRLP4-myc and HA-tagged ubiquitin (HA-UBQ) are co-expressed in N. benthamiana. Total proteins are extracted from N. benthamiana leaves and then subjected to immunoprecipitation with anti-myc beads. The ubiquitin attached to NtRLP4 is determined by immunoblot using anti-HA antibody. Leaves expressing NtRLP4-myc or UBQ-HA along are used as negative controls. The results show that a strong UBQ signal is detected in the immunoprecipitated product only when NtRLP4-myc and UBQ-HA are co-expressed (Fig. 4i). The molecular weight of UBQ-attached NtRLP4-myc is remarkably larger than NtRLP4-myc itself (Fig. 4i). These results suggest that BtRDP promotes NtRLP4 degradation through the ubiquitin system, but not the autophagy pathway.

Degrading plant RLP4s by salivary effectors is a common strategy for insects

To investigate the ability of BtRDP to degrade RLPs from other host plants of B. tabaci, Solanum lycopersicum RLP4 (SlRLP4; GenBank accession: XP_004232910) is selected for analysis. Sequence analysis reveals a similar gene structure between SlRLP4 and NtRLP4 (Fig. S25; Fig. S26a). The interaction between BtRDP and SlRLP4 is confirmed through Y2H and Co-IP assays (Fig. S26b, c). Moreover, we transiently co-express SlRLP4-myc and SlCf9-like-myc with varying concentrations of BtRDP-flag or GUS-flag, with the myc and flag tags fused to the C-terminus. The results show that BtRDP-flag exerts no influence on the accumulation of SlCf9-like-myc protein. By contrast, the protein level of SlRLP4-myc is significantly reduced in the presence of BtRDP-flag (Fig. S26d), indicating the conserved function of BtRDP in B. tabaci adaptation to different host plants.

BtRDP is an Aleyrodidae-restricted gene that does not have homologues in other species. To explore whether RLP degradation by salivary effectors is conserved among different herbivorous insects, a RLP4 homology from Oryza sativa (OsRLP4; GenBank accession: XP_015645303; Fig. 5a) and twenty salivary proteins from planthopper species without signal peptides are paired to identify the potential interactions using Y2H assays. As a result, only the salivary protein NlSP694^{- sp} (GenBank accession: MF278694) interacts with OsRLP4_(29-551) (OsRLP4 without signal peptides and transmembrane domains), while the other salivary proteins analyzed do not (Fig. 5b, c; Fig. S27). NlSP694 does not show any sequence or structure similarity to BtRDP (Fig. S28). Based on truncation analysis, NlSP694^-sp interacts with both the LRR domain and malectin-like domain of OsRLP4 (Fig. 5c), slightly different from that of the BtRDP-NtRLP4 interaction (Fig. 3c). Considering that both BtRDP and NlSP694 interact with the LRR domain of RLP4, the specific region of LRR domain responsible for NlSP694 binding is further investigated. The LRR domain of OsRLP4 is divided into three segments, in accordance with the LRR domain division in NtRLP4 (Fig. S14a). NlSP694^-sp interacts with the truncated LRR domain (OsRLP4_(480-526)) (Fig. 5c). By contrast, none of the truncated LRR domain of NtRLP4 interact with BtRDP^-sp (Fig. S14a). These results suggest that the binding sites of salivary proteins to RLP4 may be different between N. lugens and B. tabaci.

Fig. 5.

NlSP694 is a planthopper-specific salivary protein that is nearly exclusively expressed in salivary glands (Fig. 5d; Fig. S29). The function of NlSP694 or its gene homologs in other planthopper species has not been well investigated previously. Silencing NlSP694 does not impact insect survivorship (Fig. 5e). However, significant reductions in honeydew excretion and impaired fecundity are observed in dsNlSP694-treated N. lugens (Fig. 5f, g), indicating the importance of NlSP694 in insect feeding.

Subsequently, we transiently co-express OsRLP4-myc with a complete coding region of NlSP694 (C-terminal flag tagged, NlSP694-flag). Another salivary protein NlSP7-flag and NtCf9-myc are used as negative controls. The results demonstrate that NlSP694-flag specifically and significantly decreases the protein level of OsRLP4-myc, which differs significantly from that of the NlSP7-flag (Fig. 5h). Additionally, we demonstrate that NlSP694-flag fails to decrease the protein level of NtCf9-myc (Fig. 5h). These findings suggest that N. lugens potentially evolves salivary effectors to target rice RLPs.

Discussion

Plants have evolved a complex innate immune system to protect themselves against pathogen/insect invasion, and the cell surface-localized PRRs function on the frontline [2]. To counteract this response, the pathogens have been reported to develop various methods to avoid plant PRR-mediated PTI for their own advantage [36]. However, it remains unknown whether insects take the similar strategy to sabotage plant immunity. In this study, we demonstrate that the RLP4/SOBIR1 complexes confer plant resistance against herbivorous insects. In response, both planthoppers and whiteflies secrete salivary effectors that facilitate the degradation of defensive RLP4, leading to attenuated plant immunity. Our results suggest that manipulation of PRRs may be a conserved strategy for herbivorous insects.

PRRs recognize modified-self or non-self patterns and activate a series of immune signaling, which subsequently regulate downstream defense responses and confer plant resistance against invaders [2, 3]. While various insect salivary components have been reported to trigger plant immune responses, the specific PRRs responsible for this process are elusive. Our study demonstrates that RLP4 exerts a significant impact on insect performance by modulating plant hormones and ROS (Fig. 3f-h; Fig. S16-S17), highlighting the critical role of this gene in plant immunity. Nevertheless, the exact ligand for RLP4 remains unknown. It is possible that malectin-LRR receptor RLP4 responds to the DAMPs or HAMPs that are released during insect feeding [37], which deserved further investigation. RLP4 lacks the intracellular kinase domain, and forms bimolecular receptor kinases by interacting with SOBIR1 (Fig. S15). Overexpression of NtRLP4 or NtSOBIR1 enhances insect feeding, while silencing of either gene exerts the opposite effect (Fig. Fig. S16-S19). The association between RLPs and SOBIR1 protein kinases is a structural and functional equivalent to genuine receptor kinases [4]. Therefore, disrupting the stability of the RLP4/SOBIR1 complex can potentially undermine the immune recognition process.

As pivotal regulators in immune signaling, PRRs are frequently targeted by multiple pathogens through diverse mechanism. For example, Pseudomonas syringae hampered the functionality of receptor complexes by disrupting the ligand-dependent association between receptor-like kinase BAK1 and the flagellin receptor FLS2 [22], while Phytophthora infestans fostered alternative splicing of the host RLPKs to enhance infection [21]. PRRs are usually subjected to degradation via post-translational modifications [38]. In plants, the activity and stability of PRRs undergo stringent regulation since excessive activation of PRRs can be detrimental [20, 39]. One prevalent regulatory mechanism involves ubiquitination followed by subsequent degradation mediated by the 26S proteasome [38]. Mounting evidence has demonstrated the exploitation of such mechanisms by bacteria, fungi, viruses, and oomycetes to facilitate PRR degradation [40–43]. Nevertheless, the occurrence of effector-mediated PRR degradation in insects is unreported. Our study reveals that the salivary effectors BtRDP from B. tabaci and NlSP694 from N. lugens interact with plant RLP4s, leading to their degradation in an ubiquitin-dependent manner (Fig. 4f-i). Notably, BtRDP and NlSP694 share no sequence or structural similarity and lack resemblance to known eukaryotic ubiquitin-ligase domains. Their interaction with RLP4s occurs in the extracellular space (Fig. 3d; Fig. 5c), whereas the ubiquitin-proteasome system primarily functions in the cytosol and nucleus [44]. Furthermore, NtRLP4 reduction is observed only in oeBtRDP transgenic plants, not in oeBtRDP^-sp plants (Fig. 4f), suggesting that BtRDP exerts its influence on NtRLP4 in the extracellular space. These findings might exclude the possibility that BtRDP/NlSP694 harbors E3 ligase activity that directly ubiquitinates RLP4s in plant cells. It is likely that salivary proteins indirectly influence the post-translational modification of RLP4, increasing its degradation rate, which warrants further investigation.

Herbivorous insects use a combination of salivary components to facilitate successful feeding [29]. Previous studies have shown that B. tabaci suppresses plant defenses by leveraging SA-JA crosstalk, with the salivary protein Bt56 playing a pivotal role during this process [12]. Bt56 triggers the SA-signaling pathway by directly interacting with a KNOTTED 1-like homeobox transcription factor [12]. This study investigates the effects of another salivary effector, BtRDP, on plant hormones, and finds that it elicits similar effects to Bt56, albeit primarily impacting the signal recognition process (Fig. 3h; Fig. S16e). This leads us to speculate that different salivary effectors from the same insect species might simultaneously target the same signaling pathway to ensure effective immune suppression. Our analysis reveals that BtRDP interacts with other host genes in addition to NtRLP4 (Table S5), and silencing NtRLP4 cannot completely rescue the impaired feeding performance of dsBtRDP-treated insects (Fig. 4a). Salivary proteins exert a crucial role in mediating the interactions between herbivores and plants. The dynamic interplay between plants and insects might be highly complex, one salivary protein can target multiple plant pathways, and one plant pathway is influenced by multiple salivary proteins.

In conclusion, this study shows that suppressing PRR-mediated plant immunity may be the common strategy for the successful feeding of herbivorous insects. Whiteflies and planthoppers independently evolve salivary effectors that facilitate the degradation of to degrade defensive RLP4 of host plants in an ubiquitin-dependent manner, which can therefore disrupt the RLP4/SOBIR1 complexes for signal recognition (Fig. 6).

Fig. 6.

Material and methods

Insects and plants

A colony of B. tabaci (cryptic species MED) was originally collected from a soybean field in Suzhou, Anhui Province, China. The insects were subsequently maintained on N. tabacum cv. K326 plants under controlled laboratory conditions (25 ± 1 °C, 50–70% relative humidity, 16:8 h light:dark photoperiod). For experimental use, N. benthamiana and N. tabacum plants were grown in a climate-controlled growth chamber (23 ± 1 °C, 16:8 h light: dark cycle).

Phylogenetic analysis of RDP

To identify potential RDP homologs across insect species, we performed a BLASTp search using the BtRDP protein sequence as a query against the predicted proteomes of 30 insect species. Hits were filtered using an E-value cutoff of 10^-5. The retrieved RDP homologs were aligned using MAFFT (v7.310) with default parameters [45]. To improve alignment quality, we removed ambiguously aligned regions using Gblocks with default settings. The optimal amino acid substitution model (JTT+G) was determined using ModelTest-NG (v0.1.6). Phylogenetic relationships were inferred using maximum likelihood analysis in RAxML (v0.9.0) with 1000 bootstrap replicates to assess branch support [46]. The final tree was visualized and annotated using iTOL (https://itol.embl.de/).

Preparation of anti-BtRDP and anti-NtRLP4 serum

The coding sequence of BtRDP (excluding the signal peptide) was PCR-amplified and cloned into the pET-28a vector (Novagen) to generate a C-terminal 6×His-tagged fusion protein. The recombinant plasmid was transformed into Escherichia coli Transetta (DE3) competent cells (TransGen Biotech, Beijing, China). Protein expression was induced with 0.5 mM isopropyl-β-D-thiogalactoside (IPTG) at 28°C for 8 h. Bacterial cells were harvested by centrifugation and lysed via ultrasonication for 30 min. The recombinant BtRDP^-sp-His protein was purified under native conditions using Ni-NTA agarose (Qiagen, Hilden, Germany) according to the manufacturer’s protocol.

Purified BtRDP^-sp-His protein was used to immunize rabbits (Hua’an Biotechnology, Hangzhou, China) for custom polyclonal antibody production. The anti-NtRLP4 serum, prepared by immunizing rabbits with peptide GQNSDRPISTKNST, was produced via the custom service of Genscript Biotechnology Company (Nanjing, China). The anti-BtRDP serum was subsequently conjugated to Alexa Fluor™ 488 NHS Ester (#A20000, Thermo Fisher Scientific) following standard labeling protocols.

The specificity of anti-BtRDP serum was tested by comparing dsGFP- and dsBtRDP-treated Bemisia tabaci. The results showed that dsBtRDP significantly reduced the target gene. Similarly, the specificity of anti-NtRLP4 serum was tested by comparing RNAi-EV and RNAi-NtRLP4 plants, which showed a reduced NtRLP4 expression in RNAi-NtRLP4 plants.

Immunohistochemistry (IHC) staining

Dissected B. tabaci heads with attached salivary glands and Parafilm-embedded salivary sheaths were fixed in 4% paraformaldehyde (#E672002, Sango Biotechnology, Shanghai) for 30 min. Salivary sheaths adherent to Parafilm were similarly excised and fixed under identical conditions. After fixation, samples were washed three times with PBST (PBS containing 0.1% Tween-20) and blocked with 10% FBS for 2 h at room temperature. The tissues were then incubated overnight at 4°C with Alexa Fluor 488-conjugated anti-BtRDP serum (1:200 dilution), followed by co-staining with phalloidin-rhodamine (1:500, Thermo Fisher Scientific) for 30 min and 4’,6-diamidino-2-phenylindole (DAPI) solution (#ab104139, Abcam, Cambridge, USA) for nuclear visualization. Fluorescent images were acquired using a Leica SP8 confocal microscope (Leica Microsystems) to examine BtRDP localization patterns

Protein extraction and Western-blotting assays

To identify secreted salivary proteins in N. tabacum plants, approximately 200 B. tabaci adults were confined in a 3-cm diameter feeding cage and allowed to feed on host plants for 24 h. Following feeding, the infested tobacco leaves were collected, immediately frozen in liquid nitrogen, and ground to a fine powder. The powdered tissue was then homogenized in RIPA Lysis Buffer (#89900, Thermo Fisher Scientific) for protein extraction. To examine the effects of insect feeding on NtRLP expression, groups of approximately 200 B. tabaci were allowed to feed on 4-5 true-leaf stage tobacco plants for 12, 24, or 48 h, with uninfested plants serving as controls. For silencing/overexpression efficiency tests, leaf samples were collected at 48 h post-Agrobacterium infiltration, along with transgenic tobacco plants at the 4-5 true-leaf stage. For insect tissue analysis, various B. tabaci organs were carefully dissected in 1×PBS buffer, including carcasses (20), fat bodies (20), guts (40), salivary glands (60), and ovaries (10), with sample sizes indicated in parentheses. All dissected tissues were transferred to lysis buffer for protein extraction. Total protein concentrations in both insect and plant homogenates were quantified using a BCA Protein Assay Kit (#CW0014S, CwBiotech, Taizhou, China). Equal protein amounts were separated by SDS-PAGE and electro-transferred to PVDF membranes.

Membranes were probed with the following primary antibodies: anti-FLAG (1:10,000, #MA1-91878, Thermo Fisher Scientific), anti-MYC (1:10,000, #MA1-21316, Thermo Fisher Scientific), anti-GFP (1:10,000, #MA5-15256, Thermo Fisher Scientific), anti-NtRLP4 (1:5,000, GenScript Biotechnology, Nanjing, China), and anti-BtRDP serum (1:5,000, Huaan Biotechnology, Hangzhou, China). Following primary antibody incubation, membranes were treated with HRP-conjugated secondary antibodies: goat anti-mouse IgG (1:10,000, #31430, Thermo Fisher Scientific) or goat anti-rabbit IgG (1:10,000, #31460, Thermo Fisher Scientific). Protein bands were visualized using an AI 680 image analyzer (Amersham Pharmacia Biotech, Buckinghamshire, UK), with Ponceau S staining used to verify equal protein loading.

Quantitative real-time PCR (qRT-PCR) analysis

To prepare tissue-specific samples from B. tabaci, adult females were dissected to isolate carcasses (10), fat bodies (10), guts (20), salivary glands (40), and ovaries (10). Additionally, different developmental stages, including eggs (50), nymphs (20), pupae (20), and adult males (20) and females (20), were collected for analysis. The number of insects in each sample is indicated in parentheses. To extract RNA from N. tabacum plants, the samples were grinded using liquid nitrogen after collection. Total RNA was extracted using the TRIzol Total RNA Isolation Kit (#9109, Takara, Dalian, China) following the manufacturer’s instructions. RNA purity and concentration were assessed using a NanoDrop spectrophotometer (Thermo Fisher Scientific). First-strand cDNA was synthesized from equal amounts of RNA (1 µg) using the HiScript II Q RT SuperMix (#R212-01, Vazyme, Nanjing, China). Gene expression analysis was performed using a Roche LightCycler® 480 system (Roche Diagnostics, Mannheim, Germany) with SYBR Green Supermix (#11202ES08, Yeasen, Shanghai, China). The thermal cycling protocol consisted of initial denaturation at 95°C for 5 min, and amplification for 40 cycles of 95°C for 10 s and 60°C for 30 s. Primers were designed with Primer Premier v6.0 (Table S7). B. tabaci Tubulin and N. tabacum Tubulin served as internal controls for normalization. Relative gene expression was calculated using the 2^−ΔΔCt method [47]. Three to six independent biological replicates, each repeated twice, were performed.

RNA interference

The target gene DNA sequences were amplified using primers listed in Table S7 and subsequently cloned into the pClone007 Vector (#TSV-007, Tsingke Biotechnology, Beijing, China). For dsRNA production, PCR-generated DNA templates containing T7 promoter sequences were transcribed in vitro using the T7 High Yield RNA Transcription Kit (#TR101-01, Vazyme Biotech, Nanjing, China). The RNAi experiments were performed according to established methods [48]. Briefly, newly emerged adult whiteflies were anesthetized with carbon dioxide for 5-10 seconds. Approximately 50 nL of dsRNA solution (2 μg/μL) was microinjected into the mesothorax of each insect using a FemtoJet microinjection system (Eppendorf, Hamburg, Germany). Following injection, the whiteflies were maintained on healthy N. tabacum leaves under standard rearing conditions. Only actively moving insects were selected for subsequent analyses. Gene silencing efficiency was evaluated on day 4 post-injection using both qRT-PCR and Western-blotting approaches.

Generation of transgenic plants

We generated two transgenic N. tabacum lines expressing different BtRDP variants: one containing the full coding sequence (oeBtRDP) and another lacking the signal peptide (oeBtRDP^{- sp}), which may affect protein secretion. The target sequences, along with the full-length NtRLP4 coding sequence and an NtRLP4 hairpin RNAi construct, were each cloned into the pBWA(V)HS vector with C-terminal FLAG tags. All constructs were transformed into A. tumefaciens strain GV3101 and used to generate transgenic tobacco plants (K326 background) through Agrobacterium-mediated transformation performed by Wuhan Boyuan Biological Company. Transgenic lines were verified by qRT-PCR and western blotting, with empty vector-transformed plants serving as controls, and two independent lines for each construct were selected for further experiments.

Target gene silencing in plants

The hairpin RNAi construct targeting NtSOBIR1 was generated by cloning the specific fragment into the pBWA(V)HS plasmid using BsaI and Eco31I restriction enzymes. The recombinant vector was then transformed into A. tumefaciens strain GV3101 and subsequently infiltrated into N. tabacum leaves using the agroinfiltration method. Following a 48-h incubation period under standard growth conditions, silencing efficiency was quantitatively evaluated by qRT-PCR analysis.

Transcriptomic sequencing

To investigate the transcriptional effects of NtRLP4, BtRDP, and BtRDP^-sp in N. tabacum, leaf samples from EV, oeRLP4#1, oeBtRDP#1, and oeBtRDP^-sp#1 transgenic plants were homogenized in TRIzol Reagent (#10296018, Invitrogen, Carlsbad, CA, USA) for total RNA extraction following the manufacturer’s protocol. Afterwards, total RNA was extracted according to the instructions of the manufacturer, and the RNA samples were sent to Novogene Institute (Novogene, Beijing, China) for transcriptomic sequencing. Briefly, poly(A)+ RNA was isolated from 20 μg pooled total RNA using oligo(dT) magnetic beads, followed by fragmentation in divalent cation buffer at 94°C for 5 min. First-strand cDNA synthesis was performed using N6 random primers, with subsequent second-strand synthesis to generate double-stranded cDNA. After end-repair and Illumina adapter ligation, the products were PCR-amplified (15 cycles) and purified using a QIAquick PCR Purification Kit (#28104, Qiagen, Hilden, Germany). The resulting cDNA libraries were quantified and quality-checked before paired-end sequencing on an Illumina NovaSeq 6000 platform. All raw sequencing data were deposited in the National Genomics Data Center under accession number PRJCA025907.

Analysis of transcriptomic data

The raw sequencing reads were quality-filtered using Illumina’s internal software, and the resulting clean reads from each cDNA library were aligned to the N. tabacum reference genome in Sol Genomics Network (https://solgenomics.net/ftp/genomes/Nicotiana_tabacum/edwards_et_al_2017/) using HISAT v2.1.0 [49]. Low-quality alignments were removed using SAMtools (v1.7) with default parameters [50]. Transcript abundance was quantified as Transcripts Per Million (TPM) using Cufflinks (v2.2.1) [51]. Differential expression analysis was performed with DESeq2 (v2.2.1) [52], identifying significantly DEGs using thresholds of |log2(fold change)| > 1 and adjusted p-value < 0.05. PCA of transcriptome profiles was conducted using a custom R script plotPCA (https://github.com/franco-ye/TestRepository/blob/main/PCA_by_deseq2.R). For functional annotation, KEGG pathway enrichment analysis was performed in TBtools (v2.083) using a one-sided hypergeometric test: , [53]. In this software, enriched P-values were calculated according to one-sided hypergeometric test: with N representing the number of gene with KEGG annotation, n standing for the number of DEGs in N, M indicating the number of genes in each KEGG term, and m suggesting the number of DEGs in each KEGG term.

Insect bioassays

To survivorship analysis, newly emerged B. tabaci adults were microinjected with dsRNA and placed on N. tabacum leaves. After 24 h, any dead individuals (potentially due to injection injury) were removed. For each treatment, 20-30 surviving adults were transferred to leaf cages, and mortality was recorded daily for 10 days. Three independent biological replications were performed.

For fecundity analysis, dsRNA-treated virgin females were paired with untreated males in leaf cages. Five such pairs were maintained per cage (n=10-12 cages per treatment). Egg deposition was quantified daily for 3 consecutive days under controlled conditions. Each male-female pair was considered an experimental replicate.

For host preference test, detached N. tabacum leaves with moist cotton-wrapped petioles were arranged in a 50-cm diameter choice arena with a central release chamber. Forty adult whiteflies were released per chamber, and settling preference was recorded at 3, 6, 12, 24, 36, and 48 h post-release. The experiment included 10-12 replicates per treatment, with leaf positions randomized between replicates to eliminate positional bias.

EPG recording and analysis

Feeding behavior of B. tabaci was monitored using a GiGA-8d EPG amplifier (Wageningen Agricultural University; 10 TΩ input resistance, <1 pA input bias current) with 50× gain and ±5 V output range. Prior to recording, dsRNA-treated whiteflies were starved for 12 h on moist filter paper, briefly anesthetized with CO₂ (10 s), and connected to the amplifier via a gold wire electrode (20 μm diameter × 5 cm length) attached to the pronotum using conductive silver glue. Plant electrodes consisted of copper wires (2 mm diameter × 10 cm length) inserted into the soil of potted N. tabacum. Recordings were conducted for 8 h in a Faraday cage (120×75×67 cm; Dianjiang, Shanghai), with only insects surviving the full duration retained for analysis.

Recorded waveforms were processed using PROBE 3.4 software (Wageningen Agricultural University) and categorized into four distinct feeding phases: nonpenetration (np), pathway duration (C), phloem salivation (E1), and phloem ingestion (E2) as defined in a previous study [30]. Each treatment included ≥ 10 biological replicates, with dsGFP-injected and untreated whiteflies serving as controls.

Agrobacterium-mediated plant transformation in N. tabacum and N. benthamiana

The recombinant expression vectors were introduced into A. tumefaciens strain GV3101 via heat shock transformation. Transformed cells were selected on LB agar plates supplemented with 50 μg/mL kanamycin and 10 μg/mL rifampicin, followed by incubation at 28°C for 60 h. Positive colonies were inoculated into liquid LB medium with the same antibiotics and grown to log phase. Bacterial cells were then pelleted by centrifugation at 2,400 g for 2 min and resuspended in induction buffer (10 mM MgCl₂, 10 mM MES [pH 5.6], 200 μM acetosyringone) adjusted to specific OD₆₀₀ values depending on the experiment.

For standard experiments, bacterial suspensions were normalized to OD₆₀₀ = 1.0, while degradation assays used three concentrations (OD₆₀₀ = 0.05, 0.3, and 1.0). Equal volumes of selected bacterial suspensions were mixed and infiltrated into leaves of 4–5-week-old N. tabacum and N. benthamiana plants using a needleless syringe. Infiltrated plants were maintained under controlled conditions for subsequent analysis.

Diaminobenzidine (DAB) staining

To detect H₂O₂ accumulation in N. tabacum leaves, leaf samples were immersed in 1 mg/mL 3,3’-diaminobenzidine tetrahydrochloride (DAB-HCl, #D8001, Sigma-Aldrich, St. Louis, MO, USA) solution (pH 3.8) for 6 hours at room temperature, allowing H₂O₂-dependent brown polymerization to occur. Following incubation, the DAB solution was replaced with 100% ethanol for chlorophyll removal, and samples were decolorized overnight at 65°C. The destained leaves were then photographed under standardized conditions using a Canon EOS 80D digital camera (Canon Inc., Tokyo, Japan).

Interaction assays between two proteins

In the Y2H screening assay, the coding sequence of BtRDP^-sp was cloned into the pGBKT7 bait vector (Clontech, USA), while a N. benthamiana cDNA library was constructed in the pGADT7 prey vector (Biogene Biotech, Shanghai, China). These recombinant vectors were co-transformed into yeast strain Y2HGold and plated on QDO medium (SD/-Ade/-His/-Leu/-Trp, #630428, Takara) for 3 days at 30°C. Positive colonies were cultured in QDO liquid medium, and plasmids were extracted using a yeast DNA kit (#DP112-02, TIANGEN, Beijing, China) before transformation into E. coli DH5α (#TSC-C14, Tsingke) for sequencing (YouKang Biotech, China) to identify interacting partners.

In the Y2H point-to-point verification assay, BtRDP^-sp and NlSP694^-sp, and mutant variants of NtRLP4/SlRLP4/ OsRLP4 were cloned into either pGBKT7 or pGADT7 vectors using primers from Table S7. Yeast co-transformants were first selected on DDO medium (SD/-Leu/-Trp) (#630417, Takara) for 3 days at 30°C, then spotted onto QDO medium to confirm interactions through growth after another 3 days of incubation.

In the Co-IP assays, Full-length sequences of BtRDP, NlSP694, NtSOBIR1, NtRLP4, SlRLP4, and OsRLP4 were cloned into lic-myc or lic-flag vectors for C-terminal tagging, with BtFTSP-flag serving as a negative control. Five-week-old N. benthamiana leaves were agroinfiltrated with these constructs or collected from oeBtRDP transgenic plants. Total proteins were extracted from 1 g frozen tissue in IP lysis buffer (#87788, Thermo Scientific) with protease inhibitors (#56079200, Roche, Switzerland), followed by centrifugation at 1,000 × g for 20 min. Cleared lysates were incubated with anti-flag beads (#L00432-1, GenScript, Nanjing, China) for 4 h at 4°C, washed with PBS, and eluted in 2× SDS-PAGE buffer (500 mM Tris-HCl, pH=6.8, 50% glycerin, 10% SDS, 1% bromophenol blue, and 2% β-mercaptoethanol) for Western-blotting assay.

In the BiFC assay, BtRDP^-sp, BtFTSP^-sp, NtRLP4, and NtCf9-like were cloned into the pCV-cYFP or pCV-nYFP vectors, respectively. After transforming A. tumefaciens GV3101, pairwise combinations were co-infiltrated into N. benthamiana leaves. Following 36-48 h incubation in a growth chamber, reconstituted YFP fluorescence was visualized using a Leica SP8 confocal microscope with standard filter sets for YFP detection.

RLP degradation by salivary effectors among different herbivorous insects

To examine salivary protein effects on RLPs, we co-infiltrated A. tumefaciens GV3101 strains carrying various constructs into N. benthamiana leaves. The recombinant vectors expressed C-terminally tagged proteins: BtRDP-flag, NlSP694-flag, GUS-flag, NlSP7-flag, and myc-tagged RLPs (NtRLP4, SlRLP4, OsRLP4, SlCf9, NtCf9). Bacterial suspensions were mixed at predetermined OD₆₀₀ ratios before infiltration, with leaf samples collected 48 hours post-infiltration for subsequent analysis.

For confocal microscopy studies, N. benthamiana leaves were co-infiltrated with A. tumefaciens carrying: (1) NtRLP4-GFP plus BtRDP-mCherry, (2) NtRLP4-GFP plus RFP-mCherry, or (3) NtRLP4-GFP plus BtFTSP-mCherry. The samples were imaged by confocal microscopy at 48 h post injection. The same parameters were used in photographing.

This study also conducted a treatment on N. benthamiana plants that were overexpressing NtRLP4-myc using purified proteins. Briefly, the BtRDP^-sp and GFP coding sequences were cloned into pET-28a for 6×His-tagged expression in E. coli Transetta (TransGen Biotech). Protein expression was induced with 0.5 mM IPTG at 25°C for 8 h, followed by purification using Ni-NTA agarose and buffer exchange into PBS using a 3-kDa molecular-weight cutoff Amicon Ultra-4 Centrifugal Filter Device (Millipore, MA, USA). Different concentrations of purified proteins were infiltrated into leaves of N. benthamiana plants stably expressing NtRLP4-myc, with samples collected 24 h post-treatment.

To investigate degradation pathways, we infiltrated N. benthamiana leaves with: 50 μM MG132 (#M7449, Sigma-Aldrich, Steinheim, Germany), 5 μM 3-MA (#M9281, Sigma-Aldrich,), or 10 nM BAF (#BML-CM110-0100, Enzo Life Sciences, Farmingdale, USA) at 24 h post-agroinfiltration. The samples were incubated for additional24 h. Protein levels were detected using Western-blotting assay.

In planta ubiquitination assay

Ubiquitination detection in N. benthamiana was performed according to the methods previously reported [54]. The recombinant vector for transient expression of the HA-fused ubiquitin protein was kindly provided by Dr. Chao Zheng (Ningbo University). Myc-tagged proteins were co-expressed with HA-UBQ by agroinfiltration, and 50 μM MG132 was injected into the N. benthamiana leaves one day post A. tumefaciens infiltration. Immunoprecipitation was performed using anti-myc beads (ytma-20, ChromoTek , Planegg-Martinsried, Germany). Ubiquitination signals were detected by immunoblot with anti-HA antibody (HT301, TransGen Biotech, Beijing, China).

Scanning electron microscopy (SEM)

B. tabaci adults were allowed to feed on artificial diet solutions for 24 h. Parafilm sections containing salivary sheaths were carefully excised and rinsed with 1× PBS (pH 7.4) to remove residual diet. The samples were mounted on aluminum stubs, desiccated under vacuum, and sputter-coated with gold before imaging with a TM4000 II Plus SEM (Hitachi, Tokyo, Japan). Salivary sheath length (from apex to base) was measured, with 20 sheaths analyzed per treatment to ensure statistical reliability.

Statistical analysis

The log-rank test (SPSS Statistics 19, Chicago, IL, USA) was used to determine the statistical significance of survival distributions. Two-tailed unpaired Student’s t-test (comparisons between two groups) or one-way ANOVA test followed by Tukey’s multiple comparisons test (comparisons among three groups) was used to analyze the results of qRT-PCR, fecundity analysis, and host choice analysis. The EPG data were first checked for the normality and homogeneity of variance, and those not fitting a normal distribution were log₁₀ transformed before Student’s t test, as described in previous studies [30]. Data were graphed in GraphPad Prism 9.

Data availability

All data is included in the manuscript and/or supporting information. The transcriptomic data have been submitted to the National Genomics Data Center under accession number PRJCA025907. The original blot images have been displayed in Fig. S30-S33.

Acknowledgements

This project has received funding from the National Key Research and Development Plan in the 14th five-year plan (2021YFD1401100: H.J.H, C.X.Z.), National Natural Science Foundation of China (32422075: H.J.H.; U23A6006: J.P.C and C.X.Z.), and the Natural Science Foundation of Zhejiang Province (LDQ24C140001: H.J.H).

Additional information

Author contributions

H-JH, J-ML, C-XZ, and J-PC planned and designed the research. XW, J-BL, Y-ZW, H-JH, H-BL, Z-LZ, and Z-XY performed experiments and analyzed data. X-HZ provided the EV plants. H-JH and XW drafted the manuscript.

Additional files

Supplementary file 1