A rhabdovirus accessory protein inhibits jasmonic acid signaling in plants to attract insect vectors

Abstract

Plant rhabdoviruses heavily rely on insect vectors for transmission between sessile plants. However, little is known about the underlying mechanisms of insect attraction and transmission of plant rhabdoviruses. In this study, we used an arthropod-borne cytorhabdovirus, Barley yellow striate mosaic virus (BYSMV), to demonstrate the molecular mechanisms of a rhabdovirus accessory protein in improving plant attractiveness to insect vectors. Here, we found that BYSMV-infected barley (Hordeum vulgare L.) plants attracted more insect vectors than mock-treated plants. Interestingly, overexpression of BYSMV P6, an accessory protein, in transgenic wheat (Triticum aestivum L.) plants substantially increased host attractiveness to insect vectors through inhibiting the jasmonic acid (JA) signaling pathway. The BYSMV P6 protein interacted with the constitutive photomorphogenesis 9 signalosome subunit 5 (CSN5) of barley plants in vivo and in vitro, and negatively affected CSN5-mediated deRUBylation of cullin1 (CUL1). Consequently, the defective CUL1-based Skp1/Cullin1/F-box ubiquitin E3 ligases could not mediate degradation of jasmonate ZIM-domain proteins, resulting in compromised JA signaling and increased insect attraction. Overexpression of BYSMV P6 also inhibited JA signaling in transgenic Arabidopsis (Arabidopsis thaliana) plants to attract insects. Our results provide insight into how a plant cytorhabdovirus subverts plant JA signaling to attract insect vectors.

A plant rhabdovirus accessory protein subverts JA signaling of host plants for insect vector attractiveness and efficient virus transmission.

Introduction

Arthropod-borne viruses cause many diseases in human, animals, and plants (Reynolds et al., 2017; Gallet et al., 2018). Transmission of plant arboviruses relies on their insect vectors, which involves tritrophic interactions of virus–plant–insect. Intriguingly, arbovirus infections usually induce some host changes and increase host attractiveness to insect vectors (Dader et al., 2017). Plant arboviruses can modulate leaf colors and volatile organic compounds, which in turn increases insect attractiveness and virus acquisition efficiency (Fereres and Moreno, 2009; Mwando et al., 2018). In addition, viruses manipulate plant defense responses, such as jasmonic acid (JA) signaling, ethylene signaling, cellulose deposition, reactive oxygen burst, and phloem clogging, to increase insect vector attractiveness and fitness (Zhang et al., 2012; Li et al., 2014; Casteel et al., 2015; Wu et al., 2019; Wu and Ye, 2020). Although accumulating evidence documents virus-modified plant parameters for virus transmission, the underlying molecular mechanisms, especially for plant negative-stranded RNA (NSR) viruses, are just beginning to be revealed.

The family Rhabdoviridae is taxonomically classified in the order Mononegavirales, and most members contain an undivided negative-sense RNA genome. Rhabdoviruses have a broad range of hosts including vertebrates, invertebrates, and plants. Plant rhabdoviruses infect many agriculturally important monocot and dicot plants. In the field, plant rhabdoviruses are primarily transmitted by hemipteran insects including aphids, planthoppers, and leafhoppers (Mann and Dietzgen, 2014; Whitfield et al., 2018). However, little is known about the interactions of plant rhabdoviruses with their host plants and insect vectors due to lack of reverse genetic systems. Recently, a breakthrough has been achieved for reverse genetics analyses of plant NSR viruses, including rhabdoviruses and orthotospovirus (Wang et al., 2015; Gao et al., 2019; Feng et al., 2020; Verchot et al., 2020; Fang et al., 2022b), making it possible to investigate tritrophic interactions of virus–plant–insect (Jackson and Li, 2016; German et al., 2020; Zang et al., 2020; Li and Zhao, 2021).

Barley yellow striate mosaic virus (BYSMV), a member of the genus Cytorhabdovirus, infects cereal plants through transmission of the small brown planthopper (SBPH, Laodelphax striatellus) in a propagative manner (Cao et al., 2018). The BYSMV genome consists of 12,706 nucleotides and the antigenome contains 10 open reading frames (ORFs) in the order 3′-N-P-P3-P4/P5-P6-M-G-P9-L-5′, in which nucleoprotein (N), phosphoprotein (P), matrix protein (M), glycoprotein (G), and polymerase protein (L) are virus structural proteins. Additionally, P3, P4, P5, P6, and P9 are accessory proteins (Yan et al., 2015). We have recently established reverse genetic systems of BYSMV, including minireplicon and full-length infectious cDNA clones (Fang et al., 2019; Gao et al., 2019), allowing us to investigate interactions between BYSMV with its host plants and insect vectors (Qiao et al., 2022; Xu et al., 2022). Using these genetic systems, we have revealed that the BYSMV P protein can recruit host carbon catabolite repression 4 and casein kinase 1 (CK1) into viroplasm-like bodies for modulating virus replication (Gao et al., 2020; Zhang et al., 2020; Fang et al., 2022a). Recently, we found that host mitogen-activated protein kinases (MAPKs) inhibit BYSMV infections through directly phosphorylating the viral nucleoprotein (Ding et al., 2022). In contrast with the structural proteins, functions of rhabdovirus accessory proteins are largely unknown despite their potential importance in virus transmission, infection, and host immunity responses.

In plants, phytohormone jasmonic acid (JA) mediates plant defenses against herbivores and pathogens, especially necrotrophic pathogens (Zhang et al., 2017). Upon attacks of herbivores or pathogens, plant jasmonoyl-L-isoleucine (JA-Ile) is synthesized and binds to the coronatine-insensitive 1 (COI1) receptor, causing degradation of JA ZIM-domain protein (JAZ) transcriptional repressors and induction of plant defense-related genes (Thines et al., 2007; Song et al., 2014; Chen et al., 2019). JA-induced defense genes upregulate plant volatile biosynthesis and enhance plant defense against herbivores (Howe et al., 2018). Interestingly, several plant viruses have been reported to manipulate the JA-induced volatile biosynthesis to attract insect vectors (Li et al., 2014; Wu et al., 2019). Cucumber mosaic virus disrupts JA signaling to promote insect attraction (Lewsey et al., 2010; Mauck et al., 2015), through CMV 2b-dependent repression of JA-induced degradation of host JAZ proteins (Wu et al., 2017). Therefore, it is interesting to reveal how plant rhabdoviruses, as the biggest group of insect-transmitted plant NSR viruses, manipulate plant defense pathways for efficient virus transmission.

In this study, we found that BYSMV P6 increased host plant attractiveness to insect vectors in BYSMV-infected barley (Hordeum vulgare L.) plants and P6 transgenic plants. We further showed that BYSMV P6 physically interacted with HvCSN5, a catalytic subunit of the constitutive photomorphogenesis 9 (COP9) signalosome (CSN) complex (Wei et al., 2012, 2018). We further showed that the P6 protein inhibited the CSN5-dependent removal of the RUB moiety from Skp1/Cullin1/F-box (SCF) complex, the most abundant E3 ligases (Feng et al., 2003). Finally, the SCF^COI1 complex-dependent degradation of JAZs was inhibited (Thines et al., 2007), which subverts the JA signaling pathway to reduce host defense responses against insect vectors. Our results uncover the mechanism underlying how a rhabdovirus accessory protein modulates plant JA signaling to increase insect attractiveness and benefit for virus transmission.

Results

BYSMV P6 negatively regulates trichome development and JA accumulation

Rhabdoviruses encode multiple accessory proteins with unknown functions. Therefore, investigation of these accessory proteins will increase our understanding of plant rhabdovirus–plant–insect interactions. To investigate the effect of BYSMV P6 on host plant development, we generated two stable transgenic wheat (Triticum aestivum) lines (P6^OX12 and P6^OX17) and confirmed overexpression of the P6 protein through PCR and immunoblotting (Supplemental Figure S1). P6^OX12 and P6^OX17 exhibited considerably shorter height and smaller roots compared with nontransgenic (NT) wheat plants (Supplemental Figure S1). We first examined response of NT, P6^OX12, and P6^OX17 to JA signaling. In agreement with previous studies (Traw and Bergelson, 2003; Yoshida et al., 2009), trichome numbers on the edges of NT wheat leaves were significantly increased at 7 days after treatment of 200-μM MeJA compared with those of mock-treated plants (Figure 1, A and B). In contrast, MeJA treatment did not increase trichome numbers in P6^OX12 and P6^OX17 wheat leaves (Figure 1, A and B). Using a scanning electron microscope, we further found that leaf trichome numbers of P6^OX12 and P6^OX17 were substantially less than those of NT wheat plants after MeJA treatment (Figure 1, C and D). To examine whether BYSMV infection affected trichome development, we inoculated barley plants with BYSMV-RFP, a recombinant BYSMV virus expressing the reporter gene RFP as described previously (Gao et al., 2019). At 7 days after treatment of 200-μM MeJA, trichome numbers on the edges of newly emerging leaves of BYSMV-RFP-infected barley plants were significantly decreased compared with those of MeJA-treated healthy plants (Figure 1 E and F). Collectively, these results indicate that both the transgenic P6 protein and BYSMV infection negatively affect JA-involved trichome development (Traw and Bergelson, 2003; Hua et al., 2021).

Figure 1.

BYSMV P6 negatively regulates trichome development and JA concentration. A, The trichome phenotype on edges of NT and two transgenic wheat plants, P6^OX12 and P6^OX17. The newly emerging leaves of wheat plants were used for trichome observation under stereomicroscope at 7 days post treatment of mock buffer or 200-μM MeJA. Scale bar = 500 μm. B, Quantification of trichome numbers per 1 mm on the edges of leaves as shown in (A). C, Scanning electron micrographs of leaf epidermis of NT, P6^OX12, and P6^OX17 wheat plants treated with 200-μM MeJA. Scale bar = 100 μm. D, Quantification of trichome numbers on leaf epidermis in a view (704 μm × 528 μm) as shown in (C). E, The trichome phenotype on edges of barley leaves treated with mock buffer and BYSMV-RFP infection. The mock or BYSMV-RFP-infected plants were treated with 200-μM MeJA at 14 dpi and the newly emerging leaves were used for trichome observation 7 days later. RFP fluorescence indicates BYSMV-RFP infection. Scale bar = 500 μm. F, Quantification of trichome numbers per 1 mm on the edges of leaves as shown in (E). G, Specificity of immunological detection of JA in leaves of NT, P6^OX12, and P6^OX17 wheat plants. Scale bar = 50 μm. H, Specificity of immunological detection of JA in leaves of barley leaves infected by mock buffer and BYSMV-RFP at 14 dpi. Scale bar = 50 μm. In (G) and (H), leaves of wheat plants were wounded with forceps at 2 h before incubated with 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride for immunofluorescence labeling using the anti-JA antibodies as described in the “Materials and methods.” In (B), (D), and (F), Error bars indicate standard error of mean (sem). Letters indicate significant differences determined by one-way ANOVA followed by Tukey’s multiple comparison tests (P < 0.05). At least 13 plants for each experiment were scored.

We next examined whether BYSMV P6 affected JA accumulation of host plants. Immunocytochemical detection of JA concentration in tissues could be visualized by green fluorescence that results from anti-JA rabbit antibodies and the secondary antibody conjugated with the fluorescent dye AlexaFluor488 (Mielke et al., 2011; Wojciechowska et al., 2018). Leaves of NT, P6^OX12, and P6^OX17 wheat plants were wounded with forceps at 2 h before immunolabeling. As a negative control, the NT wheat leaf did not exhibit green fluorescence signal without addition of anti-JA antibodies, indicating that green fluorescence specifically indicated JA signal (Figure 1G). As expected, green fluorescence signal in NT wheat leaves showed dramatically higher intensity than in P6^OX12 and P6^OX17 wheat leaves (Figure 1G; Supplemental Figure S2). Moreover, BYSMV-RFP infection also reduced the green fluorescence intensity in barley leaves (Figure 1H; Supplemental Figure S2). Collectively, these results demonstrate that BYSMV infection and P6 overexpression efficiently suppress JA accumulation and trichome development.

BYSMV P6 inhibits JA signaling and enhances attractiveness to insect

We next examined whether BYSMV P6 affected the JA signaling pathway. The ethylene response factor (ERF) and the basic helix–loop–helix leucine zipper MYC2 transcription factors are strongly induced by JA in Arabidopsis (Arabidopsis thaliana) (Lorenzo et al., 2003, 2004). Consistently, the barley putative orthologs of ERF (HvERF) and MYC2 (HvMYC2) were strongly induced by MeJA treatment in healthy plants (Figure 2A). In contrast, MeJA-mediated induction of HvERF and HvMYC2 was strikingly suppressed in BYSMV-infected plants (Figure 2B). Induction of two JA-responsive genes lipase and lipoxygenase 2 (Ding et al., 2016) was abolished in P6^OX12 and P6^OX17 plants compared to NT plants (Figure 2C), indicating that BYSMV P6 inhibits JA signaling in plants.

Figure 2.

BYSMV infection and overexpression of BYSMV P6 inhibit JA signaling and increases attractiveness to SPBHs. A, Relative accumulation levels of HvERF and HvMYC2 mRNAs in barley seedlings treated with mock buffer or 200-μm MeJA for 4 h. B, Relative accumulation levels of HvERF and HvMYC2 mRNAs in mock or BYSMV-RFP-infected barley seedlings at 14 dpi. Both mock and BYSMV-RFP-infected barley seedlings were treated with 200-μm MeJA, followed by collection for RT-qPCR assays 4 h later. C, Relative accumulation levels of TaLipase and TaLOX2 mRNAs in NT wheat (NT, cultivar Fielder) and two BYSMV P6-Flag transgenic lines (P6^OX12 and P6^OX17) with 200-μm MeJA treatments for 4 h. D, Schematic diagram for field-like feeding preference bioassays. Plants were placed in an insect proof net and SBPHs were released from the center of the net for insect choice for 8 h. SBPH numbers on each plant were counted and analyzed. E, Percentages recaptured SBPHs out of 50 in total released on mock or BYSMV-RFP-infected barley seedlings as shown in (B). F, Percentages recaptured SBPHs out of 50 in total released on NT wheat and P6^OX12 or P6^OX17. G, Schematic diagram for Y-tube bioassay (detailed information in methods). H, Y-tube bioassays showing SBPHs percentages attracted to mock or BYSMV-RFP-infected barley seedlings. I, Y-tube bioassays showing SBPHs percentages attracted to NT wheat and P6^OX12 or P6^OX17. Error bars indicate ±sem from three biological replicates. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, Student’s t test.

Since JA signaling is implicated in insect resistance, we therefore examined effects of BYSMV infection and P6 overexpression on host attractiveness to SBPHs (L.striatellus). In the two-choice assay, healthy or infected barley were placed into an insect-free net in the dark for eliminating light interference. Then, 50 nonviruliferous SBPHs were released to the net center (Figure 2D). After 8 h, the numbers of SBPHs on healthy or infected barley plants were counted and analyzed. Interestingly, BYSMV-infected plants attracted ∼76.7% of SBPHs, which was significantly more than 23.3% of SBPHs on mock plants (P < 0.01) (Figure 2E). Likewise, two P6 overexpressing wheat plants P6^OX12 and P6^OX17 attracted obviously increased percentages (70.7%, P < 0.01 and 63.4%, P < 0.001) of SBPHs compared with NT (29.3% and 36.6%) wheat plants (Figure 2F).

We further used a Y-tube olfactometer bioassay to validate the insect attraction results. Volatiles emitted from BYSMV-infected barley plants or mock inoculated plants flowed through two Y-tube arms. Then, individual SBPH was released from the Y-tube base to allow individual insect choice (Figure 2G). As expected, BYSMV-infected plants attracted significantly more SBPHs (65.6%, P < 0.05) than mock-treated plants (34.4%) (Figure 2H). Transgenic plants overexpressing BYSMV P6 were more attractive to SBPHs (70.6%, P < 0.01 and 63.3%, P < 0.05) than NT wheat plants (29.4% and 36.7%) (Figure 2I). Collectively, these results demonstrate that BYSMV P6 increases plant attractiveness to insect vectors by inhibiting JA signaling.

BYSMV P6 interacts with the barley CSN5 protein in vivo and in vitro

To reveal the underlying molecular mechanisms of BYSMV P6 in JA signaling, we performed yeast two-hybrid (Y2H) screen assays to screen the P6 interacting proteins from an Arabidopsis cDNA yeast library. Surprisingly, the Arabidopsis CSN5A protein (AtCSN5A) was identified as a candidate in the P6 interacting proteins. Since Arabidopsis is not a natural host plant of BYSMV, we cloned the cDNA of the barley CSN5 protein (HvCSN5) that shares amino acid sequence identity of 77.0% with the AtCSN5A (Supplemental Figure S3).

To verify the interaction of BYSMV P6 and HvCSN5, we first performed bimolecular fluorescence complementation (BiFC) assays in Nicotianabenthamiana leaves. To this end, the BYSMV P6 and HvCSN5 ORFs were fused to the C (Y^C) and N (Y^N) halves of sYFP, respectively. P6-Y^C and HvCSN5-Y^N were co-expressed in N. benthamiana leaves, resulting in reconstitution of YFP fluorescence signal throughout the cytoplasm and nucleus at 3-day post infiltration (dpi) (Figure 3A; Supplemental Figure S4). To determine the key amino acids responsible for the P6–HvCSN5 interaction, a series of deletion mutants of P6 were fused to Y^N, and further co-expressed with HvCSN5–Y^C in N. benthamiana leaves (Supplemental Figure S5). The BiFC assays indicated that the ¹⁶ITITS²⁰ domain of P6 was required for the P6–HvCSN5 interaction (Supplemental Figure S5). Subsequently, amino acids of ¹⁶ITITS²⁰ were individually mutated to alanine, and the resulting point mutants were fused to Y^N, and co-expressed with HvCSN5–Y^C. We could not detect YFP fluorescent signal in P6^I16A-Y^C and HvCSN5–Y^N co-expressing leaves, suggesting that the Ile16 was essential for the P6–HvCSN5 interaction (Figure 3A; Supplemental Figure S6). All negative controls including the N. benthamiana CK1 (NbCK1) and the Rubisco protein (RbcL) did not show YFP fluorescence although all the proteins were expressed in immunoblotting analyses (Supplemental Figures S4–S6).

Figure 3.

The barley HvCSN5 protein interacts with the BYSMV P6 protein in vitro and in vivo. A, BiFC analysis of interactions between HvCSN5 with P6 or P6^I16A protein in epidermal cells of N. benthamiana leaves. The NbCK1 protein served as a negative control. Scale bar = 20 μm. B, Co-IP analysis examining protein interactions between HvCSN5 and P6 or P6^I16A in vivo. At 3 dpi, total proteins were extracted from N. benthamiana leaves expressing indicated proteins were precipitated with anti-Flag beads and analyzed by immunoblotting with anti-GFP (a-GFP) and anti-Flag (a-Flag) antibodies. C, Y2H assays for determining protein interactions between HvCSN5 with P6 or P6^I16A protein. AD, GAL4 activation domain; BD, GAL4 DNA-binding domain. EV of pGADT7 and pGBKT7 served as negative controls. D, GST pull-down assays showing HvCSN5–P6 interactions in vitro. HvCSN5-His was incubated with GST-P6, GST-P6^I16A or GST and immunoprecipitated with glutathione-Sepharose beads. The pull-down and input proteins were detected by immunoblotting with anti-GST (α-GST) and anti-His (α-His) antibodies.

We further examined the P6–HvCSN5 interaction using co-immunoprecipitation (Co-IP) assays. To this end, HvCSN5-Flag was co-expressed with GFP, P6-GFP, or P6^I16A-GFP in N. benthamiana leaves by agroinfiltration. At 3 dpi, total proteins from agro-infiltrated leaves were precipitated with anti-Flag beads and analyzed with immunoblotting assays. The results showed that HvCSN5-Flag was efficiently co-immunoprecipitated with P6-GFP, but not GFP or P6^I16A-GFP (Figure 3B). Y2H assays were carried out to validate the P6–HvCSN5 interaction. In these experiments, BYSMV P6 and P6^I16A were fused with the GAL4 activation domain, and HvCSN5 was fused to GAL4 DNA-binding domain. Yeasts harboring the combination of AD-P6 and BK–HvCSN5 were able to proliferate on SD/Trp–Leu–His–Ade selection plates with 5-mM 3-amino-1, 2, 4-triazole (3-AT) (Figure 3C). In contrast, combinations of AD-P6^I16A/BK–HvCSN5 and other negative controls could not grow on SD/Trp–Leu–His–Ade selection plates with 3-AT (Figure 3C).

The direct interaction between BYSMV P6 and HvCSN5 was examined by in vitro GST pull-down assays. In this experiment, HvCSN5-His and GFP/P6/P6^I16A fused with a GST tag (GST-GFP/GST-P6/GST-P6^I16A) were purified from Escherichia coli for GST pull-down assays. The results revealed that HvCSN5-His could interact with GST-P6, but not with GST-GFP or GST-P6^I16A in vitro (Figure 3D). Taken together, these results demonstrate that BYSMV P6 interacts with HvCSN5 in vivo and in vitro and the P6 Ile16 is essential for their interaction.

BYSMV P6 inhibits deRUBylation of CUL1 and degradation of JAZ proteins

CSN5 is an important catalytic subunit of CSN complex and responsible for removing the ubiquitin-like protein RUB1 from the cullin subunit (CUL1) of the Cullin-RING ligase family of E3 complexes, which is essential for activation of the E3 ligase functions in vivo (Wei et al., 2012; Jin et al., 2014; Wei et al., 2018). We hypothesized that the P6–HvCSN5 interaction probably interfered with CSN5-mediated CUL1 deRUBylation. To test this hypothesis, we first cloned the ORFs of CUL1 putative orthologs of barley (HvCUL1) and wheat (TaCUL1) plants, which shared high similarity with the Arabidopsis CUL1 in amino acid sequence (Supplemental Figure S7). To examine RUBylation/deRUBylation of HvCUL1, we transiently expressed the 3 × Flag-fused CUL1 putative ortholog of barley plants (HvCUL1-Flag) with empty vector (EV), HA-tagged P6 (HA-P6), or HA-P6^I16A in N. benthamiana leaves. At 3 dpi, total proteins were immunoprecipitated with anti-Flag beads and subjected to immunoblotting analyses. The results showed that the RUBylated HvCUL1 accumulated to a dramatically higher level in HA-P6 samples compared with that of EV or HA-P6^I16A (Figure 4A; Supplemental Figure S8A). In agreement with HvCUL1, the RUBylation form of TaCUL1 was also increased by HA-P6, but not by EV or HA-P6^I16A (Figure 4B; Supplemental Figure S8B).

Figure 4.

BYSMV P6 inhibits CSN5-mediated deRUBylation of CUL1 and JAZ protein degradation. A and B, Immunoblotting analyses detecting accumulation of the barley CUL1 (HvCUL1, A) and the wheat CUL1 (TaCUL1, B) transiently co-expressed with EV, HA-P6, or HA-P6^I16A in N. benthamiana leaves. HvCUL1-3Flag and TaCUL1-3Flag were precipitated with anti-Flag beads and examined by immunoblotting analyses with anti-Flag antibodies. Positions of RUBylated (CUL1^RUB) and deRUBylated cullin1 (CUL1) proteins were indicated. Total proteins were used for detection of HA-P6, or HA-P6^I16A with anti-HA antibodies. RbcL served as a loading control. C, Immunoblotting analyses detecting accumulation of HvCUL1 in barley leaves infected by BYSMV (BY) or mock buffer (M) with anti-CUL1 antibodies. The BYSMV P6 protein was detected with anti-P6 antibodies. D, Immunoblotting analyses detecting accumulation of TaCUL1 in NT and P6 transgenic wheats (P6^OX12 and P6^OX17) with anti-CUL1 antibodies. E, Confocal images showing fluorescence of HvJAZ-GFP, NbJAZ3-GFP, and GFP in N. benthamiana leaves with treatments of mock, co-expression with EV, HA-P6, or HA-P6^I16A, MG132, and MeJA as indicated. At 72-h post agroinfiltration, infiltrated leaves were sprayed with 200-μM MeJA or mock buffer for 2 h, and examined under a confocal microscope. The MG132 treatment was carried out at 12 h before observation. Scale bar = 50 μm. F, Immunoblotting analyses detecting accumulation of HvJAZ3-GFP, TaJAZ12-GFP, and GFP in the leaves as shown in (E) with anti-GFP antibodies. HA-P6 and HA-P6^I16A were detected with anti-HA antibodies.

We further examined the RUBylation levels of endogenous HvCUL1 in BYSMV-infected or mock-treated barley plants using antibodies against AtCUL1. The results showed that BYSMV infection induced higher levels of RUBylated-HvCUL1 compared to mock treatment (Figure 4C). Overexpression of BYSMV P6 alone also resulted in increased levels of RUBylated HvCUL1 in P6^OX12 and P6^OX17 transgenic plants compared to NT plants (Figure 4D; Supplemental Figure S8C). Thus, these results demonstrate that BYSMV infection and P6 overexpression interfere with CUL1 deRUBylation.

JA induction leads to ubiquitination and proteasome degradation of JAZ repressors by the SCF^COI1 ubiquitin E3 ligases (Xie et al., 1998; Chini et al., 2007; Thines et al., 2007). Therefore, we assumed that the P6-mediated inhibition of CUL1 deRUBylation might affect degradation of JAZ proteins. To test the assumption, we cloned HvJAZ3 and TaJAZ12 from barley and wheat plants, respectively (Supplemental Figure S9), and fused them to the N-terminus of GFP (HvJAZ3-GFP and TaJAZ12-GFP). HvJAZ3-GFP and TaJAZ12-GFP were transiently expressed in N. benthamiana leaves by agroinfiltration. Upon treatment of 200-μM MeJA at 70-h post infiltration (hpi), GFP fluorescent signal of HvJAZ3-GFP and TaJAZ12-GFP (the EV column) showed lower intensity compared with mock treatment (Figure 4E). However, pretreatment of proteasome inhibitor MG132 at 60 hpi rescued high GFP signal intensity (Figure 4E), indicating MeJA-mediated turnover of HvJAZ3-GFP and TaJAZ12-GFP was dependent on proteasome degradation. Interestingly, co-expression of HA-P6, rather than HA-P6^I16A, also recovered high GFP signal intensity from MeJA-mediated turnover of HvJAZ3-GFP and TaJAZ12-GFP (Figure 4E). Immunoblotting analyses were performed to verify the GFP fluorescence results, showing that accumulation of HvJAZ3-GFP and TaJAZ12-GFP was suppressed by MeJA treatment, which was in turn inhibited by MG132 pretreatment or HA-P6 co-expression (Figure 4F). As negative controls, fluorescence intensity and protein accumulation of free GFP were not affected by MeJA, MG132, or HA-P6 (Figure 4F).

Taken together, these results demonstrate that BYSMV infection and P6 overexpression interfere with CSN5-mediated CUL1 deRUBylation and inhibit proteasome-mediated degradation of JAZ proteins.

BYSMV P6 inhibits deRUBylation of CUL1 and JAZ degradation in Arabidopsis

Given that the JA signaling is a highly conserved pathway in land plants including monocot and dicot plants, we next examined whether BYSMV P6 interfered the JA signaling through interacting with the AtCSN5A. Using BiFC assays, we found that co-expression of AtCSN5A-Y^N and P6-Y^C led to production of GFP fluorescence, while AtCSN5A-Y^N/P6^I16A-Y^C or AtCSN5A-Y^N/RbcL-Y^C could not (Figures 5A). Protein expression of the BiFC assays was verified by immunoblot analyses (Supplemental Figure S10). In vitro pull-down assays revealed that the AtCSN5A-His protein was only co-immunoprecipitated with GST-P6, rather than with GST or GST- P6^I16A (Supplemental Figure S10). Therefore, these results demonstrate that BYSMV P6 directly interacts with different CSN5 proteins from both monocot and dicot plants.

Figure 5.

BYSMV P6 inhibits deRUBylation of CUL1 and JAZ degradation in Arabidopsis. A, BiFC analysis of interactions between the AtCSN5A with P6 or P6^I16A in epidermal cells of N. benthamiana leaves expressing proteins as indicated. RbcL served as a negative control. Scale bar = 20 μm. B, Root length of 7-day-old Col-0, AtP6, and AtP6^I16A overexpressing Arabidopsis lines grown on MS plates containing mock buffer or 50-μM MeJA. Scale bar = 1 cm. C, Statistical analysis of root length ratio of MeJA/Mock treatments (n > 20). D, Immunoblotting analyses detecting accumulation of endogenous Arabidopsis CUL1 in Col-0, AtP6, and AtP6^I16A with anti-CUL1 antibodies. The positions of RUBylated (CUL1^RUB) and deRUBylated CUL1 (CUL1) were indicated. E, In vitro degradation assays of purified MBP-AtJAZ1 incubated with crude extract from AtP6^I16A/OX1 and AtP6^OX5. F, Degradation assays of purified MBP-AtJAZ1 incubated with crude extract from Col-0 plants with 10-µM MG132 or Mock buffer. G, Degradation assays of purified MBP-AtJAZ1 with GST-P6^I16A or GST-P6 in crude extract from Col-0 plants. H, Degradation assays of purified MBP-AtJAZ1 with GST-P6^I16A or GST-P6 in crude extract from coi1-1 plants. RbcL served as a loading control. In (C), error bars show mean ± sem, and different letters indicate significant differences determined by one-way ANOVA followed by Tukey’s multiple comparison tests (P < 0.05).

We further generated transgenic Arabidopsis plants stably expressing BYSMV P6 or P6^I16A to examine how P6 regulated the JA signaling in Arabidopsis (Supplemental Figure S11). Root growth of 7-day-old Columbia-0 (Col-0) plants was severely inhibited by MeJA treatment (50 μM) compared with mock buffer (Figure 5, B and C). In contrast, two P6 overexpressing Arabidopsis lines, AtP6^OX5 and AtP6^OX8, exhibited shorter roots than Col-0 in mock treatment, but were insensitive to MeJA treatment (Figure 5, B and C). However, two P6^I16A transgenic lines, AtP6^I16A/OX1 and AtP6^I16A/OX2, exhibited relative long roots but were significantly suppressed by MeJA (Figure 5, B and C). These results indicate that BYSMV P6, but not P6^I16A, inhibits JA signaling in Arabidopsis.

We further examined RUBylation of endogenous AtCUL1 in Col-0, AtP6^OX, and AtP6^I16A plants. The results showed that BYSMV infection induced higher levels of RUBylated AtCUL1 (CUL1^RUB) compared to Col-0 and AtP6^I16A plants (Figure 5D; Supplemental Figure S12).

We next examined whether P6 inhibited JA-induced degradation of JAZ in vitro. To this end, GST-P6, GST-P6^I16A, and MBP-AtJAZ1 were expressed and purified from E. coli. First, MBP-AtJAZ1 was incubated with total crude protein extract from AtP6^OX5 or AtP6^I16A/OX1 plants at room temperature for 0, 10, 20, and 30 min. Immunoblotting assays showed that MBP-AtJAZ1 degraded at a slower speed in extract of AtP6^OX5 than that of AtP6^I16A/OX1, indicating that P6, rather than P6^I16A, inhibited in vitro degradation of MBP-AtJAZ1 (Figure 5E). Using the similar strategy, the Col-0 extract with MG132 inhibits proteasome degradation of MBP-AtJAZ1 (Figure 5F). Furthermore, the total crude protein extract from Col-0 plants was incubated with purified GST-P6 or GST-P6^I16A at 4°C for 1 h, and then MBP-AtJAZ1 protein was added at 25°C for 0, 10, 20, and 30 min. Immunoblotting assays showed that the MBP-AtJAZ1 protein was degraded at a higher speed in GST-P6^I16A samples than in GST-P6 samples (Figure 5G). In contrast, MBP-AtJAZ1 was not degraded in the extract from coi1-1 mutant plants with GST-P6 or GST-P6^I16A (Figure 5H), indicating degradation of MBP-AtJAZ1 was dependent on JA signaling.

Collectively, proteasome degradation of MBP-AtJAZ1 is dependent on JA signaling, which is inhibited by transgenic P6 and purified GST-P6, but not P6^I16A.

BYSMV P6 inhibits JA signaling and enhances insect attractiveness in Arabidopsis

We next examined whether BYSMV P6 affected expression of JA-responsive genes in Arabidopsis. As expected, overexpression of P6 significantly suppressed induction of JA-responsive genes (Wasternack and Song, 2017), including plant defensin 1.2 (PDF1.2), vegetative storage protein 2 (VSP2), and tyrosine aminotransferases 1 (TAT1) in two transgenic plants (AtP6^OX5 and AtP6^OX8) compared with wild-type Col-0 plants (Figure 6A). In contrast, induction of PDF1.2, VSP2, and TAT1 was not affected in two P6^I16A transgenic lines (AtP6^I16A/OX1 and AtP6^I16A/OX2) (Figure 6B).

Figure 6.

BYSMV P6 inhibits JA signaling and increases attractiveness in Arabidopsis. A, RT-qPCR assays examined relative accumulation of AtPDF1.2, AtVSP2, and AtTAT1 in Col-0 and two P6 overexpressing Arabidopsis lines (AtP6^OX5 and AtP6^OX8). Plants were treated with 200-μm MeJA for 4 h. B, Relative accumulation of AtPDF1.2, AtVSP2, and AtTAT1 in Col-0 and two P6^I16A overexpressing Arabidopsis lines (AtP6^I16A/OX1 and AtP6^I16A/OX2). Plants were treated with 200-μm MeJA for 4 h. C, Schematic diagram for circular-dish assay bioassays. Different leaves were placed around a circular dish alternately. In each experiment, more than 40 aviruliferous aphids were released from the dish center. After 4 h, numbers of aphids choosing plants were counted and the ratios for the paired plants were calculated. D, Aphid percentages recaptured out of 40 in total released between Col-0 and AtP6^OX5/AtP6^OX8 plants. E, Aphid percentages recaptured out of 50 in total released between Col-0 and AtP6^I16A/OX1/AtP6^I16A/OX2 plants. F, Schematic diagram for Y-tube bioassay (detail information in methods). G, Y-tube bioassays showing aphid percentages attracted to Col-0 or AtP6^OX5/AtP6^OX8 plants. H, Y-tube bioassays showing aphid percentages attracted to Col-0 with AtP6^I16A/OX1 or AtP6^I16A/OX2 plants. In (A, B, D, E, G, and H), error bars showing mean ± sem was from three biological replicates. ^*P < 0.05, ^**P < 0.01, ^***P < 0.001, no significance (ns) P > 0.05, Student’s t test.

We further performed feeding preference assays using P6 and P6^I16A transgenic lines. In the circular-dish bioassay, leaves from Col-0 or P6 transgenic plants were arranged next to each other in a circle (Figure 6C). In the center of the dishes, 40 apterous aphids (Myzus persicae) were set free and recorded based on their crawling traces (Figure 6C). As expected, leaves of AtP6^OX5 and AtP6^OX8 attracted significantly higher percentages of aphids compared with those of Col-0 plants (Figure 6D). In contrast, comparable percentages of aphids approached P6^I16A overexpressing leaves and Col-0 leaves (Figure 6E). The Y-tube olfactometer bioassays (Figure 6F) consistently showed that P6 overexpression lines attracted >60% of total aphids compared to <40% in Col-0 plants (Figure 6G). However, the P6^I16A overexpressing leaves did not exhibit increased aphid attractiveness (Figure 6H). Collectively, these results demonstrate that BYSMV P6 overexpression enhances aphid attractiveness through inhibiting JA signaling in Arabidopsis.

In summary, BYSMV P6 interacts with the conserved CSN5 protein and inhibited removal of the RUB moiety from CUL1 in both monocot and dicot plants. Consequently, the suppressed CUL1 was defective in proteasome degradation of JAZ proteins, thereby inhibiting expression of JA-responsive genes and defense against insects. Therefore, P6 transgenic plants and BYSMV infection exhibit enhanced insect attractiveness, which is beneficial for virus acquisition by insect vector (Figure 7).

Figure 7.

Model for BYSMV P6 inhibition of JA signaling in plants to attract insect vectors. In healthy plants, upon insect feeding or wounding, CSN5 remove the RUB1 from CUL1, resulting in proteasome degradation of JAZ proteins and upregulation of JA responsive genes to repel insect vectors. In BYSMV-infected plants, however, BYSMV P6 interacts with host CSN5 and compromises deRUBylation of CUL1, leading to suppression of JAZ protein degradation and inhibited JA signaling. Thus, BYSMV-infected plants are compromised in JA-mediated resistance to insect vectors, which in turn enhances attractiveness to insect vectors for virus acquisition.

Discussion

To defend against diverse biotic and abiotic stresses, sessile plants have evolved highly regulated and sophisticated systems to induce massive defense-related genes and secondary metabolites. The lipid-derived JAs, including JA and its derivatives, mediate broad-spectrum defense against various herbivorous insects in plants (Wasternack and Song, 2017; Zhang et al., 2017). Increasing evidence has demonstrated that JA signaling is implicated in the tripartite interactions among plant viruses, host plants, and insect vectors (Zhang et al., 2017; Wang et al., 2019; Wu and Ye, 2020). For instance, the CMV 2b protein directly interacts with and inhibits degradation of JAZ proteins (Wu et al., 2017). The nonstructural protein (NSs) of tomato spotted wilt orthotospovirus and the βC1 of tomato yellow leaf curl China virus interacts with MYC2 to inhibit JA-mediated activation of resistance genes (Li et al., 2014, Wu, et al, 2019). In addition, the C2 protein of tomato yellow leaf curl virus interacts with plant ubiquitin and then compromises JAZ1 degradation to suppress plant defense against insects (Li et al., 2019). These results suggest that plant viruses evolved different effectors to target distinct processes of the JA signal pathway to increase host plant attractiveness to insect vectors.(NSs)

Plant rhabdoviruses are heavily dependent on transmission by arthropod vectors in a circulative-propagative manner (Ammar el et al., 2009). However, little is known about the underlying molecular mechanisms for the tripartite interactions among plant rhabdovirus viruses, host plants, and insect vectors. In this study, we demonstrated that a cytorhabdovirus accessory protein enhances insect attractiveness by inhibiting JA signaling. We reveal that the BYSMV P6 physically interacts with plant CSN5 proteins and prevents deRUBylation of SCF-type E3 ubiquitin ligases, thereby inhibiting JAZ degradation and induction of JA-responsive genes. In addition to inhibition of CSN5, P6 may alter CUL1 assembly into the SCF complex, which remains to be investigated in our future studies. Collectively, our results provide evidence to show that the upstream CSN5 protein in the JA signal pathway is a target of the P6 protein of a plant rhabdovirus for virus transmission.

Strikingly, the CSN complex is a common target of animal and plant viruses. In animal viruses, the interactions between virus proteins and CSN components mainly mediate a redirection of proteasome degradation (Oh et al., 2006; Tanaka et al., 2006; Hsieh et al., 2007). The C2 protein of geminiviruses interacts with CSN5 and affects deRUBylating activity of the CSN complex, resulting in inhibition of JA signaling (Lozano-Duran et al., 2011). Recently, the P5-1 protein of rice black-streaked dwarf virus, a double-stranded RNA virus, interacts with the CSN5A protein of rice (Oryza sativa) and inhibits the ubiquitination activity of SCF E3 ligases, which suppresses the JA response pathway and facilitates virus infection (He et al., 2020). Nonetheless, these studies focused on effects of the virus effector–CSN5 interactions on the arm races between plants and viruses, but not on the insect attractiveness. In contrast, we define the P6 protein from an NSR virus as a virus effector to enhance the attractiveness of their vectors by downregulating JA signaling.

Another interesting finding in this study is that BYSMV P6 increased transgenic Arabidopsis attractiveness to aphids. Although BYSMV only infect monocot plants naturally, BYSMV P6 interacts with the CSN5 proteins of both monocot and dicot plants. These results demonstrate that the CSN complexes are strongly conserved and common targets of plant viruses in either monocot or dicot plants. Moreover, BYSMV P6 mainly inhibits removal of RUB/NEDD8 from CUL1, and accumulation of CUL1-RUB inhibits the activity of SCF-type E3 ubiquitin ligases, which would further affect several hormone signaling pathways (Schwechheimer et al., 2001; Feng et al., 2003; Gusmaroli et al., 2004; Moon et al., 2007). Besides, CSN5A or CSN5B also contribute to deRUBylation of CUL3 and CUL4 in Arabidopsis (Gusmaroli et al., 2007), which are involved in plant development and environmental response (Fonseca and Rubio, 2019; Ban and Estelle, 2021). Accordingly, it remains to be explored whether the P6–CSN5 interaction affects deRUBylation of CUL3 and CUL4. Furthermore, development and stress responses related with CUL1, CUL3, and CUL4 will be investigated in future studies of the tripartite interactions among virus, plants, and insect vectors.

Rhabdoviruses contain many economically important RNA viruses infecting a wide range of hosts. A multitude of accessory genes encode potentially informative proteins but most of them are poorly understood (Walker et al., 2011). Therefore, exploring the functions of these accessory proteins will benefit understanding of rhabdovirus transmission, infection, and immunity response. Plant rhabdoviruses usually encode viral movement proteins between the P and M genes (Huang et al., 2005; Jackson et al., 2005; Min et al., 2010; Zhou et al., 2019). Interestingly, the BYSMV antigenome contains four small genes (P3, P4, P5, and P6) at this position (Yan et al., 2015). Here, we found a function of rhabdovirus accessory proteins, BYSMV P6, in increasing plant attractiveness to insects through manipulating deRUBylation of SCF-type E3 ubiquitin ligases and inhibit JA signaling.

In summary, we define BYSMV P6 as a virus effector to downregulate JA signaling and induce plant attractiveness to insects. When JA signal is induced in healthy plants, JA-Ile binds to COI1-JAZ and then triggers degradation of JAZ transcriptional repressors, which requires the activation of the SCF complex via removal of RUB1 from CUL1 by CSN5. The BYSMV P6, expressing from either virus infection or transgenic plants, interacts with CSN5 and inhibits deRUBylation of CUL1, which interferes with the degradation of JAZs and induction of JA signaling, and leads to increasing insect attractiveness (Figure 7). Our results will benefit understanding of tripartite interaction among plant viruses, host plants, and insect vectors, and provide a potential target for controlling virus transmission in the future. Finally, our results expand fascinating evidence to show that virus infections can regulate plant resistance responses and vector behavior to increase virus transmission.

Materials and methods

Plant materials

Nicotiana benthamiana, barley (H.vulgare cv Golden promise), and wheat (T.aestivum cv Fielder) plants were grown at 25 ± 2°C and a photoperiod of 14 h:10 h, light:dark. The Arabidopsis (Arabidopsis thaliana) plants are in the background of Col-0 and were grown under 14 h:10 h (light:dark) photoperiod conditions at 23°C in growth chambers. The coi1-1 mutant has been described as previously (Xie et al., 1998). For BYSMV P6 transgenic wheat plants, the ORF of P6-Flag was amplified and cloned into the pMWB110 vector, transferred into the EHA105 strain and transformed into wheat plants (cultivar Fielder). To obtain P6 or P6^I16A transgenic Arabidopsis lines, the ORFs of P6 or P6^I16A were inserted into the pMDC32-3 × Flag vector, and transformed into Col-0 plants as described previously (Tong et al., 2021). All positive transgenic plants were verified by PCR and immunoblotting analyses. All primers were listed in Supplemental Table S1.

Virus inoculation and insect rearing

The wild-type BYSMV was maintained in barley plants through transmission by the SBPHs (L.striatellus) as described previously (Yan et al., 2015; Zhang et al., 2020). The recombinant BYSMV-RFP virus with an RFP reporter gene was obtained from a BYSMV infectious clone as described previously (Gao et al., 2019). SBPHs (L. striatellus) were reared on rice (O.sativa) seedlings at 26°C with 12 h:12 h (light:dark) photoperiod in chambers. Wingless aphids (M.persicae) were maintained on Raphanus sativus under 14 h:10 h (light:dark) photoperiod conditions at 23°C in growth chambers.

Leaf trichome analyses

To analyze density of leaf trichome, 2-week-old transgenic wheat leaves were treated with 200-μM MeJA. After 7 days, newly emerging leaves were used for trichome observation under scanning electron micrographs. In addition, trichomes on edges of transgenic wheat leaves were analyzed under stereomicroscope. At 14 dpi, the BYSMV-RFP-infected or mock barley plants were treated with 200-μM MeJA and then were examined for trichome density at 7 days after MeJA treatment. For each treatment, about eight images were taken for trichome analyses.

Detection of relative JA content by immunolocalization

The relative JA content was detected by immunofluorescence as described previously (Mielke et al., 2011; Wojciechowska et al., 2018) with modification. In these experiments, leaves of 2-week-old transgenic wheat plants and the BYSMV-RFP-infected barley plants at 14 dpi were used for immunolocalization. The tested leaves were crushed with a forcep at 2 h before fixing in the PBS buffer containing 4% (v/v) 1-ethyl-3-(3-dimethylaminopropyl)-carbodiimide hydrochloride, and subjected to vacuum application for 20 min. Subsequently, leaf samples in fixative buffer were frozen in liquid nitrogen, and incubated at 50°C for thawing. After two freeze–thawing cycles, the leaf samples were washed 3 times with PBS buffer, and incubated with 5% (w/v) BSA for blocking at 37°C for 1 h. The leaf samples were incubated with primary anti-JA rabbit antibodies (1:1,000) (Agrisera, Sweden, AS11-1,799) at 4°C overnight. After washed in 0.1% BSA in PBS 3 times, leaf samples were incubated with anti-rabbit-IgG conjugated with AlexaFluor488 (1:500) in PBS buffer containing 5% BSA at 37°C for 40 min. After washed 3 times, leaf samples were examined for GFP fluorescence under a Leica TCS SP8 confocal microscope (Germany). GFP was excited with 488 nm (2% laser intensity) and emissions were captured between 500 and 540 nm (gain value, 700). The negative control was directly incubated with rabbit-IgG-AlexaFluor488 without addition of anti-JA rabbit antibodies. All the images are representative results from at least three independent repeats.

Reverse transcription quantitative polymerase chain reaction (RT-qPCR)

RT-qPCR assays were performed as described previously (Chai et al., 2020). Briefly, total RNA was extracted from leaf samples using TRIzol reagent (Sigma-Aldrich, St Louis, MO, USA), and treated with DNase I (TaKaRa, Beijing, China) to remove contaminated DNA. Next, 1–2 μg total RNA was used as a template for reverse transcription with oligo (dT) using HiScript II Reverse Transcriptase (Vazyme, Nanjing, China). Quantitative PCR was carried out on the Bio-Rad CFX96 Real-Time PCR System using 2 × SsoFast EvaGreen Supermix (Bio-Rad, Beijing, China). The wheat actin3 (XM_037579409) and Arabidopsis actin1 (NM_179953), and the barley EF1α gene (Z50789) served as endogenous controls. Three independent repeats were collected for biological statistics. All primers were listed in Supplemental Table S1.

Insect attraction bioassays

All tested plants were pretreated with 200-μM MeJA at 2 h before insect attraction bioassays. For the Y-tube dual-choice assay, two hermetical-sealed glass chambers (40 cm in height and 23  cm in diameter) containing tested plants were connected to two branch arms of a glass Y-tube olfactometer (24 cm in length for each arm), respectively. Purified airflow was equally and continually pumped from the glass chambers to the olfactometer at 300 mL min⁻¹. The Y-tube olfactometer was exchanged after each replicate to avoid positional bias. Approximately 50 aviruliferous second/third-instar nymphs of SBPHs were individually released from the Y-tube base to allow insects choice. For field-like insect attraction assay, different tested plants were placed in two sides of an insect-proof net (30 cm × 15 cm × 30 cm, length × width × height). Approximately 60 aviruliferous second-instar nymphs of SBPHs were released from the net center. After 8 h, the numbers of SBPHs on healthy or infected barley plants were counted and analyzed. For circular-dish assay, leaves of plants for each indicated pair-wised comparisons were placed around a circular dish alternately. In each experiment, at least 40 aviruliferous aphids were released from the dish center. After 4 h, numbers of aphids choosing plants were counted and the ratios for the paired plants were calculated.

Y2H assays

The Y2H cDNA library from Arabidopsis was used for screening interaction proteins of BYSMV P6 in plants followed the protocol (Clontech, Mountain View, CA, USA). The P6 ORF was introduced into the pGBKT7 vector as a bait vector. To verify the interaction of P6 and HvCSN5, the ORFs of P6 and P6^I16A were individually cloned into the pGADT7 vector, and the ORF of HvCSN5 was introduced into the pGBKT7 vector. Plasmids expressing AD-P6/BK-HvCSN5, AD-P6^I16A/BK-HvCSN5, or other negative controls were transformed into AH109 yeast cells and grown in Trp/Leu double-deficiency yeast plates. The positive co-transformed cells were diluted and dropped on Trp/Leu double-deficiency and Trp/Leu/His/Ade quadruple-deficiency yeast plates and incubated at 30°C for 3–5 days.

Immunoblotting analysis

To generate anti-AtCUL1 antibodies, the cDNA fragment corresponding to the Arabidopsis CUL1 protein fragment (aa 1–382) was introduced into the pET-30a (+) vector to express the AtCUL1-6 × His protein in E.coli strain BL21. The AtCUL1-6 × His protein was obtained using the Ni-NTA agarose (Bio-Rad, Hercules, CA, USA, 780-0801), and then was used to immunize rabbits for the anti-AtCUL1 antibodies (Beijing Genomics institution, Beijing, China). Immunoblotting analyses were carried out as described previously (Gao et al., 2020). Briefly, total proteins were extracted from different plant samples in SDS extraction buffer (100-mM Tris (pH 6.8), 10% β-mercaptoethanol, 4% sodium dodecyl sulfate (SDS), 20% glycerol, and 0.2% bromophenol blue), then separated in SDS–PAGE (polyacrylamide gel) gels and transferred to nitrocellulose or PVDF (Polyvinylidene Fluoride) membranes. Then, membranes were blocked with 5% skim milk powder, and incubated with the polyclonal antiserum specific to the BYSMV N (1:3,000), P6 (1:3,000), RFP (1:2,000), HA (1:10,000), Flag (1:5,000), GFP (1:1,500), MBP (1:10,000), GST (1:5,000), His (1:5,000), Myc (1:10,000), or AtCUL1(1:1,000) antibodies at 37°C for 1 h. After washing 3 times, membranes were incubated with goat anti-rabbit or anti-mouse IgG horseradish peroxidase conjugate (1:30,000), and then added with Pierce ECL Plus chemiluminescent substrate, followed by exposure to X-ray films. For detection of endogenous Cul1 proteins, anti-rabbit IgG conjugated with alkaline phosphates (AP-A) served as secondary antibodies and then the substrate BCIP/NBT was added for detection. The uncropped images of immunoblotting analyses were showed in Supplemental Figures S13–S15.

Antigen affinity purification of antibodies for immunoblot of CUL1 proteins

For purification of CUL1 antibodies, the purified AtCUL1 protein from E. coli was separated in SDS–PAGE gels and transferred to nitrocellulose membranes. Then the membranes were incubated with serum waiting for purification (mixed with TBST in equal proportions) at 25°C for 3 h. After washing 2 times by TBST and 2 times by water, the membranes were washed by washing buffer (100-mM glycine [pH 2]) for 20 min at 25°C. The washing buffer was collected and added 1/10 volume neutralization buffer (1M Tris [pH 6.8]). The resultant washing buffer containing purified CUL1 antibodies was used as primary antibodies to detect endogenous CUL1 proteins of wheat and Arabidopsis plants.

BiFC assays

For BiFC assays, the full-length ORFs of HvCSN5, AtCSN5A, BYSMV P6, and P6 truncated and point mutants were introduced into the pSPYNE-35S or pSPYCE-35S vectors (Walter et al., 2004). The CK1 of N. benthaminana plants and RbcL served as a negative control, and the related Y^N and Y^C vectors have been described previously (Gao et al., 2020). Agrobacterium tumefaciens EHA105 strains containing various BiFC plasmids and the tomato bushy stunt virus P19-expressing plasmid were mixed and co-infiltrated into N. benthamiana leaves. At 3 dpi, YFP fluorescence was monitored with a Zeiss LSM710 confocal microscope. YFP was excited with 488 nm (2% laser intensity) and emissions were captured between 500 and 540 nm (gain value, 700). Immunoblotting assays were carried out to verify protein accumulation. All the images are representative results from at least three independent repeats.

Subcellular localization assay

For the subcellular localization study, the ORFs of BYSMV-P6 and CSN5 were cloned into the pGDG vector to generate P6-GFP and CSN5-GFP, then transformed individually into A. tumefaciens EHA105 strains. The cultures were infiltrated into N. benthamiana leaves. Three days after infiltration, the leaves were imaged with a Zeiss LSM710 confocal microscope. GFP was excited with 488 nm (2% laser intensity) and emissions were captured between 500 and 540 nm (gain value, 700). All experiments were repeated 3 times with similar results. All the images are representative results from at least three independent repeats.

Co-IP assay

Co-IP assays were performed as described previously (Zhang et al., 2020). The ORF of HvCSN5-Flag was cloned into the pMDC32 vector, and the ORFs of P6 and P6^I16A were introduced into the pGDG vector (Goodin et al., 2002) for expression of P6-GFP and P6^I16A-GFP, respectively. Nicotianabenthamiana leaves were agroinfiltrated with EHA105 strains harboring expression vectors of CSN5-Flag with P6-GFP, P6^I16A-GFP, or GFP. Total proteins were isolated from agroinfiltrated leaves into extraction buffer (25 mM Tris–HCl [pH 7.5], 150-mM NaCl, 1-mM EDTA, 10% [v/v] glycerol, 1% Tween-20, 5-mM DTT, 2% (w/v) polyvinylpolypyrrolidone, and protease inhibitor cocktail). After centrifugation, total protein supernatant was incubated with anti-Flag M2 agarose beads (Sigma-Aldrich) for 3 h at 4°C. After being washed 3 times with IP buffer (25-mM Tris–HCl pH 7.5, 150-mM NaCl, 1-mM EDTA, 10% [v/v] glycerol, 1% Tween-20, and 10-mM PMSF), the bound proteins were boiled at 100°C for 5 min and analyzed by immunoblotting analyses with anti-Flag or anti-GFP antibodies.

GST pull-down assay

The ORFs of HvCSN5 and AtCSN5A were cloned into the pET-30a (+) vector for expressing HvCSN5-6 × His and AtCSN5A-6 × His protein, respectively. The ORFs of P6 and P6^I16A were engineered into the pGEX-KG vector for expressing GST-P6 and GST-P6^I16A, respectively. Fusion proteins were expressed from the E. coli strain BL21 and induced by 0.1-mM isopropyl-D1-thiogalactopyranoside in 18°C for 16 h. In GST pull-down assays, GST, GST-P6, or GST-P6^I16A proteins were incubated with HvCSN5-6 × His or AtCSN5A-6 × His proteins in 500-µL binding buffer (50-mM Tris–HCl, pH 7.5, 250-mM NaCl, 0.6% Triton X-100, 0.1% glycerol, 1 × cocktail, 5 mM DTT) with 30-µL glutathione Sepharose4Bbeads (GEHealthcare, Cleveland, OH, USA) at 4°C for 2  h. After centrifugation at 800g for 1 min, beads were washed 5 times with different concentration gradients washing buffer (50-mM Tris–HCl pH7.5, 150–250-mM NaCl, 0.6% Triton X-100, 1 × cocktail), boiled in SDS buffer for immunoblotting analyses using anti-GST (1:5,000), and anti-His (1:5,000) antibodies.

JAZ protein degradation assay in vivo and in vitro

The degradation assays of JAZ proteins were carried out as described previously (He et al., 2020). For in vivo degradation assays, ORFs of HvJAZ3 and TaJAZ12 were introduced into the pSuper vector for expressing HvJAZ3-GFP and TaJAZ12-GFP, respectively. The cDNAs of HA-P6 and HA-P6^I16A were cloned into the pGD vector for expressing HA-P6 and HA-P6^I16A, respectively. HvJAZ3-GFP, TaJAZ12-GFP, or free GFP were co-expressed with EV, HA-P6, or HA-P6^I16A. At 72-h postagroinfiltration, infiltrated leaves were sprayed with 200-μM MeJA or mock buffer. Two hours later, the leaves were examined for GFP fluorescence using a Zeiss LSM710 confocal microscope or immunoblotting analyses. GFP was excited with 488 nm (2% laser intensity) and emissions were captured between 500 and 540 nm (gain value, 700). Furthermore, 50-μM MG132 was infiltrated into the leaves co-expressing HvJAZ3-GFP, TaJAZ12-GFP, or free GFP with EV 12 h before observation.

The degradation assays of AtJAZ1 were performed as described previously (Wu et al., 2017). The AtJAZ1 ORF was introduced into the pMAL-C2X vector for expressing MBP-AtJAZ1 protein in the E. coli strain BL21. The purified MBP-AtJAZ1 protein (5 µg) was incubated with 1-mL degradation buffer (50-mM Tris–HCl, pH 7.8, 100-mM NaCl, 10% (v/v) glycerol, 0.1% (v/v) Tween-20 and 20-mM β-mercaptoethanol) that contained crude total proteins (1 μg μL⁻¹) extracted from indicated plants. The mixture was incubated at room temperature and collected for protein detection by immunoblotting analyses at indicated time points. Each experiment was repeated independently at least 3 times.

Statistical analysis

Statistical analyses were performed using Graphpad Prism version 8. Normally distributed data were then analyzed using two tailed, paired t test. Distributed nonparametric data were assessed by Mann–Whitney test. One-way ANOVA （analysis of variance） followed by Tukey’s multiple comparison tests were used for analyses of the different treatments (P < 0.05). Asterisks indicate statistical significance (^*P < 0.05, ^**P < 0.01, ^***P < 0.001, and ^****P < 0.0001).

Accession numbers

The accession numbers for the genes discussed in this article are as follows: The accession numbers for the genes discussed in this article can be found on NCBI (National Center of Biotechnology Information) website: TaLipase (XM_037617170), TaLOX2 (XM_037583088), HvERF (KAE8816301), HvMYC2 (AK357816), HvCSN5 (AK372228), AtCUL1 (NM_001203732), AtJAZ1 (NM_001332386), HvJAZ3 (AK250045), TaJAZ12 (AK454215), TaCUL1 (AK332446.1), HvCUL1 (AK252626), Hvef1α (Z50789), Atactin1 (NM_179953), Taactin3 (XM_037579409), AtCSN5A (NM_102139), AtPDF1.2 (NM_123809), AtVSP2(NM_122386), AtTAT1 (BT000307), and NbCK1 (XM_019393615.1).

Supplemental data

The following materials are available in the online version of this article.

Supplemental Figure S1. PCR, protein detection, and phenotype of P6 in transgenic wheat plants.

Supplemental Figure S2. The replicates of the immunological detection of JA in wheat and barley in Figure 1, G and H.

Supplemental Figure S3. Phylogenetic tree and sequence alignment of plant CSN5 putative orthologs.

Supplemental Figure S4. Protein detection of BiFC assays and subcellular localization of P6 and CSN5.

Supplemental Figure S5. BiFC analyses of HvCSN5 interactions with the truncated P6.

Supplemental Figure S6. Alanine scanning identifies the P6 key amino acid for interaction with HvCSN5.

Supplemental Figure S7. Phylogenetic tree and sequence alignment of plant CUL1 putative orthologs.

Supplemental Figure S8. The repeats of the immunoblotting in Figure 4, A, B, and D.

Supplemental Figure S9. The phylogenetic tree of HvJAZ3, TaJAZ12, and Arabidopsis JAZ proteins.

Supplemental Figure S10. BiFC analysis and GST pull-down showing interactions of AtCSN5A–P6.

Supplemental Figure S11. PCR and protein detection of P6 in transgenic Arabidopsis lines.

Supplemental Figure S12. The repeats of the immunoblotting in Figure 5D.

Supplemental Figure S13. The uncropped images of immunoblotting in Figure 3.

Supplemental Figure S14. The uncropped images of immunoblotting in Figure 4.

Supplemental Figure S15. The uncropped images of immunoblotting in Figure 5.

Supplemental Table S1. The list of primers used in this study.

 

X.B.W. and D.M.G. conceived the original research plans. D.M.G. performed the majority of the experiments and was assisted by Z.J.Z., L.X., X.D.F., J.H.Q., Y.Z., Q.G., Z.H.D., W.Y.X., Y.Z.Y., X.B.W., and D.M.G. analyzed the data and drafted the manuscript. X.B.W. and Y.W. proofread and finalized the manuscript. X.B.W. agrees to serve as the author responsible for contact and ensures communication.

The author responsible for distribution of materials integral to the findings presented in this article in accordance with the policy described in the Instructions for Authors (https://academic.oup.com/plphys/pages/general-instructions) is Xian-Bing Wang (wangxianbing@cau.edu.cn).