Introduction

Autophagy is an essential and conserved process that leads to the degradation of intracellular components, including soluble proteins, misfolded proteins, organelles, and macromolecular complexes (Yu et al. 2018). It is crucial for cell homeostasis maintenance, growth and development, and environmental stress responses (Prasanth et al. 2011)and (Choi et al. 2013). According to the mechanism of action, there are three types of autophagy: macroautophagy, microautophagy, and chaperone-mediated autophagy (CMA). Only macroautophagy and microautophagy have been described in plants, and there is no evidence of CMA in plants (Yang and Liu 2022). Macroautophagy (hereafter referred to as autophagy) is the most common type of autophagy and has been extensively studied (Li and Vierstra 2012). In brief, autophagy is a catabolic process in which substrates are sequestered within double-membraned vesicles termed autophagosomes. The mature autophagosomes are then delivered to vacuoles or lysosomes for degradation, and the vehicles are released back into the cytosol for recycling. The complex series of processes underlying autophagosome initiation and maturation depends on the coordinated action of a conserved set of autophagy-related (ATG) proteins (Mizushima et al. 2011; Rubinsztein et al. 2012). ATG8/LC3 (autophagy-related protein 3/light chain 3) family proteins have emerged as central players in autophagosome biogenesis and cargo recruitment (Iman et al. 2017; Slobodkin and Elazar 2013). ATG8 lipidation is mediated by two ubiquitin-like conjugation pathways involving the E1-like ligase ATG7, the E2-ligase ATG3, and the E3-like ligase ATG5-ATG12-ATG16 complex (Yu et al. 2018). Cytosolic glyceraldehyde-3-phosphate dehydrogenases (GAPCs), which serve as negative regulators, interact with ATG3 to suppress autophagy (Han et al. 2015).

Recent studies reveal that autophagy, as an essential physiological process, participates in a variety of stress responses, including nutrient deprivation and immune activation (Avin-Wittenberg 2019; Chen et al. 2021). When suffering from nutritional starvation, plants enhance autophagic activity by promoting the expression of autophagy genes, boosting the metabolism and circulation of nutrients, and ensuring their survival (Masclaux-Daubresse et al. 2017). Moreover, autophagy plays a vital role in the interaction between plants and pathogens. However, its roles are complex and diverse in that autophagy can either enhance or inhibit plant defense responses (Leary et al. 2018; Gallegos 2018). Studies have shown that autophagy is activated in response to various DNA and RNA viruses with negative consequences for virus accumulation, suggesting the integration of autophagic mechanisms in basal antiviral defenses (Hafrén et al. 2017; Haxim et al. 2017; Li et al. 2018). Many viral proteins are targeted by the autophagy machinery for degradation. For example, viral silencing suppressors, such as HC-Pro encoded by tobacco etch virus and 2b encoded by cucumber mosaic virus, are degraded by autophagy, resulting in the suppression of virus accumulation (Nakahara et al. 2012; Jeon et al. 2017). Moreover, the RNA-dependent RNA polymerase (RdRp) of turnip mosaic virus (TuMV) is degraded by the autophagy pathway via direct interaction with ATG6/Beclin1, which is proposed to act as a cargo receptor (Li et al. 2018). Some other viral proteins are also subjected to autophagic degradation, such as the virulence-associated protein βC1 from cotton leaf curl Multan virus (CLCuMuV) (Haxim et al. 2017) and movement protein (MP) of citrus leaf blotch virus (CLBV) (Niu et al. 2021). On the other hand, viruses have also developed machinery to suppress autophagic activity to conquer the antiviral response. For instance, barley stripe mosaic virus (BSMV) subverts antiviral autophagy with the help of the γb protein, which disrupts ATG7-ATG8 interaction and thus impairs autophagosome formation through ATG8 binding (Yang et al. 2003). N. benthamiana and O. sativa contains three GAPCs in the cytosol. We found that all NbGAPCs (NbGAPC1, NbGAPC2, and NbGAPC3) and two OsGAPCs (OsGAPC2 and OsGAPC3) interacted with RSV CP (Fig. 2), implying that the interaction with GAPCs is important for viral infection. Whether RSV CP recruits GAPCs to regulate or hijack the glycolytic metabolic pathway of the host to promote viral replication remains unknown and needs to be investigated further. It was observed that stress conditions can induce the translocation of GAPC to the nucleus (Kim et al. 2022). It is also interesting to further examine whether RSV CP affects subcellular localization and enzymatic activities of GAPC and the possible consequence in the course of RSV infection.

We found that RSV CP could elevate autophagy in plants (Figs. 1 and 4); however, it was stable against autophagy and proteasomal degradation (Fig. 7A and B). In contrast, the accumulation of MP and NS3 was increased with E64d treatment (Fig. 7D and E), suggesting that MP and NS3 were subjected to selective autophagic degradation. The plant protein NbP3IP was previously identified as guiding the autophagic degradation of NS3 (Jiang et al. 2021). Interestingly, the degradation of MP and NS3 was inhibited by RSV CP expression (Fig. 7D and E). The CaMV gene VI product (P6), a major component of viral factory inclusion, protects P4 against autophagic degradation by sequestering it and coordinating particle assembly and storage (Hafrén et al. 2017). RSV MP was reported to directly bind to RSV CP (Zhang et al. 2008). Whether RSV CP protects RSV MP and NS3 through direct interaction is unclear. The mechanism by which RSV CP suppresses the autophagic degradation of other RSV proteins needs to be investigated further.

We also found that RSV CP-triggered autophagy could promote the autophagic degradation of NbREM1 (Fig. 7G), suggesting that RSV employs autophagy to benefit viral infection. Furthermore, RSV CP-triggered autophagy could elevate the degradation of an unrelated viral protein (MP) encoded by CLBV (Fig. 7C). The molecular mechanism by which RSV CP induces selective autophagy to target antiviral components but not its own viral products remain unclear and warrants further investigation.

Materials and methods

Plant materials and virus inoculation

N. benthamiana plants were grown in a greenhouse at 25 °C, with 70% relative humidity and 16 h of daylight. RSV was obtained from Nan**g City, Jiangsu province, China, kindly provided by Dr. Tong Zhou (Jiangsu Academy of Agricultural Sciences). The mechanical inoculation of RSV on N. benthamiana plants was performed as described previously (Kong et al. 2014). Briefly, the leaves of N. benthamiana plants at the six-leaf stage were dusted with carborundum powder and mechanically rubbed with viral inoculum prepared from RSV-infected rice leaves in 20 mM sodium phosphate buffer at pH 7.0.

Plasmid constructs

Total RNA was extracted from RSV-infected N. benthamiana leaves using Trizol (Invitrogen) and supplied as a template for reverse transcription (RT) using ReverTra Ace reverse transcriptase (Toyobo, Japan). The DNA fragments of RSV genes were amplified by PCR (polymerase chain reaction) using PrimeSTAR® HS DNA Polymerase (Takara Bio) and cloned into the responsive plasmid using the ClonExpress II One Step Cloning Kit (Vazyme, Nan**g, China). All plasmid constructs generated in this study are described in Supplementary Table 1. The Flag peptide was added at the N-terminal of ATG3. GFP-ATG8f has been described previously (Niu et al. 2022). All primers used in this study are listed in Supplementary Table 2.

Agrobacterium infiltration

The plasmid constructs were transformed into Agrobacterium tumefaciens strain GV3101. Agroinfiltration was performed as described previously.

Co-immunoprecipitation and mass spectrometry analysis

The RSV CP-GFP protein was transiently expressed and extracted from N. benthamiana. The GFP protein was prepared in parallel as a control. Co-immunoprecipitation (Co-IP) assays using GFP-Trap beads (ChromoTek, Germany) were performed as described previously (Sun et al. 2006). LC–MS/MS (Liquid Chromatograph Mass Spectrometer) and bioinformatics analyses for protein identification were performed by Shanghai Applied Protein Technology Co., Ltd.

Fluorescent protein expression and visual observation

The GFP-tagged and bimolecular fluorescence complementation (BiFC) proteins were transiently co-expressed on 4-week-old N. benthamiana leaves. At 3 dpi, the fluorescent proteins were observed using an FV3000 confocal microscope (Olympus, Japan). The fluorescence signals were visualized with laser excitation/emission filters of 488/500–510 nm for GFP and 514/580–600 nm for YFP (BiFC).

Western blot analysis

Western blot analysis was performed as described previously (Niu et al. 2022). Anti-GFP (1:5000, Sigma, Cat. No. F1804), anti-Flag (1:5000, EASYBIO, Cat. No. BE7001), anti-HA, anti-His, and anti-maltose-binding protein (MBP; all 1:5000, Bei**g Protein Innovation Co., Ltd.) antibodies, as well as secondary goat anti-mouse immunoglobulin G-horseradish peroxidase (IgG-HRP; 1:10,000, Proteintech) were used for the detection of GFP-, HA-, Flag-, His-, GST-, and MBP-tagged proteins. The actin protein was detected using a primary anti-actin antibody (1:5000, Kangwei). The detection of ATG8 and phosphatidylethanolamine-conjugated ATG8 (ATG8-PE) was carried out as described previously (Niu et al. 2022) using an anti-ATG8 primary antibody (1:2000, Abcam).

Competitive pull-down assay

The fusion proteins of MBP-ATG3, MBP-CP, GST-ATG3, and NbGAPC2-His were expressed in the Escherichia coli strain BL21 and purified using Maltose-Binding Glutathione Sepharose TM 4 Fast Flow (GE Healthcare) or Ni–NTA agarose (Qiagen) according to the manufacturer’s instructions. His pull-down assays were performed as described previously (Sun et al. 2013). Finally, samples were analyzed by western blotting with anti-GST (1:5000), anti-His (1:5000), and anti-MBP (1: 5000) antibodies.

Competitive Co-IP assay

The competitive protein RSV CP was fused with MBP and expressed in E. coli BL21 (DE3). The MBP-CP, MBP-CPN4A, and MBP proteins were purified using amylose resin with gradient column buffer as described previously (Sun et al. 2013). NbGAPC2-GFP and HA-ATG3 were transiently co-expressed in N. benthamiana leaves. The total proteins containing NbGAPC2-GFP and HA-ATG3 proteins were extracted from N. benthamiana leaves and mixed well with the purified MBP or MBP-CP (40 μg/ml), respectively, and subjected to immunoprecipitation using 20 μL GFP-Trap beads (ChromoTek, Germany) as described previously. Precipitates were washed five times with a wash buffer. Finally, samples were analyzed by western blotting with anti-GFP (1:5000), anti-HA (1:5000), and anti-MBP (1: 5000) antibodies.

Transmission electron microscope (TEM)

N. benthamiana leaf tissues with or without RSV infection (14 dpi) were prepared, subjected to vacuum infiltration, and fixed immediately with 2.5% glutaraldehyde (Sigma, G5882) in 0.1 M PBS overnight at 4 °C. Samples were washed three times with PBS, post-fixed with 1% OsO4 (Sigma, O5500), rinsed three times with PBS again, dehydrated in a graded ethanol series followed by the replacement of ethanol with acetone, and embedded in SPI-PON812 resin (SPI Science, 90,529–77–4). The ultrathin sections were stained with 2% (w/v) uranyl acetate (Polysciences, 21,447–25) and 2.6% (w/v) lead citrate (Sigma, 15,326). Images were observed and captured using a transmission electron microscope (TEM; Hitachi H-7650, Japan) at 80 kV (Guan et al. 2022).