Rice black-streaked dwarf virus P10 promotes phosphorylation of GAPDH (glyceraldehyde-3-phosphate dehydrogenase) to induce autophagy in Laodelphax striatellus

ABSTRACT

Macroautophagy/autophagy is an important innate and adaptive immune response that can clear microbial pathogens through guiding their degradation. Virus infection in animals and plants is also known to induce autophagy. However, how virus infection induces autophagy is largely unknown. Here, we provide evidence that the early phase of rice black-streaked dwarf virus (RBSDV) infection in Laodelphax striatellus can also induce autophagy, leading to suppression of RBSDV invasion and accumulation. We have determined that the main capsid protein of RBSDV (P10) is the inducer of autophagy. RBSDV P10 can specifically interact with GAPDH (glyceraldehyde-3-phosphate dehydrogenase), both in vitro and in vivo. Silencing of GAPDH in L. striatellus could significantly reduce the activity of autophagy induced by RBSDV infection. Furthermore, our results also showed that both RBSDV infection and RBSDV P10 alone can promote phosphorylation of AMP-activated protein kinase (AMPK), resulting in GAPDH phosphorylation and relocation of GAPDH from the cytoplasm into the nucleus in midgut cells of L. striatellus or Sf9 insect cells. Once inside the nucleus, phosphorylated GAPDH can activate autophagy to suppress virus infection. Together, these data illuminate the mechanism by which RBSDV induces autophagy in L. striatellus, and indicate that the autophagy pathway in an insect vector participates in the anti-RBSDV innate immune response.

Abbreviations3-MA: 3-methyladenine; AMPK: AMP-activated protein kinase; ATG: autophagy-related; co-IP: co-immunoprecipitation; DAPI: 4ʹ,6-diamidino-2-phenylindole; dpf: days post-feeding; dsRNA: double-stranded RNA; GAPDH: glyceraldehyde-3-phosphate dehydrogenase; GST: glutathione-S-transferase; RBSDV: Rice black-streaked dwarf virus; TEM: transmission electron microscope.

Introduction

Rice viruses are prevalent in many rice-growing countries and often cause serious damages to rice production. Among them, rice black-streaked dwarf virus (RBSDV), a member of the genus Fijivirus of family Reoviridae, causes important agricultural losses. RBSDV virions contain ten double-stranded RNA (dsRNA) segments, named S1 to S10 according to their molecular weight (small to large). RBSDV is transmitted by the small brown planthopper Laodelphax striatellus in a persistent and propagative manner. Therefore, elucidation of the mechanism(s) underlying the interactions between RBSDV and L. striatellus is crucial for the development of effective RBSDV control methods. In the present study, our goals are to investigate whether RBSDV infection causes autophagy, how RBSDV infection induces autophagy in L. striatellus, and how autophagy affects RBSDV infection in this insect vector.

Autophagy is a conserved cellular regulatory mechanism responsible for delivering cytoplasmic components to lysosomes/vacuoles for recycling [1]. Autophagy was initially found in yeast cells grown under starvation conditions [2], and can be divided into three types: microautophagy, macroautophagy, and chaperone-mediated autophagy [3]. Many ATG (autophagy-related) genes have been found to regulate autophagosome formation in cells as well as other autophagy processes [4]. In addition to ATGs, many other genes have also been shown to regulate various autophagy signaling pathways. For example, GAPDH (glyceraldehyde-3-phosphate dehydrogenase), encoding a glycolytic enzyme, has been shown to regulate autophagy in mouse embryonic fibroblasts to protect against cell death after it has been relocated from the cytoplasm to the nucleus [5]. Autophagy has also been reported as an important component of innate immunity against pathogen invasion [6]. Several studies have revealed that a number of viruses can be cleared from the cytoplasm in mouse embryonic fibroblasts and HeLa cells through autophagy [7]. Autophagy has also been shown to deliver viral pathogen-associated molecular patterns to endosomal pattern-recognition receptors to amplify the production of inflammatory cytokines, and to induce other antiviral adaptive immune responses [8,9]. Because autophagy plays many roles in host antiviral defense, viruses have also evolved strategies to subvert autophagy, inhibiting autophagosome formation and maturation during different stages of the virus life cycle [10,11].

To date, only two reports on autophagy in plant virus-infected insect cells have been published. Wang and others have reported that tomato yellow leaf curl virus (TYLCV) infection in whitefly can activate the autophagy pathway, leading to inhibition of virus transmission [12]. In 2017, Chen and colleagues reported that rice gall dwarf virus (RGDV) can induce formation of autophagosomes carrying RGDV virions to facilitate virus spread within its insect vector and transmission to host plants [13].

In this present study, we found that at 2 to 4 d post-feeding (dpf) on RBSDV-infected rice plants, RBSDV enters and then replicates in the midgut epithelial cells of L. striatellus. We also found that RBSDV infection can activate autophagy in these insect cells. Furthermore, the main capsid protein (also known as P10) of RBSDV alone can induce autophagy in both Sf9 and L. striatellus midgut cells. Our yeast two-hybrid (Y2H), glutathione-S-transferase (GST) affinity-isolation, and co-immunoprecipitation (co-IP) assays confirmed that RBSDV P10 can interact with GAPDH in vivo and in vitro. A previous study had indicated that GAPDH phosphorylation and the subsequent accumulation of GAPDH in the cell nucleus are crucial for the initiation of autophagy under glucose deprivation conditions in mouse embryonic fibroblasts [5]. In this study, we determined that both RBSDV infection and RBSDV P10 alone can promote phosphorylation of AMP-activated protein kinase (AMPK), resulting in GAPDH phosphorylation and relocation of GAPDH from the cytoplasm to the nucleus in both L. striatellus midgut cells and Sf9 cells. Once inside the nucleus, the phosphorylated GAPDH can activate autophagy to suppress virus infection. These results illuminate the mechanism by which RBSDV induces autophagy in L. striatellus, and indicate that the autophagy pathway in this insect vector participates in the anti-RBSDV innate immune response.

Results

RBSDV infection induces autophagic vesicle formation in RBSDV-viruliferous L. striatellus midgut epithelial cells

Autophagy in RBSDV-infected L. striatellus midgut cells was analyzed using the following three approaches: analysis of ATG8-II accumulation, observation of autophagosome, and localization of ATG8. Western blot results showed that the accumulation of ATG8-II in RBSDV-infected L. striatellus increased at 2 dpf, further increased at 3 dpf, declined at 4 dpf, and then disappeared at 6 dpf (Figure 1(A)). We then observed autophagosomes in the cells of RBSDV-viruliferous or non-viruliferous L. striatellus and counted the number of autophagosomes under a transmission electron microscope (TEM). The results showed that autophagic vesicles were present in L. striatellus midgut epithelial cells, and there were significantly more autophagosome-like vesicles in the RBSDV-viruliferous cells than that in the non-viruliferous cells at 3 dpf (Figure 1(B, C)). Confocal microscopy analysis also revealed that ATG8 puncta existed in the RBSDV-viruliferous midgut cells at 4 dpf (Figure 2). These data indicate that RBSDV infection can induce autophagy in RBSDV-viruliferous L. striatellus at the early infection stage.

Figure 1.

RBSDV infection induces autophagy in L. striatellus. (A) Changes in ATG8-II accumulation in viruliferous L. striatellus at various days post-feeding on RBSDV-infected rice plants. The accumulation of ATG8-II was determined by western blot assay using an ATG8-specific antibody. GAPDH was detected using a GAPDH-specific antibody and served as a control for sample loading. (B) The average number of autophagosomes in RBSDV-viruliferous and non-viruliferous L. striatellus midgut cells. Ten cells in each of ten sections were counted for each treatment. **, p < 0.01. (C) Morphology of autophagosomes in midgut cells of RBSDV-viruliferous and non-viruliferous L. striatellus was observed under a transmission electron microscope. Images on the right are the enlargements of the boxed areas on the left. Red arrows indicate autophagosomes.

Figure 2.

RBSDV infection-induced autophagy in L. striatellus midguts at various days post-feeding. Non-viruliferous or viruliferous L. striatellus midguts were harvested at 2, 4, and 12 d post-feeding (dpf), fixed, and immunolabeled for autophagosomes using an ATG8-specific antibody, followed labeling with goat anti-mouse IgG-conjugated Alexa Fluor 555 (red). RBSDV P10 was detected using an anti-RBSDV P10 IgG-FITC conjugate (green), and nuclei in midgut cells were visualized by DAPI staining (blue). Ten midguts were analyzed at each time point. A representative midgut imaged under a confocal microscope is shown. Scale bars: 100 μm.

Effects of autophagy on RBSDV invasion and accumulation in L. striatellus

To investigate the effect of autophagy on RBSDV accumulation in L. striatellus, we treated L. striatellus with an autophagy inhibitor, 3-methyladenine (3-MA), or an autophagy activator, rapamycin, for 24 h. The treated L. striatellus were allowed to feed on the RBSDV-infected rice plants for 3 d and then on the healthy rice seedlings for 10 d. Dimethylsulfoxide (DMSO)-treated L. striatellus were used as controls. The levels of RBSDV P10 protein and P10 RNA transcript in various samples were examined using western blot assay and RT-qPCR, respectively. The results showed that levels of RBSDV P10 protein and viral genome RNA transcripts were significantly increased in the 3-MA-treated L. striatellus, but significantly decreased in the rapamycin-treated L. striatellus compared with those in the DMSO-treated control (Figure 3(A, B)). In a separate experiment, dsRNAs representing partial sequences of the autophagy-related genes ATG3 (referred to as dsATG3), ATG5 (dsATG5), and ATG8 (dsATG8) were synthesized in vitro, and then injected individually into L. striatellus to silence the expression of these genes. L. striatellus injected with dsGFP were used as controls. The RT-qPCR results showed that the expression levels of ATG3, ATG5, and ATG8 in the dsATG3-, dsATG5-, and dsATG8-injected L. striatellus were reduced by about 80%, respectively, compared with those in the dsGFP-injected controls (Figure 4(A)). The dsATG3-, dsATG5-, and dsATG8- injected L. striatellus were then fed on RBSDV-infected rice plants, and we found that the percentage of RBSDV-viruliferous L. striatellus from these three treatments was notably higher compared with that of the dsGFP-injected control (Figure 4(B)). In addition, the survival rates of the dsATG3-, dsATG5-, and dsATG8-injected L. striatellus fed on the RBSDV-infected rice plants were significantly lower at 3 dpf compared with those of L. striatellus fed on uninfected rice plants (Figure 4(C)), suggesting that silencing of autophagy-related genes can negatively affect the L. striatellus immune system and that this weakened immune system may affect L. striatellus survival after RBSDV infection.

Figure 3.

Autophagy affects RBSDV replication in L. striatellus. (A) Western blot analysis showing the effects of 3-MA and rapamycin on RBSDV P10 accumulation in L. striatellus at 10 dpf on infected rice plants. L. striatellus treated with DMSO was used as a control (CK-). (B) RT-qPCR analysis showing the effects of 3-MA and rapamycin on the transcriptional level of the RBSDV P10 gene in L. striatellus at 10 dpf. Values are the means ± standard deviations (SD). *, p < 0.05; **, p < 0.01.

Figure 4.

Silencing of autophagy-related gene expression increases the viruliferous rate and reduces the survival rate of L. striatellus fed on RBSDV-infected rice plants. (A) Relative expression levels of ATG3, ATG5, and ATG8 in L. striatellus microinjected with LsATG3-dsRNA, LsATG5-dsRNA, and LsATG8-dsRNA, respectively, or GFP-dsRNA at 48 h post-injection. (B) Effect of silencing autophagy-related genes on the viruliferous rate of L. striatellus fed on RBSDV-infected rice plants. (C) Effect of silencing autophagy-related genes on the survival rate of L. striatellus fed on RBSDV-infected rice plants at 4 h post-feeding (hpf) and 3 dpf and on mock-infected plants (mock). Ns, not significant; *, P < 0.05; **, P < 0.01.

RBSDV P10 alone can induce autophagy in Sf9 insect cells

Because RBSDV infection can induce autophagy in L. striatellus, we decided to investigate if a single RBSDV protein is responsible for this. We individually expressed all the open reading frames (ORFs) of RBSDV in Sf9 insect cells using recombinant baculovirus expression vectors, and then analyzed the cells for autophagy activity using western blot assays and TEM. In the Sf9 cells expressing RBSDV P10, the accumulation of ATG8-II was clearly upregulated compared with that in Sf9 cells expressing other RBSDV ORFs or GFP, and in glucose-starved and P10-expressing Sf9 cells treated with 3-MA (Figure 5(A) and S1). The accumulation of ATG8-II in the P6-expressing Sf9 cells was similar to that in the GFP-expressing Sf9 cells, suggesting that the baculovirus expression vector can induce a low level of autophagy (Figure 5(A)). At the same time, our results indicated that glucose starvation treatment can induce autophagy in Sf9 cells (Figure 5(A)). We then examined the formation of autophagosomes in the P10-expressing Sf9 cells, non-transfected Sf9 cells, and Sf9 cells transfected with the empty baculovirus vector under TEM. We found that the average number of autophagosomes in the P10-expressing Sf9 cells was much higher than that in the empty baculovirus vector-transfected Sf9 cells or in the non-transfected Sf9 cells (Figure 5(B, C)). These data indicate that RBSDV P10 is responsible for inducing autophagy.

Figure 5.

RBSDV P10 can induce autophagy in Sf9 cells. (A) Expression of ATG8-II in RBSDV P10-, P6-, or GFP-expressing Sf9 cells, glucose (Glu)-starved Sf9 cells, and non-transfected (WT) Sf9 cells was analyzed by western blot assay using various antibodies. GAPDH accumulation was used to show sample loading. (B) Morphology of autophagosomes in RBSDV P10-expressing Sf9 cells and in control Sf9 cells (empty vector [EV]-transfected or non-transfected [WT]) was observed under a TEM. (C) Average numbers of autophagosome-like vesicles in the RBSDV P10-expressing and control Sf9 cells. Ten cells in ten sections were counted for each treatment. Values are the mean ± SD. *, p < 0.05.

Identification of L. striatellus proteins that can interact with RBSDV P10

To investigate how RBSDV P10 induces autophagy in L. striatellus, we performed a Y2H assay to screen an L. striatellus cDNA library. In this screen, a GAPDH-encoded protein segment (140–332 aa) was found to interact with RBSDV P10. To validate this finding, we cloned the full-length LsGAPDH gene from L. striatellus total RNA using RT-PCR. Unfortunately, in a subsequent Y2H assay, full-length LsGAPDH did not interact with RBSDV P10 (Figure 6(A)). Therefore, we performed an affinity-isolation assay and co-IP assay to further test the interaction. After the purified GST-tagged P10 (GST-P10) protein was incubated with the purified MBP-tagged LsGAPDH (MBP-LsGAPDH) protein and glutathione-agarose beads, the samples were analyzed by western blot assays using an anti-MBP antibody or an anti-P10 monoclonal antibody. The results revealed that purified GST-P10 did interact with MBP-LsGAPDH (Figure 6(B, C)). No interactions were found between GST-P10 and the MBP tag or between MBP-LsGAPDH and the GST tag. Sf9 cells co-expressing LsGAPDH-GFP and RBSDV P10-HA or GFP and RBSDV P10-HA were used for a co-IP assay with anti-HA antibody beads and western blot with anti-HA and anti-GFP antibodies. The co-IP assay also demonstrated that HA-P10 could be co-immunoprecipitated with LsGAPDH-GFP, but not with GFP alone (Figure 6(D)), indicating that RBSDV P10 can specifically interact with LsGAPDH in vitro and in Sf9 cells.

Figure 6.

LsGAPDH interacts with RBSDV P10 in vitro and in Sf9 cells. (A) Yeast two-hybrid analysis of the interaction between LsGAPDH and RBSDV P10. Full-length LsGAPDH and its 140–332 aa truncated mutant were separately cloned into pGADT7, and RBSDV P10 was cloned into pGBKT7. Serial dilutions of yeast cells co-transfected with two recombination vectors were plated on SD-Trp-Leu-His-Ade medium. Cells co-transfected with pGADT7-T and pGBKT7-p53 or with pGADT7-T and pGBKT7-Lam were used as positive and negative controls, respectively. Positive interactions are indicated by the growth of the yeast cells. (B) LsGAPDH interacts with RBSDV P10, as determined by GST affinity-isolation assay. Recombinant protein GST-P10 or GST tag was incubated with MBP-LsGAPDH and glutathione-sepharose beads. Affinity-isolated products were analyzed by western blot using an anti-MBP antibody. (C) LsGAPDH interacts with RBSDV P10, as determined by MBP affinity-isolation assay. Recombinant protein MBP-LsGAPDH or MBP tag was incubated with GST-P10 and MBP binding beads. Affinity-isolated products were analyzed by western blot using an anti-RBSDV P10 monoclonal antibody. (D) LsGAPDH interacts with RBSDV P10, as determined by co-IP assay. Sf9 cells co-expressing LsGAPDH-GFP and RBSDV P10-HA or GFP and RBSDV P10-HA for 48 h. Whole-cell lysates were used for co-IP assay with anti-GFP antibody beads and western blot with anti-HA (upper and lower panels) and anti-GFP (middle panel) antibodies. IP: Immunoprecipitation. IB: Immunoblotting.

RBSDV P10 co-localizes with LsGAPDH to induce autophagy in L. striatellus midgut cells

To determine the subcellular localization patterns of RBSDV P10 and LsGAPDH in L. striatellus midgut cells, we allowed L. striatellus to feed on RBSDV-infected rice plants for 2 d. At 4, 7, and 12 dpf, the digestive organ of L. striatellus was collected and analyzed for subcellular localization of RBSDV P10 and LsGAPDH by confocal microscopy using a P10- or GAPDH-specific antibody. The results showed that P10 and GAPDH co-localized in the L. striatellus midgut epithelial cells. This co-localization was the strongest at 4 and 7 dpf, and then gradually diminished (Figure 7).

Figure 7.

Immunofluorescence assay showing co-localization of RBSDV P10 with LsGAPDH in viruliferous L. striatellus midguts at 4, 7, and 12 dpf. Non-viruliferous and viruliferous insect midguts were immunostained with anti-RBSDV P10 monoclonal antibody-FITC (green) and anti-GAPDH mouse antibody followed by labeling with goat anti-mouse IgG conjugated with Alexa Fluor 555 (red). Nuclei of midgut cells were stained blue using DAPI (blue). White arrows indicate the position of RBSDV P10 and co-localization of RBSDV P10 and LsGAPDH. Second row is the enlarged images of the boxed areas in the images in the first row at 4 dpf. Scale bars: 50 μm.

To further confirm this finding, we examined whether the recombinant RBSDV P10 protein also colocalized with GAPDH in L. striatellus midgut epithelial cells. Purified recombinant P10-GST or GST (control) was added to an artificial diet, which was used to feed L. striatellus. Confocal images showed that P10-GST (green) can enter L. striatellus midgut epithelial cells and accumulate in them, but GST cannot (Figure 8). This finding indicates that P10 is capable of passing through the membrane of L. striatellus midgut epithelial cells to reach the cytoplasm. Importantly, the recombinant RBSDV P10 colocalized with LsGAPDH in the midgut epithelial cells (Figure 9(A)). To determine whether recombinant P10 can also induce autophagy in L. striatellus midgut cells, we analyzed the midguts from L. striatellus fed with an artificial diet containing P10-GST through TEM and western blot assay. We found that the number of autophagosomes in the midgut epithelial cells of L. striatellus fed with the artificial diet containing P10-GST was much higher than that in cells of L. striatellus fed with an artificial diet containing GST (Figure 9(B, C)). In addition, western blot assays showed that more autophagy was induced in the midguts of L. striatellus fed with P10-GST compared with that in L. striatellus fed with GST (Figure 9(D)). These results indicate that RBSDV P10 can cause GAPDH gathering in L. striatellus midgut epithelial cells and induce autophagy.

Figure 8.

Recombinant RBSDV P10 can pass through the L. striatellus cell membrane to enter the cytoplasm. Immunolabeling assay showing that RBSDV P10 can enter midgut cells after L. striatellus is fed with an artificial diet containing purified recombinant RBSDV P10. Midguts were separated from L. striatellus fed with RBSDV P10-GST or GST and analyzed by immunolabeling using an anti-RBSDV P10 monoclonal antibody conjugated with FITC (green) or an anti-GST antibody conjugated with FITC (the negative control). The samples were examined and photographed under a confocal microscope. The same midguts were also labeled with Phalloidin-Alexa Fluor 680 (red) for visualization of microfilaments and stained with DAPI (blue) for visualization of nuclei. The middle panel contains enlarged images of the boxed areas in the upper images. Scale bars: 50 μm.

Figure 9.

Recombinant RBSDV P10 colocalizes with LsGAPDH and induces autophagy in L. striatellus midgut cells. (A) Immunolabeling assay showing that recombinant RBSDV P10 can co-localize with LsGAPDH in L. striatellus midgut cells. L. striatellus midguts were immunolabeled with an anti-RBSDV P10 monoclonal antibody conjugated with FITC (green) or an anti-GST antibody conjugated with FITC (the negative control) and a GAPDH-specific mouse antibody followed by labeling with goat anti-mouse IgG conjugated with Alexa Fluor 555 (red). White arrows indicate the labeled RBSDV P10, LsGAPDH, or both. Scale bars: 50 μm. (B) Morphology of autophagosomes in midgut cells of L. striatellus fed with an artificial diet containing P10-GST or GST was observed under a transmission electron microscope. (C) Average numbers of autophagosome-like vesicles in midgut cells of L. striatellus fed with an artificial diet containing P10-GST or GST. Ten cells in ten sections were counted for each treatment. *, p < 0.05. (D) Western blot analysis of ATG8-II accumulation in L. striatellus midguts at 48 h post-feeding of L. striatellus with an artificial diet containing purified recombinant RBSDV P10-GST or GST. L. striatellus fed on RBSDV-infected rice plants (viruliferous) or on non-infected rice plants (non-viruliferous) were used as positive and negative controls, respectively. Accumulation of GAPDH was detected using a GAPDH-specific antibody and used to show sample loading.

RBSDV P10 induces autophagy by promoting AMPK and GAPDH phosphorylation and subsequent nuclear translocation of GAPDH

GAPDH has been shown to interact with ATG3 in Nicotiana benthamiana cells to suppress autophagy [14]. In mouse cells, GAPDH has been shown to be phosphorylated by AMPK in the cytoplasm and then translocated to the nucleus to activate autophagy under glucose deprivation conditions [5]. To investigate how GAPDH regulates autophagy in RBSDV-viruliferous L. striatellus, we examined the interaction between GAPDH and ATG3. Y2H and MBP affinity-isolation assay results demonstrated that LsGAPDH did interact with LsATG3 (Figure 10(A, B)). To further confirm that GAPDH is involved in autophagy induction in insect cells, we silenced GAPDH expression in Sf9 and L. striatellus cells through transfection or injection with in vitro-synthesized GAPDH dsRNA (dsGAPDH). The efficiency of GAPDH silencing and the accumulation of ATG8-II were determined by RT-qPCR and western blot assays, respectively. The results of RT-qPCR showed that at 48 h after dsGAPDH treatment, the expression of GAPDH in the Sf9 and L. striatellus cells was significantly downregulated (Figure 10(C, E)), and western blot assays revealed that protein accumulation was also downregulated (Figure 10(D, F)). Although our results showed that LsGAPDH could interact with LsATG3, silencing of GAPDH expression failed to induce autophagy in the Sf9 or L. striatellus cells (Figure 10(D, F)), indicating that the interaction between GAPDH and ATG3 alone is not required to negatively regulate autophagy in insect cells as it is in N. benthamiana cells [14]. In another experiment, dsGAPDH-injected L. striatellus were allowed to feed on the RBSDV-infected rice plants for 3 d and then analyzed for ATG8-II accumulation by western blot assay. The result showed that, compared with dsGFP-injected L. striatellus fed on RBSDV-infected rice plants, the accumulation of ATG8-II in dsGAPDH-injected L. striatellus fed on RBSDV-infected rice plants was significantly lower (Figure 10(F)), indicating that silencing of GAPDH expression in L. striatellus cells can suppress the autophagy induced by RBSDV infection.

Figure 10.

Silencing GAPDH does not induce autophagy in insect cells. (A) Results of a yeast two-hybrid assay identifying the interaction between LsGAPDH and LsATG3. LsGAPDH and LsATG3 were cloned individually into pGAD-T7 (AD) and pGBK-T7 (BD). Serial dilutions of yeast cells co-transfected with two combined vectors were plated on SD-LTHA medium. (B) MBP affinity-isolation assay identifying the interaction between LsGAPDH and LsATG3. Recombinant MBP-LsGAPDH or MBP tag was incubated with His-LsATG3 and MBP binding beads. Affinity-isolated products were analyzed by western blot using an anti-His antibody (upper panel). The middle panel shows His-tagged protein inputs in the affinity-isolation assay. Equal volumes of MBP binding beads carrying MBP-LsGAPDH or the MBP tag were determined by SDS-PAGE and stained with Coomassie Brilliant Blue (lower panel). IP: Immunoprecipitation. IB: Immunoblotting. (C) Relative expression level of SfGAPDH determined by RT-qPCR in GAPDH-dsRNA- or GFP-dsRNA-transfected Sf9 cells at 48 h post-transfection. (D) Detection of autophagy in GAPDH-dsRNA- or GFP-dsRNA-transfected Sf9 cells by western blot assays of ATG8-II accumulation. GAPDH was detected with a GAPDH-specific antibody. (E) Relative expression level of LsGAPDH determined by RT-qPCR in LsGAPDH-dsRNA- or GFP-dsRNA-microinjected L. striatellus at 48 h post-injection (hpi). (F) Detection of autophagy in LsGAPDH-dsRNA- or GFP-dsRNA-microinjected viruliferous and non-viruliferous L. striatellus at 3 dpi by western blot assays of ATG8-II accumulation. Accumulation of TUBB was detected using a TUBB-specific antibody and used to show sample loading.

Sequence alignment of the predicted GAPDH amino acid sequences from L. striatellus, mouse, and Sf9 cells showed that the LsGAPDH and Sf9GAPDH sequences had the same phosphorylation motif containing Ser95 (Figure 11(A)), which was reported to contain a Ser phosphorylation site in mouse GAPDH [5]. We then analyzed the intracellular distribution of GAPDH in Sf9 cells grown on a glucose starvation medium by confocal microscopy, and compared this pattern with that in cells grown on a medium supplemented with glucose. We found that in the cells grown in glucose starvation medium, GAPDH was redistributed from the cytoplasm to the nucleus (Figure 11(B)). To further ascertain the relationship between phosphorylation of GAPDH and its nuclear translocation, two GAPDH mutants were constructed in which the Ser95 residue was replaced by alanine (for non-phosphorylated GAPDH) or aspartic acid (for phospho-mimic GAPDH). GAPDH mutants were transfected into Sf9 cells and the location of mutated GAPDH was observed. The non-phosphorylated mutant GAPDH^S95A was located in the cytoplasm, but the phospho-mimic mutant GAPDH^S95D was predominantly localized in the nucleus (Figure 11(C)). Consistent with these observations, significantly more accumulation of ATG8-II was observed in GAPDH^S95D-expressing Sf9 cells grown on nutrient medium than that in GAPDH^S95A-expressing Sf9 cells (Figure 11(D)). These data suggested that phosphorylation of GAPDH at Ser95 induced by the phosphorylated AMPK mediates nuclear translocation and autophagy initiation. We then expressed RBSDV P10 in Sf9 cells under non-starvation conditions and analyzed the distribution of GAPDH by confocal microscopy. We observed that in the RBSDV P10-expressing Sf9 cells, GAPDH was also relocated from the cytoplasm to the nucleus (Figure 12(A, B)). In addition, the phosphorylation level of AMPK and GAPDH in RBSDV P10-expressing Sf9 cells was significantly higher than that in GFP-expressing Sf9 cells (Figure 12(C, D)). co-IP assay demonstrated the interaction between GAPDH and AMPK, and in RBSDV P10-expressing Sf9 cells, more AMPK proteins were co-immunoprecipitated with GAPDH (Figure 12(E)). Furthermore, in vitro phosphorylation assay showed that the phosphorylated AMPK from Sf9 cells expressing RBSDV P10 can phosphorylate purified prokaryotic expression GAPDH (Figure 12(F)). We then further validated these results in L. striatellus cells. Confocal images showed that RBSDV infection caused the nuclear translocation of LsGAPDH in the midgut epithelial cells of L. striatellus (Figure 13(A)). This nuclear translocation phenomenon could be observed in the midgut epithelial cells of L. striatellus fed with an artificial diet containing recombinant RBSDV P10-GST (Figure 13(B)). Western blot assay results further demonstrated that the AMPK phosphorylation level in RBSDV-infected L. striatellus increased at 3 dpf, while the AMPK phosphorylation level in non-infected L. striatellus did not change (Figure 13(C, D)), confirming that GAPDH does play a positive role in the RBSDV infection-induced autophagy in L. striatellus cells.

Figure 11.

Phosphorylation at Ser95 mediates GAPDH nuclear translocation in Sf9 cells. (A) Alignment of mouse, L. striatellus, and S. frugiperda GAPDH amino acid sequences. Black box indicates the phosphorylation site (Ser95). (B) Glucose starvation results in the nuclear translocation of GAPDH. Sf9 cells grown under glucose starvation or non-starvation conditions were probed with an anti-GAPDH mouse antibody followed by labeling with goat anti-mouse IgG conjugated with Alexa Fluor 555 (red). Nuclei in these cells were stained with DAPI (blue). Scale bars: 10 μm. (C) Nuclear translocation analyses of a phospho-mimic form of GAPDH^S95D mutation under glucose conditions and a non-phosphorylated form of GAPDH^S95A mutation under glucose starvation conditions. These Sf9 cells were probed with an anti-His tag antibody conjugated with Alexa Fluor 555 (red). Nuclei in these cells were stained with DAPI (blue). Scale bars: 10 μm. (D) Detection of autophagy in GAPDH^S95A- and GAPDH^S95D-expressing Sf9 cells by western blot analysis of ATG8-II accumulation. Accumulation of TUBB was detected using a TUBB-specific antibody and used to show sample loading.

Figure 12.

RBSDV P10 induces autophagy via phosphorylation of GAPDH by AMPK and subsequent nuclear translocation. (A) Confocal images of Sf9 cells transfected with a vector expressing GFP (GFP) or RBSDV P10-GFP (P10-GFP) under non-starvation conditions. These Sf9 cells were probed with an anti-GAPDH mouse antibody followed by labeling with goat anti-mouse IgG conjugated with Alexa Fluor 555 (red). Nuclei in these cells were stained with DAPI (blue). Scale bars: 10 μm. (B) Western blot analyses of GAPDH accumulation in the nuclear (N) or cytoplasmic (C) fraction from Sf9 cells expressing GFP or RBSDV P10-GFP. Histone H3 was detected using an anti-Histone H3-specific antibody and used as a marker for the nuclear fraction. TUBB was detected using an anti-TUBB-specific antibody and used as a marker for the cytoplasmic fraction. (C) Western blot analyses of AMPK phosphorylation level in Sf9 cells expressing RBSDV P10 or GFP, and in Sf9 cells grown under glucose starvation or non-starvation conditions. (D) Western blot analyses of GAPDH phosphorylation level in Sf9 cells expressing RBSDV P10 or GFP, and in Sf9 cells grown under glucose starvation or non-starvation conditions. IP: GAPDH indicates GAPDH immunoprecipitated with a GAPDH-specific antibody. (E) Co-immunoprecipitation of AMPK with GAPDH in RBSDV P10-GFP or GFP expressing Sf9 cells. Whole-cell lysates were used for co-IP assay with anti-GAPDH antibody and the co-immunoprecipitated AMPK and GAPDH were detected by western blot assays with anti-AMPK and anti-GAPDH. IP: GAPDH indicates GAPDH immunoprecipitated with a GAPDH-specific antibody. (F) In vitro phosphorylation assay of GAPDH by the phosphorylated AMPK. The phosphorylated AMPK was immunoprecipitated from RBSDV P10-expressing Sf9 cells using anti-AMPK antibody. The phosphorylation of GAPDH was analyzed by western blot using an anti-phospho-(Ser/Thr) antibody.

Figure 13.

RBSDV infection or feeding with recombinant RBSDV P10 regulates the nuclear translocation of GAPDH in L. striatellus midgut epithelial cells. (A) Viruliferous or non-viruliferous insect midguts were immunostained with anti-RBSDV P10 monoclonal antibody-FITC (green) and anti-GAPDH mouse antibody followed by labeling with goat anti-mouse IgG conjugated with Alexa Fluor 555 (red). Nuclei of midgut cells were stained using DAPI (blue). Second row is the enlarged images of the boxed areas of images in the upper row. Scale bars: 50 μm. (B) Feeding recombinant RBSDV P10 regulates the nuclear translocation of GAPDH in L. striatellus midgut epithelial cells. Confocal images of midgut cells after feeding L. striatellus with an artificial diet containing the purified recombinant RBSDV P10-GST or GST tag. Insect midguts were immunostained with anti-RBSDV P10 monoclonal antibody-FITC (green) or an anti-GST antibody-FITC (the negative control) and anti-GAPDH mouse antibody followed by labeling with goat anti-mouse IgG conjugated with Alexa Fluor 555 (red). Nuclei of midgut cells were stained using DAPI (blue). Second row is the enlarged images of the boxed areas of images in the upper row. White arrows indicate the labeled LsGAPDH proteins, nuclei, or both. Scale bars: 50 μm. (C and D) Western blot analyses of AMPK phosphorylation level in viruliferous (C) and mock-infected (D) L. striatellus at various days post-feeding on RBSDV-infected and healthy rice plants, respectively.

Because AMPK is the upstream phosphorylation kinase of GAPDH, we silenced AMPKα expression in Sf9 cells treated with dsAMPKα and then expressed RBSDV P10 in those treated cells. Using immunofluorescence microscopy, we found that silencing of AMPK in Sf9 cells inhibits the redistribution of GAPDH from the cytoplasm to the nucleus caused by RBSDV P10 (Figure 14(A)). Moreover, the phosphorylation level of GAPDH and the accumulation of ATG8-II in the AMPK-silenced Sf9 cells were significantly lower than those in the dsGST control (Figure 14(B)). These data further indicate that AMPK can phosphorylate GAPDH, leading to the nuclear translocation of GAPDH and autophagy. In addition, the affinity-isolation assay and bio-layer interferometry (BLI) binding analysis results also showed that RBSDV P10 interacted with AMPK (Figure 14(C, D)). Together, our above data indicated that RBSDV P10 alone can promote phosphorylation of AMPK, resulting in GAPDH phosphorylation and its nuclear translocation, and then autophagy activation in midgut cells of L. striatellus or Sf9 insect cells.

Figure 14.

Silencing AMPKα in Sf9 cells suppresses GAPDH phosphorylation and its nuclear translocation, and autophagy activation caused by RBSDV P10. (A) Silencing of AMPKα expression in Sf9 cells suppresses the nuclear translocation of GAPDH caused by RBSDV P10. Confocal images of Sf9 cells treated with dsAMPKα or dsGST (the negative control), then transfected with a vector expressing RBSDV P10-GFP (P10-GFP). These Sf9 cells were probed with an anti-GAPDH mouse antibody followed by labeling with goat anti-mouse IgG conjugated with Alexa Fluor 555 (red). Nuclei in these cells were stained with DAPI (blue). Scale bars: 10 μm. (B) Western blot analyses of GAPDH phosphorylation and ATG8-II accumulation levels in AMPK-silenced Sf9 cells expressing RBSDV P10. IP: GAPDH indicates GAPDH immunoprecipitated with a GAPDH-specific antibody. (C) GST affinity-isolation assay determining the interaction between LsAMPK and RBSDV P10. Recombinant GST-P10 or GST tag was incubated with His-LsAMPK and glutathione-sepharose beads. Affinity-isolated products were analyzed by western blot assay using an anti-His antibody (upper panel). The middle panel shows His-tagged protein inputs in the affinity-isolation assay. Equal volumes of GST binding beads carrying GST-P10 or the GST tag were determined by SDS-PAGE and stained with Coomassie Brilliant Blue (lower panel). IP: Immunoprecipitation. IB: Immunoblotting. (D) Bio-layer interferometry (BLI) analysis of recombinant His-AMPK and GST-P10 using anti-His biosensors. The red line represents the BLI curve between AMPK and RBSDV P10, and the green line represents the BLI curve between AMPK and GST tag as the negative control.

Discussion

Autophagy is an essential and conserved cellular degradation pathway in eukaryotes and plays important roles during immune responses to invading pathogens or immune signaling molecules [15]. Autophagy has also been shown to function in eukaryotic cell defense against virus invasion. A wide range of DNA and RNA viruses were found to regulate the autophagy pathway. Accumulating evidence indicates that autophagy plays an antiviral during virus infection. For example, the cotton leaf curl Multan virus βc1 protein can induce autophagy by disrupting the interaction of ATG3 with GAPDH dehydrogenases [16], and the plant autophagic machinery can target the βc1 protein for degradation through its interaction with ATG8 [17]. Herpes simplex virus can induce autophagy in both immature and mature dendritic cells, which is crucial for the induction of antiviral immune responses [18]. Autophagy also plays an essential role in protecting against tissue injury and cell death during viral infection. Orvedahl et al. reported that autophagy protects against SIN pathogenesis through a cell-autonomous mechanism that facilitates viral protein clearance and prevents virus-induced cell death [19]. Moreover, some viruses have evolved strategies for utilizing the components of the autophagic pathway to facilitate their replication [20]. For example, human rhinovirus (HRV) type 2 or type 14 infection can induce the accumulation of both early and late autophagic vacuoles [21], and HRV subverts the autophagic machinery to promote viral replication [22]. Barley stripe mosaic virus γb protein can subvert autophagy-mediated antiviral defense by disrupting the ATG7-ATG8 interaction to promote plant RNA virus infection [23]. Wang et al. reported that Influenza A virus replication requires an autophagy pathway to enhance viral RNA synthesis via PB2 (a subunit of IAV polymerase complex)-HSP90AA1 interaction by modulating HSP90AA1 expression and the AKT-MTOR signaling pathway in host cells [24]. Hepatitis C virus (HCV) infection activates MTOR signaling and autophagy via upregulation of BECN1 (beclin 1) [25], and induces the unfolded protein response, which in turn activates the autophagic pathway to promote HCV RNA replication in human hepatoma cells [26]. Foot-and-mouth disease virus triggers autophagy and utilizes the autophagy pathway to facilitate virus replication [27]. Zhang et al. demonstrated that autophagy was induced in encephalomyocarditis virus (EMCV)-infected host cells by monitoring the presence of autophagosomes and the modification of LC3 [28]; in addition, TMEM39A (transmembrane protein 39A) was shown to interact with EMCV capsid proteins to positively regulate the EMCV replication through an autophagy-dependent pathway [29,30].

It is well known that most plant viruses that cause large-scale damage are transmitted by insect vectors. Accumulating evidence indicates that some plant viruses can induce autophagy in their insect vectors [12,31]. To date, the mechanism underlying plant virus infection-induced autophagy in insects remains largely unknown. Here, we present evidence that RBSDV infection in L. striatellus can rapidly induce autophagy in midgut cells. Our results also demonstrate that RBSDV P10 alone can induce autophagy in cells, resulting in an antiviral response in RBSDV-viruliferous L. striatellus. RBSDV P10 was further found to interact with AMPK and GAPDH and promote AMPK and GAPDH phosphorylation. The phosphorylated GAPDH was then translocated into the nucleus to activate the autophagy pathway. This finding agrees with previously published results that phosphorylation and nuclear translocation of GAPDH are required for AMPK activation-evoked autophagy in mouse cells under glucose starvation conditions [5]. This study is the first report on the mechanism by which plant viruses cause autophagy in their insect vectors.

In this work, we also found that RBSDV infection in L. striatellus can increase the accumulation of ATG8-II and the number of double-membrane autophagosomes at 2–4 d post-feeding on RBSDV-infected rice plants (Figure 1(A), 2). This finding indicates that the induction of autophagy occurs soon after the invasion of RBSDV into L. striatellus midgut cells. Our autophagy activation and suppression experiments demonstrated that autophagy can strongly inhibit RBSDV accumulation in L. striatellus midgut cells (Figure 3(A)). To determine whether RBSDV-encoded proteins can induce autophagy, we individually expressed all RBSDV-encoded proteins or GFP in Sf9 cells and analyzed the autophagy induction in these cells. The results showed that only RBSDV P10 was able to induce autophagy in Sf9 cells (Figure 5 and S1). This finding was supported by the results from feeding experiments using an artificial diet containing the purified recombinant RBSDV P10 (Figure 9(B, D)). To investigate how RBSDV P10 regulates autophagy in cells, we screened a L. striatellus library using RBSDV P10 as a bait, and identified LsGAPDH as an interactor. This interaction was further confirmed through affinity-isolation and co-IP assays using full-length LsGAPDH and RBSDV P10 (Figure 6). Confocal microscopy analysis showed that RBSDV P10 colocalizes with LsGAPDH in L. striatellus midgut cells (Figures 7 ,9(A)).

GAPDH is a highly conserved enzyme that can convert glyceraldehyde-3-phosphate to D-glycerate 1, 3-bisphosphate. GAPDH is not only found in the cytosol but is also associated with the plasma membrane, nuclear membrane, endoplasmic reticulum, Golgi, and nucleus [27]. More and more pieces of evidence suggest that GAPDH is related to autophagy. Han et al. [14] showed that ATG3 interacts with GAPDH in N. benthamiana plants, and that silencing of GAPDH significantly activates ATG3-dependent autophagy. This demonstrates that ATG3 interacts with GAPDH to negatively regulate autophagy in N. benthamiana [14]. Furthermore, the ßC1 protein of cotton leaf curl Multan betasatellite interacts with GAPDH to induce autophagy in N. benthamiana plants by disrupting GAPDH-ATG3 interaction [16]

In our study, we showed that LsGAPDH does interact with LsATG3 (Figure 10(A, B)). However, silencing of GAPDH did not induce autophagy in Sf9 or L. striatellus cells (Figure 10(D, F)). Furthermore, silencing GAPDH expression in L. striatellus suppressed autophagy induced by RBSDV infection, indicating that the interaction between GAPDH and ATG3 alone is not required to negatively regulate autophagy in insect cells as it is in N. benthamiana cells [14]. In addition, RBSDV P10 induces autophagy in Sf9 and L. striatellus cells by promoting AMPK and GAPDH phosphorylation and subsequent GAPDH translocation to the nucleus. Thus, the mechanism of RBSDV P10-induced autophagy in L. striatellus cells is entirely different from that of CLCuMuB ßC1-induced autophagy in N. benthamiana cells.

A previous study showed that GAPDH is a key enzyme involved in autophagy signal transduction, and that nuclear translocation of GAPDH is necessary for the induction of autophagy in mammalian cells [5]. GAPDH is also a regulator of AMPK-driven SIRT1 activation in mice [5]. Under glucose starvation conditions, cytoplasmic GAPDH is phosphorylated at Ser122 by the phosphorylated AMPK, leading to its translocation into the nucleus [5]. Inside the nucleus, phosphorylated GAPDH activates SIRT1 through direct interaction to cause autophagy [5]. In this study, we also found that GAPDH could be phosphorylated by the phosphorylated AMPK and subsequently translocated into the nucleus to activate autophagy in RBSDV P10-expressing Sf9 insect cells (Figure 11(B, E)), and that Ser95 is the key phosphorylation site of GAPDH (Figure 11(C, D)). Furthermore, the phosphorylation level of AMPK in RBSDV P10-expressing Sf9 cells and viruliferous L. striatellus cells was significantly higher compared with that in GFP-expressing Sf9 cells and non-viruliferous L. striatellus cells, resulting in GAPDH phosphorylation and subsequent nuclear translocation (Figures 12 ,13). Our assays also demonstrated that phosphorylated AMPK can interact with GAPDH to induce GAPDH phosphorylation in Sf9 cells expressing RBSDV P10 (Figure 12). Silencing of AMPK expression in Sf9 cells can affect phosphorylation and nuclear translocation of GAPDH and autophagy activation caused by RBSDV P10 (Figure 14). Therefore, we conclude that RBSDV P10 is a key regulator of autophagy in L. striatellus midgut cells, which can in turn suppress RBSDV accumulation in L. striatellus. Based on our results, a working model of RBSDV P10-induced autophagy in L. striatellus cells is proposed (Figure 15)

Figure 15.

A working model of RBSDV P10-induced autophagy in L. striatellus cells. RBSDV infection and RBSDV P10 alone promote AMPK phosphorylation, causing phosphorylation of the AMPK-interacting protein, LsGAPDH, in L. striatellus cells. The phosphorylated GAPDH relocates from the cytoplasm to the nucleus to activate the downstream autophagy pathway. The induced autophagy in L. striatellus cells can inhibit RBSDV invasion and replication.

In summary, our study has demonstrated that RBSDV P10 can promote LsAMPK phosphorylation, which leads to LsGAPDH phosphorylation, and that the phosphorylated LsGAPDH is translocated from the cytoplasm to the nucleus to activate the autophagy pathway in L. striatellus. Furthermore, autophagy can suppress RBSDV invasion and accumulation in L. striatellus midgut cells. These findings illuminate the mechanism by which RBSDV induces autophagy in L. striatellus, and indicate that the autophagy pathway in an insect vector participates in the anti-RBSDV innate immune response.

Materials and methods

Sources of virus, test plants, and the insect vector

Seeds of rice cv. Huainan No. 5 and L. striatellus Fallén (small brown planthopper) were obtained from Institute of Plant Protection, Jiangsu Academy of Agricultural Sciences, Nanjing, China. Rice seedlings were grown inside an insect-free greenhouse set at 25°C and a 16 h light and 8 h dark photoperiod. Ten-day-old rice seedlings were used for further assays. The RBSDV Kaifeng isolate was from a previously reported source and maintained in rice plants in the laboratory [32]. L. striatellus were propagated and maintained on healthy rice seedlings inside an insect culture room set at 25°C and a 16 h light and 8 h dark photoperiod. To establish an RBSDV-viruliferous L. striatellus population, 100 first-instar nymphs of L. striatellus were allowed to feed on RBSDV-infected rice plants for 2 d and then transferred onto healthy rice seedlings. The healthy rice seedlings were replaced once every 3 d. Twelve days later, L. striatellus were tested for RBSDV infection by RT-PCR and/or dot-ELISA. The confirmed viruliferous L. striatellus insects were then used to transmit RBSDV to new rice seedlings as previously described [33].

Western blot assay

For western blot assays, total protein was extracted from various samples in cell lysis buffer (Beyotime Technology, P0013) and then separated by electrophoresis in 12% or 15% sodium dodecyl sulfate-polyacrylamide (SDS-PAGE) gels. After transferring proteins onto polyvinylidene difluoride membranes (GE healthcare, 10600021), the membranes were air dried, blocked with 0.01 M phosphate-buffered saline (PBS, 140 mM NaCl, 2.7 mM KCl, 10 mM KH₂PO₄, 10 mM Na₂HPO₄, 1.8 mM KH₂PO₄, pH 7.4), containing 5% dried skimmed milk (Sangon Biotech, A600669) for 1 h, and then incubated in a diluted primary antibody solution for 1 h at 37°C. After three times rinses in PBS with 0.05% Tween-20 (Sangon Biotech, A100777; PBST), the membranes were incubated for 1 h in the secondary antibody solution. The detection signal was then visualized using the Immobilon Western Chemiluminescent HRP Substrate as instructed (Millipore, WBKLS0100). The monoclonal antibody against RBSDV P10 was produced in the author’s laboratory [32], the anti-TUBB antibody was purchased from Proteintech (10068-1-AP), and the anti-GABARAP antibody and the anti-GAPDH antibody were purchased from Abcam (ab109364) and Beyotime Technology (AF0006), respectively.

Transmission electron microscopy

To investigate the morphological changes of L. striatellus midgut cells, we collected organs from RBSDV-viruliferous or non-viruliferous L. striatellus and fixed them in 0.1 M phosphate buffer (PB) with 2.5% glutaraldehyde, pH 7.0, for 4 h at room temperature (RT) and then in PB with 1% OsO₄ for 2 h at RT. After a series of ethanol dehydration steps, the midguts were transferred to acetone for 20 min and then embedded in Spurr resin (SPI Supplies, 90529–77-4). The embedded samples were sectioned, stained with a uranyl acetate solution, and then with an alkaline lead citrate solution followed by examination under a Hitachi H-7650 TEM. For Sf9 cells (Gibco, 12659017), the cells were grown as a monolayer on coverslips and then transfected with various recombinant bacmids using the Cellfectin™ II Reagent (Gibco, 10362100) as instructed. The transfected cells were allowed to grow for another 72 h, fixed, and then embedded in Spurr resin as described above.

Immunofluorescence microscopy

Sf9 cells were transfected with various recombinant bacmids or grown on nutrient medium or glucose starvation medium (20 mM HEPES [Sangon Biotech, E607018], pH 6.4, 140 mM NaCl, 1 mM CaCl₂, 1 mM MgCl₂ and 1% BSA [Sangon Biotech, A602440]). Then Sf9 cells or L. striatellus midguts were collected, fixed in PB with 4% paraformaldehyde for 2 h at RT, and then permeabilized in PBS containing 0.2% Triton X-100 (Sangon Biotech, A110694) for 10 min (for Sf9 cells) or in PBS with 2% Triton X-100 for 30 min (for midguts). The Sf9 cells and midguts were probed with a murine anti-GAPDH antibody (Beyotime Biotechnology, AF0006), followed by an anti-mouse IgG antibody conjugated to Alexa Fluor 555 (Sangon Biotech, D110102). Accumulation of RBSDV P10 protein in the Sf9 cells or in the midgut cells was detected using an anti-RBSDV P10 monoclonal antibody conjugated to FITC as described previously [5]. Nuclei in the Sf9 cells or in the midgut cells were stained with DAPI (Beyotime Biotechnology, C1006) as instructed. The samples were then examined and imaged under an inverted Zeiss LSM780 confocal laser-scanning microscope (Carl Zeiss AG, Oberkochen, Germany).

Production of double-stranded RNAs (dsRNAs)

To produce dsRNAs, five PCR products representing partial sequences of LsGAPDH (514 bp), LsATG3 (490 bp), LsATG5 (579 bp), LsATG8 (296 bp), AMPKα (520 bp) and GFP (414 bp) were generated. These five PCR products all contained a T7 RNA polymerase promoter sequence (TAATACGACTCACTATAGGG) at the 5ʹ terminus. DsRNAs were in vitro transcribed from individual PCR products using the MEGAscript T7 Transcription Kit as instructed (Invitrogen, AM1333). Quality and size of the dsRNAs were checked by electrophoresis in 1% agarose gels (BioFroxx, 1110GR100).

Production of recombinant proteins

Sf9 cells were transfected with recombinant baculovirus vectors (deposited in the authors’ laboratory) expressing RBSDV P6, RBSDV P10, or GFP as described previously [34]. Full-length RBSDV ORFs, and GAPDH were PCR-amplified from RBSDV-infected rice cDNA, Sf9 cell cDNA, and the pGFP plasmid (Clontech, 632370), respectively, and inserted individually into a pFastBac donor vector (Invitrogen, 10584027) to produce pFastBac-P1, pFastBac-P2, pFastBac-P3, pFastBac-P4, pFastBac-P5-1, pFastBac-P5-2, pFastBac-P7-1, pFastBac-P7-2, pFastBac-P8, pFastBac-P9-1, pFastBac-P9-2, pFastBac-GAPDH^S95A, and pFastBac-GAPDH^S95D. The recombinant donor vectors were individually transformed into Escherichia coli DH10Bac cells (Invitrogen, 10361012) to produce recombinant bacmid vectors. Sf9 cells were transfected with individual recombinant bacmid vectors in the presence of Cellfectin™ II Reagent as instructed. The transfected cells were grown inside a 27°C incubator for 72 h, and the expression of individual recombinant proteins in the transfected cells was determined through examination under a Zeiss LSM780 confocal laser-scanning microscope or through western blot assays.

Yeast two-hybrid assay (Y2H)

The L. striatellus cDNA library for Y2H was previously constructed by the author’s laboratory [34]. The cDNA library was screened using RBSDV P10 protein in vector pGBKT7 (Clontech, 630443) as the bait. Yeast transformation and screening were done as described in the manufacturer’s instructions (Clontech, PT3024-1). Y2H assays were performed using the Matchmaker GAL4 Two-hybrid System as instructed (Clontech, PT3024-1). The full-length RBSDV P10 and LsGAPDH genes were RT-PCR amplified and cloned into the bait vector pGADT7 (Clontech, 630442) and the prey vector pGBKT7 to produce pGADT7-P10 and pGBKT7-LsGAPDH, respectively. These two expression vectors were co-transformed into Saccharomyces cerevisiae Gold strain cells (Clontech, 630498) using the small-scale lithium acetate method as instructed. Yeast cells co-transformed with pGADT7-T and pGBKT7-p53, pGADT7-T and pGBKT7-Lam, pGADT7-P10 and pGBKT7, or pGADT7 and pGBKT7-LsGAPDH were used as controls. The transformed yeast cells were grown on a minimal selection medium (Clontech, 630412) without adenine, histidine, leucine, and tryptophan (SD -Ade -His -Leu -Trp) at 30°C for 72 h.

Affinity-isolation assay

Affinity-isolation assays were performed as described previously [34]. Briefly, LsGAPDH-MBP, RBSDV P10-GST, LsAMPK-His and LsATG3-His were individually expressed in BL21 (DE3) cells. The expressed proteins were purified using Glutathione Sepharose® 4 Fast Flow (GE Healthcare, 17513201) or Ni-NTA His*bind resin as instructed (Millipore, 69670). Two proteins (e.g., LsGAPDH-MBP and RBSDV P10-GST, LsAMPK-His and RBSD P10-GST or LsATG3-His and LsGAPDH-MBP) were mixed (1 mg each) in 1 mL buffer A (50 mM Tris-HCl, pH 7.5, 100 mM NaCl, 0.1 mM EDTA, 0.1 mM EGTA, 0.2% Triton X-100, 0.1% beta-mercaptoethanol [Sigma-Aldrich, 07604], and 1 mM PMSF [Sangon Biotech, A610425]) supplemented with protease inhibitor cocktail (Sigma-Aldrich, S8830), and incubated at 4°C for 3 h. The resins were rinsed three times in ice-cold buffer B (50 mM HEPES, pH 7.5, 100 mM NaCl, 0.1 mM EDTA, and 1 mM PMSF) supplemented with protease inhibitor cocktail at 4°C. The rinsed resins were resuspended in 2x SDS-PAGE loading buffer, boiled for 10 min, and the proteins were separated in SDS-PAGE gels followed by western blot assays using specific antibodies.

Co-immunoprecipitation (co-IP) assay

For co-IP assays, total protein was extracted from the transformed Sf9 cells in ice-cold immunoprecipitation buffer (25 mM Tris-HCl, pH 7.5, 10% glycerol [Hushi, 10010618], 0.15% nonidet P-40 [Sangon Biotech, A600385], 150 mM NaCl) supplemented with protease inhibitor cocktail. The resulting protein extracts were incubated with a GFP-specific antibody (Abcam, ab32146) or GAPDH-specific antibody for 2 h at 4°C, and then incubated with PureProteome^TM Protein A/G Mix Magnetic Beads (Millipore, LSKMAGAG10) for 2 h at 4°C. After incubation, the beads were rinsed three times in ice-cold immunoprecipitation buffer at 4°C and then the co-IP proteins were analyzed by western blot assays.

Quantification of gene expression and viral RNA by RT-qPCR

Total RNA was isolated from L. striatellus or Sf9 cells using TRIzol reagent (Invitrogen, 15596026) according to the manufacturer’s protocol. First strand cDNA was synthesized from total RNA using ReverTra Ace® qPCR RT Master Mix with gDNA Remover (TOYOBO, FSQ-301) following the recommended protocol. RT-qPCR was performed with gene-specific primers for LsATG3, LsATG5, LsATG8, LsGAPDH, SfGAPDH, Sf9Actin, and LsActin (primers named qPCR F/R in Table S1) using the ChamQ SYBR Color qPCR Master Mix (Vazyme, Q441-02) according to the manufacturer’s protocol. All RT-qPCR experiments were done in triplicate using three independent samples as described previously [35].

Prokaryotic expression of RBSDV P10, GAPDH, and ATG3

The full-length RBSDV P10 ORF was RT-PCR amplified from total RNA isolated from an RBSDV-infected rice plant and cloned into the pGEX-4 T-3 vector (GE Healthcare, 27–4583-01) to produce pGEX-4 T-3-P10. The expression vector was transformed into E. coli BL21 (DE3) cells (Stratagene, 200131) and then induced in a 0.5 mM isopropyl β-D-thiogalactoside (Sigma-Aldrich, I6758) solution for 16 h at 16°C. The bacterial cells were pelleted by centrifugation and then disrupted by sonication. Supernatants of individual bacterial lysates were collected, incubated with the Glutathione Sepharose® 4 Fast Flow for 2 h, and then rinsed three times in ice-cold PBS. Proteins were eluted from the beads using GST elution buffer (50 mM Tris-HCl, pH 8.0, 10 mM glutathione reduced [Sigma-Aldrich, G4251]). The full-length GAPDH and ATG3 genes were RT-PCR amplified from total RNA isolated from L. striatellus and Sf9 cells, respectively, and cloned separately into the pMal-C2G vector (New England BioLabs, E8201S) or the pET-28a vector (Novagen, 69864). The expression vectors were transformed into E. coli BL21 (DE3) cells and protein expression was induced as described above. The GAPDH-MBP fusion protein was purified using amylose resin as instructed (New England BioLabs, E8021S) and the ATG3-His fusion protein was purified using Ni-NAT His*bind resin as instructed (Novagen, 70666).

Feeding L. striatellus with different artificial diets

To deliver dsRNA or purified proteins to L. striatellus, L. striatellus were reared on artificial diet D-97 as described previously [36]. Briefly, 10 μg dsRNA or 20 μg purified proteins were added to 100 μL of artificial diet, and the mixed diet was placed between two layers of stretched Parafilm membranes secured at one open end of a glass cylinder. After placing L. striatellus into the cylinder, the other open end of the cylinder was covered with a nylon mesh. The cylinder was then covered with a wet black cotton cloth, except for end of the cylinder with the artificial diet. The mixed artificial diet was replaced twice a day. After 2 d of feeding, the insects were collected and used for subsequent analysis.

GAPDH and AMPK phosphorylation test

Total protein was extracted from Sf9 cells in cell lysis buffer as described above and then used for the test. Four microliters of anti-GAPDH antibody was added to each cell lysate (200 μL) and the samples were incubated at 4°C for 2 h. After incubation, protein A/G-agarose beads were added into each sample and then incubated again for 2 h at 4°C. The agarose beads were washed three times in lysis buffer, and the phosphorylation state of GAPDH was determined by western blot assay using an anti-phosphor-serine/threonine antibody. Total GAPDH protein in the samples was measured using a murine anti-GAPDH antibody and used to control for sample loading. For the AMPK phosphorylation assay, total protein was extracted from Sf9 cells or L. striatellus, and the phosphorylation level of AMPK was determined by western blot assay using an anti-PRKAA1/PRKAA2 (phospho-Thr183/Thr172) rabbit polyclonal antibody (Sangon Biotech, D151212). Total AMPK protein in the samples was detected using an anti-PRKAA1 rabbit polyclonal antibody (Sangon Biotech, D161273).

In vitro phosphorylation assay

For the phosphorylation of GAPDH assay, the phosphorylated AMPK was obtained by immunoprecipitation using anti-AMPK antibody (Sangon Biotech, D191051) from RBSDV P10-expressing Sf9 cells, then incubated in the presence or absence of purified prokaryotic expression GAPDH with 20 mM Tris-HCl, pH 7.5, 10 mM MgCl₂, 1 mM Na₃VO₄, 1 mM DTT, 20 mM ATP for 30 min at 30°C. The phosphorylation of the GAPDH was determined by western blot analysis using anti-phospho-(Ser/Thr) antibody.

Bio-layer interferometry (BLI) analysis

A Gator instrument (Shanghai, China) was used for bio-layer interferometry to analyze the interaction between AMPK and RBSDV P10 according to the Gator’s protocol. Prokaryotic expression His-AMPK, GST-P10, GST tag were purified as described above. His-AMPK was immobilized to the surface of anti-His biosensors (Gator). After washing and baseline step (150 s) with PBS (pH 7.4), biosensor tips were immersed into the wells containing 20 μg/mL GST-P10 or GST tag for the associate step (200 s). A dissociation step was then performed (100 s). All steps were performed at 30°C with 1000 rpm vibration.

Quantification and statistical analysis

ImageJ (https://imagej.nih.gov/ij/) was applied for the quantitation of western blot assay results. Results of RT-qPCRs are expressed as mean ± SD (standard deviation) of three independent experiments. Statistical analysis was performed using the Student'st-test in GraphPad Prism 8, and *P < 0.05 and **P < 0.01 were considered significant.

Funding Statement

This work was supported by the National Natural Science Foundation of China (No. 31972234; 31772125), the National Key Research and Development Project of China (No. 2016YFD0300706; 2017YFD0201604; YS2017YFGH000548), and the Earmarked Fund for Modern Agro-industry Technology Research System (nycytx-001).

Disclosure statement

The author(s) declare that they have no competing interests.

Supplementary material

Supplemental data for this article can be accessed here