(C) PLOS One [1]. This unaltered content originally appeared in journals.plosone.org.
Licensed under Creative Commons Attribution (CC BY) license.
url:https://journals.plos.org/plosone/s/licenses-and-copyright

------------



Turnip mosaic virus co-opts the vacuolar sorting receptor VSR4 to promote viral genome replication in plants by targeting viral replication vesicles to the endosome

['Guanwei Wu', 'State Key Laboratory For Managing Biotic', 'Chemical Threats To The Quality', 'Safety Of Agro-Products', 'Institute Of Plant Virology', 'Ningbo University', 'Ningbo', 'Zhaoxing Jia', 'College Of Plant Protection', 'Nanjing Agricultural University']

Date: 2022-02 

Accumulated experimental evidence has shown that viruses recruit the host intracellular machinery to establish infection. It has recently been shown that the potyvirus Turnip mosaic virus (TuMV) transits through the late endosome (LE) for viral genome replication, but it is still largely unknown how the viral replication vesicles labelled by the TuMV membrane protein 6K2 target LE. To further understand the underlying mechanism, we studied the involvement of the vacuolar sorting receptor (VSR) family proteins from Arabidopsis in this process. We now report the identification of VSR4 as a new host factor required for TuMV infection. VSR4 interacted specifically with TuMV 6K2 and was required for targeting of 6K2 to enlarged LE. Following overexpression of VSR4 or its recycling-defective mutant that accumulates in the early endosome (EE), 6K2 did not employ the conventional VSR-mediated EE to LE pathway, but targeted enlarged LE directly from cis-Golgi and viral replication was enhanced. In addition, VSR4 can be N-glycosylated and this is required for its stability and for monitoring 6K2 trafficking to enlarged LE. A non-glycosylated VSR4 mutant enhanced the dissociation of 6K2 from cis-Golgi, leading to the formation of punctate bodies that targeted enlarged LE and to more robust viral replication than with glycosylated VSR4. Finally, TuMV hijacks N-glycosylated VSR4 and protects VSR4 from degradation via the autophagy pathway to assist infection. Taken together, our results have identified a host factor VSR4 required for viral replication vesicles to target endosomes for optimal viral infection and shed new light on the role of N-glycosylation of a host factor in regulating viral infection.

A key feature of the replication of positive-strand RNA viruses is the rearrangement of the host endomembrane system to produce a membranous replication organelle. Recent reports suggest that the late endosome (LE) serves as a replication site for the potyvirus Turnip mosaic virus (TuMV), but the mechanism(s) by which TuMV replication vesicles target LE are far from being fully elucidated. Identification of the host factors involved in this transport process could lead to new strategies to combat TuMV infection. In this report, we provide evidence that TuMV replication depends on functional vesicle transport from cis-Golgi to the enlarged LE pathway that is mediated by a specific VSR family member, VSR4, from Arabidopsis. Knock out of VSR4 impaired the targeting of TuMV replication vesicles to enlarged LE and suppressed viral infection, and this process depends on the specific interaction between VSR4 and the viral replication vesicle-forming protein 6K2. We also showed that N-glycosylation of VSR4 modulates the targeting of TuMV replication vesicles to enlarged LE and enhances viral infection, thus contributing to our understanding of how TuMV manipulates host factors in order to establish optimal infection. These results may have implications for the role of VSR in other positive-strand RNA viruses.

Funding: This work was supported by grants from the National Natural Science Foundation of China (32070165) to G.W, Chinese Agriculture Research System (CARS-24-C-04) to F.Y., the Ningbo Major Special Projects of the Plan "Science and Technology Innovation 2025" (2021Z106) to H.Z, and sponsored by the K.C. Wong Magna Fund in Ningbo University. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

In this study, we demonstrate that the Arabidopsis thaliana VSR4 (AtVSR4) interacts with TuMV 6K2 specifically and is an essential proviral host factor for TuMV infection. Moreover, we reveal that AtVSR4 can be asparagine (N)-glycosylated and monitor 6K2 traffic to PVC/MVB/LE. We further explore the role played by AtVSR4 and its N-glycosylation in TuMV infection.

Vacuolar sorting receptors (VSRs) are arguably the most studied vacuolar trafficking factors in plants, but it is still unclear how they function in the plant endomembrane system [ 32 – 36 ]. The genome of Arabidopsis thaliana contains seven VSR isoforms (AtVSR1-7) with highly conserved amino acid sequences especially in their C-terminal domains [ 37 ]. All seven AtVSRs localize to PVC/MVB/LE [ 38 ]. AtVSRs are expressed in most Arabidopsis tissue types, including root, leaf, stem, flower, pollen, and seed [ 39 ], but the different AtVSRs are not equally expressed in each of these tissues, suggesting that they might have distinct functions. For example, AtVSR1 and 4 are predominantly expressed in leaves, while AtVSR5, 6, and 7 are only expressed in roots, and AtVSR3 is expressed specifically in guard cells [ 33 , 39 ]. VSRs have an N-terminus luminal domain, a single transmembrane domain (TMD) and a C-terminus cytoplasmic tail (CT). The CT carries the sorting information for their own transportation and harbours a YXXΦ motif ( 610 YMPL 613 ) and an acidic dileucine-like motif ( 602 EIRAIM 607 ). The YXXΦ motif was interpreted as an anterograde TGN to PVC signal, and an AMPA mutant was shown to accumulate in TGN [ 40 , 41 ]. In contrast, the acidic dileucine-like motif was suggested to act as a retrograde sorting motif from PVC to TGN, but also as an endocytosis signal, and AIRAAA mutants led to strong labelling of the vacuole [ 41 , 42 ]. According to a frequently cited model, the cargo-VSR complex is sequestered into clathrin coated vesicles at the TGN and delivered to the PVC. However, evidence is accumulating showing that VSR-cargo interactions occur much earlier in the endomembrane system. According to textbooks and most recent reviews, VSRs recognize their ligands in the ER [ 43 ]. The receptor-ligand complexes are then transported in COPII-coated vesicles to Golgi stacks, and eventually reach the TGN/EE via cisternal progression [ 34 ]. The ligands then dissociate from the VSRs, which are recycled back to cis-Golgi cisternae mediated by the retromer complex [ 34 , 44 , 45 ]. Interestingly, the VSR CT associates with a PVC/MVB/LE SNARE protein VTI11 (VESICLE TRANSPORT V-SNARE11) [ 46 ], which also copurifies with TuMV 6K2 and is required for 6K2 vesicle trafficking to PVC/MVB/LE [ 10 ]. Thus, we questioned whether VSR participates in 6K2 targeting of PVC/MVB/LE during TuMV infection.

The genus Potyvirus is the largest genus of known plant RNA viruses, comprising more than 180 species, constituting about 15% of all identified plant viruses [ 24 , 25 ]. Potyviruses include some of the most economically important plant pathogens, including TuMV, Soybean mosaic virus, Potato virus Y, and Plum pox virus [ 26 ]. Potyviruses have a single-stranded positive-sense (+ss) RNA genome of approximately 10, 000 nucleotides that encodes a long open reading frame (ORF) and a small ORF P3N-PIPO resulting from RNA polymerase slippage [ 24 , 27 ]. The large ORF is proteolytically processed by three viral proteinases into 10 mature proteins, one of which is the membrane-associated 6K2 protein. 6K2 can remodel the host ER for the formation of viral replication vesicles [ 28 , 29 ]. These vesicles contain viral RNA and viral proteins as well as host components [ 18 , 29 ]. These vesicles may mature into replication-competent single membrane vesicles, which can fuse with chloroplasts for efficient replication [ 28 , 30 ]. These vesicles are believed to take a Golgi by-pass unconventional pathway and reach PVC/MVB/LE for efficient virus infection [ 10 ]. It has not been established how 6K2 vesicles reach PVC/MVB/LE but they cannot utilize the post-Golgi trafficking pathway [ 8 , 31 ].

Eukaryotic cells are compartmentalized by various membrane-bound organelles that form a complex endomembrane network to perform diverse fundamental functions critical for cell survival. This dynamic endomembrane system is maintained through the continuous flux of vesicles to exchange lipids and proteins. Plant cells all have an endomembrane system, including plasma membrane (PM), nuclear envelope, the endoplasmic reticulum (ER), cis-Golgi apparatus, trans-Golgi network or early endosome (TGN/EE), prevacuolar compartment/multi-vesicular body or late endosome (PVC/MVB/LE) and vacuole. The secretory and endocytic pathways are two major transport routes of the plant endomembrane system [ 1 ]. These two pathways merge in the TGN/EE and their cargoes are passed on to the PVC/MVB/LE by different sorting machineries. PVC/MVB/LE serve as intermediate compartments between the TGN/EE and the vacuole, and enable proteins to recycle before fusion with the vacuole [ 2 ]. Plant viruses are obligate parasites with genomes that encode very few proteins and they exploit the host intracellular machinery to establish infection [ 3 – 5 ]. In the past decade, an essential role for the endomembrane system (including related host factors) in plant virus infection has been recognised [ 6 – 20 ]. In recent reports, PVC/MVB/LE were observed to be labelled with double stranded RNA (dsRNA) during infection by Turnip mosaic virus (TuMV; genus Potyvirus) [ 9 , 10 ], suggesting that TuMV replication vesicles transit through PVC/MVB/LE. The endosomal sorting complexes required for transport (ESCRT) family proteins, normally required for PVC/MVB/LE formation, are also hijacked by Brome mosaic virus (BMV, genus Bromovirus) and Tomato bushy stunt virus (TBSV, genus Tombusvirus) to facilitate VRC assembly [ 5 , 17 , 21 – 23 ], indicating an important role of PVC/MVB/LE during plant virus infection.

Results

AtVSR4 is required for TuMV infection To verify whether VSR proteins are involved in TuMV infection, we obtained homozygous knockout (ko) transfer DNA lines for each of five AtVSR homologs (AtVSR1, AtVSR2, AtVSR4, AtVSR5 and AtVSR7) and used them in viral infection assays (S1 Fig). All these atvsr mutants had similar growth and developmental phenotypes to the wild-type Col-0 plants. Homozygous atvsr ko and Col-0 plants were mechanically inoculated with TuMV-GFP and then monitored for green fluorescence using a hand-held UV lamp. TuMV-GFP fluorescence was first seen in the inoculated Col-0 and mutants’ leaves at 4 days post inoculation (dpi). We recorded the GFP fluorescence in all plants at 7 dpi. Plants of the atvsr4 line had much milder symptoms and weaker GFP fluorescence than the Col-0 and other atvsr plants (Fig 1A). Immunoblots revealed substantially reduced levels of TuMV coat protein (CP) at 7 dpi in atvsr4 than in Col-0 and the other atvsr mutants (Fig 1B). At both 7 dpi and 14 dpi, TuMV RNA levels in atvsr4 were significantly lower that those in Col-0 plants (Fig 1C), and at 16 dpi the symptoms in atvsr4 remained much milder than those in Col-0 plants (Fig 1D). These combined results indicate that AtVSR4 is required for TuMV infection. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 1. AtVSR4 is required for TuMV infection. (A) Representative images of TuMV-GFP infected wild-type Col-0 and atvsr mutant plants at 7 dpi under hand-held UV lamp. Scale bar = 1.5 cm. (B) Immunoblots detecting TuMV CP in the systemic leaves at 7 dpi. The relative TuMV CP signals were quantified by ImageJ software. Coomassie Brilliant Blue (CBB) R-250-stained RuBisco large subunit serves as a loading control. TuMV CP was detected with anti-TuMV CP polyclonal antibody. (C) qRT-PCR analysis of the TuMV RNA levels in the wild-type Col-0 and atvsr4 mutant plants at 7 and 14 dpi. Actin II was used as an internal control. Error bars represent the standard deviation of three biological replicates of a representative experiment. Statistical analysis was performed using Student’s t-test (**, P < 0.01). (D) Phenotypes of the wild-type Col-0 and atvsr4 plants inoculated with TuMV and the mock infection controls at 16 dpi. Scale bar = 5.0 cm. https://doi.org/10.1371/journal.ppat.1010257.g001

AtVSR4 colocalizes with the viral 6K2-induced viral replication complex (VRC) and is required for the targeting of 6K2 to enlarged LE We then questioned whether VSR4 colocalized with 6K2 in planta. Irrespective of the particular fluorescent protein fused to its N or C terminus, AtVSR4 formed many granules close to the PM (S4A Fig). Many of these granules were tightly associated with, or inside, the endosomes highlighted by FM4-64 (Fig 3A), and rarely colocalized with cis-Golgi (S4B Fig, upper panels). Moreover, AtVSR4 also formed ER-like reticular pattern structures (S4A Fig) and merged well with ER marker mCherry-HDEL (S4B Fig, lower panels). We further examined the colocalizations of AtVSR4 and 6K2 in plant cells. AtVSR4-labelled punctate bodies colocalized with the granules formed by 6K2, and these colocalization signals were tightly associated with or within FM4-64 labelled enlarged LE (Fig 3B). Thus, the distribution of AtVSR4 was dynamic and the AtVSR4-6K2 complex colocalized at enlarged LE. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. AtVSR4 is recruited into the viral replication complex induced by TuMV 6K2 and required for enlarged PVC targeting of 6K2. (A) Colocalization assay of YFP-AtVSR4 with endosomal compartments labelled by FM4-64 in N. benthamiana cells at 2 dpai. Scale bar = 20 μm. (B) Colocalization assay of YFP-AtVSR4 with 6K2-CFP and endosomal compartments labelled by FM4-64 in N. benthamiana cells at 2 dpai. White arrows indicate that AtVSR4-labelled punctate bodies colocalized with the granules formed by 6K2, and these colocalization signals were tightly associated with, or within, enlarged LE stained by FM4-64. (C) AtVSR4 colocalized with 6K2-induced vesicles in TuMV-infected N. benthamiana cells. Localizations of GFP-tagged AtVSR4 in the leaf cells inoculated with TuMV/6K2mCherry at 48 or 72 h post agroinfiltration are shown. (D) AtVSR4 colocalized with TuMV viral replication complex in the dsRNA reporter N. benthamiana B2-GFP cells. Images show the localization of GFP-tagged AtVSR4 in leaf cells inoculated with TuMV/6K2gus at 72 h post agroinfiltration. Representative colocalization signals are indicated with white arrows. Scale bar = 20 μm. Chl, auto-fluorescent chloroplasts. BF, bright field. (E) Knockout of AtVSR4 interferes with the trafficking of TuMV 6K2 to enlarged LE. Colocalization of TuMV 6K2-GFP with FM4-64 or enlarged PVC marker mCherry-AtARA7Q69L in Arabidopsis protoplasts from wild-type Col-0 and atvsr4 plants are shown. Representative colocalizations are indicated with white arrows. Images were taken at 20 h post transfection. Scale bar represents 10 μm. (F) Quantification of colocalization between 6K2 and FM4-64 or enlarged LE marker in Col-0 or atvsr4 protoplasts by calculation of the Pearson’s correlation coefficient (PCC) values. PCC was measured from 20 protoplasts. Mean values ± standard deviation from three independent experiments are shown. Statistical analysis was performed using Student’s t-test (**, P < 0.01; ***, P < 0.001). https://doi.org/10.1371/journal.ppat.1010257.g003 To investigate whether AtVSR4 is recruited by the 6K2-induced viral replication complex (VRCs) for TuMV infection, we transiently expressed AtVSR4-GFP in N. benthamiana leaf tissues co-agroinfiltrated with a modified recombinant TuMV infectious clone TuMV::6K2-mCherry, where an extra 6K2-mCherry fusion is inserted between P1 and HC-Pro. It is known that potyvirus replication takes place in the 6K2-induced vesicle structures [28,29]. We found that AtVSR4-induced punctate bodies colocalized with the 6K2-induced membranous structures during TuMV infection at two examined time points (Fig 3C). To confirm that AtVSR4 indeed targeted 6K2-containing VRCs, we further expressed RFP-AtVSR4 transiently in the B2:GFP double stranded RNA (dsRNA) reporter N. benthamiana leaf tissues [50], infected with TuMV. DsRNA signals resulting from VRCs containing 6K2 clearly colocalized with RFP-tagged AtVSR4 (Fig 3D). RFP-AtVSR4 did not colocalize with dsRNA signals in the absence of viral infection (S4C Fig). These results suggest that AtVSR4 is recruited into the TuMV VRC. Finally, to explore whether AtVSR4 is required for enlarged LE targeting of 6K2, we investigated the subcellular distribution of 6K2 in the atvsr4 mutant. In wild-type Arabidopsis Col-0 protoplasts, some FM4-64 stained enlarged LE were labelled by TuMV 6K2 but in atvsr4 protoplasts, almost no enlarged LE colocalized with TuMV 6K2 (Fig 3E). We further used an enlarged LE marker mCherry-ARA7Q69L [51], to colocalize with 6K2 in Col-0 or atvsr4 protoplasts, which gave similar results. The Pearson’s correlation coefficient (PCC) values [52], which provide a quantitative estimate of colocalization, were calculated and confirmed these observations (Fig 3F). These results suggest that AtVSR4 is required for the targeting of TuMV 6K2 to enlarged LE.

AtVSR4 facilitates TuMV 6K2 to target enlarged LE for viral replication Recent studies have shown that VRCs of TuMV induced by 6K2 were associated with PVC/MVB/LE and that this required a host factor VTI11 [9,10]. Since VSR can also target endosomes and is associated with VTI11, we hypothesised that 6K2 may utilize the VSR-mediated trafficking pathway. We used a dominant negative inhibition approach to analyze VSR traffic during TuMV infection. As described in the introduction, the VSR4 CT harbors YXXΦ (610YMPL613) and dileucine-like (602EIRAIM607) motifs for VSR transportation [33] (Fig 4A). When the YXXΦ motif is mutated to AMPA, the anterograde TGN to PVC signal is disrupted, and this mutant mainly accumulates in TGN [40]. Mutation of the dileucine-like motif to AIRAAA interferes with the PVC to TGN signal, leading to strong labelling of this mutant in the vacuole [42]. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. Effects of mutations in the YXXΦ or acidic dileucine-like motif of AtVSR4 on TuMV infection. (A) Diagram showing the mutated amino acids of two conserved motifs in the CT region of AtVSR4 used in this study. The complete amino acid sequence of VSR4 CT is shown. The YXXΦ motif YMPL and acidic dileucine-like motif EIRAIM are highlighted with black rectangular boxes, and the mutated amino acids in these motifs are underlined. (B) Effect of wild type AtVSR4 or its mutants on 6K2 localization in N. benthamiana cells. YFP-tagged 6K2 was co-expressed with AtVSR4, its mutants or GUS control fused with a c-myc tag. The YFP-6K2 signals were stained with FM4-64. White arrows indicate the enrichment of YFP-6K2 in FM4-64 labelled single membrane vesicle-like structures (SMVLs). Scale bar represents 20 μm. BF, bright field. (C) Colocalization assay of AtVSR4 or its mutants with TuMV 6K2-induced vesicles in N. benthamiana cells. Localizations of GFP-tagged AtVSR4 and mutants in the leaf cells inoculated with TuMV/6K2mCherry at 72 h post agroinfiltration are shown. Chl, auto-fluorescent chloroplasts. (D) Phenotypes of the wild-type Col-0, and transgenic plants overexpressing AtVSR4 mutants in the atvsr4 background, inoculated with TuMV at 14 dpi. Scale bar = 2.0 cm. (E) qRT-PCR analysis of the TuMV RNA levels in the wild-type Col-0 and the transgenic plants overexpressing AtVSR4 mutants at 14 dpi. Actin II was used as an internal control. Error bars represent the standard deviation of three biological replicates of a representative experiment. Statistical analysis was performed using Student’s t-test (**, P < 0.01; NS, not significant). https://doi.org/10.1371/journal.ppat.1010257.g004 We transiently co-expressed AtVSR4 and the mutants AMPA, AIRAAA, or double mutant (designated as DM) with TuMV::GFP in N. benthamiana leaves and monitored the viral infection by measuring both RNA and protein accumulation. Under UV light, TuMV-GFP fluorescence was brighter in the leaf tissues co-infiltrated with TuMV infectious clone and AtVSR4 or its mutants, compared with the control (S5A Fig). A qRT-PCR assay of TuMV genomic RNA [either sense (+) or negative sense (-)] was further performed to evaluate viral replication levels at 60 h post agroinfiltration (hpai). TuMV replicates in the primarily infected cells and viral cell-to-cell movement usually does not occur until 96 hpai [53]. Consistently, significantly higher TuMV (+) or (-) RNA levels were detected in the leaf tissues agroinfiltrated with TuMV together with either VSR4 or its mutants at 60 hpai, compared with that in the control agroinfiltrated with TuMV and partial GUS (ß-glucuronidase)-myc (S5B Fig). Moreover, overexpression of AMPA and DM mutants led to significantly higher viral RNA levels than those in the VSR4 and AIRAAA mutant treatments. Further immunoblotting analysis was performed to evaluate the viral CP accumulation level at 72 hpai. Consistently, much stronger TuMV CP signals (1.5–2.5-fold increase) were detected in the leaf samples overexpressing VSR4 or its mutants compared with the control (S5C Fig), and AMPA and DM mutants also showed significantly higher viral CP accumulation levels than VSR4 and AIRAAA. These results suggest that VSR4 has a pro-viral role in supporting TuMV replication, and also that 6K2 does not hijack the canonical TGN to LE pathway mediated by VSR proteins. To verify if these VSR4 mutants affected 6K2 distribution, TuMV 6K2 (with an N-terminal yellow fluorescent protein tag) was transiently co-expressed with GUS control or the VSR4 mutants (with a C-terminal myc tag) in N. benthamiana leaves. Confocal microscopy showed that in comparison with the control or AIRAAA mutant treatment, transient overexpression of the AMPA or DM remarkably increased the amount of 6K2-induced punctate bodies in the cytoplasm (S6A and S6B Fig). Immunoblotting analysis confirmed that co-expression of the VSR4 mutants did not affect the accumulation levels of 6K2 protein (S6C Fig). Co-IP results revealed that all three VSR4 mutants can still bind with 6K2, although AIRAAA had a relatively weaker binding capacity than wild type VSR4 (S6D Fig). We further colocalized YFP-6K2 with the endocytic-tracking styryl FM4-64 or cis-Golgi marker Man49-mCherry [49]. Confocal results showed that AMPA and DM treatment increased the chance of 6K2 localizing at cis-Golgi (S7 Fig). The PCC values confirmed our observations (S7 Fig). This result is consistent because AMPA interferes with the anterograde TGN to PVC signal, and so more 6K2 was retained in cis-Golgi. Moreover, these 6K2-induced punctate bodies actually entered into enlarged PVC/MVB/LE, but not granule-like endosomes, labelled by FM4-64 (Figs 4B and S7). We also examined whether AMPA, AIRAAA and DM affect localization at TuMV 6K2-induced vesicles. Confocal results showed that several small vesicles formed by VSR4, AMPA and DM were tightly associated or colocalized with 6K2-induced vesicles, while AIRAAA had fewer small vesicles associated with 6K2 vesicles (Fig 4C). Finally, to further confirm the effect of VSR4 mutants on TuMV infection, stable transgenic lines carrying the YFP-AtVSR4 mutant fusions under control of the 35S promoter in the atvsr4 background were generated, and YFP fluorescence was observed by confocal microscopy (S8A Fig). We opted to employ the 35S promoter, because lines under the control of a 1.2-kbp fragment upstream of the AtVSR4 CDS did not lead to detectable expression. Three independent T0-positive lines of each mutant were identified by immunoblotting (S8B Fig). We selected two homozygous T2 lines for each mutant for TuMV infection assays. In comparison with the wild-type Col-0 plants, the atvsr4 plants overexpressing AMPA or DM were more susceptible to TuMV infection and developed more severe symptoms (yellowish and dwarf phenotype) at 14 dpi (Fig 4D). Consistently, high levels of viral genomic RNA were also detected in the plants overexpressing APMA and DM (Fig 4E). The atvsr4 plants overexpressing AIRAAA had similar symptoms and viral accumulation levels to the Col-0 plants. Taken together, the results suggest that VSR4 promotes TuMV replication by promoting formation of 6K2-induced punctate bodies and further targets them into enlarged PVC/MVB/LE via an unconventional pathway.

AtVSR4 is N-glycosylated in vivo Protein trafficking is known to be triggered by post-translational modifications including N-glycosylation [54,55]. AtVSR1 was previously identified as an N-glycosylated protein, and N-glycosylation modification is critical for ligand binding and vacuolar protein trafficking [56]. The NetNGlyc server (http://www.cbs.dtu.dk/services/NetNGlyc/) predicts that AtVSR4 contains three potential N-glycosylation sites at residues Asn148, Asn294 and Asn434 in the luminal domain (Fig 5A). To determine whether AtVSR4 is N-glycosylated at these predicted residues, we treated N. benthamiana leaf tissues transiently expressing AtVSR4-myc with tunicamycin, a known N-glycosylation inhibitor, followed by immunoblotting analysis. AtVSR4-myc also only harbors these three predicted N-glycosylation sites. We found that AtVSR4-myc migrated faster in SDS-PAGE extracts from leaf tissues treated with N-glycosylation inhibitor tunicamycin than in those from the DMSO-treated controls (Fig 5B), indicating that AtVSR4 indeed harbours N-linked oligosaccharides. To analyze which sites are subjected to N-glycosylation, each of the potential asparagine (N) residues of AtVSR4-myc was mutated into glutamine (Q), enabling analysis of AtVSR4 with one, two or three of the potential glycosylation sites mutated. These mutants were transiently expressed in N. benthamiana leaf cells, followed by Western blot analysis. Mutation on any one or two of the three N-glycosylation sites did not result in apparent band shift (Fig 5C). However, mutation of all three sites, the triple mutant (TM), shifted obviously faster than the wild type AtVSR4-myc (Fig 5C). To exclude the possibility that there might be further N-glycosylation sites in addition to those predicted, we further treated the leaf tissues expressing AtVSR4-myc or AtVSR4TM-myc with tunicamycin, followed by Western blot analysis. As shown in Fig 5D, AtVSR4TM-myc had the same molecular weight with or without tunicamycin treatment. AtVSR4-myc protein showed a smaller molecular weight protein band in tunicamycin-treated cells, with the same size as the AtVSR4TM-myc, indicating the absence of N-glycosylation in AtVSR4TM-myc. These results demonstrate that these three predicted N-glycosylation sites in AtVSR4 harbor N-linked glycans. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 5. AtVSR4 is N-glycosylated. (A) Diagram showing the predicted N-glycosylated sites N148, N294, N434 in AtVSR4. The triple mutant with amino acid mutations N148Q, N294, and N434Q, designated as TM, is shown. (B) AtVSR4 size migration analysis of protein extracts from N. benthamiana cells expressing AtVSR4-c-myc treated with N-glycosylation inhibitor tunicamycin or DMSO. (C) AtVSR4 size migration analysis of protein extracts from N. benthamiana cells expressing different glycosylation site mutants. (D) AtVSR4 and TM size migration analysis of protein extracts from N. benthamiana cells expressing AtVSR4-c-myc or TM treated with tunicamycin or DMSO. (E) AtVSR4 and TM size migration analysis of protein extracts from N. benthamiana cells expressing AtVSR4-c-myc or TM treated with different glycosidases. Coomassie Brilliant Blue (CBB) R-250-stained RuBisco large subunit serves as a loading control. https://doi.org/10.1371/journal.ppat.1010257.g005 To analyze the glycosylation pattern of AtVSR4, we treated AtVSR4-myc transiently expressed in N. benthamiana cells with various glycosidases, including endoglycosidase H (Endo H) and peptide-N-glycosidase F (PNGase F) [57]. Endo H cleaves high-mannose glycans, which are added to proteins in the ER. Subsequent trimming and addition reactions in the Golgi body yield complex glycans, which are resistant to Endo-H. PNGase F removes most N-linked glycans. Following either Endo H or PGNase F treatment, the AtVSR4-myc migrated to the same position as AtVSR4TM-myc (Fig 5E). Moreover, AtVSR4TM-myc protein showed no shift upon treatment with either of these enzymes (Fig 5E), reflecting the size of the fully deglycosylated protein. To exclude the possibility that AtVSR4 and AtVSR4TM may have O-glycosylation sites, we further treated AtVSR4 and AtVSR4TM-myc with O-Glycosidase & Neuraminidase Bundle, which can simultaneously remove terminal sialic acid residues and O-linked glycans. Western blot analysis showed no shift change for either AtVSR4-myc or AtVSR4TM-myc (Fig 5E), suggesting that AtVSR4 harbors no O-glycosylation sites. These results indicate that AtVSR4 can only be N-glycosylated and that it is mature at ER with high-mannose glycans.

Non-glycosylated AtVSR4 shows a stronger ability to promote TuMV infection To determine whether the removal of N-glycans from AtVSR4 would affect TuMV infection, we first tested the protein-protein interaction between non-glycosylated AtVS4 and 6K2. Both Co-IP and luciferase complementation imaging (LCI) assays showed that non-glycosylated AtVS4 had a weak interaction signal with 6K2, no doubt because there was less protein accumulation of AtVSR4TM in cells, compared with AtVSR4 (Fig 6A and 6B). These results indicate that AtVSR4TM can still interact with 6K2. Then we investigated the effect of transient overexpression of AtVSR4TM on TuMV accumulation. N. benthamiana leaves were co-infiltrated with an TuMV::GFP infectious clone and with an AtVSR4TM-myc, AtVSR4-myc or Gus-myc control expression construct. As expected, overexpression of AtVSR4 enhanced the intensity of GFP fluorescence directly resulting from TuMV infection compared with the control treatment (Fig 6C). Interestingly, co-expression of AtVSR4TM resulted in remarkably higher intensity of GFP fluorescence than with AtVSR4 (Fig 6C). Virus accumulation was assessed in infiltrated leaves at 48 and 60 hpai by qRT-PCR and Western blot, respectively. Consistently, we found a significant increase (~2.0 fold) of TuMV genomic (+) and (-) RNA in plants co-infiltrated with AtVSR4TM compared to those with AtVSR4 (Fig 6D). The Western blot assay also showed s substantial increase (~1.7 fold) in TuMV CP accumulation in AtVSR4TM compared with AtVSR4 (Fig 6E). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 6. Effect of non-glycosylated AtVSR4 on its interaction with 6K2 and TuMV infection. (A) Co-IP assay of the interaction between 6K2 and AtVSR4 or non-glycosylated AtVSR4TM. Different cell lysates were immunoprecipitated with anti-Flag M2 gel beads, separated by SDS-PAGE and immunoblotted with anti-Flag monoclonal antibody (@Flag MAb), or anti-myc monoclonal antibody (@myc MAb). White asterisks denote myc-fusion AtVSR4 or AtVSR4TM bands. (B) luciferase complementation imaging (LCI) assays of interaction between 6K2 and AtVSR4 or AtVSR4TM. C-luc-6K2 and VSR4-N-luc or TM-N-luc were transiently co-expressed in N. benthamiana. VSR1-N-luc was used as negative control. (C) GFP fluorescence in N. benthamiana plants inoculated with TuMV-GFP together with GUS (control), AtVSR4, or TM. Plants were photographed under a hand-held UV lamp at 3 dpai. (D) Results of qRT-PCR showing the quantification of positive-strand viral genomic RNA [(+)RNA] or negative-strand viral genomic RNA [(-)RNA] accumulation in N. benthamiana plants agroinfiltrated with different combinations of plasmids from (C). The infiltrated leaf tissues were collected and pooled at 60 hours post agroinfiltration (hpai) for RNA purification. The purified RNA was analyzed by qRT-PCR with TuMV nib-specific primers using the actin II transcript level as an internal control. (E) Immunoblotting analysis of the accumulated TuMV CP levels in the infiltrated leaf tissues from N. benthamiana plants in (C) at 60 hpai. The relative TuMV CP signals were quantified by Image J software. Coomassie Brilliant Blue R-250-stained RuBisco large subunit serves as a loading control. TuMV CP was detected with anti-TuMV CP polyclonal antibody. AtVSR4 and TM was detected with anti-c-Myc monoclonal antibody. (F) Phenotypes of the wild-type Col-0 and transgenic plants overexpressing AtVSR4 or TM after inoculation with TuMV under a hand-held UV lamp at 8 dpi or under regular light at 16 dpi. Scale bar = 2.0 cm. (G) qRT-PCR analysis of the TuMV RNA levels in the wild-type Col-0 and transgenic plants overexpressing AtVSR4 or TM at 16 dpi. Actin II was used as an internal control. Data represent means with SD of three independent experiments. Statistical significance was determined by Student’s t test (**, P < 0.01; ***, P < 0.001). https://doi.org/10.1371/journal.ppat.1010257.g006 To further confirm the effect of overexpression of AtVSR4TM on TuMV infection, transgenic Arabidopsis lines overexpressing the fusion proteins Flag-4×myc-AtVSR4 or Flag-4×myc-AtVSR4TM were generated. Eight AtVSR4 and seven AtVSR4TM independent overexpression (oe) lines were examined by immunoblotting to detect the fusion protein (S9 Fig). Three T2 lines of each construct (AtVSR4oe#1, #3, #4 and AtVSR4TMoe#3, #8, #9) that expressed relatively higher levels of the fusion proteins were selected for further analysis. Wild-type Col-0, AtVSR4oe, and AtVSR4TMoe Arabidopsis plants were mechanically inoculated using sap from N. benthamiana leaves infected by TuMV::GFP. At 8 dpi, both AtVSR4TMoe and AtVSR4oe showed a remarkably higher intensity of GFP fluorescence than Col-0 plants (Fig 6F). At 16 dpi, symptoms were more severe in AtVSR4TMoe than in AtVSR4oe and Col-0 plants (Fig 6F). qRT-PCR results showed that there were significantly higher levels of TuMV RNA in AtVSR4TMoe than in AtVSR4oe plants (Fig 6G). AtVSR4oe plants also had significantly higher viral RNA accumulation than Col-0 plants. Taken together, these results indicate that non-glycosylated AtVSR4 promotes TuMV infection.

Non-glycosylated AtVSR4 promotes TuMV 6K2 to form punctate bodies and target enlarged LE We next wanted to know the mechanism whereby non-glycosylated AtVSR4 promotes TuMV infection. We therefore first investigated the effect of overexpressing AtVSR4TM on 6K2 distribution in N. benthamiana leaves. As shown in Fig 7A and 7B, AtVSR4TM overexpression significantly increased the formation of 6K2 punctate bodies at 2 dpai, compared with co-expression of AtVSR4 or expression of 6K2 alone. Immunoblotting analysis confirmed that co-expression of AtVSR4 or AtVSR4TM did not affect the protein accumulation level of YFP-6K2 (Fig 7C). Moreover, the formation of 6K2 punctate bodies apparently decreased (S10 Fig), and there was little expression of AtVSR4 and AtVSR4TM at 3 dpai (Fig 7C), suggesting a tight relationship between the formation of 6K2 punctate bodies and accumulation levels of AtVSR4 or AtVSR4TM protein. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 7. Effect of non-glycosylated AtVSR4 on subcellular localization of 6K2. (A) Subcellular localization of 6K2 in N. benthamiana cells when expressed alone and when co-expressed with VSR4 or TM. Pictures were taken at 2 dpa. Scale bar represents 20 μm. (B) Number of 6K2-induced punctate bodies in the cytoplasm when YFP-6K2 was expressed alone or when VSR4-myc and TM-myc were co-expressed (10 cells per construct were investigated at 2 dpai and the number was calculated using Image J software). Values represent the mean number of punctate bodies ±SD per 10 cells from three independent experiments. Statistical significance was determined by Student’s t test (**, P < 0.01; ***, P < 0.001). (C) Immunoblotting of total protein extracts from the N. benthamiana leaves treated in (A). The membrane was probed with anti-GFP monoclonal antibody (@GFP MAb), or Myc MAb (@Myc MAb). (D) Subcellular localization of punctate bodies induced by 6K2-induced when VSR4 or TM were overexpressed in N. benthamiana leaves. The YFP-6K2 signals were stained with FM4-64. White arrows show the enrichment of YFP-6K2 in FM4-64 labelled single membrane vesicle-like structures (SMVLs). Scale bar represents 20 μm. BF, bright field. https://doi.org/10.1371/journal.ppat.1010257.g007 To verify the distribution of these 6K2-induced punctate bodies under AtVSR4TM treatment, we colocalized them with either FM4-64 or cis-Golgi marker. As expected, these punctate bodies did not enter into FM4-64 labeled granule-like endosomes when 6K2 was expressed alone or with AtVSR4 or TM (S11A Fig). However, we observed that punctate bodies induced by 6K2 colocalized with enlarged LE labelled by FM4-64 (Fig 7D), and also with cis-Golgi marker Man49-mCherry when AtVSR4 or AtVSR4TM were overexpressed (S11B and S11C Fig). These results indicate that non-glycosylated AtVSR4 can promote formation of 6K2-induced punctate bodies from cis-Golgi, which then target the enlarged LE.

[END]

[1] Url: https://journals.plos.org/plospathogens/article?id=10.1371/journal.ppat.1010257

(C) Plos One. "Accelerating the publication of peer-reviewed science."
Licensed under Creative Commons Attribution (CC BY 4.0)
URL: https://creativecommons.org/licenses/by/4.0/


via Magical.Fish Gopher News Feeds:
gopher://magical.fish/1/feeds/news/plosone/