(C) PLOS One
This story was originally published by PLOS One and is unaltered.
. . . . . . . . . .



Umbravirus-like RNA viruses are capable of independent systemic plant infection in the absence of encoded movement proteins [1]

['Xiaobao Ying', 'Department Of Cell Biology', 'Molecular Genetics', 'University Of Maryland', 'College Park', 'Maryland', 'United States Of America', 'Sayanta Bera', 'Jinyuan Liu', 'Roberto Toscano-Morales']

Date: 2024-05

The signature feature of all plant viruses is the encoding of movement proteins (MPs) that supports the movement of the viral genome into adjacent cells and through the vascular system. The recent discovery of umbravirus-like viruses (ULVs), some of which only encode replication-associated proteins, suggested that they, as with umbraviruses that lack encoded capsid proteins (CPs) and silencing suppressors, would require association with a helper virus to complete an infection cycle. We examined the infection properties of 2 ULVs: citrus yellow vein associated virus 1 (CY1), which only encodes replication proteins, and closely related CY2 from hemp, which encodes an additional protein (ORF5 CY2 ) that was assumed to be an MP. We report that both CY1 and CY2 can independently infect the model plant Nicotiana benthamiana in a phloem-limited fashion when delivered by agroinfiltration. Unlike encoded MPs, ORF5 CY2 was dispensable for infection of CY2, but was associated with faster symptom development. Examination of ORF5 CY2 revealed features more similar to luteoviruses/poleroviruses/sobemovirus CPs than to 30K class MPs, which all share a similar single jelly-roll domain. In addition, only CY2-infected plants contained virus-like particles (VLPs) associated with CY2 RNA and ORF5 CY2 . CY1 RNA and a defective (D)-RNA that arises during infection interacted with host protein phloem protein 2 (PP2) in vitro and in vivo, and formed a high molecular weight complex with sap proteins in vitro that was partially resistant to RNase treatment. When CY1 was used as a virus-induced gene silencing (VIGS) vector to target PP2 transcripts, CY1 accumulation was reduced in systemic leaves, supporting the usage of PP2 for systemic movement. ULVs are therefore the first plant viruses encoding replication and CPs but no MPs, and whose systemic movement relies on a host MP. This explains the lack of discernable helper viruses in many ULV-infected plants and evokes comparisons with the initial viruses transferred into plants that must have similarly required host proteins for movement.

Competing interests: I have read the journal’s policy and the authors of this manuscript have the following competing interests: JL and AES are on the CY1 (CYVaV) patent PCT/US19/60945 filed November 2019. AES is also a co-founder of Silvec Biologics, which has licensed the CY1 technology from the University of Maryland for development of a virus induced gene silencing vector, and has a financial interest in its success. SY is a Principal Research Scientist and JH is a Research Technician at Silvec Biologics and have a financial interest in its success.

Funding: This work was funded by the following grants: National Science Foundation EAGER 1912025 to AES, National Science Foundation MCB-1818229 to AES, United States Department of Agriculture 308291-00001 To AES, United States Department of Agriculture: The NIFA Citrus Disease Research and Extension grant AP18PPQS&T00C223 to AES. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Copyright: © 2024 Ying et al. This is an open access article distributed under the terms of the Creative Commons Attribution License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

To evaluate if ULVs can infect plants in the absence of an encoded MP and/or helper virus, we generated full-length clones of CY1 and CY2 and examined their abilities to systemically infect the model plant Nicotiana benthamiana. We report that both CY1 and CY2 were capable of independent systemic infection in a phloem-limited fashion when delivered by agroinfiltration. ORF5 was dispensable for infection of CY2, but was associated with faster symptom development. Unexpectedly, the CY2 ORF5 protein (ORF5 CY2 ) has characteristics more similar to luteoviruses/poleroviruses/sobemovirus CPs than to 30K MPs, which all share a similar single jelly-roll domain. In addition, CY2-infected plants contained small virus-like particles (VLPs), suggesting that CY2 uses ORF5 as a CP. As with viroid HSVd, CY1 interacted with phloem protein PP2 in sap, and formed a high molecular weight complex with sap proteins in vitro that was partially resistant to RNase treatment. When CY1 was used as a virus-induced gene silencing (VIGS) vector to reduce PP2 transcripts, CY1 accumulation was also reduced in systemic leaves, supporting the possibility that PP2 is the MP for CY1 and CY2. ULVs thus represent the first plant viruses, to our knowledge, whose systemic movement relies on a host MP, evoking comparisons to ancient viruses that must have similarly relied on host proteins for movement.

Surprisingly, several Group 2 ULVs were reported to exist in plants in the apparent absence of a helper virus [ 40 – 42 , 44 ]. These include members of Class 1 that infect papaya and babaco trees throughout Ecuador, and Class 2 opuntia umbra-like virus (OULV) found in opuntia plants. Likewise, CY1, which was discovered only once in Southern California limequat trees with yellow vein symptoms in the 1950’s, is currently present without a discernable helper virus in citrus trees that were infected with grafts from the original source material [ 40 ]. These reports of ULVs infecting plants in the apparent absence of a helper virus, despite some only encoding proteins that are known to be involved in replication in related members of the Tombusviridae [ 45 , 46 ], suggest that some ULVs exist autonomously in plants without encoding an MP.

(A) Genome organization of ULVs. Group 2/Class 1 members only encode 2 replication-required proteins and have extensive 3′ UTRs. Group 2/Class 2 members that infect dicots (except CY1) have an additional ORF (ORF5) that overlaps with the end of the RdRp ORF in the −1 frame. Monocot-infecting Class 2 members have an extra embedded ORF of different lengths. Group 2/Class 3 members have at least 1 additional ORF that are unrelated to each other and ORF5. (B) Maximum-likelihood phylogenetic tree based on RdRp nucleotide sequences. Branch numbers indicate bootstrap support in percentage out of 1,000 replicates. The scale bar denotes nucleotide substitutions per site. The tree is mid-point rooted. Dicot-infecting and monocot-infecting Class 2 ULVs separate into different clades with the exception of PULV, which is closer to an ancestral viral molecule that possibly gave rise to all Class 2 ULVs. (C) Left, secondary structure of CY1 [ 39 ]. CY2 shares a similar overall secondary structure and contains 2 large segments (in orange) missing in CY1. Green dots denote start codons and yellow dots denote stop codons. Nucleotide differences between CY2 and CY1 are denoted with red dots. ORF, open reading frame; PULV, parsley umbra-like virus; RdRp, RNA-dependent RNA polymerase; ULV, umbravirus-like virus.

A number of viruses have been recently identified that encode an umbravirus-related RdRp that is generated by −1 ribosomal frameshifting, and have umbravirus-related 3′ terminal RNA structures, but do not encode umbravirus-like MPs (Figs 1A and S1 ) [ 39 , 40 ]. Based on phylogenetic analyses, umbravirus-like viruses (ULVs) have been divided into 2 groups. Group 1 is a catch-all grouping of viruses that are more related to umbraviruses than Group 2 (Figs 1B and S1A ). Group 2 members subdivide into 3 classes ( Fig 1A and 1B ): Class 1 members only contain ORFs 1 and 2 while Class 2 members have an additional ORF (ORF5) that overlaps with the end of the RdRp ORF [ 41 , 42 ]. The one exception in Class 2 is citrus yellow vein associated virus (CY1, also known as CYVaV), which no longer encodes ORF5 due to 2 large deletions and other alterations ( Fig 1C ) [ 40 ]. CY2, a close relative of CY1 that is found in hemp [ 43 ], still contains ORF5. Class 2 members that infect monocots have an additional ORF of varying lengths in the −2 frame that is nearly completely embedded within ORF5. All Class 3 members have their own unique additional ORF(s) that are unrelated to ORF5 [ 39 ].

Umbraviruses (family Tombusviridae) are unusual in that they do not encode CPs or silencing suppressors. Rather, umbraviruses rely on a helper virus from the luteovirus/enamovirus/polerovirus genera for silencing suppression and for supplying CP for gRNA encapsidation [ 32 ]. Umbraviruses have a 4.0 to 4.5 kb monopartite (+)RNA genome containing 4 ORFs [ 33 – 36 ]. The two 5′ proximal ORFs (ORF1 and ORF2) encode replication-required proteins including the RdRp, which is produced by −1 ribosomal frameshifting just upstream of the ORF1 termination codon [ 37 ]. The other 2 ORFs, which are overlapping and located downstream of an intergenic region, encode a long-distance MP (ORF3) and a 30K class cell-to-cell MP (ORF4), both of which are translated from a bicistronic subgenomic (sg)RNA [ 38 ].

Unlike host MPs, viral MPs have been extensively studied [ 11 , 12 ]. Viral MPs are nonspecific RNA-binding proteins that assist in the transport of large RNP complexes while simultaneously protecting viral RNAs from degradation by host defenses [ 11 ]. Viral MPs, which range in size from 8 kDa to 58 kDa, associate with the cytoskeleton and ER networks to move viral RNA complexes intracellularly to the PD, where most can increase the PD size exclusion limit (SEL) in a process known as “gating” [ 11 – 14 ]. At the PD, MPs mediate gating either without extensive modification of PD structure or by forming homomeric tubules through which virions transit between cells [ 11 , 12 ]. The most extensive class of MPs are the 30K MPs, named after the tobacco mosaic virus 30 kDa MP, which have a single jelly roll central core domain comprising 7 or 8 β-strands that contains a nearly invariant aspartic acid residue that is critical for MP function [ 8 , 15 ]. The N-terminal portion of 30K MPs is involved in PD targeting [ 16 , 17 ], while the C-terminal region can interact with CPs and/or can support long-distance movement through the vascular system [ 18 – 20 ]. Alterations in conserved MP residues restrict a virus to initially infected cells [ 21 , 22 ], and viruses with multiple MPs or chimeric viruses with heterologous MPs have an extended host range [ 23 – 26 ], supporting the critical role of encoded MPs in the infection cycle of plant viruses. Additionally, CPs can regulate movement of viral gRNAs by increasing the RNA-binding affinity of 30K MPs [ 13 , 27 ] and can also be indispensable for reaching vascular tissues without a requirement for virion formation [ 28 – 31 ].

Plant viruses likely originated from viruses that were transferred into the plant by plant-feeding nematodes or arthropods, and which at minimum would have coded for an RdRp and at least 1 CP [ 2 ]. To achieve a systemic infection in plants, ancient viruses would have needed to take advantage of a plant’s intercellular communication pathway that currently transports endogenous RNAs through the vascular system to other organs in poorly defined RNP complexes [ 3 – 6 ]. Modern plant virus MPs may have evolved from host MPs, as implied by the MP of red clover necrotic mosaic virus having a pumpkin RNA-binding paralog that can mediate transport of RNA through PD [ 7 ], or by duplication of a CP open reading frame (ORF) that then evolved MP functionality, as was recently proposed based on the high structural similarity of some MPs and CPs in their core domain [ 8 ]. Since encoding an MP is the key defining feature of modern plant viruses, these virus-encoded MPs must have afforded an advantage in transporting the viral genome over endogenous MPs. Of all modern-day infectious RNAs, only circular noncoding viroids are thought to still use host MPs. For example, hop stunt viroid (HSVd) uses phloem protein 2 (PP2), an enigmatic, dimeric, chitin-binding lectin encoded by a large gene family that is highly abundant in phloem sap and is considered to be a nonspecific RNA-binding protein [ 9 ]. PP2 has also been implicated in plant virus transmission, as purified PP2 mixed with virions and fed to aphids can increase the virus transmission rate as well as virion stability in vitro [ 10 ].

To be a functional plus-strand (+)RNA plant virus, genomes must be translated upon cell entry, replicated, and nascent genomes transported through intercellular plasmodesmata (PD) connections into adjacent cells as ribonucleoproteins (RNP) or after packaging into virions. After cell-to-cell movement, RNPs or virions enter the vascular system for long-distance movement through phloem sieve elements, and then exit and continue cell-to-cell spread. Some viruses are confined to the vascular system and only replicate within phloem-associated nucleated cells. Regardless of tropism, newly synthesized (+)RNA genomes require encapsidation into virions for plant-to-plant transmission either by vectors or by mechanical means [ 1 ]. To accomplish their infection cycle, plant (+)RNA viruses are expected to code for: (i) one or more replication proteins, including the core RNA-dependent RNA polymerase (RdRp); (ii) one or more movement proteins (MPs) to interact with PD components for cell-to-cell and/or long-distance movement; (iii) RNA silencing suppressor(s) to suppress the plant’s innate defense system against viruses; and (iv) capsid proteins (CPs) for encapsidation of the genomic (g)RNA(s) [ 1 ].

Results

CY1 systemically infects N. benthamiana and Mexican lime in the absence of a helper virus We previously determined that transcripts synthesized from a full-length clone of CY1 replicated to similar levels as umbravirus pea enation mosaic virus 2 (PEMV2) in Arabidopsis thaliana protoplasts [39]. This result was unexpected as CY1 only contains the umbravirus replication-associated ORFs and thus lacks umbravirus ORF3. The ORF3 protein, in addition to supporting long-distance movement, also functions to inhibit nonsense-mediated decay, which is necessary for efficient replication of PEMV2 in protoplasts [47]. The lack of a detectable helper virus in CY1-infected citrus trees [40] suggested that CY1 is also capable of independent systemic infection despite the lack of an encoded MP. To determine the validity of this hypothesis, N. benthamiana plants were agroinfiltrated with CY1 and examined daily for symptom development. Of the initially infiltrated plants, 10% developed an abnormal heart-shaped leaf at 21 days post-infiltration (dpi), and subsequent leaves emerging at the apex were uniformly small and narrow, with severely curled margins (Fig 2A, lower panel, and 2B). Clusters of small leaves began emerging from nodes at about 30 dpi, and plants also were substantially stunted (Fig 2B and 2D). Examination of roots at late stages of infection revealed the presence of 1 or more large galls that were absent from uninfected plants. Total RNA extracted from systemic leaves of newly infiltrated seedlings at 19 dpi and hybridized with CY1-specific probes revealed the presence of CY1 gRNA, with levels increasing in leaves near the apex (Fig 2C). CY1 gRNA was also detected in stems and root galls at 50 dpi (Fig 2D). PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 2. CY1 systemically infects N. benthamiana in the absence of a helper virus. (A) Leaves 3 and 4 were agroinfiltrated at the 4-leaf stage with a Ti plasmid expressing full-length CY1 transcripts. Initial heart-shaped leaf (leaf 9; black arrow) was visible at 21 dpi. Subsequent emerging leaves were small and curled down at the margins (red arrow). (B) Typical phenotype of infected plants at 50 dpi. Multiple leaves emerged at each node and roots contained multiple galls. Infected plants were stunted compared with uninfected plants of the same age (right panel). Infected plants senesced within 3 to 4 months post-infiltration. (C) Detection of CY1 in systemic upper leaves by northern blot at 19 dpi. Total RNA was extracted from specified leaves (leaf numbering shown in A and subjected to northern blot analysis). Ethidium-bromide stained rRNA was used as a loading control. (D) Stem and 3 sections of the root gall were excised from a 50 dpi plant (shown in left panel) and pressed onto nitrocellulose, which was probed for CY1 (right panel). G1-3, galls. dpi, days post-infiltration. https://doi.org/10.1371/journal.pbio.3002600.g002 To determine if CY1 can also independently infect citrus, Mexican lime was inoculated with CY1 either by direct agroinfiltration or by using dodder to transfer the virus from infected N. benthamiana (S2A Fig). Total RNA that was extracted from selected plants 12 or 15 months later, respectively, contained CY1 gRNA that was absent from uninfected plants (S2B Fig, right panel). Unlike N. benthamiana, citrus infected with CY1 did not show any discernable symptoms (S2B Fig, left panel). These results strongly suggest that CY1 is capable of independent systemic infection of plants despite the lack of encoded MPs.

CY1 is phloem-limited in infected stems and roots CY1 distribution in N. benthamiana was examined by fluorescence in situ hybridization (FISH). CY1 in apical stems was limited to phloem parenchyma cells (PPCs), sieve tubes (SE), and companion cells (CC), while scattered fluorescent signals in lower stems were only found in phloem tissue (Fig 3A). CY1 was also detected in root sections, with signals observed in both immature (nucleated) and mature (enucleated) SE and in adjacent phloem-associated cells (Fig 3B). These results indicate that CY1 movement is bidirectional in N. benthamiana and strongly suggests that CY1 infection is vascular-limited. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 3. Systemic infection of CY1 is limited to vascular tissue. FISH analysis showing accumulation of CY1 RNA in stem and root sections of infected N. benthamiana at 60 dpi. (A) Transverse section of the bottom stem, and transverse and longitudinal sections of apical stem of healthy and CY1-infected plants. Cy3-labeled CY1 probe produced the fluorescent signal. Arrowheads denote signals in specific phloem locations. DAPI-stained DNA and xylem tissue fluoresce blue. All sections were stained with ProLong Gold Antifade Mountant (Invitrogen, USA). Bars = 60 μm. PPC, phloem parenchyma cells; SE, sieve tube elements; CC, companion cells. (B) Similar analysis of root sections. dpi, days post-infiltration; FISH, fluorescence in situ hybridization. https://doi.org/10.1371/journal.pbio.3002600.g003

ORF5 CY2 is dispensable for replication and movement of CY2 Since the evolutionary loss of ORF5 did not preclude CY1 systemic infection, we next investigated CY2 infection in the presence and absence of its presumed MP, ORF5 CY2 . Unlike CY1, CY2 contains a perfect carmovirus consensus sequence (CCS; GGGUAAAUA) just upstream of the ORF5 CY2 initiation codon (Fig 4A, in green). CCSs are present at the 5′ ends of gRNAs and sgRNAs in carmoviruses, umbraviruses, and members of some other genera within the Tombusviridae [38,48,49], and sgRNA promoters are located just upstream of the CCS [50]. Two sets of mutations were generated within and upstream of the CCS to engineer CY2 variants that were incapable of generating sgRNA, and thus should preclude synthesis of ORF5 CY2 (Fig 4A). CY2sgm1 contains mutations that altered the CCS central guanylate and an upstream residue (to maintain local RNA structure) and also eliminated the ORF5 CY2 initiation codon; CY2sgm2 contains 3 alterations in the CCS and a fourth alteration in a nearby downstream residue. Since there was a possibility that ORF5 CY2 could still be synthesized by internal initiation, similar to carmovirus turnip crinkle virus CP [51], an additional construct was generated (CY2T5) with a single alteration that introduced a UGA stop codon at position 2203, which would produce a truncated, 33 amino acid ORF5 CY2 . None of the alterations affected the sequence of the RdRp or the presumptive secondary structure of the gRNA. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 4. CY2 replicates and moves systemically in the absence of ORF5. (A) Secondary structures of a portion of CY1 (left) and the corresponding region of CY2 (right). Sequence shaded light green in CY2 is a “carmovirus consensus sequence” (G 2-3 A/U 5-9 ) located at the 5′ ends of carmovirus and umbravirus gRNAs and sgRNAs. Red residues differ between CY1 and CY2. ORF5 initiation codon is circled and shaded yellow. CY2 mutants CY2sgm1 and CY2sgm2 have color-coded alterations in the sgRNA promoter region. CY25T contains a truncated ORF5 protein (33 amino acids remaining). (B) Accumulation of (+)- and (-)-strands of CY1, CY2, and CY2 mutants in Arabidopsis protoplasts at 24 h and 48 h post-transfection assayed by northern blots. CY1 accumulation was low but detectable. CY1 with an RdRp active site alteration (CY1-GDD) served as a negative control. Mock, no added RNA. Data was obtained from 3 independent experiments. Standard deviation is shown. No significant difference in accumulation of the gRNA was found for CY2 and the CY2 mutants using JMP Pro 16, Student’s t, α = 0.05. (C) Competition assays between CY1 and WT and mutant CY2. Symptomatic systemic leaves were collected at 3-, 4-, and 6-wpi and total RNA was extracted and subjected to northern analysis. CY1 and CY2 size differences were used to visually assess levels of the different gRNAs. Three independent experiments were conducted with very similar results. Note that while WT CY2 is mainly detected at 3-wpi in co-infiltrated plants, both CY1 and CY2 are present at 4- and 6-wpi (left panel). In contrast, CY1 is mainly detected at all time points in plants co-infiltrated with the ORF5 deficient or defective mutants (middle and right panels). ●, denotes low level band that is not detected in CY2sgm1 protoplasts infections that is near the size of the sgRNA (831 nt) and is also the size of a defective (D)-RNA (921 nt) that commonly arises in CY1/CY2 infections. ORF, open reading frame; RdRp, RNA-dependent RNA polymerase; wpi, weeks post-infiltration; WT, wild type. https://doi.org/10.1371/journal.pbio.3002600.g004 Arabidopsis protoplasts were transfected with CY1, CY1 GDD (CY1 with an altered RdRp active site), CY2, CY2sgm1, CY2sgm2, and CY2T5. CY2 (+)- and (-)-strands accumulated to substantially higher levels than CY1 at 24 and 48 hours post-transfection (hpi) (Fig 4B). As expected, only protoplasts infected with CY2 and CY2T5 were associated with a prominent sgRNA of the expected size (Fig 4B). No significant difference was found for the accumulation of CY2 and the CY2 ORF5 mutants (+)- or (-)-strands in protoplasts. The reason for the increased accumulation of CY2 gRNA compared with CY1 is not known; however, these data show that CY2sgm1 and CY2sgm2 do not detectably express the CY2 sgRNA and that none of the CY2 variants have impaired gRNA replication. To determine if systemic infection requires ORF5 expression, N. benthamiana plants were infiltrated with the same constructs. Plants infiltrated with wild-type (WT) CY2 produced symptoms similar to CY1; however, the heart-shaped leaf appeared at 2 weeks post-infiltration (wpi) in the majority of plants, which was 1 week earlier than found for CY1-infiltrated plants (Table 1). In addition, CY2-infected plants exhibited greater stunting, likely a consequence of earlier symptom development. All 3 ORF5 CY2 -defective mutants produced similar symptoms as CY2, but timing of symptoms and degree of stunting was more similar to that of CY1 (Table 1). To determine if there was a fitness cost associated with the absence of ORF5, competition assays were performed between CY1 and WT or mutant CY2. In plants co-infiltrated with CY1 and CY2, CY2-sized gRNA was more prominent at 3 wpi (Fig 4C, left panel). At 4 and 6 wpi, however, both CY2- and CY1-sized gRNAs were discernable, suggesting that CY1 and CY2 had similar fitness despite CY1’s lack of ORF5 and slower initial infection. In contrast, CY1 accumulation was clearly favored at 4- and 6-wpi in plants co-infiltrated with either CY2sgm1 or CY25T (Fig 4C, middle and right panels). This result suggests that CY1 is more fit than CY2 ORF5 mutants, which may reflect adaption of CY1 within long-lived citrus in the absence of ORF5. Altogether, these results indicate that ORF5 CY2 is not required for systemic infection of CY2 or CY1, but rather plays a supportive role that decreases the timing to initial symptoms. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Table 1. Percentage of N. benthamiana plants showing symptoms at different times post-infiltration of CY1, CY2, and CY2 ORF5 mutants. https://doi.org/10.1371/journal.pbio.3002600.t001 To determine if ORF5 CY2 alters the vascular-limited tropism of CY1 in infected plants, transverse sections of apical stems, bottom stems, petioles, and roots from a healthy plant as well as CY1-, CY2-, and CY2sgm1-infected plants were examined by confocal microscopy (S3 Fig). Fluorescent signal distribution was similar for all viruses and variants, which remained limited to phloem-associated cells. These results suggest that the presence of ORF5 CY2 does not affect tissue tropism of either CY2 or CY1 in infected plants.

ORF5 CY2 also accelerates infection of CY1 Enhanced accumulation of CY1 in plants co-infiltrated with CY2sgm1 and CY25T suggested that CY1 has adapted to compensate for the loss of ORF5. To determine if CY1 infection can be accelerated by ORF5 CY2 , transgenic N. benthamiana plants expressing ORF5 CY2 under the cauliflower mosaic virus (CaMV) 35S promoter (N.b.-ORF5 CY2 ) were generated. Transgenic N. benthamiana expressing OULV ORF5 (N.b.-ORF5 OULV ) were also generated in case CY1 was silenced by the more closely related CY2 ORF5 sequence, most of which is still present in CY1. Transgenic plants, which were confirmed to be expressing the respective ORF5 proteins (S4 Fig), along with WT N. benthamiana were infiltrated with CY1 and timing to first symptoms determined (Table 2). At 2 wpi, no WT plants infiltrated with CY1 developed symptoms, whereas 22% to 48% of transgenic plants displayed CY1-specific symptoms. At 3 wpi, 32% of WT plants were symptomatic compared with 61% to 86% of infected transgenic plants. Finally, at 4 wpi, 59% of WT N. benthamiana plants were symptomatic compared with 78% to 90% of transgenic plants. These data strongly suggest that CY1 infects more rapidly in the presence of ORF5 proteins. Altogether, our results indicate that Class 2 ULVs can systemically infect the vascular system of plants while encoding only umbravirus-related replication proteins and that the ORF5 protein functions to expedite the initial infection. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Table 2. CY1 infects more rapidly in transgenic plants expressing CY2 or OULV ORF5. https://doi.org/10.1371/journal.pbio.3002600.t002

ORF5 CY2 cannot substitute for tobacco mosaic virus MP or CP Earlier symptoms in CY2-infected N. benthamiana (Table 1) and CY1-infected transgenic N. benthamiana (Table 2) compared with CY1 and CY2 mutant-infected plants suggested that ORF5 CY2 might be functioning as an MP. However, the dispensability of ORF5 for movement suggested that if ORF5 protein is an MP, it is not a canonical MP. To further examine if ORF5 CY2 is a viral MP, we investigated the ability of ORF5 CY2 to substitute for the TMV MP or CP in a GFP-expressing TMV vector (S5 Fig). Two different TMV vectors were designed containing either a truncated MP or a deleted CP, both of which were incapable of systemic movement in infiltrated N. benthamiana. Replacing either the TMV MP or CP with ORF5 CY2 did not restore systemic movement (S5 Fig), leaving the possibility that ORF5 CY2 is not functioning as a canonical MP.

ORF5 CY2 shares structural similarity with 30K MPs and luteovirus/polarovirus/sobemovirus CPs To further investigate the function of the ORF5 protein, a structural model of ORF5 CY2 was generated using Alphafold2, which predicted with high confidence the presence of a jelly-roll domain containing 8 β-strands (Fig 5A). In addition, the N-terminal region of ORF5 CY2 (and all other Class 2 ORF5 proteins) contains an unstructured, arginine (R)-rich sequence that was predicted to contain a nucleolar localization signal (NoLS) (S6A Fig). To support the presence of a functional ORF5 NoLS, N. benthamiana leaves were co-infiltrated with a plasmid expressing an ORF5 CY2 -GFP fusion protein and a second plasmid expressing RFP-NoLS. Infiltrated leaf samples collected at 2 dpi and observed under a confocal microscope showed that nearly all red nucleoli that were sequestering RFP-NLS also contained green/yellow dots indicating the co-localization of GFP-tagged ORF5 CY2 (Fig 5B). These data support the presence of a functional NoLS in the N-terminal region in ORF5 CY2 . PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 5. ORF5 CY2 is structurally similar to 30K MPs and luteovirus/polerovirus/sobemovirus CPs. (A) Structure modeling of full-length ORF5 CY2 using Alphafold2. Color coding from blue to red indicates high to low structural confidence, respectively. Unstructured N-terminal region contains a predicted nucleolar localization signal (NoLS). (B) Overlay images of ORF5 CY2 -GFP and RFP-NoLS in N. benthamiana. Leaves were co-infiltrated with Agrobacteria expressing ORF5 CY2 -GFP and RFP-NoLS from CaMV 35S promoters. Infiltrated leaves were collected at 2 dpi and subjected to confocal microscopy. (C) Single jelly-roll domains of ORF5 CY2 , OuMV MP (MP OuMV ), and BYDV P3 CP (CP BYDV ). Full-length structures of MP OuMV and CP BYDV are boxed. CP BYDV also contains an N-terminal NoLS. (D) Maximum-likelihood phylogenetic tree based on amino acid sequences of Class 2 ORF5 (blue), representative 30K MPs (red), and SeMV (sobemovirus), BYDV (luteovirus), and PLRV (polerovirus) CPs (green). Branch numbers indicate bootstrap support in percentage out of 1,000 replicates. The letters beneath the branch numbers indicate the node names. The scale bar denotes amino acid substitutions per site. The tree is mid-point rooted. Protein sequences were aligned through PROMALS3D and subjected to phylogenetic analysis in MEGA X. BYDV, barley yellow dwarf virus; CP, capsid protein; dpi, days post-infiltration; MP, movement protein; ORF, open reading frame; OuMV, ourmia melon virus; PLRV, potato leafroll virus; SeMV, sesbania mosaic virus. https://doi.org/10.1371/journal.pbio.3002600.g005 Three-dimensional structural models were recently generated for members of the 30K MP superfamily and were shown with high confidence to contain a conserved central region containing a similar 7 or 8 β-strand jelly-roll domain that also resembles the jelly-roll topologies of some CPs [8]. As CPs can assist in viral RNA movement, a bioinformatic approach was taken to manually curate the relatedness of the structural model for ORF5 CY2 with representative 30K MPs and CPs. The MP most closely resembling ORF5 CY2 was the 30K class MP of ourmia melon virus (MP OuMV ) (Fig 5C, top). Superimposition of the structural models for ORF5 CY2 and MP OuMV showed good alignment of the jelly-roll domains and partial alignment of α-helices. The structural model for ORF5 CY2 was then subjected to further analysis using the DALI server, which searches for proteins in the PDB database with a similar structure. The best hit was the 22 kDa CP of luteovirus barley yellow dwarf virus (BYDV), which had a significant Z score of 10 (Fig 5C, bottom) (note that no MPs were identified due to the absence of any high-resolution MP structures in the PDB database). The BYDV CP is the same size as ORF5 CY2 and also contains an NoLS sequence in its unstructured N-terminal region [52] (S6B Fig). If ORF5 proteins are 30K class MPs, then they should contain the critical aspartic acid residue within the jelly-roll domain that is required for function. Alignment of ORF5 proteins with selected 30K MPs indicated the absence of this aspartic acid residue (S7A Fig). We also examined the charge distribution across ORF5 CY2 and other ORF5 proteins, since unlike MPs, CPs are known to have an unequal charge distribution with charged residues concentrated in the N-terminal region [8]. As shown in S7B Fig, ORF5 CY2 has an unequal charge distribution similar to CPs. To determine if there is a closer evolutionary link between ORF5 proteins and luteovirus/polerovirus/sobemovirus (LPS) genera CPs than with 30K MPs, a maximum-likelihood phylogenetic tree was generated using the aligned amino acid sequences produced in PROMALS3D [53], which uses structural and biophysical characteristics of proteins for alignment. Phylogenetic analyses suggested that ORF5 proteins form a strong clade with LPS CPs (node B), which is a sister clade to MP OuMV. The tree also suggested that MP OuMV , ORF5 proteins, and LPS CPs share a similar jelly-roll domain derived from a common ancestral protein (Fig 5D, node A), demonstrating an evolutionary relationship between LPS CPs and the MP 30K superfamily, as was recently suggested [8]. Altogether, these results suggest that ORF5 proteins are more related to CPs than MPs.

ORF5 CY2 -infected plants contain macromolecular structures resembling VLPs Since ORF5 proteins are more closely related to LPS CPs, we wanted to determine if the presence of ORF5 was associated with the presence of VLPs. To examine if CY2-infected plants contained ORF5-associated VLPs, CY1- and CY2-infected N. benthamiana tissue was subjected to a gentle virion purification protocol [54], and the 70% and 25% sucrose layers and interface were subjected to electrophoresis through 0.8% agarose gels. The 70% sucrose fraction from CY2-infected plants contained high molecular weight material that stained with ethidium bromide and did not migrate into the gels (Fig 6A). Northern blots using CY2-specific probes detected the presence of CY2 gRNA and sgRNA in the 70% sucrose fraction (Fig 6C), and western analysis confirmed the additional presence of ORF5 CY2 (Fig 6B). High molecular weight material was also present in the 70% sucrose fraction from CY1-infected plants, but this material did not contain detectable full-length CY1 RNA. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 6. ORF5 CY2 forms VLPs in infected N. benthamiana. (A) CY1- and CY2-infected plants were subjected to ultracentrifugation to gently purify any VLPs. Material from CY1- and CY2-infected N. benthamiana in the 70% sucrose cushion, interface (Int), and 25% sucrose cushion was examined following electrophoresis through a 0.8% agarose gel. Nb, N. benthamiana. (B) Western blot detection of ORF5 CY2 . Total proteins from healthy N. benthamiana and CY2-infected N. benthamiana, and 70% sucrose fractions of CY1- and CY2-infected N. benthamiana were separated by 10% PAGE. ORF5 CY2 -specific antibody was used to detect ORF5 CY2 (arrow). (C) Northern blot detection of CY2 gRNA and sgRNA in material from the 70% sucrose fraction of CY2-infected N. benthamiana. Total RNA from uninfected N. benthamiana and CY2-infected N. benthamiana served as negative and positive controls, respectively. (D) Samples of 70% sucrose cushions from uninfected N. benthamiana, CY1-infected N. benthamiana, and CY2-infected N. benthamiana were exchanged to PBS buffer and observed by EM. Particles of 3 different sizes were found in CY2-infected N. benthamiana that were absent from uninfected and CY1-infected N. benthamiana. Yellow arrows point to CY2-infected N. benthamiana particles with a diameter of 14 nm that were the most prevalent. Red arrow denotes a 24 nm particle and blue arrow denotes a 35 nm particle. EM, electron microscopy; N b., N. benthamiana; ORF, open reading frame; VLP, virus-like particle. https://doi.org/10.1371/journal.pbio.3002600.g006 Seventy-percent sucrose fraction samples from uninfected N. benthamiana and CY1- and CY2-infected plants were visualized by electron microscopy (EM). CY2-infected samples contained numerous macromolecular assemblies, the majority of which were uniformly approximately 14 nm in size (Fig 6D, bottom panels). Two additional larger particles were also occasionally found that were 24 nm and 35 nm in diameter (Fig 6D, lower right panel). Samples from CY1-infected plants contained complexes of irregular size and shape without the sharp contrast boundary observed for the particles from CY2-infected plants. None of the CY2 or CY1 complexes were visible in uninfected N. benthamiana samples. These results indicate that CY2-infected plants contain macromolecular assemblies resembling small VLPs that are associated with ORF5 CY2 and CY2 RNA, supporting the hypothesis that ORF5 CY2 has features more consistent with CPs than MPs.

CY1 and a CY1-derived defective (D)-RNA bind phloem protein 2 (PP2) in vitro and in vivo and form large RNP complexes with cucumber sap that confers partial protection from RNase digestion MPs are required for protecting the viral gRNA from degradation and for transport through plasmodesmata [11]. Since CY1 and ORF5 CY2 -deficient CY2 were able to systemically infect N. benthamiana (Fig 2 and Table 1) in the absence of an encoded MP or helper virus, the most likely explanation is that CY1 and CY2, similar to viroids, are using a host RNA MP. Since cucumber is an excellent model system for studying the movement of endogenous RNAs in the phloem due to the availability of large quantities of phloem sap [55,56], we agroinfiltrated cucumber with CY1 to assess its suitability as a host. RT-PCR of systemic leaves from infiltrated cucumber revealed the presence of CY1 (+)- and (-)-strand RNAs (S8 Fig), suggesting that cucumber would be a good system for identifying host proteins involved in CY1 systemic movement. Northwestern assays were performed using cucumber phloem sap and uniformly labeled (i) CY1; (ii) a CY1 defective RNA (D-RNA; 921 nt) composed of positions 1–671 joined to positions 2442–2693 that is frequently found in CY1- and CY2-infected N. benthamiana; and (iii) umbravirus PEMV2 (Fig 7A). All viral RNAs bound to a single location in the gels, corresponding with the position of the major 26 kDa variant of PP2, the protein responsible for HSVd movement. Approximately 7.5-fold more CY1 and 10-fold more D-RNA bound at this location compared with umbravirus PEMV2, suggesting that the protein migrating at this position is binding more efficiently to CY1 RNAs. PPT PowerPoint slide

PNG larger image

TIFF original image Download: Fig 7. CY1 and a CY1-derived defective (D)-RNA binds to cucumber PP2 in vitro and in vivo, which confers partial protection against RNase. (A) Northwestern analyses of CY1-binding proteins in cucumber sap. Cucumber phloem exudates (25 and 100 μg) were subjected to electrophoresis on 15% SDS-PAGE gels that were stained with Coomassie blue (left), or transferred to nitrocellulose membranes and probed with uniformly labeled CY1, CY1 D-RNA, or PEMV2 transcripts (right). Numbers to the right denote relative binding in 3 independent experiments with standard deviations. (B) EMSAs. Left, transcripts were mixed with increasing amounts of cucumber phloem sap extracts and analyzed on 1.5% non-denaturing agarose gels containing ethidium bromide. Right, EMSAs of sap+transcripts that were treated (+) or not treated (-) with RNAse A at room temperature for 15 min. BSA in place of sap was used as a negative control. (C) Phloem sap exudates from uninfected (Un) and CY1-infected cucumber (2 samples) at 12 dpi were collected (Input) and used for RNA-pulldown assays using streptavidin beads without (No Probe) and with attached 5′-biotinylated CY1 probes (Probe). Pulldown samples were resolved by SDS-PAGE. Top panel: gel stained with Coomassie blue. Second panel, western blot using anti-cucumber PP2 antibody (α-CsPP2). Arrow denotes the 26 kDa band corresponding to PP2. Third panel: membrane from western blot stained with Ponceau S. Fourth panel, RNAs were extracted and subjected to RT-PCR using CY1-specific primers. dpi, days post-infiltration; EMSA, electrophoretic mobility shift assay; PEMV2, pea enation mosaic virus 2; PP2, phloem protein 2. https://doi.org/10.1371/journal.pbio.3002600.g007 To further investigate CY1/CY1 D-RNA specific interaction with a sap protein, electrophoretic mobility shift assays (EMSAs) were conducted using varying amounts of cucumber sap and sufficient RNA for detection using ethidium-bromide-stained gels. CY1 elicited a sharp transition to a high MW complex in the presence of 20 μg of sap that was unable to migrate into the 0.8% agarose gels (Fig 7B). Addition of 40 μg of sap shifted all detectable CY1 transcripts to the high MW complex. CY1 D-RNA transitioned nearly completely to the high MW complex with the addition of only 10 μg of sap. EMSAs of PEMV2 were similar to those of CY1, despite binding more weakly to the interacting protein(s) in the northwestern assays. To examine whether the high MW complex protects the associated RNA from treatment with RNase, EMSAs were repeated with and without added RNase A. A fraction of the CY1 complex and the majority of the CY1 D-RNA complex were protected from 15 min of RNase treatment at all sap concentrations (Fig 7B, right panel). In contrast, the complex containing PEMV2 showed no protection, even at the highest sap concentration. These results suggest that a protective RNP complex forms between sap proteins and CY1 RNAs. To determine if CY1 is interacting with PP2 in vivo, RNA pulldown assays were performed. Sap from uninfected and CY1-infected cucumber at 12 dpi was collected and treated with a biotin-labeled probe complementary to CY1. Several proteins were extracted along with CY1, including a major protein of 24 to 26 kDa (Fig 7C, upper panel). Western blots demonstrated that this interacting protein cross-reacted with anti-cucumber PP2 antibody and RT-PCR confirmed that the pull-down complex contained CY1 RNA (Fig 7C, bottom panels). These results suggest that PP2, a 26 kDa protein with a vascular expression profile [57] that forms RNP complexes with HSVd in vitro [58], is also associating with CY1 in phloem sap during infection.

[END]
---
[1] Url: https://journals.plos.org/plosbiology/article?id=10.1371/journal.pbio.3002600

Published and (C) by PLOS One
Content appears here under this condition or license: Creative Commons - Attribution BY 4.0.

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

You are viewing proxied material from magical.fish. The copyright of proxied material belongs to its original authors. Any comments or complaints in relation to proxied material should be directed to the original authors of the content concerned. Please see the disclaimer for more details.