Viral Genome-linked Protein Research Articles

Comparisons of the genetic structure of populations of Turnip mosaic virus in West and East Eurasia

The genetic structure of populations of Turnip mosaic virus in Eurasia was assessed by making host range and gene sequence comparisons of 142 isolates. Most isolates collected in West Eurasia infected Brassica plants whereas those from East Eurasia infected both Brassica and Raphanus plants. Analyses of recombination sites (RSs) in five regions of the genome (one third of the full sequence) showed that the protein 1 (P1 gene) had recombined more frequently than the other gene regions in both subpopulations, but that the RSs were located in different parts of the genomes of the subpopulations. Estimates of nucleotide diversity showed that the West Eurasian subpopulation was more diverse than the East Eurasian subpopulation, but the Asian-BR group of the genes from the latter subpopulation had a greater nonsynonymous/synonymous substitution ratio, especially in the P1, viral genome-linked protein (VPg) and nuclear inclusion a proteinase (NIa-Pro) genes. These subpopulations seem to have evolved independently from the ancestral European population, and their genetic structure probably reflects founder effects.

Mutations in potato virus Y genome-linked protein determine virulence toward recessive resistances in Capsicum annuum and Lycopersicon hirsutum.

The recessive resistance genes pot-1 and pvr2 in Lycopersicon hirsutum and Capsicum annuum, respectively, control Potato virus Y (PVY) accumulation in the inoculated leaves. Infectious cDNA molecules from two PVY isolates differing in their virulence toward these resistances were obtained using two different strategies. Chimeras constructed with these cDNA clones showed that a single nucleotide change corresponding to an amino acid substitution (Arg119His) in the central part of the viral protein genome-linked (VPg) was involved in virulence toward the pot-1 resistance. On the other hand, 15 nucleotide changes corresponding to five putative amino acid differences in the same region of the VPg affected virulence toward the pvr2(1) and pvr2(2) resistances. Substitution models identified six and five codons within the central and C terminal parts of the VPg for PVY and for the related potyvirus Potato virus A, respectively, which undergo positive selection. This suggests that the role of the VPg-encoding region is determined by the protein and not by the viral RNA apart from its protein-encoding capacity.

A cysteine-rich plant protein potentiates Potyvirus movement through an interaction with the virus genome-linked protein VPg.

We have identified a cellular factor that interacts with the virus genome-linked proteins (VPgs) of a diverse range of potyviruses. The factor, called Potyvirus VPg-interacting protein (PVIP), is a plant-specific protein with homologues in all the species examined, i.e., pea, Arabidopsis thaliana, and Nicotiana benthamiana. The sequence of PVIP does not identify a specific function, although the existence of a "PHD finger" domain may implicate the protein in transcriptional control through chromatin remodeling. Deletion analysis using the yeast two-hybrid system showed that the determinants of the interaction lay close to the N terminus of VPg; indeed, the N-terminal 16 amino acids were shown to be both necessary and sufficient for the interaction with at least one PVIP protein. From a sequence comparison of different potyvirus VPg proteins, a specific amino acid at position 12 was directly implicated in the interaction. This part of VPg is distinct from regions associated with other functional roles of VPg. Through mutation of Turnip mosaic virus (TuMV) at VPg position 12, we showed that the interaction with PVIP affected systemic symptoms in infected plants. This resulted from reduced cell-to-cell and systemic movement more than reduced virus replication, as visualized by comparing green fluorescent protein-tagged wild-type and mutant viruses. Furthermore, by using RNA interference of PVIP in Arabidopsis, we showed that reduced expression of PVIP genes reduced susceptibility to TuMV infection. We conclude that PVIP functions as an ancillary factor to support potyvirus movement in plants.

Detection of a natural point mutation in Potato virus A that overcomes resistance to vascular movement in Nicandra physaloides., and studies on seed‐transmissibility of the mutant virus

SummaryIsolate M of Potato virus A (PVA‐M; genus Potyvirus) is avirulent in Nicandra physaloides L. (family Solanaceae). The inoculated leaves are infected but no systemic infection is observed. Forty plants of ‘Black Pod’ (BP) and ‘Black Pod Alba’ (BPA), two variants of N. physaloides described in this study, were inoculated with PVA‐M. Two plants of BP and one plant of BPA were systemically infected. Mosaic, blistering and dark green islands developed on the systemically infected leaves, and flowers showed colour‐break symptoms. PVAprogeny were sequence‐characterised for the 6K2 protein and viral genome‐linked protein (VPg) encoding regions known to control the long distance movement of PVA in N. physaloides. All virus progeny (designated as PVA‐Mm) in the systemically infected leaves of the plants inoculated with PVA‐M contained only a single amino acid substitution (Vail 16Met) in the central part of VPg due to a nucleotide substitution G6033A, as compared to PVA‐M. Other PVA isolates that infected N. physaloides systemically also contained Metll6 in VPg. In a previous study using chimeric viruses, Metl16 in VPg was shown to be a major determinant for vascular movement of PVA in N. physaloides, and this study reveals that the mutation for Metl16 can occur in vivo during replication of the avirulent PVA‐M in infected plants. Immunolocalisation studies on BP and BPA plants showed that the pods (berries) and seed coat contained PVA‐Mm in the developing seeds, but no virus was detected in embryons. Up to 27% of the mature seeds contained PVA‐Mm but no transmission to seedlings was observed in a total of 450 seeds tested, and no test plants were infected following mechanical inoculation with extracts prepared from the seeds.

Localization of a Potyvirus and the Viral Genome-Linked Protein in Wild Potato Leaves at an Early Stage of Systemic Infection

The upper noninoculated 'sink' leaves of the wild potato species, Solanum commersonii, were studied for distribution of Potato virus A (PVA) at an early stage of systemic infection. Viral RNA was detected by in situ hybridization, and five viral proteins were localized using immunohistochemical staining in leaf sections. Initial systemic infection foci were found at the vicinity of major and minor veins. In these infection foci, the viral coat protein, cylindrical inclusion protein, and helper component-proteinase colocalized with viral RNA in parenchyma and mesophyll cells, but none of these were detected in companion cells (CC). In contrast, VPg, which is the N-proximal half of the NIa protein (separated from the C-terminal proteinase domain, NIapro, by an autocatalytic cleavage) and acts as a viral genome-linked protein, was detected in CC in the infection foci, but only at an early stage of virus unloading. Outside the infection foci, conspicuous signals for VPg were readily and exclusively detected in CC of many veins in all vein classes in the absence of signals for NIapro, other viral proteins, and viral RNA. Taken together, our data indicate that both major and minor veins may unload PVA in the sink leaves of potato. The data suggest that VPg is translocated from inoculated source leaves to the sink leaves, where it accumulates in CC at an early stage of systemic infection. These findings suggest that VPg may be a 'phloem protein' that specifically acts in CC in the sink leaves to facilitate virus unloading.

A new pathotype of Pea seedborne mosaic virus explained by properties of the P3-6k1- and viral genome-linked protein (VPg)-coding regions.

A fourth pathotype of Pea seedborne mosaic virus, a member of the genus Potyvirus, was identified by analysis of the infection profile on a panel of Pisum sativum lines. The new pathotype, designated P-3, was able to overcome resistance specified by the sbm-1 resistance gene but could not overcome resistance specified by the sbm-2 resistance gene. This infection profile distinguished P-3 from previously described pathotypes, P-1, P-2, and P-4. Analysis of chimeric viruses demonstrated that properties of the P3-6k1- and viral genome-linked protein (VPg)-coding regions accounted for the infection profile of the new pathotype.

Viral genome-linked protein (VPg) controls accumulation and phloem-loading of a potyvirus in inoculated potato leaves.

The viral protein covalently linked to the 5' end of the plus-sense, single-stranded RNA genome of potyviruses (genus Potyvirus) can be an avirulence determinant in incompatible potyvirus-host combinations in which the resistance prevents systemic virus infection. The mechanism is not well known. This study shows that virus strain-specific resistance to systemic infection with Potato virus A (PVA) in Solanum commersonii is overcome by a single amino acid (aa) substitution, His118Tyr, in the viral genome-linked protein (VPg). Virus localization and other experiments revealed that Tyr118, controls phloem loading of PVA. The critical boundary may be constituted in phloem parenchyma, companion cells, or both. Tyr118 also controls the cellular level of virus accumulation in infected leaves, including phloem cells. Amino acid substitutions at three additional positions of the central part (aa 116) and C terminus (aa 185) of the VPg and of the N terminus of the 6K2 protein (aa 5) affect virus accumulation and rate of systemic infection but are not sufficient for phloem loading of PVA. These data, together with previous studies, indicate that the PVA VPg aa residues crucial for systemic infection are host specific. Also, our data and previous studies on other potyvirus-host species combinations indicate that the central part of the VPg is a domain with universal importance to virus-host interactions required for systemic invasion of plants with potyviruses.

VPg unlinkase, the phosphodiesterase that hydrolyzes the bond between VPg and picornavirus RNA: a minimal nucleic moiety of the substrate.

VPg unlinkase is an unusual eukaryotic enzyme that catalyzes hydrolysis of the phosphodiester bond between residues of the unique tyrosine of VPg (viral protein genome-linked) and the 5;-terminal uridylic acid of picornavirus RNA. Cellular targets of the VPg unlinking enzyme are yet unknown. To determine an essential nucleic part of the covalent linkage unit that is necessary for the VPg unlinkase reaction, the following derivatives of the encephalomyocarditis virus (EMCV) VPg-RNA complex were used: [125I]Kp-pUpUpGp, [125I]Kp-pUp, and [125I]Kp-pU (Kp is residual peptides bound to RNA after proteinase K treatment of VPg-RNA). [125I]K-peptides were unlinked from [125I]Kp-pUpUpGp and [125I]Kp-RNA with similar velocity, but [125I]Kp-pUp was split much slower. Under the same conditions [125I]Kp-pU was not dissociated at all. Thus, pUp is a minimal part of picornavirus RNA that is necessary for VPg unlinkase. We speculate that cellular substrates of the enzyme are phosphodiesters of oligo(poly)ribonucleotides and tyrosine or tyrosine peptides. In no case [125I]VPg-pU, [125I]VPg-pUp, and [125I]VPg-pUpUpGp were hydrolyzed by VPg unlinkase, in contrast with [125I]VPg-RNA and [125I]VPg-pUpUpGpApApApGp. We conclude that the whole VPg, when bound to trinucleotide (but not to heptanucleotide), protects the inter-polymeric phosphodiester bond against hydrolysis of the covalent linkage unit. We speculate that VPg unlinkase might repair covalent complexes of RNA and topoisomerases and trigger degradation process of the picornavirus RNA.

Characterization and complete nucleotide sequence of Strawberry mottle virus: a tentative member of a new family of bipartite plant picorna-like viruses.

An isolate of Strawberry mottle virus (SMoV) was transferred from Fragaria vesca to Nicotiana occidentalis and Chenopodium quinoa by mechanical inoculation. Electron micrographs of infected tissues showed the presence of isometric particles of approximately 28 nm in diameter. SMoV-associated tubular structures were also conspicuous, particularly in the plasmodesmata of C. quinoa. DsRNA extraction of SMoV-infected N. occidentalis yielded two bands of 6.3 and 7.8 kbp which were cloned and sequenced. Gaps in the sequence, including the 5' and 3' ends, were filled using RT-PCR and RACE. The genome of SMoV was found to consist of RNA1 and RNA2 of 7036 and 5619 nt, respectively, excluding a poly(A) tail. Each RNA encodes one polyprotein and has a 3' non-coding region of approximately 1150 nt. The polyprotein of RNA1 contains regions with identities to helicase, viral genome-linked protein, protease and polymerase (RdRp), and shares its closest similarity with RNA1 of the tentative nepovirus Satsuma dwarf virus (SDV). The polyprotein of RNA2 displayed some similarity to the large coat protein domain of SDV and related viruses. Phylogenetic analysis of the RdRp region showed that SMoV falls into a separate group containing SDV, Apple latent spherical virus, Naval orange infectious mottling virus and Rice tungro spherical virus. Given the size of RNA2 and the presence of a long 3' non-coding region, SMoV is more typical of a nepovirus, although atypically for a nepovirus it is aphid transmissible. We propose that SMoV is a tentative member of an SDV-like lineage of picorna-like viruses.

Homology-dependent resistance to tomato ringspot nepovirus in plants transformed with the VPg–protease coding region1

The coding region for the VPg (genome-linked viral protein) and protease (Pro) of a raspberry isolate (Rasp-2) of tomato ringspot virus (ToRSV) was cloned into a plant expression vector and transformed into Nicotiana benthamiana using Agrobacterium tumefaciens. Transgenic lines were also produced that expressed two mutant derivatives of the VPg–Pro coding region (i.e., containing point mutations in the putative catalytic triad and substrate-binding pocket of the protease). A large proportion of the R1 transgenic lines displayed extreme resistance to ToRSV (raspberry isolate Rasp-1, 99% homology to the transgene at the nucleotide level), especially when transformed with the wild-type VPg–Pro coding region. Several transgenic lines displaying extreme resistance to ToRSV were further tested at the R2 generation. These R2 lines were resistant to Rasp-1 isolate but were susceptible to GYV isolate (79% homology to the transgene). These transgenic lines were also resistant to a potato virus X (PVX) hybrid containing the VPg–Pro coding region of ToRSV (Rasp-2 isolate) but were susceptible to the wild-type PVX. These results imply a role for homology-dependent silencing as a mechanism for resistance in these lines.

Identification of the Amino Acid Residue Involved in Rabbit Hemorrhagic Disease Virus VPg Uridylylation

The virus genome-linked protein (VPg) coding region from rabbit hemorrhagic disease virus (RHDV) (isolate AST/89) was expressed in Escherichia coli by using a glutathione S-transferase-based vector. The recombinant polypeptide could be purified in good yields and was uridylylated in vitro from [alpha-32P]UTP in a reaction catalyzed by the recombinant RNA-dependent RNA polymerase from RHDV in the absence of added template RNA. The use of deletion and point mutants allowed the identification of Tyr-21 as the residue involved in uridylylation and consequently in the linkage between VPg and the viral genome. These data constitute the first report on the identity of the amino acid residue involved in VPg uridylylation in a member of the Caliciviridae family.

Mutations in pea seedborne mosaic virus genome-linked protein VPg after pathotype-specific virulence in Pisum sativum.

Pisum sativum plant introduction (PI) line 269818 is resistant to potyvirus pea seedborne mosaic virus (PSbMV) isolates, categorized as pathotype P1, and is susceptible to pathotype P4 isolates. This difference in infectivity is determined by the viral genome-linked protein (VPg) cistron. Mutational analysis of VPg of PSbMV isolates DPD1 and NY representing pathotypes P1 and P4 revealed that codon changes affecting amino acids 105 to 117 in the central region of VPg influenced virulence on PI 269818. In contrast, infectivity on pea cultivar Dark Skinned Perfection, which is susceptible to both pathotypes, was not affected by the mutations. Mutants overcoming resistance in PI 269818 were analyzed for changes in the VPg coding region upon passage through PI 269818 and Dark Skinned Perfection. Adaptive changes were observed only upon passage through PI 269818 and only at codons from amino acid 105 to 117. Expression of DPD1 VPg in PI 269818 did not affect infection by NY, which suggests that VPg from DPD1 is not an elicitor of a general resistance response. The results are compatible with the hypothesis that viral amplification depends upon the interaction between VPg and a host factor.

Proteins attached to viral genomes are multifunctional

Publisher Summary This chapter discusses the nature and function of eukaryotic viral genome-linked proteins (VPgs) and outlines some of the remarkable features of these proteins. The first indications that a viral protein could form a stable complex with the genome of eukaryotic viruses were based on electron microscopy analyses and various biochemical techniques and were obtained with Adenoviridae. In RNA viruses, the term viral genome-linked protein was coined for the poliovirus protein and has been adopted for virtually all the proteins linked to RNA viruses. For the sake of simplification and to facilitate comprehension, the RNA and the DNA viruses whose genome possesses a covalently bound protein have been treated separately. In positive single-stranded RNA viruses, the VPg is always synthesized as part of a larger polyprotein from which it is excised by a viral proteinase that is also part of the polyprotein. The genomes of certain phages and viruses have long been known to covalently bind a viral protein. This feature was subsequently shown to occur also in cell systems. It is thus not an exclusive property of viruses. Indeed, extensive work on cellular topoisomerases, and the tumor suppressor p53 has demonstrated that these proteins can be covalently bound to a nucleic acid.

Recessive and dominant genes interfere with the vascular transport of Potato virus A in diploid potatoes.

Resistance to Potato virus A (PVA) was examined in a diploid cross involving Solanum tuberosum subsp. andigena as a resistance source. Hypersensitive resistance (HR) to PVA cosegregated with extreme resistance (ER) to Potato virus Y conferred by the dominant gene Ry(adg) on chromosome XI. Hence, HR to PVA was controlled by a novel, dominant resistance gene closely linked to Ry(adg), or Ry(adg) recognized both viruses but conferred a different type of resistance to each virus. The HR prevented systemic infection with PVA following mechanical inoculation but not following graft inoculation. Another, recessive gene, ra, that may be linked or even allelic with Ry(adg) fully blocked vascular transport of PVA in graft-inoculated plants. Hence, a possibility exists that the genes for the three types of resistance to potyviruses may reside at the same, resistance gene-rich chromosome region syntenic in solanaceous species and might be related. The gene ra acted against all of the three PVA strains tested and, therefore, the avirulence determinants could not be mapped. However, also, PVA strain-specific resistance was found in the progeny. It was overcome by mutations introduced into the viral genome-linked protein and the helper component proteinase and/or the coat protein.

The 6K2 protein and the VPg of potato virus A are determinants of systemic infection in Nicandra physaloides.

Infection with the isolate PVA-M of potato virus A (PVA; genus Potyvirus) is restricted to the inoculated leaves of Nicandra physaloides (Solanaceae), whereas the isolate PVA-B11 infects plants systemically by 10 days post inoculation. Resistance to systemic infection was shown to develop during plant growth. A recombinant virus (B11-M) in which a 1,208-nucleotide sequence of the full-length cDNA clone of PVA-B11 was replaced with the corresponding sequence from PVA-M displayed a phenotype similar to that of PVA-M. The replaced sequence contained four amino acid differences between the two isolates: one in the 6K2 protein and three in the viral genome-linked protein (VPg). Site-directed mutagenesis of the cDNA clones and inoculation of the mutants to N. physaloides indicated that the amino acid substitutions of Met5Val in the 6K2 protein or Leu185Ser in the VPg permitted vascular movement and systemic infection. However, resistance was only partially overcome by these changes, since systemic infection proceeded at a slower rate than with PVA-B11. The amino acid substitution Val116Met in the VPg alone was sufficient to overcome resistance and recover the phenotype of the isolate PVA-B11. These data indicated that both the 6K2 protein and the VPg were avirulence determinants of PVa-M in N. physaloides and suggested a possibly coordinated function of them in the vascular movement of PVA.

Biochemical and Genetic Evidence for Interactions between Potato A Potyvirus-Encoded Proteins P1 and P3 and Proteins of the Putative Replication Complex

Interactions of the first and third proteins (P1 and P3) of the potato A potyvirus (PVA) with the other six main proteins of PVA were studied using Escherichia coli-expressed recombinant proteins in two in vitro interaction assays and a genetic assay yeast two-hybrid system (YTHS). In overlay blotting and binding assays in liquid, P1 and P3 interacted with each other and with proteins of the putative replication complex of potyvirus: RNA-helicase (CI), viral protein genome-linked (VPg), NIa proteinase part (NIaPro), and RNA-dependent-RNA-polymerase (NIb). In addition, P1 self-interaction and interaction with helper-component proteinase (HC-Pro) also were detected. Neither P1 nor P3 interact with coat protein (CP) or with various control proteins. In the YTHS, P1 interacted only with CI and P3 with NIb. The different results obtained using the two test systems may reflect changes in interactions at different stages of potyvirus infection: in the virus genome replication and the virion accumulation stages when nonstructural proteins form inclusions. Our data are consistent with previous functional data, indicating that P1 and P3 proteins are involved in potyvirus genome amplification and provide the first direct evidence that these proteins interact with the proteins that have been shown to be part of the replication complex.

Proteolytic processing of tomato ringspot nepovirus 3C-like protease precursors: definition of the domains for the VPg, protease and putative RNA-dependent RNA polymerase.

Tomato ringspot nepovirus (TomRSV) RNA-1 encodes a putative NTP-binding protein (NTB), a putative viral genome-linked protein (VPg), a putative RNA-dependent RNA polymerase (Pol) and a serine-like protease (Pro), which have been suggested to be involved in viral RNA replication. Proteolytic processing of protease precursors containing these proteins was studied in Escherichia coli and in vitro. The TomRSV protease could cleave the precursor proteins and release the predicted mature proteins or intermediate precursors. Although processing was detected at all three predicted cleavage sites (NTB-VPg, VPg-Pro and Pro-Pol), processing at the VPg-Pro cleavage site was inefficient, resulting in accumulation of the VPg-Pro intermediate precursor in E. coli and in vitro. In addition, the presence of the VPg sequence in the precursor resulted in increased cleavage at the Pro-Pol cleavage site in E. coli and in vitro. Direct N-terminal sequencing of the genomic RNA-linked VPg, of the mature protease purified from E. coli extracts and of radiolabelled mature polymerase purified from in vitro translation products revealed the sequences of the NTB-VPg, VPg-Pro and Pro-Pol dipeptide cleavage sites to be Q/S, Q/G and Q/S, respectively. In vitro processing at the NTB-VPg and Pro-Pol cleavage sites was not detected upon mutation or deletion of the conserved glutamine at the -1 position of the cleavage site. These results are discussed in light of the cleavage site specificity of the TomRSV protease.

Sequencing, genomic localization and initial characterization of the VPg of pea enation mosaic enamovirus.

The amino acid sequence of the genome-linked viral protein (VPg) of pea enation mosaic enamovirus (PEMV) has been determined. The VPg is encoded by nt 1811-1894 within ORF1 of RNA1 downstream of the proteinase motif. Direct N terminus sequencing of intact and endoproteinase Asp-N-digested VPg combined with electrospray mass spectroscopy confirmed that the VPg is composed of 28 amino acids with a molecular mass of 3138 Da. The context of the N and C terminus residues as well as the position and size of the VPg suggest that the mature VPg may be generated via post-translational proteolytic processing of the polyprotein arrangement of membrane anchor-proteinase-VPg-polymerase encoded by ORFs 1 and 2. Computer comparisons did not reveal any significant similarity between the VPg of PEMV and any other sequences including those of the VPgs of related subgroup II luteoviruses.

Expression of poliovirus P3 proteins using a recombinant vaccinia virus results in proteolytically active 3CD precursor protein without further processing to 3C pro and 3D pol

The expression of the poliovirus genome occurs by the translation of a single open reading frame to generate a long polyprotein which is subsequently processed by viral encoded proteases. The initial proteolytic cleavages result in the production of a P1 polyprotein which contains the capsid proteins, and the P2 and P3 polyproteins which contain proteins required for replication. The P3 polyprotein consists of the 3AB protein (containing the viral genome-linked protein, VPg), the viral protease, 3C pro, and RNA polymerase, 3D pol. To further study the expression and proteolytic processing of poliovirus P3 proteins in vivo, we have utilized recombinant vaccinia virus vectors to express nucleotides 5240–7400 containing the P3 region proteins of poliovirus. The P3 protein expressed from the recombinant vaccinia virus VV-P3 exhibited in vivo proteolytic activity as evident by processing of the polyprotein to generate the 3CD protein, consisting of a fusion between the 3C pro and 3D pol proteins. Further processing of the 3CD protein to 3C pro and 3D pol, however, was not detected in cells infected with VV-P3. Subcellular fractionation of VV-P3-infected cells demonstrated that the 3CD protein was present in both the soluble and membrane fractions. Finally, the 3CD protein expressed from VV-P3 was stable in cells co-infected with VV-P3 and poliovirus and no further processing to 3D pol was detected. These results are discussed with regards to in vivo studies which suggest that the 3CD polyprotein is not a precursor to 3D pol in poliovirus-infected cells.

Biologically active transcripts from cloned cDNA of genomic grapevine fanleaf nepovirus RNAs.

Transcripts were produced in vitro by run-off transcription from full-length cDNA of RNA1 and RNA2 of grapevine fanleaf nepovirus (GFLV; isolate F13) cloned downstream from a bacteriophage RNA polymerase promoter. These transcripts, which possess a 5' terminal cap structure and a non-viral G residue instead of the naturally occurring genome-linked viral protein (VPg), are infectious to Chenopodium quinoa protoplasts when inoculated by electroporation. Synthetic RNA1 alone replicated in protoplasts. Inoculation of C. quinoa plants with synthetic RNA1 plus RNA2 produced symptoms similar to, but weaker, than those observed in plants infected with natural GFLV 6 to 8 days post-inoculation. Co-inoculated RNA1 and RNA2 were able to replicate and spread systemically in plants but RNA1 alone produced no symptoms and was not detected in non-inoculated leaves, suggesting that virus spread requires RNA2. Analysis of the genomic RNAs in plants infected with transcripts showed that the non-viral G at their 5' ends was not retained in the progeny.