Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Potato leafroll virus : a classic pathogen shows some new tricks

Potato leafroll virus : a classic pathogen shows some new tricks INTRODUCTION The key advances in the understanding of the disease of potato leafroll have been reviewed in detail by Barker (1992 ). Probably one of the earliest records of disease problems caused by a plant virus was that of potato leafroll resulting from infection with Potato leafroll virus (PLRV), thus, infection with PLRV was the cause, or at least a factor in, a condition affecting potato termed the Curl recorded in Lancashire (UK) in the 1760s where it threatened potato cultivation in the 1770s. The infectious nature of potato leafroll disease was first described in 1916. The important discovery that aphids transmit PLRV was made in the 1920s. PLRV particles were first purified in the 1960s and an enzyme‐linked immunosorbent assay was developed in the 1970s to detect virus using polyclonal antisera. As with other plant viruses, the publication of genome sequences has led to much research on virion and genome structure and genome function. This review is focused mainly on the most recent developments in these areas. Taxonomy and sequence variation PLRV is often described as a luteovirus. This refers to its classification in family Luteoviridae ( D’Arcy ., 2000 ), which is largely on the basis of its transmission properties and the molecular and structural features of its virion. Within the family there are two main genera ( Luteovirus and Polerovirus ) that differ in the nature of the RNA‐dependent RNA polymerase and certain genome features. PLRV is the type species of the genus Polerovirus . Other well studied poleroviruses are Beet western yellows virus (BWYV), Cereal yellow dwarf virus‐RPV (CYDV‐RPV) (formerly Barley yellow dwarf virus , BYDV‐RPV), Beet mild yellowing virus (BMYV) and Cucurbit aphid‐borne yellows virus (CABYV). Many of the molecular features of PLRV are likely also to be broadly applicable to these poleroviruses. In contrast to those of some other poleroviruses, isolates of PLRV differ little in sequence from each other. In a recent study among 12 complete sequences ( Guyader and Ducray, 2002 ), the identity was 94%−98% over all open reading frames and comparisons among ORF 3 (coat protein, CP) sequences of these and a further seven isolates did not alter the pattern. Similarly, in comparisons among isolates obtained in Peru there was little variation in ORF 3, and no greater variation when they were compared with published sequences (MAM and L. Torrance, unpublished). In contrast, Guyader and Ducray (2002 ) found that comparisons among 19 isolates in the region of ORF 0 and the 5′ part of ORF 1 resolved them into three groups, one of Australian isolates, one containing a Peruvian and three European isolates and a third containing an isolate from Australia, a Cuban isolate and several from Europe. Thus there was no geographical correlation with sequence variation. The lack of sequence variation suggests either that PLRV only recently diverged from an ancestral virus (for example by acquiring an ability to infect potato) or that it has been subject to very strong selection constraints. The narrow genetic base of potatoes in cultivation may also be a factor in restricting PLRV variation ( Guyader and Ducray, 2002 ). The virus that most closely resembles PLRV in CP sequence is Sweet potato leaf speckles virus , which is also from South America ( Fuentes ., 1996 ). It is 70% identical to PLRV whereas other poleroviruses are no more than 65% identical. ORGANIZATION AND EXPRESSION OF THE PLRV GENOME The genome of PLRV comprises eight ORFs ( Fig. 1 ). It is densely packed, with ORF 1 overlapping much of ORF 0, and ORF 4 being entirely within ORF 3. In common with genomes of all luteoviruses ( Mayo and Miller, 1999 ), almost all types of modulation mechanism (frameshift, initiation bypass, termination suppression, production of subgenomic (sg) RNA, proteolysis of primary translation products) are used during the expression of the different ORFs. The three 5′‐proximal ORFs (ORFs 0–2) are expressed from the genomic RNA and five 3′‐proximal ORFs (ORFs 3–7) are expressed by translation from subgenomic mRNAs that are 3′ co‐terminal with genomic RNA. Thus luteoviruses present amongst the best examples of the versatility of expression strategy and the economy in use of coding sequence charactersitic of small RNA virus genomes. 1 The genome of PLRV showing different expression strategies: production of subgenomic mRNAs (sg1 and sg2), initiation at the first AUG (P0, P3 and P6), bypass and initiation at the second AUG (P1, P4 and P7), initiation as for P1 and frameshift (P1 + P2), initiation as for P3 and readthrough suppression of the UAG of ORF 3, proteolysis of the primary translation product (P1 and possibly P1 + P2). Lines represent RNA molecules, open boxes represent open reading frames (ORFs) and grey bars represent translation products. The first nine nucleotides of the 5′ non‐translated leader sequences are the same in genome RNA and sgRNA1 ( Miller and Mayo, 1991 ), but those of sgRNA2 are different. In vitro , leader sequences have effects on translation of the RNAs but are not translational enhancers and the non‐translated sequence had no effect on translation ( Juszczuk ., 2000 ). The initiation codon of ORF 4 is a few nucleotides downstream of that of ORF 3 ( Fig. 1 ). The ratio between P3 and P4 produced by translation of sgRNA1 has been reported to be c . 1 : 1 ( Juszczuk ., 2000 ) or c . 1 : 7 ( Tacke ., 1990 ). The functions of some PLRV proteins are known, but for others functions are speculative (reviewed in Mayo and Ziegler‐Graff, 1996 ). P0 is involved in symptom development and is a suppressor of gene silencing (see below); P1 has proteinase functions and contains the genome‐linked protein (VPg) ( Van der Wilk ., 1997 ), and possibly has other functions such as a helicase; P2 is an RNA‐dependent RNA polymerase; P3 is the capsid protein of c . 23 kDa; P4 is a putative movement protein; suppression of the ORF 3 amber stop codon results in translation of an ORF 3/ORF 5 read‐through protein that is a minor structural protein of c . 80 kDa that may be involved in aphid transmission. ORFs 6 and 7 have been demonstrated more recently and encode proteins of unknown function but P7 displays nucleic acid binding properties ( Ashoub ., 1998 ). Particle structure Protein composition In electron microscope preparations PLRV particles are isometric, 24 nm in diameter and sometimes appear to have small projections on the vertices. The sedimentation rate of particles ( s 20w = 115 S ), their diameter and the size of the major coat protein (P3) suggest that particles contain 180 copies of P3 in a T = 3 structure. About 1% to 5% of the P3 molecules in protein extracted from purified virus particles carry a C‐terminal extension derived from P5 ( Bahner ., 1990 ), although only a minor fraction of the particle protein, P5 is immunodominant. In one panel of monoclonal antibodies, 2 of 10 antibodies were P5‐specific ( Massalski and Harrison, 1987 ; MAM unpublished) and in another panel most were specific for P5 (L Torrance, V Flores and MAM unpublished). The size of the P3 + P5 protein in purified particles is c . 52 kDa but in extracts from infected cells the fusion protein is c . 80 kDa. P3 + P5 was observed to degrade from c . 80 kDa to c . 50 kDa after PLRV‐infected protoplasts were lysed ( Bahner ., 1990 ). The usual explanation for this discrepancy in size is that part of P5 is lost during purification, but this interpretation is not consistent with the observation that antibody panels raised to purified particles contain antibodies that react with parts of P5 close to the C‐terminal end ( Jolly and Mayo, 1994 ). It appears that the cleaved portions of P5 remain attached to purified PLRV particles but are removed when particles are denatured in SDS. If true, this suggests a relatively high affinity between P5 sequences and the surface portions of P3. The functional significance of this unusual interaction is unclear. Structure predictions Attempts have been made to predict the secondary structure of P3 from an analysis of its amino acid sequence. The prediction that the structure contained a basic N‐terminal region followed by patches of β‐sheet was made shortly after sequences were published ( Torrance, 1992 ) and subsequent predictions based on more refined software packages have supported this model ( Mayo and Ziegler‐Graff, 1996 ). The most recent of these ( Terradot ., 2001 ), and by far the most sophisticated, was made following the observation of a weak (17% identity, 33% similarity), but significant, resemblance between the sequences of the coat proteins (P3) of PLRV and the sobemovirus Rice yellow mottle virus (RYMV). Because a crystal structure was known for RYMV particles, the alignment of the PLRV and RYMV CPs allowed the construction of a model for PLRV CP and those of other luteoviruses. The interesting features to emerge are models for some epitopes and a clustering of acidic residues on the three‐fold axis of the virus particles. These are illustrated in Fig. 2 . 2 Models of the predicted structure of a reconstituted trimer of PLRV coat protein. With permission from Terradot . (2001 ; Academic Press Inc.). (A) Surface (left) or section (right) views. Monomer units are in light green, dark green or blue. Epitope 5 (residues 83–89) is shown in red, epitope 120 (residues 172–178) is shown in yellow. (B) Surfaces coloured according to the properties of the constituent amino acids: D, E—magenta; R, H, K—blue; A, V, L, I, P, M, F, W—orange; G, S, T, C, N, Q, Y—green. The acidic patch on the trimer surface is circled. Mapping the known epitopes in the PLRV CP ( Torrance, 1992 ) on to the structure model located epitope 5 as the loop between β‐sheets B and C, and epitope 10 as the G‐H loop and both map on to the reconstituted trimer ( Fig. 2A ). Epitopes are regions of sequence that are intrinsically mobile and are therefore candidates for surface regions that contribute to biological properties. The reconstituted trimer for all polerovirus CPs contains an acidic patch where E and D residues come into close proximity (magenta shading in Fig. 2B ). Possibly the acidic patch is involved in the retention of parts of P5 on the surface of virus particles. Another possibility is that it is involved in the switch of the N‐terminal sequence of the CP (a major epitope) from inside the particle to outside, a change postulated to be linked with particle disassembly ( Torrance, 1992 ). The patch is less prominent in the analogous model of CP trimers of BYDV‐MAV and BYDV‐PAV (genus Luteovirus ) and is not present in models of sobemovirus CP trimers. When expressed in insect cells, P3 did not form particles, appearing to denature into amorphous aggregates. However when the P3 molecule carried an N‐terminal extension that included six histidine residues, the protein formed virus‐like particles (VLP) and these formed into crystal‐like aggregates in the nuclei of the insect cells expressing the protein ( Lamb ., 1996 ). The particles proved very difficult to extract from the cells and seemed to be less stable during storage at −70 °C than were PLRV particles. The VLP contained nucleic acid unrelated to PLRV RNA suggesting that there is no absolute requirement for an initiation‐of‐assembly signal, but that polynucleotide material is needed for particle stability. BIOLOGY—SYMPTOMS AND PATHOLOGY PLRV is found wherever potatoes are grown, but its importance and relative abundance varies. Symptoms may be more or less severe depending on the isolate, cultivar, growing conditions and age of the plant when it was infected. Infection of potato can cause severe yield loss and quality reduction of the harvested tubers, and in some regions infection levels can be high and economic losses can be serious. In plants grown from infected tubers (secondary infection), shoots are stunted and an upward rolling of leaflets is readily apparent, especially those on lower leaves, which break easily when crushed and may be chlorotic. Symptoms of primary infection (current season infection) are usually less severe unless plants became infected early in the season. In normal circumstances, the virus is confined to phloem tissue and cannot be transmitted by inoculation of sap. Replication probably occurs in phloem companion cells and virus moves long‐distance within the phloem sieve elements ( Taliansky and Barker, 1999 ). VIRUS TRANSMISSION—VECTOR RELATIONSHIPS Several potato colonizing aphid species transmit PLRV. Myzus persicae is regarded as the most efficient vector and larval and adult stages of winged and non‐winged forms transmit PLRV. Transmission is persistent and non‐propagative (see for review, Herrbach, 1999 ). Thus, virus is ingested with phloem contents during feeding and particles move from the gut to the salivary glands via the haemolymph (virus acquisition). There is no evidence that PLRV multiplies within the aphid vectors and presumably PLRV particles remain intact throughout the route through the aphid body. The minimum feeding times for virus acquisition and transmission are each about 1 h, although transmission frequency increases with increase in the acquisition period up to 2 days or more. Vector efficiency depends not only on aphid species, but also on clones, biotypes, morphs and instars, with as much variation between as within aphid species (reviewed by Robert and Lemaire, 1999 ). Although much is known about the complex series of processes of the virus acquisition/transmission, many details and questions remain unclear or unanswered ( Reavy and Mayo, 2002 ). Specific interactions between the PLRV particles and receptors in the aphid vectors probably occur at several stages, and all of these are likely to affect the outcome of this process. The P5 ( Guilley ., 1994 ) of luteoviruses has been proposed to play a role in such a specific interaction, however, the position regarding the role of the P5 in aphid transmission is complex and full of contradictions. What is clear is that mutants lacking the P5 are not aphid‐transmissible. It has been suggested that the longevity of infective virus particles in the aphid haemolymph may be mediated by an association with symbionin, a protein produced by the endosymbiotic bacterium ( Buchnera sp.) of M. persicae. Symbionin has been proposed to bind to PLRV particle protein and protect it from proteolysis ( Hogenhout ., 1998 ), however, the binding of symbionin to virus particles or to the P5 protein has only been demonstrated in vitro and the exact role of symbionin in virus transmission needs to be resolved ( Reavy and Mayo, 2002 ). Moreover, the VLPs of PLRV CP formed in Spodoptera frugiperda cells infected with a baculovirus carrying the CP (P3) gene of PLRV fused to a histidine tag lacking RT protein, were able to move through aphids from the gut to the salivary duct lumen ( Gildow ., 2000 ). Thus the dependency of infection of the presence of P5 may be better explained by P5 being involved in the initiation of infection within the plant. However, experiments with mutants of BWYV showed that those lacking the P5 did not bind to accessory salivary glands ( Reinbold ., 2001 ), suggesting that P5 has a role in this process. Although the P5 of PLRV and BWYV may differ, it seems more probable that P5 has more than one role. It is clear that different regions of the PLRV particle (such as the acidic patch on P3 trimers or P5) are important in recognition events at different stages of the transmission process, although the precise role and function of the proteins in the processes of aphid transmission and establishment of infection in plants remains to be elucidated. Therefore, an important area to investigate will be the nature and role of specific aphid virus receptor molecules that mediate the attachment and specific movement of PLRV particles through the various cellular layers and membranes within the aphid as virus moves from the gut to eventually emerge in the saliva to infect a new plant. The identification and investigation of these receptors may provide clues on how to interfere with this process and prevent virus transmission. VIRUS MOVEMENT Typically, after initial infection, plant viruses replicate and spread locally, from cell to cell through the plasmodesmata until they reach the phloem. The virions then enter the host vascular system, invade different types of phloem‐associated cells such as phloem parenchyma and companion cells, penetrate into the sieve elements and rapidly move through them (from leaf to leaf) and exit to infect non‐vascular tissues. This is long‐distance movement (reviewed in Oparka and Santa Cruz, 2000 ; Rhee ., 2000 ). Cell‐to‐cell movement usually involves one or more virus‐encoded movement proteins (MPs) and host components, such as tubulins, protein kinases and pectin methylesterases ( Boyko ., 2000 ; Chen ., 2000 ; Dorokhov ., 1999 ; Waigmann ., 2000 ), but less is known about long‐distance transport. Luteoviruses represent a special situation, because their accumulation and spread are mainly limited to the phloem cells (see below). PLRV, like other luteoviruses, cannot be transmitted mechanically, and infects plants only when delivered into phloem tissues (by aphids, agroinoculation or grafting). Within the phloem, PLRV can invade phloem parenchyma and companion cells and spread through the sieve elements. Thus, the process of luteovirus movement may resemble just one phase of the spread of other plant viruses, namely phloem‐associated long‐distance movement and thus represents a ‘pure’ experimental system for studying phloem‐associated movement without the complication of effects from short‐distance mesophyll cell‐to‐cell movement. However, the long‐distance movement function of PLRV has long been assigned to a 17‐kDa protein encoded by ORF 4 (P4, see above), that has properties strikingly similar to those of the Tobacco mosaic virus (TMV) 30 kDa MP involved in cell‐to‐cell movement, arguably the best studied of MPs. These properties for both of the proteins are single‐stranded RNA binding, protein dimerization, phosphorylation by a membrane‐bound protein kinase, plasmodesmal localization in both virus‐infected and transgenic plants and ability to increase plasmodesmal size exclusion limits ( Hofius ., 2001 ; Schmitz ., 1997 ; Sokolova ., 1997 ; Tacke ., 1991 ; Tacke ., 1993 ). Nevertheless, in spite of the similarities in properties between the TMV 30 kDa protein and PLRV P4, PLRV, unlike TMV, is unable to spread out of the plant vasculature (see below) and it is assumed that P4 mediates virus movement only between cells within the phloem tissues. This role for the PLRV P4 is supported by the observation that its analogue in the genome of another luteovirus, BYDV‐PAV was shown by mutational analysis to be a movement protein ( Chay ., 1996 ). Direct evidence for this role comes from experiments by Lee . (2002 ) using PLRV mutants in which ORF 4 encoding P4 was rendered either untranslatable or was modified to express the protein but lacking the first four amino acids. Both mutants were able to replicate and accumulate in agroinoculated leaves of potato and Physalis floridana, but were unable to move into vascular tissues and initiate a systemic infection in these plants, indicating the PLRV P4 is strictly required for virus movement. However, in two other plant species, Nicotiana benthamiana , and Nicotiana clevelandii , the same PLRV mutants were able to spread systemically. Moreover, Ziegler‐Graff . (1996 ), found that the BWYV analogue of the PLRV P4 was not essential for movement in N. clevelandii. These results suggest that the requirement for the P4 movement protein is host‐dependent and that there is a P4‐independent mechanism for PLRV movement that operates at least in some plants. Mutants lacking the 3′ parts of the P5 gene were found to be less infective and to accumulate to a lesser extent than unmodified PLRV (MAM and B. Reavy unpublished results). These observations mirror the results of more extensive studies with BWYV showing that mutations in different parts of the P5 gene had similar effects ( Bruyére ., 1997 ). With PLRV, direct indication that P5 is also involved in systemic spread, came from experiments using PLRV tagged with green fluorescent protein (GFP) ( Nurkiyanova ., 2000 ). The PLRV genome was modified by insertion of the gene for GFP into ORF 5 near its 3′ end. The chimeric PLRV‐GFP was able to replicate, express the GFP and form virus particles that contained P5 in the form of a fusion with GFP, in protoplasts or primarily infected cells. These virus particles were aphid‐transmissible but the GFP‐tagged PLRV was unable to spread systemically when agroinfiltrated or transmitted by aphids into plant hosts. Only naturally occurring mutants that lacked some or all of the inserted GFP sequences, were able to spread. These results support the idea that PLRV P5 is involved in virus movement and suggest that some putative ‘transport’ domain(s) in the P5 may be functionally disrupted by fusion to GFP. If so, it can be assumed that different functional domains are involved in aphid transmission and systemic spread of PLRV. PLRV virions have been detected in specialized plasmodesmata connecting sieve elements to companion and phloem parenchyma cells ( Esau and Hoefert, 1972 ; Shepardson ., 1980 ), which suggests that virions are the infectious entity engaged in vascular transport. Ziegler‐Graff . (1996 ) suggested that BWYV P5 domains on the surface of virus particles act as movement proteins to mediate the P4‐independent penetration into and translocation through sieve elements of virions. The P4‐dependent pathway, on the other hand, may not require P5 and presumably involves cell‐to‐cell transport of the virus through the plasmodesmata connecting nucleate phloem (companion and parenchyma) cells. However, this is clearly a host‐dependent process, because PLRV P4‐deficient mutants are unable to spread in potato or P. floridana (see above). PLRV AND RNA SILENCING Higher plants possess a specific defence mechanism against viruses that is similar to post‐transcriptional gene silencing (PTGS). PTGS or RNA silencing is a type of a sequence‐specific degradation of RNA molecules, a key element of which is the involvement of a dsRNA that corresponds to the target RNA. Such a dsRNA is produced during replication of RNA viruses. During silencing, dsRNA is cleaved into short interfering RNAs (siRNA), 21–25 nucleotides in length, and these are thought to mediate the specificity of RNA degradation (see for reviews, Carrington, 2000 ; Vance and Vaucheret, 2001 ; Voinnet, 2001 ). To counteract RNA silencing, some viruses encode specific proteins that suppress different steps in the silencing pathways (see for reviews, Carrington, 2000 ; Vance and Vaucheret, 2001 ; Voinnet, 2001 ) . For example, the effect of potyvirus helper component‐proteinase (HC‐Pro) is to reverse silencing in cells already silenced, suggesting that this protein suppresses a function required for the maintenance of the silenced state. The Potato virus X 25 kDa protein appears to target and interfere with the spread of the silencing signal. The Cucumber mosaic virus (CMV) 2b protein affects only the initiation of silencing, by inactivating the mobile silencing signal ( Guo and Ding, 2002 ). Although the P0 has never been detected in PLRV‐infected plants, there is strong genetic evidence that it is functionally important. Mutations that prevented expression of this protein in vitro , completely abolished PLRV accumulation in plant cells ( Sadowy ., 2001 ). Recently PLRV P0 (like HC‐Pro) has been shown to display significant silencing suppressor activity in a transient expression system based on its ability to inhibit PTGS of GFP in agro‐infiltrated leaves of transgenic plants containing a GFP transgene when coexpressed with silencing inducer (GFP‐coding nucleotide sequence) ( Pfeffer ., 2002 ). PLRV P0 can also act as an RNA silencing suppressor when expressed in Drosophila cells (B. Reavy and S. MacFarlane, personal communication). The P0 encoded by the other luteoviruses BWYV and CABYV have also been shown to suppress silencing ( Pfeffer ., 2002 ). The results of further experiments allowed the authors to suggest that P0 had interacted with a component of the silencing machinery responsible for cell‐autonomous silencing but did not interfere with production and delivery of the mobile silencing signal. However, marked PTGS suppressor activity was observed only when the P0 was expressed from monocistronic constructs. Full‐length PLRV (our unpublished results) or BWYV ( Pfeffer ., 2002 ) displayed much lower suppression activity, if any. Pfeffer . (2002 ) suggested that low PTGS suppression activity was a consequence of the low accumulation of the BWYV P0 protein resulting from the suboptimal translational context of its start codon in viral RNA. Moreover, mutations designed to optimize translation initiation efficiency in BWYV RNA were not stable during virus multiplication in plants, suggesting that there is a selection against over‐expression of the P0 and presumably its silencing suppression activity. Further experimentation is required to determine how the expression of the P0 is regulated during PLRV infection and its role in counter‐defence against RNA silencing. A new insight into the mechanisms of overcoming RNA silencing that may be exploited by luteoviruses, and PLRV in particular, came from experiments with transgenic plants that expressed the P5 sequence of BWYV ( Brault ., 2002 ). Following infection of these plants with BWYV or related viruses such as PLRV, transgenic P5 and its mRNA accumulated to only low levels and siRNAs corresponding to the transgene were generated. This suggested that RNA silencing had been induced by the infections (BWYV and PLRV nucleotide sequences contain some stretches of identical sequence that account for the ability of PLRV to induce silencing). As expected, the transgene (P5) mRNA, both within and outside of the phloem, was targeted for RNA silencing‐specific degradation. The viral RNA, however, was not subject to significant silencing and the transgenic plants were readily infected with either BWYV or PLRV. Thus a mechanism must exist that can discriminate between the viral and transgene RNA for degradation by the silencing machinery ( Brault ., 2002 ). Known virus‐encoded silencing suppressor proteins cannot differentiate viral and mRNA, rescuing both of them from silencing. Therefore, the suggestion that the putative P0 suppressor is involved in selective silencing suppression of the viral RNA (not affecting the degradation of transgene RNA) does not seem very likely. It is more likely that luteoviruses and PLRV in particular, exploit a novel and as yet unrecognized mechanism for the protection of viral RNA molecules from silencing by sequestering them in replication, translation, movement or other complexes, or as virions. MECHANISMS FOR PHLOEM RESTRICTION An intriguing feature of PLRV (and other luteoviruses), is the phloem restriction of the virus. PLRV cannot be transmitted mechanically, and infects plants only when delivered into phloem tissues (by aphids, agroinoculation or grafting). PLRV can replicate in mesophyll protoplasts and in primarily inoculated epidermal or mesophyll cells. Also, in plants PLRV is to some extent able to exit from the phloem into mesophyll cells when the PLRV‐infected plants are also infected with certain other plant viruses ( Atabekov ., 1984 ; Barker, 1987, 1989 ). These findings suggest that the normal restriction of PLRV to the phloem is because PLRV movement functions do not operate in the epidermis and mesophyll tissues (see for review, Taliansky and Barker, 1999 ) and/or because PLRV cannot suppress RNA silencing host defence in non‐vascular tissues ( Voinnet ., 1999 ; Waterhouse ., 1999 ). To test this hypothesis we designed a novel experimental approach to allow PLRV to spread out of the phloem and to become mechanically transmissible as a result of complementation by umbraviruses ( Mayo ., 2000 ; Ryabov ., 2001 ). In plants infected with PLRV and the umbraviruses Pea enation mosaic virus ‐2, or Groundnut rosette virus (GRV), PLRV accumulated in clusters of mesophyll cells in both inoculated and systemically infected leaves. No PLRV mechanical transmissions were obtained by coinoculation with CMV or some other viruses, although PLRV was transmissible from mixtures with a recombinant CMV that contained the MP gene (ORF 4) of GRV [CMV(ORF4)] ( Ryabov ., 2001 ). This suggested that the umbravirus‐dependent mechanical transmission and spread of PLRV in non‐vascular tissues depends on the umbravirus‐specific cell‐to‐cell movement function, however, a mutant of CMV(ORF4), in which the CMV 2b gene was untranslatable, was unable to help PLRV transmission and movement. Because the 2b protein is an RNA silencing suppressor ( Guo and Ding, 2002 ), we suggested that suppression of silencing is essential for the spread of PLRV in non‐vascular tissues. This is consistent with the recent observation that in transgenic plants that express full‐length PLRV cDNA in only a very limited number of the mesophyll cells, the expression of a potyvirus HC‐Pro resulted in a marked increase in the number of cells that accumulated PLRV ( Barker ., 2001 ; Franco‐Lara ., 1999 ). The key feature of these experiments was that PLRV RNA transcript was produced in each plant cell and infection did not depend on virus movement from cell to cell. In contrast, Savenkov and Valkonen (2001 ) showed that HC‐Pro, expressed in PLRV‐infected plants (i.e. in conditions when cell‐to‐cell movement is required for PLRV invasion into non‐vascular tissues), enhanced titres of PLRV only in the phloem system but not outside, suggesting that silencing suppression is not enough to allow PLRV to invade mesophyll tissues. Taken together, these results suggest that both the movement function and the overcoming of silencing are required for PLRV to become mechanically transmissible. Thus, confinement of PLRV, and presumably luteoviruses in general, to phloem tissues could be explained by (i) the failure of cell‐to‐cell movement and (ii) lack of mechanisms to overcome RNA silencing outside the phloem. In other words, we can assume that the PLRV‐encoded MPs (P4 and/or P5) are unable to fulfil their functions in the mesophyll tissues. Similarly, mechanism(s) used by PLRV for counter‐defence against RNA silencing (P0‐mediated suppression and/or protection against RNA degradation, see above) operate only in the phloem but not in non‐vascular tissues. Further studies will be required to clarify the nature of this tissue specificity. RESISTANCE Several potato ( Solanum tuberosum ) cultivars are known to have genetically controlled resistance to PLRV, but no cultivar is immune to infection and there are no known sources of major gene resistance. Several forms of resistance are expressed as a less than normal level of virus accumulation in infected plants ( Solomon‐Blackburn and Barker, 2001 ). The transfer of valuable traits from tuber‐bearing and non‐tuber‐bearing wild potato species and cultivated diploids to cultivated S. tuberosum potato has been a major effort of several breeding programmes. At least seven species are known to contain resistance to PLRV including Solanum brevidens, S. etuberosum, S. chacoense and S. raphanifolium that have very high levels of resistance ( Barker and Waterhouse, 1999 ). However, none of these forms of resistance have yet been transferred successfully to cultivated potato. Genetically engineered resistance to PLRV was first reported by Kawchuk . (1990 ) who transformed potato with the PLRV CP gene. Subsequently there have been many other examples of pathogen‐derived resistance to PLRV ( Barker and Waterhouse, 1999 ). Several of these are also expressed as resistance to virus accumulation and have a similar phenotype to naturally occurring forms of host resistance to PLRV. Transgenic potatoes resistant to PLRV were released commercially in the USA in 1997, although their use has declined recently. PROSPECTS Extensive studies of PLRV and plant diseases caused by this virus, have a very long history that makes PLRV a ‘classic’ pathogen. This paper describes some achievements in the understanding of the mechanisms used by PLRV to establish a systemic infection in plants, however, before assuming that we have a detailed picture of PLRV infection, it is important to consider the gaps in the current state of knowledge. Some of the gaps are listed in the sections above. Future research can provide both novel information on the molecular mechanisms of pathogenesis induced by PLRV as well as new tools for further studies. Indeed, this classic pathogen still displays some tricks making it an attractive experimental system for studies of some general biological processes such as macromolecular trafficking through the plant's long‐distance transport system, the phloem, and mechanisms of RNA silencing and silencing suppression. For example, if PLRV proves to move through the phloem by exploiting and modifying endogenous plant pathways for macromolecular movement, the virus would represent a ‘pure’ system that allows the analysis of the phloem‐associated transport without any possible effects from cell‐to‐cell movement. Another novel aspect of work with PLRV is the use of amplicons (virus‐based transgenes) as a tool for the study of RNA silencing. Amplicons were first described by Angell and Baulcombe (1997 ) for PVX, and then for other viruses including PLRV by Franco‐Lara . (1999 ) and Barker . (2001 ). An amplicon is a transgene that comprises a full‐length (biologically active) copy of a viral genome. In plants that contain an amplicon, transgene‐derived transcripts replicate like RNA viruses and this process could potentially take place in every cell of the plant, however, such replication also induces very effective silencing of the amplicon. More recently we have made an improved version of the PLRV amplicon expressing GFP to monitor virus expression. The advantages of the PLRV amplicon‐mediated RNA silencing system are: (i) the inability of PLRV to suppress or avoid silencing in non‐vascular tissues, and (ii) the lack of virus movement from cell to cell that allows the study of silencing of viral replicating RNAs in the absence of effects from silencing suppression and virus movement. PLRV amplicons will be used for studying mechanisms of RNA silencing and its suppression by different biotic or abiotic factors. ACKNOWLEDGEMENTS This work was supported by a grant‐in‐aid from the Scottish Executive Environment and Rural Affairs Department. http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Molecular Plant Pathology Wiley

Potato leafroll virus : a classic pathogen shows some new tricks

Loading next page...
 
/lp/wiley/potato-leafroll-virus-a-classic-pathogen-shows-some-new-tricks-dWviY04vjY

References (59)

Publisher
Wiley
Copyright
Copyright © 2003 Wiley Subscription Services, Inc., A Wiley Company
ISSN
1464-6722
eISSN
1364-3703
DOI
10.1046/j.1364-3703.2003.00153.x
Publisher site
See Article on Publisher Site

Abstract

INTRODUCTION The key advances in the understanding of the disease of potato leafroll have been reviewed in detail by Barker (1992 ). Probably one of the earliest records of disease problems caused by a plant virus was that of potato leafroll resulting from infection with Potato leafroll virus (PLRV), thus, infection with PLRV was the cause, or at least a factor in, a condition affecting potato termed the Curl recorded in Lancashire (UK) in the 1760s where it threatened potato cultivation in the 1770s. The infectious nature of potato leafroll disease was first described in 1916. The important discovery that aphids transmit PLRV was made in the 1920s. PLRV particles were first purified in the 1960s and an enzyme‐linked immunosorbent assay was developed in the 1970s to detect virus using polyclonal antisera. As with other plant viruses, the publication of genome sequences has led to much research on virion and genome structure and genome function. This review is focused mainly on the most recent developments in these areas. Taxonomy and sequence variation PLRV is often described as a luteovirus. This refers to its classification in family Luteoviridae ( D’Arcy ., 2000 ), which is largely on the basis of its transmission properties and the molecular and structural features of its virion. Within the family there are two main genera ( Luteovirus and Polerovirus ) that differ in the nature of the RNA‐dependent RNA polymerase and certain genome features. PLRV is the type species of the genus Polerovirus . Other well studied poleroviruses are Beet western yellows virus (BWYV), Cereal yellow dwarf virus‐RPV (CYDV‐RPV) (formerly Barley yellow dwarf virus , BYDV‐RPV), Beet mild yellowing virus (BMYV) and Cucurbit aphid‐borne yellows virus (CABYV). Many of the molecular features of PLRV are likely also to be broadly applicable to these poleroviruses. In contrast to those of some other poleroviruses, isolates of PLRV differ little in sequence from each other. In a recent study among 12 complete sequences ( Guyader and Ducray, 2002 ), the identity was 94%−98% over all open reading frames and comparisons among ORF 3 (coat protein, CP) sequences of these and a further seven isolates did not alter the pattern. Similarly, in comparisons among isolates obtained in Peru there was little variation in ORF 3, and no greater variation when they were compared with published sequences (MAM and L. Torrance, unpublished). In contrast, Guyader and Ducray (2002 ) found that comparisons among 19 isolates in the region of ORF 0 and the 5′ part of ORF 1 resolved them into three groups, one of Australian isolates, one containing a Peruvian and three European isolates and a third containing an isolate from Australia, a Cuban isolate and several from Europe. Thus there was no geographical correlation with sequence variation. The lack of sequence variation suggests either that PLRV only recently diverged from an ancestral virus (for example by acquiring an ability to infect potato) or that it has been subject to very strong selection constraints. The narrow genetic base of potatoes in cultivation may also be a factor in restricting PLRV variation ( Guyader and Ducray, 2002 ). The virus that most closely resembles PLRV in CP sequence is Sweet potato leaf speckles virus , which is also from South America ( Fuentes ., 1996 ). It is 70% identical to PLRV whereas other poleroviruses are no more than 65% identical. ORGANIZATION AND EXPRESSION OF THE PLRV GENOME The genome of PLRV comprises eight ORFs ( Fig. 1 ). It is densely packed, with ORF 1 overlapping much of ORF 0, and ORF 4 being entirely within ORF 3. In common with genomes of all luteoviruses ( Mayo and Miller, 1999 ), almost all types of modulation mechanism (frameshift, initiation bypass, termination suppression, production of subgenomic (sg) RNA, proteolysis of primary translation products) are used during the expression of the different ORFs. The three 5′‐proximal ORFs (ORFs 0–2) are expressed from the genomic RNA and five 3′‐proximal ORFs (ORFs 3–7) are expressed by translation from subgenomic mRNAs that are 3′ co‐terminal with genomic RNA. Thus luteoviruses present amongst the best examples of the versatility of expression strategy and the economy in use of coding sequence charactersitic of small RNA virus genomes. 1 The genome of PLRV showing different expression strategies: production of subgenomic mRNAs (sg1 and sg2), initiation at the first AUG (P0, P3 and P6), bypass and initiation at the second AUG (P1, P4 and P7), initiation as for P1 and frameshift (P1 + P2), initiation as for P3 and readthrough suppression of the UAG of ORF 3, proteolysis of the primary translation product (P1 and possibly P1 + P2). Lines represent RNA molecules, open boxes represent open reading frames (ORFs) and grey bars represent translation products. The first nine nucleotides of the 5′ non‐translated leader sequences are the same in genome RNA and sgRNA1 ( Miller and Mayo, 1991 ), but those of sgRNA2 are different. In vitro , leader sequences have effects on translation of the RNAs but are not translational enhancers and the non‐translated sequence had no effect on translation ( Juszczuk ., 2000 ). The initiation codon of ORF 4 is a few nucleotides downstream of that of ORF 3 ( Fig. 1 ). The ratio between P3 and P4 produced by translation of sgRNA1 has been reported to be c . 1 : 1 ( Juszczuk ., 2000 ) or c . 1 : 7 ( Tacke ., 1990 ). The functions of some PLRV proteins are known, but for others functions are speculative (reviewed in Mayo and Ziegler‐Graff, 1996 ). P0 is involved in symptom development and is a suppressor of gene silencing (see below); P1 has proteinase functions and contains the genome‐linked protein (VPg) ( Van der Wilk ., 1997 ), and possibly has other functions such as a helicase; P2 is an RNA‐dependent RNA polymerase; P3 is the capsid protein of c . 23 kDa; P4 is a putative movement protein; suppression of the ORF 3 amber stop codon results in translation of an ORF 3/ORF 5 read‐through protein that is a minor structural protein of c . 80 kDa that may be involved in aphid transmission. ORFs 6 and 7 have been demonstrated more recently and encode proteins of unknown function but P7 displays nucleic acid binding properties ( Ashoub ., 1998 ). Particle structure Protein composition In electron microscope preparations PLRV particles are isometric, 24 nm in diameter and sometimes appear to have small projections on the vertices. The sedimentation rate of particles ( s 20w = 115 S ), their diameter and the size of the major coat protein (P3) suggest that particles contain 180 copies of P3 in a T = 3 structure. About 1% to 5% of the P3 molecules in protein extracted from purified virus particles carry a C‐terminal extension derived from P5 ( Bahner ., 1990 ), although only a minor fraction of the particle protein, P5 is immunodominant. In one panel of monoclonal antibodies, 2 of 10 antibodies were P5‐specific ( Massalski and Harrison, 1987 ; MAM unpublished) and in another panel most were specific for P5 (L Torrance, V Flores and MAM unpublished). The size of the P3 + P5 protein in purified particles is c . 52 kDa but in extracts from infected cells the fusion protein is c . 80 kDa. P3 + P5 was observed to degrade from c . 80 kDa to c . 50 kDa after PLRV‐infected protoplasts were lysed ( Bahner ., 1990 ). The usual explanation for this discrepancy in size is that part of P5 is lost during purification, but this interpretation is not consistent with the observation that antibody panels raised to purified particles contain antibodies that react with parts of P5 close to the C‐terminal end ( Jolly and Mayo, 1994 ). It appears that the cleaved portions of P5 remain attached to purified PLRV particles but are removed when particles are denatured in SDS. If true, this suggests a relatively high affinity between P5 sequences and the surface portions of P3. The functional significance of this unusual interaction is unclear. Structure predictions Attempts have been made to predict the secondary structure of P3 from an analysis of its amino acid sequence. The prediction that the structure contained a basic N‐terminal region followed by patches of β‐sheet was made shortly after sequences were published ( Torrance, 1992 ) and subsequent predictions based on more refined software packages have supported this model ( Mayo and Ziegler‐Graff, 1996 ). The most recent of these ( Terradot ., 2001 ), and by far the most sophisticated, was made following the observation of a weak (17% identity, 33% similarity), but significant, resemblance between the sequences of the coat proteins (P3) of PLRV and the sobemovirus Rice yellow mottle virus (RYMV). Because a crystal structure was known for RYMV particles, the alignment of the PLRV and RYMV CPs allowed the construction of a model for PLRV CP and those of other luteoviruses. The interesting features to emerge are models for some epitopes and a clustering of acidic residues on the three‐fold axis of the virus particles. These are illustrated in Fig. 2 . 2 Models of the predicted structure of a reconstituted trimer of PLRV coat protein. With permission from Terradot . (2001 ; Academic Press Inc.). (A) Surface (left) or section (right) views. Monomer units are in light green, dark green or blue. Epitope 5 (residues 83–89) is shown in red, epitope 120 (residues 172–178) is shown in yellow. (B) Surfaces coloured according to the properties of the constituent amino acids: D, E—magenta; R, H, K—blue; A, V, L, I, P, M, F, W—orange; G, S, T, C, N, Q, Y—green. The acidic patch on the trimer surface is circled. Mapping the known epitopes in the PLRV CP ( Torrance, 1992 ) on to the structure model located epitope 5 as the loop between β‐sheets B and C, and epitope 10 as the G‐H loop and both map on to the reconstituted trimer ( Fig. 2A ). Epitopes are regions of sequence that are intrinsically mobile and are therefore candidates for surface regions that contribute to biological properties. The reconstituted trimer for all polerovirus CPs contains an acidic patch where E and D residues come into close proximity (magenta shading in Fig. 2B ). Possibly the acidic patch is involved in the retention of parts of P5 on the surface of virus particles. Another possibility is that it is involved in the switch of the N‐terminal sequence of the CP (a major epitope) from inside the particle to outside, a change postulated to be linked with particle disassembly ( Torrance, 1992 ). The patch is less prominent in the analogous model of CP trimers of BYDV‐MAV and BYDV‐PAV (genus Luteovirus ) and is not present in models of sobemovirus CP trimers. When expressed in insect cells, P3 did not form particles, appearing to denature into amorphous aggregates. However when the P3 molecule carried an N‐terminal extension that included six histidine residues, the protein formed virus‐like particles (VLP) and these formed into crystal‐like aggregates in the nuclei of the insect cells expressing the protein ( Lamb ., 1996 ). The particles proved very difficult to extract from the cells and seemed to be less stable during storage at −70 °C than were PLRV particles. The VLP contained nucleic acid unrelated to PLRV RNA suggesting that there is no absolute requirement for an initiation‐of‐assembly signal, but that polynucleotide material is needed for particle stability. BIOLOGY—SYMPTOMS AND PATHOLOGY PLRV is found wherever potatoes are grown, but its importance and relative abundance varies. Symptoms may be more or less severe depending on the isolate, cultivar, growing conditions and age of the plant when it was infected. Infection of potato can cause severe yield loss and quality reduction of the harvested tubers, and in some regions infection levels can be high and economic losses can be serious. In plants grown from infected tubers (secondary infection), shoots are stunted and an upward rolling of leaflets is readily apparent, especially those on lower leaves, which break easily when crushed and may be chlorotic. Symptoms of primary infection (current season infection) are usually less severe unless plants became infected early in the season. In normal circumstances, the virus is confined to phloem tissue and cannot be transmitted by inoculation of sap. Replication probably occurs in phloem companion cells and virus moves long‐distance within the phloem sieve elements ( Taliansky and Barker, 1999 ). VIRUS TRANSMISSION—VECTOR RELATIONSHIPS Several potato colonizing aphid species transmit PLRV. Myzus persicae is regarded as the most efficient vector and larval and adult stages of winged and non‐winged forms transmit PLRV. Transmission is persistent and non‐propagative (see for review, Herrbach, 1999 ). Thus, virus is ingested with phloem contents during feeding and particles move from the gut to the salivary glands via the haemolymph (virus acquisition). There is no evidence that PLRV multiplies within the aphid vectors and presumably PLRV particles remain intact throughout the route through the aphid body. The minimum feeding times for virus acquisition and transmission are each about 1 h, although transmission frequency increases with increase in the acquisition period up to 2 days or more. Vector efficiency depends not only on aphid species, but also on clones, biotypes, morphs and instars, with as much variation between as within aphid species (reviewed by Robert and Lemaire, 1999 ). Although much is known about the complex series of processes of the virus acquisition/transmission, many details and questions remain unclear or unanswered ( Reavy and Mayo, 2002 ). Specific interactions between the PLRV particles and receptors in the aphid vectors probably occur at several stages, and all of these are likely to affect the outcome of this process. The P5 ( Guilley ., 1994 ) of luteoviruses has been proposed to play a role in such a specific interaction, however, the position regarding the role of the P5 in aphid transmission is complex and full of contradictions. What is clear is that mutants lacking the P5 are not aphid‐transmissible. It has been suggested that the longevity of infective virus particles in the aphid haemolymph may be mediated by an association with symbionin, a protein produced by the endosymbiotic bacterium ( Buchnera sp.) of M. persicae. Symbionin has been proposed to bind to PLRV particle protein and protect it from proteolysis ( Hogenhout ., 1998 ), however, the binding of symbionin to virus particles or to the P5 protein has only been demonstrated in vitro and the exact role of symbionin in virus transmission needs to be resolved ( Reavy and Mayo, 2002 ). Moreover, the VLPs of PLRV CP formed in Spodoptera frugiperda cells infected with a baculovirus carrying the CP (P3) gene of PLRV fused to a histidine tag lacking RT protein, were able to move through aphids from the gut to the salivary duct lumen ( Gildow ., 2000 ). Thus the dependency of infection of the presence of P5 may be better explained by P5 being involved in the initiation of infection within the plant. However, experiments with mutants of BWYV showed that those lacking the P5 did not bind to accessory salivary glands ( Reinbold ., 2001 ), suggesting that P5 has a role in this process. Although the P5 of PLRV and BWYV may differ, it seems more probable that P5 has more than one role. It is clear that different regions of the PLRV particle (such as the acidic patch on P3 trimers or P5) are important in recognition events at different stages of the transmission process, although the precise role and function of the proteins in the processes of aphid transmission and establishment of infection in plants remains to be elucidated. Therefore, an important area to investigate will be the nature and role of specific aphid virus receptor molecules that mediate the attachment and specific movement of PLRV particles through the various cellular layers and membranes within the aphid as virus moves from the gut to eventually emerge in the saliva to infect a new plant. The identification and investigation of these receptors may provide clues on how to interfere with this process and prevent virus transmission. VIRUS MOVEMENT Typically, after initial infection, plant viruses replicate and spread locally, from cell to cell through the plasmodesmata until they reach the phloem. The virions then enter the host vascular system, invade different types of phloem‐associated cells such as phloem parenchyma and companion cells, penetrate into the sieve elements and rapidly move through them (from leaf to leaf) and exit to infect non‐vascular tissues. This is long‐distance movement (reviewed in Oparka and Santa Cruz, 2000 ; Rhee ., 2000 ). Cell‐to‐cell movement usually involves one or more virus‐encoded movement proteins (MPs) and host components, such as tubulins, protein kinases and pectin methylesterases ( Boyko ., 2000 ; Chen ., 2000 ; Dorokhov ., 1999 ; Waigmann ., 2000 ), but less is known about long‐distance transport. Luteoviruses represent a special situation, because their accumulation and spread are mainly limited to the phloem cells (see below). PLRV, like other luteoviruses, cannot be transmitted mechanically, and infects plants only when delivered into phloem tissues (by aphids, agroinoculation or grafting). Within the phloem, PLRV can invade phloem parenchyma and companion cells and spread through the sieve elements. Thus, the process of luteovirus movement may resemble just one phase of the spread of other plant viruses, namely phloem‐associated long‐distance movement and thus represents a ‘pure’ experimental system for studying phloem‐associated movement without the complication of effects from short‐distance mesophyll cell‐to‐cell movement. However, the long‐distance movement function of PLRV has long been assigned to a 17‐kDa protein encoded by ORF 4 (P4, see above), that has properties strikingly similar to those of the Tobacco mosaic virus (TMV) 30 kDa MP involved in cell‐to‐cell movement, arguably the best studied of MPs. These properties for both of the proteins are single‐stranded RNA binding, protein dimerization, phosphorylation by a membrane‐bound protein kinase, plasmodesmal localization in both virus‐infected and transgenic plants and ability to increase plasmodesmal size exclusion limits ( Hofius ., 2001 ; Schmitz ., 1997 ; Sokolova ., 1997 ; Tacke ., 1991 ; Tacke ., 1993 ). Nevertheless, in spite of the similarities in properties between the TMV 30 kDa protein and PLRV P4, PLRV, unlike TMV, is unable to spread out of the plant vasculature (see below) and it is assumed that P4 mediates virus movement only between cells within the phloem tissues. This role for the PLRV P4 is supported by the observation that its analogue in the genome of another luteovirus, BYDV‐PAV was shown by mutational analysis to be a movement protein ( Chay ., 1996 ). Direct evidence for this role comes from experiments by Lee . (2002 ) using PLRV mutants in which ORF 4 encoding P4 was rendered either untranslatable or was modified to express the protein but lacking the first four amino acids. Both mutants were able to replicate and accumulate in agroinoculated leaves of potato and Physalis floridana, but were unable to move into vascular tissues and initiate a systemic infection in these plants, indicating the PLRV P4 is strictly required for virus movement. However, in two other plant species, Nicotiana benthamiana , and Nicotiana clevelandii , the same PLRV mutants were able to spread systemically. Moreover, Ziegler‐Graff . (1996 ), found that the BWYV analogue of the PLRV P4 was not essential for movement in N. clevelandii. These results suggest that the requirement for the P4 movement protein is host‐dependent and that there is a P4‐independent mechanism for PLRV movement that operates at least in some plants. Mutants lacking the 3′ parts of the P5 gene were found to be less infective and to accumulate to a lesser extent than unmodified PLRV (MAM and B. Reavy unpublished results). These observations mirror the results of more extensive studies with BWYV showing that mutations in different parts of the P5 gene had similar effects ( Bruyére ., 1997 ). With PLRV, direct indication that P5 is also involved in systemic spread, came from experiments using PLRV tagged with green fluorescent protein (GFP) ( Nurkiyanova ., 2000 ). The PLRV genome was modified by insertion of the gene for GFP into ORF 5 near its 3′ end. The chimeric PLRV‐GFP was able to replicate, express the GFP and form virus particles that contained P5 in the form of a fusion with GFP, in protoplasts or primarily infected cells. These virus particles were aphid‐transmissible but the GFP‐tagged PLRV was unable to spread systemically when agroinfiltrated or transmitted by aphids into plant hosts. Only naturally occurring mutants that lacked some or all of the inserted GFP sequences, were able to spread. These results support the idea that PLRV P5 is involved in virus movement and suggest that some putative ‘transport’ domain(s) in the P5 may be functionally disrupted by fusion to GFP. If so, it can be assumed that different functional domains are involved in aphid transmission and systemic spread of PLRV. PLRV virions have been detected in specialized plasmodesmata connecting sieve elements to companion and phloem parenchyma cells ( Esau and Hoefert, 1972 ; Shepardson ., 1980 ), which suggests that virions are the infectious entity engaged in vascular transport. Ziegler‐Graff . (1996 ) suggested that BWYV P5 domains on the surface of virus particles act as movement proteins to mediate the P4‐independent penetration into and translocation through sieve elements of virions. The P4‐dependent pathway, on the other hand, may not require P5 and presumably involves cell‐to‐cell transport of the virus through the plasmodesmata connecting nucleate phloem (companion and parenchyma) cells. However, this is clearly a host‐dependent process, because PLRV P4‐deficient mutants are unable to spread in potato or P. floridana (see above). PLRV AND RNA SILENCING Higher plants possess a specific defence mechanism against viruses that is similar to post‐transcriptional gene silencing (PTGS). PTGS or RNA silencing is a type of a sequence‐specific degradation of RNA molecules, a key element of which is the involvement of a dsRNA that corresponds to the target RNA. Such a dsRNA is produced during replication of RNA viruses. During silencing, dsRNA is cleaved into short interfering RNAs (siRNA), 21–25 nucleotides in length, and these are thought to mediate the specificity of RNA degradation (see for reviews, Carrington, 2000 ; Vance and Vaucheret, 2001 ; Voinnet, 2001 ). To counteract RNA silencing, some viruses encode specific proteins that suppress different steps in the silencing pathways (see for reviews, Carrington, 2000 ; Vance and Vaucheret, 2001 ; Voinnet, 2001 ) . For example, the effect of potyvirus helper component‐proteinase (HC‐Pro) is to reverse silencing in cells already silenced, suggesting that this protein suppresses a function required for the maintenance of the silenced state. The Potato virus X 25 kDa protein appears to target and interfere with the spread of the silencing signal. The Cucumber mosaic virus (CMV) 2b protein affects only the initiation of silencing, by inactivating the mobile silencing signal ( Guo and Ding, 2002 ). Although the P0 has never been detected in PLRV‐infected plants, there is strong genetic evidence that it is functionally important. Mutations that prevented expression of this protein in vitro , completely abolished PLRV accumulation in plant cells ( Sadowy ., 2001 ). Recently PLRV P0 (like HC‐Pro) has been shown to display significant silencing suppressor activity in a transient expression system based on its ability to inhibit PTGS of GFP in agro‐infiltrated leaves of transgenic plants containing a GFP transgene when coexpressed with silencing inducer (GFP‐coding nucleotide sequence) ( Pfeffer ., 2002 ). PLRV P0 can also act as an RNA silencing suppressor when expressed in Drosophila cells (B. Reavy and S. MacFarlane, personal communication). The P0 encoded by the other luteoviruses BWYV and CABYV have also been shown to suppress silencing ( Pfeffer ., 2002 ). The results of further experiments allowed the authors to suggest that P0 had interacted with a component of the silencing machinery responsible for cell‐autonomous silencing but did not interfere with production and delivery of the mobile silencing signal. However, marked PTGS suppressor activity was observed only when the P0 was expressed from monocistronic constructs. Full‐length PLRV (our unpublished results) or BWYV ( Pfeffer ., 2002 ) displayed much lower suppression activity, if any. Pfeffer . (2002 ) suggested that low PTGS suppression activity was a consequence of the low accumulation of the BWYV P0 protein resulting from the suboptimal translational context of its start codon in viral RNA. Moreover, mutations designed to optimize translation initiation efficiency in BWYV RNA were not stable during virus multiplication in plants, suggesting that there is a selection against over‐expression of the P0 and presumably its silencing suppression activity. Further experimentation is required to determine how the expression of the P0 is regulated during PLRV infection and its role in counter‐defence against RNA silencing. A new insight into the mechanisms of overcoming RNA silencing that may be exploited by luteoviruses, and PLRV in particular, came from experiments with transgenic plants that expressed the P5 sequence of BWYV ( Brault ., 2002 ). Following infection of these plants with BWYV or related viruses such as PLRV, transgenic P5 and its mRNA accumulated to only low levels and siRNAs corresponding to the transgene were generated. This suggested that RNA silencing had been induced by the infections (BWYV and PLRV nucleotide sequences contain some stretches of identical sequence that account for the ability of PLRV to induce silencing). As expected, the transgene (P5) mRNA, both within and outside of the phloem, was targeted for RNA silencing‐specific degradation. The viral RNA, however, was not subject to significant silencing and the transgenic plants were readily infected with either BWYV or PLRV. Thus a mechanism must exist that can discriminate between the viral and transgene RNA for degradation by the silencing machinery ( Brault ., 2002 ). Known virus‐encoded silencing suppressor proteins cannot differentiate viral and mRNA, rescuing both of them from silencing. Therefore, the suggestion that the putative P0 suppressor is involved in selective silencing suppression of the viral RNA (not affecting the degradation of transgene RNA) does not seem very likely. It is more likely that luteoviruses and PLRV in particular, exploit a novel and as yet unrecognized mechanism for the protection of viral RNA molecules from silencing by sequestering them in replication, translation, movement or other complexes, or as virions. MECHANISMS FOR PHLOEM RESTRICTION An intriguing feature of PLRV (and other luteoviruses), is the phloem restriction of the virus. PLRV cannot be transmitted mechanically, and infects plants only when delivered into phloem tissues (by aphids, agroinoculation or grafting). PLRV can replicate in mesophyll protoplasts and in primarily inoculated epidermal or mesophyll cells. Also, in plants PLRV is to some extent able to exit from the phloem into mesophyll cells when the PLRV‐infected plants are also infected with certain other plant viruses ( Atabekov ., 1984 ; Barker, 1987, 1989 ). These findings suggest that the normal restriction of PLRV to the phloem is because PLRV movement functions do not operate in the epidermis and mesophyll tissues (see for review, Taliansky and Barker, 1999 ) and/or because PLRV cannot suppress RNA silencing host defence in non‐vascular tissues ( Voinnet ., 1999 ; Waterhouse ., 1999 ). To test this hypothesis we designed a novel experimental approach to allow PLRV to spread out of the phloem and to become mechanically transmissible as a result of complementation by umbraviruses ( Mayo ., 2000 ; Ryabov ., 2001 ). In plants infected with PLRV and the umbraviruses Pea enation mosaic virus ‐2, or Groundnut rosette virus (GRV), PLRV accumulated in clusters of mesophyll cells in both inoculated and systemically infected leaves. No PLRV mechanical transmissions were obtained by coinoculation with CMV or some other viruses, although PLRV was transmissible from mixtures with a recombinant CMV that contained the MP gene (ORF 4) of GRV [CMV(ORF4)] ( Ryabov ., 2001 ). This suggested that the umbravirus‐dependent mechanical transmission and spread of PLRV in non‐vascular tissues depends on the umbravirus‐specific cell‐to‐cell movement function, however, a mutant of CMV(ORF4), in which the CMV 2b gene was untranslatable, was unable to help PLRV transmission and movement. Because the 2b protein is an RNA silencing suppressor ( Guo and Ding, 2002 ), we suggested that suppression of silencing is essential for the spread of PLRV in non‐vascular tissues. This is consistent with the recent observation that in transgenic plants that express full‐length PLRV cDNA in only a very limited number of the mesophyll cells, the expression of a potyvirus HC‐Pro resulted in a marked increase in the number of cells that accumulated PLRV ( Barker ., 2001 ; Franco‐Lara ., 1999 ). The key feature of these experiments was that PLRV RNA transcript was produced in each plant cell and infection did not depend on virus movement from cell to cell. In contrast, Savenkov and Valkonen (2001 ) showed that HC‐Pro, expressed in PLRV‐infected plants (i.e. in conditions when cell‐to‐cell movement is required for PLRV invasion into non‐vascular tissues), enhanced titres of PLRV only in the phloem system but not outside, suggesting that silencing suppression is not enough to allow PLRV to invade mesophyll tissues. Taken together, these results suggest that both the movement function and the overcoming of silencing are required for PLRV to become mechanically transmissible. Thus, confinement of PLRV, and presumably luteoviruses in general, to phloem tissues could be explained by (i) the failure of cell‐to‐cell movement and (ii) lack of mechanisms to overcome RNA silencing outside the phloem. In other words, we can assume that the PLRV‐encoded MPs (P4 and/or P5) are unable to fulfil their functions in the mesophyll tissues. Similarly, mechanism(s) used by PLRV for counter‐defence against RNA silencing (P0‐mediated suppression and/or protection against RNA degradation, see above) operate only in the phloem but not in non‐vascular tissues. Further studies will be required to clarify the nature of this tissue specificity. RESISTANCE Several potato ( Solanum tuberosum ) cultivars are known to have genetically controlled resistance to PLRV, but no cultivar is immune to infection and there are no known sources of major gene resistance. Several forms of resistance are expressed as a less than normal level of virus accumulation in infected plants ( Solomon‐Blackburn and Barker, 2001 ). The transfer of valuable traits from tuber‐bearing and non‐tuber‐bearing wild potato species and cultivated diploids to cultivated S. tuberosum potato has been a major effort of several breeding programmes. At least seven species are known to contain resistance to PLRV including Solanum brevidens, S. etuberosum, S. chacoense and S. raphanifolium that have very high levels of resistance ( Barker and Waterhouse, 1999 ). However, none of these forms of resistance have yet been transferred successfully to cultivated potato. Genetically engineered resistance to PLRV was first reported by Kawchuk . (1990 ) who transformed potato with the PLRV CP gene. Subsequently there have been many other examples of pathogen‐derived resistance to PLRV ( Barker and Waterhouse, 1999 ). Several of these are also expressed as resistance to virus accumulation and have a similar phenotype to naturally occurring forms of host resistance to PLRV. Transgenic potatoes resistant to PLRV were released commercially in the USA in 1997, although their use has declined recently. PROSPECTS Extensive studies of PLRV and plant diseases caused by this virus, have a very long history that makes PLRV a ‘classic’ pathogen. This paper describes some achievements in the understanding of the mechanisms used by PLRV to establish a systemic infection in plants, however, before assuming that we have a detailed picture of PLRV infection, it is important to consider the gaps in the current state of knowledge. Some of the gaps are listed in the sections above. Future research can provide both novel information on the molecular mechanisms of pathogenesis induced by PLRV as well as new tools for further studies. Indeed, this classic pathogen still displays some tricks making it an attractive experimental system for studies of some general biological processes such as macromolecular trafficking through the plant's long‐distance transport system, the phloem, and mechanisms of RNA silencing and silencing suppression. For example, if PLRV proves to move through the phloem by exploiting and modifying endogenous plant pathways for macromolecular movement, the virus would represent a ‘pure’ system that allows the analysis of the phloem‐associated transport without any possible effects from cell‐to‐cell movement. Another novel aspect of work with PLRV is the use of amplicons (virus‐based transgenes) as a tool for the study of RNA silencing. Amplicons were first described by Angell and Baulcombe (1997 ) for PVX, and then for other viruses including PLRV by Franco‐Lara . (1999 ) and Barker . (2001 ). An amplicon is a transgene that comprises a full‐length (biologically active) copy of a viral genome. In plants that contain an amplicon, transgene‐derived transcripts replicate like RNA viruses and this process could potentially take place in every cell of the plant, however, such replication also induces very effective silencing of the amplicon. More recently we have made an improved version of the PLRV amplicon expressing GFP to monitor virus expression. The advantages of the PLRV amplicon‐mediated RNA silencing system are: (i) the inability of PLRV to suppress or avoid silencing in non‐vascular tissues, and (ii) the lack of virus movement from cell to cell that allows the study of silencing of viral replicating RNAs in the absence of effects from silencing suppression and virus movement. PLRV amplicons will be used for studying mechanisms of RNA silencing and its suppression by different biotic or abiotic factors. ACKNOWLEDGEMENTS This work was supported by a grant‐in‐aid from the Scottish Executive Environment and Rural Affairs Department.

Journal

Molecular Plant PathologyWiley

Published: Mar 1, 2003

There are no references for this article.