M J Gibbs#$, A Ziegler*, D J Robinson*, P M Waterhouse$ and J I Cooper#.
# Natural Environment Research Council, Institute of Virology and Environmental Microbiology, Mansfield Road Oxford OX1 3SR, United Kingdom.
$ Commonwealth Scientific and Industrial Research Organisation, Division of Plant Industry, Black Mountain Laboratories, GPO Box 1600, Canberra ACT 2601, Australia.
* Scottish Crop Research Institute, Invergowrie, Dundee DD2 5DA, United Kingdom.
Corresponding author: <\p>
Mark Gibbs, Plant Science CRC, GPO Box 475, Canberra ACT 2601, Australia
telephone: 61 6 2465135, fax: 61 6 2465000, email: firstname.lastname@example.org
Accepted: 11 November, 1996
Carrot mottle umbravirus (CMoV) has always been found co-infecting plants with carrot red leaf luteovirus (CRLV) and in carrot (Daucus carota) these co-infections are associated with carrot motley dwarf disease (CMD). CMD occurs wherever carrots are grown. Hence, CMoV was believed to have a corresponding global distribution. However, little or no hybridisation was detected between cDNA generated from the sequenced Australian isolate of CMoV (CMoV-A) and RNA from the much studied Scottish isolate of CMoV (CMoV-S). A weak hybridisation signal was obtained using cDNA to a conserved part of the RNA-dependent RNA polymerase gene of CMoV-A, but when cDNAs to other parts of the CMoV-A genome were used as probes there was no detectable hybridisation with CMoV-S RNA. This lack of hybridisation suggests that the two virus isolates have relatively divergent genomes and that they should be regarded as distinct virus species. Both viruses are transmitted by Cavariella aegopodii, but only with the help of CRLV, and they yield almost identical double-stranded RNA profiles. For these reasons, we propose that the CMoV isolate from Australia be renamed carrot mottle mimic umbravirus (CMoMV). cDNA to CMoMV RNA hybridised with RNA from an isolate from New Zealand, whereas cDNA to CMoV-S RNA hybridised with RNA from isolates from England and Morocco but not to RNA from the isolate from New Zealand. Although preliminary, these data suggest that CMoV and CMoMV may have different global distributions.
The 4201 base pair long genome of the Australian isolate of carrot mottle umbravirus (CMoV-A) encodes four genes including an RNA-dependent RNA polymerase gene and a movement protein gene (Gibbs et al., 1996). Sequence comparisons show the virus to belong to the carmo-like grouping and to be most closely related to groundnut rosette umbravirus(GRV; Taliansky et al., 1996) and pea enation mosaic enamovirus RNA-2 (PEMV-RNA2; Demler et al., 1993).
CMoV is associated with carrot motley dwarf disease (CMD; Stubbs, 1948) which is one of six plant diseases believed to be caused by a combination of a luteovirus (or luteo-like virus) and an umbravirus (Smith, 1946; Watson et al., 1964 a, b; Hull and Adams, 1968; Adams and Hull, 1972; Falk et al., 1979a; Cockbain et al., 1986). A further four diseases are suspected also to be caused by complexes of this kind (Murant et al., 1995). Umbraviruses are transmitted by aphids only from plants that are also infected with their luteovirus helpers (Murant et al., 1995), and it seems likely that in nature umbraviruses are transmitted only in this way, a feature that probably maintains their associations with luteoviruses.
Studies with a Scottish isolate of CMoV (CMoV-S) and with lettuce speckles mottle umbravirus (LSMV) have shown that umbraviruses achieve dependent transmission when their genomic RNA is encapsidated in the coat protein of the helper luteovirus (Falk et al., 1979b; Waterhouse and Murant, 1983). The idea that umbraviruses rely completely on this interaction with their helpers for transmission is further supported by evidence that umbraviruses do not encode virion proteins or produce particles of their own (Watson et al., 1964b; Murant et al., 1969; Falk et al., 1979b; Gibbs et al., 1996). Not surprisingly, this last feature has hindered the identification of umbraviruses. Antisera cannot be raised to them and hence umbravirus species cannot be distinguished serologically. Instead, nucleic acid-based identification methods seem appropriate (reviewed in Waterhouse and Chu, 1995), and here we report an application of such methods to distinguish two umbravirus species.
Isolates, presumed to be of CMoV, were obtained from carrots (Daucus carota) with CMD-like symptoms grown in Cambridgeshire, England (CMoV-E), in the Australian State of New South Wales (CMoV-A; Gibbs et al., 1996; GenBank Accession Number U57305) and in New Zealand (CMoV- NZ). Another isolate, presumed to be CMoV, was obtained from carrots with CMD-like symptoms found in a market in Morocco (CMoV-M). The Scottish isolate of CMoV (CMoV-S; Murant et al., 1969) was originally obtained from a carrot with CMD growing in Carnoustie, Angus. CMoV-S(p) was an isolate obtained by passage of CMoV-S in parsley (Petroselinum crispum). Each of these virus isolates was propagated in N. clevelandii or N. benthamiana by serial mechanical inoculation (Murant et al., 1969). CMoV-A was also propagated in N. megalosiphon. A CMoV-free isolate of carrot red leaf luteovirus (CRLV) was obtained from carrots growing in the Australian Capital Territory (Waterhouse, 1985). The satellite-free culture MC1 (Murant and Kumar, 1990) of groundnut rosette umbravirus (GRV; Hull and Adams, 1968) was used.
Healthy carrot and coriander (Coriandrum sativum) were mechanically inoculated with sap from CMoV-A-infected N. clevelandii, and two days later these same plants were inoculated with CRLV using the aphid Cavariella aegopodii. Two weeks later, C. aegopodii were allowed to feed on the infected plants for two days and then transferred to healthy carrot and coriander test plants. Aphid transmission of CMoV-A was then assayed by mechanically inoculating N. clevelandii with sap from the test plants. As a control, the experiment was repeated using coriander inoculated with CMoV-A, but not with CRLV.
Extraction of double-stranded RNA
Fresh leaf tissue (100 g) from N. clevelandii was ground in liquid N2. Total nucleic acids were extracted in 10 mM Tris-Cl, 1mM EDTA, pH 8.0, 50% phenol and 0.01% SDS. The mixture was stirred for one hour, centrifuged at 12000 g for 10 minutes and nucleic acids were recovered by ethanol precipitation. Double-stranded RNA (dsRNA) was isolated by cellulose column chromatography (Dodds and Bar-Joseph, 1983). Preparations were analysed by agarose gel electrophoresis after treatment with ribonuclease A, in water or in buffer containing 50 mM Tris-Cl pH 7.3, 300 mM NaCl and 10 mM MgCl2 (Hirs et al., 1953), or after treatment with deoxyribonuclease I (Melgar and Goldthwait, 1968).
Three Northern hybridisation methods were used.
1. Aliquots of dsRNA (about 200 ng) from CMoV-A-infected N. clevelandii and from CMoV-S- infected N. clevelandii were subjected to electrophoresis in non-denaturing agarose gels and transferred to nylon membranes by blotting. Gels were stained with ethidium bromide before and after blotting to monitor each step. Hybridisation buffer (90 mM tri-sodium citrate pH 7, 900 mM NaCl (6 x SSC), 50 % formamide, 0.1 % Ficoll, 0.1 % polyvinylpyrrolidone (PVP), 0.1 % bovine serum albumin (BSA), 0.5 % SDS, and 100 mg/ml sheared salmon sperm DNA) was prepared by heating to 80oC for 10 minutes and then cooled on ice for 5 minutes. Prehybridised membranes were transferred to hybridisation buffer containing nick-translated cDNA (Rigby et al., 1977), generated from CMoV-A dsRNA (Gibbs et al., 1996), and incubated overnight at 42oC. After incubation, membranes were washed five times for five minutes each in 2 X SSC at 42oC.
2. Total nucleic acids (1 mg) from CMoV-A-infected N. clevelandii and from CMoV-S-infected N. benthamiana were lyophilised, resuspended in a solution containing 50 % v/v de-ionised formamide, 7 % v/v formaldehyde, 5 mM NaH2PO4, 5 mM Na2HPO4 and 1 mM EDTA (pH 8.0), and denatured by incubating at 60oC for 15 minutes. The samples were cooled on ice and subjected to electrophoresis in agarose gels. Both the gel and the buffer reservoirs contained 8.8 % v/v formaldehyde, 5 mM NaH2PO4, 5 mM Na2HPO4, and 1 mM EDTA (pH 8.0). Separated RNA species were transferred to nylon membranes by blotting and incubated with cDNA probes as described above. After incubation, membranes were washed three times for 10 minutes each in 0.1 X SSC with 0.1 % SDS.
3. Some Northern hybridisation experiments with dsRNA were done as described by Murant et al. (1988). Duplicate gel tracks were stained with silver to verify that the samples contained the expected dsRNA species, and gels were also stained after electroblotting to confirm transfer. A cDNA probe was made by reverse transcription of dsRNA-1 purified from plants infected with CMoV-S. Hybridisation was done overnight at 65oC in 5 X SSC, 0.8 % BSA, 0.8 % Ficoll, and 0.8 % PVP and membranes were washed four times for 15 minutes each at 65oC in 2 X SSC with 0.1 % SDS.
Two dot-blot hybridisation methods were used.
1. Aliquots of dsRNA (about 50 ng) from CMoV-A-infected N. clevelandii and from CMoV-S-infected N. clevelandii were spotted onto nylon membranes as described by White and Bancroft (1982). A cDNA probe was prepared from CMoV-S dsRNA as described above (Northern hybridisation method 3). Prehybridisation, hybridisation and washing were done as described in Northern hybridisation method 1.
2. Aliquots (6 ml) of total single-stranded RNA, extracted from infected N. benthamiana as described by Blok et al. (1994), were spotted onto nitrocellulose. A cDNA probe was prepared as described by Feinberg and Vogelstein (1984). Prehybridisation and hybridisation were done as described by Blok et al. (1994). After incubation the membrane was washed twice for 15 minutes at 65oC in 2 X SSC, 0.1 % SDS, and twice for 15 minutes at 65oC in 0.1 X SSC, 0.1 % SDS.
Isolation and characterisation of carrot mottle mimic umbravirus
Carrots with a disease similar to CMD were found in a commercial plot in the Australian State of New South Wales. When N. clevelandii were mechanically inoculated with sap from the affected carrots, their leaves became slightly distorted and developed pale necrotic patches, suggesting that a virus had been isolated.
Coriander were inoculated with sap from these symptom-bearing N. clevelandii, and two days later were inoculated with CRLV using C. aegopodii. These plants developed symptoms similar to CMD, and the disease could then be efficiently transmitted by C. aegopodii to uninfected carrots and coriander. Aphid transmission of the mechanically-transmissible agent isolated from the original carrots (CMoV-A) was confirmed when symptoms developed on the leaves of N. clevelandii inoculated with sap from this last set of carrots and coriander with CMD-like symptoms. CMoV-A has since been maintained with CRLV in carrot and coriander for nine years using C. aegopodii to transmit both viruses.
Coriander inoculated with CMoV-A, but not with CRLV, became mottled and N. clevelandii inoculated with sap from these plants developed the typical symptoms of CMoV-A. However, C. aegopodii that were allowed to feed on these plants failed to transmit CMoV-A to healthy coriander or carrot as shown by mechanically inoculating N. clevelandii with the sap of these test plants. Carrots inoculated with CMoV-A did not develop symptoms and the virus could not be transmitted from these plants by mechanical inoculating N. clevelandii with sap extracts.
Two dsRNA species were isolated from the mechanically-inoculated N. clevelandii (Gibbs et al., 1996). Northern hybridisations suggested that the smaller dsRNA represents a dsRNA form of a 3' co-terminal subgenomic mRNA (Gibbs et al., 1996). As shown in Figure 1, the dsRNAs from the N. clevelandii infected with CMoV-A were almost identical in size to the two dsRNA species found in CMoV-S-infected N. clevelandii. As judged against dsDNA size markers, the species from plants infected with CMoV-A corresponded in size to about 4.5 kilobase pairs (kbp) and 1.3 kbp. These size estimates are very close to those obtained for CMoV-S by Murant et al. (1985) and Halk et al. (1979). However, sequencing has shown the larger CMoV-A dsRNA to be 4201 bp long (Gibbs et al., 1996), suggesting that previous reports tended to over-estimate the sizes of CMoV-S dsRNAs.
As shown in Figure 2, only very weak hybridisation was detected when CMoV-S dsRNA was probed with cDNA representing nucleotides 1884 to 2233 in the CMoV-A genome (clone 658; Gibbs, 1995; Gibbs et al., 1996). This hybridisation was done using the least stringent Northern hybridisation method (Northern hybridisation method 1). The CMoV sequence represented by the cDNA of clone 658 encodes part of the RNA-dependent RNA polymerase gene of CMoV-A, including the highly conserved GDD amino acid sequence motif (Gibbs et al., 1996), and has 63 % identity with the equivalent sequence in pea enation mosaic RNA2 (Demler et al., 1993) and 66 % identity with the equivalent sequence in the genome of GRV (Taliansky et al., 1996). Weak hybridisation was also detected in a dot-blot hybridisation where cDNA synthesised by random-primed reverse transcription from CMoV-S dsRNA was used to probe membrane bound CMoV-A RNA (Figure 3). Low stringency hybridisation conditions were also used in this experiment (dot-blot method 1).
When more stringent conditions were used no cross-hybridisation was detected between CMoV-A or CMoV-S RNA species and cDNA generated from the dsRNAs of these viruses. In dot-blot experiments (dot-blot method 2), a probe prepared from clone 369, which corresponds to nucleotides 741 to 1006 in the genomic RNA of CMoV-A, hybridised with RNA prepared from CMoV-NZ-infected N. benthamiana as well as with CMoV-A RNA, but not with RNA from N. benthamiana infected with CMoV-S or with GRV (Figure 4). Similarly, in a Northern hybridisation (Northern hybridisation method 2), a probe prepared from clone 684, which corresponds to the 3' terminal 530 nucleotides of the genome of CMoV-A, hybridised with four species, two ssRNAs and two dsRNAs, in total nucleic acid preparations from CMoV-A-infected N. clevelandii, but not with total nucleic acid preparations from CMoV-S-infected N. clevelandii (data not shown).
As shown in Figure 5, cDNA synthesised by random-primed reverse transcription from CMoV-S dsRNA-1 hybridised with dsRNA species present in preparations from N. clevelandii infected with CMoV-E and CMoV-M as well as with CMoV-S and CMoV-S(p), but not with dsRNA prepared from N. clevelandii infected with CMoV-NZ. This was done using Northern hybridisation method 3.
Previous work on CMD in Britain showed that it was caused by a complex comprising CMoV and CRLV (Watson et al., 1964a, b). The results reported here suggest a more complicated picture. A mechanically-transmissible virus (CMoV-A) was isolated from carrots with CMD-like symptoms growing in Australia (Gibbs et al., 1996). Like the isolates of CMoV from Britain, CMoV-A could be transmitted by C. aegopodii with the help of CRLV. Furthermore, plants infected with either CMoV-A or the well studied Scottish isolate, CMoV-S, had almost identical dsRNA profiles. However, nucleic acid hybridisation experiments show CMoV-S and CMoV-A to be distinct. Umbraviruses have been recognised on the basis of their transmission characteristics and the presence in infected plants of two dsRNA species with sizes close to those found in CMoV-S-infected plants (Murant et al., 1995). For these reasons we believe CMoV-A to be an umbravirus and have renamed it carrot mottle mimic umbravirus (CMoMV).
CMoMV and CMoV-S induced similar symptoms in the propagation host plants with one notable exception. CMoMV induced moderate to severe systemic necrotic and chlorotic lesions in N. megalosiphon (Figure 6), whereas CMoV-S produced no symptoms in this host.
Weak hybridisation was detected when CMoV-S RNA was probed with a cDNA corresponding to a highly conserved part of the CMoMV genome. Similarly weak hybridisation was detected when CMoMV RNA was probed with cDNA made by random-primed reverse transcription of CMoV-S dsRNA. These results suggests that there is sufficient similarity between the viral genomes, in some places, for moderately stable base-pairing. However, relatively low stringency hybridisation conditions were used in both experiments and hence, the signals could also represent non-specific hybridisation.
The hybridisation experiments also showed that CMoMV occurs in New Zealand, and a similar experiment, using clone 369, has shown CMoMV to occur in California (B.W. Falk, University of California Davis, personal communication). Other experiments, using cDNA to CMoV-S dsRNA as a probe, showed that CMoV occurs in Britain and Morocco. Thus our results indicate that either CMoV or CMoMV, presumably in combination with a CRLV-like luteovirus, is associated with CMD-like symptoms at different localities. The first of these combinations is found in Britain and Morocco whereas the second is found in some states on the Pacific rim. These results cast doubt on the former belief that CMoV, like CMD, occurs wherever carrots are grown (Murant 1974, 1975), but it must be emphasised that these data can only be considered preliminary. Both viruses may be globally distributed or one may have a more limited distribution than the other. It is also possible that some plants at some localities are infected by CRLV and both umbraviruses. Moreover, infection of carrots by CRLV alone also occurs in the field and causes the leaf reddening or yellowing characteristic of CMD but with little stunting (Watson et al., 1964b; Waterhouse, 1985).
Although we do not know the nature of the virus that acts as the natural helper of CMoMV, four observations suggest that CRLV has this role. First, CRLV is the only luteovirus known to infect Umbelliferae. Second, CMoMV transmission by C. aegopodii with the help of CRLV is efficient (P.M.W unpublished observations). Third, CRLV has been found in eastern Australia (Waterhouse, 1985). Fourth, the carrots from which CMoMV was isolated had CMD-like symptoms and these symptoms were replicated when test carrots were co-inoculated with CMoMV and CRLV. In view of this last observation, previous assumptions concerning the aetiology of CMD should be re-examined.
Falk and Duffus (1981) warned of the possible difficulties in recognising and identifying umbraviruses and other transmission-dependent viruses. The experiences reported here, where two distinct umbravirus species occupy apparently identical niches and probably rely on similar or identical helper luteoviruses, can only reinforce that warning.
Cockbain AJ, Jones P, Woods RD, 1986. Transmission characteristics and some other properties of bean yellow vein-banding virus, and its association with pea enation mosaic virus. Annals of Applied Biology 108, 59-69.
Falk BW, Duffus JE, 1981. Epidemiology of helper-dependent persistent aphid transmitted virus complexes. In: Maramorosch K, Harris KF, eds. Plant Diseases and Vectors. New York: Academic Press, 162-179.
Murant AF, Kumar IK, 1990. Different variants of the satellite RNA of groundnut rosette virus are responsible for the chlorotic and green forms of groundnut rosette disease. Annals of Applied Biology 117, 85-92.
Murant AF, Rajeshwari R, Robinson DJ, Raschké JH, 1988. A satellite RNA of groundnut rosette virus that is largely responsible for symptoms of groundnut rosette disease. Journal of General Virology 69, 1479- 1486.
Murant AF, Robinson DJ, Gibbs MJ, 1995. Genus Umbravirus. In: Murphy FA, Fauquet CM, Bishop DHL, Ghabriel SA, Jarvis AW, Martelli GP, Mayo MA, Summers MD, eds. Virus taxonomy the classification and nomenclature of viruses. New York: Springer-Verlag.
Murant AF, Waterhouse PM, Raschke JH, Robinson DJ, 1985. Carrot red leaf and carrot mottle viruses: Observations on the composition of the particles in single and mixed infections. Journal of General Virology 66, 1575-1579.
During the course of this work MJ Gibbs was a student of the University of Oxford Department of Plant Sciences. Research at the Scottish Crop Research Institute is grant-aided by the Scottish Office Agriculture and Fisheries Department. CMoV-NZ and CMoV-M were held at SCRI under a licence issued by the Scottish Office Agriculture and Fisheries Department.