Evidence for Lettuce big-vein associated virus as the causal agent of a syndrome of necrotic rings and spots in lettuce

Authors


E-mail: martin.verbeek@wur.nl

Abstract

Lettuce big-vein associated virus (LBVaV, genus Varicosavirus) was shown to be responsible for characteristic necrotic symptoms observed in combination with big-vein symptoms in lettuce breeding lines when tested for their susceptibility to lettuce big-vein disease (BVD) using viruliferous Olpidium virulentus spores in a nutrient film technique (NFT) system. Lettuce plants showing BVD are generally infected by two viruses: Mirafiori lettuce big-vein virus (MiLBVV, genus Ophiovirus) and LBVaV. New mechanical inoculation methods were developed to separate the two viruses from each other and to transfer both viruses to indicator plants and lettuce. After mechanical inoculation onto lettuce plants MiLBVV induced vein-band chlorosis, which is the characteristic symptom of BVD. LBVaV caused a syndrome of necrotic spots and rings which was also observed earlier in lettuce plants inoculated in the NFT system, resembling symptoms described for lettuce ring necrosis disease (RND). This observation is in contrast with the idea that LBVaV only causes latent infections in lettuce. De novo next-generation sequencing demonstrated that LBVaV was the only pathogen present in a mechanically inoculated lettuce plant with symptoms, providing evidence that LBVaV was the causal agent of the observed necrotic syndrome and thus fulfilling Koch’s postulates for this virus. The necrotic syndrome caused by LBVaV in lettuce is referred to as LBVaV-associated necrosis (LAN).

Introduction

Lettuce big-vein disease (BVD) is characterized by vein clearing and banding, and sometimes by ‘bumpy’ leaves and retarded growth of the heads of lettuce (Lactuca sativa) (Jagger & Chandler, 1934; Huijberts et al., 1990). Unsightliness of the foliage and lighter or even underdeveloped heads greatly affect the market value and cause considerable economic losses (Zink & Grogan, 1954; Walsh, 1994).

Since the first report of BVD in the USA (Jagger & Chandler, 1934), the identity of the causal agent remained uncertain for many years. BVD is transmitted by the chytrid fungus Olpidium virulentus (formerly O. brassicae) (Campbell, 1962; Tomlinson & Garrett, 1962; Campbell & Fry, 1966; Lot et al., 2002; Sasaya & Koganezawa, 2006; Hartwright et al., 2010; Maccarone et al., 2010) and can survive in resting spores of this fungus for more than 20 years (Campbell, 1985), which makes the disease extremely difficult to control. Although BVD was considered a virus-like soilborne disease for decades, it took until 1983 for the first virus-like particles to be observed by electron microscopy in BVD-infected plants (Kuwata et al., 1983). These particles were rod-shaped with a length of about 300–320 nm and resembled those of Tobacco stunt virus (TStV) (Kuwata & Kubo, 1986). It was assumed that the virus found in lettuce and TStV were closely related but distinct, because on the one hand TStV was not able to infect lettuce and on the other hand tobacco was not susceptible to the lettuce virus (Kuwata & Kubo, 1986). However, this was reconsidered when highly similar nucleotide sequences of both viruses became available (Sasaya et al., 2004, 2005). The rod-shaped virions previously observed turned out to be very labile and difficult to purify. Partially purified virus preparations were used to raise an antiserum which was used in ELISA and immunosorbent electron microscopy (ISEM) (Vetten et al., 1987). Huijberts et al. (1990) demonstrated mechanical transmission to various indicator plants and back-transmission of the BVD-causing agent to lettuce using zoospores of O. virulentus.

The genome of this rod-shaped virus consists of two negative-sense single-stranded (ss) RNA molecules, approximately 6·8 kb (RNA-1) and 6·1 kb (RNA-2) in size (Sasaya et al., 2001, 2002, 2004). This virus was believed to be the causal agent of BVD and named Lettuce big-vein virus (LBVV), and designated the type member of the genus Varicosavirus (Mayo, 2000).

In 2000 another virus, assigned to the genus Ophiovirus, that appeared to be related to BVD was found. In Italy this ophiovirus, initially named Mirafiori lettuce virus and now known as Mirafiori lettuce big-vein virus (MiLBVV), was isolated from a BVD-infected lettuce plant (Roggero et al., 2000). Although this virus proved to be extremely labile, virions could be purified without loss of infectivity, viral RNA was isolated and the complete genome sequence, consisting of four ssRNA molecules, was elucidated (Van der Wilk et al., 2002). This virus was able to induce the characteristic lettuce big-vein symptoms after vector-mediated transmission to lettuce plants. While lettuce inoculated with LBVV only showed latent infections, it was concluded that MiLBVV was the causal agent of BVD (Lot et al., 2002; Sasaya et al., 2008). During field surveys BVD-affected lettuce plants were found to contain both MiLBVV and LBVV, while plants with single infections were rarely found. However, no correlation was found between the occurrence of BVD symptoms and the presence or absence of LBVV (Roggero et al., 2003). After recognition of MiLBVV as the causal agent of BVD, LBVV was renamed Lettuce big-vein associated virus (LBVaV) (Fauquet et al., 2005).

Breeding companies use different methods to test lettuce breeding lines for susceptibility or resistance to BVD. One of these methods comprises a nutrient film technique (NFT) containing viruliferous spores of O. virulentus. One of the breeding companies using the NFT method reported that, besides the characteristic symptoms related to BVD, a syndrome of necrotic rings and patches could often be observed. Infected plants from this system usually test positive for both MiLBVV and LBVaV and always negative for the ophiovirus Lettuce ring necrosis virus (LRNV) (J. Schut, Rijk Zwaan, De Lier, The Netherlands, personal communication). This paper describes the separation of MiLBVV and LBVaV from a dual-infected lettuce plant originating from this NFT system by means of mechanical inoculation using novel buffers. In addition, this study reports the use of mechanical inoculation experiments and de novo next-generation sequencing (Metzker, 2010) to show, for the first time, that necrotic symptoms in lettuce could be attributed to LBVaV. The name LBVaV-associated necrosis (LAN) was chosen for this disease.

Materials and methods

Virus isolates and maintenance

A crisphead lettuce plant, infected using the NFT system and showing clear BVD and LAN symptoms, was a kind gift of J. Schut (Rijk Zwaan, De Lier, The Netherlands). This isolate was designated Ls301 and added to the virus collection of Plant Research International in 1999. It was maintained by transferring lettuce plants cv. Summertime to a 50% mixture of fresh soil and air-dried soil containing root parts from infected plants. The plants were kept in a greenhouse at 16°C under low light intensities. LRNV isolate Ls247, originating from 1988 (Bos & Huijberts, 1996), was kept in air-dried soil at 4°C and was revived from the collection by planting lettuce cv. Patty plants in this soil and keeping them in a greenhouse under the same conditions as isolate Ls301.

Mechanical inoculation

In mechanical inoculation studies various buffers were used with pH ranging from 5 to 9. Two successful buffers were used in further studies: 0·1 m HEPES (4-(2-hydroxyethyl)piperazine-1-ethanesulphonic acid) buffer (Sigma-Aldrich) pH 7·8 and 0·1 m MES (4-morpholineethanesulphonic acid) buffer (Sigma-Aldrich) pH 5·5, both containing 20 mm Na2SO3, 10 mm Na-DIECA and 5 mm Na-EDTA. Infected leaves were ground in a 1:1 (w/v) ratio in ice-cold buffer. Mortars and pestles were chilled and kept on ice during the inoculation procedure. Indicator plants Chenopodium quinoa, Nicotiana benthamiana, N. hesperis 67A, N. megalosiphon, Noccidentalis P1 and N. occidentalis 37B were dusted with carborundum and inoculated. The plants were grown in a greenhouse at 20°C, with a day length of 16 h.

Lettuce plants of cvs Patty (butterhead type), Valmaine (romaine or cos type), Winnie (Batavian type) and Summertime (iceberg or crisphead type) were grown in a greenhouse at 16°C without additional light. When the lettuce plants reached the two- or three-leaf stage, they were given a dark period of 2 days by placing them under cardboard boxes. The plants were dusted with carborundum and inoculated. The inoculation procedure was repeated after a 2-day interval, while in between the two inoculations the lettuce plants were kept in the dark. The plants were then grown in the same greenhouse for symptom development without using additional light.

Virus purification and antiserum production

Virus particles of both LBVaV and MiLBVV were purified essentially as described previously by Roggero et al. (2000) but using 0·1 m Tris–HCl pH 8 containing 20 mm sodium sulphite, 10 mm Na-DIECA and 5 mm Na-EDTA as homogenization buffer. The pellets obtained after the sucrose cushion step were resuspended in 1 mL 0·1 m Tris–HCl pH 8, layered onto a continuous 10–40% Cs2SO4 gradient and subjected to isopycnic gradient centrifugation for 17 h at 200 000 g in a Beckman SW60 rotor. Virus bands were collected and concentrated by centrifugation for 3 h at 150 000 g. The purity of each virus preparation was checked by electron microscopy and SDS-PAGE. Antisera were raised by giving a rabbit two subcutaneous injections of an emulsion of purified virus (total amount 100 μg) and Freund’s incomplete adjuvant with a 3-week interval. Blood serum was collected 3 weeks later and ELISA-grade coating and conjugate were produced. Both are available through Prime Diagnostics, Wageningen, the Netherlands.

Virus detection

Mechanically inoculated plants were carefully monitored for single infections by double antibody sandwich (DAS)-ELISA, using the antisera against LBVaV and MiLBVV (Prime Diagnostics). The plants were checked for the presence or absence of LRNV in a triple antibody sandwich (TAS)-ELISA with monoclonal antibodies kindly provided by H.-J. Vetten, JKI, Braunschweig, Germany.

Alternatively, total RNA was extracted from systemically infected leaf material with the aid of the RNeasy Plant Mini kit (QIAGEN) following the manufacturer’s instructions. RT-PCR was used subsequently to verify LBVaV infections with primers CPp and CPn (Sasaya et al., 2001) and MiLBVV and LRNV infections with generic ophiovirus primers OP1 and OP2 (Vaira et al., 2003).

De novo next-generation sequencing

Total RNA was extracted from 1 g systemically LBVaV-infected leaf material of a lettuce plant cv. Patty showing necrotic symptoms after mechanical inoculation. The leaf material was placed in an extraction bag (BIOREBA), frozen in liquid nitrogen and ground to a powder which was resuspended in 4 mL 0·1 m MES pH 5·5 containing 20 mm Na2SO3, 10 mm Na-DIECA and 5 mm Na-EDTA. The suspension was centrifuged for 10 min at 10 000 g in an Eppendorf centrifuge. One millilitre of the supernatant was mixed with 500 μL of a 24% polyethylene glycol (PEG) 6000/6·9% NaCl solution in an Eppendorf tube. After incubation on ice for 1 h, the suspension was centrifuged for 20 min at 20 000 g. The pellet was resuspended in 200 μL nuclease-free water and subsequently centrifuged for 10 min at 10 000 g to remove the PEG. RNA was isolated from 100 μL suspension using the RNeasy Plant Mini kit (QIAGEN) according to the manufacturer’s instructions. DNA was removed from the samples by the On-Column DNase Digestion with the RNase-Free DNase Set (QIAGEN) according to the manufacturer’s protocol. RNA was eluted from the columns with 50 μL nuclease-free water and the concentration was measured using the Quant-iT RiboGreen RNA Reagent and Kit (Invitrogen). cDNA was synthesized using random primers, tagged and amplified according the protocol of Adams et al. (2009). The tagged sample was sequenced using an FLX Titanium, Roche/454 genome sequencer (Business Unit Bioscience, Plant Research International). Sequence reads were assembled after removing the tag and primer sequences using the CLC genomics workbench (CLC Bio). blastn (Zhang et al., 2000) and blastx (Altschul et al., 1997) searches with the contigs were performed against the GenBank non-redundant protein and nucleotide databases.

Results

Separation of the viruses from the BVD complex

Initial experiments with mechanical inoculations from BVD-infected lettuce plants to C. quinoa, using various buffers ranging from pH 5 to pH 9, showed two optima when local lesions were counted (data not shown). The first optimum was found with a buffer of pH 7·8 (0·1 m HEPES) and the second optimum was obtained using a buffer with pH 5·5 (0·1 m MES). Remarkably, two types of local lesions were observed in this test plant: (i) greenish rings and (ii) small necrotic lesions (Fig. 1). The latter were more numerous when the MES buffer was used.

Figure 1.

 Local lesions in Chenopodium quinoa after mechanical inoculation with (a) Mirafiori lettuce big-vein virus (Ls301-O, 14 days post-inoculation, dpi) and (b) Lettuce big-vein associated virus (Ls302, 7 dpi).

Both buffers were then used to inoculate Ls301-infected lettuce samples to different indicator plants. When the HEPES buffer was used for repeated passages through N. occidentalis P1, LBVaV was lost while MiLBVV persisted. The MES buffer, however, seemed to favour transmission of LBVaV, although MiLBVV was also transmitted using this buffer. Inoculations from lettuce to a number of Nicotiana species resulted in mixed infections. However, N. megalosiphon was shown to be a moderate host for both viruses. Of five inoculated plants, one became infected with MiLBVV and another one with LBVaV. Neither of these two plants showed characteristic symptoms, but the single infections were detected using ELISA. From the LBVaV-infected N. megalosiphon the virus was successfully transmitted to N. benthamiana. Thus, the two viruses present in isolate Ls301 were separated and treated further as different isolates: MiLBVV isolate Ls301-O (Van der Wilk et al., 2002) and LBVaV isolate Ls302. Ls301-O was maintained in the greenhouse in N. occidentalis P1 and Ls302 in N. benthamiana or N. occidentalis 37B.

Symptoms of Ls301-O and Ls302 in indicator plants

The local lesion host C. quinoa reacted to Ls301-O with chlorotic spots appearing 1 week after inoculation, later changing into green rings (Fig. 1a). Nicotiana occidentalis P1, N. hesperis 67A and N. occidentalis 37B reacted in 5–7 days, with necrotic local lesions appearing as light brown rings. After another week systemic symptoms appeared: veinal necrosis, mottling and necrotic spots. Ls301-O was able to infect N. benthamiana, but this plant reacted very poorly, with faint systemic chlorosis or no visible symptoms at all.

After inoculation of Ls302 from N. megalosiphon to N. benthamina, local lesions appeared after 5 days as dark brown rings with a light centre, rapidly increasing in diameter and spreading into the veins of the local leaf. After 7–9 days systemic necrosis was observed, leading to curling of the younger leaves, necrosis of leaf stems and stem, stunting and in some cases eventual death of the plant. Using systemically infected N. benthamiana leaves with necrotic symptoms as inoculum, the varicosavirus was readily transmitted to other indicator plants. Chenopodium quinoa reacted upon LBVaV inoculation with pinpoint-like necrotic local lesions, which appeared 6–7 days post-inoculation (Fig. 1b). Nicotiana occidentalis P1 and N. occidentalis 37B showed dark brown necrotic rings after 4–6 days, resembling those on N. benthamiana. Systemic infection in N. occidentalis P1 resembled infection with the ophiovirus Ls301-O, but resulted in more stunting of the plant after 3–4 weeks. Systemic symptoms in N. occidentalis 37B appeared after 2 weeks and could be recognized as yellow leaves or parts of the leaves in which dark brown or black necrotic spots were present. Chenopodium amaranticolor, N. tabacum cv. White Burley and N. glutinosa were not susceptible to the Ls302 isolate.

Mechanical inoculation of lettuce

Ls302 was mechanically inoculated from N. benthamiana with symptoms onto lettuce cvs Patty, Valmaine, Winnie and Summertime. Inoculations were conducted in multiple experiments and the results are combined in Table 1. Only 3% of the cv. Patty plants became infected, 10% of cvs Winnie and Summertime, and 16% of the inoculated Valmaine plants. When LBVaV-infected Valmaine was used as a virus source to inoculate Valmaine and Winnie, infection rates increased to 48% and 25%, respectively (Table 1). In all four lettuce cultivars the LBVaV-infected plants showed symptoms. Symptoms in cv. Patty developed approximately 20 days post-inoculation and consisted of yellow spots containing brown necrotic spots and occasionally rings (Fig. 2b), resembling the symptoms observed in lettuce plants inoculated in the NFT system (Fig. 2a). The infection became systemic and caused a pattern of necrotic spots throughout all leaf stages. In Valmaine (Fig. 2d) necrotic pinpoint local lesions could be observed 10–14 days post-inoculation, which later changed into necrotic spots and rings. Systemic infection was visible in the youngest leaves as chlorotic spots, later becoming necrotic spots and rings. Also, on the underside of the leaves light brown or orange rings appeared. Cultivar Winnie (Fig. 2e) showed small necrotic local lesions and systemic necrotic spots and ring-like patterns, also on the underside of the leaves. Lettuce cv. Summertime (Fig. 2f) developed larger ring-like necrotic patterns on the underside of the middle leaves of the plant.

Table 1. Number of plants infected and showing symptoms after mechanical inoculation of lettuce cultivars with Mirafiori lettuce big-vein virus (MiLBVV) or Lettuce big-vein associated virus (LBVaV)
VirusLettuce cultivar
PattyValmaineWinnieSummertime
InfectedSymptomsInfectedSymptomsInfectedSymptomsInfectedSymptoms
  1. n.t.: not tested.

  2. aInoculum source: Nicotiana occidentalis P1.

  3. bNumber of infected plants/number of inoculated plants.

  4. cInoculum source: Nicotiana benthamiana.

  5. dInoculum source: lettuce cv. Valmaine.

MiLBVVa3/12b0/122/161/161/101/101/160/16
LBVaVc1/301/302/122/121/101/101/101/10
LBVaVdn.t.n.t.10/2110/214/164/16n.t.n.t.
Figure 2.

 (a) Symptoms of Lettuce big-vein associated virus (LBVaV)-associated necrosis on the underside of a lettuce leaf after inoculation with lettuce big-vein disease in the NFT system at Rijk Zwaan’s facilities (courtesy of Rijk Zwaan). (b, c) Lettuce cv. Patty showing symptoms caused by LBVaV (b) and Lettuce ring necrosis virus (LRNV; c) after mechanical inoculation; symptoms of both viruses appear very similar – sometimes symptoms of LRNV tends to be more ring-like and symptoms of LBVaV more patchy. (d–f) Systemic symptoms caused by LBVaV in lettuce cvs Valmaine (d), Winnie (e) and Summertime (f) after mechanical inoculation.

MiLBVV isolate Ls301-O could be transferred from N. occidentalis P1 to lettuce cvs Patty, Summertime, Valmaine and Winnie. Infection of these cultivars did not lead to symptoms, with the exception of one Valmaine and one Winnie plant which showed only vein-band chlorosis, the characteristic symptom of BVD. This observation supports other reports that MiLBVV is the causal agent of BVD (Lot et al., 2002; Sasaya et al., 2008).

As the symptoms of LBVaV Ls302 are similar to the ones described for lettuce ring necrosis disease, plants of lettuce cv. Patty infected with Ls302 were compared with plants inoculated with the LRNV isolate Ls247 (Bos & Huijberts, 1996). The latter also showed necrotic rings, but they were clearly coalescing into large ring-like patterns all over the leaf (Fig. 2c). Samples of plants infected with isolate Ls247 did not react with the LBVaV antiserum in DAS-ELISA, but did give positive reactions in the LRNV-specific TAS-ELISA (Fig. 3).

Figure 3.

 Specificity of antisera used in DAS- and TAS-ELISA in this study. Absorbance readings (A405) after 30 min of substrate incubation at room temperature are plotted for healthy lettuce and for lettuce samples infected with Lettuce ring necrosis virus (LRNV, isolate Ls247), Mirafiori lettuce big-vein virus (MiLBVV isolate Ls301-O), Lettuce big-vein associated virus (LBVaV, isolate Ls302) or LBVaV+MiLBVV. All samples were tested in duplicate. Error bars represent standard deviation.

All Ls302-inoculated lettuce plants showing necrotic symptoms reacted positively in a LBVaV-specific DAS-ELISA. All plants reacted negatively in ELISAs with antisera against MiLBVV and LRNV. RT-PCR tests were positive with specific primers for LBVaV, and negative with generic primers for the genus Ophiovirus (including MiLBVV and LRNV), confirming that the infected lettuce plants were infected by LBVaV and not by MiLBVV and LRNV. Back-inoculation of the infected lettuce plants onto N. benthamiana resulted in the characteristic necrotic local lesions and systemic necrosis. None of the inoculated lettuce plants showed latent infection with Ls302.

To demonstrate that no other unknown viral agents were present in the Ls302-infected lettuce plants, one sample of a plant of cv. Patty with symptoms was subjected to 454 sequencing (so-called deep sequencing). This gave 21 058 reads with a maximum length of 559 nts.

De novo assembly of these reads resulted in 1003 contigs with lengths varying from 123 to 6719 nt. blastn and blastx searches against the GenBank non-redundant protein and nucleotide databases were done with all contigs. Only three contigs contained viral sequences, while all other contigs contained sequences from plant origin.

The first contig contained 2765 reads and had a length of 6719 nt. This contig showed 95% sequence identity with LBVaV RNA 1 (AB075039). The second contig contained 625 reads and had a length of 5066 nt. It showed 94% sequence identity with LBVaV RNA 2 (AB114138). The third contig contained 61 reads and was 998 nt long. This contig showed 96% sequence identity with LBVaV RNA 2 (AB114138). Alternative assembly of all Ls302 reads using LBVaV RNA1 (6797 nt) and RNA2 (6081 nt) (AB075039 and AB114138, respectively) as reference sequences resulted in two consensus sequences. The Ls302-RNA1 partial sequence had a length of 6791 nt and showed 95% identity with AB075039. In comparison with the reference sequence 3 nt were missing at the 5′ end and 4 nt at the 3′ end. The Ls302-RNA2 partial sequence had a length of 6075 nt and showed 95% identity with AB114138. In comparison with the reference RNA2 sequence 3 nt were missing at both ends. No other sequences, besides those of plant origin, were found in this sample. The nucleotide sequences reported in this article are available under GenBank accession numbers JN710440 (LBVaV-Ls302-RNA1) and JN710441 (LBVaV-Ls302-RNA2).

Discussion

The rod-shaped particles of LBVaV are almost always found in BVD-infected lettuce plants and have long been thought to be the causal agents of BVD (Kuwata et al., 1983; Vetten et al., 1987; Huijberts et al., 1990). Only recently were the ophiovirus-like particles of MiLBVV, which were overlooked for a long time because of their poor visibility in electron microscopy, associated with BVD (Roggero et al., 2000). Later, when it became clear that MiLBVV causes big-vein symptoms, it became a general belief that LBVaV does not cause symptoms in lettuce (Lot et al., 2002). Because of the labile character of LBVaV and the difficulty of mechanical transfer to indicator plants or lettuce it has taken a long time for the characteristics of this virus to be described. LBVaV can cause a syndrome in lettuce that is easily confused with that of lettuce ring necrosis disease (RND) for which LRNV is the causal agent (Torok & Vetten, 2002). In the present study, LRNV-infected lettuce plants sometimes showed necrotic rings all over the leaf, whereas LBVaV-infected plants of the same lettuce cultivar showed necrotic spots with a more patch-like appearance. However, this difference in appearance was not always clear and visual discrimination between the two viruses seems to be very difficult (Fig. 2b,c). It is likely that the symptoms of LBVaV are commonly confused with symptoms of LRNV, especially because LBVaV, MiLBVV and LRNV may occur in mixed infections.

Interestingly, in the field, plants are usually infected with both MiLBVV and LBVaV and necrotic symptoms do not appear to be common in these plants. Obviously, the presence of LBVaV in mixed infections does not automatically lead to necrotic symptoms and it is likely that other factors such as virus titre, climate conditions and substrate play a role in disease development, as may yet-unknown interactions between MiLBVV and LBVaV. A better understanding of the precise effects of mixed infections and environmental conditions on symptom expression and development may aid in minimizing economic losses but will require further investigations.

Two other lettuce diseases reported in the past, ‘maladie de taches orangées’ (Martin & Gauvrit, 1989; Campbell & Lot, 1990) and ‘nécrose annulaire de la laitue’ (Verhoyen et al., 1985), were described as soilborne, transmitted by O. virulentus, and causing necrotic symptoms in lettuce. There was also a single report from the USA of a virus-like disease in romaine lettuce vectored by O. virulentus causing chlorosis and necrosis of the older leaves (Patterson & Grogan, 1986). Because of the similarity in the symptoms described and the fact that the described diseases were transmitted via O. virulentus, it might be possible that LBVaV (and/or LRNV) is involved in the aetiologies of the above-mentioned diseases.

Mechanical inoculation methods for BVD-associated viruses reported to date were commonly based on phosphate buffer with a pH range from 7 to 8·3 (Huijberts et al., 1990; Roggero et al., 2000; Lot et al., 2002), although it is likely that those experiments were carried out with mixtures of LBVaV and MiLBVV. In the present study a 0·1 m MES buffer with pH 5·5 was appropriate for maintaining LBVaV in N. benthamiana and a 0·1 m HEPES buffer with pH 7·8 worked well for the maintenance of MiLBVV in N. occidentalis P1, in both cases resulting in high virus titres and enabling mechanical back-inoculation to lettuce. Successful inoculation experiments to infect lettuce plants with the BVD causal agent were reported for the first time in 2002 where it was demonstrated that MiLBVV is the causal agent of big-vein symptoms, but infection with LBVaV was latent (Lot et al., 2002). Mechanical transmission of both viruses to lettuce was not possible when too many passages in alternative host plants were made (Lot et al., 2002). In the present paper it was demonstrated that by using novel inoculation buffers several passages through indicator plants did not affect the ability of both viruses to infect lettuce, although infection rates were very low. Four different cultivars of lettuce were tested, belonging to different lettuce types, and LBVaV induced similar necrotic symptoms in these four cultivars. The infection rate seemed to differ between the cultivars; however, the experiments were not set up to test differences in the susceptibility of lettuce cultivars to MiLBVV and LBVaV.

After successfully transferring LBVaV to lettuce plants, observing symptoms and testing the inoculated plants with ELISA and PCR for presence of viruses known to be involved in O. virulentus transmitted diseases, there was still some doubt whether any unknown agent could be involved in causing the necrotic symptoms in lettuce. After all, Koch’s postulates were not fulfilled so far, as purified virus particles have lost infectiousness. When next-generation sequencing became available, it became possible to rule out the involvement of other pathogens in the observed disease. Using the 454 sequencing technology on RNA isolated from a mechanically inoculated LBVaV-infected lettuce plant of cv. Patty, only viral sequences belonging to LBVaV were found. Interestingly, almost the full genomic sequence of both RNAs of the LBVaV Ls302 isolate could be reconstituted from the obtained sequencing reads. Through the induction of the typical necrotic symptoms by mechanical inoculation of LBVaV onto lettuce plants, followed by determination of the presence of both LBVaV RNA sequences and the absence of viral sequences other than those of LBVaV in these plants, it was shown that isolate Ls302 of LBVaV is the causal agent of LAN disease in lettuce. Accordingly, with the aid of next-generation sequencing, Koch’s postulates have now been fulfilled for this virus.

Acknowledgements

Part of this work was supported by the European Commission under contract number QLRT-1999-01471. The authors thank all their colleagues contributing to the DISCOVAR project for their joint effort to understand better the impact of ophioviruses and varicosavirus in lettuce. We are grateful to J. Schut and J. Folders of Rijk Zwaan for sharing materials and information on the symptomatology of LBVaV in their inoculation system. We thank E. Schijlen, W. te Lintel Hekkert and H. C. van de Geest for their help with 454 sequencing and data analysis.

Ancillary