Watermelon mosaic virus

Potyviruses have been badly affecting crop yields in most parts of the world, with Zucchini yellow mosaic virus (ZYMV), Watermelon mosaic virus (WMV) and Papaya ringspot virus (PRSV) being of particular economic importance. Watermelon mosaic virus (WMV) causes severe economic losses in cucurbitaceous, leguminous, malvaceous and chenopodiaceous plants in temperate and Mediterranean regions. It produces chlorosis, mottling, blisters on leaves and fruits, leaf distortion and stunting in watermelon, muskmelon, squash, pumpkin and cucumber. WMV has been shown to infect experimentally, more than 170 plant species belonging to 27 plant families. The biological variability of WMV has been well-documented. Serologically, it is close to Soybean mosaic virus (SMV) and Papaya ringspot virus (PRSV), but distantly related to Potato virus Y (PVY) and Bean yellow mosaic virus (BYMV). The genome of the reported WMV isolates is more than 10kb, flanked by untranslated regions at both the ends. The large open reading frame (ORF) encodes a putative polyprotein of 3217 aa, with a calculated Mr. of 366,904. Sequence analyses of WMV isolates revealed close relationship with the reported isolates of SMV (84.7% to 85.8% aa identity). However, the N-terminal P1 protein encoding region was substantially different, presenting less than 35% identity. SimPlot analysis revealed that WMV arose through an ancestral event of interspecific recombination between SMV and Bean common mosaic virus (BCMV)/ Peanut stripe virus (PSV)-related potyviruses. Very little genetic material resistant to WMV-2 is available. Cultural practices, crop rotation, cross-protection and genetic resistance have effectively been used against WMV. Coat protein transgenic resistance to WMV has also been reported in squash and cantaloupe. _____________________________________________________________________________________________________________


INTRODUCTION
The genus Potyvirus is by far the largest of the known plant virus groups and contains nearly 200 definite and tentative species (Fan et al. 2003;Fauquet et al. 2005).Viruses in this genus are 680-900 nm in length, 11-13 nm in diameter and encapsidate a genome of about 10 kb comprising multiple copies of a single protein species of 30-47 kDa (Shukla et al. 1994).They are transmitted by aphids in a non-persistent manner using helper components.Some of the members are seed-transmitted.Flexuous particles contain (+)sense ssRNA with a 5c VPg, 5c non-coding region, single open reading frame (ORF) encoding a single poly-protein and 3c untranslated region (UTR).The polyprotein is later processed into 10 functional proteins by virus-encoded proteinases (Shukla et al. 1994).On a worldwide basis, potyviruses are badly affecting crop yields, with Zucchini yellow mosaic virus (ZYMV), Watermelon mosaic virus (WMV) and Papaya ringspot virus (PRSV) being of particular economic importance (Lecoq et al. 2001).
WMV, a pathogen of worldwide importance, causes economic losses in Cucurbitaceous, Leguminous, Malvaceous and Chenopodiaceous plants (Purcifull et al. 1984).The biological variability of WMV is well documented (Purcifull and Hiebert 1979;van Regenmortel 1971).Lindberg et al. (1956) carried out the first detailed study of viral isolates from cucurbits and classified the isolates into two groups, namely melon mosaic and squash mosaic groups.Subsequently, WMV isolates were classified as WMV-1 and WMV-2 (Purcifull and Hiebert 1979;Yeh et al. 1984).Specifically, isolates that failed to infect non-cucurbitaceous plants were designated WMV-2, while isolates that infected plants outside the Cucurbitaceae were designated WMV-1, although the latter is now considered to be a strain of PRSV (Purcifull et al. 1984).Purcifull and Heibert (1979) also reported a third isolate that did not react with antisera against either WMV-1 or WMV-2, and which has now been given the status of a distinct potyvirus species, the Moroccan watermelon mosaic virus (MWMV).Soybean mosaic virus (SMV) and Blackeye cowpea mosaic virus (BlCMV) are both closely related to, but distinct from, WMV (Fauquet et al. 2005).Although WMV has long been known in many parts of the world, it has not yet been characterized extensively at the molecular level.The complete genome sequences of only three isolates have been reported very recently from France, Pakistan and China.

ECONOMIC IMPORTANCE
WMV is a virus of world-wide importance in temperate and Mediterranean regions (Lecoq et al. 1998).It presents a broader host range than most other potyviruses, causing agronomic problems in important crops, mostly cucurbits, but also peas (Inouye 1964;Schroeder and Provvidenti 1971) and orchards such as Vanilla fragrans (Wang et al. 1993) and Habenaria radiata (Gara et al. 1997).Experimentally, WMV has been shown to infect more than 170 plant species belonging to 27 families (Shukla et al. 1994).In France, WMV epidemics were observed every year from 1981 to 2002 (Lecoq et al. 2003).It has been reported to be most prevalent in melon-growing regions of the Central Valleys of California (Grafton-Cardwell et al. 1996), Brazil (Yuki et al. 2000) and Spain (Luis-Arteaga et al. 1998).Surveys carried out in cucurbit crops from Pakistan (Ali et al. 2004), South Carolina (Sammons et al. 1989), Israel (Antignus et al. 1989), Lebanon (Abou-Jawdah et al. 2000) and Turkey (Sevik and Arli-Sokmen 2003) revealed WMV-2 as the dominant virus.

SYMPTOMATOLOGY
WMV presents a broader host range than most potyviruses and produces a variety of symptoms.It induces chlorotic spots on Chenopodium amaranticolor and Chenopodium quinoa (Fig. 1A).Systemic mottling or mosaic symptoms are produced on Lagenaria siceraria (Fig. 1B), Cucumis sativus and Citrullus lanatus (Fauquet et al. 2005).Dark green vein banding and leaf distortion are observed in Nicotiana benthamiana, but the symptoms are not very severe.N. benthamiana and Cucurbita pepo have been recommended as suitable propagative hosts for virus purification, owing to the high virus titers in these plants.Systemic mottling and necrosis, accompanied by wilting and premature death, occurs in Pisum sativum.Systemic mottling is also observed in Vicia faba.Severe mosaic symptoms (Fig. 1C), puckering and distortion of the leaves and a shoestring appearance, especially at the margin of the leaves, are the prominent symptoms in C. pepo (Fig. 1D) and Cucumis melo var.flexuosus.No infection occurs when Luffa acutangula, Vigna unguiculata, Nicotiana rustica, N. glutinosa and N. samsun are inoculated with WMV (Fauquet et al. 2005).

MOLECULAR CHARACTERIZATION
The viral genomes of all reported WMV isolates are greater than 10 kb, flanked at both ends by UTRs (Desbiez and Lecoq 2004;Ali et al. 2006).The viral genome encodes a putative polyprotein of 3217 aa with a calculated Mr of 366,904.Nine putative proteinase cleavage sites are present that allow processing of the polyprotein into 10 smaller putative functional proteins by 3 viral-encoded proteinases designated P1, HC-Pro and NIa-Pro.The amino acid sequence contains all the characteristic features of potyviruses (Table 1), including the conserved sequence motifs KLSC and PTK in HC-Pro and DAG in CP, which are essential for transmission by aphids.The highly conserved sequence motif KITC, which is required for interactions with stylet canals, has diverged to KLSC, but the lysine residue regarded as crucial for HC-Pro activity remains unchanged.It is interesting to note that most potyviruses closely related to WMV, including Bean common mosaic virus (BCMV), Bean common mosaic necrosis virus (BCMNV), Cowpea aphid borne mosaic virus (CABMV), SMV, ZYMV and Dasheen mosaic virus (DMV), have their KITC motif diverged to KLSC.This observation supports previous findings that mutations in the KITC motif have no effect on aphid transmissibility and HC-Pro activity (Atreya and Pirone 1993).It has also been documented that highly basic residues, such Lys or Arg, at this position in HC-Pro play key role in potyviral helper component proteinase activity (Atreya and Pirone 1993).
The available sequence data, which are mostly confined to partial CP, P1 and CI regions from only a few parts of the world, reveal very little diversity among WMV isolates.Neighbour-joining trees of different regions of the genome (P1, CI and CP) show little variability (Ali et al. 2006).The phylogenetic relationship with other members of the family  (Desbiez and Lecoq 2004).This subgroup is further containing two clusters, one comprising SMV isolates, including WMV, and the other containing the majority of BCMV isolates, BCMNV and CABMV.Although the subgroups have no formal taxonomic status, they make it easier to deal with related viruses and can be useful for diagnosis and plant breeding.
Since there are reports of new data entries in GenBank that were not used in previous analyses, we drew a new phylogenetic tree based on the CP sequence.This resulted in three clusters (Fig. 2): the first cluster contains isolates reported from Pakistan, France, Spain, Australia and Israel (showing little divergence); the second cluster comprises isolates from USA, Tonga, New Zealand and Spain (BAD and SG isolates); and the third cluster includes Japanese (Habeneria) and Chinese isolates.The Japanese isolate do not fit well in the cluster and show a little more divergence.The Spanish isolates are distributed in two clusters showing some diversity, and can be divided into two genetic strains, I and II, as reported by Moreno et al. (2004).Since the reported sequences of the P1 and CP regions are from different isolates reported at different periods of time, no correlation can be established on the basis of this phylogenetic comparison.Moreno et al. (2004) found no evidence of spatial differentiation of the Spanish WMV population.Furthermore, the Spanish WMV population has no metapopulation structure with local extinction and recolonization, and shows different evolutionary dynamics for different regions of the WMV genome.At least 7% of the Spanish isolates are recombinants between genetic strains I and II, presenting two interesting features: (1) crossover points are not detected over the whole genome, they mostly occur between the analyzed regions in the CI and CP cistrons, and not between the P1 and CI cistrons; (2) crossover points are not observed within the analyzed regions encoding the P1, CI and CP proteins (Moreno et al. 2004).
Sequence analyses revealed that WMV is very closely related to reported isolates of SMV (84.7-85.8%aa identity), although the N-terminal P1 protein-encoding regions differ substantially (<35% aa identity).SimPlot analyses (Ray 1998) using the WMV-Pk isolate as a query sequence and three other isolates from the BCMV group revealed that WMV arose through an ancestral event of interspecific recombination between SMV and BCMV/Peanut stripe virus (PSV)-related potyviruses (Desbiez and Lecoq 2004;Ali et al. 2006).High homology in the 5c coding region (nt 134 to 950) of WMV genome has been observed for PSV and BCMV, after which the trend shifts toward SMV, indicating that only the N-terminal P1 region was subjected to recombination (Ali et al. 2006).Although the available WMV sequence analysis results confirm that this virus is closely related to SMV (Desbiez and Lecoq 2004;Ali et al. 2006), there is a discrepancy in its position regarding its molecular and biological properties.According to the criteria proposed by Shukla et al. (1994), WMV should be considered as a strain of SMV (Yu et al. 1989).However, high variability is observed in the 5c part of the genome between the two isolates in terms of both the percent identity and genome length.Phylogenetic trees drawn for the N-and Cterminal P1, HC-Pro and 5c UTR regions clearly indicate that WMV-Pk converges more toward SMV isolates in the 5c UTR, C-terminal P1 and HC-Pro regions, but is quite distant in the N-terminal P1 region, where the sequence identity is more toward BCMV/PSV isolates, thereby strengthening the assumption that WMV originated as a result of recombination between SMV and BCMV/PSV (Desbiez and Lecoq 2004).The putative recombination spot for the amino acid sequence has been identified (Desbiez and Lecoq 2004).In the 5c UTR, the nucleotide sequence percent identity is again more closely related to SMV than to PSV/BCMV, but the variability in size and sequence makes it impossible to explore any other recombination events.Nevertheless, the percent identities between WMV and other potyviruses remain below 70%, whereas they are above 90% between strains of SMV (Desbiez and Lecoq 2004).This recombination event may have occurred only once in the emergence of the virus, since the P1 sequences of WMV strains from different geographical origins present the same characteristics with a unique putative recombination point.No WMV strains with SMV-like P1 proteins have been observed to-date.
The relatively important divergences between WMV and BCMV or SMV in different parts of the genome indicate that the recombination event took place very early during the evolution and differentiation of potyviruses.
The host range has been widely used in the past to classify WMV strains, but the variability in host range and symptomatology (Dijkstra 1992), cross-reactivity of antisera and ambiguous serological relationships between isolates of the same or different viruses (Shukla et al. 1992) make the situation very ambiguous.Putative homologous recombination, which has been recognized for a number of potyviruses, including Plum pox virus (PPV) (Glasa and Kudela 2001), Potato virus Y (PVY) (Glais et al. 2002), BCMV (Revers et al. 1996), Turnip mosaic virus (TuMV) (Ohshima et al. 2002) and Lettuce mosaic virus (LMV) (Krause-Sakate et al. 2004) adds to the problem of defining reliable criteria for demarcating the taxonomic units.Shukla et al. (1994) proposed that species of the same virus should have a 3c UTR sequence identity of >75%.It has also been proposed that the species demarcation criteria should be <76% nt identity and <82% aa identity, and that the corresponding threshold for individual genes should range from 58% (P1) to 74-78% (other genes).A genus demarcation criterion for the entire ORF of <46% nt identity has been proposed (Adams et al. 2005).For CP, 76-77% nt identity was suggested as the optimal species demarcation criterion (Adams et al. 2005).According to these criteria, WMV-Pk has a 3c UTR sequence identity of 81.4% and a CP sequence identity of 86.5% with the SMV-Svr strain, which are sufficiently high to consider WMV as a strain of SMV.However, unlike SMV, WMV has a broad host range, which is rare for an individual potyvirus.Even the biological and serological properties of WMV and SMV are different.Although phylogenetic analyses have revealed that WMV clustered with SMV isolates in the BCMV subgroup, its full-length genome is longer than those of reported SMV isolates and the P1 region is closer to BCMV/PSV than to SMV, thereby presenting a very confusing situation.Hence, it has been recommended that WMV should be kept as an independent species in the genus Potyvirus rather than categorizing it as a strain of SMV.The CI gene, which most accurately reflects the taxonomic status according to the complete ORF, has been proposed as the best region for diagnostic and taxonomic studies.However, such classification of viruses on the basis of only a fraction of their genomic information requires further reconsideration, and proper weightage should also be given to other taxonomic criteria.

CONTROL
Various strategies for controlling WMV have been used in the past.These stylet-borne non-persistently transmitted viruses are difficult to control through the use of insecticides (Nameth et al. 1986).Two other methods used to protect crops from aphids are reflective mulches (Nameth et al. 1986;Brown et al. 1993;Orozco et al. 1994;Summers et al. 1995) and oil sprays (Simmons and Zitter 1980).Toscano et al. (1979) reported a 90% reduction in viral incidence through the use of aluminum foil mulches and up to 26% reduction using oil sprays.However, aluminum foils are expensive and slow the growth of seedlings, while oil sprays are a potential problem when sprayed under high temperatures and can result in plant injuries.Walters (2003) reported that using row-covers or black or white mulches had greater influences in reducing the incidence of WMV with high total marketable yields.Culture practices, such as weed control and effective rotation, have also been exploited for controlling WMV.Use of resistant varieties is the most effective and inexpensive method of controlling a viral disease.However, very little genetic material resistant to WMV is available.Some tolerance has been reported in the Cantaloupe variety Top-mark (Nameth et al. 1986), breeding line 91213 of C. melo (Moyer et al. 1985), four accessions of C. colocynthis from Iran and one from Morocco.High resistance has been reported in Egun PI 164708 from India (Gillaspie and Wright 1993) and PI 494528 and PI 494532 of C. colocynthis (Provvidenti 1986).Accessions C-885 and C-769 are potential sources of multiple resistance to PRSV-W, WMV-2 and ZYMV (Diaz et al. 2003).Moderate resistance was found in PI 595203 (C.lanatus var.lanatus), an Egusi type originally collected in Nigeria (Xu et al. 2004).The C. melo accession TGR-1551 was also resistant (Diaz et al. 2003).Cross-protection has been used effectively against WMV (Kosaka and Fukunishi 1997).Transgenic resistance is one of the most effective methods of controlling viral diseases.CP transgenic resistance to WMV has been reported in squash and cantaloupe (Namba et al. 1992;Clough and Hamm 1995).The transgenic squash CZW-3 expressing the CP genes of CMV, ZYMV and WMV has also been reported to show high resistance to these three aphid-borne viruses (Fuchs et al. 2004).

FUTURE PERSPECTIVES
The biological variability of WMV is well documented (Purcifull and Hiebert 1979;van Regenmortel 1971), but it has not yet been characterized extensively at the molecular level.The complete genome sequences of only three isolates have been reported and we recommend more sequencing of isolates collected from entirely different agro-ecological zones of the world to make the taxonomic status more clear.According to the reported criteria, WMV could easily be considered a strain of SMV.However, unlike SMV, WMV has a broad host range, which is rare for an individual potyvirus.Even the biological and serological properties of WMV and SMV are different.Sequence analysis suggests the event of recombination and the relatively important divergences between WMV and BCMV or SMV in different parts of the genome indicate that this recombination event took place very early during the evolution and differentiation of potyviruses.Based on these confusing results, it has been proposed to place WMV as separate species in the genus, but to support this hypothesis there is a dire need to sequence more and more isolates.No WMV strains with SMV-like P1 proteins have been observed to-date.As WMV presents a broader host range than most other Potyviruses, experiments must be planed to explore the genomic region, which enabled the virus to have such a large host range.Those viruses which are transmitted by large species of vectors usually have large host range.The interaction of the virus vector also needs to be explored.

Fig. 1 B
Fig. 1 Symptomology of Watermelon mosaic virus.(A) Chlorotic spots on C. amaranticolor.(B) mosaic symptoms induced on L. siceraria leaves, (C) Leaf of Cucurbita pepo showing severe mosaic symptoms infected with WMV and (D) C. pepo showing dark green enation like symptom, severe puckering and distortion of leaves and shoe string appearance, especially at the margin of the leaves.

Table 1
Conserved sequence motifs identified in WMV-Pk polyprotein.