Low genetic diversity of Squash vein yellowing virus in wild and cultivated cucurbits in the U.S. suggests a recent introduction

Virus Research 163 (2012) 520–527

Contents lists available at SciVerse ScienceDirect

Virus Research journal homepage: www.elsevier.com/locate/virusres

Low genetic diversity of Squash vein yellowing virus in wild and cultivated cucurbits in the U.S. suggests a recent introduction夽 Craig G. Webster, Scott Adkins ∗ United States Department of Agriculture-Agricultural Research Service, U.S. Horticultural Research Laboratory, Fort Pierce, FL 34945, United States

a r t i c l e

i n f o

Article history: Received 9 November 2011 Received in revised form 19 November 2011 Accepted 20 November 2011 Available online 28 November 2011 Keywords: Ipomovirus Potyviridae Squash vein yellowing virus Recombination

a b s t r a c t Squash vein yellowing virus (SqVYV) isolates were collected from cultivated and weedy cucurbits representing major hosts and locations in the U.S. and analyzed to better understand the diversity and population structure. No differences in symptoms were observed in field-collected isolate source plants or subsequently inoculated greenhouse plants, and the complete genome of an SqVYV isolate from a wild cucurbit host (smellmelon, Cucumis melo var. dudaim) was highly similar (99.4% nucleotide identity, 99.3% amino acid identity) to the previously published type isolate from squash. Although analysis of the coat protein (CP) and two serine proteases (P1a and P1b) sequences for 41 isolates showed little diversity across seven years of sampling, it revealed two distinct groups of SqVYV isolates with low intra-group diversity. Our analyses also suggested that recombination had occurred between SqVYV isolates, similar to other ipomoviruses. Selection pressures on the genome regions analyzed were negative indicating purifying selection was occurring. The magnitude of negative selection in SqVYV was consistent with what has been reported for other ipomoviruses, and was greatest for the CP and least for the P1b. The observed genetic diversity was similar to that reported for Cucumber vein yellowing virus but less than that reported for Sweet potato mild mottle virus, Cassava brown streak virus and Ugandan cassava brown streak virus. Collectively, these results indicate that the current U.S. population of SqVYV has undergone a recent genetic bottleneck and was introduced from elsewhere. Published by Elsevier B.V.

1. Introduction The family Potyviridae is one of the largest and most economically important groups of plant viruses and is currently classified into six genera based on vector specificity and genetic relatedness. Members of the genus Ipomovirus [named for the type species Sweet potato mild mottle virus (SPMMV)] are transmitted by whiteflies (e.g. Adkins et al., 2007; Cohen and Nitzany, 1960; Maruthi et al., 2005). Five species are now recognized as members of the genus Ipomovirus although significant differences exist in their genome organization. Squash vein yellowing virus (SqVYV) causes watermelon vine decline in Florida (Adkins et al., 2007). Cucumber vein yellowing virus (CVYV) induces foliar chlorosis and fruit symptoms in cucurbits in the Mediterranean Basin (Cohen and Nitzany, 1960; Cuadrado et al., 2001; Lecoq et al., 2000; Louro et al., 2004; Janssen et al., 2002) and Africa (Desbiez et al., 2001; Yakoubi et al., 2007).

夽 Disclaimer: Mention of a trademark, warranty, proprietary product or vendor does not constitute a guarantee by the U.S. Department of Agriculture and does not imply its approval to the exclusion of other products or vendors that also may be suitable. ∗ Corresponding author. Tel.: +1 772 462 5885; fax: +1 772 462 5986. E-mail address: [email protected] (S. Adkins). 0168-1702/$ – see front matter. Published by Elsevier B.V. doi:10.1016/j.virusres.2011.11.017

Sweet potato mild mottle virus (SPMMV) causes foliar mottle in sweet potato in Africa (Hollings et al., 1976; Tairo et al., 2005). Cassava brown streak virus (CBSV) and Ugandan cassava brown streak virus (UCBSV) cause cassava brown streak disease, characterized by vein yellowing of leaves and necrosis on stems and tuberous roots of cassava (Manihot esculenta; e.g. Alicai et al., 2007; Hillocks and Jennings, 2003; Legg et al., 2011). Both CBSV and UCBSV are present in East Africa although the two species show genetic and geographic separation (Mbanzibwa et al., 2009b, 2011a; Monger et al., 2010; Winter et al., 2010). The currently unassigned Potyviridae species Tomato mild mottle virus (TomMMoV) found infecting solanaceous plants in Africa and the Middle East (Hiskias et al., 2001; Monger et al., 2001; Walkey et al., 1994) was recently reported as another ipomovirus (Abraham et al., 2011). Ipomoviruses use a single positive sense RNA genome encoding one polyprotein that is processed into ten mature viral proteins with an additional protein (PIPO) predicted to be expressed as a fusion with the N-terminal region of the P3 protein (Chung et al., 2008). The genome encodes all proteins necessary for virus replication, movement and transmission, similar to other members of the Potyviridae (reviewed in Fauquet et al., 2005). Apart from these similarities, however, major variations in the genome organization of current ipomoviruses occur. The type species, SPMMV, is most similar to the aphid transmitted members of the genus Potyvirus

C.G. Webster, S. Adkins / Virus Research 163 (2012) 520–527

with a single serine type protease (P1). The SPMMV P1 protein functions as a suppressor of silencing (Giner et al., 2010) instead of the helper component protease (HC-Pro), which functions for this purpose in other members of the Potyviridae (Anandalakshmi et al., 1998). Interestingly, SqVYV and CVYV encode two P1 proteins (P1a and P1b) and lack an HC-Pro (Janssen et al., 2005; Li et al., 2008; Mbanzibwa et al., 2009a; Valli et al., 2006). The P1b of CVYV was demonstrated to function as a suppressor of silencing with a significantly different mode of action from that reported for the SPMMV P1 (Valli et al., 2008; Giner et al., 2010). Sequence identity and net charge of the two P1 proteins of SqVYV and CVYV are different with P1a showing higher identity with those of the aphid transmitted potyviruses, whereas P1b is more related to the P1 protein of the tritimoviruses (Li et al., 2008; Valli et al., 2007). CBSV and UCBSV use a third genome organization strategy and encode a single P1 protease (like SPMMV), lack an HC-Pro (like SqVYV and CVYV) and encode a Maf/HAM1 pyrophosphatase (that is unique among all other ipomoviruses and most other plant viruses), which has been suggested as a mechanism for reducing viral mutagenesis in these two species (Mbanzibwa et al., 2009a). These differences in genome organization of current ipomoviruses suggest that future taxonomic reorganization of these species may be necessary. SqVYV has been widespread in cucurbit crops in southwest and west-central regions of Florida since 2002 (Adkins et al., 2010), and the virus has also been reported in Indiana (Egel and Adkins, 2007). The virus is an economic problem in watermelon production where it causes watermelon vine decline, a severe disease characterized by wilting, stem and rind necrosis, and complete plant death (Adkins et al., 2007, 2011). Reservoirs hosts of SqVYV include the cucurbit weeds Balsam apple (Momordica charantia; Adkins et al., 2008) and smellmelon (Cucumis melo var. dudaim; Adkins et al., 2009, 2011), and cultivated cucurbits like squash. In fact, all known hosts of SqVYV are cucurbits unlike for the closely related CVYV, which has several reported non-cucurbit hosts (Janssen et al., 2002; Morris et al., 2006) in addition to economically important cucurbit crop hosts (e.g. cucumber and watermelon). We examined SqVYV isolates collected from the currently known geographic range of the virus and compared the genetic diversity and selection pressures exerted on different coding regions. The results of this study show that SqVYV isolates from the U.S. have low genetic diversity within both cultivated and weedy cucurbit host species, which probably indicates the population has recently gone through a genetic bottleneck. This suggests that SqVYV was likely introduced from elsewhere and has since become established in cucurbits in Florida. 2. Materials and methods 2.1. Virus sources and maintenance Virus isolates were collected over seven years from a variety of hosts and locations, and named based on these parameters as shown in Table 1. Mechanical inoculations of squash (Cucurbita pepo) cv. Prelude II (Seminis Seeds, Oxnard, CA) and watermelon (Citrullus lanatus) cv. Crimson Sweet were made using 20 mM sodium phosphate buffer (pH 7.0) with 1% (w/v) Celite. Plants were maintained in a temperature controlled greenhouse.

521

1997) as previously described (Adkins et al., 2007). Reverse transcription (RT)-PCR of the 5 and 3 regions of the polyprotein used SqVYV-specific primers for the coat protein (CP; Adkins et al., 2008), P1a [P1av (5 -ATGGCTCAAGTTTACGACTTTAAA3 ) and P1avc (5 -GTATTCATCAATGCAAGTAATTCTCT-3 )] and P1b [P1bv (5 -TCAAGTACGGGCATGAAGTGCATA-3 ) and P1bvc (5 -ATATAAGTCAATACTGCAAGCTCTTC-3 )]. To amplify and sequence the complete genome of isolate SM2008cHe, originally collected from an infected smellmelon (Adkins et al., 2009), double-stranded (ds) cDNA was made using primers FLICv (5 CATAGGATTTAGGTCACACTATAGGGATCCAAAATAAACATTACATGAAC3 ) and FLICvc (5 -TAATACGACTCACTATAGGGAGAGGATCCT18 GG3 ). Short (∼900–1000 bp), overlapping regions of the genome were amplified using primers as previously described (Li et al., 2008). In addition, primers corresponding to the P1a and P1b regions (Li et al., 2008) were tested on isolates for which no amplicon was produced using the above primers. The 5 and 3 termini were amplified using T7 or SP6 primers (which hybridize to FLICv or FLICvc sequences at the termini of the ds cDNA) in combination with SqVYV-specific primers corresponding to the PIa (5 terminus; Li et al., 2008) or CP (3 terminus; Adkins et al., 2008) regions. Amplicons were analyzed on 1% agarose gels and excised, purified and ligated into pGEM-T (Promega, Madison, WI). Ten clones per amplicon were sequenced in both directions on an ABI3730XL automated sequencer at the USHRL DNA Sequencing Support Laboratory. 2.3. Sequence analysis and phylogenetic reconstruction Sequences from this study were deposited in GenBank, after primer sequences were removed, and assigned accession numbers as shown in Table 1. Alignment and phylogenetic analysis of nucleotide (nt) and deduced amino acid (aa) sequences were as previously described (Chellemi et al., 2011). Briefly, p-distance models were used to calculate nt and aa sequence diversity with both neighbor-joining and maximum parsimony methods used for phylogenetic trees, and 1000 bootstrap replications in MEGA 4.1 (Tamura et al., 2007). Selection pressures on genes were estimated from the ratio of the nonsynonymous (dN ) to synonymous (dS ) substitution rates (ω) for each individual gene, as well as for each codon using the Datamonkey online positive selection interface (Kosakovsky Pond and Frost, 2005a) available at http://www.datamonkey.org/. Negative (or purifying) selection was indicated by a ratio of ω < 1, neutral selection was indicated by a ratio of ω = 1 and positive selection was indicated by a ratio of ω > 1. Three methods of detecting selection were used: single likelihood ancestor counting (SLAC), random effects likelihood (REL) and fixed effects likelihood (FEL) (Kosakovsky Pond and Frost, 2005b) with default conditions including the significance level. Concatenated sequences of the P1a, P1b and CP genes were examined for recombination using the GARD program available at the Datamonkey server. The HKY85 nucleotide substitution bias model was selected for all analyses using the automatic model selection tool provided. Sliding window comparisons of nucleotide sequences were made using SimPlot (Version 3.5; Lole et al., 1999) using a window and step size of 200 nt and 20 nt, respectively. 3. Results

2.2. Amplification, cloning and sequencing

3.1. Comparison of symptoms induced by SqVYV isolates

Total RNA was extracted using an RNeasy Plant Mini kit (Qiagen, Valencia, CA) and used to make first strand cDNA with a degenerate potyvirus reverse primer (Gibbs and Mackenzie,

Decline of watermelon vines caused by SqVYV has been observed in watermelon production areas of Florida since 2002. Symptomatic watermelon plants collected from commercial fields

522

C.G. Webster, S. Adkins / Virus Research 163 (2012) 520–527

Table 1 Squash vein yellowing virus isolates analyzed in this study. Isolatea

Habitat description

Symptomsb

P1ac

P1bc

CPc

Mo2005aHi Mo2005bHi Mo2007aCo Mo2007bCo Mo2007aHe Mo2007bHe Ca2008Ma Pu2009aCo SM2008aHe SM2008bHe SM2008ciHe SM2008ciiHe SM2008diHe SM2008diiHe SM2008eiHe SM2008eiiHe SM2008fHe SM2008gHe SM2008hiHe SM2008hiiHe SM2008hiiiHe Sq2003Hi WM2004aiMa WM2004aiiMa WM2004biMa WM2004biiMa WM2005aCo WM2005bCo WM2005aiHi WM2005aiiHi WM2005aiiiHi WM2005bHi WM2005cHi WM2005diHi WM2005diiHi WM2005eHi WM2007He WM2007Ma WM2007GA WM2007IN WM2008 WM2008aCo WM2008bCo WM2008aHa WM2008bHa WM2008cHa WM2008aHe WM2008aMa WM2008bMa WM2008cMa WM2009aCo WM2009bCo

Edge watermelon field Edge watermelon field Edge watermelon field Experimental watermelon field Edge watermelon field Edge watermelon field Commercial cantaloupe field Experimental cucurbit field Edge watermelon field Edge watermelon field Edge watermelon field Edge watermelon field Edge watermelon field Edge watermelon field Edge watermelon field Edge watermelon field Edge watermelon field Edge watermelon field Edge watermelon field Edge watermelon field Edge watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field Commercial watermelon field

NS NS NS NS NS NS VY VY NS NS NS NS NS NS NS NS NS NS NS NS NS VY SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN SWN

JF894251 JF894252 –e –e –e JF894253 JF894254 JF894255 JF894256 JF894257 JF897996 – JF894258 – JF894259 – JF894260 JF894261 JF894262 – – EU259611 –e –e –e –e JF894263 JF894264 JF894265 – – JF894266 –e JF894267 – JF894268 JF894269 –e –e JF894270 –e –e –e –e –e –e JF894273 JF894271 –e JF894272 JF894274 JF894275

JF897923 JF897924 –e –e JF897925 JF897926 JF897927 JF897928 JF897929 JF897930 JF897996 JF897931 JF897932 JF897933 JF897934 JF897935 JF897936 JF897937 JF897938 – – EU259611 –e –e –e –e –e JF897939 JF897940 JF897941 JF897942 JF897943 –e JF897944 – –e JF897945 –e –e JF897946 –e –e –e –e –e –e JF897947 JF897948 –e JF897949 JF897950 JF897951

JF897952d JF897953 JF897954 JF897955 –e JF897956 JF897957 JF897958 JF897959 JF897960 JF897996 – JF897961 – JF897962 – JF897963 JF897964 JF897965 JF897966 JF897967 EU259611 JF897968 JF897969 JF897970 JF897971 JF897972 JF897973 JF897974 – – JF897975 JF897976 JF897977 JF897978 JF897979 JF897980 JF897981 JF897982 JF897983 JF897984 JF897985 JF897986 JF897987 JF897988 JF897989 JF897990 JF897991 JF897992 JF897993 JF897994 JF897995

a Isolate named based on host, year and location of collection. Host species: (Mo) Balsam apple, Momordica charantia; (WM) watermelon, Citrullus lanatus; (SM) smellmelon, Cucumis melo var. dudaim; (Ca) cantaloupe, Cucumis melo; (Pu) pumpkin, Cucurbita pepo; (Sq) squash, Cucurbita pepo. Locations: Co, Collier County; Ha, Hardee County; He, Hendry County; Hi, Hillsborough County; and Ma, Manatee County, Florida; GA, Georgia; IN, Indiana. Multiple isolates collected from the same host and location were given an extra letter in the name to distinguish them (e.g. SM2008cHe, to denote the isolate from the third smellmelon plant collected at one location in Hendry County in 2008), and where multiple sequence variants were found within an isolate from a single plant a lowercase Roman numeral was added to the name to distinguish each variant (e.g. SM2008diHe, to denote the first sequence variant of the isolate from the fourth smellmelon plant collected at one location in Hendry County in 2008). Each sequence variant was given a unique GenBank accession number. b Symptoms observed in field-collected isolate source plants: NS, no symptoms; VY, vein yellowing; SWN, systemic wilt and necrosis. c GenBank accession numbers. d Only 540 bp of CP amplified using nested PCR (Adkins et al., 2008). e In replicate reactions an amplicon of the expected size could not be produced.

for this study all displayed the systemic wilt and necrosis that has been previously observed (data not shown). No SqVYV was isolated from symptomless watermelon plants in these same fields. However, cucurbit weeds smellmelon and Balsam apple collected adjacent to fields of declining watermelon showed no symptoms of infection as previously reported (Adkins et al., 2008). Other cucurbit crops (cantaloupe, pumpkin and squash) displayed only foliar vein yellowing when infected with SqVYV, consistent with previous results (Table 1; Adkins et al., 2007).

Squash plants mechanically inoculated with isolates originally collected from watermelon in southwest (WM2007He) and westcentral (Sq2003Hi, WM2007Ma) Florida developed typical virus symptoms including vein yellowing on upper, non-inoculated leaves. Watermelon plants inoculated with these same isolates developed necrosis along the vines and petioles, and died within 21 days. No differences in the progression or severity of symptoms induced by virus isolates collected from cultivated or weedy cucurbits in the different regions surveyed were observed.

C.G. Webster, S. Adkins / Virus Research 163 (2012) 520–527

523

Table 2 Comparison of smellmelon and type Squash vein yellowing virus isolates. Regiona

nt

nt identityb

aa

aa identityb

Complete genome 5 UTRc P1a P1b P3 PIPO 6K1 CI 6K2 NIa-Vpg N1a-Pro NIb CP 3 UTR

9836 118 1611 963 879 240 156 1887 159 576 699 1512 1074 202

99.4% (56) 100% (0) 99.8% (4) 98.9% (11) 99.2% (6) 99.8% (1) 99.4% (1) 99.4% (12) 100% (0) 98.6% (7) 99.6% (3) 99.3% (11) 99.9% (1) 100% (0)

3172 – 537 321 293 80 52 629 53 192 233 504 358 –

99.3% (21) – 99.4% (4) 97.8% (7) 99.0% (3) 100% (1) 100% (0) 99.7% (2) 100% (0) 99.5% (1) 100% (0) 99.4% (3) 99.7% (1) –

a Genome regions of SqVYV. 5 and 3 untranslated regions (UTR); P1a and P1b, P1 proteinases; P3, third protein; PIPO, pretty interesting potyvirus protein; 6K1 and 6K2, first and second 6 kDa proteins; CI, cytoplasmic inclusion protein; NIa-VPg/-Pro, nuclear inclusion a protein viral genome linked domain (VPg) and proteinase (Pro); NIb, nuclear inclusion b protein; CP, coat protein. b Percent nucleotide (nt) and amino acid (aa) identity of smellmelon isolate SM2008cHe to type isolate Sq2003Hi (GenBank accession no. EU259611) with the number of polymorphisms shown in parentheses. c The 20 most 5 bases of smellmelon isolate came from the FLICv primer used to synthesize cDNA.

3.2. Full length sequence of isolate SM2008cHe The complete nucleotide sequence of the coding region of smellmelon isolate SM2008cHe was found to be almost identical to the type SqVYV isolate (Sq2003Hi). Nucleotide and aa identities with the type SqVYV isolate were >99% across the whole genome and each mature protein, with the exception of the P1b (98.9% nt and 97.8% aa identity) and NIa-Pro (98.6% nt and 99.5% aa identity; Table 2). The organization of the genome was identical to the type SqVYV isolate with ten mature proteins (P1a, P1b, P3, 6K1, CI, 6K2, NIa-VPg, NIa-Pro, NIb and CP) predicted from nine hepta-peptide polyprotein processing sites, the sequences of which were identical to the type SqVYV isolate, as were other conserved motifs previously identified (Li et al., 2008). In addition, a conserved G2 A7 motif was identified in the P3 coding region, which through ribosomal frameshifting or transcriptional slippage is predicted to produce the PIPO fusion protein (Chung et al., 2008), of which the deduced 80 aa PIPO region shared 100% identity with the corresponding region of the type SqVYV isolate genome. 3.3. Phylogenetic analysis of the SqVYV genome P1a, P1b and/or CP genes were sequenced for 41 isolates of SqVYV collected from diverse locations and hosts over a period of seven years (Table 1) to more fully examine the genetic diversity of SqVYV isolates currently present in the U.S. SqVYV was identified by CP gene sequence in a watermelon plant from Georgia (isolate WM2007GA), marking the first report of the virus in this state. In addition, some isolates were identified which contained sequence variants among the ten clones sequenced for each isolate. Such polymorphisms were only seen in the PIb and CP genes, with five sequence variants identified in each region. All variants of such isolates were included in further analyses (Fig. 1). Generally, only one or two polymorphisms were present and were often synonymous. However, isolate WM2005aHi, which contained three sequence variants (WM2005aiHi, WM2005aiiHi and WM2005aiiiHi), showed 24 polymorphisms in the P1b region. Interestingly, each sequence variant had different polymorphisms but variant ii had the most (20) unique polymorphisms; however, the ten CP and ten P1a clones of WM2005aHi were identical in sequence. Nucleic acid sequencing gave useful sequence data for most isolates tested. However, in several instances we failed to amplify some of the regions studied including P1a and/or P1b of 14 isolates (see Table 1 for details). This was despite multiple adjustments of reaction

conditions (e.g. reduced annealing temperatures, alternative primers) and multiple amplification attempts. Analysis of all SqVYV sequences showed the U.S. population consists of two genetically distinct groups with very low intragroup but greater inter-group diversity. This was evidenced by the long branch lengths separating the two groups but relatively short branch lengths within each group (Fig. 1). The level of genetic diversity between these groups was greatest in the P1a regions as seen by long branch lengths, and less in the CP and P1b regions. In fact, diversity in the P1b region was very low as seen from the relatively short branch lengths of the P1b phylogenetic tree (Fig. 1b; note the same scale is used in Fig. 1a–c). Isolate WM2005aHi was the only isolate from the phylogenetically distinct group for which all three regions could be amplified. As noted above, three sequence variants were observed in the P1b of this isolate, with variants i and iii showing the highest identity with the type isolate (Fig. 1b), whereas variant ii had the P1b region most different from the type isolate. The CP of five other isolates from Hardee, Manatee or Hillsborough Counties also had a genetically distinct CP. Isolates of SqVYV collected from different locations, years and hosts did not group in any significant way by these parameters in the phylogenetic trees. For instance, isolates Sq2003Hi, Pu2009aCo and WM2007IN were almost identical in sequence across all three genome regions (3495 nt) examined despite being collected up to seven years apart from different hosts (squash, pumpkin and watermelon) and from distinct geographic regions [Collier County (southwest Florida), Hillsborough County (west-central Florida) and Indiana]. The isolates collected from the two cucurbit weeds known to be SqVYV hosts were nearly identical in sequence to those collected from watermelon and other cucurbit crops. There was no evidence of greater diversity in isolates collected from weeds. 3.4. Selection pressure in the SqVYV genome The genetic diversity, and strength and direction of selection pressure on the SqVYV genome was estimated for each region, as well as for a concatamer of all regions sequenced (Table 3). Negative (or purifying) selection was estimated to be occurring across each region of the genome by both methods comparing the global rate of changes (ω and dN − dS ) as well as at the level of individual codons (Table 3). The strength of the negative selection was greatest in the CP coding region (ω = 0.0258, dN − dS 0.0563, 27 negatively selected codons) compared to the P1b (ω = 0.3490, dN − dS 0.4185, 2 negatively selected codons) and P1a (ω = 0.1717, dN − dS 0.1967, 5

524

C.G. Webster, S. Adkins / Virus Research 163 (2012) 520–527

Fig. 1. Unrooted neighbor-joining phylogenetic trees of the Squash vein yellowing virus (a) P1a, (b) P1b or (c) coat protein gene sequences with primer sequences removed prior to analysis. Details of isolate host, location and GenBank accession are given in Table 1. Trees were constructed in MEGA 4.1. Bootstrap percentages (1000 replicates) are given at the branch nodes when greater than 70% with genetic distances included on the branches with a scale of 0.01 substitutions per site. Significance of groupings was confirmed by comparison to maximum parsimony based trees (data not shown). The type isolate (Sq2003Hi) is shaded in gray, and recombinant isolates (WM2005aHi and WM2005bHi, and sequence variants) are underlined.

negatively selected codons) coding regions, which were only under moderately strong negative selection pressure. When the isolates of only the type group of SqVYV isolates (Fig. 1) were compared, the strength of negative selection was greatly reduced [e.g. ω of the CP was 0.2871 instead of 0.0258 (Table 2)]. However, all methods of analysis still indicated negative selection. The reason for this change in the strength of selection pressure was due to the greatly reduced rate of synonymous

changes (0.1549 vs. 0.0195) between the two groups. This decrease was also seen for both the P1a and concatenated segments (Table 2). 3.5. Recombination within the SqVYV genome Grouping of isolates in the phylogenetic trees was inconsistent between genomic regions for two watermelon isolates (WM2005aHi and WM2005bHi) suggesting recombination had

C.G. Webster, S. Adkins / Virus Research 163 (2012) 520–527

525

Table 3 Genetic diversity and selection pressures in different regions of Squash vein yellowing virus genome. Character

Genome region P1a

P1b

CP

Concatamera

ntb Number of isolatesc

1566 27 (0)

912 26 (5)

1017 41 (5)

3495 25 (8)

nt diversityd

0.0280 ± 0.0015 0.0087 ± 0.0013

0.0089 ± 0.0015 0.0076 ± 0.0015

0.0269 ± 0.0022 0.0058 ± 0.0009

0.0335 ± 0.0014 0.0093 ± 0.0008

aa diversityd

0.0358 ± 0.0036 0.0170 ± 0.0032

0.0144 ± 0.0037 0.0125 ± 0.0037

0.0094 ± 0.0022 0.0056 ± 0.0016

0.0335 ± 0.0024 0.0129 ± 0.0019

dN e

0.0199 ± 0.0019 0.0013 ± 0.0014

0.0067 ± 0.0017 0.0053 ± 0.0016

0.0040 ± 0.0010 0.0024 ± 0.0007

0.0182 ± 0.0014 0.0055 ± 0.0008

dS e

0.1159 ± 0.0130 0.0142 ± 0.0033

0.0192 ± 0.0050 0.0177 ± 0.0049

0.1549 ± 0.0192 0.0195 ± 0.0035

0.1287 ± 0.0086 0.0244 ± 0.0026

ωe

0.1717 0.5130

0.3490 0.7062

0.0258 0.2871

0.1414 0.2254

dN − dS f +ve codonsg −ve codonsg

0.1967 0 (0.0%) 5 (0.958%)

0.4185 0 (0.0%) 2 (0.658%)

0.0563 0 (0.0%) 27 (7.96%)

0.1896 2 (0.172%) 204 (17.5%)

a

P1a, P1b and CP sequences were concatenated for isolates from which all three regions were amplified. Primers used to amplify PIa, P1b and CP anneal within coding region and primer sequences were removed prior to analysis. c The number of virus isolates analyzed with the number of additional sequence variants found within a single host plant indicated in parentheses. d The average nucleotide (nt) and amino acid (aa) diversity of SqVYV isolates as calculated by the p-value method with standard errors shown from 1000 bootstrap replications. Values in italics only include the type group of SqVYV isolates. e Frequency of non-synonymous (dN ), synonymous (dS ) changes along with the ratio (ω) as determined by the method of Li et al. (1985) in MEGA 4.1. Values in italics only include the diversity value calculated from the type group of SqVYV isolates. f Estimate of dN − dS by single likelihood ancestor counting (SLAC) method. g The number of positively and negatively selected codons estimated by the Datamonkey online software package with percentages shown in parentheses. Only codons with evidence of selection by two or more models (SLAC, random effects likelihood and fixed effects likelihood; Kosakovsky Pond and Frost, 2005b) are included. b

occurred (Fig. 1). For instance, isolate WM2005aHi showed P1a and CP regions distinct from the type isolate (Sq2003Hi), whereas the P1b region, that contained three sequence variants, was either highly similar to (variant i and iii) or distinct from (variant ii) the type isolate. Similarly, WM2005bHi had P1a and P1b regions highly similar to the type isolate, whereas the CP was distinct, and highly similar to WM2005aHi. Sliding window analysis was used to further examine recombination in concatenated genomes by comparing the type isolate (Sq2003Hi) to isolates from smellmelon (SM2008ciHe) and

watermelon (WM2005aHi and WM2005bHi). This analysis supported the occurrence of recombination in both watermelon isolates (Fig. 2). For example, all three sequence variants of WM2005aHi showed >98% identity to Sq2003Hi across the P1b region but only 15–90% identity across the P1a and CP regions (P1a and CP sequences were identical for all three variants). Included in this region of low identity in P1a was a 6 nt (2 aa) insertion in WM2005aHi compared to Sq2003He. The potential breakpoints of the recombination event were located at ∼1800 and ∼2300 ± 100 nt (Fig. 2) given the 200 nt window size used in the

Fig. 2. Sliding window analysis of a 3495 nt concatamer of the Squash vein yellowing virus (SqVYV) P1a, P1b and CP coding regions. Comparisons of isolates SM2008cHe, WM2005aHi and WM2005bHi were made to the type SqVYV isolate (Sq2003Hi). The second P1b sequence variant was used for WM2005aHi. Locations of the coding regions are indicated at the top of the graph. SimPlot (Version 3.5) was used with a 200 nt window and 20 nt step size.

526

C.G. Webster, S. Adkins / Virus Research 163 (2012) 520–527

analysis. WM2005bHi showed >98% identity to Sq2003Hi across P1a and P1b regions but considerably less across the CP region. In fact, WM2005bHi and WM2005aHi (the most divergent isolate analyzed in this study) were identical across the CP region. This recombination event was also located at ∼2300 ± 100 nt (Fig. 2) and likely identical in location to the recombination detected in WM2005aHi. Predictions from the GARD program further supported these two recombination events.

4. Discussion Our investigation into the diversity of SqVYV has demonstrated that U.S. isolates from cultivated and weedy cucurbits form two distinct genetic groups with low intra-group diversity (Table 3; Fig. 1). Symptoms induced in watermelon and squash by members of both groups were indistinguishable, with watermelons developing necrosis along the vine leading to plant death and squash developing vein yellowing, both of which are identical to symptoms induced by the type isolate (Sq2003He; Adkins et al., 2007). Throughout the emergence of SqVYV in Florida, we have found no mild isolates or asymptomatic infections in watermelon. The set of isolates sequenced in this study represents the major locations, hosts and time periods of SqVYV infections currently known in the U.S., including the first report of SqVYV in Georgia. Diversity varied across the genome with the P1a region showing the highest diversity between groups (>0.17 substitutions per site, Fig. 1) but low diversity within the group containing the type isolate (Sq2003Hi). This suggests that other, as of yet undetected, more diverse isolates may occur. However, the CP primers previously described for detecting SqVYV (Adkins et al., 2008) remain effective for all currently known isolates. Recombination was strongly suggested by our analysis to have occurred between isolates of the U.S. population with two watermelon isolates (WM2005aHi and WM2005bHi) varying in genetic relatedness across the genomes. This is the first report of recombination in SqVYV although it appears to occur in most ipomoviruses including CBSV, UCBSV and SPMMV (Tugume et al., 2010; Mbanzibwa et al., 2011b). The overall low level of diversity observed in the P1b (Fig. 1b) across all isolates examined in this study also seems contradictory to the moderate negative selection pressure in this region (Table 3). However, a recombination in the genome of the most diverse isolate (WM2005aHi) explains both of these pieces of evidence. The low level of genetic diversity we observed within the U.S. population of SqVYV, including within weed reservoirs such as smellmelon and Balsam apple, can be explained either by negative (purifying) selection, or a genetic bottleneck event from recent founder effects (García-Arenal et al., 2001). Such founder effects could be caused by introduction of the virus to a new host or to a new location. The low diversity observed in SqVYV isolates from weed hosts makes introduction to a new location the more likely explanation. The moderate negative selection observed in this study was insufficient to explain the observed low genetic diversity. Stronger negative selection was observed in the CP (ω = 0.0258, Table 2), which is typical for arthropod-vectored viruses (Chare and Holmes, 2004). Collectively, these data suggest that the U.S. population of SqVYV has arisen from a recent introduction of at least two distinct lineages of SqVYV represented by isolates Sq2003Hi and WM2005aHi. This information fits well with observations of the SqVYV epidemic in Florida watermelons, where the first symptoms of systemic wilt and necrosis in commercial watermelons were observed in 2002. Once introduced, SqVYV rapidly became established in cucurbit growing regions of the state and beyond, which would explain the genetic and symptom similarity of isolates collected from a wide range of locations in

Florida and one location each in Indiana and Georgia over the seven years of sampling (Fig. 1). Low genetic diversity has been reported in both ipomoviruscucurbit crop systems, SqVYV in the U.S. (this report) and CVYV in Spain (Janssen et al., 2007), which share a common genome organization. However, this is not the case for the three other members of the genus. Greater genetic diversity and importantly similar negative selection pressure has been observed in CBSV and UCBSV isolates from cassava (Mbanzibwa et al., 2011b) and SPMMV isolates from sweet potato (Mukasa et al., 2003; Tairo et al., 2005; Tugume et al., 2010). Therefore, the low genetic diversity of SqVYV and CVYV may be an exception for ipomoviruses, and likely reflects the effects of a genetic bottleneck from recent introductions to the U.S. and Spain, respectively. Additional sampling and analysis are required to determine whether the limited diversity we report here for SqVYV and previously reported for CVYV (Janssen et al., 2007) extends to isolates of these viruses collected from a wider geographic area. Higher diversity may be expected for both SqVYV and CVYV at their center of origin as this is typical for viruses that coevolved with wild plants over a long period of time, including other members of the family Potyviridae (e.g. Spetz et al., 2003; Webster et al., 2007; Coutts et al., 2011).

Acknowledgements We thank Carrie Vanderspool, Bridget Burns, Cassandra Bradley, Nicole Miller and Jeff Smith for their excellent technical assistance; Gary Vallad, Jim Mertely, Phyllis Gilreath and many growers and scouts for providing cucurbit samples and access to commercial fields; and Olufemi Alabi and Rayapati Naidu for their helpful advice on phylogenetic analysis.

References Abraham, A., Menzel, W., Vetten, H.J., Winter, S., 2011. Analysis of the tomato mild mottle virus genome indicates that it is the most divergent member of the genus Ipomovirus (family Potyviridae). Arch. Virol., doi:10.1007/s00705-011-1167-9. Adkins, S., Webb, S.E., Achor, D., Roberts, P.D., Baker, C.A., 2007. Identification and characterization of a novel whitefly-transmitted member of the family Potyviridae isolated from cucurbits in Florida. Phytopathology 97, 145–154. Adkins, S., Webb, S.E., Baker, C.A., Kousik, C.S., 2008. Squash vein yellowing virus detection using nested polymerase chain reaction demonstrates that the cucurbit weed Momordica charantia is a reservoir host. Plant Dis. 92, 1119–1123. Adkins, S., Webb, S.E., Roberts, P.D., Kousik, C.S., Stansly, P.A., Bruton, B.D., Achor, D., Muchovej, R.M., Baker, C.A., 2010. A review of Ipomoviruses and watermelon decline in Florida. In: Stansley, P.A., Naranjo, S.E. (Eds.), Bemisia: Bionomics and Management of a Global Pest. Springer, New York, pp. 333–337. Adkins, S., Webster, C.G., Baker, C.A., Weaver, R., Rosskopf, E.N., Turechek, W.W., 2009. Detection of three whitefly-transmitted viruses infecting the cucurbit weed Cucumis melo var. dudaim in Florida. Plant Health Prog., doi:10.1094/PHP2009-1118-01-BR. Adkins, S., Webster, C.G., Kousik, C.S., Webb, S.E., Roberts, P.D., Stansly, P.A., Turechek, W.W., 2011. Ecology and management of whitefly-transmitted viruses of vegetable crops in Florida. Virus Res. 159, 110–114. Alicai, T., Omongo, C.A., Maruthi, M.N., Hillocks, R.J., Baguma, Y., Kawuki, R., Bua, A., Otim-Nape, G.W., Colvin, J., 2007. Re-emergence of cassava brown streak disease in Uganda. Plant Dis. 91, 24–29. Anandalakshmi, R., Pruss, G.J., Ge, X., Marathe, R., Mallory, A.C., Smith, T.H., Vance, V.B., 1998. A viral suppressor of gene silencing in plants. Proc. Natl. Acad. Sci. U.S.A. 95, 13079–13084. Chare, E.R., Holmes, E.C., 2004. Selection pressures in the capsid genes of plant RNA viruses reflect mode of transmission. J. Gen. Virol. 85, 3149–3157. Chellemi, D.O., Webster, C.G., Baker, C.A., Annamalai, M., Achor, D., Adkins, S., 2011. Widespread occurrence and low genetic diversity of Colombian Datura virus in Brugmansia suggest an anthropogenic role in virus selection and spread. Plant Dis. 95, 755–761. Chung, B.Y.-W., Miller, W.A., Atkins, J.F., Firth, A.E., 2008. An overlapping essential gene in the Potyviridae. Proc. Natl. Acad. Sci. U.S.A. 105, 5897–5902. Cohen, S., Nitzany, F.E., 1960. A whitefly transmitted virus of cucurbits in Israel. Phytopathol. Mediterr. 1, 44–46. Coutts, B.A., Kehoe, M.A., Webster, C.G., Wylie, S.J., Jones, R.A.C., 2011. Indigenous and introduced potyviruses of legumes and Passiflora spp. from Australia: biological properties and comparison of coat protein nucleotide sequences. Arch. Virol. 156, 1757–1774.

C.G. Webster, S. Adkins / Virus Research 163 (2012) 520–527 Cuadrado, I., Janssen, D., Velasco, L., Ruiz, L., Segundo, E., 2001. First report of Cucumber vein yellowing virus in Spain. Plant Pathol. 85, 336. Desbiez, C., Delécolle, B., Wipf-Scheibel, C., Lecoq, H., 2001. Le Cucumber vein yellowing virus, virus transmis par l’aleurode Bemisia tabaci, est un member des Ipomovirus, Potyviridae. 8émes Recontres de Virologie Végétale. CNRS, Aussois (FR). Egel, D.S., Adkins, S., 2007. Squash vein yellowing virus identified in watermelon (Citrullus lanatus) in Indiana. Plant Dis. 91, 1056. Fauquet, C.M., Mayo, M.A., Maniloff, J., Desselberger, U., Ball, L.A., 2005. Virus taxonomy: classification and nomenclature of viruses. In: Report of the International Committee on Taxonomy of Viruses, 8th edn. Elsevier, San Diego, p. 1259. García-Arenal, F., Fraile, A., Malpica, J.M., 2001. Variability and genetic structure of plant virus populations. Annu. Rev. Phytopathol. 39, 157–186. Gibbs, A., Mackenzie, A., 1997. A primer pair for amplifying part of the genome of all potyvirids by RT-PCR. J. Virol. Methods 63, 9–16. Giner, A., Lakatos, L., García-Chapa, M., López-Moya, J.J., Burgyán, J., 2010. Viral protein inhibits RISC activity by Argonaute binding through conserved WG/GW motifs. PLoS Pathog. 6, e1000996, doi:10.1371/journal.ppat.10009960. Hillocks, R.J., Jennings, D.K., 2003. Cassava brown streak disease: a review of present knowledge and research needs. Int. J. Pest Manag. 49, 225–234. Hiskias, Y., Lesemann, D.-E., Vetten, H.J., 2001. Biological characteristics of tomato mild mottle virus: a potyvirus isolated from tomato and thorn apple in Ethiopia. Afr. Crop Sci. J. 9, 517–525. Hollings, M., Stone, O.M., Bock, K.R., 1976. Purification and properties of sweet potato mild mottle, a whitefly borne virus from sweet potato (Ipomoea batatas) in East Africa. Ann. Appl. Biol. 82, 511–528. Janssen, D., Martín, G., Velasco, L., Gómez, P., Segundo, E., Ruiz, L., Cuadrado, I.M., 2005. Absence of a coding region for the helper component-proteinase in the genome of cucumber vein yellowing virus, a whitefly-transmitted member of the Potyviridae. Arch. Virol. 150, 1439–1447. Janssen, D., Ruiz, L., Velasco, L., Segunda, E., Cuadrado, I.M., 2002. Noncucurbitaceous weed species shown to be natural hosts of Cucumber vein yellowing virus in south-eastern Spain. Plant Pathol. 51, 797. Janssen, D., Velasco, L., Martín, G., Segundo, E., Cuadrado, I.M., 2007. Low genetic diversity among Cucumber vein yellowing virus isolates from Spain. Virus Genes 34, 367–371. Kosakovsky Pond, S.L., Frost, S.D.W., 2005a. Datamonkey: rapid detection of selective pressures on individual sites of codon alignments. Bioinformatics 21, 2531–2533. Kosakovsky Pond, S.L., Frost, S.D.W., 2005b. Not so different after all: a comparison of methods for detecting amino acid sites under selection. Mol. Biol. Evol. 22, 1208–1222. Lecoq, H., Desbiez, C., Delécolle, B., Cohen, S., Mansour, A., 2000. Cytological and molecular evidence that the whitefly-transmitted Cucumber vein yellowing virus is a tentative member of the family Potyviridae. J. Gen. Virol. 81, 2289–2293. Legg, J.P., Jeremiah, S.C., Obiero, H.M., Maruthi, M.N., Ndyetabula, I., Okao-Okuja, G., Bouwmeester, H., Bigirimana, S., Tata-Hangy, W., Gashaka, G., Mkamilo, G., Alicai, T., Lava Kumar, P., 2011. Comparing the regional epidemiology of the cassava mosaic and cassava brown streak virus pandemics in Africa. Virus Res. 159, 161–170. Li, W., Hilf, M.E., Webb, S.E., Baker, C.A., Adkins, S., 2008. Presence of P1b and absence of HC-Pro in Squash vein yellowing virus suggests a general feature of the genus Ipomovirus in the family Potyviridae. Virus Res. 135, 213–219. Li, W.-H., Wu, C.-I., Luo, C.-C., 1985. A new method for estimating synonymous and nonsynonymous rates of nucleotide substitution considering the relative likelihood of nucleotide and codon changes. Mol. Biol. Evol. 2, 150–174. Lole, K.S., Bollinger, R.C., Paranjape, R.S., Gadkari, D., Kulkarni, S.S., Novak, N.G., Ingersoll, R., Sheppard, H.W., Ray, S.C., 1999. Full-length human immunodeficiency virus type 1 genomes from subtype C-infected seroconverters in India, with evidence of intersubtype recombination. J. Virol. 73, 152–160. Louro, D., Quinot, A., Neto, E., Fernandes, J.E., Marian, D., Vecchiati, D., Caciagli, P., Vaira, A.M., 2004. Occurrence of Cucumber vein yellowing virus in cucurbitaceous species in southern Portugal. Plant Pathol. 53, 241. Maruthi, M.N., Hillocks, R.J., Mtunda, K., Raya, M.D., Muhanna, M., Kiozia, H., Rekha, A.R., Colvin, J., Thresh, J.M., 2005. Transmission of Cassava brown streak virus by Bemisia tabaci (Gennadius). J. Phytopathol. 153, 307–312.

527

Mbanzibwa, D.R., Tian, Y., Mukasa, S.B., Valkonen, J.P.T., 2009a. Cassava brown streak virus (Potyviridae) encodes a putative Maf/HAM1 pyrophosphatase implicated in reduction of mutations and P1 proteinase that suppresses RNA silencing but contains no HC-Pro. J. Virol. 83, 6934–6940. Mbanzibwa, D.R., Tian, Y., Tugume, A.K., Mukasa, S.B., Tairo, F., Kyamanywa, S., Kullaya, A., Valkonen, J.P.T., 2009b. Genetically distinct strains of Cassava brown streak virus in the Lake Victoria basin and the Indian Ocean coastal area of East Africa. Arch. Virol. 154, 353–359. Mbanzibwa, D.R., Tian, Y.P., Tugume, A.K., Mukasa, S.B., Tairo, F., Kyamanywa, S., Kullaya, A., Valkonen, J.P.T., 2011a. Simultaneous virus-specific detection of the two cassava brown streak-associated viruses by RT-PCR reveals wide distribution in East Africa, mixed infections, and infections in Manihot glaziovii. J. Virol. Methods 171, 394–400. Mbanzibwa, D.R., Tian, Y., Tugume, A.K., Patil, B.L., Yadav, J.S., Bagewadi, B., Abarshi, M.M., Alicai, T., Changadeya, W., Mkumbira, J., Muli, M.B., Mukasa, S.B., Tairo, F., Baguma, Y., Kyamanywa, S., Kullaya, A., Maruthi, M.N., Fauqeuet, C.M., Valkonen, V.P.T., 2011b. Evolution of cassava brown streak disease-associated viruses. J. Gen. Virol. 92, 974–987. Monger, W.A., Alicai, T., Ndunguru, J., Kinyua, Z.M., Potts, M., Reeder, R.H., Miano, D.W., Adams, I.P., Boonham, N., Glover, R.H., Smith, J., 2010. The complete genome sequence of the Tanzanian strain of Cassava brown streak virus and comparison with the Ugandan strain sequence. Arch. Virol. 155, 429–433. Monger, W.A., Spence, N.J., Foster, G.D., 2001. Molecular evidence that the aphidtransmitted Tomato mild mottle virus belongs to the Potyviridae but not the Potyvirus genus. Arch. Virol. 146, 2435–2441. Morris, J., Steel, E., Smith, P., Boonham, N., Spence, N., Barker, I., 2006. Host range studies for Tomato chlorosis virus, and Cucumber vein yellowing virus transmitted by Bemisia tabaci (Gennadius). Eur. J. Plant Pathol. 114, 265–273. Mukasa, S.B., Rubaihayo, P.R., Valkonen, J.P.T., 2003. Sequence variability within the 3 -proximal part of the Sweet potato mild mottle virus genome. Arch. Virol. 148, 487–496. Spetz, C., Taboada, A.M., Darwich, S., Ramsell, J., Salazar, L.F., Valkonen, J.P.T., 2003. Molecular resolution of a complex of potyviruses infecting solanaceous crops at the centre of origin in Peru. J. Gen. Virol. 84, 2565–2578. Tairo, F., Mukasa, S.B., Jones, R.A.C., Kullaya, A., Rubaihayo, P.R., Valkonen, J.P.T., 2005. Unraveling the genetic diversity of the three main viruses involved in sweet potato virus disease (SPVD), and its practical implications. Mol. Plant Pathol. 6, 199–211. Tamura, K., Dudley, J., Nei, M., Kumar, S., 2007. MEGA4: Molecular Evolutionary Genetics Analysis (MEGA) software version 4.0. Mol. Biol. Evol. 24, 1596–1599. Tugume, A.K., Mukasa, S.B., Kalkkinen, N., Valkonen, J.P.T., 2010. Recombination and selection pressure in the ipomovirus sweet potato mild mottle virus (Potyviridae) in wild species and cultivated sweetpotato in the centre of evolution in East Africa. J. Gen. Virol. 91, 1092–1108. Valli, A., Dujovny, G., García, J.A., 2008. Protease activity, self interaction, and small interfering RNA binding of the silencing suppressor P1b from Cucumber vein yellowing Ipomovirus. J. Virol. 82, 974–986. Valli, A., López-Moya, J.J., García, J.A., 2007. Recombination and gene duplication in the evolutionary diversification of P1 proteins in the family Potyviridae. J. Gen. Virol. 88, 1016–1028. Valli, A., Martín-Hernández, A.M., López-Moya, J.J., García, J.A., 2006. RNA silencing suppression by a second copy of the P1 serine protease of Cucumber vein yellowing Ipomovirus, a member of the family Potyviridae that lacks the cysteine protease HCPro. J. Virol. 80, 10055–10063. Walkey, D.G.A., Spence, N.J., Clay, C.M., Miller, A., 1994. A Potyvirus isolated from solanaceous hosts. Plant Pathol. 43, 931–937. Webster, C.G., Coutts, B.A., Jones, R.A.C., Jones, M.G.K., Wylie, S.J., 2007. Virus impact at the interface of an ancient ecosystem and a recent agroecosystem: studies on three legume-infecting potyviruses in the southwest Australian floristic region. Plant Pathol. 56, 729–742. Winter, S., Koerbler, M., Stein, B., Pietruszka, A., Paape, M., Butgereitt, A., 2010. Analysis of cassava brown streak virus reveals the presence of distinct species causing cassava brown streak disease in East Africa. J. Gen. Virol. 91, 1365–1372. Yakoubi, S., Desbiez, C., Fakhfakh, H., Wipf-Scheibel, C., Marrakchi, M., Lecoq, H., 2007. Occurrence of Cucurbit yellow stunting disorder virus and Cucumber vein yellowing virus in Tunisia. J. Plant Pathol. 89, 417–420.