HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS		QUICK SEARCH:   [advanced] Author: Keyword(s): Year:  Vol:  Page:

This Article Abstract Full Text (PDF) Supplementary table Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Hall, G. Search for Related Content PubMed PubMed Citation Articles by Hall, G. Agricola Articles by Hall, G.

Selective constraint and genetic differentiation in geographically distant barley yellow dwarf virus populations

Cornell University, Department of Ecology and Evolutionary Biology, Corson Hall, Ithaca, NY 14853, USA

Correspondence
Gerod Hall
gsh8{at}cornell.edu

	ABSTRACT

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Numerous studies have documented molecular variability in plant virus populations, but few have assessed the relative contribution of natural selection and genetic drift in generating the observed pattern of diversity. To this end, gene function, environment and phylogenetic history were examined to observe the effect on genetic diversity and population structure of the PAV and PAS species of Barley yellow dwarf virus (family Luteoviridae). Three functional classes of gene were analysed: transcription-related (RdRp), structural (CP) and movement-related (MP). The results indicate that there were no inherent differences, in terms of total diversity or diversity at synonymous or non-synonymous nucleotide sites, between functional classes of genes or populations. Rather, selective constraints on a gene may be more or less relaxed depending on its function and the phylogenetic history of the population sampled. The CP of the PAS species, but not the PAV species, was differentiated genetically between regions. This is probably due to genetic drift, as there was no evidence that any gene deviated from a neutral model of evolution or is under positive selection. In general, the MP was under considerably less functional constraint than structural or replication-related proteins and four positively selected codon sites were identified. Mutations at these sites differentiate species and geographical subpopulations, so presumably they have aided the virus in adaptation to its host environment and contributed to intra- and interspecies diversification.

A table showing the geographical region of origin and GenBank accession numbers of virus isolates analysed in this study is available as supplementary material in JGV Online.

	INTRODUCTION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Plant viruses may affect the fitness of their hosts by reducing host survivorship, fecundity or competitive ability relative to uninfected individuals in the population. Virus strains may differ markedly in the severity of symptoms that they induce (Anderson et al., 1991; Bencharki et al., 1999) and their ability to infect a given host (Moury et al., 2001; Sacristán et al., 2005). Variation among strains in virulence, infectivity and transmission may affect patterns of disease spread and, thereby, host population dynamics in natural systems (Raybould et al., 1999) or crop yield in agricultural systems. The behaviour and constraints of a virus upon host infection surely have a genetic underpinning. Thus, it is necessary to acquire knowledge of genetic diversity in pathogen populations to understand better the role that they play in ecological processes and as impediments to agricultural production. This study investigates qualitative (genetic differentiation) and quantitative (genetic variation) differences in genomic content among geographically distant populations of two species of Barley yellow dwarf virus (BYDV). This is a first step in understanding what evolutionary forces are acting on the viral genome.

Barley yellow dwarf (BYD) disease is caused by members of the family Luteoviridae in the genera Luteovirus and Polerovirus. Each virus species has a distinct aphid-transmission phenotype and the acronym for the species is derived from this specificity. Species in the family Luteoviridae commonly isolated from grain crops include GAV (Luteovirus), which is transmitted most efficiently by Schizaphis graminum and Sitobion avenae (Wang et al., 2001), MAV (Luteovirus), transmitted most efficiently by S. avenae (formerly Macrosiphum avenae), PAV (Luteovirus), transmitted most efficiently by Rhopalosiphum padi and S. avenae, SGV (unassigned to a genus within the family), transmitted most efficiently by S. graminum, and RPV (Polerovirus) transmitted most efficiently by R. padi (Rochow, 1969; Rochow & Muller, 1971). BYD disease has significant impacts in agricultural and natural plant communities. It is the most economically damaging viral disease of grain crops worldwide (Lister & Ranieri, 1995) and in grasslands it may contribute to shifts in community composition due to asymmetrical fitness effects on exotic and native grass species (Malmstrom et al., 2005a, b). Among the species listed above, PAV is the most widely distributed and economically important. To reflect significant variation in coat protein (CP) sequence among isolates, PAV has been divided into two species, PAV and PAS (Mayo, 2002). Sympatric populations of both species have been identified in Morocco (Bencharki et al., 1999), New York state (Chay et al., 1996a) and France (Mastari et al., 1998). The BYDV genome is composed of a single-stranded, positive-sense RNA with six open reading frames (ORFs) in total (Miller et al., 2002). Three viral genes were analysed in this study: ORFs 2, 3 and 4. ORF2 encodes the viral RNA-dependent RNA polymerase (RdRp) and is responsible for the replication of all viral RNAs (Koev et al., 2002). ORF3 encodes the major component of the CP, which is required for virion assembly (Mohan et al., 1995) and is thereby a prerequisite for aphid transmission (Gildow, 1987, 1993) and systemic plant infection (Filichkin et al., 1994). ORF4 encodes the movement protein (MP), which is required for the virus to spread systemically in the host (Chay et al., 1996b). ORF4 is embedded completely within ORF3, but is translated +1 bp out of the CP reading frame (Dinesh-Kumar & Miller, 1993).

There is currently a limited understanding of what factors influence genetic diversification and genome evolution in BYDV. This is due in part to a lack of studies that investigate variation in non-structural genes and explore how the genetic variation found in natural virus populations affects gene function. Through an analysis of diversity and selective constraint in multiple types of gene and across virus subpopulations, the present study was able to separate out the effects of gene function, environment and phylogenetic history in creating the pattern of diversity observed in the virus population. Thus, this study evaluates the role of selection in promoting or limiting genetic variability and provides a bridge between descriptive studies of population diversity and studies of protein structure and function (Moury, 2004).

	METHODS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Virus isolates.
The New York state virus isolates analysed in this study were obtained from field-caught R. padi or from field-collected corn (Zea mays), wheat (Triticum aestivum) and reed canary grass (Phalaris arundinacea) during the time period 1998–2003. Nucleotide sequences of members of the family Luteoviridae from other geographical regions were obtained from GenBank.

Immunocapture, RT-PCR and DNA sequencing.
Virions were isolated from plant extracts by first grinding tissue in a mixture of dry ice and PBS, followed by immunocapture in anti-BYDV-PAV antibody-coated microfuge tubes. In the same tubes, RT-PCR was used to amplify the CP nucleotide sequence (603 bp) or a portion of the RdRp gene (654 bp). RT-PCR of the CP gene was performed by using oligonucleotide primers designed by Bencharki et al. (1999). The Primer Select program (DNASTAR software package) along with full-length PAV sequences of isolates PAS-129 and PAV-Aus, available in GenBank, were used to design primers to amplify the RdRp gene. Reverse transcription was performed by using oligonucleotide 5'-TTAGGGTCAACTCCGAATGATTC-3' and PCR using the former and oligonucleotide 5'-GCGCCTAAGTGGGAACACG-3'. Single-step RT-PCR was carried out in a 50 µl reaction volume containing 25 mM Tris (pH 8.8), 10 mM KCl, 10 mM dithiothreitol, 10 mM (NH₄)₂SO₄, 2.5 mM MgSO₄, 1 µM each primer, 0.4 ng BSA ml^–1, 200 µM dNTPs, 30 units RNase inhibitor, 50 units SuperScript II RNase H^– RT and 5 units Taq polymerase (all enzymes were from Invitrogen). Thermocycling conditions were: one cycle of 45 min at 42 °C for reverse transcription, one cycle of 2 min at 95 °C for inactivation of reverse transcriptase, 36 cycles of 30 s at 94 °C, 30 s at 53 °C for the coat protein or 56 °C for the RdRp, 1 min at 72 °C, and a final extension of 10 min at 72 °C. PCR products were purified by using a Qiagen PCR clean-up kit, then submitted to the Cornell University Bioresource Center for direct sequencing. Sequencing was carried out in both the forward and reverse directions with the primers listed above.

Phylogenetic and nucleotide diversity analyses.
Sequences were aligned with the CLUSTAL W algorithm of the MEGALIGN program (DNASTAR software package). Alignments of the RdRp, CP and MP genes were unambiguous, but manual adjustments were made to the alignment of the complete-genome sequences. In PAUP version 4.0b10 (Swofford, 2000), maximum-likelihood (ML) trees for the RdRp, CP and MP genes (PAV and PAS isolates only) were constructed by using the 2 ST model of nucleotide substitution with base frequencies and the shape of the gamma distribution estimated from the data. The resultant tree topologies were used for ML analysis of codon substitution. To examine phylogenetic relationships among BYDV species explicitly, a second set of trees was constructed by using the procedure outlined above, except that MAV, GAV, SGV, SbDV (Soybean dwarf virus, family Luteoviridae) and RPV (used as outgroup) sequences were included in the analysis. The addition of these sequences did not alter the topology of the PAV/PAS region of the tree. Thus, excluding the additional sequences, the trees depicted in Fig. 1(a–c) are identical to those used for the PAML analysis. For each gene, robustness of the nodes of the phylogenetic tree was assessed by bootstrap percentages computed after 100 resamplings.

View larger version (27K): [in this window] [in a new window]   Fig. 1. ML trees of isolates of species in the family Luteoviridae, calculated by using (a) partial RdRp nucleotide sequences (ML score, 3126), (b) complete CP nucleotide sequences (ML score, 3101), (c) complete MP nucleotide sequences (ML score, 2099) and (d) complete-genome sequences (except SGV, ORFs 2, 3 and 4 only) (ML score, 44053). PAV isolates are shown in italics and PAS isolates in bold. All nodes with <70 % bootstrap support out of 100 replicates were collapsed to polytomies.

Tests of selective neutrality and population differentiation.
Estimation of population parameters (s, k, ) and Tajima's D (Tajima, 1989) and Fu & Li's F* and D* (Fu & Li, 1993) tests of selective neutrality were performed with DnaSP (Rozas et al., 2003). Tajima's D test compares nucleotide diversity with the number of segregating sites, which are expected to be equal if mutations are selectively neutral. Fu & Li's D* statistic is based on differences between the number of singletons (mutations appearing only once among the sequences) and the total number of mutations. Fu & Li's F* statistic is based on the differences between the number of singletons and the mean number of nucleotide differences between pairs of sequences. SNAP Workbench (Price & Carbone, 2003) was used to implement the programs SEQTOMATRIX (Hudson et al., 1992), which facilitated the conversion of aligned CP sequences to a distance matrix, and PERMTEST (Hudson et al., 1992), which tested for geographical subdivision in the virus populations. PERMTEST calculates Hudson's K_ST statistic of genetic differentiation. K_ST is equal to 1–K_S/K_T, where K_S is a weighted mean of K1 and K2 (mean number of differences between sequences in subpopulations 1 and 2, respectively) and K_T represents the mean number of differences between two sequences regardless of their subpopulation. The null hypothesis of no genetic differentiation will be rejected (P<0.05) when K_S is small and K_ST is close to 1.

Tests of positive selection and selective constraint.
The ratio of non-synonymous (dN) to synonymous (dS) nucleotide substitutions (=dN/dS) can be used to measure the degree of functional constraint for the maintenance of the encoded protein (Li, 1993). An ratio >1 indicates that non-synonymous substitutions have a higher probability of fixation than synonymous substitutions and, presumably, offer a fitness advantage to the protein (positive selection). An ratio close to 0 indicates that non-synonymous substitutions are less likely than synonymous substitutions and the gene will be conserved at the amino acid level (negative selection). An ratio equal to 1 indicates neutral evolution. Two approaches were taken to determine the mode and strength of selection acting on the viral genome. To test for positive selection in the RdRp, CP and MP, ML models of codon substitution were implemented in PAML version 3.14 (Yang, 1997). I then compared selective constraint on these genes among PAV and PAS isolates collected in New York state and Morocco.

The ML models employed allow to vary among codon sites, but remain constant across lineages in the phylogeny. The models implemented were M₀, which assumes =1 for all codon sites, M₁, which fixes sites as either invariant (₀=0) or neutral (₁=1), M2, which adds a third site class to the neutral model to allow for positive selection (₂>1), M7, which assumes a beta distribution for (0<<1), and M8, which adds another site class to the beta model to allow for positively selected sites (Yang et al., 2000). A likelihood-ratio test (LRT) was used to determine whether the positive-selection models (M2, M8) fit the data significantly better than the neutral models (M1, M7) (Yang et al., 2000). To construct the LRT statistic, twice the log-likelihood difference between the general model and the null model was compared with a ² test, with degrees of freedom equal to the difference in the number of parameters between the two models. Bayes empirical Bayes (BEB) analysis was used to infer to what class, conserved, neutral or positively selected, a codon site belongs (Yang et al., 2005). Codon sites with values >1 and where posterior probabilities summed to be >95 % were identified as potentially being under positive selection.

To assess the magnitude of the selective constraint acting on the RdRp, CP and MP genes, dS and dN values were generated from pairwise sequence comparisons by using the method of Nei & Gojobori (1986) implemented in PAML version 3.14. To avoid dividing by 0, two approaches were used: a constant was added to dS for one synonymous substitution, dN/(dS+constant), and the ratio dN/(dN+dS) was calculated (Mishmar et al., 2003). Friedman's test and a series of pre-planned comparisons were used to compare dN/(dS+constant) or dN/(dN+dS) between species and geographical regions. I first tested the null hypotheses that, within a given species, selective constraint does not differ between the RdRp and CP or MP genes and, in different species, selective constraint does not differ between homologous genes. I then tested the null hypothesis that selective constraint on the CP or MP gene does not vary across geographical regions or species.

	RESULTS

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Phylogenetic relationships among isolates
ML methods were used to determine the phylogenetic relationship among PAV and PAS isolates collected from a number of geographical regions. CP sequence identity between New York state isolates was similar to that described for populations sampled in other geographical regions (Bencharki et al., 1999; Bisnieks et al., 2004) (Table 1). For the RdRp gene, there is low within-species diversity, but between species, diversity is greater than that found in the CP gene (Table 1). The RdRp tree (Fig. 1a) places PAV and MAV in a monophyletic clade that excludes PAS, but the CP (Fig. 1b) and MP (Fig. 1c) phylogenies place PAV and PAS in a monophyletic clade, with MAV sister to both species. The placement of CN-PAV also differs between gene trees. It clusters on a branch sister to all PAV and PAS isolates in the MP tree, but its RdRp is clearly derived from the same lineage as PAS. Analysis of the complete genomes of a number of different species in the family Luteoviridae supports the topology of the RdRp tree, i.e. MAV and PAV form a monophyletic clade and CN-PAV and PAS form a monophyletic clade (Fig. 1d). It also shows that SGV, which is unassigned to a genus in the family Luteoviridae, forms a monophyletic clade with luteovirus isolates and is only distantly related to the polerovirus RPV. Disagreement between RdRp and CP phylogenies could be the result of selection and/or genetic drift that has led to divergence of the CN-PAV and MAV CP sequences. It is also possible that ancestral luteoviruses have experienced recombination in the CP. The position of isolate ASL-1 in the CP and RdRp trees also warrants further investigation. ASL-1 shares greater nucleotide sequence identity with PAS isolates in its CP, but greater identity with PAV isolates in its RdRp sequence. ASL-1 may be a product of recombination between PAV and PAS or, alternatively, CP similarity could be the result of convergent evolution.

View this table: [in this window] [in a new window]   Table 2. Nucleotide diversity per site () in the RdRp, CP and MP genes of BYDV-PAV and PAS isolates collected in New York and Morocco Numbers in parentheses are standard deviations of estimates. DnaSP was used to estimate as the mean of pairwise comparisons among sequences in a population (Nei, 1987). Freidman's non-parametric test of population-central values was used to compare values. Values with the same letter were included in the same analysis and capitalization denotes values found to be significantly different (P<0.05).

View this table: [in this window] [in a new window]   Table 3. ML analysis of codon substitution (Yang, 1997) for detection of positive selection in the RdRp, CP and MP of PAV and PAS isolates collected in New York and Morocco

	DISCUSSION

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

Geographical subdivision has been reported for the global populations of Tomato spotted wilt virus (TSWV) (Tsompana et al., 2005), Cucurbit yellow stunting disorder virus (Rubio et al., 2001) and Turnip mosaic virus (Tomimura et al., 2003), as well as for regional populations of Rice yellow mottle virus (RYMV) (Pinel et al., 2000), Sweet potato chlorortic stunt virus (Alicai et al., 1999) and Kenneyda yellow mosaic virus (Skotnicki et al., 1996). Several of these studies invoke virus adaptation in response to new hosts or transmission modalities to explain the observed pattern of variation, but few present evidence that geographical variants differ in their biology or that there is a correlation between positively selected codon sites and the geographical region from where the population was sampled. In the absence of such evidence, one could also attribute geographical subpopulation structure to transmission bottlenecks followed by genetic drift. As selection and genetic drift can both lead to a decrease in diversity within populations and an increase in diversity between populations, it is often difficult to distinguish the effects of these processes. However, selection and drift need not be mutually exclusive forces. A study by Choi et al. (2001) found considerable divergence between the American and Mexican populations of Wheat streak mosaic virus (WSMV), mostly at synonymous nucleotide sites. They concluded from these data that negative selection and drift acting simultaneously have contributed to the evolution of WSMV strains. The present study found significant genetic differentiation between the CP of the Moroccan and New York PAS populations, but no support for the hypothesis that either population deviates from a neutral model of evolution. Thus, it appears that for PAS, like WSMV, genetic drift rather than adaptation is responsible for divergence between the two populations.

Genetic diversity and selective constraint
Total nucleotide diversity was estimated to be 0.028 and 0.029 in the CP and 0.017 and 0.020 in the RdRp for the New York PAV and PAS populations, respectively. These values are similar to or lower than total nucleotide diversity values reported for other insect-transmitted plant RNA viruses, such as 0.035 for TSWV (Tsompana et al., 2005), 0.068 for Citrus tristeza virus, 0.07 for Groundnut rosette virus and 0.194 for RYMV (García-Arenal et al., 2001). Diversity did not differ significantly across genes, species or geographical regions. The only exception was the Moroccan PAS population, which has much lower nucleotide diversity in its CP (0.006) when compared with the Moroccan PAV population (0.028) or the New York PAS population (0.025). This study did not find a significant difference in selective constraint on the New York or Moroccan PAS CP. Thus, it is likely that the Moroccan population has been through a genetic bottleneck more recently than the New York population, such that there has been less time for population expansion and re-establishment of diversity. In New York, selective constraint did not differ between the PAV CP (0.055), the PAV RdRp (0.052) and the PAS RdRp (0.035), but was significantly greater for the PAS CP (0.44). This indicates that genetic variation in either gene class is not necessarily more constrained due to the function of the encoded protein. I also found that selective constraint does not differ between the PAS CP in New York and Morocco and it is less constrained than the PAV CP in both environmental contexts. Because PAV and PAS are sympatric in these locations and share the same vector specificity, it might be the case that relaxed selective constraints on the PAS CP are due to differences in plasticity of the protein rather than differences in the source or magnitude of the selection pressure acting on the virus population.

Utilizing statistical models that allow heterogeneous ratios among codon sites, researchers have found evidence of diversifying selection in genes of CMV (Moury, 2004), Potato virus Y (Moury et al., 2002) and TSWV (Tsompana et al., 2005). Positively selected amino acid sites were detected in structural proteins of each of these viruses and in transcription-related proteins of CMV and TSWV. For BYDV-PAV and PAS, the MP was under the least selective constraint when compared with genes in other functional classes, except in the Moroccan PAS population, where it seemed to be more constrained than any gene sampled from any other population. As discussed above, this could be the result of a recent population bottleneck. Four codon sites in the MP are putatively affected by positive selection. Three of these sites are located in the C-proximal part of the protein, which is known to be the nucleic acid-binding domain of the potato leaf roll virus (PLRV) (family Luteoviridae, genus Polerovirus) MP (Tacke et al., 1991). The other site is located in the N-terminal part of the protein. Experimental manipulation of the PLRV nucleotide sequence suggests that protein–protein interactions occur in this domain (Tacke et al., 1993). In interpreting the results of the PAML analysis, it is necessary to keep in mind that synonymous sites in the MP correspond to non-synonymous sites in the CP. This may influence estimates of for the MP by reducing the rate of synonymous substitutions. Relatively high estimates have been shown for overlapping reading frames in CMV (Moury et al., 2002) and PLRV (Guyader & Ducray, 2002).

In summary, the present study found that PAV and PAS are related more distantly with respect to their RdRp than their CP genes. It is not well understood how genetic variation in the BYDV RdRp relates to virus accumulation in host tissues, host range and disease severity. Future research must address this issue, as the RdRp is a significant source of genetic variation at the intra- and interspecies levels. The results of this study indicate that genetic variation in the RdRp is not correlated with the aphid-transmission phenotype of the virus, because analysis of the RdRp gene alone cannot differentiate between BYDV species. This reflects the different selection pressures acting on the RdRp and CP. Despite the different roles that they play in the virus infection cycle, selective constraint experienced by the RdRp or CP did not differ based solely upon their function. Rather, functional constraint was determined by gene function, phylogenetic history and selection pressures in a particular environment. Given our findings that New York PAS populations are genetically distinct from the Moroccan populations, it should not be assumed that isolates of the same species in different geographical regions will have similar biological characteristics. It may be necessary to independently evaluate the infection properties of isolates present in each area where BYDV hinders grain production. This information will aid growers in the development of grain varieties with the appropriate resistance and will link features in the viral genome to virus-transmission dynamics and effects on host fitness.

	ACKNOWLEDGEMENTS

 
I thank Micheal Milgroom for reviewing previous drafts of this manuscript, Alison Power for providing materials and facilities necessary for maintaining virus culture plants and aphid colonies, and Kai Blaisdell, Steve Bogdanowicz, Erik Dopman and Damon Little for their assistance in developing molecular protocols and performing cladistic analyses. All molecular work was performed in the Evolutionary Genetics Core Facility in the Department of Ecology and Evolutionary Biology at Cornell University. Research was supported by the National Institutes of Health Ruth L. Kirschstein individual research fellowship and the Andrew W. Mellon student research grant.

	REFERENCES

TOP ABSTRACT INTRODUCTION METHODS RESULTS DISCUSSION REFERENCES

 
Alicai, T., Fenby, N. S., Gibson, R. W., Adipala, E., Vetten, H. J., Foster, G. D. & Seal, S. E (1999). Occurrence of two serotypes of sweet potato chlorotic stunt virus in East Africa and their associated differences in coat protein and HSP70 homologue gene sequences. Plant Pathol 48, 718–726.[CrossRef]

Anderson, E. J., Qiu, S. G. & Schoelz, J. E. (1991). Genetic analysis of determinants of disease severity and virus concentration in cauliflower mosaic virus. Virology 181, 647–655.[CrossRef][Medline]

Bencharki, B., Mutterer, J., El Yamani, M., Ziegler-Graff, V., Zaoui, D. & Jonard, G. (1999). Severity of infection of Moroccan barley yellow dwarf virus PAV isolates correlates with variability in their coat protein sequences. Ann Appl Biol 134, 89–99.

Bisnieks, M., Kvarnheden, A., Sigvald, R. & Valkonen, J. P. T. (2004). Molecular diversity of the coat protein-encoding region of Barley yellow dwarf virus-PAV and Barley yellow dwarf virus-MAV from Latvia and Sweden. Arch Virol 149, 843–853.[CrossRef][Medline]

Chay, C. A., Smith, D. M., Vaughan, R. & Gray, S. M. (1996a). Diversity among isolates within the PAV serotype of barley yellow dwarf virus. Phytopathology 86, 370–377.

Chay, C. A., Gunasinge, U. B., Dinesh-Kumar, S. P., Miller, W. A. & Gray, S. M. (1996b). Aphid transmission and systemic plant infection determinants of barley yellow dwarf luteovirus-PAV are contained in the coat protein readthrough domain and 17-kDa protein, respectively. Virology 219, 57–65.[CrossRef][Medline]

Choi, I.-R., Hall, J. S., Henry, M., Zhang, L., Hein, G. L., French, R. & Stenger, D. C. (2001). Contributions of genetic drift and negative selection on the evolution of three strains of wheat streak mosaic tritimovirus. Arch Virol 146, 619–628.[CrossRef][Medline]

Dinesh-Kumar, S. P. & Miller, W. A. (1993). Control of start codon choice on a plant viral RNA encoding overlapping genes. Plant Cell 5, 679–692.[Abstract/Free Full Text]

Filichkin, S. A., Lister, R. M., McGrath, P. F. & Young, M. J. (1994). In vivo expression and mutational analysis of the barley yellow dwarf virus readthrough gene. Virology 205, 290–299.[CrossRef][Medline]

Fu, Y.-X. & Li, W.-H. (1993). Statistical tests of neutrality of mutations. Genetics 133, 693–709.[Abstract]

García-Arenal, F., Fraile, A. & Malpica, J. M. (2001). Variability and genetic structure of plant virus populations. Annu Rev Phytopathol 39, 157–186.[CrossRef][Medline]

Gildow, F. E. (1987). Virus-membrane interactions involved in circulative transmission of luteoviruses by aphids. Curr Top Vect Res 4, 93–120.

Gildow, F. E. (1993). Evidence for receptor-mediated endocytosis regulating luteovirus acquisition by aphids. Phytopathology 83, 270–277.[CrossRef]

Guyader, S. & Ducray, D. G. (2002). Sequence analysis of Potato leafroll virus isolates reveals genetic stability, major evolutionary events and differential selection pressure between overlapping reading frame products. J Gen Virol 83, 1799–1807.[Abstract/Free Full Text]

Hudson, R. R., Boos, D. D. & Kaplan, N. L. (1992). A statistical test for detecting geographic subdivision. Mol Biol Evol 9, 138–151.[Abstract]

Koev, G., Liu, S., Beckett, R. & Miller, W. A. (2002). The 3'-terminal structure required for replication of barley yellow dwarf virus RNA contains an embedded 3' end. Virology 292, 114–126.[CrossRef][Medline]

Li, W.-H. (1993). Unbiased estimation of the rates of synonymous and nonsynonymous substitution. J Mol Evol 36, 96–99.[CrossRef][Medline]

Lister, R. M. & Ranieri, R. (1995). Distribution and economic importance of barley yellow dwarf. In Barley Yellow Dwarf: 40 Years of Progress, pp. 29–53. Edited by C. J. D'Arcy & P. A. Burnett. St Paul, MN: American Phytopathological Society.

Lucio-Zavaleta, E., Smith, D. M. & Gray, S. M. (2001). Variation in transmission efficiency among Barley yellow dwarf virus-RMV isolates and clones of the normally inefficient aphid vector, Rhopalosiphum padi. Phytopathology 91, 792–796.[Medline]

Malmstrom, C. M., Hughes, C. C., Newton, L. A. & Stoner, C. J. (2005a). Virus infection in remnant native bunchgrasses from invaded California grasslands. New Phytol 168, 217–230.[CrossRef][Medline]

Malmstrom, C. M., McCullough, A. J., Johnson, H. A., Newton, L. A. & Borer, E. T. (2005b). Invasive annual grasses indirectly increase virus incidence in California native perennial bunchgrasses. Oecologia 145, 153–164.[CrossRef][Medline]

Mastari, J., Lapierre, H. & Dessens, J. T. (1998). Asymmetrical distribution of barley yellow dwarf virus PAV variants between host plant species. Phytopathology 88, 818–821.[Medline]

Mayo, M. A. (2002). ICTV at the Paris ICV: results of the plenary session and the binomial ballot. Arch Virol 147, 2254–2260.[CrossRef]

Miller, W. A., Liu, S. & Beckett, R. (2002). Barley yellow dwarf virus: Luteoviridae or Tombusviridae? Mol Plant Pathol 3, 177–183.[CrossRef]

Mishmar, D., Ruiz-Pesini, E., Golik, P. & 10 other authors (2003). Natural selection shaped regional mtDNA variation in humans. Proc Natl Acad Sci U S A 100, 171–176.[Abstract/Free Full Text]

Mohan, B. R., Dinesh-Kumar, S. P. & Miller, W. A. (1995). Genes and cis-acting sequences involved in replication of barley yellow dwarf virus-PAV RNA. Virology 212, 186–195.[CrossRef][Medline]

Moury, B. (2004). Differential selection of genes of cucumber mosaic virus subgroups. Mol Biol Evol 21, 1602–1611.[Abstract/Free Full Text]

Moury, B., Cardin, L., Onesto, J.-P., Candresse, T. & Poupet, A. (2001). Survey of Prunus necrotic ringspot virus in rose and its variability in rose and Prunus spp. Phytopathology 91, 84–91.[Medline]

Moury, B., Morel, C., Johansen, E. & Jacquemond, M. (2002). Evidence for diversifying selection in Potato virus Y and in the coat protein of other potyviruses. J Gen Virol 83, 2563–2573.[Abstract/Free Full Text]

Nei, M. (1987). Molecular Evolutionary Genetics. New York: Columbia University Press.

Nei, M. & Gojobori, T. (1986). Simple methods for estimating the numbers of synonymous and nonsynonymous nucleotide substitutions. Mol Biol Evol 3, 418–426.[Abstract]

Pinel, A., N'Guessan, P., Bousalem, M. & Fargette, D. (2000). Molecular variability of geographically distinct isolates of rice yellow mottle virus in Africa. Arch Virol 145, 1621–1638.[CrossRef][Medline]

Power, A. G. & Gray, S. M. (1995). Aphid transmission of barley yellow dwarf viruses: interactions between viruses, vectors, and host plants. In Barley Yellow Dwarf Virus: 40 Years of Progress, pp. 259–289. Edited by C. J. D'Arcy & P. A. Burnett. St Paul, MN: American Phytopathological Society.

Price, E. W. & Carbone, I. (2003). SNAP Workbench. Department of Plant Pathology, North Carolina State University.

Raybould, A. F., Maskell, L. C., Edwards, M.-L., Cooper, J. I. & Gray, A. J. (1999). The prevalence and spatial distribution of viruses in natural populations of Brassica oleracea. New Phytol 141, 265–275.[CrossRef]

Rochow, W. F. (1969). Biological properties of four isolates of barley yellow dwarf virus. Phytopathology 59, 1580–1589.[Medline]

Rochow, W. F. & Muller, I. (1971). A fifth variant of barley yellow dwarf virus in New York. Plant Dis 55, 874–877.

Rozas, J., Sánchez-DelBarrio, J. C., Messeguer, X. & Rozas, R. (2003). DnaSP, DNA polymorphism analyses by the coalescent and other methods. Bioinfomatics 19, 2496–2497.

Rubio, L., Abou-Jawdah, Y., Lin, H.-X. & Falk, B. W. (2001). Geographically distant isolates of the crinivirus Cucurbit yellow stunting disorder virus show very low genetic diversity in the coat protein gene. J Gen Virol 82, 929–933.[Abstract/Free Full Text]

Sacristán, S., Fraile, A., Malpica, J. M. & García-Arenal, F. (2005). An analysis of host adaptation and its relationship with virulence in Cucumber mosaic virus. Phytopathology 95, 827–833.[Medline]

Skotnicki, M. L., Mackenzie, A. M. & Gibbs, A. J. (1996). Genetic variation in populations of kennedya yellow mosaic tymovirus. Arch Virol 141, 99–110.[CrossRef][Medline]

Smith, G. R., Borg, Z., Lockhart, B. E. L., Braithwaite, K. S. & Gibbs, M. J. (2000). Sugarcane yellow leaf virus: a novel member of the luteoviridae that probably arose by inter-species recombination. J Gen Virol 81, 1865–1869.[Abstract/Free Full Text]

Swofford, D. L. (2000). PAUP*: Phylogenetic analysis using parsimony (*and other methods), version 4.0b10. Sunderland, MA: Sinauer Associates.

Tacke, E., Prüfer, D., Schmitz, J. & Rohde, W. (1991). The potato leafroll luteovirus 17K protein is a single-stranded nucleic acid-binding protein. J Gen Virol 72, 2035–2038.[Abstract/Free Full Text]

Tacke, E., Schmitz, J., Prüfer, D. & Rohde, W. (1993). Mutational analysis of the nucleic acid-binding 17 kDa phosphoprotein of potato leafroll luteovirus identifies an amphiphatic alpha-helix as the domain for protein/protein ineractions. Virology 197, 274–282.[CrossRef][Medline]

Tajima, F. (1983). Evolutionary relationship of DNA sequences in finite populations. Genetics 105, 437–460.[Abstract/Free Full Text]

Tajima, F. (1989). Statistical method for testing the neutral mutation hypothesis by DNA polymorphism. Genetics 123, 585–595.[Abstract/Free Full Text]

Tomimura, K., Gibbs, A. J., Jenner, C. E., Walsh, J. A. & Ohshima, K. (2003). The phylogeny of Turnip mosaic virus; comparisons of 38 genomic sequences reveal a Eurasian origin and a recent ‘emergence’ in east Asia. Mol Ecol 12, 2099–2111.[CrossRef][Medline]

Tsompana, M., Abad, J., Purugganan, M. & Moyer, J. W. (2005). The molecular population genetics of the Tomato spotted wilt virus (TSWV) genome. Mol Ecol 14, 53–66.[CrossRef][Medline]

Wang, X., Chang, S., Jin, Z., Li, L. & Zhou, G. (2001). Nucleotide sequences of the coat protein and readthrough protein genes of the Chinese GAV isolate of barley yellow dwarf virus. Acta Virol 45, 249–252.[Medline]

Yang, Z. (1997). PAML: a program package for phylogenetic analysis by maximum likelihood. Comput Appl Biosci 13, 555–556.[Free Full Text]

Yang, Z., Nielsen, R., Goldman, N. & Krabbe Pedersen, A.-M. (2000). Codon-substitution models for heterogeneous selection pressure at amino acid sites. Genetics 155, 431–449.[Abstract/Free Full Text]

Yang, Z., Wong, W. S. W. & Nielsen, R. (2005). Bayes empirical Bayes inference of amino acid sites under positive selection. Mol Biol Evol 22, 1107–1118.[Abstract/Free Full Text]

Received 12 January 2006; accepted 30 May 2006.

This Article Abstract Full Text (PDF) Supplementary table Alert me when this article is cited Alert me if a correction is posted Citation Map Services Email this article to a friend Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Citing Articles Citing Articles via CrossRef Citing Articles via Google Scholar Google Scholar Articles by Hall, G. Search for Related Content PubMed PubMed Citation Articles by Hall, G. Agricola Articles by Hall, G.