Skip to main page content

• HOME
• Current Issue
• Archive
• Alerts
• About ASM
• Contact us
• Tech Support
• Journals.ASM.org

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Moonan, F. Articles by Mirkov, T. E. Search for Related Content PubMed PubMed Citation Articles by Moonan, F. Articles by Mirkov, T. E.

 Previous Article  |  Next Article 

Journal of Virology, February 2002, p. 1339-1348, Vol. 76, No. 3
0022-538X/01/$04.00+0     DOI: 10.1128/JVI.76.3.1339-1348.2002
Copyright © 2002, American Society for Microbiology. All Rights Reserved.

Analyses of Genotypic Diversity among North, South, and Central American Isolates of Sugarcane Yellow Leaf Virus: Evidence for Colombian Origins and for Intraspecific Spatial Phylogenetic Variation

Francis Moonan and T. Erik Mirkov^*

Department of Plant Pathology and Microbiology, Texas A&M University System Agricultural Experiment Station, Weslaco, Texas 78596

Received 8 May 2001/ Accepted 23 October 2001

Top Abstract Introduction Materials and Methods Results Discussion References

 
We have analyzed the genotypic diversity of Sugarcane yellow leaf virus (SCYLV) collected from North, South, and Central America by fingerprinting assays and selective cDNA cloning and sequencing. One group of isolates from Colombia, designated the C-population, has been identified as residing at the root node between a separable superpopulation structure of SCYLV and other members of the family Luteoviridae, indicating that the progenitor viruses of the North, South, and Central American isolates of the SCYLV superpopulation most likely arose from a C-population structure. From a model of intrafamilial evolution (F. Moonan et al., Virology 269:156–171, 2000), a prediction could be made that within the SCYLV species, the capacity of genomic sequence divergence would range from lowest in the capsid protein open reading frame 3 (ORF 3) to highest in a region spanning across the carboxy-terminal end of the RNA-dependent RNA polymerase ORF. We have demonstrated the validity and applicability of this intrafamilial model for the prediction of intraspecies SCYLV diversity. Analysis of spatial phylogenetic variation (SPV) within the SCYLV isolates could not be assessed by application of a "partial likelihoods assessed through optimization" (PLATO)-derived intraspecies model alone. However, application of a PLATO-derived intrafamilial model with the intraspecies-derived model allowed distinction of three forms of SPV. Two of the SPV forms identified correspond to the extremes in a continuum of sequence evolution displayed in a SCYLV superpopulation structure, and the third form was diagnostic of a C-population structure. The application of these types of models has value in terms of predicting the types of SCYLV intraspecies diversity that may exist worldwide, and in general, may be useful in application for more informed design of transgenes for use in the elicitation of homology-dependent virus resistance mechanisms in transgenic plants.

Top Abstract Introduction Materials and Methods Results Discussion References

 
In the Americas, Sugarcane yellow leaf virus (SCYLV) is associated with a disease referred to as yellow leaf syndrome (YLS [3, 5, 6, 47, 55]), although a similar disease via phytoplasma infection may produce otherwise identical symptomology and is also currently referred to as YLS disease in certain areas of the world (6, 45). In sugarcane, losses of as high as 50% have been estimated to have occurred in field sites as the result of virus-induced YLS (55).

The complete SCYLV genome has been sequenced and characterized and most closely resembles viruses of the family Luteoviridae in the genus Polerovirus (36, 48). The results of these studies indicate that SCYLV, like Polerovirus members of the family Luteoviridae, encode at least six definable open reading frames (ORFs) typically listed as ORFs 0 to 5 (4, 7, 31, 32, 35, 36, 43, 48). ORFs 0 and 1 are thought to produce peptides via alternate translational start sites from the monopartite genomic RNA of SCYLV, and ORFs 3 and 4 are thought to produce peptides via alternate translation from a subgenomic RNA. These studies also indicate that a fusion peptide comprising a sequence encoded by ORF 2 is produced by a -1 frameshift during ORF 1 translation and that a fusion peptide comprising ORF 5 is produced via a translational readthrough of the peptide encoded by ORF 3, a finding consistent with the genome organization of well-characterized members of the genus Polerovirus (4, 7, 31, 32, 35, 36, 43, 48). The SCYLV peptides encoded by ORFs 0 and 1 are of unknown function and a protease, respectively, and the peptides encoded by ORFs 3, 4, and 5 appear to be structural proteins comprising the virion particle, with the ORF 3 sequence the primary capsid protein (36, 48). The peptide sequence encoded by ORF 2 appears to be multifunctional, including sequence for both an RNA-dependent RNA polymerase (RdRp) and a putative genome linked viral protein (VPg), and the VPg peptide is thought to be processed from this multifunctional peptide by proteolytic cleavage and covalently conjugated to the 5' terminus of the SCYLV genome, similar to that in members of the genus Polerovirus (4, 7, 31, 32, 35, 36, 43, 48).

SCYLV as defined by the Seventh Report of the International Committee on the Taxonomy of Viruses (54) is an unclassified member of the family Luteoviridae, although its overall genome structure most closely resembles that of members of the genus Polerovirus (36, 48). The Luteoviridae, as currently defined, include the genera Luteovirus, Polerovirus, and Enamovirus (54). The structure and classification of the sequences encoding the RNA-dependent RNA polymerase of the Luteoviridae have served as the primary basis for generic distinction within the family (7, 54). Based upon the classification system of Koonin and Dolja (25), Luteovirus members have genomes that encode RNA-dependent RNA polymerases classified as supergroup II; Polerovirus members have RdRps classified as supergroup I; and the sole member of the Enamovirus genus, Pea enation mosaic virus 1 (PEMV-1), has an RdRp sequence that contains sequences of both RdRp supergroup I and II origin (36). The genomes of PEMV-1, Soybean dwarf virus, and SCYLV exhibit spatial phylogenetic variation (SPV [18]) that is thought to have arisen via recombination between polerovirus and luteovirus ancestors after the divergence of these two progenitor groups (4, 7, 17, 18, 31, 32, 35, 36, 43, 54).

Our previous comparisons of the SCYLV genome with other members of the family Luteoviridae involved the development of an intrafamilial model of SPV (36). Based upon this model, we predicted that the SCYLV genomic sequence diversity would be lowest in a region spanning the capsid protein ORF 3, low to intermediate in a region spanning the protease ORF 1 to within the RdRp-encoded ORF 2, and highest in a region spanning from the RdRp to an untranslated sequence located 5' proximal to an ORF 3 and 4 polycistron. To test the predictions of this model, we collected SCYLV isolates from field sites throughout North, South, and Central America and analyzed the genotypic diversity among these isolates. The results of our analyses not only demonstrate the validity and applicability of this model, in terms of predicting the types of SCYLV intraspecific diversity, but have also aided us in identifying a population of isolates from Colombia, which we refer to as the C-population, that most likely represents the ancestral population of all other sampled American SCYLV isolates.

Top Abstract Introduction Materials and Methods Results Discussion References

 
Plant materials, SCYLV infectivity diagnosis, and amplicon production. Plant tissues samples were collected from sugarcane plants with leaf-yellowing symptoms diagnostic of YLS. RNA extraction and purification from sugarcane tissue samples for Northern blot analyses were performed essentially as described by Ingelbrecht et al. (23) by using a double-stranded DNA probe derived from pFM261 (36). A summary of the samples collected that were positive for SCYLV infection is presented in Table 1. Amplicons were produced as diagnostic fragments by using reverse transcription (RT)-PCR methods and were designated PROREP (protease to replicase), REPUTR (replicase to untranslated region and beginning of capsid protein), and CPRT (capsid protein to beginning of capsid readthrough protein region). The positions of the PROREP, REPUTR, and CPRT amplicons, in relation to the SCYLV-A genome, are shown in Fig. 4. SCYLV first-strand cDNA was produced by RT of 3 µg of total plant RNA with a Universal Riboclone cDNA Synthesis System (Promega, Palo Alto, Calif.), in a 25-µl volume at 42°C and with 5 U of avian myeloblastosis virus reverse transcriptase and 50 ng of oligonucleotide primers, according to the manufacturer’s instructions. The following oligonucleotide primers were used for the RT reactions: oFM359 (5'-GCTCTCCACAAAGCTATCT-3'), oFM387 (5'-CTGACATTCCTTCGTGAGC-3') and oFM361 (5'-TGTTTTCACGATGTGGTTC-3'). RT reaction mixtures were diluted with an additional 12.5 µl of water and then utilized in subsequent PCRs. For each accession, PCRs were done in triplicate for each of the PROREP, REPUTR, and CPRT amplicons. PCRs were done in 50-µl reaction volumes of 1x Qiagen Taq Polymerase Buffer with MgCl₂, 2 mM deoxynucleoside triphosphates, and 1 to 5 µl of first-strand cDNA mixture. PCRs were for 30 cycles of 95°C for 1 min, 52°C (PRO-REP) or 58°C (REPUTR and CPRT) for 2 min, and 72°C for 2 min. PCRs were done to produce the following amplicons with the following primer combinations: PROREP with oFM359 and oFM323 (5'-CAGACATTGCTGATTAC-3'), REPUTR with oFM387 and oFM386 (5'-AGATAGCTTTGTGGAGAGC-3'), and CPRT with oFM361 and oFM366 (5'-GCTCACGAAGGAATGTCAG-3'). Amplicons were size fractionated on agarose gels, and amplicons were gel purified with a Geneclean II Kit (Bio 101, Carlsbad, Calif.) according to the manufacturer’s instructions prior to their use in fingerprinting or cloning.

View this table: [in this window] [in a new window]

TABLE 1. Origins of North, South, and Central American isolates of SCYLV and nature of the genotypic data collected for this studya

View larger version (37K): [in this window] [in a new window]

FIG. 4. Analysis of SPV in isolates of SCYLV from North, South, and Central America, corresponding to nt 1518 to 4352 of the SCYLV-A genome (A), and the tree topology types exhibited in this SPV (B). (A, top to bottom) Relative position of the clonal inserts used to derive the SCYLV sequences for this region; genomic organization of the SCYLV ORFs corresponding to this region, corresponding isolate data-derived PLATO output ranges shown as boxed ranges corresponding to the SCYLV genome; corresponding PLATO output ranges derived from an intrafamilial model; distribution of tree topologies corresponding to regions of SPV; and grouping of isolates into three groups (groups A, B, and C), corresponding to the exhibited patterns of SPV. The SCYLV-A genome used as a reference, the direct extrapolation of the PLATO intrafamilial analysis-based output, and a description of this method of SPV analysis are derived from Moonan et al. (36).

Production and analysis of CAPS fingerprint gels. Amplicons generated in triplicate, from independent PCRs, were used for DNA fingerprinting. Gel-purified amplicon fragments were digested with both Sau3AI and TaqI restriction enzymes to generate the SCYLV-cleaved amplified polymorphic sequence (CAPS), and these were radiolabeled by 5' overhang end-filling reactions with Klenow enzyme (Promega) and dATP, dCTP, dTTP, and [-³²P]dCTP. CAPS reactions were fractionated on 6% polyacrylamide sequencing gels under standard denaturing conditions (44). Dephosphorylated HinfI-digested X174 DNA was labeled with [-³²P]ATP and polynucleotide kinase and used as size markers. A total of 65 CAPS allelic fragments were scored and ranged between 27 and 384 nucleotides (nt) in length: 17 CAPS allelic fragments for the CPRT amplicons, 22 for the REPUTR amplicons, and 26 for the PROREP amplicons.

cDNA cloning, sequencing, and computational analyses. CAPS markers scored as binary data were analyzed by unweighted pair group method with arithmetic mean clustering methodology with Paup, version 4.0 beta 4 (51), and the haploid data analysis components of Popgene (57). Analyses of Nei’s (37) gene diversity (h) and Nei’s (39) pairwise genetic distances were calculated with Popgene and analyzed individually, as a whole, and as geographic groupings. Nucleotide and deduced peptide alignments and CAPS-associated files are deposited at Treebase (www.herbaria.harvard.edu/treebase/index.html). REPPRO, REPUTR, and CPRT amplicons derived from one or two of the three available independent PCRs were ligated into pCR4-TOPO (Invitrogen) to generate independent clones for each of the amplicons. Only one of the CPRT amplicons of the L1, B1, and C4 accessions was sequenced (Table 1, pFM623, pFM629, and pFM624). All other amplicons were cloned and sequenced in duplicate. The consensus sequences of the SCYLV isolates are deposited in GenBank and are listed in Table 1. The T6 accession represents the source material for SCYLV-A (36). Source material for SCYLV-F (48) from variety CP65-357 was obtained and designated the F1 isolate of this study. The sequence from a Brazilian isolate described by Maia et al. (29) was designated the B0 isolate. Outgroup comparisons, nucleotide alignments of the SCYLV data set with other species of the Luteoviridae, and the description of the origins of these genomic sequences are reported in the study by Moonan et al. (36). DNA sequence generation was performed on a contract basis by the DNA Sequencing Facility at Iowa State University, Ames. Plasmid insert sequences were assembled with Seqman II (DNAstar, Inc., Madison, Wis.). Nucleotide multiple sequence alignments were analyzed with Paup 4.0 beta 4 or beta 8 (51), MEGA 2.0 (26), PHYLIP 3.57 (15), and the neighbor-joining (NJ) method of CLUSTAL X (52, 53). Deduced peptide sequences were analyzed with the NJ method of CLUSTAL X and the quartet maximum-likelihood-based method of PUZZLE (50). PUZZLE analyses were done with both the Dayhoff and Jones, Taylor, and Thornton phylogenetic models with 1,000 quartets for the partial deduced peptide sequences of ORFs 1 and 5, as well as ORFs 3 and 4, and 10,000 quartets for ORF 2. For phylogenetic analysis of nucleotide sequences with Splitstree 2.4 (22), distance data was generated with the Hasegawa-Kishino-Yano phylogenetic model (20) of Paup 4.0 beta 8. Analyses for potential recombination sites were done by "likelihood analysis of recombination in DNA" (LARD [21]) and the Recombination in DNA Program (RDP [30]). Nucleotide sequences were analyzed for SPV by the PLATO (18) and DNAml (15) maximum-likelihood methods and the NJ method of CLUSTAL X. Because the consensus sequence for SCYLV-A contains three alternate nucleotide substitutions in the region spanning the PROREP amplicon-spanning region and six alternate nucleotide substitution patterns in the REPUTR amplicon-spanning region, the 8 and 64 sequentially derived sequences for SCYLV-A were initially analyzed with the 14 individual sequences of the other isolates assayed. The results (not shown) indicated that there was no overlap of the different SCYLV-A combinations with the sequences of the other isolates. By using composited pairs of the #1 and #8 PROREP and #1 and #64 REPUTR-spanning regions, which represented all individual nucleotide substitutions in SCYLV-A, two 2,835-nt SCYLV-A sequences were generated and referenced as T6-1 and T6-2. The transition/transversion ratio used in both the PLATO and DNAml was calculated with 14 generated composite 2,835-nt clonal sequences for each of the B1, N6, L1, G2, C1, C3, and C4 accessions, along with the T6-1 and T6-2 sequences. The overall transition/transversion ratio for this sequence range was calculated to be 1.5182, and this ratio was used with the DNAml (15) and the HKY phylogenetic model with PLATO.

Top Abstract Introduction Materials and Methods Results Discussion References

 
Of 35 accessions collected, 16 samples tested positive for SCYLV infection and are listed in Table 1. Accessions T3, T4, and T5 (not listed) functioned as negative controls and represent plants independently generated via meristem culture from the T6 accession via previously described methods (46). Numerous anecdotal reports suggest that SCYLV is probably distributed worldwide. Our survey of the Americas confirms the previous reports that SCYLV is endemic in all of the major sugarcane-growing regions in the continental United States (5, 36, 45, 48), as well as in Brazil (29, 55), but also indicates that the geographic range of this virus may also now be formally expanded to include Guatemala, Colombia, and Argentina.

Fingerprint analyses of isolates. A total of 65 CAPS allelic fragments were scored in binary from the amplicons sampled, and genotypic clustering was performed with the UPGMA methodology of Paup. The resulting dendrogram is shown in Fig. 1A. Fingerprint patterns for the T6, F1, F2, G3, G6, G7, G8, and G9 amplicons were identical in triplicate, while fingerprint patterns for the B2 and N6 isolates were identical in triplicate, and these are represented in the dendrogram in Fig. 1A as a vertical bar subtended by concave branches. An identical fingerprint patterning in triplicate implies an effective Nei (37, 38, 39) pairwise genetic distance of zero, indicating that groups with identical fingerprints are most likely representative of the same genotype. The Cali, Colombia, isolate members C1, C3, and C4, which resolved as a putative cluster in the PAUP-derived UPGMA dendrogram in Fig. 1A, exhibited differences in gene diversity as a group in comparison to the other isolates. The overall gene diversity for all of the isolates was calculated as h = 0.0471. In contrast, the C1/C3/C4 group gene diversity was calculated as h = 0.0547, while the remaining isolates were calculated as h = 0.00248, indicating that the bulk of the gene diversity from the samples assayed was represented primarily by the members of the C1/C3/C4 group of isolates. Analyses of Nei (39) pairwise genetic distance ranges within the C1/C3/C4 group and within the remaining isolates were 0.0313 to 0.1313 and 0 to 0.0473, respectively, which reflect this difference in calculated gene diversity measurements.

View larger version (22K): [in this window] [in a new window]

FIG. 1. Phylogenetic relationships assessed with both a UPGMA dendrogram derived from fingerprinting data (A) and phylograms derived from nucleotide (B and C) and deduced peptide (D and E) data. The designations of the isolates and the origins of the sequence or fingerprint data used in the diagrams are given in Table 1. Nucleotide sequence alignments were analyzed with the F84 phylogenetic model of DNAml of the PHYLIP package, and deduced peptide sequence alignments were analyzed with the maximum-likelihood based quartet puzzling method of PUZZLE. The nucleotide sequence ranges analyzed for the isolate phylograms (B and C), respectively, correspond to nt 1518 to 4352 and nt 3144 to 4238 of the SCYLV-A genome described by Moonan et al. (36). Relationships between the deduced peptide sequences of RdRp ORF 2 of SCYLV are shown for both the complete RdRp peptide sequences (D) and a partial RdRp sequence corresponding to amino acid residue positions 471 to 572 of the RdRp of SCYLV-A (E).

Sequence analyses of isolates. Based upon the analysis of the genotypes represented in Fig. 1A, amplicons for B1, N6, L1, G2, C1, C3, and C4 were selected for DNA sequencing. From the five or six independent amplicons cloned and sequenced for each accession, which represented paired data sets, a 2,832- or 2,835-nt consensus sequence for each isolate was then derived, which corresponded to nt 1518 to 4352 of the SCYLV-A and T6 isolate genome (36). Alignment of these data sets indicated that one predominant insert/deletion (INDEL) could be identified in the alignment, which corresponded to one gap of 3 nt, in the six individual paired data sets from the REPUTR amplicons of the C1, C3, and C4 isolates. This INDEL corresponded to the nt 3512 to 3514 of the SCYLV-A genome, which is positioned in an untranslated region of the genome, between ORF 2 and ORFs 3. Pairwise genetic distances were calculated with the Jukes-Cantor (J-C) method of MEGA 2.0 with 1,000 replicates, and mean sequence divergence measurements were calculated. The MEGA 2.0-calculated mean sequence diversity for the set of nine consensus sequences, excluding the B0 data, was 0.01835, with a standard error (SE) of 0.0166; the mean diversity of the C1/C3/C4 and T6/F1/L1/G2/B1/N6 sequences as groups was 0.00557 (SE = 0.0084); and between these same groups the nucleotide diversity was 0.01278 (SE = 0.00159). The calculated F84 and J-C pairwise genetic distances within the C1, C3, and C4 isolates, respectively, ranged from 0 0.0018 to 0.0440 and 0.0018 to 0.0160, and within the T6, F1, L1, G2, B1, and N6 isolates they ranged from 0.0036 to 0.0107 and 0.0036 to 0.0107.

Both the NJ method of CLUSTAL X and the maximum-likelihood method of DNAml were used to generate phylograms. These were compared to the trees generated by the paired sample sets. In either of these paired sample sets, or the consensus sequences, there was little difference in the tree topologies in the most common and most likely trees generated by these methods. Figure 1B shows the resulting DNAml phylogram with the consensus sequence data, which may be compared to the fingerprint derived data shown in Fig. 1A. A comparison of the dendrogram in Fig. 1A and the phylogram in Fig. 1B indicates that the results of the fingerprint and sequence analysis demonstrate the same pattern of similarities between isolates, supporting a contention that the C1/C3/C4 group of isolates sampled represents a population structure separable from the T6/F1/F2/L1/G2/B1/B2/N6 group of isolates sampled.

Partial sequences from a isolate of SCYLV from Campinas, São Paulo, Brazil, listed in Table 1 as a B0 isolate were analyzed in a similar fashion and represented nt 3144 to 4238 of the SCYLV-A (T6 isolate) genome. These B0 data, produced by Maia et al. (29), included the sequence spanning a small portion of SCYLV ORF 2 and the complete ORFs 3 and 4 of SCYLV, which are encoded in a polycistron within the SCYLV genome. The generated alignments indicated an INDEL of a single A nucleotide corresponding to a nucleotide position between nt 3623 and nt 3624 of the SCYLV-A (T6 isolate) genome, which is positioned in the untranslated region between ORF 2 and ORF 3. The B0 resulting DNAml phylogram for the SCYLV-A corresponding nucleotide range from 3144 to 4238 is shown in Fig. 1C. As shown in Fig. 1C, the C1, C3, and C4 isolates from this analysis resolved as a clade most closely associated with the B0 isolate. Based upon both the fingerprint and sequence-derived analyses, the Cali, Colombia, C1/C3/C4 sampled group could be considered as a separable and distinct geographic population set from the sampled T6/F1/F2/L1/G2/B1/B2/N6 geographic population set. Arbitrarily, we assigned the designation of C-population to the C1/C3/C4 population set to reflect its Colombian origins and the designation "superpopulation" to the T6/F1/F2/L1/G2/B1/B2/N6 population set to reflect its larger size from the pool of the overall samples of isolates studied, the apparent genetically contiguous characteristics of the isolates sampled and analyzed, as well as the widespread geographic distribution of the isolates sampled and studied, which spanned across geographic locales in both North and South America.

To determine the degree to which the differences in nucleotide sequence corresponded to differences in deduced peptide sequence composition of the complete and partial ORF sequences encoded, we compared the deduced peptide sequences of partial ORFs 1 and 5 and the complete ORFs 2, 3 and 4 with alignments in which the B0 isolate derived data was excluded (Fig. 1D) or included (Fig. 1E). Trees were generated by the NJ method, as well as by the maximum-likelihood method of PUZZLE. For the PUZZLE-generated trees, both the Dayhoff and JTT phylogenetic models were used, but there was no significant difference in tree topology seen with these two methods. The trees from the ORF 3 and 4 deduced peptide data, in which the B0 sequences were included or excluded exhibited no significant differences, with the SE of the branchtree lengths overlapping, thus represented the tree topology of a single clade for all of these assessments (data not shown). The trees generated by analyses of the deduced peptide sequence of ORF 2, however, consistent with our pairwise comparisons, indicated that a separate clade composed of sequences derived from the C1, C3, and C4 isolates could be delimited from the remaining isolate sequences. Figure 1D shows the PUZZLE-derived phylogram from the B0 excluded alignment, and Fig. 1E shows the results from inclusion of the B0 isolate data. As shown in Fig. 1D and E, the individual deduced peptide sequences of the six C1, C3, and C4 sampled sequences resolved in phylograms as a clade separable from the remaining sequences, as did the phylograms generated by the nucleotide sequence-generated phylograms (Fig. 1B and C).

The isolates’ pairwise genetic distances were calculated with the same nucleotide sequence alignments used to produce Fig. 1B and 1C, with the HKY phylogenetic model of Paup 4.0 beta 8, and the resulting data were further analyzed with Splitstree 2.4. As shown in Fig. 2, Splitstree output from the B0 excluded alignment, which represented the nucleotide range from nt 1518 to nt 4352 of the SCYLV-A genome, produced a network diagram in which a series of short quadrangles are generated that extend from the T6 isolate to the N6 isolate. This quadrangle set is linked to a long quadrangle, which at the extreme is associated with the C1, C3, and C4 isolate sequences. In the three short quadrangles represented in Fig. 2, the T6 isolate is positioned from one quadrangle which is opposed by a B1/N6-associated quadrangle, in which a third quadrangle may be joined to these two, to which the G2/F1/L1 isolates are associated. The C1 isolate is placed in this network within the quadrangle split that leads directly to a structure of three short quadrangles within the SCYLV superpopulation. Splitstree output from a B0 inclusive alignment which represented nucleotides ranging from nt 3144 to nt 4238 of the SCYLV-A genome also produced a network diagram similar to that shown in Fig. 2 (data not shown). Although the C-population cluster from this B0 inclusive output was also anchored at a node at which the C1 isolate was placed, rather than a quadrangle extending from the C-population to the superpopulation cluster, a single branch linked the two. In addition, rather than a set of networked quadrangles at the superpopulation basal node, the superpopulation isolates were distributed in a radial tree topology extending from a central superpopulation node and included the B0 isolate sequence.

View larger version (16K): [in this window] [in a new window]

FIG. 2. Splitstree network diagram derived from nt 1518 to 4352 of the SCYLV-A genome, illustrating the most likely possible phylogenetic relationships between isolates of SCYLV. The network was produced in Splitstree from an HKY phylogenetic model produced with PAUP and corresponds to the data used to produce the DNAml phylogram illustrated in Fig. 1B. The designations of the isolates and the origins of the sequences used to produce the network are given in Table 1.

Sequences represented in Fig. 1B and 2 were aligned with those of other members of the family Luteoviridae, with an alignment that has been previously described used as a guide (36). Pairwise genetic distances were calculated with the J-C model of MEGA 2.0, the J-C and F84 phylogenetic models of DNAml, and the HKY model of Paup 4 beta 8. Data from the Paup 4.0 beta 8-produced model were analyzed with Splitstree 2.4. As shown in Fig. 3, output from Splitstree placed the C1 isolate sequence at a node of a quadrangle from which the C3 and C4 isolates branched and at a node placed directly at a branch that extends into a node from which predominantly Polerovirus genomic sequences were generated. This C1 isolate node, as shown in Fig. 3, also leads directly to a node from which two short branches lead to individually separable but internally unresolvable clusters of the T6/F1/L1/G2 and the B1/N6 isolate sequences.

View larger version (29K): [in this window] [in a new window]

FIG. 3. Splitstree network diagram derived from nt 1518 to 4352 of the SCYLV genome, illustrating the most likely phylogenetic relationships between the SCYLV isolates and other members of the family Luteoviridae. The network was produced in Splitstree from an HKY phylogenetic model produced with PAUP. The designations of the isolates and the origins of the SCYLV sequences used to produce the network are given in Table 1, and the origins of the Luteoviridae sequences are described in Moonan et al. (36). Isolate clusters formally designated as belonging to the Enamovirus, Polerovirus, and Luteovirus genera are encircled, as are the postulated superpopulation and C-population clusters of SCYLV. Within the network diagram, assignment of the B0 isolate of Maia et al. (29) (in diamond) is based on analyses of partial sequence information data shown in Fig. 1C and E, and assignment of other isolates of SCYLV to the superpopulation (in box) is based upon UPGMA analysis of fingerprint data, as shown in Fig. 1A.

Analyses of SPV. Using the paired data sets of 16 sequences, we analyzed the 2,835-nt sequence range of SCYLV with both LARD and RDP. LARD was initially used for detection of recombination and SPV in Dengue virus (21), and RDP is a program for aiding in the detecting of recombination among a set of aligned viral sequences (30). Utilizing combinations of parental sequences with both of these programs, no potential recombination sites could be identified. By using the same paired data set, we then performed an analysis of SPV in the same fashion described by Moonan et al. (36), except that the transition/transversion ratio utilized was 1.5182. From the intraspecies PLATO analysis output, z values of >3.57 were statistically significant and are shown in Fig. 4A, below the corresponding PLATO output from the intrafamilial derived range used with the SCYLV-A genome (36). NJ method results and DNAml trees were generated and analyzed with the intraspecies derived PLATO output ranges, but the corresponding trees for these ranges showed little congruence in tree topologies generated by these two phylogenetic methods. The sequence ranges were then reanalyzed by using the corresponding breakpoints extrapolated from the Luteoviridae sequence alignment derived intrafamilial PLATO output (36), along with the breakpoints generated from the intraspecies PLATO output. The results, shown in Fig. 4, indicate that SPV could be delimited within the SCYLV isolates by utilizing this approach. From the side-by-side individual tree comparisons derived from the NJ method and DNAml tree topologies, four tree topologies could be delimited, which are shown in Fig. 4B. Analyses of the resulting trees indicated that the overall distribution of SPV allowed the assignment of sequences to three possible groups: (i) group A, containing the T6, F1, L1, and G2 isolate sequences; (ii) group B, containing the B2 and N6 isolate sequences; and (iii) group C, containing the C1, C3, and C4 isolate sequences. These groups were delimited by the four tree topologies shown in Fig. 4B: (i) an I-tree ("I" for indistinguishable), in which there was no statistical clustering of a sequence range into any cluster beyond a single clade; (ii) a B-tree, in which B group designates clustered significantly different from the other isolates, which among themselves showed no significant statistical difference in terms of genetic distances; (iii) a C-tree, in which C group designates clustered significantly different from the other isolates, which among themselves showed no significant statistical difference in clustering; and (iv) an X-tree, in which the group A, group B, and group C isolate clusters were separated in a trichotomy.

Top Abstract Introduction Materials and Methods Results Discussion References

 
SCYLV population structures may be delimited, and their evolutionary relationships inferred, by utilizing phylogenetic models. In terms of plant viral strain discrimination, the most common molecular criterion used is based on comparisons of nucleotide sequence identity, as well as the derived deduced peptide sequence similarities of complete ORFs. The most commonly utilized deduced peptide sequences used in the discrimination of strains encode the RdRp and capsid proteins. Our sampling of genotypic diversity of SCYLV has thus initially involved the cloning and characterization of regions of the genome spanning these ORFs, and ca. 48% of the SCYLV genome has been sampled in these analyses. In part, we collected and analyzed SCYLV accessions from North, South, and Central America to determine whether we could discriminate, by using molecular criteria, new strains of the virus. The partial ORF 1 and 5 deduced peptide sequence comparisons revealed up to 5% amino acid sequence differences (data not shown), but interpretations from the small lengths of the deduced peptides from these partial sequences should be considered dubious. Pairwise comparisons of the deduced peptide sequences of ORF 2 indicated some differences which might be indicative of an amino acid identity difference of as high as 5% between members of the SCYLV superpopulation and the C-population. However, the estimates of SEs from the within-accession measurements and between-accession estimates indicated overlap by an SE in each of the instances, and therefore interpretations of strain distinction from our sampling are questionable. Our results suggest that isolates of the C-population represent a different strain from the SCYLV superpopulation, but resolution of this issue will require either increased sampling or the addition of pertinent biological data.

The use of phylogenies based on molecular data is a commonly accepted approach to defining population structures (16, 19, 24, 37–41), and the advantages of using phylogenies in this fashion have been described (14, 40). Migration of RNA viruses has been shown to fix classes of genomes that may exhibit variable degrees of fitness (11). In terms of RNA viruses, most migrations involve either vector transmissions or dispersal of small samples of infected organisms or virus-carrying vectors to new geographic locales (8–11). The lineage of these viruses may be traced by various means, including the identification of dispersed recombinant virus forms (18, 21). In the Luteoviridae, evidence exists that RNA recombination has occurred episodically at two levels. The first-level episodic RNA recombination event(s) accommodates the divergence of the two major classes of Luteoviridae genomes that are currently represented by the Luteovirus and Polerovirus genera: Luteovirus members produce an RNA-dependent RNA polymerase protein, lack a VPg protein, and have a 5' gene organization most closely allied with members of the Sobemovirus genus, whereas Polerovirus members express a VPg, as well as an RdRp most closely allied with the RdRp proteins expressed by members of the Dianthovirus or Carmovirus genus (4, 7, 31, 32, 35, 36, 43). At a secondary level, evidence indicates that episodic recombination between polerovirus-polerovirus (17) and luteovirus-polerovirus ancestors (36, 43, 48) has occurred since the divergence of these two groups (4, 7, 31, 32, 35).

Overall, the phylogenetic analysis-based data we present in this study indicates that a majority of the isolates we studied are associated with a superpopulation structure of SCYLV. The results also indicate that the C-population of SCYLV represents a population structure from which the SCYLV superpopulation most likely emerged. Two primary hypotheses might be considered to explain the phylogenetic relationship between the SCYLV populations we refer to as the superpopulation and the C-population: (i) the superpopulation was derived from a separable and distinct C-population via either founder effects or genetic drift or (ii) the C-population was derived from the superpopulation via recombination with some unknown Luteoviridae family member that exists at a closer genetic distance to the main Polerovirus clade, to which SCYLV is most closely allied. The evidence primarily supports the first conclusion. First, our attempts to use different methods of computational technology, i.e., to determine whether the SPV we observed is derived from recombinatorial events, indicated that there was no significant evidence for detectable recombination events with the isolates sampled and analyzed. Second, a comparison of the translated ORF 2 (RdRp) gene products from the C-population in relation to the sampled nucleotide sequences indicated that the majority of those sequence changes in the C-population are silent in reference to the same homologous sites in the SCYLV superpopulation. Third, our current sampling methods indicate that the isolates that constitute the C-population have higher estimates of gene diversity than the isolates from the superpopulation, which would be a predictable observation if founder effects were involved in a hypothesized evolution of the SCYLV superpopulation from a C-population member or progenitor. The mean Nei (37) gene diversity estimates derived from fingerprint data are h = 0.0248 for the superpopulation set and h = 0.0547 for the C-population set for all of the CAPS markers. The range of genetic distances measured from the nucleotide data also supports this conclusion.

Based upon the SPV shown in Fig. 4, the dendrogram in Fig. 1A, and the Splitstree network in Fig. 3, the B0 and B2 isolates could be assigned in a classification scheme as a population subgroup, group B, which along with the B1 and N6 isolates represents one end of a derived continuum of sequence evolution within the SCYLV superpopulation structure. Based primarily on the SPV shown in Fig. 4, the T6, F1, F2, L1, G2, G3, G6, G7, G8, and G9 isolates could be considered as group A, constituting the other end of a continuum of isolates in the superpopulation. This partitioning scheme may be represented in the Splitstree network in Fig. 3 by the placement of the B0 (in diamond) and F2, G3, and G6 to G9 isolates (in box). This group A-group B partitioning scheme should be considered artificial. The positioning of the L1/G2 sequences in the Splitstree network in Fig. 2 in a quadrangle adjoining the F1/T6 and N6/B1 isolates suggests that the SCYLV superpopulation members more accurately represent a continuum of sequence evolution. The value of a group A-group B partitioning scheme is in its merits as a guide toward developing molecular epidemiologic models for the geographic migration and dispersal of SCYLV. If a parsimonious philosophy (i.e., that the lowest number of steps leading to a result is the most likely path taken to arrive at that result) is applied to this somewhat artificial group classification, a progression of SCYLV evolution in which C-population isolates have given rise to group B-type isolates, from which group A-type isolates are further evolved, is the simplest conclusion. If we utilize the group A-group B superpopulation and C-population forms of classification, our survey would indicate that group A isolates are distributed between North America and Guatemala, that group B isolates are distributed between Argentina and Brazil, and that the C-population isolates have only been identified from Colombia. None of our geographic sites yielded genotypes assignable to more than one grouping from this somewhat artificial form of classification. Although these observations may be due to a lack of extensive enough sampling, they could also be the result of founder effect involvement in the intraspecies evolution of SCYLV.

In this study, we have examined viral isolate relationships in part with network diagrams generated by Splitstree (22). The diagrams generated by an analysis with Splitstree utilize a form of split decomposition analysis (1, 13) that is thought to provide a more reliable interpretation of phylogenetic data, because evolutionary data often contain conflicting phylogenetic data from which alternate and also highly likely phylogenetic relationships might not otherwise be illustrated (12, 13, 22, 27, 28, 33). For example, in the network diagram in Fig. 2, a series of quadrangles are linked in relation to an observed continuum of sequence evolution of a superpopulation of SCYLV that extends from the T6 isolate to the N6 isolate. Three main quadrangles are represented, of which the T6 isolate is positioned from one quadrangle which is opposed by a B1/N6 associated quadrangle, and in which a third quadrangle may be joined to these two, in which the G2/F1/L1 isolates are associated. Figure 2 also shows that although the C1 isolate may be placed within a quadrangle node that leads to the central quadrangles of the superpopulation structure, effectively placed at what is presumed to be the shortest genetic distance and therefore most likely intersection with this SCYLV superpopulation structure, a second and also highly likely alternative leads to the T6 associated quadrangle. This is in contrast to the typical type of phylogram, such as that in Fig. 1B. Utilizing the same alignment, but a different phylogenetic model, the L1 and G2 isolates might be interpreted from Fig. 1B as being more closely allied with the B1 and N6 isolates, whereas the network diagram in Fig. 2 suggests that when other highly likely alternatives are considered, this relationship is more ambiguous. When SCYLV isolates are analyzed with alignments generated with other members of the family Luteoviridae, as shown in Fig. 3, a similar relationship between isolates may be interpreted. In both Fig. 2 and Fig. 3, the C1 isolate is positioned at a node within a quadrangle from which the C3 and C4 isolates are generated. Based upon these data, an assumption that any one of the three Columbian isolates could represent the closest C-population founder representative to the SCYLV superpopulation is also a reasonable interpretation.

The implications of our observations of SCYLV intraspecies SPV. Our analysis of SPV within North, South, and Central American isolates of SCYLV has been based in part on data derived from an intrafamilial model of SPV within the Luteoviridae (36). A direct extrapolation of data from this intrafamilial model to the SCYLV-A genome is shown in Fig. 4A. Based on this intrafamilial model, primers for the PROREP, REPUTR, and CPRT amplicons were selected from corresponding regions of the SCYLV genome, which yielded amplicons by annealing to sites with a low capacity for sequence evolution and yet produced amplicons that could each be predicted to have different degrees of sequence diversity. Among the predictions that could be made from our original intrafamilial analysis (36) are that the CPRT-spanning region of the SCYLV genome should exhibit a low sequence diversity, the PROREP-spanning regions should exhibit moderate diversity, and the REPUTR-spanning region should exhibit a higher diversity than the other two regions. As shown by our intraspecies analysis of SPV shown in Fig. 4A, the predictability of this intrafamilial analysis-based model of SPV is supported by the SPV that we detect from the SCYLV field isolates we have studied. The CPRT region exhibits low diversity, the PROREP region shows moderate diversity, and the REPUTR region exhibits high diversity. SPV models of this type have utilitarian value. They may be useful as criteria involved in the production of virus-resistant plants, utilizing transgenic methodologies that employ homology-dependent posttranscriptional gene-silencing mechanisms as a primary component. Because this method of plant protection is reliant upon transgene sequence homology with the targeted viral genome (2), the capacity for viral sequence evolution in the region represented by a transgene is a significant factor in developing ecologically safe and economically viable long-term virus resistance (34, 42, 49). For example, sugarcane plants expressing untranslated viral capsid sequences of Sorghum mosaic virus (SrMV) strain SCH (SrMV-SCH), challenged with SrMV viruses of strains SCM (SrMV-SCM) and SCI (SrMV-SCI) and Sugarcane mosaic virus strain D (SCMV-D), show various levels of virus resistance that correlated with the percentage of sequence identity of the transgenes to the sequence of the challenging virus (23). The corresponding homologous sequences of SrMV-SCM, SrMV-SCI, and SCMV-D, respectively, have 95, 95, and 75% identity with the equivalent SrMV-SCH sequence range (56). Challenge experiments with these same viruses protected, respectively, 17, 18, and 3 of 25 sugarcane plants compared to 22 of 25 plants protected by the SrMV-SCH virus-challenged plants (23). In many instances, sequence diversity data for a virus species is unavailable, but genomic information from related viruses is available. In these instances, an intrafamilial model could be constructed to predict genomic sequence regions expected to have low diversity, which would be more appropriate choices to use as transgene sequences.

 
We thank James Miller (USDA-ARS Sugarcane Breeding Unit, Canal Pt., Fla.) for providing the F1 isolate source material used by Smith et al. (48) and James Irvine (Sugarcane Breeding Unit, Texas A&M Agricultural Experiment Station) for assistance in isolate collections during his visits to South and Central American locales and for helpful discussions on sugarcane biology and yellow leaf syndrome. We thank Tom Sappington (USDA-ARS Subtropical Agricultural Research Center, Weslaco, Tex.) for helpful discussions on population genetic analyses with molecular data. We also thank Grant Smith (Bureau of Sugar Experiment Stations, Indooropilly, Queensland, Australia) and Mike Irey (U.S. Sugar Corp., Clewiston, Fla.) for helpful discussions on SCYLV and YLS.

This work was supported by grant 99-15 from the International Consortium of Sugarcane Biotechnologists.

 
* Corresponding author. Mailing address: Department of Plant Pathology and Microbiology, Texas A&M University System Agricultural Experiment Station, 2415 East Highway 83, Weslaco, TX 78596. Phone: (956) 968-5585. Fax: (956) 969-5260. E-mail: e-mirkov{at}tamu.edu.

Top Abstract Introduction Materials and Methods Results Discussion References

 

1
1. Bandelt, H. J., and A. W. M. Dress. 1992. Split decomposition: a new and useful approach to phylogenetic analysis of distance data. Mol. Phylogenet. Evol. 1:242–252.[CrossRef][Medline]
2
2. Baulcombe, D. C. 1996. Mechanisms of pathogen derived resistance to viruses in transgenic plants. Plant Cell 8:1833–1844.[CrossRef][Medline]
3
3. Borth, W., J. S. Hu, and S. Schenk. 1994. Double-stranded RNA associated with sugarcane yellow leaf syndrome. Sugar Cane 3:5–8.
4
4. Chaloub, B. A., and H. D. Lapierre. 1996. Importance des recombinaisons ARN dans l’evolution des luteovirus. Agronomie 15:393–400.
5
5. Comstock, J. C., M. S. Irey, B. E. L. Lockhart, and Z. K. Wang. 1998. Incidence of yellow leaf syndrome in CP cultivars based on polymerase chain reaction and serological techniques. Sugar Cane 4:21–24.
6
6. Cronje, C. P. R., A. M. Tymon, P. Jones, and R. A. Bailey. 1998. Association of a phytoplasma with yellow leaf syndrome of sugarcane in South Africa. Ann. Appl. Biol. 133:177–186.
7
7. D’Arcy, C. J., and M. Mayo. 1996. Proposals for changes in luteovirus taxonomy and nomenclature. J. Gen. Virol. 142:1285–1287.
8
8. Domingo, E., C. Escarmis, N. Sevilla, A. Moya, F. Santiago, J. Q. Elena, I. S. Novella, and J. J. Holland. 1996. Basic concepts in RNA virus evolution. FASEB J. 10:859–864.[Abstract]
9
9. Domingo, E., and J. J. Holland. 1994. Mutation rates and rapid evolution of RNA viruses, p.161–189. In S. S. Morse (ed.), The evolutionary biology of viruses. Raven Press, New York, N.Y.
10
10. Domingo, E., J. J. Holland, C. Briebricher, and M. Eigen. 1995. Quasispecies: the concept and the word, p.181–191. In A. J. Gibbs, C. H. Calisher, and F. Garcia-Arenal (ed.), Molecular basis of viral evolution. Cambridge University Press, Cambridge, England.
11
11. Domingo, E., and J. J. Holland. 1997. RNA virus mutations and fitness for survival. Annu. Rev. Microbiol. 51:151–178.[CrossRef][Medline]
12
12. Dopazo, J., A. W. M. Dress, and A. von Haeseler. 1993. Split decomposition: a new technique to analyze viral evolution. Proc. Natl. Acad. Sci. USA 90:10320–10324.[Abstract/Free Full Text]
13
13. Dress, A. W. M., D. H. Huson, and V. Moulton. 1996. Analyzing and visualizing sequence and distance data using Splitstree. Discrete Appl. Math. 71:95–109.[CrossRef]
14
14. Feselstein, J. 1992. Estimating effective population size from samples of sequences: inefficiency of pairwise and segregating sites as compared to phylogenetic estimates. Genet. Res. 59:139–147.[Medline]
15
15. Feselstein, J. 1989. PHYLIP: phylogeny interference package. Cladistics 5:164–166.
16
16. Gibbs, A. 1999. Evolution and origins of tobamoviruses. Proc. Philos. Trans. Soc. Lond. B 354:517–685.
17
17. Gibbs, M. J., and J. I. Cooper. 1995. A recombinational event in the history of luteoviruses, probably induced by base-pairing between two distinct viruses. Virology 206:1129–1132.[CrossRef][Medline]
18
18. Grassly, N. C., and E. C. Holmes. 1997. A likelihood method for the detection of selection and recombination using sequence data. Mol. Biol. Evol. 14:239–247.[Abstract]
19
19. Hartl, D. L., and A. G. Clark. 1989. Principles of population genetics. Sinauer Associates, Sunderland, Mass.
20
20. Hasegawa, M., H. Kishino, and T. Yano. 1985. Dating the human-ape splitting by a molecular clock of mitochondrial DNA. J. Mol. Evol. 22:160–174.[CrossRef][Medline]
21
21. Holmes, E. C., M. Worobey, and A. Rambaut. 1999. Phylogenetic evidence for recombination in dengue virus. Mol. Biol. Evol. 16:405–409.[Abstract]
22
22. Huson, D. H. 1998. Splitstree: analyzing and visualizing evolutionary data. Bioinformatics 14:68–73.[Abstract/Free Full Text]
23
23. Ingelbrecht, I., J. E. Irvine, and T. E. Mirkov. 1999. Posttranscriptional gene silencing in transgenic sugarcane. Dissection of homology-dependent virus resistance in a monocot that has a complex polyploid genome. Plant Physiol. 119:1187–1197.[Abstract/Free Full Text]
24
24. Kimura, M. 1983. The neutral theory of molecular evolution. Cambridge University Press, Cambridge, England.
25
25. Koonin, E. V., and V. V. Dolja. 1993. Evolution and taxonomy of positive-strand RNA viruses: implications of comparative analysis of amino acid sequences. Crit. Rev. Biochem. 28:375–430.
26
26. Kumar, S., K. Tamura, and M. Nei. 1993. MEGA: molecular evolutionary genetic analysis. Pennsylvania State University Press, University Park.
27
27. Lockhart, P. J., M. A. Steel, M. D. Hendy, and D. P. Penny. 1994. Recovering an evolutionary tree under a more realistic model of sequence evolution. Mol. Biol. Evol. 11:605–612.
28
28. Lockhart, P. J., D. Penny, and A. Meyer. 1995. Testing the phylogeny of swordtail fishes using split decomposition and spectral analysis. Mol. Evol. 41:666–674.
29
29. Maia, I. G., M. C. Goncalves, P. Arruda, and J. Vega. 2000. Molecular evidence that Sugarcane yellow leaf virus (ScYLV) is a member of the Luteoviridae. Arch. Virol. 145:1009–1019.[CrossRef][Medline]
30
30. Martin, D. P., and E. P. Rybicki. 2000. RDP: detection of recombination amongst aligned sequences. Bioinformatics 16:562–563.[Abstract/Free Full Text]
31
31. Martin, R. R., P. K. Keese, M. J. Young, P. M. Waterhouse, and W. L. Gerlach. 1990. Evolution and molecular biology of luteoviruses. Annu. Rev. Phytopathol. 28:341–363.
32
32. Mayo, M. A., and V. Ziegler-Graff. 1996. Molecular biology of luteoviruses. Adv. Virus Res. 46:416–468.
33
33. McLenachan, P. A., P. J. Lockhart, H. R. Faber, and B. C. Mansfield. 1996. Evolutionary analysis of the multigene pregnancy specific ß1-glycoprotein family: separation of historical and nonhistorical signals. J. Mol. Evol. 42:273–280.[CrossRef][Medline]
34
34. Miller, W. A., G. Koev, and B. R. Mohan. 1997. Are there risks associated with transgenic resistance to luteoviruses? Plant Dis. 81:700–710.
35
35. Miller, W. A., S. P. Dinesh-Kumar, and C. P. Paul. 1995. Luteovirus gene expression. Crit. Rev. Plant Sci. 14:179–211.
36
36. Moonan, F., J. Molina, and T. E. Mirkov. 2000. Sugarcane yellow leaf virus: an emerging virus that has evolved by recombination between luteoviral and poleroviral ancestors. Virology 269:156–171.[CrossRef][Medline]
37
37. Nei, M. 1973. Analysis of gene diversity in subdivided populations. Proc. Natl. Acad. Sci. USA 70:3321–3323.[Abstract/Free Full Text]
38
38. Nei, M. 1978. Estimation of average heterozygosity and genetic distance from a small number of individuals. Genetics 89:583–590.[Abstract/Free Full Text]
39
39. Nei, M. 1972. Genetic distance between populations. Am. Nat. 106:283–292.[CrossRef]
40
40. Nei, M., and S. Kumar. 2000. Molecular evolution and phylogenetics. Oxford University Press, Oxford, England.
41
41. Page, D. M., and E. C. Holmes. 1998. Molecular evolution: a phylogenetic approach. Blackwell Science, Cambridge, England.
42
42. Patron, N. J., M. A. Mayo, H. Barker, M. S. Liney, and H. G. Smith. 1998. Assessments of risk associated with growing virus resistant transgenic sugarbeet. Aspects Appl. Biol. 52:287–293.
43
43. Rathjen, J. P., L. E. Karageorgos, N. Habili, P. M. Waterhouse, and R. H. Symons. 1994. Soybean dwarf luteovirus contains the third variant genome type in the luteovirus group. Virology 198:671–679.[CrossRef][Medline]
44
44. Sambrook, J., E. F. Fritsch, and T. Maniatis. 1989. Molecular cloning: a laboratory manual, 2nd ed. Cold Spring Harbor Laboratory Press, Cold Spring Harbor, N.Y.
45
45. Scagliusi, S. M., and B. E. L. Lockhart. 2000. Transmission, characterization, and serology of a Luteovirus associated with yellow leaf syndrome of sugarcane. Phytopathology 90:120–124.[Medline]
46
46. Schenck, S., and A. T. Lehrer. 2000. Factors affecting the transmission and spread of Sugarcane yellow leaf virus. Plant Dis. 84:1085–1088.
47
47. Schenck, S., J. S. Hu, and B. E. Lockhart. L. 1997. Use of a tissue immunoassay to determine the distribution of sugarcane yellow leaf virus in Hawaii. Sugar Cane 4:5–8.
48
48. Smith, G. R., Z. Borg, B. E. L. Lockhart, K. S. Braithwaite, and M. J. Gibbs. 2000. Sugarcane yellow leaf virus: a novel member of the Luteoviridae that probably arose by interspecies recombination. J. Gen. Virol. 81:1865–1869.[Abstract/Free Full Text]
49
49. Smith, H., and N. Patron. 2000. Risk assessment of virus resistant transgenic sugar beet. Sugar Beet Rev. 68:47–50.
50
50. Strimmer, K., and A. von Haesler. 1996. Quartet puzzling: a quartet maximum likelihood method for reconstructing tree topologies. Mol. Biol. Evol. 13:964–969.
51
51. Swofford, D. L. 1999. PAUP, version 4.04b. Sinauer Associates, Sunderland, Mass.
52
52. Thompson, J. D., D. G. Higgins, and T. J. Gibson. 1994. CLUSTAL W: improving the sensitivity for progressive multiple sequence alignment through sequence weighting, position-specific gap penalties, and weight matrix choice. Nucleic Acids Res. 22:673–680.
53
53. Thompson, J. D., T. J. Gibson, F. Plewniak, F. Jeanmougin, and D. G. Higgins. 1997. The ClustalX windows interface: flexible strategies for multiple sequence alignment aided by quality analysis tools. Nucleic Acids Res. 24:4876–4882.
54
54. van Regenmortel, M. H. V., C. M. Fauquet, D. H. L. Bishop, E. B. Carstens, M. K. Estes, S. M. Lemon, J. Maniloff, M. A. Mayo, D. J. McGeoch, C. R. Pringle, and R. B. Wickner. 2000. Virus taxonomy. Seventh report of the International Committee on Taxonomy of Viruses. Academic Press, New York, N.Y.
55
55. Vega, J., S. M. Scagliusi, and E. C. Ulian. 1997. Sugarcane yellow leaf disease in Brazil: evidence for associating with a luteovirus. Plant Dis. 81:21–26.
56
56. Yang, Z. N., and T. E. Mirkov. 1997. Sequence and relationships of Sugarcane mosaic virus and Sorghum mosaic virus strains and development of RT-PCR-based RFLPs for strain discrimination. Phytopathology 87:932–939.[Medline]
57
57. Yeh, F. C., and T. J. B. Boyle. 1997. POPGENE, the user-friendly shareware for population genetic analysis. Molecular Biology and Biotech Centre, University of Alberta, Edmonton, Alberta, Canada.

---

Journal of Virology, February 2002, p. 1339-1348, Vol. 76, No. 3
0022-538X/01/$04.00+0     DOI: 10.1128/JVI.76.3.1339-1348.2002
Copyright © 2002, American Society for Microbiology. All Rights Reserved.

This Article Abstract Full Text (PDF) Alert me when this article is cited Alert me if a correction is posted Services Similar articles in this journal Similar articles in PubMed Alert me to new issues of the journal Download to citation manager Reprints and Permissions Copyright Information Books from ASM Press MicrobeWorld Citing Articles Citing Articles via Google Scholar Google Scholar Articles by Moonan, F. Articles by Mirkov, T. E. Search for Related Content PubMed PubMed Citation Articles by Moonan, F. Articles by Mirkov, T. E.