| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
1 CIRAD, UMR 53 PVBMT CIRAD-Université de la Réunion, Pôle de Protection des Plantes, 7 Chemin de l'IRAT, 97410 Saint Pierre, La Réunion, France
2 Institute of Infectious Diseases and Molecular Medicine, University of Cape Town, Observatory 7925, South Africa
3 Electron Microscopy Unit, University of Cape Town, Rondebosch 7701, South Africa
4 Museum National d'Histoire Naturelle, Dept RDDM, USM 501, CNRS UMR 5166, Evolution des Régulations Endocriniennes, 57 rue Cuvier, CP 32, 75005 Paris, France
Correspondence
J.-M. Lett
lett{at}cirad.fr
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Supplementary material is available with the online version of this paper.
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
). Usually multiple begomovirus species have emerged simultaneously in a given region, with the ensuing species complexes causing diseases in a wide variety of plant species, including many of great agricultural importance.
The rate at which new species are emerging is perhaps best exemplified by the diversity of the almost 700 full begomovirus DNA-A sequences currently deposited in public sequence databases. Given the 89 % identity threshold of the International Committee on Taxonomy of Viruses (ICTV), these genomes represent more than 200 species (Fauquet et al., 2007).
While providing a major contribution to the richness of currently observed begomovirus species diversity, recombination continues both to fuel begomovirus diversification and complicate the classification of new species. The important contribution of recombination to geminivirus evolution is now well established (Umaharan et al., 1998; Padidam et al., 1999) and it is suspected that it is directly responsible for the emergence of many of the most agriculturally damaging begomovirus species complexes (Zhou et al., 1997; Monci et al., 2002; Garcia-Andres et al., 2006). Despite this, very little is actually known either about why recombination seems to contribute to the emergence of species complexes, or how recombinants with enhanced pathogenicity arise and proliferate. Furthermore, both the biochemical processes that determine the kinds of recombinant genomes produced, and the evolutionary processes that determine which of these survive, remain a complete mystery. However, some studies have indicated that recombination hot-spots may exist within begomovirus genomes (Stanley, 1995; Ndunguru et al., 2005; Fauquet et al., 2005; Garcia-Andres et al., 2007). Identifying the locations of any recombination hot- and cold-spots within begomovirus genomes sampled from nature would certainly be a valuable first step towards understanding the underlying processes controlling the generation and spread of recombinants within species complexes.
We decided to quantitatively evaluate the importance of recombination in the genetic diversification of begomoviruses within a newly discovered monopartite begomovirus species complex indigenous to the south-west Indian Ocean (SWIO) islands off the coast of Africa. Despite the pace at which begomovirus diversity has been explored in the past few years, very few full-length African begomovirus DNA-A sequences other than those of the African cassava mosaic disease (CMD) pathosystem are presently available. To increase the richness of the available African begomovirus genome sequence data, we therefore extended previous preliminary surveys of monopartite begomovirus species on the islands of Madagascar, Comoros and Seychelles archipelagos (Lefeuvre et al., 2007; Delatte et al., 2005b). We describe the molecular diversity and taxonomic relationships of 14 SWIO island begomovirus isolates, including seven new species, causing recent plant disease epidemics in the SWIO islands. Importantly, when analysed together with African and Mediterranean begomovirus sequences, we find solid statistical evidence of recombination hot- and cold-spots within the DNA-A components of these viruses. This result may indicate how and why recombination makes such a substantial contribution to begomovirus diversity in general.
|
METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
) and stored dried (Bos, 1977). Total DNA was extracted using a DNeasy Plant miniprep kit (Qiagen) according to the manufacturer's instructions.
|
), and VD360-CD1266 (Delatte et al., 2005b). PCR reactions were carried out as described in Delatte et al. (2005b). The presence/absence of a DNA-B genome component and DNA-
molecules were also assessed for each of the isolates using, respectively, the PCR primer sets PBL1v2040-PCRc1 (Rojas et al., 1993) and Beta 1-Beta 2 (Briddon et al., 2002).
Cloning strategies.
Circular viral DNA molecules were amplified using a TempliPhi kit (GE Healthcare) as described by Inoue-Nagata et al. (2004). Full genomes were cloned into the vector pBC-KS in the HindIII restriction site for AM701758
[GenBank]
, AM701759
[GenBank]
, AM701766
[GenBank]
, AM701767
[GenBank]
and AM491778
[GenBank]
and in the BamHI restriction site for all others. A complete DNA-A-like component for each isolate was sequenced by gene walking using the Macrogen sequencing service (Korea).
Phylogenetic analysis.
Full DNA-A-like sequences from 14 isolates (this study) were arranged so that the first nucleotide in the sequence corresponded to the first base (adenine) of virion strand replication (Laufs et al., 1995). Forty-one other full DNA-A and DNA-A-like sequences of related viruses were obtained from public sequence databases using TaxBrowser (http://www.ncbi.nlm.nih.gov/) on May 2006. Multiple sequence alignments were constructed using partial order graphs (POA) (Lee et al., 2002), the CLUSTAL W (Thompson et al., 1994) based subalignment tool available in MEGA 3.1 (Kumar et al., 2004) and manual editing.
The optimal model of sequence evolution defined by ModelTest (Posada, 2006) was used for phylogenetic reconstruction (GTR+I+G). The maximum-likelihood (ML) tree was determined from a preliminary neighbour-joining (NJ) analysis using PAUP* with the heuristic search algorithm. In addition to these analyses, we performed Bayesian phylogenetic reconstruction on the full dataset using MrBayes 3.1.2 (Ronquist & Huelsenbeck, 2003). Four runs with six Markov chains were conducted simultaneously for 1 000 000 generations starting from random initial trees, and sampled every 100 generations. Variation in the ML scores in this sample was examined graphically with Tracer (Rambaut & Drummond, 2004). The trees generated prior to stabilization of ML scores were discarded with the consensus phylogeny and posterior probability of their nodes being determined with a burn-in of 25 %. The method of Shimodaira & Hasegawa (1999) implemented in PAUP* was used to test whether the ML scores of the NJ, ML and Bayesian phylogenetic reconstructions fell within the same confidence limits.
Recombination analyses.
Detection of potential recombinant sequences, identification of likely parental sequences and localization of possible recombination breakpoints was carried out on a 178-sequence alignment (170 begomovirus, seven curtovirus and one topocuvirus sequences) using the RDP (Martin & Rybicki, 2000), GENECONV (Padidam et al., 1999), BOOTSCAN (Martin et al., 2005a), MAXIMUM CHI SQUARE (Smith, 1992), CHIMAERA (Martin et al., 2005b) and SISTER SCAN (Gibbs et al., 2000) recombination detection methods as implemented in RDP3 (Martin et al., 2005b), available from http://darwin.uvigo.es/rdp/rdp.html (see the RDP project file submitted as supplementary material for full details of program settings). The analysis was performed with default settings for the different detection methods and a Bonferroni corrected P-value cut-off of 0.05. The breakpoint positions and recombinant sequence(s) inferred for every detected potential recombination event were manually checked and adjusted where necessary using the extensive phylogenetic and recombination signal analysis features available in RDP3. Once a set of unique potential recombination events was identified, we compiled a breakpoint map by plotting the positions of all clearly identifiable breakpoints. A breakpoint density plot was then constructed from this map and the statistical significance of potential breakpoint hot- and cold-spots was tested as described in Heath et al. (2006). Briefly, the statistical analysis used takes the observed distribution of polymorphic sites in an alignment and randomly maps all the observed recombination events to this distribution, such that the real and randomly mapped events all involve exchanges of sequence tracts containing the same numbers of polymorphic sites. Doing this accounts for the fact that uneven distribution of polymorphic sites along the length of an alignment makes the identification of breakpoints in certain alignment regions more difficult than in others. This random mapping process is then repeated 1000 times and the actual distribution of breakpoints is compared to that of the 1000 permuted mappings using two tests. The first is a ‘global’ test which determines whether there are breakpoint clusters in the real distribution with more breakpoints than generally occur in the distributions determined from the permuted datasets. This analysis is highly conservative as it ignores the fact that it will be far harder to detect a genuinely significant breakpoint cluster in regions of conserved sequence than it will be to detect one in regions of more diverse sequence (as mentioned above, breakpoints are most easily and accurately detectable where diversity is high). Therefore, a second, less conservative, ‘local’ test compares corresponding portions of the real and permuted breakpoint distributions and determines whether local regions of the real distribution contain significantly more breakpoints than generally occur in corresponding regions of the permuted datasets. The P-values associated with both the global and local tests are simply the proportions of permuted datasets with greater breakpoint clusters. Whilst we judged P-values <0.05 to be significant for the conservative global test, to guard against false positives, we judged P-values <0.01 as being significant for the less conservative local test.
Species distinction analysis.
Sequence identity was computed from the precedent multiple sequence alignments without curtovirus and topocuvirus sequences (170 sequences) using the dna.dist function available in the R package, APE (Paradis et al., 2004). We identified genotypes belonging to different species using the ICTV-recommended 89 % complete DNA-A/DNA-A-like sequence identity threshold for species demarcation. To take into account possible influences of discovery order on species number estimates, we repeated the species identification operation 1000 times using the sequences in a random order. The mean and standard deviation of identified species numbers were calculated from the results of these permutations.
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
; Fig. 1). While PCR amplification and cloning of apparently full-length DNA-A-like components was possible from all symptomatic leaf samples, DNA-B and DNA- specific PCRs yielded no amplification products. This implied that the 14 viruses were most likely all monopartite, as it has been shown previously for four SWIO species with agroinfectious clones (Delatte et al., 2005b). These DNA-A-like sequences were all of typical monopartite begomovirus size, ranging from 2742 to 2781 nt. Most of the sequences had predicted genes typical of monopartite begomoviruses in terms of both size and position. However, for some sequences, deduced protein sequences contained potential translation errors, as indicated in Table 1. All sequences are available in GenBank/EMBL/DDJP database under the accession numbers given in Table 1.
|
), 7 of the 14 new sequences represent new species. Species names for these viruses, based on host plant and region of origin, are proposed in Table 1. The remaining seven sequences share >89 % identity with DNA-A-like sequences of previously described species such as Cotton leaf curl Gezira virus (CLCuGV), Tomato leaf curl Comoros virus (ToLCKMV), Tomato leaf curl Mayotte virus (ToLCYTV), Tomato leaf curl Madagascar virus (ToLCMGV) and Tobacco leaf curl Zimbabwe virus (TbLCZV). We assembled a dataset of 170 African, Mediterranean and SWIO begomovirus DNA-A-like sequences containing 51 ICTV-designated species and the seven tentative new species described here. We applied our species-counting algorithm test to this dataset and determined that, depending on the discovery order, 49 species (±1.3) should be defined as such. While this number is not substantially different from 51 classified by the ICTV, our dataset contains an additional seven sequences that are almost certainly legitimate novel species. We further attempted to determine which parts of the virus genomes mostly contribute to pairwise distance scores that are currently the primary taxonomic measure used for species demarcation. Using alignments of individual ORFs indicated that 32±1.1 (V1 ORF), 25±1.3 (V2 ORF), 43.5±1.1 (C1 ORF), 29.5±1.3 (C2 ORF), 29.5±1.1 (C3 ORF) and 43.5±1.3 (C4 ORF) groups of sequences might be defined if we consider only these subgenomes and a 89 % identity threshold. This clearly indicated that sequences of the C1 and C4 ORFs are the primary source of taxonomic signal in begomoviruses.
Phylogenetic analysis
Phylogenetic reconstruction was achieved under the sequence evolutionary model GTR+I+G. The SH-test performed on NJ, ML and Bayesian trees concluded that both the ML and Bayesian phylogenetic reconstructions were congruent and had the greatest likelihood. Most of the nodes of the Bayesian phylogenetic tree had probabilities values greater than or equal to 0.95, indicating that branches are relatively stable (Fig. 2).
|
). The SWIO isolates are found in three of these phylogroups (G1, G3 and G4). Viruses widely sampled from various host species (chayote, cotton, hollyhock, pepper, tobacco and tomato) throughout Africa are found in G1, which also contains four SWIO isolates [of which, Tomato leaf curl Diana virus (ToLCDiaV) and Tomato leaf curl Toliara virus (ToLCTolV) are new species; Table 1]. Phylogroup G2 contains Mediterranean tomato yellow leaf curl virus (TYLCV) isolates and closely related tomato infecting virus species, including African ToLCVs from Sudan and Mali. However, none of the currently described indigenous SWIO isolates fall into phylogroup G2. Phylogroup G3 contains East African cassava mosaic virus (EACMV) and other closely related species, such as South African cassava mosaic virus (ACMV), East African cassava mosaic Zanzibar virus (EACMZV) and East African cassava mosaic Kenya virus (EACMKV). G3 also contains ToLCMGV isolates from the west coast of Madagascar. Finally, the forth phylogroup, G4, contains ACMV, Tomato leaf curl Uganda virus-[Iganga] (ToLCUGV-[Iga]) and twelve SWIO begomovirus isolates including five new species: Tomato leaf curl Moheli virus (ToLCMohV), Tobacco leaf curl Comoros virus (TbLCKMV), Tomato leaf curl Seychelles virus (ToLSCV), Tomato leaf curl Anjouan virus (ToLCAnjV) and Tomato leaf curl Antsiranana virus (ToLCAntV). Importantly, G3 and G4 contain both monopartite and bipartite begomoviruses.
Analysis of recombination
We analysed evidence of recombination in a 178-sequence alignment containing 170 full-length SWIO, African and Mediterranean begomovirus DNA-A and DNA-A-like sequences, and eight curtovirus and one topocuvirus full genome sequences. It was apparent from this analysis that collectively the SWIO isolates bear detectable evidence of at least 22 past recombination events (Fig. 3). Only CLCuGV-Be[An:For:01] was not detectably recombinant. Among the recombination events that were detected, many were between different species: the TbLCKMV-[GC:Fou99:05] and TbLCKMV-[GC:Sim18:04] isolates have apparently obtained almost their entire CP ORF from a virus resembling TbLCZV-[ZW] (event ‘p’ in Fig. 3), whereas the rest of their genome resembles that of the tomato infecting virus ToLCAntV-[MG:Mia1:01] (AM701767
[GenBank]
). Another very striking recombination event was detected in the ToLCSCV-[Mah:VE77:04] sequence from the Seychelles archipelago (event ‘v’ in Fig. 3). We were surprised to find that part of the Rep ORF of this virus was apparently derived from a divergent begomovirus resembling Sweet potato leaf curl virus (SPLCGV). However, upon closer analysis, it is probably more feasible that both SPLCGV and ToLCSCV-[Mah:VE77:04] have obtained large portions of their C1 ORFs from a curtovirus-like source.
|
). This analysis revealed one large ‘globally’ significant recombination hot-spot (global P-values <0.05 across its length) and two smaller ‘locally’ significant hot-spots (local P-values <0.01). Whereas the large global hot-spot encompasses almost the entire intergenic region (IR) between the C1 start codon and approximately 50 nucleotides 5' of the V2 ORF start codon, the locally significant hot-spots occur at the V1–C3 interface and in the centre of the C1 ORF. In addition to these hot-spots, the analysis also revealed two locally significant recombination cold-spots (local P-value <0.01). These occurred in the V2 ORF and in the third quarter of the V1 ORF (Table 2).
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
) represent new species. Interestingly, the ten viral species described in the SWIO islands are distributed amongst three of the four major phylogenetic groups identified within the African/Mediterranean begomovirus cluster. The presence of SWIO monopartite begomoviruses in two of the groups containing the bipartite cassava mosaic begomoviruses (G3 and G4) supports the hypothesis that either there have been multiple DNA-B component losses to produce the three different African monopartite virus lineages, or there have been multiple acquisitions of DNA-B components to produce the bipartite virus lineages (Saunders et al., 2002; Mansoor et al., 2003).
Our attempts at provisional classification of the novel virus genotypes described in this study led us to examine begomovirus species demarcation criteria. Given the prevailing demarcation criteria, we determined that, amongst the 170 sequences examined, the 89 % ICTV begomovirus species demarcation criterion implies that only 49 of these should be classified as species. This analysis revealed something quite interesting about the begomovirus genomes examined. Whereas the genomes contained approximately 43.5±1.1 distinct rep genes, they contained far fewer different kinds of other genes (ranging from 25±1.3 distinct V2 ORFs to 32±1.1 distinct cp genes). There is the equivalent of 36 % more rep genes in circulation than all other SWIO/African/Mediterranean begomovirus genes. Our results clearly indicate that recombination is almost certainly the driving force behind this apparent proliferation of rep genes. This simple result clearly illustrates how recombination confounds the definition of useful taxonomic criteria (Seal et al., 2006; Fauquet et al., 2003).
The phylogenetic analysis performed in this study clearly demonstrates that the breadth of begomovirus diversity found on the SWIO islands is qualitatively similar to that identifiable across the entire African continent. Also, besides six isolates classified as belonging to the G1 and G3 groups, the island isolates are all most closely related to one another and only share a distant common ancestor with the mainland viruses. This probably indicates that the SWIO islands have, with the exception of infrequent transmission events from mainland Africa, been epidemiologically isolated for a long time.
Phylogeographically the results are also intriguing: for the G4 group, there is a well supported cluster of SWIO isolates with ACMV. There is evidence here that either (i) ACMV, a lonely outlier amongst the other African viruses, originated on the SWIO islands, or (ii) the SWIO isolates are an extant and thriving population of an ancestral lineage that, besides ACMV and ToLCUGV-[Iga], has largely disappeared on the African mainland. As with cassava, the plants sampled in this study are exotic introduced species. One would expect original (and possibly still the natural) hosts of these viruses to be indigenous uncultivated plants. Further studies should aim to characterize begomovirus diversity in these hosts.
Our recombination analysis clearly indicates the presence of breakpoint hot- and cold-spots within SWIO/African/Mediterranean begomovirus genomes. This indicates either that DNA breakage and repair do not occur randomly in begomoviruses or that, if breakpoints do occur randomly, selection has preferentially culled recombinants with breakpoints in certain positions while permitting the survival of recombinants with breakpoints in other positions. That all recombinants are not created equal has been clearly demonstrated with laboratory constructed geminivirus recombinants (Liu et al., 2001; Martin & Rybicki, 2002) and one would expect that many, if not most, natural recombinants would experience serious fitness deficits. There are in fact two well-supported explanations as to why recombinants are generally less fit than their parents. First, protein engineers have discovered that hybrid genes with bits of sequence from distantly related sources tend to encode proteins that do not fold properly – probably due to disruptions of co-evolved amino acid contacts within their structures (Voigt et al., 2002; Saraf & Maranas, 2003). Second, when genes are transferred wholesale into distantly related genetic backgrounds, they appear only to function well either when they do not interact with a lot of other genes, or when they are co-transferred with the other genes with which they do interact (Jain et al., 1999; Martin et al., 2005c; Escriu et al., 2007). It is therefore likely that, whereas the recombination hot-spots we have detected represent genomic regions where breakpoints are both biochemically permissive and highly ‘survivable’, the cold-spots represent regions where breakpoints are either particularly deleterious or are biochemically very unlikely to occur.
Although we detected a large number of unique recombination breakpoints (i.e. breakpoints that occurred during different recombination events) across the entire 3' portion of rep spanning the C4 ORF, this region is bounded by two recombination hot-spots. This pattern of recombination is almost certainly due to both a biochemical predisposition to recombination in these sequences, and a high tolerance for recombination in the proteins encoded in this region. Importantly, experimental analyses of recombination in geminiviruses (Schnippenkoetter et al., 2001; Stenger et al., 1991; Garcia-Andres et al., 2007) and the replicational release mechanisms put into practice during agroinoculation of geminiviruses, have indicated that the origin of virion strand replication is a biochemically predisposed recombination hot-spot. While there is a clear breakpoint distribution peak detected at the virion strand ori, the highest breakpoint distribution peak is 5' of the ori, close to the rep start codon. This region corresponds to the most variable region of begomovirus genomes. It is probable that at least part of the reason why so many breakpoints are detected here is that this is the genome region where breakpoints are easiest to detect. Nevertheless, the statistical test used to detect hot spots takes this increased variability into account and has still identified that there are an improbably large number of breakpoints in this region. We propose first, that the IR-wide breakpoint hot-spot is a consequence of recombinants with breakpoints outside of genes generally being fitter than those with breakpoints within genes. This possibility is supported by the fact that the V1–C3 interface, the only other genome region where breakpoints are possible outside of genes, is also a recombination hot-spot.
Importantly, there exists direct experimental support for our observation that the V1–C3 interface is a recombination hot-spot because recombination at this point does not incur a significant fitness cost. In experimental recombination in controlled mixed TYLCSV and TYLCV-Mld infection, the most prevalent (and hence probably the most fit) emergent recombinant had one breakpoint within 100 nucleotides of the V1–C3 ORF interface and another at precisely the virion strand ori (Garcia Andres et al., 2007). That this particular recombinant genotype is highly fit is further evidenced by its close resemblance to the widespread natural TYLCSV–TYLCV-Mld recombinant, Tomato yellow leaf curl Malaga virus (Monci et al., 2002). The problem remains, however, to explain the recombination hot-spot in the middle of the rep gene. Our second proposal is therefore that the N-terminal portion of Rep and any protein expressed from the C4 ORF are exceptionally tolerant of recombination, with the most tolerable breakpoint positions (i.e. those that disrupt Rep folding the least) occurring near the centre of the gene around the recombination hot-spot.
The presence of recombination cold-spots within the V2 ORF and the third quarter of the V1 ORF is consistent with our first proposal that recombination breakpoints within coding regions are generally more damaging than those outside of coding regions. However, the fact that the detectable breakpoint cold-spots are within the virion sense ORFs, whereas the greatest number of breakpoints are within the complementary sense ORFs, leads us to a third proposal: the uneven distribution of recombination breakpoints is possibly due, at least in part, to clashes between virion strand replication and gene transcription. Whereas replication and virion strand transcription proceed in the same direction and are therefore unlikely to interfere with one another, transcription of the complementary strand ORFs tends to disrupt replication forks moving in the opposite direction. Analysis of replicating begomoviral DNA intermediates has revealed a wide distribution of so-called heterogeneous length linear dsDNA forms (hDNA), possibly created during such clashes. The ends of these hDNA molecules tend to map most frequently to the V-ori and either the C2/C3 transcription promoter near the hot-spot we detected in the centre of rep, or the C2/C3 terminator near the hot-spot we detected at the V1–C3 ORF interface (Jeske et al., 2001). Completion of replication from displaced, partially replicated virion strands would then proceed via the recombination-dependent replication pathway (Preiss & Jeske, 2003), which in the presence of potential template DNAs with different sequences could result in detectable recombination events. Completion of replication would result in a recombinant virion strand with one breakpoint at the point where replication was initially disrupted and the other at the virion sense ori where replication was completed.
Novel environments, such as the new host species offered to begomoviruses by invasive polyphagous vector biotypes, are possibly the defining force driving begomovirus evolution worldwide. For example, introduction into Reunion of the polyphagous B. tabaci biotype B is believed to be responsible for severe TYLC disease epidemics on the island in the late 1990s (Peterschmitt et al., 1999; Delatte et al., 2005a). Spread of this biotype to other SWIO islands may (i) facilitate host switching into cultivated crops of uncharacterized begomoviruses that currently only infect weeds and (ii) induce an overlap of exotic TYLCV and indigenous begomovirus distributions. Given the propensity of begomoviruses to recombine, emergence of new recombinants with increased virulence and/or modified host ranges are to be expected. An emergent TYLCV–Tomato yellow leaf curl Sardinia virus (TYLCSV) recombinant lineage in Spain (Monci et al., 2002; Garcia-Andres et al., 2006) demonstrates that the probability of such an occurrence is high, especially as the genetic distance between TYLCV and the SWIO indigenous ToLCVs is similar to the distance between TYLCV and TYLCSV.
By highlighting the extraordinary diversity of begomoviruses on the SWIO islands, we have provided a detailed description of their phylogenetic and recombinant histories. The phylogenetic association between the monopartite SWIO isolates and both monopartite and bipartite mainland African isolates indicate that they are probably indigenous to the islands. The large number of unique recombination events that we have detected amongst the SWIO isolates and their nearest mainland relatives reiterates the pivotal role of this process in begomovirus evolution. It is, however, apparent from our breakpoint distribution analysis that purifying selection and/or varying biochemical predispositions to recombination in different parts of begomovirus genomes place substantial constraints on the degree of evolutionary innovation that is possible by recombination.
|
ACKNOWLEDGEMENTS |
|---|
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMETHODSRESULTSDISCUSSIONREFERENCES |
|---|
Briddon, R. W., Bull, S. E., Mansoor, S., Amin, I. & Markham, P. G. (2002). Universal primers for the PCR-mediated amplification of DNA beta: a molecule associated with some monopartite begomoviruses. Mol Biotechnol 20, 315–318.[CrossRef][Medline]
Delatte, H., Holota, H., Naze, F., Peterschmitt, M., Reynaud, B. & Lett, J. M. (2005a). The presence of both recombinant and nonrecombinant strains of Tomato yellow leaf curl virus on tomato in Reunion Island. Plant Pathol 54, 262[CrossRef]
Delatte, H., Martin, D. P., Naze, F., Golbach, R. W., Reynaud, B., Peterschmitt, M. & Lett, J. M. (2005b). South West Indian Ocean islands tomato begomovirus populations represent a new major monopartite begomovirus group. J Gen Virol 86, 1533–1542.
Escriu, F., Fraile, A. & Garcia-Arenal, F. (2007). Constraints to genetic exchange support gene coadaptation in a tripartite RNA virus. PLoS Pathog 3, e8[CrossRef][Medline]
Fauquet, C. M., Bisaro, D. M., Briddon, R. W., Brown, J. K., Harrison, B. D., Rybicki, E. P., Stenger, D. C. & Stanley, J. (2003). Revision of taxonomic criteria for species demarcation in the family Geminiviridae, and an updated list of begomovirus species. Arch Virol 148, 405–421.[CrossRef][Medline]
Fauquet, C. M., Sawyer, S., Idris, A. M. & Brown, J. K. (2005). Sequence analysis and classification of apparent recombinant begomoviruses infecting tomato in the Nile and Mediterranean Basins. Phytopathology 95, 549–555.[Medline]
Fauquet, C. M., Briddon, R. W., Brown, J. K., Moriones, E., Stanley, J., Zerbini, M. & Zhou, X. (2007). Geminivirus strain demarcation and nomenclature. Arch Virol in press
Garcia-Andres, S., Monci, F., Navas-Castillo, J. & Moriones, E. (2006). Begomovirus genetic diversity in the native plant reservoir Solanum nigrum: evidence for the presence of a new virus species of recombinant nature. Virology 350, 433–442.[CrossRef][Medline]
Garcia-Andres, S., Tomas, D. M., Sanchez-Campos, S., Navas-Castillo, J. & Moriones, E. (2007). Frequent occurrence of recombinants in mixed infections of tomato yellow leaf curl disease-associated begomoviruses. Virology 359, 302–312.[CrossRef][Medline]
Gibbs, M. J., Armstrong, J. S. & Gibbs, A. J. (2000). Sister-Scanning: a Monte Carlo procedure for assessing signals in recombinant sequences. Bioinformatics 16, 573–582.
Heath, L., van der Walt, E., Varsani, A. & Martin, D. P. (2006). Recombination patterns in aphthoviruses mirror those found in other picornaviruses. J Virol 80, 11827–11832.
Inoue-Nagata, A. K., Albuquerque, L. C., Rocha, W. B. & Nagata, T. (2004). A simple method for cloning the complete begomovirus genome using the bacteriophage 
29 DNA polymerase. J Virol Methods 116, 209–211.[CrossRef][Medline]
Jain, R., Rivera, M. C. & Lake, J. A. (1999). Horizontal gene transfer among genomes: the complexity hypothesis. Proc Natl Acad Sci U S A 96, 3801–3806.
Jeske, H., Lutgemeier, M. & Preiss, W. (2001). DNA forms indicate rolling circle and recombination-dependent replication of Abutilon mosaic virus. EMBO J 20, 6158–6167.[CrossRef][Medline]
Kumar, S., Tamura, K. & Nei, M. (2004). MEGA3: integrated software for Molecular Evolutionary Genetics Analysis and sequence alignment. Brief Bioinform 5, 150–163.
Laufs, J., Traut, W., Heyraud, F., Matzeit, V., Rogers, S. G., Schell, J. & Gronenborn, B. (1995). In vitro cleavage and joining at the viral origin of replication by the replication initiator protein of tomato yellow leaf curl virus. Proc Natl Acad Sci U S A 92, 3879–3883.
Lee, C., Grasso, C. & Sharlow, M. F. (2002). Multiple sequence alignment using partial order graphs. Bioinformatics 18, 452–464.
Lefeuvre, P., Delatte, H., Naze, F., Dogley, W., Reynaud, B. & Lett, J. M. (2007). A new tomato leaf curl virus from the Seychelles archipelago. Plant Pathol 56, 342
Liu, H., Lucy, A. P., Davies, J. W. & Boulton, M. I. (2001). A single amino acid change in the coat protein of Maize streak virus abolishes systemic infection, but not interaction with viral DNA or movement protein. Mol Plant Pathol 2, 223–228.[CrossRef]
Mansoor, S., Briddon, R. W., Zafar, Y. & Stanley, J. (2003). Geminivirus disease complexes: an emerging threat. Trends Plant Sci 8, 128–134.[CrossRef][Medline]
Martin, D. & Rybicki, E. (2000). RDP: detection of recombination amongst aligned sequences. Bioinformatics 16, 562–563.
Martin, D. P. & Rybicki, E. P. (2002). Investigation of Maize streak virus pathogenicity determinants using chimaeric genomes. Virology 300, 180–188.[CrossRef][Medline]
Martin, D. P., Posada, D., Crandall, K. A. & Williamson, C. (2005a). A modified bootscan algorithm for automated identification of recombinant sequences and recombination breakpoints. AIDS Res Hum Retroviruses 21, 98–102.[CrossRef][Medline]
Martin, D. P., Williamson, C. & Posada, D. (2005b). RDP2: recombination detection and analysis from sequence alignments. Bioinformatics 21, 260–262.
Martin, D. P., van der Walt, E., Posada, D. & Rybicki, E. P. (2005c). The evolutionary value of recombination is constrained by genome modularity. PLoS Genet 1, e51[CrossRef][Medline]
Monci, F., Sanchez-Campos, S., Navas-Castillo, J. & Moriones, E. (2002). A natural recombinant between the geminiviruses Tomato yellow leaf curl Sardinia virus and Tomato yellow leaf curl virus exhibits a novel pathogenic phenotype and is becoming prevalent in Spanish populations. Virology 303, 317–326.[CrossRef][Medline]
Ndunguru, J., Legg, J. P., Aveling, T. A., Thompson, G. & Fauquet, C. M. (2005). Molecular biodiversity of cassava begomoviruses in Tanzania: evolution of cassava geminiviruses in Africa and evidence for East Africa being a center of diversity of cassava geminiviruses. Virol J 2, 21[CrossRef][Medline]
Padidam, M., Sawyer, S. & Fauquet, C. M. (1999). Possible emergence of new geminiviruses by frequent recombination. Virology 265, 218–225.[CrossRef][Medline]
Paradis, E., Claude, J. & Strimmer, K. (2004). APE: analyses of phylogenetics and evolution in R language. Bioinformatics 20, 289–290.
Peterschmitt, M., Granier, M., Mekdoud, R., Dalmon, A., Gambin, O., Vayssières, J. F. & Reynaud, B. (1999). First report of tomato yellow leaf curl virus in Réunion Island. Plant Dis 83, 303
Posada, D. (2006). ModelTest Server: a web-based tool for the statistical selection of models of nucleotide substitution online. Nucleic Acids Res 34, W700–W703.
Preiss, W. & Jeske, H. (2003). Multitasking in replication is common among geminiviruses. J Virol 77, 2972–2980.
Rambaut, A. & Drummond, A. J. (2004). Tracer v1.3, Available from http://evolve.zoo.ox.ac.uk/software.html
Rojas, M. R., Gilbertson, R. L., Russel, D. R. & Maxwell, D. P. (1993). Use of degenerate primers in the polymerase chain reaction to detect whitefly-transmitted geminivirus. Plant Dis 77, 340–347.
Ronquist, F. & Huelsenbeck, J. P. (2003). MrBayes 3: Bayesian phylogenetic inference under mixed models. Bioinformatics 19, 1572–1574.
Rybicki, E. P. & Pietersen, G. (1999). Plant virus disease problems in the developing world. In Advances in Virus Research, vol. 53, pp. 127–178. Edited by K. Maramorosch, F. A. Murphy & A. J. Shatkin. San Diego, CA: Academic Press.
Saraf, M. C. & Maranas, C. D. (2003). Using a residue clash map to functionally characterize protein recombination hybrids. Protein Eng 16, 1025–1034.
Saunders, K., Salim, N., Mali, V. R., Malathi, V. G., Briddon, R., Markham, P. G. & Stanley, J. (2002). Characterisation of Sri Lankan cassava mosaic virus and Indian cassava mosaic virus: evidence for acquisition of a DNA B component by a monopartite begomovirus. Virology 293, 63–74.[CrossRef][Medline]
Schnippenkoetter, W. H., Martin, D. P., Willment, J. A. & Rybicki, E. P. (2001). Forced recombination between distinct strains of Maize streak virus. J Gen Virol 82, 3081–3090.
Seal, S. E., vandenBosch, F. & Jeger, M. J. (2006). Factors influencing begomovirus evolution and their increasing global significance: implications for sustainable control. Crit Rev Plant Sci 25, 23–46.[CrossRef]
Shimodaira, H. & Hasegawa, M. (1999). Multiple comparisons of log-likelihoods with applications to phylogenetic inference. Mol Biol Evol 16, 1114–1116.
Smith, J. M. (1992). Analyzing the mosaic structure of genes. J Mol Evol 34, 126–129.[Medline]
Stanley, J. (1995). Analysis of African cassava mosaic virus recombinants suggests strand nicking occurs within the conserved nonanucleotide motif during the initiation of rolling circle DNA replication. Virology 206, 707–712.[CrossRef][Medline]
Stenger, D. C., Revington, G. N., Stevenson, M. C. & Bisaro, D. M. (1991). Replicational release of geminivirus genomes from tandemly repeated copies: evidence for rolling-circle replication of a plant viral DNA. Proc Natl Acad Sci U S A 88, 8029–8033.
Thompson, J. D., Higgins, D. G. & Gibson, T. J. (1994). CLUSTAL W: improving the sensitivity of progressive multiple sequence alignment through sequence weighting, position-specific gap penalties and weight matrix choice. Nucleic Acids Res 22, 4673–4680.
Umaharan, P., Padidam, M., Phelps, R. H., Beachy, R. N. & Fauquet, C. (1998). Distribution and diversity of geminiviruses in Trinidad and Tobago. Phytopathology 88, 1262–1268.[Medline]
Voigt, C. A., Martinez, C., Wang, Z. G., Mayo, S. L. & Arnold, F. H. (2002). Protein building blocks preserved by recombination. Nat Struct Biol 9, 553–558.[Medline]
Wyatt, S. D. & Brown, J. K. (1996). Detection of subgroup III geminivirus isolates in leaf extracts by degenerate primers and polymerase chain reaction. Phytopathology 86, 1288–1293.[CrossRef]
Zhou, X., Liu, Y., Calvert, L., Munoz, C., Otim-Nape, G. W., Robinson, D. J. & Harrison, B. D. (1997). Evidence that DNA-A of a geminivirus associated with severe cassava mosaic disease in Uganda has arisen by interspecific recombination. J Gen Virol 78, 2101–2111.[Abstract]
Received 20 June 2007;
accepted 9 August 2007.
This article has been cited by other articles:
|
|
 |
|
  T. Kon, M. R. Rojas, I. K. Abdourhamane, and R. L. Gilbertson Roles and interactions of begomoviruses and satellite DNAs associated with okra leaf curl disease in Mali, West Africa J. Gen. Virol., April 1, 2009; 90(4): 1001 - 1013. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 P. Lefeuvre, J.-M. Lett, A. Varsani, and D. P. Martin Widely Conserved Recombination Patterns among Single-Stranded DNA Viruses J. Virol., March 15, 2009; 83(6): 2697 - 2707. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. van der Walt, E. P. Rybicki, A. Varsani, J. E. Polston, R. Billharz, L. Donaldson, A. L. Monjane, and D. P. Martin Rapid host adaptation by extensive recombination J. Gen. Virol., March 1, 2009; 90(3): 734 - 746. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. Varsani, D. N. Shepherd, A. L. Monjane, B. E. Owor, J. B. Erdmann, E. P. Rybicki, M. Peterschmitt, R. W. Briddon, P. G. Markham, S. Oluwafemi, et al. Recombination, decreased host specificity and increased mobility may have driven the emergence of maize streak virus as an agricultural pathogen J. Gen. Virol., September 1, 2008; 89(9): 2063 - 2074. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
F. M. Codoner and S. F. Elena The promiscuous evolutionary history of the family Bromoviridae J. Gen. Virol., July 1, 2008; 89(7): 1739 - 1747. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| INT J SYST EVOL MICROBIOL | MICROBIOLOGY | J GEN VIROL |
| J MED MICROBIOL | ALL SGM JOURNALS | |
0.01. New virus isolates presented in this study are in bold.