[Home] [ICTV Taxonomy - Index of Viruses] [Virus Descriptions] [Character List] [Picture Gallery] [Tutorial] [Online Data Retrieval & Identification] [Virus Isolate Registration & Submission] [Search]

New World Begomovirus genome Project

Background

Viruses infect every kind of living organism, including higher plants. In this project we propose to undertake the first extensive inventory of the biological diversity of begmoviruses that infect uncultivated Eudicots in the New World (Western Hemisphere). Begomoviruses are one of four genera in the Family, Geminiviridae, which is one of fourteen taxonomic families of plant viruses. This family is unique among plant viruses in having small, circular, positive-sense, single-stranded DNA (ssDNA) genomes encapsidated in paired (geminate) icosahedral particles. New World begomoviruses all have a bipartite genome arrangement (two chromosomes), whereas, both bipartite and monopartite (one chromosome) viruses occur in the Eastern Hemisphere (Old World). (Harrison, 1985; Lazarowitz, 1992). Begomoviruses infect a wide range of dicotyledonous plants, namely, Eudicots, in subtropical/mild temperate and Mediterranean climates, worldwide (Brown, 1990; 1994; 2001; Brown and Bird, 1992; Paximadis et. al., 1999;Torres-Pacheco et. al., 1996). All members of this virus genus are transmitted by a single species of whitefly Bemisia tabaci (Genn). (Bird, 1957; Bird & Maramorosch, 1978; Brown, 2001). The distribution of the whitefly Bemisia tabaci (Genn.) (Alyerodidae; Hemipteran/Homopteran) vector (Brown and Czosnek, 2002), which colonizes Eudicot species, parallels that of begomoviruses, however, the host range of this cryptic species, viewed by most as a species complex (Brown et al., 1995), is far more broad, attaining 500 species or more for the most polyphagous haplotypes (Bird & Sanchez, 1971; Cock 1986; 1993; Gill, 1992a,b; Martin, 1987; 2003). It is noteworthy that the majority of extant whiteflies (Aleyrodidae), all which have co-evolved with Eudicots and are vascular feeders, are specialized on a single or a very few woody Eudicot species. In contrast, B. tabaci, the sole arthropod begomovirus vector, is instead highly adapted to herbaceous Eudicots and a few woody species (Gill, 1992b; Martin, 2003). B. tabaci ancestors are thought to possibly have existed since pre-Gondwanaland, resulting in its present-day 'Old' and 'New World' distribution (Gill, 1992a).

Interestingly, begomoviral genomes contain remnants of prokaryotic ancestry, while they are clearly adept at encoding and utilizing genes that function in eukaryotic plant and arthropod systems, namely Eudicots and their whitefly vector, respectively. Among the extant relatives of begomoviruses are the small ssDNA Proteobacteria-infecting Microviridae, or bacterial phages. Viruses in both families have a propensity to recombine with their co-infecting counterparts. Indeed, there is also evidence of possible genetic recombination with certain host plant genomes. For example, an ancient genetic association with several solanaceous species is apparent by the apparent integration of multiple begomoviral repeated sequences into the genome of tobacco and tomato, respectively (Bejarno et. al., 1996; Idris and Brown, submitted; Paximadis et al., 1999). Their cross-kingdom versatility is further underscored by the Begomovirus origin of replication (ori) and viral replication strategy, which are similar to those of a number of ssDNA phages (Frischmuth et. al., 1990). Within the begomoviral ori is a stem loop structure that contains the sequence TAATATT*AC, at which the circular genome is nicked (T*A) to initiate rolling circle replication (Fontes et al. 1994; Hanley-Bowdoin, 1999). This sequence is also conserved across all four genera within the Geminiviridae, and certain other ssDNA viruses of bacteria and plants. In contrast, begomoviral coding regions and their promoters typically have eukaryotic features (Bisaro, 1996; Ruiz-Medrano et. al., 1999), however, introns are not present in viral transcripts.

The mechanisms of diversification employed by begomoviruses have only recently been elucidated, as a growing number of begomoviral sequences have been determined. Besides mutation due to DNA polymerase error and minimal DNA repair, begomoviruses diversify using two other mechanisms: intermolecular recombination (Bisaro, 1994; Idris and Brown, 2002; Navas-Castillo et al., 2000; Padidam et. al., 1999; Zhou et al., 1997) and chromosomal reassortment (Brown et al., 2002; Faria et al., 1994; Gilbertson et al, 1993; Hill et al., 1999; Hou et al., 1996; 1998; Pita et al., 2001; Sanz et al., 2000), both of which require infection of the plant host by two or more viruses. That there is evidence for intermolecular recombination between members of three Geminiviridae genera, and between begomoviral species and strains, indicates that geminiviruses do not 'cross-protect' against close or distant relatives. This is in great contrast to RNA plant viruses.

Why this Taxon? Begomoviruses constitute a particularly noteworthy study system in the Biodiversity Surveys and Inventories Program because they are (i) novel, emerging plant viral pathogens that widely infect Eudicots of subtropical and adjacent habitats of varying species richness,(ii) are of interest as a model ssDNA plant virus genus that thereby lends itself to the study of gene function and regulation in plants, owing in part to their ability to redirect the plant cell cycle into the replication phase (Ach et al., 1997), and (iii) do not silence other competing begomoviral genomes present in the same host, and therefore (iv) utilize mixed infections as an opportunity for vertical exchange of genetic information, a dynamic process that has likely yielded thousands of genotypes in possibly hundreds of plant hosts, worldwide (Harrison, 1999; Liu et. al., 1998). (v) In the Americas, begomoviruses utilize as hosts a rich array of native and introduced Eudicot species, collectively, the most abundant plants on earth. (vi) These viruses have evolved a tight interaction with organs of their whitefly vector, which facilitates plant-to-plant dispersal. And, (vii) Begomovirus transmission is further dependent on a 70S heat shock protein abundant in whitefly hemolymph, which is synthesized by the prokaryotic endosymbiont that is essential for meeting the nutritional requirements of the whitefly vector (Brown & Czosnek, 2002). (viii) Finally, because of the small size of Begomovirus genomes, a phylogenetically informative genome sequence can be obtained relatively easily and inexpensively, compared to substantial investments required to sequence most other microbial genomes, including large and/or multipartite RNA viruses, which are not readily amenable to amplification of the entire genome or to HTPS.

The overarching goal of this project is to conduct the first systematic biodiversity inventory of Neotropical begomoviruses associated with Eudicot plant hosts in habitats of varied species richness. This international, cross-disciplinary effort is the first of its kind, and involves virologists, systematists, and molecular biologists, and will contribute substantially to the populating of international plant and plant virus databases and allied linkages.

The project aims are to: (i) increase overall knowledge of the extent of biological diversity in begomoviruses that infect uncultivated plant species, and (ii) engage in the discovery of new viral species and strain.

The specific objectives are: (i) Collect, archive, and inventory the distribution of NW begomoviruses that infect uncultivated Eudicot species in habitats of varying species richness, taking into account extant biodiversity data for viruses and plants, together with differences in climate/weather scenarios, and geographical and geological factors; (ii) Determine the genome sequence for all unique Begomovirus species and strains; (iii) Mine the viral sequence database for evidence of mechanisms by which begomoviruses diversify, in relation to their Eudicot host(s) and to a lesser extent, their whitefly vector; (iv) Archive plant and virus biodiversity data through a seamless network of allied electronic biotic, genetic, and taxonomy databases committed to long-term data storage, and to the exploration of biodiversity and integrative molecular and biological processes, essential for learning more about the complexities of life forms and their interactions on the planet Earth.

The project addresses the following hypotheses: Compared to their cultivated counterparts (which exist in monocultures and varying ecological/geographical habitats, plant hosts are genetically uniform, are often re-distributed by humans): (i) 'wild' begmoviruses exhibit stronger phylogeographical affiliations; (ii)'wild' begomoviruses, have unique genetic signatures that reflect differences in co-evolution with their wild hosts, (iii) possibly utilizing different/distinct diversification mechanisms, or (iv) employ similar mechanisms, but with different frequencies, (v) evolve at different rates, and (vi) possibly utilize different 'hot spots' for recombination. Also, we will consider whether (vii) uncultivated plants influence virus diversification differently than do cultivated plants, and if molecular signatures of 'wild viruses' reveal distinctive patterns that may reflect their wild heritage, and if so, whether such patterns will enable predictions about 'wild' and 'cultivated' virus host ranges. These hypotheses will be tested by comparative analyses of viral genome sequences obtained herein with sequences of begomoviruses isolated from cultivated plants (for which a number of genome or partial sequences have already been determined). Other novel aspects that will be explored involve the consideration of plant-virus interactions using molecular genetics together with phylogenetic approaches to seek evidence for patterns of viral-plant co-biodiversification (host, habitat, geographical, etc.) (Savolainen et al., 2000; Soltis et al., 1999). Also, we may very well also discover new aspects of begomoviruses diversification that are novel in native or naturalized plants. This project is highly do-able and relies on cross-disciplinary, international, and local collaboration between experts, many of who now work together.

Ecological Scale, Taxonomic Breadth, and Conceptual Hypotheses.

Virus phylogenies, recombination, and reassortment. We will assume early post-Gondwanaland presence of vector (R.Gill, pers. commun.) and viruses.

First, we will uncover any NW monopartite or tripartite (A, A, B or A, B, B) viruses? Presently viral phylogenetic relationships are not strictly biogeographic for NW but are for OW viruses, possibly suggesting movement of viruses between regions? If so, can lineages be identified that correlate with host or climate/weather that reveal unexpected patterns, where strict patterns appear to be lacking.

Natural virus host range. Employing information available from the several herbaria involved in the project, information from the literature, and at floristic databases (TROPICOS, The New York Botanical Garden Specimen Database), we will ascertain the ranges of Angiospermous plants of interest that are known virus host. Then, these plants will be scanned across their ranges to assess degree of overlapping between the hosts and the viruses.

Weeds: endemic and truly native weeds vs. ruderal vs cultivated plants. The most derived viruses should be associated with cultivated plants due to founder effects and possibly adaptive evolutionary aspects (virus-host defense system interactions). An interesting component of our search will be related to the association between angiosperms with truly restricted distributions and their begomoviruses. This aspect of the project will be focused within a geographical circumscription that is fairly well known to us the Yucatan Peninsula Biotic Province (YPBP). Although the flora of the YPBP is still far from being deeply known, we have a basal knowledge that allows us to state a series of facts. One of them, which are relevant to this project, relates to the level of Angiosperm endemism in the area. This biogeographical feature was recently evaluated by Estrada Loera (1991) and Durán et al., (1998) and established them to be ~10% of the total flora. An even more recent reassessment of the YPBP endemic Angiosperms (Carnevali et al., 2003) reduced the estimate to ~7.5%. The endemic species are fairly well known (ca. 186 species), which should allow us to score all or most of them for Begomovirus infection.

Assessment of the relationships of begomoviruses in the endemic Angiosperm taxa (Holm et al., 1999) will allow us to evaluate hypotheses pertaining to Begomovirus evolution and speciation. Thus, some of the questions we will address are: do endemic angiosperms have endemic viruses? This would possibly imply virus speciation associated with the cladogenetic events in the angiosperm group, if the endemic angiosperm is a neoendemic, meaning that has evolved in situ. However, if the endemic angiosperm is a paleoendemic (meaning that its present distribution is a remnant of a formerly wider distribution), then a host-virus association is likely to be ancestral and the virus may have had a former wider distribution. Such patterns will only become evident when host and natural virus host ranges are fully ascertained, and a robust virus phylogeny is available.

Urgency.

Virus databases: no systematic effort has been undertaken to inventory the distribution, diversity, or examine the influence of uncultivated plant species virus diversification and host-distribution. Only a small number of characterized begomoviruses (~45) are archived in the ICTV database, thus one of the most important contributions of this project involves populating the database with information at the virus 'isolate' level. The ICTVdB provides insights to and from where such agents disperse or emerge, and ensures that new isolate information is immediately submitted to the database, alerting biologists and epidemiologists to the possibility of emergent pathogens, alterations of existing ones, or of the disappearance of once widespread viruses. From such sequence patterns, population diversity can be compared to previously known viral genomes, which will permit for the first time the 'tracking' of evolutionary changes. Sequencing begomoviruses primarily from cultivated species (gemini@biosci.arizona.edu) in recent years has made it necessary to overhaul the taxonomy of the genus, and extant descriptions in ICTVdB are in urgent need of revision in relation to species, strains, and populations associated with native/naturalized hosts (Stanley et al., 2004). We expect that this inventory will reveal numerous unrecognized begomovirus species and strains from uncultivated plants, thereby contributing to the longer crucial term aim to sequence representative 'isolates' at the population level.

Plant herbaria and electronic databases: this project provides a seamless mechanism to contribute to existing and developing national and international electronic plant databases, and will provide crucial voucher specimens to regional herbaria, which are badly needed to preserve botanical history and to attain the larger goal of producing an inventory and/or archive for all species of life on earth.

Mining plant pathogen sequences to identify disease determinants: an inventory the diversity of Begomovirus-wild Eudicot complexes will enhance present efforts to reduce virus damage to cultivated plants using pathogen-derived resistance strategies, by contributing an expansive inventory of conserved and virus-specific sequences. We expect that this broadly representative sequence database will also be relevant for mining conserved pathogenicity determinants involved in plant defense responses to virus infection, also lucrative targets for disease control. Finally we expect that a comprehensive inventory of begomoviruses will constitute an important point of reference for future models involving multiple kingdom inventories, including other microbial communities, which are currently under-explored (Garrido-Ramirez et al., 2000).

Taxonomic Considerations.

Begomoviruses. Two major begomoviral clades are presently represented (Fauquet et. al., 2003; Rybicki, 1994; Padidam et al., 1995). Begomovirus phylogenies are based on the DNA-A component or monopartite chromosome. Viruses originating in the Old World are overall more divergent than their NW counterparts, at ~15-30%, irrespective of a genome organization. Thus, it is supposed that certain Old World viruses constitute the basal taxa for the genus; however, the group is not monophyletic and so may have diverged long ago leaving no tangible evolutionary record. A second virus sequence of evolutionary import is the 'iterated' or inverted, directly repeating sequence in the common region (CR) (7-15 bases), which confers binding specificity to the viral REP-associated protein. 'Iterons' are identical or similar for closely related species (Arguello-Astorga et. al., 1994), thus, yielding a 'signature sequence' useful for 'tracking' molecular signatures in common viral lineages, and further, are favorable 'hot spots' for recombination.

The DNA-B component of bipartite begomoviruses is highly divergent, and has been considered less phylogenetically uninformative than the DNA-A. This may possibly reflect the need for flexibility in host adaptation (Gillette et. al., 1998; Ingham and Lazarowitz, 1993; Petty et al, 1995; Pooma et. al., 1996). Even though it is not phylogenetically useful, it is essential for other considerations. First, the intergenic region of a cognate DNA-A/DNA-B pair, or 'common region' (CR) is nearly identical because it is crucial for viral replication-associated protein recognition (Bisaro, 1996; Hanley-Bowdoin, 1999; Lazarowitz, 1992). Thus, if two components share an identical (or nearly so) CR (190-220 bases), it is possible to conclude with some confidence that they are a cognate pair, which facilitates recognition of mixed infections when non-cognate components are identified (Harrison et. al., 1997).

In general, however, the taxonomy of the genus Begomovirus has been confounded by the only recent recognition of the virus family, Geminiviridae, having been established only in 1978. This was followed by the subsequent emergence or discovery of potentially hundreds of new species or strains, particularly in the genus, Begomovirus, which incite diseases of agriculturally important crops. Initially, the guidelines for nomenclature at the species level were vague and taxonomic definitions had not been agreed upon, owing mostly to the only recent recognition of these novel ssDNA containing plant virus group, together with the paucity of genetics studies and unavailability of begomoviral sequence data. The Geminiviridae Working Group (Stanley et al. 2004), together with the International Committee on Nomenclature for Viruses (ICTV) (itself, established in 1966), have since drawn up working guidelines for lower and higher order taxa within the family. In 1995, a list of criteria was developed that allowed virologists to designate existing and newly discovered species. The list continues to be modified by the Working Group as new information so mandates. Two highly relevant discoveries that have confounded taxonomic considerations and thus have inspired recent modification of criteria are: (1) growing evidence for interspecies and intergeneric recombination in the Geminiviridae, which have become recognized as essential, but previously unrecognized, mechanisms for genetic diversification, and (2) the discovery that certain monopartite viruses require a non-viral extrachromosomal 'satellite' DNA' (satDNA) molecule for induction of disease symptoms (Briddon et al., 2000; 2001; 2003; Saunders et al., 2000; 2002). Presently, satDNAs only are known in the Eastern Hemisphere (Stanley et al., 2004). Clearly, the recombining of sequences (hot spots?) between species from different clades, and the reassortment of chromosomes between distinct strains, can confound the working species/strain definitions, and so careful diligence must be exercised to conduct thorough and appropriate sequence analyses prior to designating new begomoviral species.

The generic definition states that 'a virus species is a polythetic class of viruses that constitutes a replicating lineage and occupies a particular ecological niche' (van Regenmortel et al., 1997). As such, the properties of a virus can be compared with a set of species demarcation criteria to establish its taxonomic status. Albeit, theoretical, the species concept provides a convenient way to categorize viruses. Unlike individual virus isolates, a viral species cannot be manipulated, nor does a species have a precise genomic sequence, per se and so viral species are conceptualized by a 'type species sequence'. The working definition for a begomoviral species considers that two DNA-A components (bipartite genomes) and/or monopartite chromosomes will diverge by 11% (89% nt identity) or greater. The ICTV has not adopted strain definitions because these can vary from genus to genus, owing to differences in measurable biotic characters, which are key guiding features. But conceptually, strains are 'viruses belonging to the same species and have stable, heritable biological and molecular differences'. For begomoviruses, strains typically share >90-96% nt identity and express measurable phenotypic character(s) such as differences in host range or symptom phenotype in the same host. Finally, the Geminivirus Working Group has established and revises the rules for viral nomenclature (as these are quite complex, details will not be provided here) (Fauquet et al., 2003; Stanley et al., 2004).

The confounding effects on taxonomic considerations imparted by recombination, reassortment, or the involvement of a non-viral satellite sequence, which constitute varying proportions of the overall sequence of the infectious agent, quickly become apparent (Idris et al. 2002). For example, a recombinant containing one-fourth of the genome/DNA-A chromosome of one species may group erroneously when subjected to phylogenetic analysis. Numerous begomoviruses have been shown to exhibit such anomalies, while cases of double or triple recombination also have been discovered for which an entire DNA-A component and its counterpart DNA-B component group in entirely different clades (reviewed in Brown, 2001). Finally, there is evidence that certain satDNAs may be capable of interacting with a cognate and noncognate helper virus to induce disease symptoms (Idris and Brown, in preparation), posing a further conundrum when considering biotic phenotypes for strain and species demarcations, in the context of percentage of total genome occupied by a foreign sequence (Stanley et al., 2004).

Eudicots. Although the Eudicots are only a subgroup of flowering plants, or angiosperms, they include more than half of all living plant species, outnumbering all other plant groups combined (http://www.ucmp.berkeley.edu/anthophyta/auanthophyta.html). This is especially notable because Eudicots have only existed since the middle of the Cretaceous, about one-quarter of the lifetime of land plants. There are two competing hypotheses for angiosperm origin (Judd et al., 1999). The two hypotheses differ with respect to relationships and put forth very different pictures of angiosperm diversification. The Paleoherb hypothesis suggests that the basal lineages were herbs with rapid lifecycles, whereas the competing Magnoliid hypothesis suggests that the basal lineages were small trees with slower lifecycles. Cladistic analyses by Doyle and Donoghue (1993) favor an early angiosperm with morphology similar to living members of the Magnoliales and Laurales, which are medium-sized trees with long broad leaves and large flowers. This hypothesis is born out by molecular studies and so, also is favored by many systematic botanists. It suggests that the earliest angiosperms were under story trees and shrubs, and that the flower was not the key innovation for rapid angiosperm diversification (Barkman et al., 2000; Chase et al., 2000; Crane et al., 195; Donoghue & Doyle, 1989; Doyle, 1998; Kuzoff et al., 2000; Mathews and Donoghue, 1999; 2000; Wing et al., 1993; Zanis et al., 2002). An alternative view holds that angiosperms have an herbaceous origin (Taylor and Hickey, 1992). Using cladistics analysis, the basal angiosperms are tropical paleoherbs, flowering plants with uncomplicated flowers, and a mix of monocot and dicot features. The implication is that the key innovations of flowers, and a rapid life cycle were present in the earliest angiosperms (Barkman et al., 2000).

Plant phylogeny and species richness. Six main vascular plant families, which represent the most common species infected by begomoviruses, will be emphasized in this study. An important conundrum is that these families are not considered close relatives in recent phylogenetic studies (Judd et. al., 1999; Soltis et. al., 1999; Savolainen et. al., 2000). Instead, they belong to different orders representing all major branches of tricolpate angiosperms. The Euphorbiaceae (Malpighiales), Fabaceae (Fabales), and Cucurbitaceae (Cucurbitales) are placed in the Eurosids I clade; with the Malvaceae (Sapindales) in the Eurosids II clade; the Solanaceae (Solanales) in the Euasterids I clade, and the Asteraceae (Asterales) in the Euasterids II clade. All genera have herbaceous annuals and/or perennials species and occur in the NW, either as indigenous weedy or ruderal (early successional) species, or as exotic introductions found primarily in disturbed habitats. Many genera also include crop plants. This list (not shown) was compiled from many sources, including Holm et al., (1979) and most are widespread and occur in our areas of study.

The Euphorbiaceae (Malpighiales) is a large family (ca. 300 genera, 6,900 species) consisting of trees, shrubs, herbs, vines, and succulents found throughout the world, primarily in tropical areas. It contains few crops, of which only cassava (Manihot esculenta) is widely cultivated. Other crops include castor bean (Ricinus communis), and Cnidoscolus chayamansa, grown occasionally in the NW tropics as a vegetable. Important weedy genera include Chamaesyce, Phyllanthus, and Poinsettia. Additional genera associated with the following orders are also considered possible hosts owing to the potentially diverse range of the whitefly vector, and will be considered here. They include: Caryophyllales, Polygonales, Zygophyllaceae, Malpighiales, Oxalidales, Fabales, Myrtales, Brassicales, Malvales, Solanales, Gentianales, Lamiales, and Apiales.

The Fabaceae s.l. (Fabales) is a very large family (ca. 630 genera, 18,000 species) of trees, shrubs, and herbs found throughout the world. The majority of economically important plants-both crops and weeds-belong to the Papilionoideae subfamily, most of which are herbaceous. The majority of Mimosoideae and Caesalpinoideae are woody, but each of these subfamilies also contains herbaceous plants that are potential Begomovirus hosts. There are many crop plants, particularly pulses and forages, in this family, including species of Arachis (groundnut), Glycine (soybean), Lupinus (lupine), Medicago (alfalfa), Phaseolus (beans), Trifolium (clover), and Vigna (cowpea). After the cereals, Fabaceae are probably the most important, abundant, and widespread crops in NW agroecosystems. The Cucurbitaceae (Cucurbitales) is a comparatively small family (ca. 120 genera, 825 species) of mostly herbaceous vines, both annual and perennial, found primarily in the tropics and subtropics. This family includes many crops (melons, squash, cucumbers, gourds, chayote), but relatively few weedy species. The Malvaceae (Malvales) is another large family (ca. 1100 genera, 1,800 species) of both woody and herbaceous species found throughout the world. Recent treatments combine the Tiliaceae, Sterculiaceae, and Bombacaceae in the Malvaceae s.l.; our list includes only species generally placed in the Malvaceae s.s. The most important crop plant in the family is cotton (Gossypium); also cultivated are okra (Abelmoschus esculentus) and many ornamentals. Many are common weeds, particularly Abutilon, Malva, Mavella, and Sida species The Solanaceae (Solanales) includes nearly 150 genera and 3000 species; is worldwide in distribution but is most diverse in the NW tropics. The family includes many crops, including tobacco (Nicotiana tabacum), tomato (Lycopersicon esculentum), potato (Solanum tuberosum), peppers (Capsicum spp.), eggplant (Solanum melongena), and tomatillo (Physalis ixocarpa). Additionally, many species are cultivated as medicinal plants and ornamentals, and many are abundant and widespread weeds (e.g., Solanum). The Asteraceae (Asterales) is the world's largest family of vascular plants (ca. 1550 genera, 23,000 species). They are worldwide, but are particularly common in the arid and semiarid regions of temperate and subtropical zones. Members of the family are mostly herbaceous, but include some trees, shrubs, vines, and a few succulents. Many genera include common and abundant early successional plants and agricultural weeds, but there are relatively few crops, which include sunflower (Helianthus annuus), safflower (Carthamus tinctorius), lettuce (Lactuca sativa), and artichoke (Cynara scolymus).

ICTV Database. The Universal Virus Database (dB) (ICTVdB) of the International Committee for Taxonomy of Viruses (ICTV) is a database with searchable tools that enables cataloging and tracking of viruses, in relation to geographic and environmental ranges for virus vectors and plant hosts. The aim of the ICTVdB is to catalog the NCBI Taxon ID (used by all major sequence databases) and sequence accession numbers using the decimal code of the virus name in ICTVdB. The value of archiving data in ICTVdB for the virus host plant and vector, is that virus sequence data and the resources used to identify a virus at the time of collection/identification will broaden the linkage beyond the sequence level to single genes because the ICTVdB has a large set of fields devoted to functional genomics. Further, each viral protein is described and has active hyperlinks to protein databases. The ICTVdB has an inter-operable structure that can be exploited for a resource linkup between other database structures. The decimal virus code used throughout provides the infrastructure for the resource linkage. Early in the development of the database a Decimal Code was established and assigned to each virus. The structure of the Decimal Code simultaneously indicates the taxonomic status of a virus from the Order to the isolate level. As the database developed, it was seen that the decimal code served as more than an unequivocal identifier for taxonomically correct internal linkages. It has since been used as a filename for transposing ICTVdB to the web, and has been adopted as a surrogate accession number both by EMBL and SWISS-PROT, which link to the ICTVdB database with a code that is now widely used (Tidona et. al., 2001). 'Springer Index of Viruses' (1st Ed, Springer, Heidelberg 1511pp) provides the basic vehicle for linking between other databases. Most recently, the NCBI Taxonomy Workgroup provided the framework the linkage, "LinkOut", a protocol for filters based on XML. The XML filters are generated and updated on a regular basis by ICTVdB and submitted to NCBI. This arrangement also has been established with the EBI/Swiss-Prot Taxonomy Group and the Species 2000 projects, which employ the decimal code (Büchen-Osmond, 2003). When new data are entered into the ICTVdB, they are linked to the corresponding sequence data in major public sequence databases. Two-way linkages are provided, so that a user of GenBank can move easily from a viral genome sequence to a full Virus Description and its properties, in a single click of the mouse. The accumulation of detailed data at the isolate level will facilitate functional genomics and proteomics required to understand the biological properties of biological interactions. The user can select a set of characteristics from ICTVdB; identify genes associated with those characteristics, and move seamlessly to GenBank for homology determinations. Thus, the ICTVdB, with its links to relevant resources, will provide a tool for further elucidation of changes in viral genomes resulting from environmental and other forces, which lead to a decisive mutation and at times, the emergence of new viruses. This cross-linking between sequences and gene function will constitute a major step toward a precise justification for Begomovirus nomenclature, a problem that has become more widely recognized as a greater number of sequences have become available, while lacking sufficiently precise documentation of isolate and host origins.

Work Plan & Project Management.

Collaborators/Cooperators/Roles. USA-AZ team. JK Brown (PI) has a longstanding involvement in the ecology, epidemiology, and taxonomy of emerging begomoviruses in agroecosystems, worldwide. She also studies whitefly vector genetic diversity and its effect on begomovirus diversification, and the basis for virus-vector specificity. She has collaborated with Dr. Julio Bird in Puerto Rico for fourteen years. Julio Bird has characterized in biological terms nearly every begomovirus on the island during his career. He is also the first person to make the observation that B. tabaci is a polymorphic insect vector for the genus, opening the door for work Brown has subsequently embarked upon to understand taxonomic and vector-based differences in B. tabaci populations. S. McLaughlin is Professor, Plant Sciences Department, and Curator of the UAZ Herbarium. He is an expert in the botany and ecology of plants in the arid Southwestern US and will carry out plant identifications received/collected in conjunction with the Brown lab. N. Merchant is an expert in high throughput computing and workflow environment software development. He develops Laboratory Information Management Systems (LIMS) utilized in the UA high throughput genomics and proteomics core facilities. He is responsible for the Bioinformatics core and associated websites at UAZ. C. Büchen-Osmond has developed and manages the ICTVdB for the ICTV since 1991. She has collaborated over the past years with EBI (Swiss-Prot/UniProt) and GenBank at NCBI to establish the linkage between the currently valid virus taxonomy and its genome. Though she recently left the Biosphere, she will remain a vital member of the AZ team. Mexico team. Co-PI's: G. Carnevali Fernández-Concha works mostly in the systematics of Orchidaceae and Bromeliaceae. Since establishing in Yucatan (1996), he has specialized in the floristics and biogeography of the Yucatan Peninsula Biotic Province. He is currently developing the first comprehensive, modern, full-scale flora of the area, and is the head of the CICY herbarium, the seventh largest herbarium in México and the most important in the Yucatan Peninsula. O. Moreno-Valenzuela has expertise in Plant Biochemistry and Molecular Biology. He is involved in the study of begomoviruses in natural environments in the Yucatan Peninsula. R. Rivera Bustamante has long been involved in the study of begomoviruses as emerging plant viruses in Mexican agroecosystems. His laboratory uses geminivirus-based models to study gene expression in plants, virus-virus interactions, and plant-virus interactions. Cooperators/collectors in each locale have expertise in plant pathology/virology and/or are trained Herbaria curators. Our virologist cooperators are familiar with characteristic symptoms of begomovirus-infected native and naturalized species in their area, and have worked with one or both of the lead laboratories. Virologists will work collaboratively with Herbaria Curators to collect and archive the plant and virus collections. Support for all collectors has been requested to cover their costs for travel, lodging, per diem, tubes, glycerol, markers, bags, supplies for vouchering herbaria specimens, mailing costs, and other miscellaneous expenses they are expected to incur.

Collectors for the Arizona, Sonora/Sinaloa will be the Brown/McLaughlin labs, S. Garza, and J. Sanchez-Escalante UNSON Hermosillo; sgarza@rtn.uson.mx; and J. A. Garzon INIFAP, Sinaloa, Mexico; sin_inifap@sagar.gob.mx. Other locales/teams are: Brazil: A.L. Lourenção, Instituto Agronomico de Campinas/J.A. Caram de Souza-Dias (IAC); andre@iac.br; Dominican Republic: R.T. Martinez/Gerhard Jurgens (IDIAF); rmartinez@idiaf.org.do, Honduras & Nicaragua: FHIA team (D. Krigsvold, J. Dickson, G. Pilz, H. Aguilar; J. Melgar, A. Ruerda; dkrigsvold@fhia.org.hn; Guatemala: M. Palmieri and A. MacVean, Universidad Del Valle; palmieri@uvg.edu.gt; Panama: D. Emmen (U of Panama) and Carmen Vergara; demmen@cwpanama.net; Florida: R. McMillan, UFL rtmcm@mail.ifas.ufl.edu w/S. McLaughlin (AZ), UA; Puerto Rico: J. Bird, Prof Emeritus, UPR acbird33@hotmail.com/J. Ackerman, UPR Herbarium, San Juan PR; ackerman@upracd.upr.clu.edu; Texas: T. Isakeit/M. Dubrule Reed; Plant Pathology Dept, Texas A&M Univ, College Station; t-isakeit@tamu.edu; additional: CICY and CINVESTAV have identified teams in Argentina, Colombia, Costa Rica, Venezuela, and for the remainder of Mexico (joint proposals to CONACYT & CONABIO). Sampling strategy. Sampling of these plant/begomovirus complexes is driven by a partial understanding based on much preliminary work involving analysis of some complete and many partial begomoviral genome sequences, which suggests a gradient of biodiversity that extends from north to south in North America. At the northern most end (US Sunbelt States), numerous strains of the same species abound, and reassortment has been demonstrated for a number of species, which are now postulated to constitute the virus equivalent of a 'species complex'. In North America, the most divergent and abundant species have been found near the Mexico-Guatemala border, but the Yucatan Peninsula has not yet been explored extensively. In the Caribbean, legume-infecting begomoviruses occur in most locations, but in Puerto Rico, preliminary studies (Brown & Bird) have demonstrated that nearly every other species there is unique when compared to viruses identified in Central American-E. Caribbean and Caribbean both on the land bridge and on individual islands, including Panama. Here, there is evidence for extensive intramolecular recombination between viruses from different lineages (clades), albeit for most of these, complete genome sequences are lacking. South America appears to possibly constitute a second center of diversity beginning in Colombia, despite its proximity to the E. Caribbean and Panama. A recent study of viral coat protein genes from Brazil revealed a surprising amount of diversity there and only one virus (of legumes) with close relatives known elsewhere in the Americas/Caribbean region. Only partial sequence data are available for a few viruses from Argentina and Venezuela, some which are most closely related to Brazilian viruses, while others have close relatives in the E. Caribbean or Central America. Begomoviruses in the remainder of South America and much of Mexico and the Caribbean remain unexplored (unpublished data, JK Brown; Brown, 2000; 2001). Plant and virus collections. Teams will collect and archive plant and virus specimens of symptomatic plant species. The curator of each herbarium and the counterpart virologist will comprise a team that will collect and/or voucher plant samples in each strategically selected site. For virus samples, leaf discs will be collected from all species showing typical begomovirus symptoms. Five-ten leaf discs will be punched from the apical or lateral meristematic growth and placed into tubes containing glycerol (3/plant). The UAZ/Brown lab holds USDA-APHIS-PPQ permit #64337 (Exp Jan 28/05) and can receive plant samples collected worldwide. One tube will be archived in the country of collection, and two tubes will be forwarded to 'processing lab' where one will be archived (-80C) and the other used for laboratory studies. Voucher plant specimens will be collected in triplicate, with one specimen remaining in a herbarium of the host country, one deposited in the herbarium at Centro de Investigación Cientifica de Yucatán (CICY), and one deposited in the herbarium at UAZ (http://eebweb.arizona.edu/HERB/index.html). Voucher criteria will include: Country, Political boundary (county, municipality, district), Date, Collectors' name, Collection number [unique # for each collector], Geographic coordinates (Latitude/longitude or UTM) recorded by a GPS unit, Elevation, Plant habit (annual, perennial, vine), flower color, habitat and associated species. Approximately 50-100 collections per yr/per site will be undertaken for PCR analysis/Phase I, depending on the extent of species richness and symptom variation (A detailed 'Collection Protocol' has been developed; available upon request). Because begomoviruses infect meristems, certain symptomatic plants may be sterile, lacking the morphological characters required for their identification. In such instances, voucher specimens should also be obtained from fertile plants of the same species. When collectors are unable to determine with certainty which fertile plants represent the same species as sterile, infected plants, specimens of likely conspecifics will be collected. Voucher identification, archiving, distribution of triplicated vouchers, and the population of plant databases will be ongoing for the duration. The website/allied databases will be established in the first trimester.

Since many plant collectors that will participate in the Begomovirus project we have provided a Spanish version for the data entry procedure. The above listed documents can be accessed from the submission screen in Spanish and are translated on the fly into Spanish using ALTA VISTA's Bablefish webpage program. The questionare and submission forms on the other hand are published in Spanish.

Copyright 2002 ICTV. All rights reserved.

Comments to ICTVdB Management