Evolutionary Biology of Topsovirus

M. Tsompana, S. -H. Sin, and J. W. Moyer
Department of Plant Pathology, North Carolina State University

Introduction
The disease known as "tomato spotted wilt" was first described in Australia in 1915 and it was not until 1930 that it was shown to have a viral etiology. Since that time, viruses similar or identical to the tomato spotted wilt virus (TSWV) have caused severe epidemics in tropical, subtropical and temperate regions throughout the Northern hemisphere, Western Europe, and Asia. At first, this taxon of plant viruses was categorized as a monotypic virus group consisting of a simple virus (TSWV), but in 1991 the placement and unique status of TSWV was significantly challenged with the report of impatiens necrotic spot virus (INSV) (1). Thorough molecular biological studies of the virus revealed a taxonomic relationship with the Bunyaviridae, a recognized group of animal viruses including four distinct genera: Bunyavirus, Phlebovirus, Hantavirus and Nairovirus. As a result, the Tospovirus genus was established within the family Bunyaviridae. In addition, differences in nucleic acid homology, serological relatedness, and host range between isolates of TSWV offered sufficient evidence to warrant establishment of two distinct serogroups within the Tospovirus genus (Table I.) (2)

Table I
Tospovirus species generated during the International symposium in Tospoviruses and thrips, May 1998, Wageningen, Netherlands (2)

Molecular Biology Of Tospoviruses
I. Virus Structure And Cytopathology
The morphology of the Tospoviruses is typical of members of the family Bunyaviridae (Figure 1.): virions are 80-110 nm, quasi–spherical and are defined by a membranous envelope containing two viral–coded GPs designated G1 (78 kDa) and G2 (58 kDa). Particles are also composed of two other proteins including a putative 330 kDa replicase (L) and the 29 kDa N protein.

	Figure 1. Tospovirus virion structure
		a) diagrammatic structure		b) electron micrograph of TSWV. Open image for larger view.

Other structures of viral origin are also observed in infected cells composed of nonstructural (NSs and NSm) or structural proteins. The NSs protein accumulates in long fibrillar structures that may associate as loose bundles (e.g. TSWV) or in highly ordered paracrystalline arrays (e.g. INSV) (Figure 2) (3). NSm protein is detected in granular electron-dense inclusions. Also excess N protein is observed as electron-dense, granular material in the cytoplasm.

	Figure 2. Ultrathin sections of Nicotiana benthamiana leaf tissue infected with INSV(left) and TSWV(right).

The Tospovirus genome is partitioned among three single-stranded RNA segments labeled L, M, and S in order of decreasing size (Figure 3.) (4). Each segment is pseudocircular with a panhandle-like structure at the termini formed by base pairing of inverted complementary sequences. There is an eight-nucleotide sequence (5¢-AGAGCAAU-3¢) conserved at the termini of each segment of all TSWV isolates and of the S and M RNAs of INSV. The intergenic region (IGR) on the M and S segments is AU rich with a high inclination for base pairing. The IGR’s of the M and S segments of TSWV are reported to be of variable lengths and are regarded as the most hypervariable regions of the genome. A 33-nucleotide duplicate sequence occurring in the IGR of the S molecule of some TSWV isolates has been correlated with a loss of competitiveness in mixed infections of isolates with and without the duplication (5).

The largest RNA is approximately 9 kb and has a single open reading frame (ORF) in the viral complementary sense coding for a 330 kDa protein, which is the putative RNA-dependent RNA polymerase. Both M and S RNA’s have two ORF’s in an opposite polarity (ambisense) most like the Phleboviruses. The M RNA is approximately 4.8 kb and encodes a 34 kDa protein in the viral sense designated NSm, proposed to be involved in cell to cell movement and stimulation of tubule formation in protoplasts. The viral complementary sense ORF codes for the G1/G2 precursor protein. The G1/G2 proteins are more highly conserved between TSWV viruses than the N protein. The S segment is approximately 3 kb and contains two ORF’s in ambisense orientation separated by a large intergenic region. The ORF nearer the 5¢ end of the RNA codes for a nonstructural protein in the viral sense designated NSs (54 kDa) whose function has not been determined. The ORF nearer the 3¢ end is in the viral complementary sense and codes for the nucleocapsid protein (29 kDa) which encapsidates the viral RNA within the viral envelope.

Figure 3. Tospoviruses genome organization. Positive and negative sense ORFs are indicated by + and -, respectively. Open image for larger view.

Host Range And Symptomatology
TSWV has one of the broadest host ranges of any plant-infecting virus. The virus infects over 900 plant species which include both monocotyledonous and dicotyledonous plants (6). Important economic plants susceptible to TSWV infection include tomato, potato, tobacco, peanut, pepper, lettuce, papaya, and the ornamentals chrysanthemum, begonia, agetarium, and impatiens. Other tospoviruses have much narrower host ranges and thus the broad host range of TSWV is not characteristic of the genus. INSV is predominant in greenhouse flower crops, including cineraria, ranunculus, impatiens, New Guinea impatiens and others.

Symptoms caused by Tospoviruses are highly variable and of little diagnostic value. Necrosis on several plant parts, chlorosis, ring patterns, mottling, silvering, and local lesions are the most characteristic symptoms. In some instances, the disease can limit the feasibility of crop production. Infection of young plants of species such as tobacco, tomato, pepper and several floral crops, is often accompanied by high mortality rates, whereas infection of some plants and older plants of all species may be latent, mild, or transient. Sometimes symptoms may mimic symptoms and injuries caused by other biotic and abiotic stresses.

Transmission
TSWV is transmitted from plant to plant in nature by eight species of thrips; Frankliniella bisponsa, F. fusca, F. intonsa, F. occidentalis, F. schultzei, Thrips palmi, T. setosa and T. tabaci (2). A distinctive characteristic of its infection cycle is that virus acquisition can only occur when thrips larvae feed on infected plants. Thus, initiation of the infection cycle can occur only when female adult thrips lay eggs on TSWV-infected leaves that are suitable for egg and larval development. Once acquired, TSWV is retained by thrips through molting, pupation, and emergence to the adult stage. The primary dispersal of TSWV is by adult thrips, which disperse widely, feed on many different plant hosts and may remain viruliferous for the remainder of their life. Evidence for replication of the virus in the insect vector is based on the accumulation of non-structural protein (NSs) and the visualization of other inclusions in endothelial cells, muscle cells and the salivary glands. TSWV can also be spread by vegetative plant propagules but not through true seed. It can also be transmitted experimentally in infected tissue extracts.

Mechanisms of RNA plant virus evolution
The most remarkable paradigm of genetic variability, both within and between species, is given by plant RNA viruses. The diversity in nucleotide sequences among RNA viruses, is enormous especially when compared to the evolution of most other life creatures in earth. The driving forces of this evolution are believed to be: mutation, recombination and reassortment. Generally, these are the forces that drive evolution.

RNA viruses show extremely high mutation rates, because of the lack of proofreading ability of their replicases (7). Although, the mutation or error rate of viral RNA–dependent RNA polymerase (RdRp) has not been estimated for plant virus, it has been measured for animal RNA viruses and it is approximately 10–4, or one error per genome (8). This is the generally accepted error rate of RdRps. Despite the fact that the error rates are relatively conserved among all RNA viruses, the mutation frequencies, which refer to only those misincorporations that become established in a population, vary extremely for different viruses. It is known that the greater the mutation frequency, the higher the adaptation ability of the virus to new niches (i.e. hosts) (8).

Recombination in viruses is analogous to recombination in meiosis. This mechanism has been very important in the evolution of plant RNA viruses such as luteoviruses, nepoviruses, bromoviruses and cucumoviruses (8). Also plant pararetroviruses such as caulimoviruses have demonstrated evidence for recombination. On the contrary, recombination is rarely reported reported for negative strand viruses. Till now, no evidence for recombination has been proved for Bunyaviruses, and more specifically for the Tospoviruses. This is one of the major issues to be answered. Does recombination exist in this genus, and if yes in what extent and frequency? Are there species within the genus more prone to recombination than others, and if yes what are the factors that facilitate this difference?

Last but not least, RNA viruses with segmented genome use reassortment, the exchange of genomic segments among different parent of a virus with segmented genome (9) as a mechanism of divergence. This has long been proposed as one of the sources of variation in RNA viruses, provided that mixed infections are very common among field isolates of plant viruses. More interestingly, there are cases where recombination and reassortment coexist (10). Although, reassortment might be considered a rare event, its existence may confer a selective advantage such as adaptation to new hosts and expansion of the host range. Also genetic drift, the acquisition of genome changes through point mutations, insertions, deletion and inversions, is thought to be the second mechanism of evolution for viruses with segmented genome (11).

Genetics And Evolution of Tospoviruses
It was as early as 1946, when Norris demonstrated the existence of five strains of TSWV, which he separated from naturally occurring complexes (12). He named these strains Tip Blight (TB), Necrotic (N), Ringspot (R), Mild (M), and Very Mild (VM). In 1953, Best and Gallus separated from tomato plants six strains named as A, B, C1, C2, D, and E (13). Although, there was a difficulty in finding exact parallels between Norris’ strains and Bests’ and Gallus’, because of different experimental conditions and the theoretical lack of a "pure" strain of TSWV, these two reports were a mere proof that TSWV occurs in plants in nature ad a heterogeneous mixture of stable isolates. In 1954 and 1961 Best made the first attempt to explain the "cross protection effect" among TSWV strains with a new theory which implied transfer of character determinants from one virus particle to another (per se recombination) (14,15).

Today, it is known that what Best was trying to describe was genome reassortment, responsible for the forming of new phenotypes from the mixture. More specifically, the tripartite genome organization of the family Bunyaviridae allows closely related viruses to exchange genetic information through the reassortment of whole genome segments to form a distinct isolate. Qiu et al in 1998, were able to built a system to associate specific genome segments with viral phenotypes and to study factors influencing genome reassortment (5). They generated reassortant isolates by coinoculating a TSWV isolate, TSWV-D, with TSWV-10 or TSWV-MD. More interestingly, the SRNA from TSWV-D was dominant over the SRNA for TSWV-10, fact correlated with the presence of a net increase of 62nt, including a 33nt duplication in the IGR of the SRNA of the less competitive isolate. This duplicate sequence was highly conserved among isolates from the southeastern United States and an isolate from Bulgaria. In 1999, Qiu and Moyer determined that genome reassortment as the mechanism of TSWV-adaptation to the TSWV N-gene-derived resistance (TNDR) and that elements from two or more segments of the genome are involved in suppression of the resistance reaction (16). The genetic reservoir present in the heterogeneous virus populations combined with the reassortment mechanism provide conditions conducive for the rapid adaptation of virus populations to resistant host plants. More specifically, they generated a reassortant, so called L10M10SD, which combined the L and M RNA’s from TSWV-10 and the S RNA from TSWV-D, which was able to systemically infect TNDR Burley 21.

The determinants of adaptation to resistance in tomato and tobacco have been mapped to the M segment and in pepper in the S segment. Overcoming the SW-5 gene in tomato was linked solely to the presence of M RNA from a TSWV isolate, TSWV-A, and the ability of MA to overcome TNDR was modified by the L RNA and S RNA of TSWV-A (17) . In Capsicum annum (pepper), the Tsw gene conferring dominant resistance to TSWV was mapped to the distal portion of chromosome 10 (18). No mapped homologues of SW-5 map to this region in pepper. It was also inferred that the outcome of infection in plants carrying SW-5 and Tsw is controlled by distinct viral gene products, and thus these loci are not recently derived from a common evolutionary ancestor.

Phylogenetic Analysis
Phylogenetic analysis based on sequence information for different isolates of the same virus, is a reliable source of the evolutionary history of each strain and indicative of its geographic origin. In 1998 Pappu, H. et al, were able to sequence natural populations of TSWV infecting flue-cured tobacco in Georgia and infer a high degree of sequence conservation among the Nc genes of the tobacco isolates, and those reported from other parts of the world. More interestingly, the Georgia (GA) isolates formed a distinct cluster that was clearly resolved from the rest of the TSWV isolates (19). Multiple sequence alignment and dendrogram construction of the N gene and the IGR of the SRNA of TSWV, based on known nucleotide sequences from the GeneBank and unpublished sequences generated by Qiu, W. P. et al in 1998, reconfirmed this general conclusion (unpublished data) (Figure 4.). In 1999 Bhat, A. I. et al, drew similar conclusions after the analysis of the IGR of TSWV M RNA segment. Again, cluster analysis of the IGR sequences showed that all GA and Florida (FL) isolates are closely clustered and are distinct from TSWV isolates from other countries as well as from other Tospoviruses (20). A cluster dendrogram, developed by Hoffman, K. et al, from the the deduced amino acid sequence of the Nc gene of TSWV, showed that isolates from the Southeastern United States form a group that is distinct from isolates from Europe and other parts of the world (21). These independent analyses consistently revealed similarities at the molecular level correlated with the geographic origin of the isolates. The impact of host (plant and vector) species on natural virus populations is the object of our current investigations.

	Figure 4 a. Clustered rooted tree for the Nc gene of TSWV: Bunyamwera (Bunyavirus) was the outliner used and b. clustered rooted tree for the IGR of TSWV: Rift valley fever virus (Phlebovirus) was used as an outliner. Open image for larger view.

Anticipitated Benefits
We will then use these evolutionary studies on the formulation of a system that will allow diagnosticians to not only diagnose the virus but also to predict the source of the infection. This system can be used to determine if the virus outbreak is from within the cropping system of from an outside source such as local weeds, as well as provide some indication of the geographic origin of the virus. The sensitivity of TSWV populations to selection pressure strongly suggests that an additional opportunity exists exploiting this trait to identify management practices that would reduce risk. The result of our research will also increase our understanding how TSWV interact with its vector and could also provide the basis for other novel control measures of TSWV and INSV.