The virus determinants that specifically recognize vector species are the subject of numerous studies. These determinants may be glycoproteins of the viral membranes or capsids in arboviruses of both plants and animals (reviewed by Gray & Banerjee, 1999; Van den Heuvel et al., 1999; Hogenhout et al., 2003), amino acid motifs of the coat protein (Geminiviridae) (Hohnle et al., 2001) or coat-associated proteins (Luteo-viridae) (Gray & Gildow, 2003) in the circulative-non-propagative plant viruses. In non-circulative transmission of plant viruses, specific coat protein amino acids have been identified as determinants of vector specificity in Cucumoviridae (Pirone & Perry, 2002), and amino acid motifs of the HCs with similar function have been identified in both Potyviridae (Raccah et al., 2001) and Caulimoviridae (Blanc et al., 2001). In contrast, very little is known about the receptors in the insect vector that determine the specificity of virus transmission. For insect-infecting arboviruses, many receptors have been identified, presumably corresponding to membrane receptors involved in cell entry in various tissues of the vector (reviewed by Van den Heuvel et al., 1999; Hogenhout et al., 2003). However, receptors that account for the high degree of virus-vector specificity observed in circulative transmission remain to be characterized. Some genetic approaches have investigated the determinants of virus-vector specificity, and, in some cases, only a single vector gene may be involved (reviewed by Mellor, 2000). The development of the international sequencing programmes on mosquitoes and aphids may be beneficial in the near future by allowing genomic approaches similar to that initiated recently in the biting midge Culicoides sonorensis, infected with the Epizootic haemorrhagic disease virus (Campbell & Wilson, 2002). For Geminiviridae and Luteoviridae, which do not replicate in their vectors, the situation appears simpler since a less intimate relationship should involve less molecular interaction. However, only one report has been published on a putative receptor of BYDV (luteovirus) in the salivary glands of the aphid vector Sitobion avenae (Li et al., 2001). Finally, for non-circulative viruses, even the chemical nature of the putative receptor associated with the cuticle lining the vector mouthparts or anterior gut is unknown. Clearly, identifying and characterizing specific receptors of viruses in insect vectors is the major challenge for studies of vector transmission, since such receptors would be perfect targets for disease control strategies.
An interesting aspect is the modification of vector behaviour that could be either directly or indirectly induced by the virus. In fact, by modifying the feeding behaviour of vectors, parasites can quantitatively and qualitatively increase their chances of transmission. This phenomenon has been studied in parasites such as protozoa or bacteria (Hurd, 2003), but rarely in viruses. It has been shown that when DENV infects the central nervous system of its mosquito vectors, the time length of blood meals increases. This improves the chances of interrupted feeding, favouring immediate re-landing on a new host and increased virus transmission (Platt et al., 1997). In plant luteoviruses, volatile compounds emitted from potato plants infected with Potato leafroll virus attract and favour settlement of aphid vectors (Eigenbrode et al., 2002). These rare studies confirm that viruses have indeed evolved strategies for manipulating their insect vectors and open a relatively unexplored field of research in virus-vector relationships.
Experimental studies on the evolution biology of viruses have developed tremendously over the last 15 years. In vitro and in vivo studies have demonstrated that, at least for RNA viruses and retroviruses, virus populations are structured in a specific way that is sometimes designated quasispecies (Moya et al., 2000; Domingo, 2000). Due to the high error rate of viral polymerases, mutations constantly appear in viral genomes during replication. Rapidly, the population becomes an ensemble of mutant genomes centred on a mean, or consensus, sequence. During repeated and severe population bottlenecks, in asexual populations, the genetic drift is increased and results in an accumulation of mutations, designated Muller's ratchet. Since most mutations are deleterious the mean fitness of virus populations will in consequence dramatically decrease. This debilitating effect of population bottlenecks on virus populations has been experimentally demonstrated many times in bacteria and animal viruses, noticeably in Vesicular stomatitis virus (for review see Elena & Lenski, 2003; Garcia-Arenal et al., 2003). Transmission in general and vector transmission in particular has been suggested as a likely step inducing severe population bottlenecks for viruses (Elena et al., 2001; Sacristan et al., 2003). Some studies have estimated the minimum number of virus particles or infectious doses that a vector must ingest in order to transmit either vertebrate (for example see Fu et al., 1999) or plant (Walker & Pirone, 1972; Pirone & Thornbury, 1988) viruses. However, the proof of the existence of such bottlenecks and the evaluation of their actual size, in natural vector transmission, has never been experimentally established. This question is of great interest because, as proposed for the helper strategy (Pirone & Blanc, 1996; Froissart et al., 2002), some strategies of virus-vector interaction may have been selected because they ameliorate the putative severe genetic bottlenecks encountered by virus populations at each round of vector transmission.
Was this article helpful?