Exploitation of the genetic drift to enhance resistance durability ?

A major challenge of such studies is to disentangle the effects of genetic drift and of selection on virus populations when these two evolutionary forces vary simultaneously, i.e. in the absence of neutral markers, which is the rule for plant viruses. To date, no population genetics model is available to estimate jointly the intensities of genetic drift (through the parameter Ne) and selection (selection coefficient s) in such cases. We have developed such a model and shown, by simulating data under the Wright-Fisher model with an independent software (FFPopSim) that it was highly accurate and unbiased to estimate Ne and s.

We applied this model to evolution data of a composite population of five PVY variants in 15 contrasted pepper genotypes. This revealed that:

• both Ne and s varied greatly from one pepper genotype to the other
• Ne varied also during the course of plant infection, with a narrow bottleneck observed at the onset of systemic infection
• the ranking of the selection coefficients s of the five virus variants was similar between the 15 plant genotypes but the fitness differences between variants varied substantially from one pepper genotype to the other.

The model that we developed to estimate jointly Ne and s could be of great interest to study the evolution of parasite populations, especially haploid and clonal microorganisms, within their hosts, since it does not require neutral markers and applies to a large range of Ne and s values.

Importantly, both Ne and s, as measured from the viral populations, were shown to be highly heritable from the plant side. As a consequence, Ne and s measured through the effect of plants on the genetic composition of virus populations could be considered as agronomic traits of interest that show extensive variability and that are highly heritable and may be used to manipulate the adaptation of viruses and enhance plant resistance durability. Supporting this assumption, we showed that the two factors that explained best the frequency of pepper resistance breakdown in laboratory conditions were (i) the global virus accumulation within infected plants and (ii) Ne at the initial steps of infection (i.e. at inoculation and in the inoculated leaves).

Narrow bottlenecks affect pea seedborne mosaic virus populations

After having estimated, for the first time, Ne during vector transmission of a plant virus (ca. 1 infectious virus particle being transmitted by an individual aphid vector), we have more recently estimated, also for the first time, Ne during vertical seed transmission. Seed transmission is an economically significant issue in at least 18% of plant viruses in at least one of their host plant species. Although almost no bottleneck was observed during the colonization of the vegetative parts of pea (Pisum sativum) plants by pea seedborne mosaic virus (PSbMV; genus Potyvirus), a single virus particle contributed, on average, to the transmission of PSbMV in pea seeds. Such narrow bottlenecks drastically affect the genetic diversity in virus populations and, through the genetic drift that they induce, may alter the adaptation capacities of viruses.