Skip to main content

Evolution of plant resistance to virus infection

Final Report Summary - ERVIR (Evolution of plant resistance to virus infection)

Host-pathogen co-evolution is a central question in Biology as it is at the root of pathogen emergence, host switch and host range expansion, and the composition and dynamics of ecosystems. However, contrary to agroecosystems, evidences for plant-virus co-evolution in wild ecosystems are almost non-existent. The main objective of the project ERVIR is to advance knowledge on host-pathogen co-evolution by providing experimental support of the theoretical predictions on two major models of host-parasite interaction determining the success of infection: the gene-for-gene (GFG) and the matching-allele (MA) models. The evolution of GFG and MA models are analysed in this project through the comparison of the evolutionary dynamics of major dominant and recessive resistance genes to plant viruses, which conform to each of these models, respectively. For this, we used the wild pepper Capsicum annuum var. glabriusculum (chiltepin), which is the wild ancestor of domesticated C. annuum, and the L and pvr2 genes involved in the resistance toward tobamoviruses and potyviruses, respectively.
Specifically, the aspects to be addressed in the project ERVIR are: (1) Analyse the frequency, the diversity and the population structure of the resistance genes in chiltepin populations of Mexico; (2) Evaluate the selection pressures exerted by viruses on these genes; (3) Characterise the mechanism of virus recognition and resistance.
GFG model:
First, the correlation between the presence of the L gene and the tobamovirus resistance phenotype was confirmed, and a method was developed to detect this gene in a large collection of samples representing wild and cultivated chiltepin populations from Mexico. Interestingly, the frequency of the L gene was significantly higher in the wild populations than in the cultivated ones, which suggests that the recent cultivation of this species is not yet focused on tobamovirus resistance in Mexico. Moreover, it suggests that if selection has occurred during incipient domestivation for some traits, it has been at the cost of losing resistance. Then, the analysis of more than 100 full-length L gene sequences showed a high variability and a strong geographical structure of polymorphisms at this locus. This result suggested that local adaptations of chiltepin occurred toward different tobamovirus pathotypes and/or different environmental conditions. The chiltepin resistance against P0 and P1 virus pathotypes was then investigated at medium (23ºC) and high (above 30ºC) temperatures. It should be pointed that most resistances following the GFG model are thermosensitive, being effective only at temperatures around 25ºC. This is also the case for most alleles at the L locus of Capsicum. These assays demonstrated that the P0 pathotype resistance at high temperature is frequent in wild chiltepin population, what may be of apdative value, considering the high temperatures during the chiltepin growing season in the sampled geographic areas. In order to identify molecular signatures of this adaptation, the selection pressures exerted on the L gene was evaluated. Several co-variations and sites under positive selection were detected, mainly located in the domain involved in the virus recognition. Altogether, these results suggested that the L resistance gene is a target for selection by virus infection. In addition, viral detection analyses showed the low tobamovirus incidence in chiltepin populations, which suggests that the high prevalence of the L resistance gene and its adaptation to the environmental conditions limit efficiently the spread of the tobamovirus.
MA model:
The analysis of more than 100 pvr2 mRNA sequences showed a strong genetic structure of the chiltepin populations at the pvr2 locus. However, a very low percentage of resistance alleles was detected in wild populations, which suggested that there is no selection for potyvirus resistance in chiltepin. Nevertheless, four news pvr2 alleles were identified. In collaboration with the group of Dr. Jean-Luc Gallois (INRA Montfavet-Avignon, France), the resistance phenotype of the new alleles was investigated. First, we confirmed by complementation in yeast that the transcriptional function of pvr2 is not altered by the mutations identified in chiltepin. Then, the effect of these mutations (single or combined) on the interaction with the viral protein VPg was evaluated by yeast-two hybrid assays. These experiments showed that one new allele out of four prevented the interaction with the VPg of Potato virus Y (used as a predictor to detect resistance alleles against Potyvirus family), and identified the critical role of one position in the pvr2/VPg interaction. Finally, the phenotyping of one chiltepin population confirmed in planta the results previously obtained in yeast. In addition to these functional assays, the localisation of the mutations identified in chiltepin was performed on the 3D structure of the pvr2 protein modelled in collaboration with Pr. Luis Fernández Pacios (UPM Madrid, Spain). This study showed that the alleles were characterised by point mutations which mostly occurred in the domain of interaction with the potyviral VPg, and that the most drastic electrostatic potential changes on this domain are linked with the disruption of the pvr2/VPg interaction. However, the analysis of the selective pressure exerted on the pvr2 gene did not show a clear evidence of positive selection. Virus detection assays in chiltepin populations showed a higher incidence of Potyvirus than that of Tobamovirus, which can be explained by the low frequency of pvr2 resistance, and two species of Potyvirus were identified. Thus, in addition to the key role of pvr2 gene in plant development, the impossibility to evolve towards a resistance against several virus species, which is characteristic to the MA model (contrary to the GFG model where “universal” resistance can occur), seems to drastically limit the evolution of this resistance gene against virus infections.
Altogether, the ERVIR project contributed to test the key hypotheses related to the evolution of resistance under different plant-pathogen interaction models by fine-scaled analysis of a single host plant species. In addition to this fundamental interest, this project contributed (i) to evaluate the influence of the human crop management on the host-pathogen interactions, (ii) to valorise the biodiversity by characterization of new resistant alleles toward damageable viruses of pepper, and (iii) to identify the cohort of viruses in wild environment which can initiate disease emergence in pepper cultivation.