RESUMEN
Downy mildew is caused by Plasmopara viticola, an obligate oomycete plant pathogen, a devasting disease of grapevine. To protect plants from the disease, complex III inhibitors are among the fungicides widely used. They specifically target the mitochondrial cytochrome b (cytb) of the pathogen to block cellular respiration mechanisms. In the French vineyard, P. viticola has developed resistance against a first group of these fungicides, the Quinone outside Inhibitors (QoI), with a single amino acid substitution G143A in its cytb mitochondrial sequence. The use of QoI was limited and another type of fungicide, the Quinone inside Inhibitors, targeting the same gene and highly effective against oomycetes, was used instead. Recently however, less sensitive P. viticola populations were detected after treatments with some inhibitors, in particular ametoctradin and cyazofamid. By isolating single-sporangia P. viticola strains resistant to these fungicides, we characterized new variants in the cytb sequences associated with cyazofamid resistance: a point mutation (L201S) and more strikingly, two insertions (E203-DE-V204, E203-VE-V204). In parallel with the classical tools, pyrosequencing and qPCR, we then benchmarked short and long-reads NGS technologies (Ion Torrent, Illumina, Oxford Nanopore Technologies) to sequence the complete cytb with a view to detecting and assessing the proportion of resistant variants of P. viticola at the scale of a field population. Eighteen populations collected from French vineyard fields in 2020 were analysed: 12 showed a variable proportion of G143A, 11 of E203-DE-V204 and 7 populations of the S34L variant that confers resistance to ametoctradin. Interestingly, the long reads were able to identify variants, including SNPs, with confidence and to detect a small proportion of P. viticola with multiple variants along the same cytb sequence. Overall, NGS appears to be a promising method for assessing fungicide resistance of pathogens linked to cytb modifications at the field population level. This approach could rapidly become a robust decision support tool for resistance management in the future.
