The disertation thesis was focused on the analysis of variability of two nepoviruses genomes, Grapevine fanleaf virus -- GFLV and Arabis mosaic virus -- ArMV. The thesis described single isolates of these nepoviruses and their until recently non-explored genome portions coded by RNA1 strand, the coding regions for 1BHel and 1EPol proteins precisely. The main topic of this thesis was the identification of infected plants that were infected by nepoviruses GFLV and ArMV. Based on design of specific primers there was established sequence homology of coding sequences. Another topic was the development and optimalization of diagnostic system for the efficient GFLV and ArMV detection. This work was based on hypothesis of very high level GFLV and ArMV genomes variability. The results were recieved mainly from sequencing analyses performed by capillary automatic sequencer ABI-PRISM 310 (Applied Biosystems, Carlsbad, USA). Obtained results were assessed with CLC Main Workbench 5 software (CLC Bio, Aarhus, Denmark). Real-time PCR TaqMan multiplex was also designed using this software. This real-time tool is able to detect and quantify both nepoviruses within one PCR reaction simultaneously. There were sequenced and analysed the genetic codes at the nucleotide level for MP 2B , 2CCP, 1BHel and 1EPol proteins in this work. The first one was assessed because of the development of real-time PCR TaqMan multiplex. In this case were sequenced 290 bp from 14 GFLV and ArMV isolates, the variability level was established in frame of nucleotides 71.7-97.6%. The partial genetic code for 2CCP was assessed in frame of 6 GFLV isolates and the variability was established in 935 bp fragment to 83-86% at the nucleotide level and 81-91% at the amino acid level. The partial genetic code for 1BHel was sequenced in frame of 6 isolates GFLV and the variability was established in 379 bp in frame of all accessible GFLV and ArMV isolates, the variability was established 86.4-98.9% at the nucleotide level and 96-100% at the aminoacid level. The partial genetic code for 1EPol protein was sequenced in 6 GFLV isolates in frame of 365 bp and the variability was established for all accessible sequences in GenBank/NCBI. The variability was established at the nucleotide level 79.5-100% and at the aminoacid level 91.4-100%. All received sequences longer than 300 bp were submitted to GenBank/NCBI. The obtained partial genetic codes for 1BHel a 1EPol proteins coded in RNA1 were unique at the time of publication. This work provided the important results which were used subsequently like source information for construction of vectors intended for the preparation of transgenic plants. Then the incurred grapevine would show resistance against GFLV. Moreover, based on results of this work there was created certified methodology of the real-time PCR TaqMan multiplex method which can be standartly used as an efficient diagnostic tool for simultaneous detection of GFLV and ArMV. This real-time system was succesfully compared with ELISA what would be interesting mainly for diagnostic laboratories engaged in the certification processes of propagation material.

La maladie du court-noué est principalement causée en Europe par le Grapevine fanleaf virus (GFLV) et l’Arabis mosaic virus (ArMV). Le principal moyen de lutte contre sa dispersion consiste à certifier l’absence de ces virus dans les vignes commercialisées par des méthodes sérologiques tels que le DAS-ELISA. De par leurs propriétés biophysique et structurale exceptionnelles, les Nanobodies (Nb) issus des domaines variables d’immunoglobulines (Ig) composées uniquement de chaînes lourdes, se distinguent avantageusement des Ig conventionnelles. L’objectif majeur de ma thèse était d’établir un test de diagnostic du court-noué à base de Nb. À partir de deux collections de Nb contre le GFLV et l’ArMV, j’ai identifié des Nb reconnaissant un large spectre d’isolats viraux. Leur fusion à une protéine fluorescente ou à la phosphatase alcaline a conduit à l’obtention de réactifs de détection performants du GFLV et de l’ArMV par DAS-ELISA. La structure atomique d’un complexe Nb/GFLV résolue à 2.8 Å par cryomicroscopie électronique a permis de cartographier l’épitope en surface du virus et a révélé une couverture maximale de la particule virale par le Nb. La comparaison des tests Nb à des réactifs sérologiques commerciaux a révélé leur supériorité en terme de sensibilité et de spécificité, ouvrant ainsi la voie à la commercialisation d’un nouveau test de diagnostic des virus du court-noué de la vigne. / The grapevine fanleaf disease is mainly caused in Europe by the Grapevine fanleaf virus (GFLV) and the Arabis mosaic virus (ArMV). The principal mean to limit their spread, is to certify their absence in marketed grapevines by serological methods such as DAS-ELISA. Their unique biophysical and structural properties make the variable domains of heavy chain-only immunoglobulin, called Nanobodies (Nb) a real asset for the development of a diagnostic test against fanleaf disease viruses. I identified Nb able to detect a broad spectrum of viral isolates from two Nb collections against GFLV and ArMV. Their fusion to a fluorescent protein or to a bacterial alkaline phosphatase resulted in the production of efficient DAS-ELISA detection reagents. The atomic structure of a Nb/GFLV complex was solved at 2.8 Å by cryoelectron microscopy, allowing the precise mapping of the viral epitope. This result showed a maximum coverage of the viral particle by the Nb, leading to a maximal signal in DAS-ELISA. The full Nb tests against GFLV and ArMV were compared to commercial reagents and showed the superiority of the former in both sensitivity and specificity, opening the way for the development and commercialization of a new type of serological kits for the detection of grapevine viruses.