Résumé
Cassava is one of the most important staple crops in Africa and its production is seriously damaged by viral diseases. In this study, we identify for the first time and characterize the genome organization of novel ampeloviruses infecting cassava plants in diverse geographical locations using three high-throughput sequencing protocols [Virion-Associated Nucleotide Acid (VANA), dsRNA and total RNA], and we provide a first analysis of the diversity of these agents and of the evolutionary forces acting on them. Thirteen new Closteroviridae isolates were characterized in field-grown cassava plants from the Democratic Republic of Congo (DR Congo), Madagascar, Mayotte, and Reunion islands. The analysis of the sequences of the corresponding contigs (ranging between 10,417 and 13,752 nucleotides in length) revealed seven open reading frames. The replication-associated polyproteins have three expected functional domains: methyltransferase , helicase, and RNA-dependent RNA polymerase (RdRp). Additional open reading frames code for a small transmembrane protein, a heat-shock protein 70 homolog (HSP70h), a heat shock protein 90 homolog (HSP90h), and a major and a minor coat protein (CP and CPd respectively). Defective genomic variants were also identified in some cassava accessions originating from Madagascar and Reunion. The isolates were found to belong to two species tentatively named Manihot esculenta-associated virus 1 and 2 (MEaV-1 and MEaV-2). Phylogenetic analyses showed that MEaV-1 and MEaV-2 belong to the genus Ampelovirus, in particular to its subgroup II. MEaV-1 was found in all of the countries of study, while MEaV-2 was only detected in Madagascar and Mayotte. Recombination analysis provided evidence of intraspecies recombination occurring between the isolates from Madagascar and Mayotte. No clear association with visual symptoms in the cassava host could be identified.