Poleroviruses infect a wide range of cultivated plants such as potatoes, sugar beet and plants of Cucurbitaceae family. These viruses are restricted to phloem tissue where they replicate in nucleated cells and translocate over long distances through sieve elements. Polerovirus capsid is composed of the major coat protein (CP) and of a minor component referred to as the readthrough (RT) protein and exposed at the outside of the particles. CP and RT proteins are essential for virus movement and transmission by aphids. The aim of this study is to identify phloem proteins interacting with viral proteins and potentially involved in viral cycle, by screening an A. thaliana companion cell (CC) cDNA library with structural proteins or protein domains of CABYV. Four genes encoding for a heat shock protein (HSP), a profilin (PRF3), a glycosyl hydrolase and the protein ”Response to low sulfur ” (LSU3) were identified and interact with the C-terminal part of the RT protein (RTC‑ter) and with the RT protein for the HSP. An additional gene encoding for the protein ALY, identified in the laboratory, by screening an aphid cDNA library with structural proteins of the Turnip yellows virus (another polerovirus) was studied. This protein has four orthologues in Arabidopsis, involved in the gene silencing mechanism against Tomato Bushy Stunt Virus. Here we show that CABYV and TuYV structural proteins interact with the four orthologues of Arabidopsis. Involvement of these candidate genes was not confirmed in Arabidopsis knock-out mutants. In functional experiments, ambiguous results were obtained with PRF3 arabidopsis mutants, and this lead me to study the interaction between PRF3 protein CABYV RT c-ter domain by FLIM, but no interaction was found so far. As all candidat interact with the RTC-ter domain, we studied more precisely the role of this domain in the viral cycle and the role of the complete RT protein. We studied the in vivo RT protein processing and its consequences on systemic movement of CABYV mutants. Using a collection of point mutations introduced in the central domain of the CABYV RT protein, we approached the site of the RT processing and proposed that this process is affected by the secondary structure around the cleavage site. We also reported for the first time the generation of a polerovirus mutant able to synthesize only the RT protein and to incorporate it into the particle. This mutant was unable to move systemically. Conversely another mutant producing a full-length RT protein impaired in correct processing and incorporating a shorter version of the RT protein showed very weak systemic infection. These data are strongly in favor of a role of both RT proteins in efficient CABYV movement. An inefficient virus transport was still maintained in the absence of RT proteins suggesting an RT-independent movement pathway. Based on these results, we propose a model for CABYV long-distance transport in which the complete RT protein, or its C-terminal part, acts in trans on wild-type virions to promote their efficient long-distance transport.
