Positive strand RNA viruses, inculding tombusviruses, are known to utilize cellular membranes to assemble their replicase complexes (VRCs). Two tombusviruses , Tomato bushy stunt virus (TBSV) and Carnation Italian ringspot virus (CIRV), replicate on different organellar membranes, peroxisomes or endoplasmic reticulum (ER) for TBSV and mitochodria outer membranes in case of CIRV. I showed that both TBSV and CIRV replicase proteins could assemble VRCs and replicate viral RNA on purified microsomes (ER) and mitochondria. Different efficiencies of assembly was shown determined by multiple domains on TBSV or CIRV replication proteins. To study why VRC assembly could occur on an alternative organellar membranes, I focused on the phospholipids, key lipid components in ER or mitochondria membranes. Phospholipids directly interact with viral replicases, however, their specific roles during (+)RNA virus replication are far less understood. I used TBSV as a model (+) RNA virus, and established a cell-free TBSV replication system using artificial membranes prepared from different phospholipids. I showed that phosphatidylethanolamine (PE) is required for full cycle replication of the viral RNA.Moreover, PE is enriched at the sites of TBSV replication in plant and yeast cells, and was up-regulated during TBSV replication. Furthermore, up-regulation of total cellular PE content in yeast due to deletion of CHO2 leads to dramatically stimulated TBSV replication. Overall, I identified PE as the key lipid component of membranes required for TBSV replication, and my data highlighted that PE, an abundant phospholipid in all eukaryotic cells, not only serves as a structural component of membrane bilayers, its interaction with the viral replication proteins also stimulates (+)RNA virus replication. Further experiments indicated both early secretory pathway and endocytic pathway are involved in PE re-distribution to site of replication. In addition to lipids and subcellular membranes, certain host proteins are also involved in (+) RNA virus replication and VRC assembly. I identified Hop-like stress- inducible protein 1 (Sti1p), which interacts with heat shock protein 70, is required for the inhibition of CIRV replication. My findings indicate that Hop/Sti1 co-chaperone could act as a virus restriction factor in case of mitochondrial CIRV, but not against peroxisomal tombusvirus.

Cornichon (CNI) proteins are a conserved family of proteins among eukaryotes, from Erv14 in the yeast Saccharomyces cerevisiae to CNI homologs (CNIHs) in mammals and plants. Erv14 functions as a cargo receptor of coat protein complex II (COPII) for protein trafficking from the endoplasmic reticulum (ER) to the Golgi apparatus, en route to their final destinations. By interacting with specific cargo proteins, CNI proteins regulate key steps of embryo polarity in Drosophila, budding in yeast, and synaptic transmission in the mammalian brain. However, we have very limited understanding of plant CNIHs. Positive-strand RNA viruses assemble their viral replication complexes (VRCs) at specific host organelle membranes. With a better understanding of host factors involved in targeting viral replication proteins to the preferred organelles, we expect to block trafficking of viral replication proteins and thus, viral infection, by manipulating the required host proteins. Brome mosaic virus (BMV) is a model of positive-strand RNA viruses and its replication can be recapitulated in yeast. Importantly, BMV replication protein 1a is the only required viral protein to form VRCs at the perinuclear ER membrane in yeast. I demonstrate that Erv14 and COPII coat proteins are required for targeting BMV 1a to the perinuclear ER in yeast, suggesting a novel function of COPII vesicles in protein trafficking to the perinuclear ER membrane and in the BMV VRC formation. As for cellular functions, I show that plant CNIHs complement the defective distribution of BMV 1a in yeast mutant lacking Erv14. Taking advantage of Arabidopsis thaliana knockout mutants and knockdown of gene expression in Nicotiana benthamina, I also discover that CNIHs unexpectedly play crucial roles in pollen development, infection of a bacterial pathogen, and maintenance of ER tubules. I further confirm that CNI proteins are also required for maintaining ER tubules in yeast, suggesting a novel and conserved role in shaping ER morphology. Therefore, these findings indicate the functional diversity and redundancy of CNI proteins in key cellular processes and suggest a novel strategy to control plant pathogenic viruses and bacteria by manipulating plant CNIHs. / Ph. D.

Positive-stranded (+)RNA viruses replicate inside cells and depend on many cellular factors to complete their infection cycle. In the meanwhile, (+)RNA viruses face the host innate immunity, such as cell-intrinsic restriction factors that could block virus replication. Firstly, I have established that the plant DDX17-like RH30 DEAD-box helicase conducts strong inhibitory function on tombusvirus replication when expressed in plants and yeast surrogate host. This study demonstrates that RH30 blocks the assembly of viral replicase complex, the activation of RNA-dependent RNA polymerase function of p92pol and viral RNA template recruitment. In addition, the features rendering the abundant plant DEAD-box helicases either antiviral or pro-viral functions in tombusvirus replication are intriguing. I found the reversion of the antiviral function of DDX17-like RH30 DEAD-box helicase and the coopted pro-viral DDX3-like RH20 helicase due to deletion of unique N-terminal domains. The discovery of the sequence plasticity of DEAD-box helicases that can alter recognition of different cis-acting elements in the viral genome illustrates the evolutionary potential of RNA helicases in the arms race between viruses and their hosts. Moreover, I discovered that Xpo1 possesses an anti-viral function and exports previously characterized cell-intrinsic restriction factors (CIRFs) from the nucleus to the replication compartment of tombusviruses. Altogether, in my PhD studies, I found plant RH30 DEAD-box helicase is a potent host restriction factor inhibiting multiple steps of the tombusvirus replication. In addition, I provided the evidence supporting that the Nterminal domain determines the functions of antiviral DDX17-like RH30 DEAD-box helicase and pro-viral DDX3-like RH20 DEAD-box helicase in tombusvirus replication. Moreover, I discovered the emerging significance of the Xpo1-dependent nuclear export pathway in tombusvirus replication.

Le virus de la mosaïque jaune du navet (TYMV) possède un génome monopartite constitué d’ARN de polarité positive codant pour trois protéines, dont seule la polyprotéine 206K est indispensable à la réplication virale.Elle subit une maturation protéolytique, générant les protéines 140K et 66K, localisées au niveau de l’enveloppe des chloroplastes, siège de la réplication virale.Adressée aux chloroplastes, la protéine 140K y recrute la 66K et se comporte comme une protéine intégrale membranaire.Le domaine d’adressage aux chloroplastes (DAC) de la protéine 140K a été défini grâce à la transfection et à des protoplastes d’Arabidopsis thaliana par différentes constructions codantpour des versions délétées de la protéine fusionnées à l’EGFP, et à leur observation en microscopie confocale. Le DAC comprend deux hélices alpha amphipathiques dont la présence a été attestée par dichroïsme circulaire. Leur nécessité pour la localisation aux chloroplastes, l’association aux membranes et la réplication virale, a été étudiée. Différents patterns de distribution subcellulaire de la protéine 140K ont été observés. Ils sont corrélés au taux d’expression de la protéine. Sa dimérisation a également été démontrée.L’implication d’autres résidus du DAC dans la localisation subcellulaire, la dimérisation et la réplication virale, a également été recherchée. / Turnip yellow mosaic virus (TYMV) is a positive single-stranded RNA virus. Among the three ORFs encoded by the TYMV genome, 206K is the only protein required for viral replication. It is cleaved into 140K and 66K, which are both present at the chloroplast envelope membrane, where viral replication takes place.The 140K protein is targeted to chloroplasts, where it recruits 66K, and behaves as an integral membrane protein. The chloroplast targeting domain (DAC) of the 140K protein was defined using Arabidopsis thaliana protoplasts transfected by various constructs encoding deleted versions of 140Kfused to EGFP and subsequent confocal microscopy. The DAC comprises two amphipathic alpha helices, as confirmed by circular dichroism. Their involvement in chloroplast localisation and membrane association has been assessed, as well as their contribution to viral replication.We observed different subcellular distribution patterns of 140K protein, which correlate with the expression level of the protein. Its capability to dimerize has also been demonstrated.The involvement of other DAC residues in subcellular localisation, dimerization and viral replication has been studied.